Chemosphere 57 (2004) 1189–1199

www.elsevier.com/locate/chemosphere

Ozonation of Cationic Red X-GRL in aqueous

solution: degradation and mechanism
Weirong Zhao *, Huixiang Shi, Dahui Wang
Department of Environmental Engineering, Zhejiang University, Hangzhou 310027, PR China

Received 26 March 2003; received in revised form 11 June 2004; accepted 11 August 2004

Abstract

Ozonation of the azo dye Cationic Red X-GRL was investigated in a bubble column reactor at varying operating
parameters such as oxygen flow rate, temperature, initial Cationic Red X-GRL concentration, and pH. The conversion
of dye increased with the increasing of pH and oxygen flow rate. As the reaction rate constant and the volumetric mass
transfer coefficient increase while the ozone equilibrium concentration decreases with the temperature, there is a min-
imum conversion of dye at 25
C. The increasing of initial dye concentration leads to a decreasing conversion of dye
while the ozonation rate increases. The formation of intermediates and the variation of pH, TOC, and nitrate ion dur-
ing ozonation were investigated by the use of some analytical instruments such as GC/MS, GC, and IC. The interme-
diates of weak organic acids lower the pH value of the solution. The probable degradation mechanism of the Cationic
Red X-GRL in aqueous solution was deliberated with the aid of Molecular Orbital calculations. The N(12)–C(13) site
in Cationic Red X-GRL, instead of the N(6)–N(7) site, is found to be the principal site for ozone cycloaddition in the
degradation processes. During the degradation process, among the six nitrogen atoms of Cationic Red X-GRL, one is
transferred into a nitrate ion, one is converted into an amine compound, and the remaining four are transformed into
two molecules of nitrogen.
 2004 Elsevier Ltd. All rights reserved.

Keywords: Ozonation; Recalcitrant compounds; Degradation pathways; Decolorization

1. Introduction The chromophores of organic dyes usually are com-

Textile industries have shown a significant increase in
the use of synthetic complex organic dyes as the coloring
material. The annual world production of textiles is
about 30 million tons, requiring 700 000 tons of different
dyes per year (Talarposhti et al., 2001).
The color of dye is resulted from conjugated chains
or rings that can absorb different regions of wavelength.
*
Corresponding author. Tel.: +86 571 8795 1239; fax: +86
571 8795 2771.
E-mail address: weirong@mail.hz.zj.cn (W. Zhao).
posed of C@C bonds, N@N bonds, C@N bonds, and
aromatic and heterocyclic rings containing oxygen,
nitrogen or sulfur (Chu and Ma, 1998). The degradation
of molecules of dyes in the environment by microorgan-
isms is likely to be slow, which means that it is possible
for high levels of dye to persist, and potentially accu-
mulate. Therefore, the discharge of dye-house waste-
water into the environment is aesthetically displeasing,
impedes light penetration, damages the quality of the
receiving streams and may be toxic to treatment
processes, to food chain organisms and to aquatic life
(Ledakowicz et al., 2001). Although some treatment

