Professional Documents
Culture Documents
www.elsevier.com/locate/chemosphere
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.
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
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.
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 · 104 M; Qoxygen = 100 l h1; 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 · 104 M; pH = 5.75; Qoxygen = 100 l h1; 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 h1; Ionic strength = 0.1 M;
Pozone = 1.75 kPa.
experiments carried out under different operating seen from Table 1, at 25 C the volumetric mass transfer
conditions. coefficient kLa of bubble column reactor increases from
As the phosphate ions, used in phosphate buffer solu- 5.01 · 103 s1 at the oxygen flow rate of 40 l h1 to
tions, are weak scavengers of free radicals (much weaker 7.26 · 103 s1 at the oxygen flow rate of 70 l h1, and
than tert-butyl alcohol), the formation of radicals by 12.79 · 103 s1 at the oxygen flow rate of 100 l h1,
ozone decomposition, and hence the radical attack to respectively. In addition, the ozone mass transfer rate
dye are not very strong (Peyton and Glaze, 1988). There- is proportion to the volumetric mass transfer coefficient
fore, the conversion of dye is mainly attributed to ozone kLa of the bubble column reactor. Therefore, the conver-
direct attack, while radical attack is secondary. sion of dye increases with the oxygen flow rate.
The effect of pH is shown in Fig. 3a. As can be seen The effect of temperature on the conversion of dye is
from Fig. 3a, the increase in pH yields an increased con- complex as shown in Fig. 3c. At 6 min, the conversion of
version of dye although the differences are almost negli- dye is 88%, 86%, 85%, 87%, and 91% for reaction tem-
gible at pH 8.14 and 9.24. This effect of pH could be peratures of 15, 20, 25, 30, and 35 C, respectively. There
attributed to the followings: if pH is increased, the con- is a minimum conversion of dye at 25 C. It could be
centration of hydroxyl radical which can enhance the attributed to double effect: if temperature increases,
indirect attack rate to dye is also increased; but at the the reaction rate constant and the volumetric mass
same time, the equilibrium concentration of ozone is de- transfer coefficient kLa of bubble column reactor in-
creased (Sotelo et al., 1989) which would lower the direct creases, but the ozone equilibrium concentration (Sotelo
attack rate of ozone to dye. et al., 1989) decreases correspondingly.
The oxygen flow rates affecting fluid dynamic condi- Fig. 3d shows the evolution of the conversion of dye
tions have a positive effect on the dye degradation, as with time for experiments at different initial dye concen-
shown in Fig. 3b. At 6 min, the conversions of dye are tration. At a given time, an increase of initial dye concen-
93%, 87% and 67% for the oxygen flow rates of tration leads to a decreased conversion of dye though the
100 l h1, 70 l h1 and 40 l h1, respectively. As can be 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 · 105 M, while 96% conversion is reached if the ini- (Tomiyasu et al., 1985; Gagnon et al., 1997; Kunz
tial dye concentration is 1.32 · 104 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 · 104 M. Initial dye abatement rates are, however, during ozonation. At 10 min, when the dye was almost
1.2 · 105, 2.1 · 105, and 2.9 · 105 M min1 for initial thoroughly decolorized, only 5.7% of TOC was mineral-
dye concentrations of 7.25 · 105, 1.32 · 104, and ized, while the pH fell to 3.14 from 5.82 at 0 min and the
2.10 · 104 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 l1 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.
