Reactions of aqueous chlorine and chlorine dioxide with L- tryptophan
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/AA00003776/00001
 Material Information
Title: Reactions of aqueous chlorine and chlorine dioxide with L- tryptophan genotoxicity studies and identification of some genotoxic reaction products
Physical Description: xii, 246 leaves : ill. ; 29 cm.
Language: English
Creator: Owusu-Yaw, Joe D., 1953-
Publication Date: 1989
 Subjects
Genre: bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis (Ph. D.)--University of Florida, 1989.
Bibliography: Includes bibliographical references (leaves 216-245).
Statement of Responsibility: by Joe. D. Owusu-Yaw.
General Note: Typescript.
General Note: Vita.
 Record Information
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: aleph - 001563046
notis - AHH6772
oclc - 22711565
System ID: AA00003776:00001

Full Text













REACTIONS OF AQUEOUS CHLORINE AND CHLORINE DIOXIDE WITH L-
TRYPTOPHAN: GENOTOXICITY STUDIES AND IDENTIFICATION OF
SOME GENOTOXIC REACTION PRODUCTS










BY





JOE D. QWUSU-YAW


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1989
















ACKNOWLEDGEMENTS

I wish to express my sincere gratitude to the chairman

of my committee, Dr. Cheng-i Wei, for his guidance, support

and encouragement throughout the course of this study. I

would also like to extend my appreciation to the cochair, Dr.

Willis B. Wheeler, for his support both financially and

otherwise. My very special thanks to Drs. Jesse F. Gregory,

Claude McGowan and Simon S. Yu for their kind assistance and

advice and their willingness to serve on my committee.

I also wish to acknowledge the assistance of Dr. John P.

Toth, Mr. Walter Jones and Mr. Samuel Y. Fernando. I also

wish to acknowledge the financial assistance of Dr. Rod

McDavis, the dean for minority affairs and his group at the

graduate school. Last but not least my special thanks to my

spouse Inza Gibson Owusu-Yaw and my kids Tristian and Jocelyn

who have shared with me the frustrations and joys of academic

work; and to my parents and aunt Agnes who made all this

possible.
















TABLE OF CONTENTS



ACKNOWLEDGEMENTS ................................ .....

LIST OF TABLES........................................

LIST OF FIGURES........................................

ABSTRACT ................................................

CHAPTER

I INTRODUCTION ....................................

Background.......................................
Objectives of the Study..........................

II LITERATURE REVIEW...............................

Historical Perspectives.........................

Chemistry of Aqueous Chlorine....................
Reaction with inorganics....................
Reaction with organics......................
Chlorine disinfection........................

Chemistry of Chlorine Dioxide...................
Reaction with inorganics....................
Reaction with organics......................
Chlorine dioxide disinfection...............

Reactions of Aqueous Chlorine with Amino
Acids, Peptides and proteins................

Reactions of Chlorine Dioxide with Amino
Acids, Peptides and Proteins................

Toxicological Assessment........................
Overview................................. ....
Toxicology of chlorine and chlorine
dioxide.................................
Mutagenesis and carcinogenesis..............


iii


Page

ii

vi

ix

xi



1

1
4

6

6

8
9
13
20

25
27
29
32


35


44

49
50

53
62









III MATERIALS AND METHODS........................... 76

Chlorine Demand-Free Water...................... 76

Generation of Aqueous Chlorine.................. 76

Generation of Aqueous Chlorine Dioxide........... 77

Iodometric Titration............................ 78

Reactions of Aqueous Chlorine or Chlorine
Dioxide with L-Tryptophan................... 79
Preliminary reactions of aqueous
chlorine................................. 79
Reactions of aqueous chlorine and chlorine
dioxide.................................. 79

Chlorine Consumption............................ 81

Concentration Techniques........................ 81
Liquid-liquid extraction.................... 81
Amberlite XAD polymeric resin adsorption.... 84
Rotary evaporation.......................... 87

Fractionation of Reaction Product Extracts...... 89
Thin layer chromatography................... 89

Genotoxicity Assays............................. 91
Ames Salmonella/mammalian microsome assay... 91
Sister chromatid exchange (SCE) assay....... 94
Chromosome preparations..................... 95

Gas Chromatograph/Mass Spectrometer/Data
System...................................... 96

IV RESULTS......................................... 98

Mutagenicity of the Reaction Products of
Aqueous Chlorination of L-Tryptophan........ 98
Rotary evaporation.......................... 98
Liquid-liquid extraction.................... 102

Mutagenicity of the Reaction Products of
Aqueous Chlorine or Chlorine Dioxide
with L-Tryptophan........................... 113
Liquid-liquid extraction.................... 113
Amberlite XAD adsorption .................... 118








Fractionation of the Reaction Products of
Aqueous Chlorine or Chlorine Dioxide and
Tryptophan.................................. 135
Thin layer chromatographic analyses of the
chlorination reaction products........... 135

Induction of Sister Chromatid Exchange by the
TLC Fractions of the Reaction Products of
Aqueous Chlorine or Chlorine Dioxide
with Tryptophan............................. 152

Identification of the Reaction Products.......... 155

V DISCUSSION....................................... 161

Reactions of Aqueous Chlorine and Chlorine
Dioxide with Tryptophan..................... 162

Mutagenicity Evaluation......................... 165
Concentration of the reaction products....... 166
Fractionation of the reaction products....... 182
Comparison of liquid-liquid extraction
and Amberlite XAD 2/8 adsorption........ 185
Comparison of aqueous chlorine and chlorine
dioxide.................................. 188

Induction of Sister Chromatid Exchange by
the TLC Subfractions of the Reaction
Products of Aqueous Chlorine and
Chlorine Dioxide with Tryptophan............. 193

Identification of the Reaction Products.......... 196

Implications of the Mutagenicity to
Carcinogenicity of the Findings.............. 202

VI SUMMARY.......................................... 205

APPENDIX.............................................. 208

REFERENCES............................................ 216

BIOGRAPHICAL SKETCH................................... 246















LIST OF TABLES


Table Page

1 Some halogenated compounds formed in
drinking water as a result of water
chlorination................................... 65

2 Mutagenicity of the aqueous concentrate arising
from the reaction of equimolar concentrations
of aqueous chlorine and tryptophan using
Salmonella typhimurium strain TA98............. 100

3 Mutagenicity of the aqueous concentrate arising
from the reaction of equimolar concentrations
of aqueous chlorine and tryptophan using
Salmonella typhimurium strain TA100............. 101

4 Mutagenicity of the liquid-liquid extracts of
the reaction products of aqueous chlorination
of tryptophan using Salmonella typhimurium
strain TA98.................................... 104

5 Mutagenicity of the liquid-liquid extracts of
the reaction products of aqueous chlorination
of tryptophan using Salmonella typhimurium
strain TA100..................................... 106

6 Mutagenicity of the liquid-liquid extracts of
the reaction products arising from the
chlorination of tryptophan using Salmonella
typhimurium strain TA98........................ 110

7 Mutagenicity of the liquid-liquid extracts of
the reaction products arising from the
chlorination of tryptophan using Salmonella
typhimurium strain TA100....................... 111

8 Mutagenicity of the ethyl ether extracts arising
from the aqueous chlorination of tryptophan
using Salmonella typhimurium strain TA98....... 114

9 Mutagenicity of the ethyl ether extracts arising
from the aqueous chlorination of tryptophan
using Salmonella typhimurium strain TA100....... 115








10 Mutagenicity of the ethyl ether extracts arising
from the reaction of aqueous chlorine dioxide
with tryptophan using Salmonella typhimurium
strain TA98.................................... 116

11 Mutagenicity of the ethyl ether extracts arising
from the reaction of aqueous chlorine dioxide
with tryptophan using Salmonella typhimurium
strain TA100..................................... 119

12 Percent distribution of radioactivity after resin
adsorption in the reaction of aqueous chlorine
with tryptophan................................ 121

13 Recovery of radioactivity after resin adsorption
in the reaction of aqueous chlorine and chlorine
dioxide with tryptophan........................ 122

14 Percent distribution of radioactivity after resin
adsorption in the reactions of aqueous chlorine or
chlorine dioxide with tryptophan............... 123

15 Mutagenicity of the Amberlite XAD eluates arising
from control aqueous chlorine, chlorine dioxide
and tryptophan solutions....................... 125

16 Mutagenicity of the Amberlite XAD eluates arising
from the reaction of aqueous chlorine with
tryptophan using Salmonella typhimurium strain
TA98.......................................... 127

17 Mutagenicity of the Amberlite XAD eluates arising
from the reaction of aqueous chlorine with
tryptophan using Salmonella typhimurium strain
TA100.......................................... 131

18 Mutagenicity of the Amberlite XAD eluates arising
from the reaction of aqueous chlorine dioxide
with tryptophan using Salmonella typhimurium
strain TA98.................................... 133

19 Mutagenicity of the Amberlite XAD eluates arising
from the reaction of aqueous chlorine dioxide
with tryptophan using Salmonella typhimurium
strain TA100..................................... 137

20 Distribution of radioactivity in the TLC
subfractions of the reaction products of
aqueous chlorine or chlorine dioxide with
tryptophan...................................... 142


vii








21 Mutagenicity of the TLC subfractions obtained from
the Amberlite XAD eluates of the reaction products
of aqueous chlorination of tryptophan using
Salmonella typhimurium strain TA98............. 144

22 Mutagenicity of the TLC subfractions obtained from
the Amberlite XAD eluates of the reaction products
of aqueous chlorination of tryptophan using
Salmonella typhimurium strain TA100............ 146

23 Mutagenicity of the TLC fraction of the reaction
products of the aqueous chlorination of
tryptophan re-developed on a second plate...... 149

24 Mutagenicity of the TLC subfractions obtained from
the Amberlite XAD eluates of the reaction products
of chlorine dioxide and tryptophan using
Salmonella typhimurium strain TA98............. 151

25 Mutagenicity of the TLC subfractions obtained from
the Amberlite XAD eluates of the reaction products
of chlorine dioxide and tryptophan using
Salmonella typhimurium strain TA100............ 153

26 Sister chromatid exchange frequencies induced by
the TLC subfractions of the reaction products of
aqueous chlorine and chlorine dioxide with
tryptophan...................................... 154

27 Some physical properties of Amberlite XAD-2 and
XAD-8 resins ................................... 175


viii















LIST OF FIGURES


Figure Page

1 Chlorine substitution reactions with some
phenolic compounds and aromatic organic acids.. 18

2 Chlorine substitution reactions with
pyrimidines and purines........................ 19

3 Detailed reaction mechanism of the haloform
reaction....................................... 21

4 Reactions of chlorine with amino acids leading
to the formation of nitriles and aldehydes..... 39

5 Schematic presentation of sample preparation
for mutagenicity assessment in the preliminary
studies......................................... 83

6 Schematic presentation of sample preparation
for mutagenicity assessment of the reactions
of aqueous chlorine or chlorine dioxide with
tryptophan...................................... 88

7 Schematic presentation of sample for thin layer
chromatography and genotoxicity assessment..... 136

8 Reconstructed ion chromatogram of the reaction
products from the methylene chloride extract... 156

9 Reconstructed ion chromatogram of the thin
layer chromatography subfraction of the
reaction products of chlorine with
tryptophan...................................... 159

10 Reconstructed ion chromatogram of the thin
layer chromatography subfraction of the
reaction products of chlorine dioxide with
tryptophan...................................... 160

11 Comparison of liquid-liquid extraction and
Amberlite XAD 2/8 adsorption................... 187









12 Comparison of aqueous chlorine and chlorine
dioxide......................................... 190

13 Mechanism for the reaction of tryptophan with
aqueous chlorine............................... 197

14 Electron impact mass spectra of 1,1,3-tri-
chloropropanone and 1,1,3,3-tetrachloro-
propanone....................................... 199

15 Proposed scheme for the formation of unsatur-
ated alkanoic acids........................... 200

16 Proposed pathway for the formation of chlorin-
ated acetones and aldehydes from resorcinol.... 201

17 Proposed mechanism for the formation of
quinoline from tryptophan...................... 203















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of
the Requirements of the Degree of
Doctor of Philosophy

REACTIONS OF AQUEOUS CHLORINE AND CHLORINE DIOXIDE WITH L-
TRYPTOPHAN: GENOTOXICITY STUDIES AND IDENTIFICATION OF
SOME GENOTOXIC REACTION PRODUCTS

By

JOE D. OWUSU-YAW

AUGUST 1989


Chairman: Cheng-i Wei
Cochairman: Willis B. Wheeler
Major Department: Food Science and Human Nutrition

Aqueous chlorine and chlorine dioxide reacted readily

with tryptophan at three different molar ratios (1:1, 3:1 and

7:1 disinfectant:tryptophan) to produce complex reaction

mixtures and a dark precipitate. Evaluation of the mutagenic

activity of the reaction products using the Ames Salmonella/

microsome assay indicated that they were mutagenic at all the

three ratios used. The nonvolatile mutagenic reaction

products were mainly acidic and neutral compounds; they were

direct-acting and caused both frameshift and base-pair

substitution mutations. The reaction products were also

capable of increasing the sister chromatid exchange

frequencies in vitro in Chinese hamster ovary cells.








Amberlite XAD polymeric resin adsorption was superior to

liquid-liquid extraction for concentrating small quantities

of organic mutagens from the model reaction mixtures.

