Water Research 36 (2002) 11431154

Advanced oxidation of a reactive dyebath euent: comparison

of O3, H2O2/UV-C and TiO2/UV-A processes
Idil Arslan Alatona, Isil Akmehmet Balcioglub,*, Detlef W. Bahnemannc
a

Department of Chemistry and Biochemistry, University of South Carolina, GSRC, 631 Sumter Street, Columbia, SC 29208, USA
b
Bogazic-i University, Institute of Environmental Sciences, 80815 Bebek, Istanbul, Turkey
c
Institut fur
. Solarenergieforschung Hameln/Emmerthal ISFH, Aussenstelle Hannover, D-30165, Hannover, Germany
Received 5 September 2000; accepted 27 June 2001

Abstract
In the present study the treatment eciency of dierent AOPs (O3/OH
, H2O2/UV-C and TiO2/UV-A) were
compared for the oxidation of simulated reactive dyebath euent containing a mixture of monochlorotriazine type
reactive dyes and various dye auxiliary chemicals at typical concentrations encountered in exhausted reactive dyebath
liquors. A525 (color), UV280 (aromaticity) and TOC removal rates were assessed to screen the most appropriate
oxidative process in terms of reactive dyebath euent treatment. Special emphasis was laid on the eect of reaction pH
and applied oxidant (O3, H2O2) dose on the observed reaction kinetics. It was established that the investigated AOPs
were negatively aected by the Na2CO3 content (=867 mg/L) which is always present at high concentrations in
dyehouse euents since it is applied as a pH buer and dye xation agent during the reactive dyeing process. The
ozonation reaction exhibited almost instantaneous decolorization kinetics and a reasonable TOC reduction rate. It
appeared to be stable under the investigated advanced oxidation conditions and outranked the other studied AOPs
based on the above mentioned criteria. Besides, the electrical energy requirements based on the EE/O parameter (the
electrical energy required per order of pollutant removal in 1 m3 wastewater) was calculated for the homogenous AOPs
in terms of decolorization kinetics. In view of the electrical energy eciency, ozonation and H2O2/UV-C oxidation at
the selected treatment conditions appear to be promising candidates for full-scale dyehouse euent decolorization.
r 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Advanced oxidation; Ozonation; H2O2/UV-C; TiO2/UV-A; Reactive dyebath euent; EE/O

1. Introduction
Like many other industrial euents, textile industry
wastewater varies signicantly in quantity, but additionally in composition [1]. Textile wastewater is
strongly colored which creates an environmental as well
as aesthetic problem. As regulations are becoming ever
more stringent, the need for technically and economically more ecient means of decolorization and mineralization is obvious. Currently, there are no
*Corresponding author. Tel.: +90-212-263-1540; fax: +90212-257-5033.
E-mail address: balciogl@boun.edu.tr (I.A. Balcioglu).

economically attractive technologies to achieve color

removal. Existing physicochemical technologies such as
membrane ltration or activated carbon adsorption are
expensive and commercially unattractive. Furthermore,
these processes just transfer pollutants from one phase
to another rather than eliminating them from the water
matrix. Recovery and reuse of certain chemical compounds present in dyebath euents is currently under
investigation [2,3] but is not possible for the application
of certain dyestu classes, such as reactive dyes. In
addition, the problem of color in dyehouse euent has
become identied particularly with the dyeing of cotton
bers that contributes to almost 40% of total ber
consumption annually worldwide, and the use of

0043-1354/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 3 - 1 3 5 4 ( 0 1 ) 0 0 3 3 5 - 9

1144

I.A. Alaton et al. / Water Research 36 (2002) 11431154

Nomenclature
A0
A525
CH2 O2
DSSR
EE/O

I0
kd

(Initial) absorbance at t 0 min,

1/m
Absorbance at l 525 nm representing
color), 1/m
Concentration of H2O2, mM
Double-skin-sheet reactor
Electrical energy per order of
contaminant (color) removal
in 1 m3 wastewater,
kWh/(m3  order)
Incident UV light ux, Einstein/(m2  s)
or Einstein/(L  s)
rst order decolorization rate constant,
1/min

reactive dyestus since as much as 50% of these dyes

end up in the exhausted dyebath in their hydrolyzed and
unxed form [4,5].
The chemical limitations of conventional chemical
oxidation techniques can be overcome by the development of so-called advanced oxidation processes (AOPs)
which use strong oxidizing agents (O3, H2O2) and/or
catalysts (Fe, Mn, TiO2) in the presence or absence of an
irradiation source [6]. AOPs mainly involve the generation of a very powerful and non-selective oxidizing
agent, the hydroxyl radical (OHd), for the destruction of
refractory and hazardous pollutants found in groundwater, surface water and industrial wastewater [7].
The treatment of reactive dyes in aqueous solutions
via dierent AOPs has been extensively studied [815].
However, the advanced oxidation of reactive dyebath
euents is rather limited to a few investigations [1618].
From these studies it could be inferred that the dyebath
constituents (sodium salts, detergents, wetting agents,
sequestering agents etc.) seriously reduced the treatment
eciency of the selected AOP. Thus, the inhibiting eect
of common reactive dye auxiliary chemicals, in particularly the OHd scavenging eect of carbonate ions
applied in the form of anhydrous Na2CO3 and used at
substantially high concentrations in the reactive dyeing
process, still needs to be further questioned for dierent
AOPs. Moreover, a detailed assessment of the advanced
oxidation conditions (i.e. reaction pH and oxidant dose)
that are AOP- and pollutant-specic has to be undertaken to determine the most appropriate AOP for the
treatment of the textile euent in question.
With the aim to contribute to the clarication of the
above mentioned points, the experimental work described herein evaluates comparatively the treatment of
simulated exhausted reactive dyebath liquors prepared
from a mixture of dierent dyeing formulations by two

kUV280
OHd
rid
TFFBR
TOC
t
UV254
UV280

First order dearomatization rate

constant, 1/min
Hydroxyl radical(s)
Initial rate of decolorization
(A525 removal), 1/(m  min)
Thin-lm-xed-bed reactor
Total organic carbon (mg/L)
Treatment time, min
Absorbance at l 254 nm, 1/m
Absorbance at l 280 nm representing
aromaticity), 1/m