0045-6535/$ - see front matter  2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chemosphere.2004.08.014
1190 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199
processes, such as chemical coagulation and carbon
adsorption, may remove certain categories of dye to
about 90% (Nam et al., 2001), the main drawback of
these processes is the generation of a large amount of
sludge or solid waste, resulting in high operational costs
for sludge treatment and disposal (Banat et al., 1996).
Ozonation, an easy-operated control technology
(Chu and Ma, 1998), is very effective in treating waste-
waters containing recalcitrant compounds (Peyton and
Glaze, 1988; Beltran-Heredia et al., 2001; Koch et al.,
2002; Chu and Ching, 2003). Ozone reacts with aqueous
compounds in two ways: direct reactions of the molecu-
lar ozone with the compounds; and indirect reactions of
the hydroxyl radicals resulting from the decomposition
of ozone catalyzed by hydroxyl ion. The direct reactions
are often highly solute selective and slow, whereas the
indirect hydroxyl radical reactions are non-selective,
fast, and proceeded more rapidly with increasing pH
(Beltran-Heredia et al., 2001). Usually, different reaction
pathways contribute to the ozonation processes simulta-
neously (Koch et al., 2002). The direct reactions are
dominant in acidic solutions, while the indirect hydroxyl
radical reactions constitute a significant portion of ozo-
nation at basic pH (Tomiyasu et al., 1985). The direct
reactions are very suitable for opening aromatic rings
by means of ozone cycloaddition, while the indirect
attacks are very suitable for mineralization of total
organic carbon (TOC).
The role of ozonation and its related oxidation proc-
ess in treating several organic dyes have been examined
by different authors (Shu and Huang, 1995; Liakou
et al., 1997; Alaton et al., 2002). However, most of these
works paid little attention to the intermediates and path-
ways during ozonation.
Cationic Red X-GRL, an azo dye, was selected for
the study because it is not biodegradable by the conven-
tional activated sludge process (Zhao et al., 2003), and it
is broadly used in the textile, color solvent, ink, paint,
varnish, paper, and plastic industries.
In the present work, ozonation experiments of azo
dye were investigated by varying the oxygen flow rate,
initial dye concentration, temperature, and pH. The for-
mation of intermediates, as well as the variation of pH,
TOC, and nitrate ion was studied. The ozonation mech-
anism of Cationic Red X-GRL in aqueous solutions was
studied with the aid of the molecular geometry and elec-
tronic structures analysis of the dye.
2. Experimental
2.1. Reagents
The commercial azo dye Cationic Red X-GRL, 5-[4-
(benzyl-methyl-amino)-phenylazo]-1,4-dimethyl-4H-
[1,2,4]triazol-1-ium chloride, was purified by the methanol
N
N N N
Cl -
N N
Fig. 1. Structural formula of Cationic Red X-GRL.
recrystallization method (Zhao et al., 2003). As can be
seen in Fig. 1, its cation part connects with a chloride
ion by ionic bond. The molecular weight of the dye is
356.84 g mol
1, and its maximum absorbance wave-
length is at 530 nm.
Ionic strength of 5.0 M concentrated phosphate buf-
fer solutions were prepared previously by the reaction
of a calculated amount of sodium hydroxide solution
and a phosphoric acid solution.
2.2. Apparatus
A schematic of the experimental set-up for ozonation
is depicted in Fig. 2. Ozone was produced from dry and
pure oxygen by a CHYF-3A model ozone generator
(Rongxing Elec. Co. Ltd., Hangzhou, China) that is able
to generate a maximum of 3 g ozone h
1. The oxygen
flow rate to the generator was monitored with a rotam-
eter incorporated into the ozone generator. The 3.5 l
Pyrex glass bubble column reactor was equipped with
a few inlets for bubbling the feed gas, sampling, venting,
and measuring temperature. Excess ozone was passed
into two gas absorption bottles containing 2% KI solu-
tion. All tubes from the ozone generator to the reactor
and the gas absorption bottles were made of Neoprene
and the fittings were made of Teflon. The experiments
were carried out in a thermostatic bath to keep the tem-
perature at the desired value ±0.5
C.
The mass transfer coefficient kL and the interfacial
area a of the bubble column reactor under different
operating conditions determined in our previous work
is summarized in Table 1 (Zhao et al., 2003).
2.3. Analytical methods
As ozonation intermediates do not have absorption
at a visible region that will be illustrated afterwards, a
TU1800PC UV–VIS spectrophotometer (Pgenral Ana-
lytical Instrument Co. Ltd., Beijing, China) was used
to determine the concentrations of Cationic Red
X-GRL by Beer-LambertÕs law using a molar extinction
coefficient of 4.57 · 106 M
1 m
1 at the maximum
absorbance wavelength of 530 nm.
The TRACE GC 2000 gas chromatograph coupled
with TRACE MS mass spectrometer in the electronic
impact mode was used to confirm the identity of the
degradation products. The capillary column used was
 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199 1191

9 9

11 12 off gas
2
6
sampling
4

oxygen
1 5 7 8
3 10

Fig. 2. Schematic of the experimental set-up. (1) Ozone generator; (2) Rotameter; (3) Temperature controller; (4) Electrical heater; (5)
Thermocouple; (6) Bubble column reactor; (7) Thermostatic bath; (8) Absorption bottle; (9) Three-way valve; (10) Porous plate; (11)
To input ozone gas detection; (11) To off gas detection.

