The formation and control of disinfection by-products using chlorine dioxide

Article abstract of DOI:10.1016/S0045-6535(00)00010-2

In this study, chlorine dioxide (ClO₂) was used as an alternative disinfectant with vanillic acid, p-hydroxybenzoic acid, and humic acid as the organic precursors in a natural aquatic environment. The primary disinfection by-products (DBPs) formed were trihalomethanes (THMs) and haloacetic acids (HAAs). Under neutral conditions (pH = 7) for vanillic acid, more total haloacetic acids (THAAs) than total trihalomethanes (TTHMs) were found, with a substantial increase during the later stages of the reaction. In the case of p-hydroxybenzoic acid, the amount of THAAs produced was minimal. Raising the concentration of ClO₂ was not favorable for the control of THAAs in low concentrations of vanillic acid. ClO₂ could reduce the total amount of TTHMs and THAAs for higher concentration of vanillic acid. It was found that the humic acid treatment dosage was not significant. Under alkaline conditions (pH = 9), the control of TTHMs and THAAs for the treatment of vanillic acid was better and more economical, however, an appreciable amount of inorganic by-products were observed. Under the same alkaline condition, the control of THAA for the treatment of p-hydroxybenzoic acid was not beneficial and for the treatment of humic acid was not significant. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0045-6535(00)00010-2

Chemosphere 41 (2000) 1181±1186
The formation and control of disinfection by-products using
chlorine dioxide
a,
a
a
a
Chen-Yu Chang *, Yung-Hsu Hsieh , I-Chen Shih , Shen-Sheng Hsu ,
b
Kuo-Hua Wang
a
Department of Environmental Engineering, National Chung-Hsing University, 250 Kuo-Kuang Road, Taichung, Taiwan, ROC
Department of Environmental Engineering and Health, Yuanpei Technical College, 306 Yuanpei Street, Hsinchu, Taiwan, ROC
b
Received 13 October 1999; accepted 9 December 1999
Abstract
In this study, chlorine dioxide (ClO2) was used as an alternative disinfectant with vanillic acid, p-hydroxybenzoic
acid, and humic acid as the organic precursors in a natural aquatic environment. The primary disinfection by-products
(DBPs) formed were trihalomethanes (THMs) and haloacetic acids (HAAs). Under neutral conditions (pH ˆ 7) for
vanillic acid, more total haloacetic acids (THAAs) than total trihalomethanes (TTHMs) were found, with a substantial
increase during the later stages of the reaction. In the case of p-hydroxybenzoic acid, the amount of THAAs produced
was minimal. Raising the concentration of ClO2 was not favorable for the control of THAAs in low concentrations of
vanillic acid. ClO2 could reduce the total amount of TTHMs and THAAs for higher concentration of vanillic acid. It
was found that the humic acid treatment dosage was not signi®cant. Under alkaline conditions (pH ˆ 9), the control of
TTHMs and THAAs for the treatment of vanillic acid was better and more economical, however, an appreciable
amount of inorganic by-products were observed. Under the same alkaline condition, the control of THAA for the
treatment of p-hydroxybenzoic acid was not bene®cial and for the treatment of humic acid was not signi®cant. Ó 2000
Elsevier Science Ltd. All rights reserved.
Keywords: Disinfection by-products; Chlorine dioxide; Trihalomethanes; Halogenated acetic acids; Disinfectant
1. Introduction
fection has been proven leading to the formation of
disinfection by-products (DBPs) (Bellar et al., 1974;
Rook, 1974; Oliver, 1983; Koch and Krasner, 1989;
Peters et al., 1990). These compounds are produced by
chemical reactions between organic compounds and the
applied chlorine. These DBPs are potentially carcino-
genic and teratologic substances (National Cancer In-
stitute, 1976; Bull and Kop¯er, 1991; Craun et al., 1994).