Greek letters
Molar extinction coecient of H2O2 at
eH2 O2
l 254 nm, 1/(M  cm) wavelength, nm
l
Wavelength, nm

homogenous (O3, H2O2/UV-C) AOPs and one heterogeneous (TiO2/UV-A) AOP in batch experiments.
Optimization of the reaction pH and oxidant (O3 and
H2O2) dose were made in accordance with the highest
rate in color (absorbance at 525 nm wavelength),
aromaticity (absorbance at 280 nm wavelength) and
TOC (total organic carbon) removal. Thereafter, the
oxidative performance of the selected chemical oxidation systems was compared in terms of their electrical
energy requirements.

2. Materials and methods

2.1. Simulated reactive dyebath euent
The investigated reactive dyes and dye assisting
chemicals were kindly supplied by an integrated textile
manufacturing plant. Synthetic spent dyebath wastewater was prepared weekly according to the cotton ber
dyeing procedure of the factory. The recipe mixture was
used considering the international market share of their
reactive dyestu content and colors mostly applied to
the cotton fabrics in the dyeing stage. The synthetic
reactive dyebath euent was prepared according to the
information that usually 20% of the dyestus (in their
hydrolyzed or unxed form) and 100% of all applied
dye assisting chemicals remain in the exhausted reactive
dyebath liquor [4]. To guarantee that all dyestus in the
prepared mixture were 100% hydrolyzed, they were rst
dissolved in de-ionized boiling water together with the
dye auxiliaries, and then the dyebath was kept in a cool
room for 12 h prior to its use. The concentration of the
chemical components present in the simulated dyebath
euent and the functions of these chemicals in the
dyeing process are summarized in Table 1.

1145

I.A. Alaton et al. / Water Research 36 (2002) 11431154

Table 1
Chemical composition of the simulated reactive dyebath euent
Dyestu (color index)

Company

Reactive group

Concentration (mg/L)

Procion blue HERD

(reactive blue 160)
Procion crimson HEXL
(ref. no. PR 3930/00)a
Procion Yellow HE4R
(reactive yellow 84)
Procion navy HEXL
(ref. no. PR 3830/01)a
Procion yellow HEXL
(reactive yellow 138 : 1)

BASF

Bisaminochlorotriazine

BASF

Aminochlorotriazine

40.60

BASF

Bisaminochlorotriazine

15.00

BASF

Aminochlorotriazine

86.30

BASF

Aminochlorotriazine

33.30

6.83

Auxiliary chemical

Function

Concentration (g/L)

Acetic acid
NaCl
Na2CO3
NaOH

Neutralization
Electrolyte (aggregation of dyes onto fabric)
pH buer
Produces covalent bonds between
dyestu and fabric
Anti-creasing agent

0.79
41.00
13.00
0.51

Polyether based co-polymer

micro-dispersion
Acryl co-polymer-phosphor
mixture
Alcyl phenol polyglycol ether
a

Sequestering agent

0.85

Detergent (washing out of unxed dyestu)

0.50

Color index numbers have not been allocated.

Table 2
Environmental characterization of the 15 fold diluted reactive
dyebath euent
Parameter

Value

TOC (mg/L)
BOD5 (mg/L)
AOX (mg/L)
UV254nm (1/m)
UV280nm (1/m)
A525nm (1/m)
CO2

3 (mg/L)
Cl
(mg/L)
pH

46.8
nda
0.102
38.08
41.72
12.95
490.6
1659.0
10.9

1.20

Below the detection limit of the standard BOD5 test.

The recipe mixture was treated for 1 h in a 15-fold

dilution corresponding to the typical dyehouse euent
concentration from the washing and rinsing stages of the
textile factory. Selection of this dilution rate was also
based upon the fact that for light induced treatment
systems the very poor visible-and UV-light transmission
of the actual (undiluted) dyebath is the major drawback
for an eective treatment via AOPs. The environmental
characterization of the 15 fold diluted exhausted reactive
dyebath liquor is given in Table 2.

2.2. The ozone reactor

All ozonation experiments were performed in a
1500 mL capacity semi-batch bubble column presented
in detail elsewhere [17]. Continuous recirculation of the
reaction solution was provided by a peristaltic pump
(Cole Parmer) at a ow rate of 50 mL/min. Ozone was
produced by a corona discharge Fischer OZ 500 model
ozone generator from dry oxygen used as the feed gas.
The ozone input rate varying between 0.624.45 g/h
(applied ozone dose=5442970 mg/L) was delivered at a
gas ow rate of 100 L/h through a sintered glass sparger
and excess ozone leaving the reactor was destroyed by
two sequential 20% KI traps incorporated to the reactor
set-up. All ozonation experiments were conducted at
room temperature and no reactor cooling was provided.
2.3. The photochemical reactor
H2O2/UV-C oxidation of the simulated reactive dyehouse euent was conducted in a 2500 mL capacity
annular photoreactor at batch mode. The UV-C reactor
was a 304 SS (stainless steel) tube (L 500 mm;
d 85 mm) equipped with a 25 W low-pressure mercury
vapor UV-C lamp (L 450 mm; d 25 mm) that was
placed into a quartz envelope (L 460 mm;
d 31 mm). Both UV-C lamp and quartz lamp housing