Table 1 The input and off gas mass concentration of ozone was

The interfacial areas a, mass transfer coefficient kL and determined by iodometrically (Peyton and Glaze, 1988).
volumetric mass transfer coefficient kLa of bubble column
Then, the partial pressure of the ozone gas, Pozone, was
reactor
determined by the relationship of input and off gas mass
T,
C Qoxygen, kL · 105, a, kLa · 103, concentration of ozone.
l h
1 m s
1 m
1 s
1
15 100 8.19 110.2 9.02 2.4. Procedures
20 100 9.83 107.3 10.55
25 100 12.17 105.1 12.79 In order to investigate the influence of process varia-
30 100 13.52 100.5 13.59 bles to ozonation, the Cationic Red X-GRL solutions
35 100 14.96 96.6 14.45
were buffered by adding a certain amount of concen-
25 70 10.04 72.3 7.26
25 40 8.25 60.8 5.01
trated phosphate buffer to desired pH value and ionic
strength of 0.1 M. Samples were taken out from the reac-
tor periodically to analyze the concentration of Cationic
Red X-GRL.
an HP-5 (cross linked 5% phenyl methyl siloxane, For the preparation of GC and GC/MS samples, the
30.0 m · 320 lm · 0.25 lm). 1 ll of the solution was concentration of Cationic Red X-GRL solution was set
chromatographed under the following conditions: injec- to 500 mg l
1 in order to enhance the concentration of
tor temperature was 250
C; the initial column tempera- intermediates, and the pH value of solution was adjusted
ture was held constant at 40
C for 3 min, ramped at to three by H2SO4 to inhibit the formation of hydroxyl
10
C min
1 to 150
C and held constant for 1 min; then radical indirect attack intermediates. 200 ml samples
ramped further at 10
C min
1 to 250
C and held con- were taken out from the reactor at 0, 15, 30, 45, and
stant for 1 min. The carrier gas was helium and the 60 min, respectively. The samples were extracted four
pressure in the column head was 100 kPa. In the spec- times by 25 ml dichloromethane, under the pH condition
trometer, the source and quadruple temperature were of 3 and 11 adjusted by H2SO4 and NaOH, respectively.
kept at 250 and 100
C, respectively. The Agilent The extracted phase was purged to 1 ml using pure nitro-
6890N gas chromatograph was used to measure the peak gen gas at 25
C, and then adjusted to 5 ml by adding
areas of the intermediates using the same method as 1 ml 3-methoxy-1-thiophenol (1.006&) and dichloro-
mentioned for GC/MS. methane for GC analysis. GC analysis was conducted
The Metrohm 792 Basic IC ion chromatograph was on 1 ml samples taken at 0, 15, 30, 45, and 60 min time
used to measure the concentration of nitrate ion. 20 ll intervals.
of the sample, pre-filtrated using 0.45 lm hydrophilic
microfiltration membranes, was injected into the Metro-
sep A Supp 4 anion column for analysis. The mobile 3. Results and discussion
phase was a mixture of 1.8 mM Na2CO3, 1.7 mM NaH-
CO3, and 5% acetone. 3.1. Influence of process variables
The Shimadzu TOC analyzer was used to measure
TOC by directly injecting the aqueous solutions to char- Experiments of dye degradation were developed to
acterize the mineralization of Cationic Red X-GRL dur- modify the oxygen flow rate (Qoxygen), initial dye con-
ing ozonation. The pH value was measured by pHS-25 centration (Cdye,0), temperature, and pH. Fig. 3 shows
(Rex Analytical Instrument Co. Ltd., Shanghai, China). the conversion of dye at various intervals for a set of
1192 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199

1.0 1.0

(a) (b)
0.8 0.8
1-Cdye/Cdye,0

0.6 0.6

1-Cdye /Cdye,0
0.4 0.4

pH = 3.15
Q oxygen= 100 l h -1
pH = 5.75
0.2
pH = 8.14
0.2
Q oxygen= 70 l h -1
pH = 9.24 Q oxygen= 40 l h -1
0.0 0.0
0 1 2 3 4 5 6 0 1 2 3 4 5 6

1.0 1.0

(c) (d)
0.8 0.8

1-Cdye /Cdye,0
0.6 0.6
1-Cdye /Cdye,0

T = 15 0C 0.4
0.4
T = 20 0C
T = 25 0C Cdye,0 = 7.25 X 10-5 M
0.2
0.2
T = 30 0C Cdye,0 = 1.32 X 10-4 M
T = 35 0C Cdye,0 = 2.10 X 10-4 M
0.0 0.0
0 1 2 3 4 5 6 0 1 2 3 4 5 6
time, min time, min

Fig. 3. Effect of process variables on conversion of dye: (a) effect of pHs: Cdye,0 = 2.10 · 10
4 M; Qoxygen = 100 l h
1; T = 25
C;
Ionic strength = 0.1 M; Pozone = 1.75 kPa, (b) effect of oxygen flow rates: Cdye,0 = 1.3210 ·
4 M; pH = 8.14; T = 25
C; Ionic
strength = 0.1 M; Pozone = 1.75 kPa, (c) effect of temperatures: Cdye,0 = 1.32 · 10
4 M; pH = 5.75; Qoxygen = 100 l h
1; Ionic strength =
0.1 M; Pozone = 1.75 kPa, (d) effect of initial concentrations of dye: T = 25
C; pH = 8.14; Qoxygen = 100 l h
1; Ionic strength = 0.1 M;
Pozone = 1.75 kPa.

