PII: S0043-1354(00)00330-4

Wat. Res. Vol. 35, No. 4, pp. 10931099, 2001

# 2001 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
0043-1354/01/$ - see front matter

OZONATION OF AQUEOUS AZO DYE IN A SEMI-BATCH

REACTOR
JIANGNING WU* and TINGWEI WANG
School of Chemical Engineering, Ryerson Polytechnic University, Toronto, Ontario, Canada M5B 2K3
(First received 1 July 1999; accepted in revised form 1 June 2000)
Abstract}The ozonation of wastewater containing azo dye in a semi-batch reactor has been studied.
Results revealed that the rate of ozone transfer increased with increases in the initial dye concentration, the
applied ozone dose and temperature. A model was developed to predict the enhancement factor of ozone
mass transfer. This model enables the prediction of mass transfer coefficient of ozone from the following
parameters: initial dye concentration, applied ozone dose, temperature and concentration of dissolved
ozone in the organic-free water. This model was also valid for reactors of larger sizes. Result of kinetic
studies showed that ozonation of the azo dye was a pseudo-first-order reaction with respect to dye. The
apparent rate constant increased with the applied ozone dose and temperature. However, the apparent rate
constant declined logarithmically with increasing initial dye concentration. In addition, ozonation reduced
chemical oxygen demand and enhanced the biodegradability of the wastewater. # 2001 Elsevier Science
Ltd. All rights reserved
Key words}ozonation, azo dye, ozone mass transfer, decolorization kinetics, textile wastewater,
biodegradability

INTRODUCTION

Azo dyes are characterized by nitrogen to nitrogen

double bonds (N=N). They are the most used dyes
in textile industry, accounting for over 50% of all
commercial dyes (Waring et al., 1990). However,
large amount of azo dyes remain in the effluent after
the completion of the dyeing process and most azo
dyes are non-biodegradable (Liakou et al., 1997;
Razo-Flores et al., 1997). Therefore, subsequent
biological treatment has almost no effect on the
color of textile effluents. Consequently, the textile
wastewaters are heavily colored with a large variation
in chemical composition. In addition to the heavy
color and the potential hazard of the azo dyes, the
quality of the textile wastewaters is further deteriorated by part or all of the following: the heavy metals
associated with some of the dyes, the high chemical
oxygen demand (COD) and total organic carbon
(TOC) contents, the large amount of inorganic salts
such as NaCl , the suspended solids, etc. (Tzitzi et al.,
1994; Razo-Flores et al., 1997; Wu et al., 1998).
Heavily colored textile wastewater can be effectively treated by ozonation (Law et al., 1996;
Hitchcock et al., 1998). It has been confirmed that
the rate-limiting step in the ozonation of dye-

*Author to whom all correspondence should be addressed.

Tel.: +1-416-979-5000; fax: +1-416-979-5044; e-mail:
j3wu@acs.ryerson.ca

containing wastewater is the mass transfer of ozone

from the gas phase to the wastewater (Saunders et al.,
1983; Carriere et al., 1993; Shu et al., 1995).
However, the driving force for ozone mass transfer,
which is the difference between the concentration of
the dissolved ozone and the equilibrium ozone
concentration at the gasliquid interphase, is system-dependent. The dependence of the driving force
on the wastewater system results from the fact that
the concentration of the dissolved ozone varies
considerably with the rate of the self-decomposition
of ozone, which in turn depends on the characteristics of the wastewater. In particular, as indicated
above, textile wastewaters are extremely variable in
composition (Tzitzi et al., 1994). A reliable and
generally valid model for the quantitative prediction
of the mass transfer coefficient of ozone has not yet
been developed. Consequently, the rate of ozone
mass transfer must be determined system by system.
The dye contents of textile wastewaters and their
temperatures are significantly different depending on
the locations from which the wastewater is collected.
Wastewater from the dye-bath without dilution by
the rinse water contains large amount of dyes and has
a high temperature, while the textile mill effluent
discharging to the Public-Owned Treatment Works
(POTWs) is much more diluted and the temperature
varies with the seasons. Researchers have observed
that ozonation of some dye-containing wastewater
was a pseudo-first-order reaction with respect to dye