At the lower disinfectant doses (1:1 and 3:1), the

reaction products of chlorine dioxide with tryptophan were

either more potent mutagens or were similar to that of aqueous

chlorine; while at the highest ratio (7:1), aqueous chlorine

reaction products were consistently more mutagenic. Thin

layer chromatographic separation of the reaction products and

detection under ultraviolet light revealed the presence of

several blue and green fluorescent compounds which had not

been previously reported. The use of radiolabeled

14C-tryptophan indicated the most potent mutagens were produced

in very minute quantities and represented less than 1% of the

total starting material. Gas chromatography/mass spectrometry

of the reaction products at the 1:1 molar ratio did not

reveal the presence of any chlorinated compounds. The

compounds identified at this ratio included isatin, oxindole

and indolylacetonitrile. When an excess of chlorine (7:1

molar) was used, several chlorinated compounds were

identified. Chloral, 1,1,3,3-tetrachloropropanone and

1,1,3-trichloropropanone are some of the known mutagens and

chloroform precursors identified from these reactions.


xii















CHAPTER I
INTRODUCTION


Background

Chlorine is used on a worldwide basis as a disinfectant

for potable water and for treatment of sewage effluents. It

is also used as an anti-foulant in electric plants and as a

biocide for the treatment of industrial waters and in food

processing (White, 1972; Morris, 1971). The commercial use

of chlorine has greatly improved the quality of life and

prevented the transmission of water-borne diseases throughout

the world. In the food industry, chlorine is intentionally

added to flour for oxidizing and bleaching purposes, and to

cooling and cleaning waters for milk, fish, meats, poultry and

fruits and vegetables (Smith and Arends, 1984; Foegeding,

1983; Kotula et al., 1974; Ranken et al., 1965; Mercer and

Sommers, 1957). Dilute aqueous chlorine solutions may be used

to decontaminate fruits, eggs and vegetables (Dychdala, 1983).

Chlorine is FDA approved and may also come into contact with

foods unintentionally when it is used as a cleaning and

sterilizing agent for food processing equipment and packaging.

Concern about the use of chlorine as a disinfectant for

potable water arose when it was discovered that








2

trihalomethanes (THM) are produced as undesirable side

products as a result of water chlorination practices (Bellar

et al., 1974; Rook, 1974). These concerns were exacerbated

when the National Cancer Institute confirmed that chloroform,

a major trihalomethane, was carcinogenic in laboratory animals

(NCI, 1976). This led the U.S. Environmental Protection

Agency to make an amendment to the National Interim Primary

Drinking Water Regulations that set a maximum contaminant

level of 0.10 mg/L for total THMs (USEPA, 1983). Since this

revelation, other halogenated and non-halogenated byproducts

have been shown to be produced as a result of water

chlorination practices (Coleman et al., 1984; Christman et

al., 1983; Loper et al., 1978; Rook, 1977). In addition to

chloroform, other halogenated byproducts such as the di- and

trichloroacetonitriles and their brominated analogs which are

produced during water chlorination, have also been shown to

be genotoxic (Bull et al., 1986; Daniel et al., 1986;

Weisburger, 1977).

Free, proteinaceous and bound amino acids as well as

fulvic and humic substances present in potable water have been

implicated and confirmed as precursors of chlorination

reaction products (Bierber and Trehy, 1983; Oliver, 1983;

Lytle and Perdue, 1981; Trehy and Bierber, 1981; Christman,

1980). In fact, extracts of chlorinated drinking water, amino

acids and fulvic and humic substances have been shown to be

genotoxic in different short-term assays (Horth et al., 1987;








3

Tan et al., 1987b; Fielding and Horth, 1986; Bull and

Robinson, 1985; Meier et al., 1983; Sussmuth, 1982; Nazar and

Rapson, 1982; Loper, 1980; Nestmann et al., 1980). The

difficulty encountered in the interpretation of these findings

is that other poorly characterized compounds are also present

which may or may not contribute to the observed genotoxicity.

Epidemiological evidence has also been adduced to link certain

cancers of the colon, bladder and rectum to water chlorination

practices (Cantor et al., 1985; Cragle et al., 1985; Crump,

1983; Williamson, 1981; Cantor and McCage, 1978).

The formation of undesirable toxic side products during

chlorination to produce potable water and the need to reduce

total THM levels in drinking water to below 0.10 mg/L, has led

to a search for alternate disinfectants (USEPA, 1983).

Chlorine dioxide has been proposed as an alternative oxidant

and drinking water disinfectant in America and has received

numerous favorable reviews (Knocke et al., 1987; Lauer et al.,

1986; Lykins and Griese, 1986; Masschelein, 1985; Aieta et

al., 1980). This is because, unlike chlorine, chlorine

dioxide produces very little or no THMs in treated water and

it does not react with ammonia commonly found in water to form

chloramines (Federal Register, 1983; Symons et al., 1981;

White, 1972). Even though chlorine dioxide produces much

fewer THMs, its use is limited by cost (about 3 times that of

chlorine) and the possible health hazards associated with its

biological imapct and long-term use which are not known.










Objectives of the Study

Amino acids are ubiquitous in the environment and have

been shown to act as precursors of mutagenic and/or

carcinogenic byproducts in drinking water. Based on the

reactions of chlorine and/or chlorine dioxide with amino

acids, the following general objectives were proposed for this

study:

1. To investigate the mutagenic potential of the

reaction products of the environmentally signi-

ficant amino acid, tryptophan, with different

molar concentrations of aqueous chlorine or

chlorine dioxide using available concentration

techniques and the Ames Salmonella mutagenicity

assay.

2. To compare liquid-liquid extraction and Amberlite

XAD adsorption for concentrating the reaction

products and also to compare the mutagenic

activities of aqueous chlorine and chlorine dioxide

reaction products.

3. To separate the reaction products using different

chromatographic methods and to re-evaluate the

mutagenicity of the fractions using the Ames

Salmonella/microsome assay. In addition, to

investigate the highly mutagenic subfractions for

genotoxicity using the sister chromatid exchange

assay in cultured Chinese hamster ovary cells.








5

4. To identify the reaction products in the highly

mutagenic fractions and propose mechanisms for their

formation.














CHAPTER II
LITERATURE REVIEW


Historical Perspectives

Chlorine is an member of the halogen family and although

it is one of the most widely distributed elements on earth,

it is never found uncombined in nature. It exists in the

earth's crust in the form of soluble salts of potassium,

sodium, calcium and magnesium (Dychdala, 1983; White, 1972).

Chlorine is believed to have been known to medieval

chemists long before it was discovered in its gaseous state

in 1774 by a Swedish chemist named Steele (Baldwin, 1927).

Steele called the gas dephlogisticated muriatic acid and

observed that it was soluble in water and had permanent

bleaching effects on paper, vegetables and flowers. The gas

was liquefied by compression in 1805 and was found to act on

metals and their oxides. Davy declared the gas to be an

element and named it chlorine, derived from the Greek word

chlorouss" which has been translated to mean green, greenish

yellow or yellowish green, based on its color (Baldwin, 1927).

In 1811 Davy discovered chlorine dioxide gas when he poured

a strong acid on sodium chlorate (White, 1972). He observed

that the gas had an intense color similar to that of chlorine

and subsequently named it euchlorine.








7

Even though the bleaching action of chlorine had been

used for a long time, the disinfecting and deodorizing

properties were not known until the first part of the

nineteenth century. Chlorinated lime was used for

disinfecting and deodorizing hospitals and for sewage

treatment in London in 1854 (Dychdala, 1983). In 1881, Koch

demonstrated the bactericidal action of chlorine as

hypochlorite on pure bacterial cultures under controlled

laboratory conditions and in 1886, the American Public Health

Association issued a favorable report on the use of

hypochlorites as disinfectants. In 1894, Traube established

the first guidelines for hypochlorites in water purification

and disinfection (Dychdala, 1983; Hadfield, 1957). Chloride

of lime was first introduced to north America for water

treatment in the early part of the nineteenth century and

within a few years, over 800 million gallons of water had been

purified by chlorination (Dychdala, 1983; Race, 1918). To

date, chlorine has gained worldwide recognition as the

disinfectant of choice for the treatment of potable water.

The use of chlorine in medicine began in 1915 when Dakin

used dilute solutions of sodium hypochlorite for disinfecting

open and infected wounds during World War I. Dakin's original

formulation contained chlorinated lime, sodium carbonate and

boric acid (Dychdala, 1983). The dairy industry was perhaps

the first food processing industry to utilize the disinfecting

and deodorizing properties of chlorine (Mercer and Sommers,








8

1957). In this respect chlorine was used as a sanitizer for

milk bottles and milk equipment. Use of chlorine then spread

to other food industries and it is used in all branches of

food processing including canning, flour, poultry, fish, meat

and fruits and vegetables industries.



Chemistry of Aqueous Chlorine

In water, chlorine gas undergoes rapid hydrolysis to form

hypochlorous acid according to the following equation (Morris,

1946):

Cl2 (g) + H20 > H* + Cl + HOC1 1.

A small amount of a id H+, is produced; however, it tends to

lower the pH of the solution. Other molecular and ionic

species such as H20C1+, OC1 and Cl2 are produced; however,

HOC1 is by far the most reactive (Morris, 1978). Consistent

with industrial practice, the term chlorine will be used in

this text to refer to aqueous solutions of the active chlorine

compounds, OC1", C12, and HOC1. Molecular chlorine gas will

be referred to as such or as chlorine gas. The active

chlorine compounds, OC1 Cl2 and HOCl coexist in solution and

are difficult to distinguish at the trace level, therefore,

they are grouped together as free residual chlorine (White,

1972).

Hypochlorous acid is classified as a weak acid which

tends to undergo partial dissociation as follows:

HOC1 > H + OC1 2.








9

The dissociation constant for equation 2 ranges from 1.6 to

3.2 x 10-8 for the temperature range 0-250C (Morris, 1966).

The amounts and ratios of the different ionic and molecular

species present at any particular time are dependent upon pH,

temperature and chlorinity or dissolved solids. For example,

at pH 6.0 and a temperature of 150C, the fractions of HOC1, Cl2

and OC1 are 0.975, 3.6 x 10-5 and 0.025, respectively.

However, when the pH is increased to 9.0 under identical

conditions, the fractions become 0.038, 1 x 10-9 and 0.962,

respectively (Morris, 1978).

The above-mentioned aqueous chlorine species have

different reactivities which are related to their redox

potentials as follows (Masschelein, 1979):

C12 + 2e > 2C1 (1.36V) 3.

HOC1 + H+ + 2e" > Cl1 + H20 (1.49V) 4.

OC1 + H20 + 2e -> Cl + 20H- (0.9V) 5.

Of the three species, HOC1 is by far the most reactive in the

pH range of 6-7. Morris (1978) has shown that towards

nitrogenous substrates, HOC1 is 300, 800 and 4 x 104 more

reactive than Ci2, H2OC1 and OC1', respectively, in aqueous

solutions at 150C and neutral pH.

Reaction with inorganics

Due to its reactivity, chlorine or hypochlorous acid

readily oxidizes all organic and inorganic materials in

aqueous solutions until at some breakpointt," when demand is

fully satisfied (Dychdala, 1983). Inorganic substances such








10

as carbon, cyanide, iron, nitrite, sulfite, sulfide and

manganese react with hypochlorous acid via oxidation. Either

the chlorine atom or the oxygen atom of hypochlorous acid may

act as the center of reaction. The chlorine atom is more

electropositive, therefore, electrophilic reactions proceed

by way of the chlorine atom (Morris, 1978).

Chlorine reacts rapidly with inorganic carbon according

to the following expression:

C + 2C12 + 2H20 > 4HC1 + CO2 6.

The above reaction takes place in dechlorination processes

which usually employ granular carbon filters, and the acid

formed consumes alkalinity in treated waters (White, 1972).

Reducing substances. Hydrogen sulfide, characterized by its

obnoxious odors, is a constituent of septic sewage and is

commonly found dissolved in underground waters. During

drinking water chlorination, hydrogen sulfide is completely

oxidized to sulfate according to equation 7 (White, 1972):

H2S + 4C12 + 4H20 > H2SO4 + 8HC1 7.

Nitrites are oxidized readily by free available chlorine to

produce nitrates in a reaction similar to the above (Huburt,

1934):

NO2 + HOC1 -> N03 + H+ + Cl 8.

Chlorine reacts with iron in water supplies to remove

iron from water and to produce a coagulant for both water and

sewage treatment. Similarly, under alkaline conditions,

chlorine reacts with cyanide to form cyanate and with soluble








11

manganese compounds to produce insoluble manganese dioxide

which can be removed by filtration. The cyanide process

represents a detoxication mechanism in water treatment

operations (White, 1972).

Ammonia. The reaction of chlorine with ammonia in dilute

aqueous solutions is of practical importance in water

treatment and has been shown to produce three types of

N-chloro compounds or chloramines. In these reactions, the

electrophilic chlorine atom from HOC1 attaches itself to the

bare electron pair of the nitrogen atom with the concurrent

release of H+ from ammonia and OH- from HOC1 to rapidly form

monochloramine and water as shown in equation 9 (Morris,

1978). The reaction can proceed slowly to form dichloramine

and even more slowly to produce trichloramine or nitrogen

trichloride according to the following equations:

HOCl + NH3 > NH2Cl + H20 9.

HOC1 + NH2C1 > NHC12 + H20 10.

HOC1 + NHC12 > NCl3 + H20 11.