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 h1; T = 25 C;
Cdye,0 = 1.278 · 104 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
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 h1; T = 25 C; pH = 3; Cdye,0 = 500 mg l1; 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)– Red X-GRL, the electron of compound S4 delocalizes
C(5) of the compound D2 departed far from 120 illus- with more difficulty than the electron of Cationic Red
trate that the N(1), N(2), N(4), N(6), and N(7) atoms X-GRL. Hence, the absorption spectra of compound
of the compound S4, and N(1), N(2), and N(4) atoms S4 were not located in the visible region.
of the compound D2 are approximately sp2 hybridiza- The results in Table 4 indicate that the positive
tion. charge on Cationic Red X-GRL resides primarily on
The dihedral angles of N(4)–C(5)–N(6)–N(7), 39.055 the N(1) and N(4) atoms, and that N(2) and N(12)
in compound S4 and 40.580 in Cationic Red X-GRL, atoms have more electron rich character. Although the
illustrate that the coplanarity of 1,4-dimethyl-4H- electron densities of the N(2) atom are higher than the
[1,2,4]triazo-1-ium group with the N(6)–N(7) group in N(12) atom, under the strong positive charge effect of
compound S4 is almost the same as in Cationic Red N(1) atom, the cycloaddition of ozone takes place in
X-GRL. However, the bond lengths of N(6)–N(7), N(12)–C(13) preferably. Similarly, for compound S4,
1.248 Å in compound S4 and 1.362 Å in Cationic Red the positive charge resides primarily on the N(1) and
X-GRL, illustrate that the binding of N(6) and N(7) N(4) atoms, and the N(2) and N(6) atoms have more
atoms in compound S4 is much stronger than in Cationic electron rich character. Under the strong positive charge
Red X-GRL. Therefore, due to the lack of electron- effect of N(1) atom, the cycloaddition of ozone takes
donation groups such as the auxochrome of the Cationic place in C(5)–N(6) preferably. For compound D2, the
W. Zhao et al. / Chemosphere 57 (2004) 1189–1199 1197
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 of ozone increases; consequently, the conversion of dye
ozone cycloaddition in N(2)–C(3) sites, as the partial is enhanced. There is a minimum conversion of dye at
charges indicated in Table 4. The state S7 can be further 25 C as the reaction rate constant and the volumetric
broken into compound D4, HCOOH, and N2. mass transfer coefficient of the bubble column reactor
The oxidation of the intermediates into organic acids increases while the ozone equilibrium concentration de-
such as HCOOH can lower the pH value of the solution. creases with the temperature. An increase of initial dye
Fig. 5 shows that the overall degradation processes concentration leads to a decreased conversion of dye
are not attributed to hydroxyl radial reactions because though the ozonation rate increases.
the reactions are in the acidic solution where the indirect Ozonation rarely completely mineralizes organic
attacks are negligible (Peyton and Glaze, 1988). compounds to CO2 and H2O, but produces partial oxi-
Among the intermediates derived from the degrada- dation products such as organic acids, ketones, alde-
tion, only compounds D1 and D3 can be decomposed hydes, and alkanes where oxygen is introduced into
into carbon dioxide partially and slowly, whereas the many of the carbonaceous sites within the product mol-
TOC falls slightly even if the dye has been completed ecules. The intermediates of weak organic acids lower
decolorized (Fig. 4a). the pH value of the solution.
During the degradation, among the six nitrogen The ozonation mechanism of dye in aqueous solu-
atoms of Cationic Red X-GRL, one is transferred into tions was deduced with the aid of the molecular geome-
a nitrate ion, one is converted into an amine compound, try and electronic structures analysis of the dye. The
and the remaining four are transformed into two mole- N(12)–C(13) site instead of the N(6)–N(7) site in Cati-
cules of nitrogen. 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
4. Conclusion a nitrate ion, one is converted into an amine compound,
and the remaining four are transformed into two mole-
Ozonation of azo dye Cationic Red X-GRL was cules of nitrogen.
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- Acknowledgment
rect attack is enhanced although the equilibrium concen-
tration of ozone is decreased. As the volumetric mass The authors are grateful to the Returnee Foundation
transfer coefficient of bubble column reactor increases of China Ministry of Education (project no. 2002-247)
with the oxygen flow rate, hence the mass transfer rate for the financial support.
W. Zhao et al. / Chemosphere 57 (2004) 1189–1199 1199