1146

I.A. Alaton et al. / Water Research 36 (2002) 11431154

were centered into the reactor tube. The UV-C lamp

emitted 98% of its total irradiation at 254 nm. The
incident light ux and eective path length of the UV-C
lamp were determined by hydrogen peroxide actinometry [19] as I0 2:0  10
5 Einstein/(L  s) and 1.72 cm,
respectively. Mixing was provided by circulating the
reaction solution with a peristaltic pump (Cole Parmer)
at a rate of 50 mL/min. At the start of the H2O2/UV-C
experiments, 20% (for CH2 O2 =5 mM)-67% (for CH2 O2
=40 mM) of the UV-C light at l 254 nm was
absorbed by H2O2 (average UV254 for the initial dyebath sample=38.10 1/m; eH2 O2 at 254 nm=1900 L/
(mol  m)). Since throughout the H2O2/UV-C reactions
the H2O2 concentration remained nearly constant and
the UV254 values of the reactive dyebath decreased to
67% of their original value at the end of the oxidation
period, it can be assumed that the UV-C light entering
the reactor was eectively and dominantly absorbed by
H2O2 (7697% UV254 absorption for 5 and 40 mM
H2O2, respectively). As in the case of the ozonation
experiments all H2O2/UV-C reactions were performed at
ambient temperature and no lamp cooling was provided.
2.4. The photocatalytic reactor
The experimental set-up of the reactor employed for
the photocatalytic treatment of the simulated dyebath
euent is presented in Fig. 1a, whereas the double-skinsheet reactor (DSSR) is schematically elucidated in
Fig. 1b.
The Plexiglass (Polymethylmethacrylate, PMMA,
.
Rohm
GmbH, Darmstadt, Germany) DSSR is able to
use diuse as well as direct UV light and the high
turbulence provided in the reactor results in a good
mixing of the heterogeneous photocatalyst with the
dissolved contaminants [20]. Moreover, the DSSR has a
relatively large reaction volume compared with a thinlm-xed-bed reactor (TFFBR), which is one of its
inherent advantages. The apparatus consists of a
double-skin sheet (10 channels; 410  31  12 mm; total
volume=1.38 L; total surface area=0.13 m2) made of
PMMA as the UV-A photoreactor and a reservoir (3 L),
connected by pressure-resistant PVC-tubes (19 mm i.d.).
In all experiments the system was run in a recirculation
batch mode. The ow of the simulated dyehouse euent
was provided by an ordinary centrifugal pump at a rate
of 426 L/h, i.e., the residence time in the photocatalytic
reactor was 12 s. The reactor was illuminated from the
top with 24  15 W Phillips Cleo-R type UV-A uorescent lamp tubes with a spectrum ranging between
315 nmolo395 nm (lmax 362:5 nm) and thus closely
representing the UV-A portion of solar light. The UV-A
lamps were placed parallel to the width of the plates
at a distance of 15 cm. The average incident light ux
directly at the reactor surface was measured with a
.
UV-A sensitive luxmeter (Dr. Honle)
as 40 W/m2

Fig. 1. The experimental set-up of the TiO2/UV-A process (a)

and the schematic presentation of the DSSR (b). L=light
source; Ai=irradiated Plexiglass plate; t=supporting tray;
Vr=reservoir; P=pump.

(1.2  10
4 Einstein/m2  s) and was practically completely absorbed by the TiO2 in the TiO2+synthetic
dyebath suspension. The lamps were turned on 45 min
before the start of the photocatalytic reaction to ensure a
constant light output. The UV-A lamps were cooled
with 3 fans incorporated into the lamp housing.
2.5. Experimental procedure
Prior to the advanced oxidation process, the pH of the
dyehouse euent was adjusted to the desired value
employing dilute HCl or NaOH solutions. During the
H2O2/UV-C process, the dyehouse euent was sparged
with H2O2 (30%; Merck) at the beginning of the
treatment period and then the light was immediately
turned on, that accounted for t 0; i.e. the start of the
reaction. Residual (unreacted) H2O2 was immediately
destroyed with enzyme catalase (Sigma) originating
from bovine liver (13,000 a.u./mg protein) to prevent
its positive interference with spectrophotometric analyses at 280 nm. For the photocatalytic experiments,
Degussa P25 type TiO2 powder (BET=50 m2/g; average
particle size=30 nm; consisting of 80% anatase and
20% rutile crystal phase) was added resulting in a nal
concentration of 1 g/L in the dyehouse euent samples.
Thereafter, the reaction solution was sonicated for
15 min to obtain a well-mixed suspension and magneti-