Their presence in drinking water has given rise to health
concerns, which have translated into regulations in
several countries. This problem triggered the search for
an alternative disinfectant (AWWA, 1986; Myer, 1990;
White, 1992) to prevent or reduce the possibility of
DBPs formation in disinfected water (Krasner et al.,
1989; Arora et al., 1990). Chlorine dioxide (ClO2) was
Recently municipal drinking water has become
highly contaminated by human activity, requiring in-
creased chlorination, which leads to the formation of
harmful chlorine by-products, such as trihalomethanes
(THMs) and haloacetic acids (HAAs). Since the detec-
tion of trihalomethanes in the 1970s by the US Envi-
ronmental Protection Agency in the New Orleans La.,
USA, waterworks, the use of chlorine in water disin-
* Corresponding author. Tel.: +886-1-285-5367; fax: +886-2-
287-4469.
E-mail address: cychang@enve.ev.edu.tw (C.-Y. Chang).
0045-6535/00/$ - see front matter Ó 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 5 - 6 5 3 5 ( 0 0 ) 0 0 0 1 0 - 2

1182
C.-Y. Chang et al. / Chemosphere 41 (2000) 1181±1186
Table 1
The parameters in DBPs experiment
investigated, as one of the promising disinfectants as a
substitute for chlorine (Lykins and Griese, 1986; Narkis,
1995) for the following reasons. It is a strong disinfec-
tant that is e€ective over a wide pH range (White, 1992).
ClO2 can eliminate bad odors (White, 1992; Edwards
and Amirtharajah, 1993) iron and manganese in un-
treated water (Aieta and Berg, 1986; White, 1992). It is
very successful in killing bacteria and especially most
ecient in deactivating viruses. A smaller dosage and
less reaction time are required for ClO2 to produce the
same disinfection e€ects as chlorine. The ClO2 manu-
facturing system is also easy to install, operate and
maintain. Because of these characteristics, ClO2 was
investigated as one of the promising substitute disin-
fectants for chlorine. The National Research Council
(NRC, 1980) reported on the use of ClO2 as an alter-
native disinfectant.
Parameter
Unit
Organic acids
pH
5, 10, 20
7 Æ 0:5; 9 Æ 0:5
20
mg-DOC/L
unit
Temperature
ClO2
Reaction time
°C
mg-ClO2/L
15, 30
0±168
h
Analysis
DOC, ClO2, ClO2 , ClO3 , Cl ,
THMs, UV254
laboratory in a cooler and stored in a cold room until
they were analyzed. The analytical methods followed the
EPA 524.2. Samples were adjusted to pH 4.5 in the ®eld
and extracted with normal pentane, containing dibro-
momethane and 1,2-dibromoproane as the internal
standards. The THMs were analyzed using a HP5890II
plus gas chromatograph equipped with an electron
capture detector (GC±ECD), a one-column injector and
a J&W DB-5 capillary column. The HAAs water sam-
ples were prepared by adding 150 mL NH4Cl per 100 ml
of the sample with the pH adjusted to below 0.5. Sam-
ples were extracted with methy-tert-butylether (MTBE)
esterized using diazomethane and analyzed using GC±
ECD based on EPA standard 552.2 (see Table 1).
In this investigation, ClO2 was used to treat a simu-
lated water system containing humic acid, vanillic acid
and p-hydroxybenzoic acid, which have been detected in
natural bodies of water (Yamada and Somiya, 1979) and
were found as organic precursors in the chlorination
disinfection process (Rice, 1980). Our purpose was to
study the disinfection e€ects of ClO2 in respect to the
control of disinfection by-product formation for its po-
tential use in municipal water supplies. This provides the
theoretical base for using ClO2 as a disinfectant in wa-
terworks in Taiwan.
2.2. Quality control
All samples were collected in duplicate with control
samples included for all-target analytes. All DBPs
methods incorporated surrogate internal standards and
quanti®cation was based on response factors established
by multi-level calibration with forti®ed samples analyzed
under identical conditions. For the THMs, raw water
samples (matrix spikes; n ˆ 11) were analyzed at a for-
ti®cation level of 5 lg/L (chloroform ˆ 25 lg/L). The
overall recovery was 98:3 Æ 2:7%. The HAAs method
precision was estimated at ‹20%. The mean recovery of
HAAs was typically >92% as estimated from the re-
covery of the added MBBA internal standard. DBPs
identi®ed by GC±ECD were con®rmed by GC±MS.