experiments carried out under different operating
conditions.
As the phosphate ions, used in phosphate buffer solu-
tions, are weak scavengers of free radicals (much weaker
than tert-butyl alcohol), the formation of radicals by
ozone decomposition, and hence the radical attack to
dye are not very strong (Peyton and Glaze, 1988). There-
fore, the conversion of dye is mainly attributed to ozone
direct attack, while radical attack is secondary.
The effect of pH is shown in Fig. 3a. As can be seen
from Fig. 3a, the increase in pH yields an increased con-
version of dye although the differences are almost negli-
gible at pH 8.14 and 9.24. This effect of pH could be
attributed to the followings: if pH is increased, the con-
centration of hydroxyl radical which can enhance the
indirect attack rate to dye is also increased; but at the
same time, the equilibrium concentration of ozone is de-
creased (Sotelo et al., 1989) which would lower the direct
attack rate of ozone to dye.
The oxygen flow rates affecting fluid dynamic condi-
tions have a positive effect on the dye degradation, as
shown in Fig. 3b. At 6 min, the conversions of dye are
93%, 87% and 67% for the oxygen flow rates of
100 l h
1, 70 l h
1 and 40 l h
1, respectively. As can be
seen from Table 1, at 25
C the volumetric mass transfer
coefficient kLa of bubble column reactor increases from
5.01 · 10
3 s
1 at the oxygen flow rate of 40 l h
1 to
7.26 · 10
3 s
1 at the oxygen flow rate of 70 l h
1, and
12.79 · 10
3 s
1 at the oxygen flow rate of 100 l h
1,
respectively. In addition, the ozone mass transfer rate
is proportion to the volumetric mass transfer coefficient
kLa of the bubble column reactor. Therefore, the conver-
sion of dye increases with the oxygen flow rate.
The effect of temperature on the conversion of dye is
complex as shown in Fig. 3c. At 6 min, the conversion of
dye is 88%, 86%, 85%, 87%, and 91% for reaction tem-
peratures of 15, 20, 25, 30, and 35
C, respectively. There
is a minimum conversion of dye at 25
C. It could be
attributed to double effect: if temperature increases,
the reaction rate constant and the volumetric mass
transfer coefficient kLa of bubble column reactor in-
creases, but the ozone equilibrium concentration (Sotelo
et al., 1989) decreases correspondingly.
Fig. 3d shows the evolution of the conversion of dye
with time for experiments at different initial dye concen-
tration. At a given time, an increase of initial dye concen-
tration leads to a decreased conversion of dye though the
ozonation rate increases. Thus, at 6 min, dye has com-
 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199 1193

pletely disappeared if the starting concentration is the carbonaceous sites within the product molecules
7.25 · 10
5 M, while 96% conversion is reached if the ini- (Tomiyasu et al., 1985; Gagnon et al., 1997; Kunz
tial dye concentration is 1.32 · 10
4 M. About 83% con- et al., 2001). Thus, the decreasing pH value of unbuffered
version is reached if the initial dye concentration is solution indicates the formation of weak organic acids
2.10 · 10
4 M. Initial dye abatement rates are, however, during ozonation. At 10 min, when the dye was almost
1.2 · 10
5, 2.1 · 10
5, and 2.9 · 10
5 M min
1 for initial thoroughly decolorized, only 5.7% of TOC was mineral-
dye concentrations of 7.25 · 10
5, 1.32 · 10
4, and ized, while the pH fell to 3.14 from 5.82 at 0 min and the
2.10 · 10
4 M, respectively. molar ratio of Cnitrate and Cdye,0 increased to 0.96. At
120 min, the pH was stable at 3.14, same as at 10 min,
3.2. Variation of pH, TOC, Cdye, Cnitrate , and Cnitrate/ but the molar ratio of Cnitrate and Cdye,0 increased slightly
Cdye,0 during the degradation to 0.99 from 0.96 at 10 min, while the TOC decreased
slowly to 34.9% from 5.7% at 10 min. During the exper-
In this part, 50 mg l
1 unbuffered Cationic Red X- iments, the pH value of the solution decreased with the
GRL solution was used, and the samples were taken reaction, which indicates the formation of organic acids.
out from the reactor periodically to analyze pH, TOC, The molecule of Cationic Red X-GRL contains six
NO
3 , and concentration of Cationic Red X-GRL. The nitrogen atoms (Fig. 1), three in the form of 1,4-di-
results of experiments about the variation of pH, methyl-4H-[1,2,4]triazo-1-ium group, two in the form
TOC, concentration of dye (Cdye), concentration of ni- of azo group, and one in the form of amino group. The
trate ion (Cnitrate) and Cnitrate/Cdye,0 during the degrada- molar ratio of Cnitrate and Cdye,0 increased to 0.99 at
tion are depicted in Fig. 4. 120 min indicating only one of the six nitrogen atoms of
Ozonation rarely completely mineralizes organic dye was transformed into nitrate ion. It is worth mention-
compounds to CO2 and H2O, but produces partial oxida- ing that the nitrite ion formed from the ozonation (Fig. 5)
tion products such as organic acids, ketones, aldehydes, was not detected by IC during the experiments, as the ni-
and alkenes where oxygen is introduced into many of trite ion is apt to be oxidized into nitrate ion by ozone.

3.3. Intermediates discerned by GC/MS

(a) 20
6 The purpose of GC/MS analysis is to identify some
intermediates produced from the degradation processes
15
of the Cationic Red X-GRL. The GC/MS chromato-
-1
TOC, mg l

5 grams of the blank and the sample are displayed in Fig.

pH
pH

10 6. The peak of dichloromethane is shown on the retention

TOC period from 3 to 4.48 min, whereas the peak of 3-meth-
4 oxy-1-thiophenol is on the retention time of 11.79 min
5
(internal standard). The electron ionization mass spectra
of the four intermediates are displayed in Fig. 7. Table 2
3 0 lists out the four intermediates discerned by GC/MS.
0 20 40 60 80 100 120
time, min It is reasonable to presume that the discerned inter-
mediates must be the relatively stable state of the inter-
0.15 1.0
mediates derived from ozonation processes because the
active intermediates cannot be present under the high
0.8
oxidation potential of ozone (2.07 V).
concentration, mM