I.A. Alaton et al. / Water Research 36 (2002) 11431154

cally stirred for another 20 min in the dark to allow

equilibration of the dyehouse euentTiO2 mixture.
Dark control experiments conducted to determine the
exact initial dark adsorption-equilibration period indicated that the reactive dye wastewater already reached
its equilibrium after 15 min sonication with an average
removal of 6% TOC. Prior to analyses, TiO2 containing
samples were immediately centrifuged in a Heraeus
Sepatech Labofuge at 4000 rpm for 40 min to obtain a
clear supernatant for analyses.
2.6. Analytical measurements
10 mL sample aliquots were taken from the reactors at
regular time intervals and analyzed for UV-VIS
absorbance and TOC. The color of the initial and
treated samples was analyzed with a Shimadzu UV-120
model single beam spectrophotometer. The visible light
absorbance at the characteristic wavelength of the
samples, e.g. l 525 nm, was recorded to follow the
progress of decolorization during advanced chemical
oxidation. The changes in UV280 absorption of the
reaction solution was also followed during 1 h treatment
period to verify the formation and concomitant disappearance of aromatic advanced oxidation products.
TOC was determined by a Carlo Erba TCM 480 model
organic carbon analyzer (Fisons Instruments).
2.7. Kinetic evaluation
During the application of AOPs, decolorization of the
reacting sample exhibited an exponential decay. Assuming rst order reaction kinetics for the synthetic
dyehouse euent matrix, the decolorization rate constant was determined from the slope of the semilogarithmic plot of color (peak absorbance A at
l 525 nm) versus advanced oxidative treatment time,
respectively, in accordance with the kinetic equation;
lnA=A0 kd  t;

where kd stands for the rst order decolorization rate

constant (in 1/min). The abatement of UV280 (UV
absorbance at l 280 nm) could also be evaluated in
terms of the rst order kinetic equation;
lnUV280 =UV280;0 kUV280  t;

where kUV280 describes the rst order UV280 abatement

rate constant (in 1/min).

3. Results and discussion

3.1. General aspects
Changes in pH were rather insignicant during any
advanced oxidation of the simulated dyehouse euent,

1147

which is easily explained since the reactive dyebath

euent was highly buered. In all treatment cases,
disappearance of UV-VIS absorbance exhibited a rst
order kinetic behavior, whereas no clear kinetic trend
was observed for the TOC abatement rates.
3.2. O3 treatment
Ozone reacts with aromatic pollutants found in water
and wastewater via two dierent pathways namely direct
molecular and indirect radical chain type reaction [21].
The ozonation reaction pathway strongly depends on
the characteristics of the wastewater to be treated
(i.e. pH, concentration of initiators, promoters and
scavengers in the reacting medium). To demonstrate the
eect of pH on the treatment eciency, the pH of the
simulated reactive dyebath euent being subjected to
ozonation at a specic O3 dose of 2340 mg/L was
adjusted to 3, 7 and 11, respectively. Fig. 2ac displays
the time dependent changes in color (2a), UV280 (2b),
and TOC (2c) at varying pH values.
In all treatment cases, almost instantaneous decolorization was observed and thus kd values could not be
accurately determined for all studied reaction pH values.
From Fig. 2b and c it is also apparent that appreciably
faster UV280 and TOC removal rates were achieved at
pH 7 than at pH 11 and 3. Initial decolorization rates
(rid values, obtained from the initial slopes of the timedependent color abatement curves), rst order UV280
removal kinetics (kUV280 ) and overall per cent TOC
reduction eciencies obtained after 1 h ozonation are
summarized in Table 3.
According to Table 3 the initial decolorization rates
obtained for the ozonation of the simulated dyehouse
euent at the selected pH range were almost identical
(average rid 11:37 1/m  min). The overall UV280
removal increased from 70.7% at pH 3 to 93.1% at
pH 7 and remained almost unchanged (92.4%) at pH 11.
This observation indicated that color removal could be
simply achieved via both reaction pathways of ozone,
whereas the reaction pH had to be at least 7 to enhance
ozone decomposition for higher overall UV280 and TOC
(TOC removal was 28% at pH 7 instead of 17% at pH
3) reduction. Conducting the ozonation process at pH
11 did not further improve the treatment performance of
the ozonation process due to the fact that the additional
2

OHd scavenging eect of HCO

present in the
3 /CO3
reaction solution compensated the acceleration of ozone
decomposition to OHd at alkaline pH. As a consequence, the highest overall TOC removal rate was found
as 28% at pH 7, when the radical type reaction becomes
eective and simultaneously the inhibiting eect of
carbonate ions is not yet very pronounced.
The eect of the applied specic ozone dose was
evaluated in separate experiments where the specic
ozone input rate was varied in the range of 544

1148

I.A. Alaton et al. / Water Research 36 (2002) 11431154

Table 3
Eect of pH on the ozonation of simulated reactive dyebath
euent (applied ozone dose=2340 mg/L)
Reaction
pH

rid
(1/m  min)

kUV280
(1/min)

TOC removal
(%)

3.0
7.0
11.0

11.5
11.2
11.4

0.15
0.33
0.28

17.3
27.8
15.9

Fig. 3. Eect of O3 dose on the ozonation of simulated reactive

dyebath euent. Experimental conditions: Reaction pH=7.0;
Treatment time=1 h.

eciency continued to increase from 28% for 2340 mg/L

O3 further up to 40.5% at the highest investigated O3
dose.
3.3. H2O2/UV-C treatment
Fig. 2. Eect of pH on A525 (a), UV280 (b) and TOC (c)
abatement kinetics during ozonation of the simulated reactive
dyebath euent. Experimental conditions: O3 dose=2340 mg/
L; Treatment time=1 h.