1093

1094

Jianging Wu and Tingwei Wang

concentration and that the rate of decolorization

decreased with the increase in dye concentration
(Carriere et al., 1993; Shu et al., 1995). However, the
relationship between the apparent rate constant and
the initial dye concentration has not been studied,
although a logarithmic relationship was reported
previously by Wu and co-workers for a wastewater
containing various dyes and a number of proprietary
auxiliary ingredients (Wu et al., 1998).
Despite the fact that ozone is a very strong
oxidant, complete mineralization of azo dyes in
wastewater by ozonation is usually not achievable
(Tzitzi et al., 1994; Lopez et al., 1998). On the
other hand, biological treatment of wastewater
is the most economical method (Weber et al., 1991)
so that biological removal of the ozonation products
following ozonation is cost-effective. Therefore,
systematic investigation of the changes in biodegradability of the azo dyes after ozonation is quite
necessary.
In this study, the ozonation of aqueous azo dye in
a semi-batch reactor was investigated. An azo dye,
Reactive Black 5, was selected as the model
compound. The objectives are to develop a model
to predict the mass transfer coefficient of ozone from
the gas phase to the wastewater, to investigate the
relationships between the rate constant of ozonation
and the reaction conditions, and to examine the effect
of ozonation on the biodegradability of azo dyes
under various experimental conditions.

EXPERIMENTAL

Materials
The synthetic wastewater was prepared using deionized
water at various dye concentrations to simulate those in unrinsed dye baths and in diluted mill effluents (Wu et al.,
1998).
Reactive black 5 was purchased from Sigma-Aldrich
Canada Ltd. (Oakville, ON). All other chemicals used for
measuring dissolved ozone concentration and BOD were
reagent grade and were obtained from Sigma-Aldrich or
BDH Inc. (Toronto, ON). COD digestion vials were
supplied by VWR Canlab (Mississauga, ON).

Analytical methods
Sample color was measured by an integration method
developed previously (Law et al., 1996; Wu et al., 1998).
This method involves scanning the absorbance of a sample
from 400 to 700 nm (Beckman DU 650 spectrophotometer,
Beckman Instruments Canada Inc, Mississauga, ON) and
integrating the area under the absorbance curve. The
integrated area is expressed in integrated absorbance units
(IAU), which is directly proportional to sample color. The
integration method is simpler than the American Dye
Manufactures Institute (ADMI) tristimulus filter method
(APHA, 1992) and the two methods have been shown to
yield similar results (Wu et al., 1998).
The concentration of dissolved ozone was measured by
the indigo method (APHA, 1992, method no. 4500-O3). The
biological oxygen demand (BOD) and COD of samples were
all measured according to standard methods (APHA, 1992).

Apparatus and procedure

Ozone was generated from pure oxygen in a Model GL-1
ozone generator (PCI-WEDECO Environmental Technologies, West Caldwell, NJ). The concentrations of ozone in
the output gas of the ozone generator were measured by an
ozone monitor (Model HC-400, PCI-WEDECO Environmental Technologies). The reactor was a 0.5 l gas washing
bottle equipped with a porous gas diffuser, and other
reactors used to test the model developed in this study were
4 l and 6 l in volume, respectively. Excess ozone leaving the
reactor was destroyed by a catalytic ozone-destruct unit
filled with Carulite catalyst (Carus Chemical Company,
Peru, IL). The temperature of the reactor was controlled by
a constant temperature circulating bath (Model 1187, VWR
Canlab, Mississauga, ON).
The ozonation process began when the ozoneoxygen
mixture provided by the ozone generator was sparged into
the reactor, which was operated in semi-batch mode by
feeding the ozone-containing gas continuously. Samples
were taken at appropriate time intervals to analyze color
and concentration of dissolved ozone. In addition, the
BOD5 and COD of wastewater samples were measured
before and after 30 min ozonation.