The reactions occur in sequence and, therefore, compete with

each other. The chloramines possess some oxidative power and

are collectively referred to as combined residual chlorine.

Factors affecting the types and ratios of chloramines formed

include pH, temperature, contact time and initial chlorine to

ammonia ratio (White, 1972). For example, equation 9 will

rapidly convert all the free chlorine to monochloramine at pH

7-8 when the ratio of chlorine to ammonia is equimolar, i.e.,








12

5:1 by weight or less. When the molar ratio of chlorine to

ammonia is increased, products from equations 10 and 11 can

also be formed depending on the conditions (White, 1972).

Equations 10 and 11 are related to the breakpointt"

phenomenon used in water treatment for the control of

obnoxious tastes and odors. "Breakpoint" chlorination refers

to the use of excess amounts of chlorine beyond the water

demand, thus resulting in residual free available chlorine

(Griffin, 1946). Breakpoint chlorination leads to increased

germicidal efficiency and disappearance of foul tastes and

odors naturally present in water. At or above the breakpoint,

the initial monochloramine starts a disproportionation

reaction to form dichloramine according to equation 10, which

in turn forms trichloramine or nitrogen trichloride as the

chlorine residuals increase further (White, 1972; Griffin,

1944; Griffin and Chamberlin, 1941).

Other halogens. Apart from chlorine, bromine is perhaps the

second most important halogen with regard to water

chlorination. Bromide, a constituent of natural waters and

seawater, reacts rapidly with chlorine or hypochlorite

according equation 12:

Br + HOC1 > HOBr + Cl 12.

The HOBr is a strong electrophile and reacts much more rapidly

than HOC1 (Morris, 1978). A primary difference between

bromine and chlorine systems is that at neutral pH there are

higher concentrations of molecular bromine, Br2, than








13

molecular chlorine. Also, at any given pH, HOBr ionizes much

more rapidly than HOCl (Jolley and Carpenter, 1983). After

reaction with organic, HOBr is reduced to Br" which is

reoxidized to HOBr by residual aqueous chlorine in the water.

The overall result is enhanced reactivity of aqueous chlorine

in the presence of the bromide ion (Morris, 1978).

Upon chlorination of natural waters containing iodide,

chlorine readily oxidizes the iodide according to the

following equation which forms the basis of the iodometric

method for chlorine determination (Jolley and Carpenter, 1983;

Lister and Rosenblum, 1963).

HOC1 + 21' > 12 + Cl + OH" 13.

Molecular iodine hydrolyses rapidly to form hypoiodous acid

which undergoes dissociation according to equation 14:

HOI > H* + OI 14.

Equation 14 becomes important only at high pH and is not of

practical significance in the chlorination of either natural

or wastewater systems (Jolley and Carpenter, 1983).

Reaction with organics

Reaction of chlorine with organic in drinking water

proceeds via oxidative degradation or chlorine incorporation

through electrophilic substitution or addition reactions. The

major known products of the latter reactions are THMs whose

identification in treated water aroused serious health

concerns in the United States and elsewhere (Bellar, et al.,

1974; Rook, 1974).








14

Oxidation. The predominant type of reaction involving

chlorine and organic compounds in natural and process waters

is oxidation. Carbohydrates and related compounds constitute

a class of organic compounds likely to undergo oxidation

reactions in aqueous solutions (Jolley et al., 1978; Whistler

et al., 1966). Heterocyclic aromatic compounds such as the

purines, pyrimidines and indoles have been shown to undergo

oxidative reactions in dilute aqueous chlorine solutions (Lin

and Carlson, 1984; Hoyano et al., 1973). Indole has been

shown to yield oxindole and isatin upon reaction with chlorine

in dilute aqueous solutions (Lin and Carlson, 1984).

Other classes of compounds known to undergo oxidative

reactions with aqueous chlorine or hypochlorite solutions

include proteins and amino acids, sulfur-containing compounds,

lipids, and polynuclear aromatic hydrocarbons (Trehy et al.,

1986; Lin and Carlson, 1984; Oyler et al., 1983; Dychdala,

1983; Ghanbari et al,, 1983; Burleson et al., 1980).

Addition. Addition of chlorine or halogens to organic

compounds may occur by way of reactive double bonds. This

type of reaction is another example of electrophilic attack

in which the rate determining step is believed to be transfer

of Cl to the double bond to give a chloronium ion (Morris,

1978). Olefins and acetylenes represent a class of compounds

that are vulnerable to attack by chlorine. Their major

reaction with chlorine is by addition in which chlorine atoms

and hydroxyl groups are added to the double bonds (Ghanbari








15

et al., 1983; Carlson and Caple, 1978; Jolley et al., 1978;

Morris, 1978; Pitt et al., 1975; Leopold and Mutton, 1959).

For example, aqueous chlorination of oleic acid has been shown

to produce a mixture of chlorohydrins with water playing a

role in product formation (Carlson and Caple, 1978). Chlorine

addition to double bonds is a slow process and proceeds only

when the double bonds involved are activated (Morris, 1978).

Substitution. The two major types of substitution reactions

of chlorine and chlorine compounds involve formation of either

N- or C-chlorinated compounds.

Formation of N-chlorinated compounds. Nitrogen appears as a

natural component of foods and other substances either as

organic or inorganic nitrogen. Since inorganic nitrogen has

been mentioned earlier, the following discussion will focus

only on organic nitrogen. Sources of organic nitrogen in

natural and process waters include amides, amino acids,

indoles, proteins, nucleic acid bases, and their derivatives

as well as pesticides (Jolley et al., 1978). Organic nitrogen

in water exerts a chlorine demand and leads to the formation

of N-chlorinated compounds (White, 1972; Morris, 1967).

The major reaction mechanisms involved in the formation

of N-chlorinated compounds have been discussed by Morris

(1967). With amines, aqueous chlorine reacts to form N-chloro

compounds by a mechanism similar to its reaction with ammonia.

The electrophilic chlorine atom attaches itself to the bare

electron pair of nitrogen, thus releasing water. Conversely,








16

the reaction may be viewed as one of displacement whereby the

nucleophilic amine displaces OH from HOC1 (Morris, 1978).

Amides also appear to react via the same pattern or via the

haloform reaction but at much slower rates. A simplified

scheme of chlorine substitution with organic N-substituted

organic compounds is shown below (Morris, 1967):

R-NH2 + HOC1 > R-NHC1 + H20 15.

R-CO-NH2 + HOC1 > R-CO-NHC1 + H20 16.

In both cases, the rates of chloramination vary

proportionately with the basic strength or nucleophilicity of

the nitrogenous substrate (Morris, 1978). Proteins are more

complex than amines and amides and chlorination of organic

nitrogen from proteinaceous material is a slow and tedious

process. However, proteins may also undergo substitution with

chlorine to form N-chlorinated proteinaceous byproducts

(Jolley et al., 1978; White, 1972).

Formation of C-chlorinated compounds. Formation of

C-chlorinated compounds via chlorine substitution is perhaps

the most widespread in nature (Jolley et al., 1978; Morris,

1975; Carlson et al., 1975). The major substitution reactions

of chlorine involve substitution into aromatic or heterocyclic

compounds and the haloform reaction. Humic and fulvic

substances abound in nature and represent vegetation under

varying degrees of decay. These complex and otherwise poorly

characterized high molecular weight materials contain aromatic

and heterocyclic moeities which readily undergo chlorination








17

reactions in aqueous chlorine solutions when activated (Jolley

et al., 1978).

Some substitution reactions involving phenolic compounds

and aromatic organic acids are presented in Figure 1. As

before, the mechanism of the reaction involves an

electrophilic attack of HOC1 on the phenoxide ion. In this

case the reaction is facilitated by the presence of a negative

charge on the nucleophilic substrate and only when the ring

is "activated" by substituents such as the oxide ion, does

substitution occur (Morris, 1978). Phenols are more reactive

in dilute aqueous chlorine solutions than other aromatic

compounds possessing different substituent groups (Carlson and

Caple, 1978). Substitution may proceed at the activated ortho

or para positions not fully occupied by chlorine atoms.

Further rupture of the aromatic ring may occur by mechanisms

that are not fully understood (Morris, 1978). Heterocyclic

aromatic compounds such as the nucleic acid bases may also

undergo substitution reactions in which they are either

activated or inactivated toward electrophilic attack compared

to benzene (Morris, 1978). An example of the activation

process involves the chlorination of some purines and

pyrimidines. Aqueous chlorination of the pyrimidines,

cytosine and uracil has been shown to lead to the formation

of their corresponding 5-chloro analogs (Figure 2) and other

poorly characterized byproducts (Jolley et al., 1975; Hoyano

et al., 1973; Patton et al., 1972).














+ HOCI ---


+ HOCI ---


+ HOCI ---






OH
+ HOCI --


+ HOH






+ HOH


COOH


+ HOH
Cl





OH
+ HOH


COOH COOH
+ HOCI + HOH
COOH CI HI-a -COOH








Figure 1. Chlorine substitution reactions with some phenolic
compounds and aromatic organic acids (Adapted from
Jolley et al., 1978).










NH2


+ HOCI --


0 C
II II

HN I + HOCI HN!



0
II
HN N I
,NJ|I + HOCI -.

H H


6H, H
CH3


+ HOH






+ HOH


H HN
H H


+ HOH


3 OH C1 CH3

H N N

CH3


Figure 2. Chlorine substitution reactions with pyrimidines
and purines (Adapted from Jolley et al., 1978).


NH2








20

Similar chlorinated products may be obtained with purines

as shown in Figure 2. With the purines, actual formation of

the chlorinated analogs has not been demonstrated in dilute

aqueous chlorine solutions but rather in non-polar solvents

and in acetic acid (Jolley et al., 1978). Sources of nucleic

acid bases in potable water are plankton, bacteria, and plant

and animal matter. Like the nucleic acid bases, pyrroles are

also thought to be greatly activated to undergo chlorine

substitution reactions (Morris, 1978). Pyridines on the other

hand are thought to be sufficiently deactivated so that

reaction with dilute aqueous chlorine solutions is not

expected to occur in the ring (Morris, 1978).

Finally, formation of C-chlorinated compounds may occur

through the haloform reaction (Figure 3) with humic material

leading to the formation of volatile compounds such as

chloroform (Morris, 1975; Rook, 1974).

Chlorine disinfection

Despite much research done in this area, the actual

mechanism of germicidal action of chlorine has not been fully

elucidated. It is well known that even in trace amounts,

chlorine has fast killing action on bacteria and several

theories have been proposed to explain this killing action.

To date, none of the theories has been fully accepted due to

a lack of substantive evidence (Dychdala, 1983; White, 1972).

The proposal that the germicidal action of chlorine is due to

the liberation of nascent oxygen held for some time until it















0
11 OH
R-C-CH3 so
slow


R-C-CHX

HOX

x2


0e

- R-C=CH2]

HOX rather
r fast
X2A


Oe 0
1 OH- II
R-C=CHX ..=B R-C-CH2X
less
slow


0 0
O OOH
R-C-CHX2 least R-C-CX2 -
least
slow




0
II OH
CHX3 + R-C-OH as H +
Fast at
pH 13


0e

R-C=CX2


HOX even
or faster
X2
0
II
R-C-CX3


Figure 3.


Detailed reaction mechanism of the haloform
reaction. X=Halogen (Adapted from Morris, 1985).








22

was finally dispelled. This is because other strong nascent

oxygen producers such as hydrogen peroxide and potassium

permanganate are less effective bactericides than chlorine

(White, 1972). Also, chlorine still possesses germicidal

activity under conditions which exclude direct oxidation of

bacterial protoplasm (Mercer and Somers, 1957). Other

theories, even though attractive, have not received the full

support of researchers in this area. Baker (1926) proposed

that chlorine reacted with bacteria cell membrane proteins to

produce N-chloro compounds which interfere with bacterial cell

metabolism leading to death. The N-chloro theory was taken

a step further by Rudolph and Levine (1941) who concluded that

the bactericidal activity of chlorine or hypochlorites

proceeds in two stages: 1) the penetration of an active

ingredient into the bacterial cell, and 2) the chemical

reaction of this ingredient with the cellular protoplasm to

form toxic N-chloro compounds which destroy the organism.

Mercer and Somers (1957) stated that bacterial death by

chlorine is a poisoning process in which a form of chlorine

combines chemically with the protoplasm of the bacterial cell

to produce toxic organic complexes.

Other proposed mechanisms of chlorine disinfection were

directed at key enzymes in the target organisms. Green and

Stumpf (1946) and later Knox et al. (1948) observed

experimentally that there was a correlation between the effect

of chlorine on bacterial growth and its effect on the rate of








23

glucose oxidation by certain bacteria. They were able to show

that a bacterial suspension became sterile when the organisms

lost their ability to oxidize glucose. The mechanism of

inhibition was thought to involve irreversible oxidation of

the sulfhydryl groups of essential enzymes like

triosephosphate dehydrogenase. The sulfhydryl oxidation

theory received additional support when chlorine was shown to

oxidize sulfhydryl radicals to sulfonyl compounds and sulfides

to sulfates under conditions that simulate water chlorination

(Ingols et al., 1953; Black and Goodson, 1952; Douglas and

Johnson, 1938). Wyss (1962) postulated the theory of

unbalanced growth in which destruction of part of the enzyme

system throws the bacteria out of metabolic balance leading

to death. Studying the effect of chlorine on bacterial

enzymes, Venkobachar and coworkers (1975) observed that in the

bactericidal concentration of chlorine total dehydrogenase

activity in Eschericia coli was markedly affected and the

inhibition correlated with percent bacteria kill. They also

observed that succinyl dehydrogenase activity in the bacteria

was inactivated probably due to the oxidation of sulfhydryl

groups at lower chlorine doses.