2970 mg/(L  h) and the reaction pH was kept at pH 7

(Fig. 3).
As is evident from Fig. 3, the changes in rid and UV280
proceeded parallel to each other and the increase in
these process parameters was particularly signicant
when the ozone dose was changed from 540 to 1310 mg/
L. The initial decolorization rate remained invariant
above an ozone dose of 2340 mg/L and the highest
UV280 removal (=94.5%) that was obtained at the
highest specic O3 dose studied (2970 mg/L) did not
dier signicantly from that achieved at a concentration
of 1310 mg/L (=92.7%). From these ndings it is
apparent that the ozonation reaction becomes practically mass transfer limited in terms of color and UV280
abatement rate after a threshold O3 input concentration of 1310 mg/L. However, the overall TOC removal

Because of its inherent advantages (i.e. no sludge

formation, signicant reduction in COD at relatively
short reaction times and easy handling), the H2O2/UV-C
system remained the AOP most often applied for the
treatment of hazardous/refractory pollutants found in
water or wastewater [22]. To explore the eect of pH and
initial H2O2 concentration on the H2O2/UV-C treatment
eciency, advanced oxidation of the reactive dyebath
euent was carried out for 1 h at varying pH (3, 7 and
11) and H2O2 doses (5, 10, 20 and 40 mM) that was
initially added to the reaction solution. Further, a
control experiment was carried out in which the dyebath
euent was exposed to mere UV-C irradiation for the
same treatment period. In all H2O2/UV-C experiments,
residual H2O2 remained in the reaction solution
(>30 mg/L) indicating that it was provided in excess
and exerted no rate limiting eect on the oxidation
reaction.
As can be seen in Fig. 4a, complete color removal
could be achieved after only 10 min oxidation, whereas
UV280 removal rates were almost complete after 1 h

I.A. Alaton et al. / Water Research 36 (2002) 11431154

1149

Table 4 presents the kd ; kUV280 ; rid values as well as

overall per cent TOC removal eciencies obtained for
H2O2/UV-C oxidation of the reactive dyebath euent in
the photochemical (UV-C) reactor in the presence of
680 mg/L (20 mM) H2O2 at varying pH values.
As is apparent from this Table 4, kd and kUV280 values
increased by a factor of 2.7 and 2.6 respectively by
elevating the pH from 3 to 7. Overall UV280 removal
eciency was not seriously aected by the reaction pH
and varied between 87% and 91% after 1 h treatment. In
contrast to ozonation, further increasing the pH to 11
inhibited the color and UV280 abatement rates drastically. In a related study [23], dye degradation via H2O2/
UV-C/US (US=ultrasonication) was reported to be
favored at higher pH values. In the present study, a
completely dierent kinetic trend was particularly
observed in the case of the TOC reduction rate; overall
TOC removal eciency dropped from 30% at pH 3 to
14% at pH 7 and practically no TOC reduction was
observable at pH 11 (average TOC removal 1%). The
inhibition of the TOC reduction at pH 11 is proposed to
be the consequence of the OHd scavenging eect of
carbonate anions present in the simulated reactive
dyebath euent in the form of Na2CO3 (=867 mg/L).
As compared to ozonation, the H2O2/UV-C process was
more strongly suppressed by increasing the reaction pH.
This observation is attributed to the fact that if for the
pollutant under investigation, degradation via direct
UV-C photolysis is negligible, the eciency of the H2O2/
UV-C oxidation system will primarily depend on the
rate of OHd production via UV-C photolysis of H2O2.
Increasing the pH in the ozonation system, on the other
hand, will result in a higher OHd production rate due to
the following reactions:

Fig. 4. Eect of pH on A525 (a), UV280 (b) and TOC (c)

abatement kinetics during H2O2/UV-C oxidation of the
simulated reactive dyebath euent. Experimental conditions:
H2O2 dose=680 mg/L; Treatment time=1 h.
Table 4
Eect of pH on the H2O2/UV-C oxidation of simulated reactive
dyebath euent (applied H2O2 dose=680 mg/L)
Reaction
pH

kd
(1/min)

rid
(1/m  min)

kUV280
(1/min)

TOC removal
(%)

3.0
7.0
11.0

0.14
0.36
0.07

5.5
6.8
0.9

0.05
0.13
0.04

30.4
13.9
1.1

treatment and did not change signicantly after 30 min

oxidation (Fig. 4b) at pH 7. On the other hand,
according to Fig. 4c, TOC values decreased steadily at
pH 3, but only slightly at pH 7 and pH 11 throughout
the course of H2O2/UV-C experiments.

d
O3 OH
-Od

3 OH

d

Od

3 -O2 O

Od
H -OHd

As a consequence, the H2O2/UV-C process is expected

to be more sensitive to the scavenging eect of carbonate
at higher pH values than the ozonation, while for the
ozonation process the latter is partially compensated by
the increased OHd formation rate via Eqs. (3)(5).
Like the reaction pH, the applied H2O2 dose also
signicantly inuenced the H2O2/UV-C oxidation rate
of the reactive dyebath euent as is evident from Fig. 5
displaying the treatment performance of the H2O2/UVC process at pH 7 and varying initial H2O2 doses.
Generally speaking, the same inuence of H2O2 dose
was observed for all investigated process parameters
except for TOC. The removal rates increased considerably with increasing the initial H2O2 concentration from
0 to 680 mg/L (20 mM) reaching a maximum value for
color and UV280 abatement kinetics at a H2O2 dose of

1150

I.A. Alaton et al. / Water Research 36 (2002) 11431154

around 680 mg/L (kd 0:36 1/min; kUV280 0:13 1/min).