RESULTS AND DISCUSSION

Model development for the prediction of mass transfer

coefficient of ozone
In the case where ozone is absorbed into water and
simultaneously undergoes reaction, the mass balance
of ozone in a completely mixed semi-batch reactor at
steady state is
dC
kL aC  C
1
dt
where C is the concentration of dissolved ozone, C is
the equilibrium ozone concentration at the gasliquid
interphase, and kL a is the volumetric mass transfer
coefficient of ozone in the presence of chemical
reactions (Beltran et al., 1992). Integrating equation
(1) with boundary conditions of C 0 at t 0 and
C C at t t yields
lnC  C=C  kL a t

The values of kLa can be obtained by plotting the

left-hand side of equation (2) against the ozone
wastewater contact time.
In the absence of chemical reactions, the volumetric mass transfer coefficient of ozone is expressed
as koL a. The values of koL a can be determined in the
same way as that mentioned above for kLa, with the
exception that ozone should be bubbled into organicfree water instead of wastewater.
As for the values of C, they are functions
of experimental conditions. Since ozone is only
slightly soluble in water, the resistances to mass
transfer of ozone in gas phase and interface are
insignificant (Levenspiel, 1972). The values of C are
determined when the concentrations of dissolved
ozone stop changing with the gasliquid contact
time.
By defining E as the enhancement factor (Dankwerts, 1970; Levenspiel, 1972), E is expressed as

Ozonation of aqueous azo dye

1095

(Tufano et al., 1992, Tufano et al., 1994)

E

kL a
koL a

The enhancement factor is a measure of increase in

the rate of ozone mass transfer due to chemical
reactions.
Experiments were conducted under different conditions to obtain the values of koL a and kLa, then the
corresponding values of E were calculated using
equation (3). During ozonation, the color of the
wastewater decreased and the concentration of the
dissolved ozone increased with time. A representative
plot is shown in Fig. 1. The ozonation time used to
calculate the values of kL oa. kLa and E was
approximately 5 min before the equilibrium concentration of ozone was reached. In the case of Fig. 1,
the time used was 10 min.
Figure 2 shows the dependence of E on initial dye
concentration (Cdye), temperature (T) and applied
ozone dose (Dozone, mass of ozone applied to
the reactor per unit time per unit liquid volume).
The applied ozone dose was adjusted by changing the
ozone percentage in the gas while maintaining the gas
flowrate at 0.52 l/min and pressure at 82.73 kPa.
According to Fig. 2, E increased linearly with all
the three parameters. Consequently, the following
model was developed by the least-squares regression
of the experimental data:
E 0:3219 0:3798Cdye 0:0238Dozone 0:0279T
4
where the units of Cdye, Dozone and T are g/l, mg/l min
and 8C, respectively. Results of statistical analyses
indicated that E has a strong correlation with the
three parameters (R2=0.9788) and that each of the
three parameters has significant effects on E.
Fig. 2. Dependance of the enhancement factor on the initial
dye concentration, the applied ozone dose and temperature.

Fig. 1. Color and concentration of dissolved ozone as

functions of the ozonation time for the dye containing
wastewater (initial dye concentration=0.5 g/l, applied
ozone dose=26.1 mg/l min, T=208C). The units of IAU
and min 1 apply to 1705 and 0.1469, respectively.
* , color; * , concentration of dissolved ozone.

As indicated above in equation (4), E increases

linearly with the dye concentration. This result might
be explained from the reaction kinetics. Theoretically, the ozonation of dye is first order with respect
to ozone and the dye, respectively (Langlais et al.,
1990). Obviously, the rate of ozonation should
increase with the concentrations of both the ozone
and dye. When the ozone concentration is constant,
which was the case in this study as described later, the
ozonation becomes pseudo-first order with respect to
the dye, and the rate of the reaction increases with the
dye concentration linearly. Consequently, the values
of E increase linearly, simply because the rate of
ozone mass transfer is enhanced by the chemical
reaction. As for the linear increase of E with the
applied ozone dose, the explanation might be that
under the experimental conditions, increasing the
ozone input dosage could increase the turbulence in
the gasliquid interface, which then enhances the rate

1096

Jianging Wu and Tingwei Wang

of mass transfer. On the other hand, as for the effect

of temperature on E, it seems that in the tested
temperature range, the increase in reaction rate with
the temperature rise is more significant than the
decrease caused by the decrease in ozone solubility
due to the temperature rise.
Further experiments were conducted in reactors of
various volumes to test the validity of equation (4) in
reactors of larger volume. Firstly, the predicted E
values were calculated directly by substituting certain
values of Dozone, T and Cdye into equation (4). Then
the same values of Dozone and T as above and the
same values of Dozone, T and Cdye as above were used
in experiments conducted in the organic-free water
and in the synthetic wastewater, respectively. By
measuring the concentrations of dissolved ozone
during experiments, the values of koL a and kLa, and
ultimately, the values of experimental E were
determined.
For tests conducted in 0.5 l reactor, a gas flowrate
of 0.75 l/min, which differed from that used in
deriving equation (4), was used and the pressure
remained at 82.73 kPa. A comparison of the values of
the predicted E with the experimental E is shown in
Table 1. The values of the predicted E and experimental E are in agreement.