Using radiolabelled chlorine, Friberg (1956) observed

that the first contact oxidation reactions of chlorine with

bacterial cells was responsible for bacterial death and not

through chloramine formation. He later demonstrated that

chlorine, even in minute amounts, reacts with and results in








24

destructive permeability changes in bacterial cell walls

resulting in death (Friberg, 1957). Additional mechanisms of

chlorine disinfection have been reported to include induction

of chromosomal aberrations (Ingols, 1958); disruption of

protein synthesis (Bernade et al., 1967); reaction with

purines and pyrimidines of nucleic acids (Hoyano et al.,

1973); oxidative decarboxylation of amino acids (Pereira et

al., 1973); loss of macromolecules due to inhibition of oxygen

uptake and oxidative decarboxylation (Venkobachar et al.,

1975); production of lesions in DNA (Shih and Lederberg,

1976); inhibition of respiration and active nutrient transport

(Camper and McFeters, 1979); disruption of bacterial spore

coats and increasing spore permeability (Foegeding, 1983).

Chlorine has also been reported to inactivate poliovirus by

reacting with the outer protein coat of the virus (Tenno et

al., 1980) or by damaging the internal RNA (Olivieri et al.,

1980; O'Brien and Newman, 1979).

Of the active chlorine species, hypochlorous acid is by

far the most active disinfectant. Its germicidal efficiency

is related to the ease with which it can penetrate cell walls.

The ease of cell penetration is comparable to that of water

which is structurally similar. Also its relatively low

molecular weight and electrical neutrality allow easy cell

penetration (White, 1972). The germicidal efficiency of free

available chlorine depends on other factors such as pH,

temperature, chlorine concentration, length of contact time,








25

and type and concentration of microorganisms. The pH of the

solution is perhaps the most important of all the above

factors because an increase in pH causes a substantial

decrease in biocidal efficiency and vice versa. This is

because at higher pH, hypochlorous acid is converted to

hypochlorite ion (equation 2) which is an inferior biocide.

The hypochlorite ion obtained from hypochlorous acid

dissociation is a relatively poor disinfectant because of its

inability to diffuse through microbial cell walls. The

inability to traverse microbial cell walls is due to the

negative charge associated with it (White, 1972). Combined

available chlorine as chloramines are also less effective

bactericidal agents than chlorine. Chloramines are slow-

acting and require more than a hundred-fold increase in

contact time and much higher concentrations to achieve the

same bacterial lethality as chlorine (Kabler, 1953;

Butterfield, 1948).



Chemistry of Chlorine Dioxide

Chlorine dioxide is a polar gas that readily dissolves

in but does not react with water to produce aqueous chlorine

dioxide. It is a true dissolved gas because it does not react

with water (Masschelein, 1979). The basic reaction of Davy,

who discovered the gas by pouring a strong acid over sodium

chlorate in 1911, is still used today (White, 1972):

NaC1O3 + 2HC1 > ClO2 + NaCl + H20 + Cl 17.








26

Chlorine dioxide has an intense greenish yellow color with a

distinctive odor similar to chlorine but more irritating and

more toxic (White, 1972). One important physical property of

chlorine dioxide is that it is five times more soluble in

water than chlorine. It is extremely volatile and decays by

photodecomposition, volatilization and decay reactions. In

fact, it can easily be removed from aqueous solutions merely

by aeration. It is very unstable and explosive and as such,

cannot be transported but is usually prepared on site (White,

1972).

Chlorine dioxide does not belong to the family of

"available chlorine" compounds, i.e., those compounds that

hydrolyze to form hypochlorous acid, but due to its oxidizing

power it is referred to as having an available chlorine

content of 263 percent. It has a larger oxidation capacity

than HOC1 because it can accept 2.5 times more electrons than

HOC1 (Ingols and Ridenour, 1948). This can be deduced from

reactions involving iodine liberation from iodide in the

iodometric titration method:

C10 + 51' + 4H -> Cli + 2.5I2 + 2H20 18.

HOC1 + 21' > Cl + I2 + OH 19.

The oxidation potential of chlorine dioxide is less than

that of HOC1 as indicated by its redox potential of 0.95V

compared to that of HOC1 in equation 4. The redox potential

of chlorine dioxide and chlorite are presented below

(Masschelein, 1979):










C10 + e- > C102 (0.95V) 20.

C102 + 2H20 + 4e > Cl + 40H (0.78V) 21.

The oxidizing capacity of chlorine dioxide is not all used in

water treatment because in practice chlorine dioxide is

reduced to chlorite according to equation 20. The chlorite

ion is an effective oxidizing agent and is consumed at a much

slower rate in oxidation-reduction reactions.

Reaction with inorqanics

In the treatment of drinking water, chlorine dioxide is

commonly used for the removal of divalent ferrous iron and

manganese and sometimes sulfide through oxidation reactions

(White, 1972). Much higher temperatures are required for

oxidation of trivalent chromium salts to chromate and will

not be discussed here (Masschelein, 1979).

Oxidation of iron. In water treatment, an effective means of

iron removal is through chlorine dioxide oxidation of iron

from the moderately soluble ferrous (II) to the ferric (III)

state as an insoluble ferric hydroxide precipitate which can

be removed by filtration. The above reaction is presented in

equation 22.

ClO2 + FeO + NaOH + H20 > Fe(OH)3 + NaC102 22.

The pH optimum for the above reaction is above 7.0, and is

preferable in the range between pH 8-9. The reaction is a

slow one, and in most cases, chlorine is the chemical of

choice for this type of reaction. This is because chlorine

is cheaper and equally effective as chlorine dioxide (White,








28

1972). Chlorine dioxide has been shown to oxidize organically

bound iron (Masschelein, 1979).

Oxidation of manganese. Manganese is present in many wells

and surface waters and can cause blackness of water at high

concentrations. Even though it is not known to cause any

health problems, the maximum level permitted in water is 0.05

mg/L (Knocke et al., 1987). Chlorine is used for manganese

removal but the reaction is too slow to be desirable.

Chlorine dioxide reacts more rapidly with manganese compounds

than chlorine. In this reaction, manganese (II) is oxidized

to manganese (IV) as insoluble manganese dioxide according to

the following reaction (White, 1972):

2C102 + MnSO4 + 4NaOH > MnO2 + 2NaC102 +

Na2SO4 + 2H20 23.

The stoichiometric dosage of chlorine dioxide required for

manganese oxidation has been established as 2.45 mg/mg Mn2

assuming the reduction byproduct is chloride. However, under

laboratory conditions, much higher doses of chlorine dioxide

are required. Knocke and coworkers (1987) have recently shown

that at least twice the stoichiometric amount of chlorine

dioxide was required to lower manganese levels to less than

0.05 mg/L. Also, when the total organic carbon content of the

water was high, much more chlorine dioxide was required than

permitted by law to limit manganese content to acceptable

levels.








29

Optimum conditions for the reaction have not been fully

elucidated but it is thought that complete oxidation of

manganese occurs at pH values higher than 7 (White, 1972).

Like iron, chlorine dioxide will oxidize organically bound

manganese. The chlorite formed from chlorine dioxide

reduction can also be used to reduce manganese or iron under

similar conditions (Masschelein, 1979).

Oxidation of sulfides. Hydrogen sulfide is found dissolved

in underground waters and has been shown to react with

chlorine to produce sulfates and elemental sulfur. Chlorine

dioxide oxidation of hydrogen sulfide to sulfate occurs

extensively in the pH range 5-9 (White, 1972). The type of

products formed, either elemental sulfur or sulfate, is

dependent on pH and temperature, with higher temperatures and

pH favoring sulfate formation (Dohnalek and FitzPatrick,

1983). Reactions of chlorine with sulfur has fully been

investigated and characterized but the same does not hold true

for chlorine dioxide. Until more data becomes available on

chlorine dioxide oxidation of sulfides, chlorine remains the

chemical of choice for sulfide removal since it is cheaper and

data is available on the type of byproducts likely to be

encountered.

Reaction with organic

Chlorine dioxide is a selective oxidant which reacts with

organic matter via oxidation to produce both volatile and

non-volatile reaction products. In some cases, however, some








30

chlorinated compounds are encountered. This type of

chlorination reaction is thought to be an indirect one and

probably corresponds to a progressive reduction of the dioxide

passing through the hypochlorous acid stage (Masschelein,

1979). These chlorination reactions involving chlorine

dioxide arise presumably due to the presence of chlorine as

a contaminant in the dioxide used (Masschelein, 1979).

Chlorine dioxide does not react with ammonia to form

chloramines; therefore, chloramine toxicity is avoided with

its use. Unsubstituted aromatics, primary amines and many

other organic compounds are unreactive towards chlorine

dioxide (Raugh, 1979).

Oxidation of phenols and odor control. Phenols are ubiquitous

pollutants in water and the environment and have been used as

model compounds for chlorine reactions and haloform formation

(Aieta and Berg, 1986). Oxidation of phenols and phenolic

compounds by chlorine and chlorine dioxide has been

extensively investigated because phenolic tastes and odors are

produced in natural waters as a result of industrial waste

pollution of raw water. Petroleum and wood processing plants,

algae and decaying vegetation are principal sources of

phenolic compounds. Chlorination of polluted waters leads to

the formation of chlorophenols such as 2-chlorophenol ,

2,4-dichlorophenol, 2,4,6-trichlorophenol and others which

possess distinct offensive tastes and odors (White, 1972).








31

Chlorine dioxide has been shown to be specific for the

control of these phenolic and chlorophenolic tastes and odors

many times faster than free available chlorine. This is

related to the high oxidizing capacity of chlorine dioxide

when compared to chlorine (Masschelein, 1979; White, 1972).

Apart from control of phenolic tastes and odors, chlorine

dioxide has successfully been used against algae odors, fishy

tastes, and earthy-musty taste and odor compounds (Lalezary

et al., 1986; Wright, 1980; Medker et al., 1968).

Reaction of equal parts of chlorine dioxide with phenols

has been reported to yield p-benzoquinone and 2-chloro-

benzoquinone in the concentration range 1-1000 mg phenol/L.

Organic acids, mainly maleate and oxalate are also formed if

the ratio of chlorine dioxide to phenol is increased five-fold

(Wajon et al., 1982; Masschelein, 1979). Extended oxidation

with excess chlorine dioxide leads to the formation of

carboxylic acids and, under moderate conditions, quinones.

When excess phenols are used, chlorophenols are formed due to

the slow release of hypochlorous acid which subsequently

reacts with excess phenol (Wajon et al., 1982). Under water

treatment conditions, the chlorine dioxide to phenol ratio

will be greater than unity and the major products likely to

be encountered are p-benzoquinone and simple organic acids.

Other oxidative reactions. Carbohydrates, proteins and amino

acids and other food components have been shown to undergo

mainly oxidative reactions with chlorine dioxide and are








32

discussed later in this chapter. Environmentally significant

compounds such as polynuclear aromatic hydrocarbons (PAHs) and

their heterocylcic analogs also react with chlorine dioxide

to varying degrees. Lin and Carlson (1984) subjected some

environmental heterocycles to dilute aqueous solutions of

chlorine, chloramine and chlorine dioxide and the latter was

found to produce mainly oxygenated byproducts. For example,

chlorine dioxide oxidation of indole produced oxindole and

isatin.

Chlorination reactions. As mentioned earlier, chlorination

or chlorine incorporation is not an intrinsic reaction of

chlorine dioxide. These reactions are secondary reactions

which are thought to occur through the release of hypochlorous

acid from a primary product or from impurities in the chlorine

dioxide (Masschelein, 1979). In reactions involving chlorine

and chlorine dioxide incorporation into lipids, Ghanbari et

al. (1982b) showed that chlorine incorporation was many orders

of magnitude higher with chlorine than chlorine dioxide.

Chlorine dioxide disinfection

Even though the bactericidal properties of chlorine

dioxide were known at the beginning of this century, the

practical uses of said properties were not fully utilized

until the middle of the century. This is because chlorine

dioxide was first applied to water treatment around the middle

of the century and also increased environmental and water

pollution led to a search for alternate and varied








33

disinfectants (Masschelein, 1979; White, 1972). McCarthy

(1944) first recommended the use of chlorine dioxide for the

disinfection of water of low organic content. This is because

increasing the organic content of water decreased the

germicidal effectiveness of chlorine dioxide. Ridenour and

Ingols (1947) showed that chlorine dioxide was as effective

as chlorine for the control of E. coli and other bacterial

populations. These workers were the first to show that unlike

chlorine, the bactericidal efficiency of chlorine dioxide was

unaffected over a wide pH range (from pH 6-10), which makes

it suitable for the disinfection of high pH waters. On the

other hand, the germicidal effectiveness of chlorine dioxide

was found to decrease at low temperatures (Ridenour and

Ambruster, 1949).

Most of the studies regarding chlorine dioxide

disinfection have focused on a comparison with that of

chlorine and, as a result, very little information is

available on its mode of action. Some of the comparative

studies have shown that chlorine dioxide is a more efficient

virucide and bactericide than chlorine (Alvarez and O'Brien,

1982; Aieta et al., 1980; Longley et al., 1980; Sawyer, 1976).