Elevating the H2O2 concentration further from 680 to
1360 mg/L (40 mM) resulted in a signicant inhibition of
color and UV280 reduction rate, whereas overall TOC
removal eciency continued to increase reaching its
highest value (14.6%) in the presence of 40 mM
(1360 mg/L) H2O2. The eect of an optimum oxidant
dose for rst order decolorization rate constants when
varying the initial H2O2 concentrations were observed in
previous [12,24]. The above mentioned observations are
typical for H2O2/UV-C oxidation systems, since it is
already well established that H2O2 itself can act as an
eective OHd scavenger at concentrations that are
specic for the pollutant in question [25]. Thus, as
stated in other experimental investigations, the presence
of excess H2O2 can lower the treatment eciency of
AOPs and it is very important to optimize the applied
H2O2 dose to maximize the treatment performance of
the AOP.
Certain dyes can be photo-decolorized directly in the
presence of UV-C irradiation. Chu and Ma (1998) found
that direct UV-C photolysis of several commercial azo
and anthraquinone dyes followed a rst order decay. In
the present study, two control experiments in the
absence of UV-C light or H2O2 were separately carried
out to test the eect of H2O2 and UV-C irradiation,
respectively, on the oxidation rates. The progress of
decolorization via direct UV-C photolysis appeared to
be very slow (kd 0:002 1/min), while 4% reduction in
UV280 (kUV280 =0.0007 1/min) and no TOC removal
could be observed (Fig. 5, 0 mg/L H2O2). In a separate
dark control experiment, where 20 mM H2O2 was mixed
with the 15 fold diluted exhausted reactive dyebath
solution at pH 7 for 1 h, no degradation took place at all
(data not shown). These observations revealed that the
studied monochlorotriazine type reactive dyes were
photolytically stable and their degradation necessitates
the application of H2O2 catalyzed UV-C photolysis at

Fig. 5. Eect of initial H2O2 concentration on the H2O2/UV-C

oxidation of simulated reactive dyebath euent. Experimental
conditions: Reaction pH=7.0; Treatment time=1 h.

appropriate H2O2 doses. The same results for the dark

and mere UV-C control experiments were obtained in
the case of reactive dyestus with vinylsulphone anchor
groups [16].
From the above ndings it is obvious that increasing
the pH to the natural alkaline pH of reactive dyebath
euents had a more detrimental eect upon the
treatment performance of the H2O2/UV-C oxidation
system than of the ozonation process. Consequently, the
achievement of maximum decolorization and mineralization rates required dierent optimum reaction conditions for the investigated AOPs.
3.4. TiO2/UV-A treatment
Numerous experimental data have been published
dealing with the eect of pH on heterogeneous photocatalytic degradation of various organic compounds
[26,27]). Several studies have shown that changes in
reaction rates were rather insignicant even at extreme
pH values [28]. For charged substrates, on the other
hand, a signicant dependency of both adsorption and
photocatalytic degradation eciency upon the pH value
has often been observed. This eect was explained by the
strong dependency of the overall surface charge and
hence the adsorptive properties of TiO2 particles on the
solution pH [28]. Generally speaking, alkaline pH values
have been found to be favorable for the photocatalytic
degradation of pollutant molecules present in their
cationic form while negatively charged molecules are
more readily degraded at low pH values when the
photocatalyst
surface
is
positively
charged
(pHzpc(TiO2)D6.8). At high pH values, OH
ions
adsorbed on the surface of the photocatalyst particles
act as ecient traps for the photogenerated holes [29];

d
OH

surf h -OHsurf

In the dark equilibration adsorption tests carried

out for 1 h in the DSSR with the 15-fold simulated
reactive dyebath euent and 1 g/L suspended P25-TiO2
powder at pH 4, 7, 9 and 11, a signicant adsorption
of the anionic reactive dyes was observed in the acidic
pH. As can be seen in Fig. 6ac when comparing
the initial (t 0) values for A525 (6a), UV280 (6b) and
TOC (6c), the initial adsorption observed at pH 4
in the absence of irradiation accounted for 10.9%
initial color, 8.8% UV280 and 5.8% TOC reduction,
respectively.
The removal in all selected process parameters was
relatively slow and even retarded at the beginning of the
photocatalytic oxidation for pH 711. The length of the
induction period of the UV280 removal kinetics almost
coincided with the time of decolorization that was
practically complete after 3040 min treatment (Fig. 6a
and b). This is the expected behavior of the photocatalytic system. The dyes absorb the incident visible

1151

I.A. Alaton et al. / Water Research 36 (2002) 11431154

Table 5
Eect of pH on the TiO2/UV-A treatment of simulated reactive
dyebath euent (TiO2 dose=1000 mg/L)
Reaction
pH

kd
(1/min)

rid
(1/m  min)

kUV280
(1/min)

TOC
(%)

3.0
7.0
9.0
11.0

0.04
0.05
0.04
0.02

0.24
0.31
0.28
0.15

0.02
0.02
0.02
0.01

12.5
10.3
9.0
0.0

oxidation processes in the bulk phase rather than at the

semiconductor surface also played a major role during
the TiO2/UV-A treatment of the simulated dyehouse
euent.
Table 5 presents kd ; kUV280 ; rid values together with the
overall per cent TOC removal eciencies obtained for
1 h photocatalytic treatment of 15-fold diluted reactive
dyebath euent at dierent reaction pH.
At the end of the photocatalytic oxidation period,
overall color and UV280 removal ranged between 89.5
94.6% and 69.375.5% at varying pH values.
3.5. Comparison of electrical energy requirements

Fig. 6. Eect of pH on A525 (a), UV280 (b) and TOC (c)

abatement kinetics during TiO2/UV-A treatment of the
simulated reactive dyebath euent. Experimental conditions:
TiO2 dose=1000 mg/L; Treatment time=1 h.