Tests of equation (4) were also conducted in

reactors of larger sizes. The larger reactors were the
same type as the 0.5 l one and their volumes were 4 l
and 6 l, respectively. The gas flowrate was 2.0 l/min
at 82.73 kPa. Table 2 shows the testing results. As
can be observed from Table 2, equation (4) can
predict E values for reactors of extended sizes
successfully.
For the prediction of the mass transfer coefficient
of ozone, firstly, the value of E should be predicted
by using equation (4). Then only the organic-free
water needs to be ozonated to obtain values of koL a.
Finally, kLa can be calculated from the values of E
and koL a by using equation (3).
Decolorization kinetics
The decolorization of dyes by ozone is first order
with respect to ozone and the dye, respectively
(Langlais et al., 1990). When the amount of ozone
is in excess, the reaction is pseudo-first-order with
respect to dye. The pseudo-first-order trend was
observed in each of the experimental runs in this
study. The representative curve and the corresponding first-order correlation are shown in Fig. 1. The
same pseudo-first-order behaviors were also reported

Table 1. Results of testing equation (4) in 0.5 L reactor (Gas flowrate=0.75 l/min, gas pressure=82.73 kPa (12 psi))
Cdye (g/l)
0.30
0.80
1.25
1.75
2.25
a

Dozone (mg/l min)

T (8C)

Predicted E

Experimental E

Error in predicted value (%)a

15.1
62.9
56.6
10.1
31.4

40
10
25
15
10

1.91
2.40
2.84
1.64
2.20

1.88
2.51
2.75
1.62
2.26

1.60
4.38
3.27
1.23
2.65

Error in predicted value=((Experimental E Predicted E)/Experimental E)  100%.

Table 2. Results of testing equation (4) in 4 and 6 L reactors (gas flowrate=2.0 l/min, gas pressure=82.73 kPa (12 psi))
Cdye (g/l)

Dozone (mg/l min)

T (8C)

0.35
0.50
0.50
0.75
1.25
1.50
1.50
2.30
2.50
2.50

32.3
21.0
33.5
33.5
21.0
21.0
33.5
29.4
21.0
33.5

15
20
20
10
40
30
20
25
30
20

0.25
0.50
0.50
0.80
1.50
1.50
1.85
2.35
2.50
2.50

19.6
14.0
22.4
22.4
14.0
22.4
22.4
17.6
14.0
22.4

10
20
20
25
30
20
35
40
30
20

Experimental E

Error in predicted value (%)a

4 l reactor
1.64
1.57
1.87
1.68
2.41
2.23
2.25
2.59
2.61
2.63

1.67
1.52
1.84
1.71
2.47
2.29
2.30
2.51
2.59
2.71

1.80
3.29
1.63
1.75
2.43
2.62
2.17
3.19
0.77
2.95

6 l reactor
1.16
1.40
1.60
1.86
2.06
1.98
2.53
2.75
2.44
2.36

1.19
1.37
1.56
1.81
2.14
1.89
2.61
2.86
2.57
2.27

2.52
2.19
2.56
2.76
3.74
4.76
3.07
3.85
5.06
3.96

Predicted E

Error in predicted value=((Experimental E Predicted E)/Experimental E)  100%.

Ozonation of aqueous azo dye

elsewhere (Saunders et al., 1983; Carriere et al., 1993;

Shu et al., 1995).
The effects of the initial dye concentration, the
applied ozone dose and temperature on the apparent
rate constant were investigated and Fig. 3 illustrates
the results. According to Fig. 3, the rate constant k
increases with both the applied ozone dose and
temperature. Also, the apparent rate constant
declines logarithmically with the initial dye concentration as follows:
0:9678
k 0:0616Cdye

5
1

and g/l,
where the units of k and Cdye are min
respectively. The logarithmic relationship between
the apparent rate constant and the initial dye
concentration derived in this study is in agreement
with that observed from our previous study involving
a textile wastewater system containing azo dyes. The
previous textile wastewater was simulated according
to a formula from a local dyehouse at Georgia for