The proposed mechanisms of action of chlorine dioxide on

microbes are similar to those of chlorine; either through a

reaction with macromolecules or through gross physiological

changes. Like chlorine, there is evidence that chlorine

dioxide acts on dehydrogenase enzymes located in bacterial








34

cytoplasmic membranes but this evidence has not been

substantiated.

Chlorine dioxide is reactive towards several compounds

including phenols, some amino acids and proteins and this

reactivity has been used to explain its bactericidal action.

Fujii and Ukita (1957) demonstrated that chlorine dioxide

oxidation of tryptophan produced some indole derivatives of

tryptophan which may explain the bactericidal action of

chlorine dioxide. On the mode of killing action of chlorine

dioxide on bacteria, Bernade and coworkers (1967) concluded

that chlorine dioxide was sufficiently reactive with several

amino acids to alter their structures and thus, react with

cell walls. They also observed that protein synthesis was

abruptly inhibited by the action of chlorine dioxide. Roller

et al. (1980) have reported that under laboratory conditions,

the level of chlorine dioxide used in treatment plants does

not affect DNA or protein synthesis. Contrary to the above,

Alvarez and O'Brien (1982) in their studies on poliovirus

concluded that chlorine dioxide reacted with viral RNA leading

to virus inactivation.

Noss and others (1983) investigated the reactivity of

chlorine dioxide with nucleic acids, proteins and amino acids.

Chlorine dioxide readily and rapidly reacted with the amino

acids cysteine, tyrosine and tryptophan but not with viral

RNA. The authors postulated that the virucidal activity of

chlorine dioxide was probably due to alteration of viral








35

proteins. Further reports by the same authors suggest that

chlorine dioxide reaction with tyrosine appears to be the main

mechanism of action of F2 virus inactivation by chlorine

dioxide (Olivieri et al., 1985).

Another probable mechanism of action of chlorine dioxide

may be due to its interaction with membrane lipids and fatty

acids leading to permeability changes in membranes. Ghanbari

et al. (1982b) showed that chlorine dioxide was highly

reactive towards free fatty acids leading to the formation of

mainly oxidized byproducts. The authors observed that free

fatty acids were more reactive with chlorine dioxide than

their methyl esters. From the foregoing discussion, it is

possible that chlorine dioxide oxidation of membrane lipids

and proteins leads to microbial death. Even though some of

the reactions have been reported to occur in model systems,

the actual mechanisms that take place in the intact

microorganisms leading to death still need to be proven.



Reactions of Aqueous Chlorine with Amino Acids, Peptides
and Proteins

Proteins are complex high molecular weight constituents

of plant and animal matter made up of carbon, nitrogen,

oxygen, hydrogen and, in some cases, sulfur. Of these

elements, carbon, nitrogen and sulfur exert a chlorine demand

(White, 1972). Proteins and protein material from many

sources contribute greatly to the organic nitrogen content of








36

natural waters and pose a problem during water chlorination

(Qualls and Johnson, 1985). Amino acids are water soluble and

are typically found in natural waters at concentrations in the

range 2-400 gg/L (Le Cloirec and Martin, 1985). During water

treatment, the amount of amino acids can increase

substantially due to hydrolysis-depolymerization of proteins

or peptides, or due to microbial metabolism (Le Cloirec and

Martin, 1985). In the food industry, amino acid

concentrations are likely to be much higher in process

cleaning waters, and superchlorination of process waters is

sometimes used to control microbial contamination (Robinson

et al., 1981).

The reactions of chlorine with amino acids, proteins and

peptides have been studied in detail and reported to produce

both oxidized and chlorinated byproducts under different

treatment conditions (Jolley et al., 1978; Pereira et al.,

1973). The first report in the literature on the chlorination

of amino acids was made by Langheld (1909) who reported that

sodium hypochlorite reacted with alpha-amino acids to form

mono- or dichloroamino acids which subsequently decomposed to

form aldehydes. Langheld (1909) observed that the rate of

monochloramine decomposition was affected by the amino group;

primary amino groups decomposed to give aldehydes or ketones

as well as ammonia while secondary amino groups produced

amines instead of ammonia. Dakin et al. (1916) and Dakin

(1916) confirmed Langheld's study and showed that nitriles








37

were formed as major products when excess chlorine was used.

Dakin and colleagues showed that hydrogen cyanide was produced

from glycine, acetonitrile from alanine, isobutyl cyanide from

leucine and cyanobenzene from phenyl acetate. The general

mechanism proposed by Dakin (1916) involves the initial

formation of a dichloroamino acid which undergoes further

decomposition to form a cyanide. The chlorination of several

amino acids including those studied by Dakin and coworkers has

been further investigated (Isaac and Morris, 1983a; Margerum

et al., 1979). These studies showed that at neutral pH all

the amino acids tested readily reacted with chlorine with the

possible involvement of N-chloro compounds as intermediates.

Rydon and Smith (1952) developed a paper chromatographic

method for determining N-chloroamino acid formation during

chlorination. They showed that several amino acids including

glycine, alanine, leucine, serine, threonine, lysine,

phenylalanine and histidine liberated chlorine from the

corresponding N-chloro derivatives which reacted positively

with a starch-potassium iodide spray. Cysteine, cystine,

methionine and tyrosine failed to release chlorine due

probably to the formation of mainly oxidized products.

The importance of pH in product formation was

investigated by Wright (1936) who observed that aldehydes were

formed from alpha-amino acids on reaction with sodium

hypochlorite at neutral pH while nitriles predominated at

lower pH. From his point of view, acidity favors chlorination








38

and alkalinity favors oxidation. Wright showed that under

acidic conditions glycine was first chlorinated, and the

resulting chloroamino acid was then oxidized by any available

chlorine present; whilst under alkaline conditions,

destruction of available chlorine was mainly due to oxidation.

Jolley and Carpenter (1983) also observed that at neutral pH,

monochlorinated amino acids were formed with aqueous chlorine,

while dichloro derivatives were produced as major products

under acidic conditions. Stanbro and Smith (1979) studied the

kinetics and mechanisms of decomposition of N-chloroalanine

in aqueous solution and showed that the decomposition products

included acetaldehyde, ammonia and carbon dioxide and,

depending on pH, pyruvate. Le Cloirec and Martin (1985) have

provided additional evidence to support the general mechanisms

already proposed. A summary of the general reaction

mechanisms leading to the formation of nitriles and aldehydes

(Figure 4) has been provided by Le Cloirec and Martin (1985).

In the reactions leading to the formation of aldehydes

and nitriles, imine-type compounds are known as essential

intermediates but have not been isolated in the chlorination

of individual amino acids. However, in the case of a peptide,

DL-alanyl-DL-leucine, an imine-type compound has been isolated

and characterized (Pereira et al., 1973). Pereira et al.

(1973) also showed that nitriles are major and aldehydes are

minor products of amino acid chlorination reactions. Tan et

al. (1987a) studied the kinetics of the reactions of aqueous

















COOH
I -HC1 -HC1
R.CH -- > R.CH = NCl > R.C m N
I -C02
NC12 Nitrile


HOC1


COOH COOH
I HOC1 I -CO2 HOC1
R.CH -> R.CH CH = NH
I I -HC1
NH2 NHC1


OH
I -NH2Cl
R.CH -- > RCHO
NHC Aldehyde
NHC1 Aldehyde


Figure 4. Reactions of chlorine with amino acids leading to
the formation of nitriles and aldehydes (Adapted
from Le Cloirec and Martin, 1985)










chlorine and several amino acids in different buffers and

observed that all the amino acids were reactive towards

chlorine over a wide pH range. The type of buffer system was

found markedly to affect the reaction rates. These workers

did not identify any of the reaction products but confirmed

previous reports (Isaac and Morris, 1983a) that the reaction

rates of the sulfur-containing amino acids, cysteine and

methionine, as well as tryptophan were too rapid to be

monitored by the iodometric method of titration. Jancangelo

and Olivieri (1985) have also reported that the sulfur-

containing amino acids and the heterocyclic amino acid,

tryptophan, are the most reactive with monochloramine.

The specific functional groups of the amino acids have

been shown to affect the type of products likely to be formed

during chlorination. In the case of the sulfur amino acid,

cysteine, Pereira and colleagues (1973) have shown that the

major products of chlorination were cystine and cysteic acid

while cystine produced only cysteic acid as the end product.

Amino acids with phenyl or heterocyclic groups readily form

chlorinated analogs if the rings are sufficiently activated.

For example tyrosine and tryptophan have been shown to form

the corresponding mono- and dichloro derivatives upon aqueous

chlorination (Trehy et al., 1986; Burleson et al., 1980;

Pereira et al., 1973). Aqueous chlorination of tyrosine has

been reported to yield mono- and dichlorobenzeneacetonitrile

and mono- and dichlorobenzeneacetaldehyde (Trehy et al., 1986;








41

Burleson et al., 1980; Pereira et al., 1973). Major

chlorination products of phenylalanine were previously

reported to be phenylacetonitrile (95%) and phenylacetaldehyde

(5%) but some chlorinated derivatives have been identified as

well (Horth et al., 1987).

In the case of tryptophan, Kirk and Mitchell (1980)

observed the formation of a dark-colored precipitate when

reacted with different concentrations of aqueous chlorine.

Although the reaction products were not fully identified, the

authors isolated and partially characterized a monochlorinated

product. Using structurally similar indole compounds, they

showed that the intact heterocyclic ring was very important

in color formation. Such color changes have been used

colorimetrically to determine chlorine dioxide concentrations

in water upon reaction with tyrosine (Hodgen and Ingols,

1954). Tryptophan has also been reported to produce oxidized

indole derivatives as well as ether extractable chlorinated

analogs (Owusu-Yaw et al., 1989; Trehy et al., 1986).

Nitriles and aldehydes are also produced as a result of

aqueous chlorination of tryptophan (Owusu-Yaw et al., 1989;

Horth et al., 1987).

The products of the reactions of chlorine with peptides

and proteins have been reported to be mainly N-chloro or

N,N-dichloro derivatives (Pereira et al., 1973, Rydon and

Smith, 1952). Rydon and Smith (1952) showed that peptides

and proteins reacted with chlorine gas to form








42

N-chloropeptides which could be detected visually on paper

chromatograms using a starch-potassium iodide spray.

According to the authors, color was developed from the

liberation of chlorine from the N-chloropeptide bond which

reacted with the starch iodide-indicator. Pereira and

coworkers (1973) investigated the reactions of excess chlorine

with the four dipeptides, glycyl-L-alanine, glycyl-DL-

phenylalanine, DL-alanine-DL-leucine and L-valyl-L-valine and

showed that the corresponding N,N-dichloro-dipeptides were the

major reaction products. From their point of view, the amide

nitrogen of the peptide was not attacked, and the peptide bond

itself was relatively resistant to chlorination. These

authors were the first to show that N-chloroimines were

produced by the dehydrochlorination of the N,N-dichloro-

dipeptide DL-alanyl-DL-leucine with ethanolic sodium hydroxide

at ambient temperature.

Glycylglycine has been used as a model for the formation

of N,N-dichloroglycylglycine and N-chloroglycylglycine (Isaac

and Morris, 1985; Qualls and Johnson, 1985; Margerum et al.,

1979). These studies show that the rate of chlorination of

the dipeptide is faster than that of several amino acids and

ammonia and that the peptide will be preferentially

chlorinated over ammonia or the amino acids glycine, alanine,

serine and glutamate in water. Other peptides such as

phenylalanylglycine and glycylglycylalanine have been reported

to produce stable chloramine residuals (Horth et al., 1989).








43

From all the studies cited, it becomes clear that the fastest

initial reactions of peptides with chlorine occur at the

nucleophilic terminal amino groups or the reactive side groups

of amino acids. For example when flour was chlorinated with

excess chlorine gas, the levels of the amino acids,

methionine, threonine, cysteine and histidine in the proteins

decreased by approximately, 68, 41, 32 and 17%, respectively

(Ewart, 1968).

Tan et al. (1987a) studied the kinetics of the reactions

of aqueous chlorine and three peptides and observed that the

rates of reaction of L-glycyl-L-tryptophan and L-tryptophyl-

glycine were too rapid to be measured iodometrically. The

authors suggested that the rapid reaction rates were due

mainly to the reaction of chlorine with the heterocyclic ring

of tryptophan and not with the amino or carboxyl groups. In

the case of aspartame, even though the component amino acids,

aspartate and phenylalanine, were individually reactive

towards chlorine, the dipeptide was not. The authors did not

provide any conclusive evidence as to whether any of the

reaction products were chloramines. Apart from the formation

of N-chloro compounds, Tsen and Kulp (1971) have postulated

that amino acid side chains of proteins are progressively

cleaved by increasing concentrations of chlorine. They

observed that the amount of water-extractable proteins in

chlorinated flour increased as the chlorine concentration was

increased. Chlorine has also been shown to be readily








44

incorporated into proteins from different sources including

shrimp, meat and poultry (Johnston et al, 1983; Cunningham and

Lawrence, 1977); egg albumin, gelatin, casein, and bovine

serum albumin (Baker, 1947; Wright, 1936 and 1926; Milroy,

1916); and flour proteins (Wei et al., 1984).