and UV-A light eciently thus preventing the photons

from reaching the TiO2 particles (screening eect). Only
when the solution becomes transparent for UV-A
photons, a signicant photocatalytic eect is to be
expected thus explaining this observation. Among the
investigated pH values, the heterogeneous photocatalytic treatment rate was seriously inhibited only at pH 11.
An approximately two fold decrease in color and UV280
removal was found by increasing the pH from 7 to 11,
whereas no TOC reduction occurred at pH 11. Though
the studied dyes were not signicantly adsorbed on the
photocatalyst surface at pH 7 and 9, no parallel decrease
in the reaction rates was observed as compared to the
results obtained at acidic pH (pH 4) indicating that

In energy-intense treatment processes such as UV-C

light induced photochemical processes (H2O2/UV-C,
O3/UV-C etc.), the key design variables, i.e., exposure to
UV-C radiation and order of magnitude of contaminant
concentration removal, can be combined into a single
function, the so-called electrical energy per order of
pollutant removal (EE/O). The EE/O is a powerful
scale-up parameter and a measure of the treatment rate
obtained in a xed volume of contaminated water as a
function of the applied specic energy dose (exposure to
UV light or O3) [30]. This simple gure-of-merit make
the comparison of treatment performances of dierent
AOPs as simple as comparing the slopes of the energy
dose vs. log Ci =Cf curves. Thus, the steeper the obtained
slope, the more ecient is the selected advanced
oxidation system.
Fig. 7 presents the log A0 =A (l 525 nm) values
obtained during advanced oxidation of the 15-fold
diluted simulated reactive dyebath euent as a function
of the applied electrical energy during ozonation (a) and
H2O2/UV-C oxidation (b). For comparative purposes,
the specic O3 input (dose) was converted into applied
electrical power rate per unit volume of treated wastewater during ozonation using a conversion factor of
10 kWh/lb of ozone generated [31]. Since heterogeneous
photocatalytic reactions via semiconductor powders
such as TiO2 can be realized using solar energy thus
eliminating the electrical energy costs associated with
UV-A lamps, the EE/O concept was not applied to
TiO2/UV-A treatment of the reactive dyebath euent.

1152

I.A. Alaton et al. / Water Research 36 (2002) 11431154

Fig. 7. Decolorization of the simulated reactive dyebath euent as a function of the applied electrical energy for ozonation (a) and
H2O2/UV-C oxidation (b). Experimental conditions: O3 dose=1310 mg/L; H2O2 dose=680 mg/L; TiO2 dose=1000 mg/L; Reaction
pH=7.0; Treatment time=1 h.

Moreover, the observed induction period would make

the calculation of the EE/O very ambiguous in the latter
case.
From the results shown in Fig. 7 it is obvious that for
the ozonation process (total applied O3 dose at pH
7=1310 mg/L) decolorization was complete after 3 min
corresponding to an O3 dose of 1.43 kWh/m3 and thus
the EE/O was calculated for that specic time period. In
the case of H2O2/UV-C oxidation (20 mM H2O2; pH 7)
color removal was achieved after 10 min exposure to
UV-C irradiation equivalent to a UV-C dose of only
1.04 kWh/m3 thus resulting in a very similar EE/O value.
Here it should be pointed out that for estimation of the
total operating expenses like air preparation and H2O2
costs have also to be considered for the ozonation and
H2O2/UV-C processes, respectively.

4. Conclusions
The comparison of dierent OHd generation techniques is of interest, on the one hand, to determine the
most ecient experimental conditions for the destruction of organic pollutants present in dyehouse euents,
and, on the other hand, to provide useful information
for the understanding of the mechanism of advanced
oxidation processes. The following conclusions can be
drawn from the experimental work:
*

Among the investigated AOPs, the quickest color

(rid 11:8 1/m  min), UV280 (kUV280 0:35 1/min)
and highest overall TOC removal (40.5%) were
obtained by ozonation at pH 7 and an ozone dose
of 2970 mg/L. Further, since O3 treatment is not

I.A. Alaton et al. / Water Research 36 (2002) 11431154

merely based on OHd production, it will not be

aected as seriously by the high carbonate alkalinity
of the dyehouse euent as the H2O2/UV-C and the
TiO2/UV-A processes.
From the economical point of view, the H2O2/
UV-C process emerges as the most attractive
advanced oxidation system for reactive dyebath
euents in terms of complete decolorization
(EE/O=0.633 kWh/m3  order) very closely followed
by the ozonation process (EE/O=0.684 kWh/m3 
order).
TiO2/UV-A treatment proceeded relatively slowly
even at neutral pH where the highest decolorization
(rid 0:31 1/m  min; kd 0:047 1/min), dearomatization (kUV280 0:024 1/min at pH 7) and the highest
photocatalytic TOC reduction rate (10.26%) were
achieved. Considering operating costs, the TiO2/UVA process can only compete with O3 or H2O2/UV-C
when solar light is used instead of articial UV-A
light.

Acknowledgements
Idil Arslan is grateful to TUBITAK BAYG for the
NATO A2 scholarship. The authors also acknowledge
Bogazic-i Research Foundation for the nancial support
through grant 98Y-03.