1097

yarn dyeing, containing three dyes and a number of

proprietary auxiliary ingredients (Wu et al., 1998).
Obviously, the composition of the previous textile
wastewater are very different from and much more
complicated than that of the synthetic wastewater
used in this study. The consistency between the
results of the current and previous wastewater
systems indicated that the linear log (k)log (Cdye)
relationship is always true for the azo dyes regardless
of the co-existence of other components. However,
the validity of this linearity should be further
confirmed by applying the model to other textile
wastewater systems.
Effect of ozonation on the biodegradability of the
azo dye
Ozonation could increase the biodegradability of
some recalcitrant compounds and the ratio of BOD5
to COD is often used as a measurement of

Fig. 3. Effects of the initial dye concentration, the applied ozone dose and temperature on the apparent
rate constant: (a) Dozone=26.1 mg/l min, T=208C; (b) Cdye=0.5 g/l, T=208C; and (c) Dozone=26.1 mg/
l min, Cdye=0.5 g/l.

1098

Jianging Wu and Tingwei Wang

Table 3. COD removal and biodegradability increase of reactive black 5 after 30 min ozonation (gas flowrate=0.52 l/min, gas pressure
=82.73 kPa (12 psi))
(COD)0a
(mg/l)

(COD)30b
(mg/l)

COD
removal (%)

(BOD5)0c
(mg/l)

(BOD5)30d
(mg/l)

Increase in
BOD5 : COD
ratioe

86.5
69.3
38.9
26.6

3.01
3.32
3.64
4.11

26.9
27.3
29.4
32.9

65.5
26.8
13.2
10.9

63.8
69.7
72.8
74.8

3.32
3.32
3.32
3.32

24.4
28.5
27.8
30.1

20.3
28.3
30.7
36.0

68.7
71.0
70.6

3.32
3.32
3.32

29.1
30.2
30.0

27.9
31.4
30.8

Dozone=26.1 mg/l min, T=208C

Cdye (g/l)
0.20
0.50
1.00
2.00

110
290
585
1223

15
89
358
898
Cdye=0.5 g/l, T=208C

Dozone (mg/l min)

16.7
34.8
43.6
69.7

290
290
290
290

105
88
79
73
Cdye=0.5 g/l, Dozone=26.1 mg/l min

T (8C)
10
30
40

290
290
290

91
84
85

Sample COD before ozonation (mg/l).

Sample COD after 30 min ozonation (mg/l).
c
5-day BOD of the sample before ozonation (mg/l).
d
5-day BOD of the sample after 30 min ozonation (mg/l).
e
Increase in BOD5 : COD ratio=((BOD5)30/(COD)30) : ((BOD5)0/(COD)0).
b

biodegradability (Karrer et al., 1997; Lopez et al.,

1998). Therefore, the azo dye-containing wastewater
in this study was ozonated to examine this effect. It
was found from our experiments that after 30 min
ozonation, values of sample COD and BOD did not
change with ozonation time. Therefore, the ozonation time was set for 30 min. It was observed that
after ozonation, the COD reduction of the wastewater samples ranged 2787% and the increase in
BOD5 : COD ratio was 1166 times as demonstrated
in Table 3. These results indicate that the ozonation
of the azo dye can reduce the sample COD, and can
increase the biodegradability of the azo dye.

CONCLUSIONS

A model was developed to predict the enhancement factor of ozone mass transfer in semi-batch
reactors for the ozonation of aqueous azo dyes. By
applying this model, the volumetric mass transfer
coefficient of ozone can be predicted from the initial
dye concentration, the applied ozone dose, the
reaction temperature and the concentration of
dissolved ozone in organic-free water. This model
has been validated in reactors that vary from 0.5 to
6 l volumes.
Decolorization of the azo dye by ozone is a
pseudo-first-order reaction with respect to dye. The
apparent rate constant declines logarithmically with
increased initial dye concentration. This logarithmic
relationship is qualitatively in agreement with that
observed previously from a much complicated wastewater system containing azo dye mixtures and a
number of auxiliary ingredients.