2. Materials and methods
2.1. Procedure
In this DBP study, 5 ml of phosphate bu€er solution,
the proper amount of ClO2, and organic precursors
(humic acid, vanillic acid, p-hydroxybenzoic acid) under
controlled parameters, were added to each BOD bottle
reactor with de-ionized water and kept at 20°C in an
incubator. This mixture was sampled at di€erent time
intervals over a total of 7 days. Three bottles were col-
lected for analysis. The samples were immediately ana-
lyzed for ClO2 residual concentration. A 10 mL portion
from the 0.2 lm ®ltrate was collected to analyze the
inorganic DBPs (Cl , ClO2 , Br and ClO3 ) using ion
chromatography (DIONEX, series 4500, column AS-
12A, 4 mm (10±12), P/N 46034). In the remaining
samples, 1 g of Na2S2O3 was added to terminate the
reaction. One portion of the sample was used to deter-
mine the amount of dissolved organic carbon (DOC)
and UV absorption analysis.
3. Results and discussion
3.1. E€ect of organic acid species
3.1.1. Low dosage of ClO2 for treating low DOC in
neutral solution
The THMs sample was prepared in a 40 ml brown
glass bottle (with a Te¯on ring and screw-on cap) with
100 mg of NH4Cl as the preservative. Water was then
added to the above solution to ®ll the bottle. Samples
were capped with Te¯on-lined seals, returned to the
By comparing the unit consumption of ClO2 by
DOC, TTHMs and THAAs in Table 2, the amount of
decomposition products from the large molecular, hu-
mic acid, was much greater than that from vanillic acid
and p-hydroxybenzoic acid. A possible reason is that the

C.-Y. Chang et al. / Chemosphere 41 (2000) 1181±1186
1183
humic acidsÕ various functional groups, such as hydroxyl
group (R±OH), carbonyl (±CO), carboxyl (±COOH),
phenolic and quinoid groups react easily with ClO2.
sumption of ClO2 per unit of DOC in humic acid was
much greater in the case of vanillic acid, however the
amounts of TTHMs and THAAs were much less than
that in vanillic acid. This indicates that more ClO2 is
required to break down its large structure. TTHMs re-
formed after 16 h of reaction in vanillic acid, indicating
that TTHMs can be controlled through reaction time.
3.1.2. Low dosage ClO2 for treating of high DOC in
neutral solution
In Table 3, the unit consumption of ClO2 for DOC
was greater in the case of p-hydroxybenzoic acid than
with vanillic acid. The formation of TTHMs and
THAAs was reversed, indicating an increase for vanillic
acid and the decomposition products from p-hydroxy-
benzoic acid reacting easily with ClO2. The unit con-
3.1.3. High dosage of ClO2 for treating high DOC in
neutral solution
In Table 4, The consumption of ClO2 was greatest
for humic acid and the removal ratio of DOC was the
Table 2
Treatment of organic acids with ClO2 (DOC ˆ 5.0 mg/L, ClO2 ˆ 15.0 mg/L, pH ˆ 7)
Vanillic acid
p-Hydroxy benzoic acid
Humic acid
ClO2/DOCa
3.9
2.0
4.9
2.6
4.7
3.1
THM, lg/L
TTHMs/DOCb
THAAs, lg/L
THAAs/DOCc
0.91
2.8
1.95
4.4
1.47
16.0
7.53
l.28
3.31
Major productsd
CHCl3 and TCAA
a ClO2 consumed/DOC removal.
b Final amount of TTHMs/DOC removal.
c Final amount of THAAs/DOC removal.
d Major products of TTHMs and THAAs.
Table 3
Treatment of organic acids with ClO2 (DOC ˆ 10.0 mg/L, ClO2 ˆ 15.0 mg/L, pH ˆ 7)
Vanillic acid
p-Hydroxybenzoic acid
Humic acid
ClO2/DOCa
1.60
4.8
2.05
3.7
1.89
2.8
TTHMs, lg/L
TTHMs/DOCb
THAAs, lg/L
THAAs/DOCc
0.76
69.2
11.06
1.00
4.1
0.62
27.1
6.11
1.10
Major productsd
CHCl3 and TCAA
a ClO2 consumed/DOC removal.
b Final amount of TTHMs/DOC removal.
c Final amount of THAAs/DOC removal.
d Major products of TTHMs and THAAs.