References
[1] Correia VM, Stephenson T, Judd SJ. Characterization of
textile wastewaterFa review. Environ Technol 1994;15:
91729.
[2] Erswell A, Brouckaert CJ, Buckley CA. The reuse of
reactive dye liquors using charged ultraltration membrane technology. Desalination 1988;70:15767.
[3] Grau P. Textile industry wastewaters treatment. Water Sci
Technol 1991;24:97103.
[4] Easton JR. Color in dyehouse euent. In: Cooper P,
editor. The Society of Dyers and Colorists. Oxford: Alden
Press, 1995. p. 621.
[5] Reife A, Freeman S. Environmental chemistry of dyes and
pigments. Canada: John Wiley & Sons Inc., 1996.
[6] Legrini O, Oliveros E, Braun AM. Photochemical processes. Chem Rev 1993;93:67198.
[7] Glaze WH, Kang JW, Chapin DH. The chemistry of water
treatment processes involving ozone, hydrogen peroxide
and ultraviolet radiation. Oz Sci Eng 1987;9:33552.
[8] Arslan I, Akmehmet Balcioglu I. Degradation of commercial reactive dyestus by heterogeneous and homogenous
advanced oxidation processes: a comparative study. Dyes
Pigments 1999;43:18995.
[9] Chen G, Lei L, Yue PL. Wet oxidation of high-concentration reactive dyes. Ind Eng Chem Res 1999;38:
183743.

1153

[10] Chun H, Yizhong W. Decolorization and biodegradability

of photocatalytic treated azo dyes and wool textile
wastewater. Chemosphere 1999;39(12):210715.
[11] Herrera F, Kiwi J, Lopez A, Nadtochenko V. Photochemical Decoloration of Remazol Brilliant Blue and
Uniblue A in the presence of Fe3+ and H2O2. Environ Sci
Technol 1999;33:314551.
.
[12] Ince NH, Gonen
c- DT. Treatability of a textile azo dye by
UV/H2O2. Environ Technol 1997;18:17985.
[13] Peralta-Zamora P, Kunz A, Gomes de Moraes S, Pelegrini
R, Moleiro PC, Reyes J, Duran N. Degradation of reactive
dyes 1. A comparative study of ozonation, enzymatic and
photochemical processes. Chemosphere 1998;38(4):83552.
[14] Solzhenko EG, Soboleva NM, Goncharuk VV. Decolorization of azo dye solutions by Fentons oxidation. Water
Res 1995;29:220610.
[15] Vinodgopal K, Peller J, Makogon O, Kamat PV. Ultrasonic mineralization of a reactive textile azo dye, Remazol
Black B. Water Res 1998;32(12):364650.
[16] Arslan I, Akmehmet Balcioglu I, Tuhkanen T. Oxidative
treatment of simulated dyehouse euent by UV and nearUV light assisted Fentons reagent. Chemosphere 1999;
39(15):276783.
[17] Arslan I, Akmehmet Balcioglu I. Eect of common
reactive dye auxiliaries on the ozonation of vinylsulphone
and aminochlorotriazine containing dyehouse euents.
Desalination 2000;130:6171.
. E. Decolorization treatments of azo dye
[18] Uygur A, Kok
wastewaters including dichlorotriazinyl reactive groups by
using advanced oxidation method. JSDC 1999;115:3504.
[19] Nicole I, De Laat J, Dore M, Duguet JP, Bonnel C.
Utilisation du rayonnement ultraviolet dans le traitement
de eaux: measure du ux photonique par actinometrie
chimique au peroxyde dhydrogene. Water Res 1990;24:
15768.
[20] Goslich R, Dillert R, Bahnemann DW. Solar water
treatment: principles and reactors. Water Sci Technol
1997;15(4):13748.
[21] Glaze WH, Kang JW. Advanced oxidation processes: test
of a kinetic model for the oxidation of organic compounds
with ozone and hydrogen peroxide in a semi-batch reactor.
Ind Eng Chem Res 1989;28(11):15807.
[22] Safarzadeh-Amiri A, Bolton JR, Cater S. Ferrioxalatemediated photodegradation of organic pollutants in
contaminated water. Water Res 1997;31:78798.
[23] Poon CS, Huang Q, Fung PC. Degradation kinetics of
Cuprophenyl Yellow RL by UV/H2O2/Ultrasonication
(US) process in aqueous solution. Chemosphere 1999;
38(5):100514.
[24] Shu HY, Huang CR, Chang MC. Decolorization of monoazo dyes in wastewater by advanced oxidation process: a
case study of acid red 1 and acid yellow 23. Chemosphere
1994;29:2597607.
[25] Walling C. Fentons reagent revisited. ACC Chem Res
1975;8:12531.
[26] Bahnemann DW. In: Boule P, editor. The handbook of
environmental chemistry. Heidelberg: Springer, 1999. p.
285351.
[27] Homann MR, Martin ST, Choi W, Bahnemann DW.
Environmental applications of semiconductor photocatalysis. Chem Rev 1995;95:6996.

1154

I.A. Alaton et al. / Water Research 36 (2002) 11431154

[28] Bahnemann DW, Cunningham J, Fox MA, Pelizetti E,

Pichat P, Serpone N. In: Helz GR, Zepp RG, Crosby D.
editors. Aquatic and surface photochemistry. Boca Raton:
Lewis Publishers, 1994. p. 261316.
[29] Heimenz PC. Principles of colloid and surface photochemistry. New York: Marcel Dekker, 1986.

[30] Bolton JR, Bircher KG, Tumas W, Tolman CA. Figuresof-merit for the technical development and application of
advanced oxidation processes. J Adv Oxid Technol
1996;1(1):137.
[31] Calgon, Carbon oxidation technologies. The AOT Handbook. Ontario: Markham, 1995.