0.10 It is interesting to note that no intermediates contain-

C nitrate /Cdye,0

0.6 ing –NH2 and/or –NO2 groups broken from N@N and/
or C@N groups through the cycloaddition of molecule
0.4 ozone were found by GC/MS.
0.05
Cnitrate According to the structural formula of intermediates
Cdye 0.2 discerned by GC/MS and depicted in Fig. 5, compounds
Cnitrate/Cdye,0 D2 and D3 are the constituents of chromophore (–R1 in
(b)
0.00 0.0 Fig. 5), and compound D4 is the constituent of com-
0 20 40 60 80 100 120
pound D2. Therefore, the formation sequence of inter-
time, min
mediates could be:
Fig. 4. The variation of the pH, TOC, Cdye, Cnitrate, and Cnitrate/ O3
Cationic Red X-GRL ! D1 þ D2 þ D3 ð1Þ
Cdye,0 during the ozonation: Qoxygen = 100 l h
1; T = 25
C;
Cdye,0 = 1.278 · 10
4 M; Pozone = 1.75 kPa. (a) pH and TOC, O3
(b) Cdye, Cnitrate, and Cnitrate/Cdye,0. D2 ! D4 ð2Þ
1194 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199

4
3 N O3
5 9 10
N 11 N
N 6 7 N 21 20
2 N+ N N N
1 N 12
13
19
8 O
16
15 14 17 18
N N (S 1) O O
R2
R1
Path 1 Path 2

* R1 N H
N b CH3 R1 N H
+ N N N O O R2
O O R2
(D1) N N H O H O O H
O (S3) O (S )
2 (S9 )
O3

Organic acids, 4 O
N 9 10
3 5 6
ketones, aldehyes, N 7 R1 N R2
O3 +
alkanes, and CO2 NO3- NO2- + 2N
N 11
N 8
1 (S10 ) OH OH (S )
11
(S4) 15 14

O3 O3
O
O N O O
N O O
O
N2 + O N N * N N
N N N N
(D3)
(S6) (S5 )

4N N
6 N
* O
5
3 O O3 O O
O O O
N N N + HCOOH + N2
O N N O H
2N N O O (D4 )
(D2) 1 (S7 ) H H (S8 )

Fig. 5. The probable degradation pathways of Cationic Red X-GRL during ozonation.

3.4. Variation of peak areas determined by GC cules such as organic acids, ketones, aldehydes, alkenes,
and carbon dioxide.
The GC peak areas of the four intermediates dis- Compound D2, a fragment of the chromophore,
cerned by GC/MS during the ozonation were measured, which peak of peak areas is at the reaction time of ca.
and the results were depicted in Fig. 8. 30 min, illustrated the cleavage of the chromophore. As
According to the color emission theory of dyes, the compound D2 can be further decomposed to compound
Cationic Red X-GRL is composed of the chromophore D4, its peak areas declined after the reaction time of ca.
(–R1 in Fig. 5) and the auxochrome (–CH2–R2 in Fig. 5). 30 min.
In general, the auxochromes, which are usually the elec- It is interesting to mention that the compounds D3
tron-donating groups, can enhance the electron densities and D4 are accumulative during the reaction because
of chromophores and can displace the absorption spec- they are not oxidized by ozone easily (Peyton and Glaze,
tra of dyes to the visible region. 1988; Kunz et al., 2001).
The benzaldehyde, which has the highest peak of
peak areas and appears first among the four intermedi-
ates at the reaction time of ca. 17 min, demonstrated that 3.5. Austin model 1 molecular orbital calculations
the auxochrome leaves from dye first shifting the
absorption spectra of the Cationic Red X-GRL out of With the aid of the Austin Model 1 Molecular Orbi-
the visible region. The declination of the peak areas after tal (AM1 MO) calculations based on quantum physics
the reaction time of ca. 17 min implies the continuous (MOPAC toolbox, Cambridge Soft Corp.), fully optim-
decomposition of the benzaldehyde to the smaller mole- ized molecular geometry for the Cationic Red X-GRL,
 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199 1195

RT: 2.38 - 22.10

11.79
100

95

90

85 (a)
80

75
Relative abundance

70

65
e
c 60
n
a
d
n 55
u
b 19.08
A 50
e
vti
al 45
e
R
40

35
18.16
30 11.91 20.64

25 3.11
14.54
11.40 11.97 19.24
20 17.90
4.48 6.08 7.66 8.13 8.37 16.74 17.14 21.23
15 5.57 9.59 10.16 12.26 14.23 15.51

10

0
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

time, min
RT: 2.75 - 22.84
19.08
100

95 7.43
90
D1 (b)
85

80

75
Relative abundance

D3
70 3.11
D4 D2
65
11.79
60
11.91
55 20.64

50
14.55
19.00
45
8.68 11.41 11.99 20.05
40 17.90 22.33
3.55 4.48 10.25
35 7.66
3.69
4.95 6.68 6.85 17.15
8.38 9.83 10.45 16.74 22.21
30 9.59 15.52
12.74 14.23 16.23
25