Table 4
Treatment of organic acids with ClO2 (DOC ˆ 10.0 mg/L ClO2 ˆ 30.0 mg/L, pH ˆ 7)
Vanillic acid
p-Hydroxybenzoic acid
Humic acid
ClO2/DOCa
1.74
4.4
2.22
3.6
7.84
3.9
TTHMs, lg/L
TTHMs/DOCb
THAAs, lg/L
THAAs/DOCc
0.65
48.0
7.03
0.51
2.3
1.79
37.1
17.11
0.33
Major productsd
CHCl3 and DCAA
a ClO2 consumed/DOC removal.
b Final amount of TTHMs/DOC removal.
c Final amount of THAAs/DOC removal.
d Major products of TTHMs and THAAs.

1184
C.-Y. Chang et al. / Chemosphere 41 (2000) 1181±1186
least. This suggested that humic acid has a larger
structure than the others acids and it cannot be de-
stroyed easily. The production of THAAs was greater
in the case of p-hydroxybenzoic acid. This could be
attributed to the reaction of its functional groups on
the structure or its middle products with ClO2. By
considering the structures of vanillic acid and p-hy-
droxybenzoic acid, the presence of activating group,
o-methoxy group (±OCH3) in vanillic acid renders it to
be more active in reaction with ClO2 than p-hydroxy-
benzoic acid under this condition. In addition, the low
concentration of TTHMs and THAAs from p-hy-
droxybenzoic acid favors the control of by-product
formation.
3.2. In¯uence of ClO2 dosages
3.2.1. Various dosages of ClO2 in treating low DOC in
neutral solution
The formation of THAAs from vanillic acid was
faster and with greater amounts under a high dosage of
ClO2. This shows unfavorable control of THAAs for-
mation using vanillic acid. In the case of p-hydroxy-
benzoic acid, there was a high residual concentration of
ClO2. Increasing the ClO2 dosage increased the forma-
tion of TTHMs and THAAs. The residual percentage of
humic acid with a high dosage of ClO2 was lowered, but
the residual percentage of DOC was higher, indicating
the increased decomposition of the humic acid molecu-
lar structure by the stronger oxidizing power of ClO2 at
higher doses. Further reaction under these conditions
produced more intermediates without removing the
DOC.
3.1.4. Low dosage of ClO2 for treating high DOC in
alkaline solution
In an alkaline solution, the consumption of ClO2
by the three organic acids and the production of
TTHMs were similar, as shown in Table 5. In the for-
mation of THAAs, vanillic acid produced a greater
amount than the other two acids, and with a di€erent
trend, which indicates the diculty in controlling
THAAs formation.
3.2.2. Various dosages of ClO2 for treating high DOC in
neutral solution
Under this condition, a higher dosage of ClO2 fa-
vored the oxidation reaction in this system. The ®nal
consumption of ClO2 at a higher dosage was greater
Table 5
Treatment of organic acids with ClO2 (DOC ˆ 10.0 mg/L, ClO2 ˆ 15.0 mg/L, pH ˆ 9)
Vanillic acid
p-Hydroxybenzoic acid
Humic acid
ClO2/DOCa
1.22
3.0
1.29
2.5
2.50
2.6
TTHMs, lg/L
TTHMs/DOCb
THAAs, lg/L
THAAs/DOCc
0.44
46.7
6.3
0.36
7.9
0.74
26.2
7.6
1.1
Major productsd
CHCl3 and DCAA
a ClO2 consumed/DOC removal.
b Final amount of TTHMs/DOC removal.
c Final amount of THAAs/DOC removal.
d Major products of TTHMs and THAAs.