Reactions of Chlorine Dioxide with Amino Acids,
Peptides and Proteins

Chlorine dioxide has been proposed as an alternate to

chlorine for disinfection purposes and it is therefore

important to understand the kind of reactions and byproducts

likely to be formed with its use, especially proteinaceous

material. The reactions of chlorine dioxide under water

treatment conditions are mainly oxidative and as such, it has

been used for the control of phenolic tastes and odors in

natural waters (White, 1972). Several reports are available

in the literature on the reactions of chlorine dioxide with

amino acids under different conditions, but as pointed by

Masschelein (1979), the conditions of the reaction are very

important.

In earlier investigations, Schmidt and Braunsdorf (1922)

observed that most amino acids were not reactive towards

chlorine dioxide and that the least reactive were glycine,

leucine, serine, alanine, phenylalanine, glutamate,

asparagine, valine and hydroxyproline. The authors found

cystine to be very reactive. Bernade et al. (1967) further








45

investigated the reaction of chlorine dioxide with several

amino acids and observed that after a contact time of 30

minutes, asparagine, histidine, phenylalanine, arginine,

proline and leucine were unreactive. Contrary to the above

findings, Kennaugh (1957) noted that when a mixture containing

tryptophan, tyrosine, proline, threonine, methionine, valine

and lysine were mixed with diaphanol, a solution of 50% acetic

acid saturated with chlorine dioxide, the aromatic amino acids

disappeared after 18 hours. Also, most of the amino acids

were found to react after 3 days and phenylalanine did not

produce any ninhydrin positive material. Noss et al. (1983)

studied the reactivity of equimolar concentrations (2.5 x

104 M) of twenty-two amino acids and four nucleic acids with

chlorine dioxide at neutral pH and found only seven (cysteine,

cystine, proline, histidine, hydroxyproline, tyrosine and

tryptophan) to be reactive. Additional evidence in support

of these studies was provided by Tan et al. (1987b) who showed

that six of twenty-one amino acids were reactive with chlorine

dioxide. Of the six, the reactions of cysteine, tryptophan

and tyrosine were too rapid to be measured by the iodometric

titration method.

From the above and other studies, it is clear that the

reactivity of chlorine dioxide with amino acids and proteins

depends on pH, temperature and contact time, with higher

temperatures and contact times and lower pH favoring the

reaction. Also, specific functional groups of the amino acids








46

and proteins such as phenyl or sulfhydryl and the relative

molar concentrations of reagent and substrate affect the

reactions. For example, glycine and phenylalanine were inert

toward chlorine dioxide at room temperature when equimolar

concentrations of reagent and substrate were used (Noss et

al., 1983). However, the reactions were found to proceed when

excess molar concentrations of chlorine dioxide were used

(Taymaz et al., 1979).

The sulfur-containing amino acids and those with phenyl

or heterocyclic rings have been reported to be the most

reactive with chlorine dioxide (Masschelein, 1979). Of the

amino acids containing sulfur, cystine has been reported to

produce cysteic acid with a cystine disulfoxide intermediate.

The reaction is said to involve the breakage of the sulfur-

sulfur bond of cystine followed by further oxidation

(Masschelein, 1979). Under similar conditions, methionine is

oxidized to the sulfoxide and then to the sulfone. Other

organic sulfides and mercaptans are also known to undergo

similar oxidative reactions with chlorine dioxide.

Chlorine dioxide oxidation of tyrosine is similar to

those of phenols which are especially susceptible to

oxidation. When pure chlorine dioxide is used, no chlorinated

products are obtained (Masschelein, 1979). With chlorine

dioxide, tyrosine has been reported to produce pink

dopaquinone and red dopachrome. This reaction has been used

as a basis for the colorimetric determination of chlorine








47

dioxide concentrations in water (Hodgen and Ingols, 1954).

When excess chlorine dioxide, 3 moles per mole of tyrosine,

is used for an extended period of time, other unidentified

colored products are formed as well. Reaction of 2 molars

excess of chlorine dioxide with phenylalanine at elevated

temperatures has been reported to produce phenylacetic acid,

phenylacetaldehyde, benzoic acid, benzaldehyde and mandelic

acid (Taymaz et al., 1979). No chlorinated products were

identified from the above reaction. When the chlorine dioxide

concentration was increased to 10 excess molar under similar

conditions, no phenylalanine or above-mentioned products were

detected after 6 days. From the foregoing, it is possible

that further oxidation of the initial reaction products takes

place with excess chlorine dioxide thus giving off carbon

dioxide and causing the disappearance of substrate.

The heterocyclic group of compounds is also very

susceptible to chlorine dioxide oxidation leading to the

formation of complex oxidized products. Chlorine dioxide

oxidation of tryptophan produced indoxyl, isatin, indigo red

as well as unidentified polymerization products (Fujii and

Ukita, 1957). Structurally similar indole derivatives such

as indole itself, 3-methyltryptophan, 3-indolelactic acid and

3-methylindole have also been shown to be readily oxidized by

chlorine dioxide, sometimes leading to color and precipitate

formation (Sen et al., 1989; Lin and Carlson, 1984). Indole

is oxidized to oxindole and isatin while 3-methylindole








48

produces 3-methyloxindole and o-formamidoacetophenone (Lin and

Carlson, 1984).

Unlike amino acids, the reaction of chlorine dioxide with

proteins and peptides has not received much attention in model

systems. A review of the available literature has been

documented by Fukayama et al. (1986). Tan et al. (1987b)

studied the kinetics and mutagenicity of three peptides with

chlorine dioxide at room temperature using equimolar

concentrations of reagent and substrate. They observed that

the reactions of two dipeptides L-tryptophyl-glycine and

L-glycyl-L-tryptophan were too fast to be determined by

iodometric titration. They attributed the very fast reaction

rates to the presence of the tryptophan moieties which have

previously been shown to react rapidly with chlorine dioxide

(Fujii and Ukita, 1957). The third dipeptide, aspartame, was

found to be relatively unreactive since less than 10% of the

initial reagent was lost after 2 hours of reaction. The

authors explained that the inertness of this dipeptide with

chlorine dioxide could be attributed to the component amino

acids, aspartate and phenylalanine, which were individually

unreactive towards chlorine dioxide.

Chlorine gas and chlorine dioxide have traditionally been

used to treat flour to improve baking properties and dough

quality. Meredith et al. (1956) treated wheat flour with high

levels of chlorine dioxide and did not find any chlorinated

derivatives of amino acids and proteins. They, however, found








49

that the levels of the amino acids, cysteine and tryptophan

were decreased by 25 and 8%, respectively. Moran and

coworkers (1953) reported that methionine was oxidized to the

corresponding sulfoxides and sulfones and cysteine to cysteic

acid in flour treated with chlorine dioxide. Ewart (1968)

exposed solutions of wheat proteins to chlorine dioxide and

observed the formation of a pink-yellow color due to the

oxidation of tryptophan and tyrosine. Under similar

conditions there was an increase in chlorine dioxide-induced

solubility of flour proteins. In the treatment of wool,

chlorine dioxide has been reported to cause a dissolution of

up to 40% of the proteinaceous matter leading to losses in

weight. These weight losses have been attributed to the

reactivity of amino acid components of the wool with chlorine

dioxide (Masschelein, 1979). Tyrosine, methionine, glutamate

and tryptophan were the most reactive. The reactivity of

tryptophan with chlorine dioxide was found to be greater than

that of tyrosine and usually led to the formation of a reddish

coloration (Masschelein, 1979).



Toxicological Assessment

Evidence already presented suggests that both chlorine

and chlorine dioxide have the capacity to react with a wide

range of organic compounds under different treatment

conditions. Many organic substances in food and potable water

are capable of reacting with chlorine to produce both








50

halogenated and oxidized products. Chlorine dioxide on the

other hand produces mainly oxidized reaction products. Even

the most gentle chlorination treatments are likely to produce

chlorinated and non-chlorinated organic. Chlorination of

organic compounds has been shown to increase their lipophilic

nature and thereby increase their toxicity and/or their

ability to bioaccumulate (Kopperman et al., 1978). This

section will focus on the toxicology of chlorine and compounds

arising from the use of chlorine and chlorine dioxide

disinfection and their possible health hazards.

Overview

Chlorine has been used as the major disinfectant for

water supplies for many years and has been a major factor in

reducing the transmission and spread of numerous waterborne

diseases and infections. Concerns about the use of chlorine

for water treatment came to light when it was discovered that

THMs are produced as a result of water chlorination practices

(Bellar et al., 1974; Rook, 1974). These concerns gained

additional impetus when the National Cancer Institute reported

that chloroform, a major byproduct of chlorination, is

carcinogenic in two laboratory animals (NCI, 1976). Because

of this discovery, the United States Environmental Protection

Agency amended the National Interim Primary Drinking

Regulation which set up guidelines for the control of THM

concentrations in drinking water. A maximum contaminant level

of 0.10 mg/L was set for total THMs in finished drinking water








51

in the United States (Symons et al., 1981). In order to

comply with these guidelines, several researchers in the area

have devoted their efforts to the search for alternate

disinfectants which will produce very little or no THM in

finished drinking water (Condie, 1986; Lykins and Griese,

1986; Symons et al., 1978).

Chlorine dioxide was selected as the alternate oxidant

and disinfectant for drinking water because in most aspects

it is similar or superior to chlorine and it produced very

little or no THM (Knocke et al., 1987; Aieta and Berg, 1986;

Lykins and Griese, 1986; Lauer et al., 1986; Aieta et al.,

1980). Also, it has been in use for many years in several

European countries without any reported adverse effects

(Symons et al., 1978). According to Symons and coworkers, a

1977 survey showed that about 103 facilities were using or had

used chlorine dioxide in the United States. The suggested use

of chlorine dioxide was welcomed by several researchers and

numerous reviews are available regarding its action,

especially for THM control. One of the main disadvantages of

chlorine dioxide is its additional cost to the consumer which

is estimated at about three times that of chlorine. However,

a report by Lykins and Griese (1986) in Evansville, Indiana,

shows that apart from being suitable for THM control, chlorine

dioxide is also cost effective (about $1.77/resident/year).

On the other hand, it becomes clear that the toxicological

problems and possible health hazards associated with chlorine








52

dioxide and its disinfection byproducts are not immediately

known (Condie, 1986).

Unfortunately, THMs are not the only byproducts of

disinfection which pose possible health hazards in drinking

water. Other chlorinated and non-chlorinated organic

compounds, both volatile and non-volatile, have been reported

to be produced during water chlorination (Loper et al., 1985;

Coleman et al., 1983; Cheh et al., 1980; Christman, 1980).

Organic compounds that have been identified in drinking water

are mostly volatile and represent only about 5-10% of the

total organic compounds in drinking water (NAS, 1977). Of the

250 or more compounds previously identified in drinking water,

about 10% have been classified as being carcinogenic and a

good number remain to be assessed for genotoxicity (Kraybill,

1978). According to the latest survey, more than 1,100

organic compounds have been identified in drinking water, most

of which are at or below the gg/L level (Lucas, 1985).

Another group of compounds, haloacetonitriles (HANs),

has been identified in drinking water which has compounded

the problem (Daniel et al., 1986; Bierber and Trehy, 1983;

Oliver, 1983; Trehy and Bierber, 1981). Proteinaceous

material, especially amino acids, as well as fulvic and humic

substances have been implicated and confirmed precursors of

both THMs and HANs upon reaction with chlorine (Bierber and

Trehy, 1983; Oliver, 1983; Rook, 1980). A third class of

compounds which has received very little attention is the








53

relatively high molecular weight compounds which are not

easily amenable to gas chromatographic analysis. This class

of compounds is mostly non-volatile and several have been

shown to be mutagenic in short-term assays (Cheh et al.,

1980). The difficulty encountered in their isolation and

identification is that they are mostly non-volatile and

difficult to identify using current analytical techniques

(Meier et al., 1983). Some are polymerization products

produced from the condensation of several individual compounds

which makes identification more difficult even with up-to-date

instrumentation. Due to these analytical limitations, the

toxicology and possible health risks posed by this class of

compounds still remain unknown.

Toxicology of chlorine and chlorine dioxide

Chlorine forms hypochlorous acid (HOC1) which dissociates

to form hypochlorite ion (OC1) in solution. Under alkaline

conditions, chlorine dioxide disproportionate to form

chlorite (ClO2") and chlorate (Cl03). Most of the current

literature on the toxicology of chlorine dioxide and chlorine

deals with the undesirable byproducts arising from the

reactions of these disinfectants with organic materials in

water. As a result, very little information is available on

the toxicology of the disinfectants themselves or the

endproducts of their decomposition such as chlorite, and

hypochlorite. This section reviews the available work

concerning the toxicology of chlorine and chlorine dioxide,


1








54

as well as their decomposition byproducts, hypochlorite,

chlorate and chlorite.

Chlorine, hypochlorous acid and hypochlorite

Information dealing with the toxicity of chlorine is

somewhat conflicting. Blabaum and Nichols (1956) reported

that no gross abnormalities occurred in mice fed chlorinated

drinking water for 30 or 50 days. Also, a seven generation

study in rats administered chlorine in their drinking water

did not result in any observable adverse toxic or teratogenic

effects. Other reports on the toxicity of chlorine or

chlorinated water indicate that toxicity, if any, is not due

to chlorine itself but probably due to its reaction with

precursors in the stomach. Chang et al. (1981) have reported

that administration of chlorinated water led to the

development of fatty liver in rats with a better than two-fold

increase in liver triglycerides. The acyl groups of

triacylglycerols and phospholipids were also affected. Such

effects were completely reversible when dosing was

discontinued and the animals recovered fully after 10 days.