Ozonation of the aqueous azo dye can reduce the

sample COD 2787% and can increase biodegradability of the wastewater 1166 times.
Acknowledgements}J. W. wishes to express her gratitude
for the financial support provided by the Natural Sciences
and Engineering Research Council of Canada. T. W. was a
visiting scholar in our lab and he was financially supported
by Chinese Government (PRC).

REFERENCES

APHA (1992) Standard Methods for the Examination of

Water and Wastewater. eds. Greenberg A. E., Clesceri L.
S. and Eaton A. D. 18th ed. American Public Health
Association, Washington, DC.
Beltran F. J., Encinar J. M., Garcta Araya J. F. and
Alonnso M. A. (1992) Kinetic study of the ozonation of
some industrial wastewaters. Ozone Sci. Engng. 14,
303327.
Carriere J., Jones P. and Broadbent A. D. (1993)
Decolorization of textile dye solutions. Ozone Sci. Engng,
15, 189200.
Danckwerts P. V. (1970) GasLiquid Reactions. McGrawHill Book Company, New York, pp. 34, 105.
Hitchcock D. R., Law S. E., Wu J. and Williams P. L.
(1998) Determining toxicity trends in the ozonation of
synthetic dye wastewaters using the nematode. Caenorhabditis elegans. Arch. Environ. Contam. Toxicol. 34,
259264.
Karrer N. J., Ryhiner G. and Heinzle E. (1997) Applicability test for combined biologicalchemical treatment of
wastewater containing biorefractory compounds. Wat.
Res. 31(5), 10131020.
Langlais B., Reckhow D. A. and Brink D. R. (1990) Ozone
in Water Treatment: Application and Engineering. Lewis
Publishers, Boca Raton, FL, pp. 44 and 117121.
Law S. E., Wu J. and Eiteman M. A. (1996) Ozone
decolorization of cotton dyehouse wastewater. Proceed-

Ozonation of aqueous azo dye

ings of ASAE Annual International Meeting, July 1418,
Phoenix, AZ.
Levenspiel O. (1972) Chemical Reaction Engineering. Wiley,
New York, pp. 418419.
Liakou S., Pavlou S. and Lyberatos G. (1997) Ozonation of
azo dyes. Wat. Sci. Technol. 35(4), 279286.
Lopez A., Ricco G., Mascolo G., Tiravanti G., Pinto A. C.
D. and Passino R. (1998) Biodegradability enhancement
of refractory pollutants by ozonation: A laboratory
investigation on an azo-dyes intermediate. Wat. Sci.
Technol. 38, 239245.
Razo-Flores E., Luijten M., Donlon B., Lettinga G.
and Field J. (1997) Biodegradation of selected azo dyes
under methanogenic conditions. Wat. Sci. Technol. 36,
6572.
Saunders F. M., Gould J. P. and Southerland C. R. (1983)
The effect of solute competition on ozonolysis of
industrial dyes. Wat. Res. 17(10), 14071419.
Shu H. and Huang C. (1995) Degradation of commercial azo dyes in water using ozonation and UV
enhanced ozonation process. Chemosphere. 31(8),
38133825.

1099

Tufano V., Andreozzi R. and Caprio V. (1994) Optimal

operating conditions for lab-scale ozonation reactors.
Ozone Sci. Engng. 16, 181195.
Tufano V., Andreozzi R., Caprio V., DAmore M. G. and
Insola A. (1992) Kinetic analysis of complex gas-liquid
reactions: transport limitations in the slow-fast transition
regime. Chem. Engng. Sci. 47, 22392244.
Tzitzi M., Vayenas D. V. and Lyberatos G. (1994)
Pretreatment of textile industry wastewaters with ozone.
Wat. Sci. Technol. 28(9), 151160.
Waring D. R. and Hallas G. (1990) The Chemistry and
Application of Dyes. Plenum Press, New York, pp. 1821.
Wu J., Eiteman M. A. and Law S. E. (1998) Evaluation of
membrane filtration and ozonation processes for treatment of reactive-dye wastewater. J. Environ. Engng.
124(3), 272277.
Weber W. J. and Corseuil H. (1991) An engineered reactor
approach to integrating physicochemical and biological
processes for in-situ bioremediation of contaminated
subsurface systems. In Proceedings of the Specialty
Conference, Environmental Engineering, ASCE, Reno,
Nevada, p. 191.