Table 6
E€ect of increasing ClO2 on treatment
DOC,
mg/L
DOC
removal, %
ClO2
consumed, %
TTHMs,
lg/L
THAAs,
lg/L
Major product
of HAAs
Vanillic acid
5.0
10.0
5.0
Increased
Increased
Increased
Increased
Reduced
Increased
Reduced
b
Reduced
a;b
Increased
Reduced
Increased
TCAA
c
±
±
Increased
±
TCAA
p-hydroxy benzoic acid
Humic acid
Reduced
Reduced
Increased
b
a;d
c
10.0
5.0
±
±
b
±
TCAA
Reduced
Reduced
±
Reduced
a
c
10.0
±
±
a Di€erent trends of formation.
b Insigni®cant e€ect.
c TCAA as the major product at low dosage, DCAA as the major product at high dosage.
d Higher than which at low dosage in the former stage, lower in the later stage.

C.-Y. Chang et al. / Chemosphere 41 (2000) 1181±1186
1185
Table 7
E€ect of ClO2 treatment in alkaline solution (ClO2 ˆ 15.0 mg/L, DOC ˆ 10.0 mg/L, pH ˆ 9)
DOC removal
ClO2 consumed
TTHMs formed
THAAs formed
Major product
of HAAs
b
Vanillic acid
p-Hydroxybenzoic
acid
±
Increased
Increased
b
Reduceda
Reduced
Reduced
Increaseda
DCAAc
DCAAc
±
b
b
Humic acid
Reduced
±
Reduced
±
DCAAc
a Di€erent trends of formation in neutral and alkaline solutions.
b Insigni®cant e€ect.
c TCAA as the major product in neutral solution.
than that at a lower dosage. The removal percentage for
DOC and concentration of chlorate anion were much
higher. To a lesser extent, the concentrations of chlorite
and chloride anions were also higher.
around 1.2±10.2 lg/L and THAAs, 2.3±69.2 lg/L from
the ClO2 reactions with vanillic acid, p-hydroxybenzoic
acid and humic acid in water. According to the experi-
mental results, ClO2 is a recommendable alternative
disinfectant.
Table 6 summarizes the results, indicating that in-
creased dosages of ClO2 increase the DOC removal
percentage, decrease the ClO2 consumption and the
formation of TTHMs. The major product from a low
concentration of organic acid solution was TCAA, but
not DCAA, and the situation was reversed for higher
acid concentration.
Compared with the control factors for DBPs for-
mation, overall, there was no relationship between pH
and ClO2/DOC, but the result indicate that higher pH is
favorable for the removal of humic acid with large
molecular weight. Lower pH was favorable to the re-
moval of vanillic acid and p-hydroxybenzoic acid with
small molecular weight. Otherwise in the formation of
DBPs, lower pH reduced TTHMs production, but in-
creased the HAAs and TOX production. Increased ClO2
dosage increased the ratio of DOC removal and
TTHMs, THAAs all increased. It was suggested that a
higher ClO2 dosage increased the chance of reaction
with activating groups (±OH, ±OCH3) on the structure
of the organic precursors.
3.2.3. Low dosage of ClO2 for treating high DOC under
di€erent pH values
Using either a phosphate or boric acid bu€er solution
with vanillic acid with ClO2 to control the treatment pH,
the resulting for DOC removal was better in an alkaline
solution than in a neutral solution. The formation of
TTHMs and THAAs was substantially reduced with an
almost equal quantity of ClO2. However, the formation
of chlorite and chlorate anions was much higher than
with a lower dosage. The alkaline solution did not
control the formation of THAAs from p-hydroxyben-
zoic acid. With regard to ClO2 consumption, DOC re-
moval, formation of TTHMs and THAAs, there was
little di€erence for humic acid reactions in alkaline or
neutral solutions.
In the study of in¯uence by organic precursors, the
concentration of organic acids and the ratio of DOC
removal were increased at higher organic concentra-
tions. The formation of TTHMs and THAAs, which
were transformed from unit DOC consumption
(TTHMs/DOC, THAAs/DOC), was also increased. The
possible reason is that ClO2 only breaks down large
molecular structures into smaller structures. The above
phenomenon is encouraging in humic acid reactions.
According to these experiment observations, the reac-
tion occurred fast during the initial 24 h and a certain
level of DBPs was obtained. After 24 h, the greater the
reaction time, the greater the production, but the level
of increase was not great. This phenomenon will sup-
ply some operational options at the water treatment
facilities.