20

15

10

0
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

time, min
Fig. 6. The GC/MS spectra: Qoxygen = 100 l h
1; T = 25
C; pH = 3; Cdye,0 = 500 mg l
1; Pozone = 1.75 kPa. (a) Blank, (b) sample.

compound S4, and compound D2 were calculated as and N(7) atoms are approximately sp2 hybridization,
shown in Fig. 5. Selected bond distances, bond angles, whereas the N(12) atom is sp3 hybridization. The dihe-
dihedral angles, and partial charges associated with Cat- dral angle of 40.580
of the N(4)–C(5)–N(6)–N(7)
ionic Red X-GRL, compound S4, and compound D2 indicates the distortion of the plane of 1,4-dimethyl-
were given in Tables 3 and 4, respectively. 4H-[1,2,4]triazo-1-ium group with the plane of N(6)–
In Table 3, the bond angles of N(2)–N(1)–C(5), N(7) group due to steric hindrance of methyl.
N(1)–N(2)–C(3), C(3)–N(4)–C(5), C(5)–N(6)–N(7), and The bond angles of N(2)–N(1)–C(5), N(1)–N(2)–
N(6)–N(7)–C(8) of Cationic Red X-GRL departed far C(3), C(3)–N(4)–C(5), C(5)–N(6)–N(7), and N(6)–
from 120
illustrate that the N(1), N(2), N(4), N(6), N(7)–C(8) of the compound S4, and the bond angles
1196 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199

Fig. 7. Electron ionization mass spectra of intermediates: (a) retention time of 4.95 min, compound D4, (b) retention time of 7.43 min,
compound D1, (c) retention time of 8.68 min, compound D3 and (d) retention time of 10.25 min, compound D2.

Table 2
The ozonation intermediates determined by GC/MS
Retention time, min m/z Molecular formula Chemical name Structural
formula
4.95 87 C3H5NO2 N-formyl-N-methyl-formamide D4 in Fig. 5
7.43 106 C7H6O benzaldehyde D1 in Fig. 5
8.68 124 C7H8O2 3-methyl-hexa-2,4-dienedial D3 in Fig. 5
10.25 113 C4H7N3O 2,4-dimethyl-2,4-dihydro-[1,2,4] triazol-3-one D2 in Fig. 5

of N(2)–N(1)–C(5), N(1)–N(2)–C(3), and C(3)–N(4)–
C(5) of the compound D2 departed far from 120
illus-
trate that the N(1), N(2), N(4), N(6), and N(7) atoms
of the compound S4, and N(1), N(2), and N(4) atoms
of the compound D2 are approximately sp2 hybridiza-
tion.
The dihedral angles of N(4)–C(5)–N(6)–N(7), 39.055

in compound S4 and 40.580
in Cationic Red X-GRL,
illustrate that the coplanarity of 1,4-dimethyl-4H-
[1,2,4]triazo-1-ium group with the N(6)–N(7) group in
compound S4 is almost the same as in Cationic Red
X-GRL. However, the bond lengths of N(6)–N(7),
1.248 Å in compound S4 and 1.362 Å in Cationic Red
X-GRL, illustrate that the binding of N(6) and N(7)
atoms in compound S4 is much stronger than in Cationic
Red X-GRL. Therefore, due to the lack of electron-
donation groups such as the auxochrome of the Cationic
Red X-GRL, the electron of compound S4 delocalizes
with more difficulty than the electron of Cationic Red
X-GRL. Hence, the absorption spectra of compound
S4 were not located in the visible region.
The results in Table 4 indicate that the positive
charge on Cationic Red X-GRL resides primarily on
the N(1) and N(4) atoms, and that N(2) and N(12)
atoms have more electron rich character. Although the
electron densities of the N(2) atom are higher than the
N(12) atom, under the strong positive charge effect of
N(1) atom, the cycloaddition of ozone takes place in
N(12)–C(13) preferably. Similarly, for compound S4,
the positive charge resides primarily on the N(1) and
N(4) atoms, and the N(2) and N(6) atoms have more
electron rich character. Under the strong positive charge
effect of N(1) atom, the cycloaddition of ozone takes
place in C(5)–N(6) preferably. For compound D2, the
 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199 1197

170 The AM1 MO calculations of the partial charges in

Compound D 4 Table 4 show that the cycloaddition of ozone takes place
Compound D 1 in the N(12)–C(13) position instead of the N(6)–N(7)
160 Compound D 3
position, forming the state S1.
Compound D 2
peak areas, pA.s