The induction of fatty liver may be due to the formation of

chloroform with precursor substances in the body since

chloroform has been shown to give rise to this condition

(Cornish, 1980).

No toxic effects were observed in rats and guinea pigs

fed up to 400 ppm of sodium hypochlorite in milk and water

(Cunningham, 1980). At low dose levels, chlorine was actually








55

shown to act as a broad spectrum antibiotic and stimulated the

growth rates of the rats and guinea pigs. However, at very

high doses (2000 ppm), Cunningham observed significant

increases in liver weights probably due to the formation of

fatty liver described by Chang and coworkers (1981). In

earlier studies, Cunningham and Lawrence (1977) fed flour,

chlorinated with 2,000 and 10,000 ppm chlorine, to rats and

observed decreased growth rates and increased liver and kidney

weights. The authors reported similar effects in rats fed

chlorinated flour lipids and proteins (Cunningham and

Lawrence, 1978).

The tumor initiating and/or promoting activities of

chlorine have also been studied but the results are still

inconclusive. Hayatsu and coworkers (1971) observed that

sodium hypochlorite acts as a cocarcinogen in mice and

increased tumor yields when applied in conjunction with a

sub-threshold dose of 4-nitroquinoline-l-oxide. Contrary to

these observations, Pfeiffer (1978) reported that sodium

hypochlorite decreased the carcinogenic activity of

benzo(a)pyrene in mice.

Chlorine has been reported to induce chromosomal

aberrations in mammalian cells in vitro (Mickey and Holden,

1971), and sodium hypochlorite, which is commonly used in

water disinfection, has been shown to inhibit preferentially

the growth of DNA polymerase-deficient strains of E. coli

(Rosenkranz, 1973). These mutagenic events are due to the








56

reaction of chlorine with the DNA of living matter

(Rosenkranz, 1973). Sodium hypochlorite and not hypochlorous

acid (Thomas et al., 1987) has also been reported to be weakly

mutagenic for Salmonella typhimurium in the Ames Salmonella/

microsome assay (Wlodkowski and Rosenkranz, 1975). It

resulted in a base-pair substitution mutation in Salmonella

presumably due to a reaction with purine and pyrimidine bases

which has previously been demonstrated to occur in vitro

(Hoyano et al., 1973). Meier et al. (1983) have demonstrated

that sodium hypochlorite, and not hypochlorous acid, was

capable of inducing sperm-head abnormalities in mice when fed

various chlorine compounds by gavage. From the evidence

provided, it is not clear whether the mutagenic and/or

carcinogenic effects are due to chlorine itself, or due to the

action of some reaction byproducts. It has been reported that

chloroform, di- and trichloroacetic acid as well as di- and

trichloroacetonitrile were formed in vivo in rats following

oral administration of sodium hypochlorite (Mink et al.,

1983). Some of these byproducts may play a role in the

observed genotoxic effects and will be discussed in the latter

section.

Recent studies by Revis and coworkers (1985) have shown

that chlorine residuals similar to those found in drinking

water increased heart rate, heart fibrous tissue and plasma

cholesterol levels in rabbits and pigeons by interacting with

a diet marginal in calcium. They suggested a relationship








57

between water chlorination and hypothyroidism which they

assumed to be caused by the production of chlorinated

products. In earlier studies Revis et al. (1983) also

observed signs of myocardial hypertrophy and arteriosclerosis

in rabbits and pigeons fed chlorine as sodium hypochlorite in

their drinking water. These changes were reversed when

calcium intakes in the diets were restored to normal.

Chlorine, as chloramine, has received very little

attention in terms of its toxicological potential. It has

been reported to be weakly mutagenic in bacteria (Shih and

Lederberg, 1976) but does not produce sperm-head abnormalities

in mice (Meier et al., 1985). In a very recent study, both

mono- and dichloramines and their derivatives were shown to

be mutagenic in the Ames assay (Thomas et al., 1987). The

authors showed that S. typhimurium strain TA100 was more

sensitive to chloramine mutagenesis and that the lipophilic

dichloramines were more active than the corresponding

monochloramines. Chloramine as the synthetic disinfectant,

chloramine-T, has also been reported to cause sister chromatid

exchanges in cultured Chinese hamster ovary cells in vitro

(Weitberg, 1987). Chloramine has also been evaluated for

teratogenicity in developing rat fetuses but showed little or

no adverse effects at levels up to 100 ppm (Abdel-Rahman et

al., 1982).










Chlorine dioxide, chlorate and chlorite

The toxicology of chlorine dioxide as well its inorganic

byproducts, chlorate (C103) and chlorite (CIO'2), has been

studied extensively because of their observed effects on the

hematopoietic system. All three compounds are usually

evaluated together because both chlorine dioxide and chlorate

can rapidly be converted to chlorite, and therefore, some

aspects of their toxicology are expected to be similar

(Condie, 1986). Like chlorine, considerable disagreement

exists in the literature concerning the toxicity of chlorine

dioxide and its reaction products. Chronic administration of

chlorine dioxide and chlorate to rats in their drinking water

did not produce any observed toxic effects at doses up to 10

mg/L. However, higher mortality rates were observed in rats

administered 100 mg/L for 2 years (Haag, 1949).

Chlorite has been shown to be a potent stressor of blood

and produces methemoglobinemia and hemolytic anemia both in

vivo and in vitro (Hefferman et al., 1979a; 1979b). The dose

required to produce hemolytic anemia is much lower than that

required for methemoglobinemia. It has also been reported to

cause hemolytic anemia, increase osmotic fragility and

glucose-6-phosphate dehydrogenase activity in erythrocytes,

and reduce growth and conception rates in mice (Moore and

Calabrese, 1980a; 1980b; Moore et al., 1980). Further studies

have established that all three compounds produce the same

hemolytic effects in test animals as well as decrease








59

glutathione levels in rats (Abdel-Rahman et al., 1984;

Abdel-Rahman et al., 1979; Couri and Abdel-Rahman, 1979).

Chlorine dioxide induced reduction in glutathione levels

tended to disappear with continued exposure (Couri and

Abdel-Rahman, 1979). The mechanism of the hemolytic activity

of chlorine dioxide and its metabolites has been attributed

to the production of hydrogen peroxide which, like other

strong oxidants, produces this effect (Bull, 1983).

Chlorite and chlorate, but not chlorine dioxide have been

reported to produce hematological effects in nonhuman primates

(Bercz et al., 1982). Chlorine dioxide, on the other hand,

caused a decrease in serum thyroxin levels in monkeys. The

hypothyroid effect was unique to chlorine dioxide since its

endproducts, chlorite and chlorate, did not show this effect

even at higher doses (Bercz et al., 1982). Since chlorine

dioxide is a very reactive oxidant, it is not known whether

the decrease in thyroxin levels is due solely to this

disinfectant or due to some primary oxidation products formed

with precursor molecules in the gastrointestinal tract. Other

studies showed no significant decreases in serum thyroxin

levels in adult rats fed up to 100 mg/L chlorine dioxide

(Condie, 1986). However, administration of similar doses to

pregnant and lactating females resulted in decreased serum

thyroxin concentrations in rat pups (Condie and Bercz, 1985;

Orme et al., 1985).








60

Reproductive and teratogenic studies have also been

performed in laboratory animals. Exposure of female mice to

chlorine dioxide in their drinking water from breeding to

weaning resulted in reduced growth rates and body weights

(Moore et al., 1980), and at much higher concentrations,

increased stillbirths and fetal reabsorptions (Couri et al.,

1982). Contrary to these findings, Suh and coworkers (1983)

did not observe any maternal toxicity or gross changes in

fetal weight, litter size or skeletal anomalies in the

development of rats fed chlorine dioxide, chlorite and

chlorate in their drinking water. Further studies did not

show any measurable adverse reproductive or teratogenic

effects in rats fed sodium chlorite in drinking water (Carlton

and Smith, 1985). Carlton and Smith (1985) showed that

chronic exposure of rats to 100 mg/L sodium chlorite increased

the number of abnormal sperms and to some extent, impaired

sperm motility. This finding is in contrast to that of Meier

and coworkers (1985) who did not detect any sperm-head

abnormalities in mice dosed up to 400 mg/L chlorine dioxide,

chlorate and chlorite.

The effects of chlorine dioxide on the neurobehavioral

development of rats were investigated by Taylor and Pfohl

(1985). They observed that administration of chlorine dioxide

both postnatal and prenatally resulted in behavioral defects

and depressed brain growth. These effects were consistent

with depressed thyroid function. The authors also observed








61

delayed eye opening, decreased body and brain weights and

activities, and delayed locomotor activity. Chlorine dioxide,

and not chlorite, has been implicated as a hyper-

cholesterolemic agent which increased serum cholesterol levels

by interacting with a diet marginal in calcium (Revis et al.,

1985). No definitive conclusions have been made on the

genotoxicity of chlorine dioxide and its metabolites. Suh et

al. (1984) reported that chlorine dioxide depressed DNA

synthesis in the testis and intestinal epithelium of rats

dosed 100 mg/L chlorine dioxide in their drinking water.

Meier et al. (1985) tested the mutagenic activity of chlorine

dioxide, chlorate and chlorite in mice using chromosomal

aberrations, micronucleus and sperm-head abnormalities assays

as endpoints. No mutagenic activity was detected in any of

the disinfectants 5 days after exposure. The tumor initiating

and promoting activities of chlorate and chlorite have also

been investigated. Both chemicals did not demonstrate any

significant carcinogenic activity in Sencar mice (Kurokawa et

al., 1985; 1984).

With regard to human studies, oral administration of

chlorine dioxide, chlorate and chlorite to healthy adult male

volunteers did not produce any detrimental clinical effects

(Lubbers et al., 1985). The authors employed an exposure time

of 12 weeks with the volunteers receiving a total of 5 mg/L

chlorine dioxide or its endproducts. Again, no biochemical

or physiological effects were observed in the volunteers.








62

Caution must be taken in interpreting these findings since a

lack of effect does not necessarily imply safety. It is not

known whether long-term low-dose exposure to any of these

chemicals will induce any severe adverse physiological

effects.

Mutagenesis and Carcinogenesis

Different classes of compounds have been identified in

drinking water, some of which may be produced from

chlorination reactions. A 1976 USEPA survey showed that of

289 compounds identified in drinking water, approximately 38%

were halogenated. Another survey of five major cities in the

United States showed that approximately 50% of the volatile

hydrophobic compounds in drinking water were halogenated

(USEPA, 1975). The number could be considerably higher if

proper analytical methods for detection and quantification

were available. Also, the source and distribution of these

chemicals differ. This section will briefly review the

mutagenic and carcinogenic properties of those compounds that

are formed during water chlorination, such as THMs and HANs,

and those of precursor substances such as amino acids and

humic substances.

Drinking water

The carcinogenic properties of drinking water were first

reported by Marx (1974). Since then numerous studies have

been conducted on the mutagenic and carcinogenic properties

of drinking water in different municipalities in the United








63

States and elsewhere and, in all cases, researchers have shown

that mutagens are produced as a result of water chlorination.

Drinking water in the United States and Canada (Basu et al.,

1985; Nestmann et al., 1983; Cheh et al., 1980; Loper, 1980),

South Africa (Grabow et al., 1980), the Netherlands (Kool et

al., 1985), United Kingdom (Horth et al., 1987; Fielding and

Horth, 1986), Finland (Kronberg et al., 1988), Israel

(Guttman-Bass et al., 1987) and many other countries have been

tested and reported to be mutagenic in several short-term

assays. In most of the studies, replacement of chlorine with

chlorine dioxide leads to a reduction in the observed

mutagenic activities (Guttman-Bass et al., 1987). According

to the findings of Kool et al. (1985), chlorine treatment of

drinking water generally increases the mutagenic activity in

S. typhimurium TA98 and TA100. Also chlorine dioxide is

capable of increasing the mutagenic activity in certain cases

with the overall increase being less than that of chlorine.

The main drawback to these findings is that mutagenicity

is usually detected from a mixture of compounds of different

classes and origin, and even though compounds such as

chloroform, bromoform and tetrachloroethylene are known

carcinogens, their overall contribution to the observed

genotoxicity is not usually known. Being highly volatile,

some of these chemicals may be lost during sample

concentration procedures prior to genotoxicity evaluation.

The third problem is that some of these compounds are poorly








64

characterized non-volatiles which cannot be easily isolated

and identified, therefore, their overall contribution to the

carcinogenic and mutagenic effects remain unknown. None of

the studies in the literature has accounted for 100% of the

mutagenic activity observed in drinking water extracts.

Kronberg et al. (1988) have recently identified a very strong

mutagen,3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone

(MX) and its isomer, (E)-2-chloro-3-(dichloromethyl)-4-

oxobutenoic acid (EMX) in acidified chlorinated extracts of

humic acid and drinking water. According to the authors, MX

accounted for 50-100% of the mutagenic activity of extracts

of chlorinated humic water and for 20-50% of the activity in

drinking water extracts while EMX accounted for only about 2%

of the total mutagenic activities using the Ames Salmonella

assay. MX has also been detected in drinking water from three

locations in the U.S.A. and in the chlorination of humic acid

(Meier et al., 1987). According to the authors, MX appears

to account for a significant proportion of the mutagenic

activity observed in these samples.