Table 7 shows that the pH in¯uence was more pro-
nounced for TTHMs in the case of vanillic acid. The
production of THAAs from p-hydroxybenzoic acid was
more pronounced and concentrated. For humic acid, pH
provided almost no in¯uence on the formation of
TTHMs and THAAs.
4. Conclusions and suggestions
Acknowledgements
Under the conditions of pH 7 and 9, DOC between 5
and 10 mg/L and a chlorine dioxide dosage between 15
and 30 mg/L (higher dosage than the normal operation
conditions (0.1±1 mg ClO2/L) to simulate the dosage
after a rainstorm). The formation of TTHMs was
This work was supported by the research grant No.
NSC85-2211-E005-018 from the National Science
Council in Taiwan.

1186
C.-Y. Chang et al. / Chemosphere 41 (2000) 1181±1186
disinfection by-products in US drinking water. J. AWWA
81, 41±53.
References
Lykins Jr., B.W., Griese, M.H., 1986. Using chlorine dioxide
for trihalomethane control. J. AWWA 78, 88±93.
Myer, A.G., 1990. Evaluating alternative disinfectants for
THM control in small systems. J. AWWA 82, 77±84.
Narkis, N., 1995. Disinfection of ecients by combnations of
chlorine dioxide and chlorine. Wat. Sci. & Technol. 31,
105±114.
Aieta, E.M., Berg, J.D., 1986. A view of chlorine dioxide in
drinking water treatment. J. AWWA 78, 62±73.
Arora, H., Le Chevallier, M.W., Dixon, K.L., 1990. DBP
occurrence survey. J. AWWA 89, 60±68.
AWWA Research Foundation & KIWA, 1986. Use of Alter-
native Disinfectants, Chlorination By-Products: Production
and Control, pp. 174±188.
National Cancer Institute, 1976. Report on Carcinogenesis
Bioassay of Chloroform, Bethesda, MD.
Bellar, T.A., Lichtenberg, J.J., Korner, R.C., 1974. The
occurrence of organohalids in chlorinated drinking water.
J. AWWA 66, 703±706.
National Research Council, 1980. Drinking Water and Health,
vol. 2, National Academy Press, Washington, DC.
Oliver, B.G., 1983. Dihaloacetonitriles in drinking water: algae
and fulvic acid as precusors. Environ. Sci. Technol. 17, 80±83.
Peters, R.J.B., de Leer, E.W.B, de Galan, L., 1990. Dihalo-
acetonitriles in dutch drinking water. Water Res. 24,
797±800.
Bull, R.J., Kop¯er, F.C., 1991. Health E€ect of Disinfectants
and Disinfection By-Products. AWWA Res. Fdn., Denver,
CO.
Craun, G.F., Bull, R.J., Clark, R.M., Doull, J., Grabow, W.,
Marsh, G.M., Okun, D.A., Regli, S., Sobsey, M.D.,
Symons, J.M., 1994. Balancing chemical and microbial
risks of drinking water disinfection, part I. Ben®ts and
potential risks. J. Water SRT-Aqua 43, 192±199.
Edwards, G.A., Amirtharajah, A., 1993. Removing color
caused by humic acids. J. AWWA 77, 50±62.
Rice, R.G., 1980. The use of ozone to control trihalomethanes
in drinking water treatment. Ozone Sci. & Eng. 2, 75±99.
Rook, J.J., 1974. Formation of haloform during chlorination of
natural waters. Water Treatm. Exam. 23, 234±243.
White, G.C., 1992. The Handbook of Chlorination and
Alternative Disinfectants, 3rd ed. Van Nostrand Reinhold
Company, New York, pp. 150±151.
Koch, B., Krasner, S.W., 1989. Occurrence of distribution by-
products in a distribution system. In: Proceedings of the
Annual Conference sponsored by American Water Works
Association, Pt. 2, pp. 1203±1230.
Yamada, H., Somiya, I., 1979. In¯uence of process conditions
on the e€ect of ozone treatment of organic substance in the
water. Ozone Sci. & Eng. 1, 61±72.
Krasner, S.W., McGuire, M.J., Jacangelo, J.G., Patania, N.L.,
Reagen, K.M., Aieta, E.M., 1989. The occurrence of