There are two probable cleavage paths of O–O band

in state S1 as illustrated in Fig. 5. Because GC and GC/
80 MS detected compound D1, but not compound S11, Path
60
1 is more predominant than Path 2. The same result in
ozonation of curcumin was found by Grosjean et al.
40 (1988).
20
Under the influence of the nitrogen atom and the ac-
tive group of the azo group in state S3, the site b is acti-
0 vated. The state S3 is then transformed into compounds
0 10 20 30 40 50 60
S4 by rearrangement of methyl while releasing the nitrite
time, min
ion, which can be further oxidized into nitrate ion by
Fig. 8. The peak areas of intermediates determined by GC: ozonation. As the compound S4 is a cationic substance,
Qoxygen = 100 l h
1; T = 25
C; pH = 3; Cdye,0 = 500 mg l
1; which cannot be extracted by dichloromethane solution
Pozone = 1.75 kPa. during the preparation of samples, the GC/MS and GC
cannot detect it.
positive charge resides primarily on the N(1), N(4), and For compound S4, the cycloaddition of ozone takes
C(5) atoms, and N(2) and C(3) atoms have more elec- place in the C(5)–N(6) site instead of the N(6)–N(7) site,
tron rich character, so the cycloaddition of ozone takes as the partial charges indicated in Table 4, and then
place in N(2)–C(3) preferably. forms the state S5.
The 1,4-dimethyl-4H-[1,2,4]triazo-1-ium group of the
3.6. Degradation mechanism state S6 captures one oxygen atom and one unpaired
electron of the azo group, forming the compound D2,
Summarizing the above results, the degradation D3, and N2 by rearrangement. Because of the steric hin-
pathways of dye during ozonation can be presumed to drance of methyl and the presence of the unsatu-
be that as depicted in Fig. 5. rated N(2)–C(3) double bond, the compound D2 is not

Table 3
Bond lengths (Å), bond angles and dihedral angles (
) for AM1 optimized Cationic Red X-GRL, compound S4, and compound D2
structures
Parameter Cationic Red X-GRL Compound S4 Compound D2
N(1)–N(2) 1.469 1.467 1.373
N(2)–C(3) 1.260 1.260 1.283
C(3)–N(4) 1.462 1.462 1.385
N(4)–C(5) 1.462 1.462 1.383
N(1)–C(5) 1.300 1.300 1.369
C(5)–N(6) 1.456 1.456
N(6)–N(7) 1.362 1.248
N(7)–C(8) 1.404 1.456
N(2)–N(1)–C(5) 108.9 108.4 109.7
N(1)–N(2)–C(3) 107.0 107.2 109.1
N(2)–C(3)–N(4) 107.6 107.2 107.7
C(3)–N(4)–C(5) 112.8 113.1 110.2
N(1)–C(5)–N(4) 103.7 104.0 103.4
C(5)–N(6)–N(7) 107.5 107.5
N(6)–N(7)–C(8) 107.5 107.5
C(11)–N(12)–C(13) 117.8
N(1)–N(2)–C(3)–N(4) 0 0 0
N(4)–C(5)–N(1)–N(2)
0.316
0.316
0.316
N(4)–C(5)–N(6)–N(7) 40.580 39.055
C(5)–N(6)–N(7)–C(8) 179.553 178.210
N(6)–N(7)–C(8)–C(9)
12.432
10.759
N(12)–C(13)–C(16)–C(17) 97.911
1198 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199

Table 4
Selected partial charges of Cationic Red X-GRL, compound S4, and compound D2 computed using the AM1 methods
Atom Cationic Red X-GRL Compound S4 Compound D2
N(1) 0.659 0.652 0.520
N(2)
0.297
0.207
0.358
C(3) 0.057 0.056
0.002
N(4) 0.471 0.472 0.408
C(5) 0.193 0.033 0.265
N(6)
0.062
0.222
N(7)
0.032 0.323
C(8) 0.017 0.022
C(9)
0.057
0.028
C(10)
0.064 0.084
C(11) 0.089
0.079
N(12)
0.189
C(13)
0.061
C(14)
0.056
0.035
C(15)
0.052
0.096
C(16) 0.090
C(17)
0.060
C(18)
0.026
C(19)
0.050
C(20)
0.026
C(21)
0.062

polymerized whereas can be oxidized to state S7 by
ozone cycloaddition in N(2)–C(3) sites, as the partial
charges indicated in Table 4. The state S7 can be further
broken into compound D4, HCOOH, and N2.
The oxidation of the intermediates into organic acids
such as HCOOH can lower the pH value of the solution.
Fig. 5 shows that the overall degradation processes
are not attributed to hydroxyl radial reactions because
the reactions are in the acidic solution where the indirect
attacks are negligible (Peyton and Glaze, 1988).
Among the intermediates derived from the degrada-
tion, only compounds D1 and D3 can be decomposed
into carbon dioxide partially and slowly, whereas the
TOC falls slightly even if the dye has been completed
decolorized (Fig. 4a).
During the degradation, among the six nitrogen
atoms of Cationic Red X-GRL, one is transferred into
a nitrate ion, one is converted into an amine compound,
and the remaining four are transformed into two mole-
cules of nitrogen.
4. Conclusion
Ozonation of azo dye Cationic Red X-GRL was
investigated by varying the oxygen flow rate, initial
dye concentration, temperature, and pH. The increase
of pH yields an increased conversion of dye as the indi-
rect attack is enhanced although the equilibrium concen-
tration of ozone is decreased. As the volumetric mass
transfer coefficient of bubble column reactor increases
with the oxygen flow rate, hence the mass transfer rate
of ozone increases; consequently, the conversion of dye
is enhanced. There is a minimum conversion of dye at
25
C as the reaction rate constant and the volumetric
mass transfer coefficient of the bubble column reactor
increases while the ozone equilibrium concentration de-
creases with the temperature. An increase of initial dye
concentration leads to a decreased conversion of dye
though the ozonation rate increases.
Ozonation rarely completely mineralizes organic
compounds to CO2 and H2O, but produces partial oxi-
dation products such as organic acids, ketones, alde-
hydes, and alkanes where oxygen is introduced into
many of the carbonaceous sites within the product mol-
ecules. The intermediates of weak organic acids lower
the pH value of the solution.
The ozonation mechanism of dye in aqueous solu-
tions was deduced with the aid of the molecular geome-
try and electronic structures analysis of the dye. The
N(12)–C(13) site instead of the N(6)–N(7) site in Cati-
onic Red X-GRL is the principal site for ozone cycload-
dition. During the degradation, among the six nitrogen
atoms of Cationic Red X-GRL, one is transferred into
a nitrate ion, one is converted into an amine compound,
and the remaining four are transformed into two mole-
cules of nitrogen.
Acknowledgment
The authors are grateful to the Returnee Foundation
of China Ministry of Education (project no. 2002-247)
for the financial support.
 W. Zhao et al. / Chemosphere 57 (2004) 1189–1199
References
Alaton, I.A., Balcioglu, I.A., Bahnemann, D.W., 2002.
Advanced oxidation of a reactive dyebath effluent: compar-
ison of O3, H2O2/UV-C and TiO2/UV-A processes. Water
Res. 36, 1143–1154.
Banat, I.M., Nigam, P., Singh, D., Marchant, R., 1996.
Microbial decolorization of textile-dye-containing effluents:
a review. Bioresource Technol. 58, 217–227.
Beltran-Heredia, J., Torregrosa, J., Dominguez, J.R., Peres,
J.A., 2001. Kinetics of the reaction between ozone and
phenolic acids present in argo-industrial wastewaters. Water
Res. 35, 1077–1085.
Chu, W., Ching, M.H., 2003. Modeling the ozonation of 2,4-
dichlorophoxyacetic acid through a kinetic approach. Water
Res. 37, 39–46.
Chu, W., Ma, C.W., 1998. Reaction kinetics of UV-decolou-
rization for dye materials. Chemosphere 37, 961–974.
Gagnon, G.A., Booth, S.D.J., Peldszus, S., Mutti, D., Smith, F.,
Huck, P.M., 1997. Carboxylic acids: formation and removal
in full-scale plants. J. Am. Water Works Assoc. 89, 88–97.
Grosjean, D., Whitemore, P.M., Moor, C.D., Cass, G.R., 1988.
Ozone fading of organic colorants: products and mechanism
of the reaction of ozone and curcumin. Environ. Sci.
Technol. 22, 1357–1361.
Koch, M., Yediler, A., Lienert, D., Insel, G., Kettrup, A., 2002.
Ozonation of hydrolyzed azo dye reactive yellow 84(CI).
Chemosphere 46, 109–113.
Kunz, A., Reginatto, V., Durán, N., 2001. Combined treatment
of textile effluent using the sequence phanerochaete chrysos-
porium–ozone. Chemosphere 44, 281–287.
1199
Ledakowicz, S., Soplecka, M., Zylla, R., 2001. Biodegradation,
decolorization and detoxification of textile wastewater
enhanced by advanced oxidation processes. J. Biotechnol.
89, 175–184.
Liakou, S., Pavlou, S., Lyberatos, G., 1997. Ozonation of azo
dyes. Water Sci. Technol. 35 (4), 279–286.
Nam, S., Renganathan, V., Tratnyek, P.G., 2001. Substituent
effects on azo dye oxidation by the FeIII–EDTA–H2O2
system. Chemosphere 45, 59–65.
Peyton, G.R., Glaze, W.H., 1988. Destruction of pollutants in
water with ozone in combination with ultraviolet radiation.
3. Photolysis of aqueous ozone. Environ. Sci. Technol. 22,
183–189.
Shu, H.Y., Huang, C.R., 1995. Degradation of combination
of commercial azo dyes in water using ozonation and
UV enhanced ozonation process. Chemosphere 31, 3813–
3825.
Sotelo, J.L., Beltran, F.J., Benitez, F.J., Beltran-Heredia, J.,
1989. HenryÕs law constant for the ozone–water system.
Water Res. 23, 1239–1246.
Talarposhti, A.M., Donnelly, T., Anderson, G.K., 2001.
Colour removal from a simulated dye wastewater using a
two-phase anaerobic packed bed reactor. Water Res. 35,
425–432.
Tomiyasu, H., Fukutomi, H., Gordon, G., 1985. Kinetics and
mechanism of ozone decomposition in basic aqueous
solutions. Inorg. Chem. 24, 2962–2968.
Zhao, W., Shi, H., Wang, D., 2003. Kinetics of the reaction
between ozone and Cationic Red X-GRL. Chin. J. Chem.
Eng. 11, 388–394.