Presented in Table 1 is a of some of the compounds

identified in drinking water which includes known or suspected

carcinogens. Volatile compounds have been defined as those

that are purgeable from drinking water at room temperature

with an inert gas such as helium (Tardiff et al., 1978).

These compounds represent 5-10% of the total organic in

drinking water and the remaining 90-95% have not been fully












Table 1. Some halogenated compounds formed in drinking water
as a result of water chlorinationa


Chloroformb

Bromodichloromethaneb

Chlorodibromomethaneb

Bromoformb

Dichloromethane

Bis (2-chloroethyl) etherb

Trichloroethyleneb

Tetrachloroethyleneb

1,1,1-Trichloroethane


Trichloroacetic acid

Trichloroacetaldehyde

Chlorophenols

Alpha-chloroketones

Chlorobenzenes

Chlorinated acetonitrilesb

Chlorinated aromatic acids

Chlorinated purines

Chlorinated pyrimidines


aAdapted from Kraybill (1978)

bKnown or suspected carcinogens










identified, or are unknown (NAS, 1980). Not included are

those high molecular condensation and/or polymerization

products which are not easily amenable to gas chromatographic

analyses even with sample derivatization.

Trihalomethanes and haloacetonitriles

THM were the first class of compounds which aroused

interest in chlorination byproduct formation (Rook, 1974;

Bellar et al., 1974). Most of the purgeable organic are

formed in drinking water at a level of 1 ppb or less.

However, chloroform is usually present in both treated and

untreated water at much higher concentrations of up to 100

ppb on the average (Tardiff et al., 1978; Bellar et al.,

1974). Some of the most frequently encountered chlorinated

compounds in groundwater include trichloroethylene, vinyl

chloride, tetrachloroethylene, 1,1-dichloroethylene, carbon

tetrachloride, 1,1,1-trichloroethane and 1,1-dichlorobenzene

(Cotruvo and Vogt, 1985). Chloroform, the major THM, has been

shown to be carcinogenic and to increase the incidence of

hepatocellular carcinomas and kidney epithelial tumors when

administered to rats and mice (NCI, 1976). However, recent

studies indicate otherwise (Pereira et al., 1985). In 1983,

Pereira showed that THM, including chloroform, possess very

little or no tumor initiating activity and that their main

action in tumorigenesis is promotion. Pereira et al. showed

that chloroform, apart from low level DNA binding, did not

possess any tumor initiating property. In a fairly recent








67
study, Pereira and coworkers (1985) administered 1,800 ppm

chloroform to Swiss mice, pretreated with ethylnitrosourea,

and reported that chloroform inhibited hepatocarcinomas in the

mice. According to the authors, administration of chloroform

in the drinking water of Swiss mice inhibited both spontaneous

and induced hepatocarcinogenesis caused by ethylnitrosourea.

These results are in sharp contrast to the NCI study in which

chloroform was shown to be a hepatocarcinogen. In the NCI

study, chloroform was administered in corn oil and Pereira

and coworkers point out the possibility of a synergistic

interaction between the vehicle and chloroform leading to the

previously observed hepatocarcinogenicity.

Chloroform has been reported to be non-mutagenic in

bacteria assays including Salmonella and E. coli (Simmon et

al., 1977; Uehleke et al., 1976). Other THMs such as

bromoform, bromodichloromethane and dibromochloromethane were

only weakly mutagenic in Salmonella without metabolic

activation (Simmon et al., 1977).

The second major class of volatile organic, HANs, have

received considerable attention since they were discovered

(Oliver, 1983; Bierber and Trehy, 1983). Bierber and Trehy

(1983) reported that HANs easily escape detection because they

undergo gradual hydrolysis to form nonvolatile products.

Dichloroacetonitrile, the major HAN, is usually detected in

drinking water at about 10% that of the average THM

concentration (Oliver, 1983). Haloacetonitriles (HAN) have








68

been reported to pose some possible health problems to humans

because of their potential mutagenic and carcinogenic

properties in test animals (Meier et al., 1983). They have

been detected in the stomach contents of rats dosed with

aqueous chlorine solutions and the possibility exists for the

same effects to be observed in humans (Mink et al., 1983).

Bull and Robinson (1985) tested the mutagenic activity of the

five common HAN, chloroacetonitrile, dichloroacetonitrile,

trichloroacetonitrile, bromochloroacetonitrile, and

dibromoacetonitrile in the Ames Salmonella/microsome assay and

found that dichloroacetonitrile and bromochloroacetonitrile

were direct-acting mutagens which produced base-pair

substitution mutations in Salmonella. All five compounds

induced significant elevations in sister chromatid exchange

(SCE) frequencies in Chinese hamster ovary (CHO) cells. SCE

frequencies increased with increasing chlorine substitution

and was further enhanced when bromine was substituted for

chlorine. Three of the compounds were also capable of

producing tumors in Sencar mice. HAN have been shown to bind

covalently to liver and kidney DNA and cause DNA strand

breaks, which indicates potential carcinogenic ability

(Pereira et al., 1985). Trichloroacetonitrile was found to

be the most potent DNA strand breaker while the bromo-

derivatives were intermediate. Pereira and coworkers have

proposed a scheme for HAN metabolism which involves formation

of phosgene, a highly reactive electrophile, and cyanide.








69

The above findings were further confirmed by Daniel et

al. (1986) who reported that HANs were cytotoxic in human

cultured lymphoblast (CCRF-CEM) cells. They produced 13-60%

cell killing at a dose of 50 gmol and caused major DNA strand

breaks. It should be noted that in all of these studies,

chlorine dioxide has not been mentioned due to the fact that

its use in water treatment does not lead to THM formation.

Further studies are needed to determine the stability of HANs

in aqueous media and the possible health hazards associated

with their long-term low-dose exposure to humans.

Precursor substances

The final part of this review will focus on those

precursor substances which react with chlorine or chlorine

dioxide to form genotoxicants. Proteinaceous material and

humic and fulvic substances are known precursors of

haloacetonitrile and trihalomethane formation in natural

waters upon reaction with chlorine. Unfortunately, these

compounds are not the only ones of concern in drinking water

since other chlorinated and non-chlorinated byproducts are

formed as well. Even though chlorine dioxide produces very

little or no THM in water, it does not remove them from water.

Also, some of the oxidation byproducts of chlorine dioxide

reactions in water may also be genotoxic.

Amino acids, peptides and proteins. The genetic effects of

amino acids after chlorination were first reported by Sussmuth

(1982) who demonstrated that harmless nutrients such as amino








70

acids and yeast and meat extracts could be converted into

carcinogens and mutagens by chlorination. In this experiment,

80 mL of chlorine gas were bubbled through 250 mg of yeast

extract, meat extract or amino acids for 4 minutes and the

reaction products evaluated for genotoxic activity. Due to

the high chlorine dose, Sussmuth did not find the need to

concentrate the reaction products. Methionine was found to

be the most potent precursor of genotoxic activity both in the

Ames Salmonella assay and the rec assay in Bacillus subtilis.

Tyrosine, phenylalanine, cysteine and glycine were also

genotoxic. In the mutagenicity assay, all the above compounds

were direct-acting mutagens which did not require metabolic

activation. None of the reaction products were identified in

this study.

In another study, Horth et al. (1987) evaluated the

mutagenic activity of several amino acids in Salmonella after

reaction with sodium hypochlorite. With the exception of

glycine, mutagenic activity was detected in all the amino

acids treated with chlorine. Methionine, phenylalanine,

tyrosine and a group of heterocyclics were found to be the

most potent precursors of mutagenic activity. Of the

heterocyclics, tryptophan and proline were the most active.

The authors identified several compounds in their mutagenic

extracts. Except for dichloroacetonitrile which is a known

carcinogen, none of the other compounds identified possessed

mutagenic activity when tested separately. It was concluded








71

that the level of dichloroacetonitrile was too low and

unlikely to account for any significant portion of the

mutagenicity. In follow-up studies Horth and coworkers (1989)

tentatively identified several other compounds from

chlorination of tyrosine and phenylalanine. Using authentic

standards, the authors showed that dichloroacetonitrile,

produced from both amino acids, and benzylchloride and

2-phenyl-2,2-dichloroethenal, produced from phenylalanine,

were all mutagenic to S. typhimurium strain TA100.

Dichloroacetonitrile has also been shown to be produced from

the chlorination of aspartic acid, tyrosine and tryptophan

(Trehy et al., 1986). The overall contribution of the

unidentified compounds and those for which no authentic

standards were available still remain unknown.

In similar studies Rapson and coworkers (1985) tested

the mutagenicity of compounds produced by reaction of aqueous

chlorine and chlorine dioxide with tyrosine, at increasing

chlorine to tyrosine ratios. Tyrosine produced mutagenic

reaction products with the mutagenicity increasing with

chlorine dose. Mutagenicity peaked at 4 equivalents of

chlorine per mole of tyrosine. Substitution of chlorine with

chlorine dioxide led to a linear decline in mutagenic activity

from 2,100 revertants over background to 0 with 100% chlorine

dioxide. Similar results were obtained by Tan et al. (1987b)

who showed that replacement of chlorine with chlorine dioxide

on a molar basis decreased the mutagenic activity in








72

chlorinated tryptophan concentrates. However, in the case of

two dipeptides, L-tryptophylglycine and L-glycyl-L-tryptophan,

reaction with chlorine dioxide led to the formation of

mutagens while aqueous chlorine did not (Tan, 1986). In

contrast to the findings of Sussmuth (1982) and Horth et al.

(1987), Tan et al. (1987b) did not detect any mutagenic

activity after reaction of aqueous chlorine and methionine.

Chlorination of purines and pyrimidines as well as

nucleosides and nucleotides did not result in any mutagen

formation (Horth et al., 1987). It is evident that many

difficulties exist in the concentration, fractionation and

evaluation of individual components in drinking water or model

systems. In a majority of tests, a mixture of compounds and

not individual components are used due to analytical

difficulties encountered sample separation. Unless proper

separation and analytical methods are developed, it will be

difficult to pinpoint the individual chemical or group of

chemicals responsible for mutagenic activity in chlorination

reaction mixtures.

Fulvic and humic substances. Fulvic and humic acids are

naturally occurring metabolic endproducts of decaying

vegetation which account for the bulk of organic matter

present in surface waters. Chlorination of water containing

these substances has been reported to result in the formation

of THMs and other non-chlorinated, high molecular weight,

poorly characterized organic (Coleman et al., 1983; Rook,








73

1977), some of which are mutagens and carcinogens (Kronberg

et al., 1988; Meier et al., 1987, 1986 and 1983).

Chlorination of kraft wood pulp, catechol, lignin and

humic and fulvic acids have all been shown to give rise to

mutagen formation in the Ames Salmonella/microsome assay

(Kronberg et al., 1988; Meier et al., 1987; Holmbom et al.,

1984; Kringstad et al., 1983; Meier et al., 1983; Nazar and

Rapson, 1982; Nazar et al., 1981). The pH of the reaction is

important in mutagen formation especially when Amberlite XAD

resins are used to adsorb or concentrate the reaction

mixtures. According to Nazar and Rapson (1982), at elevated

pH of 7 or higher, mutagenic activity decays due to the

cleavage of organically bound chlorine by hydroxide ions.

These findings have been confirmed by Meier et al. (1983) who

observed a decrease in mutagenic activity with increasing pH

due in part to the alkali liability of the compounds.

Mutagenic activity of chlorinated model humic acids is similar

to that of drinking water which implies that these substances

account for a substantial portion of the mutagenic activity

of drinking water (Kronberg et al., 1988; Meier et al., 1987;

Coleman et al., 1984). Also most of the mutagenic activity

is due to non-volatile compounds since most of the volatile

organic are lost during sample concentration (Meier et al.,

1983). Coleman et al. (1984) identified more than 25

compounds from ether extracts of chlorinated humic acids and

confirmed their individual mutagenic activities using








74

authentic standard compounds. Kronberg and colleagues have

also identified and characterized a strong mutagen, MX, from

acidified extracts of chlorinated humic acid using HPLC

fractionation techniques. MX accounted for 50-100% of the

observed mutagenic activity while its isomer, EMX, accounted

for very little of the observed activity. In a separate

study, Meier et al. (1986) identified several compounds from

ether extracts of chlorinated humic acids, some of which are

known mutagens. Analysis of these extracts indicated that

they accounted for only 7% of the total activity in the

original solutions. They also identified MX and EMX which had

previously been identified (Holmbom et al., 1984). Further

studies by Meier et al. (1987) and Hemming et al. (1986)

showed that MX accounted for between 15-34% and 3-19%,

respectively, of the mutagenic activity in XAD concentrates

of drinking water. Both research groups observed that sample

acidification prior to XAD adsorption was essential for the

effective recovery of MX from reaction drinking water or

chlorination reaction mixtures.

In summary, evidence has been provided to show that

genotoxicants are produced during chlorination of water which

contains proteinaceous material and fulvic and humic

substances. A comparison of the mutagenicities of chlorinated

and non-chlorinated surface waters has always shown that

mutagens are present in the chlorinated samples. In any

environmental impact assessment study, those compounds that









75

are not easily identifiable should be taken into account in

order to ensure that they are not being ignored.















CHAPTER III
MATERIALS AND METHODS

Chlorine Demand-free Water

Chlorine demand-free water was prepared following a

modification of the procedure of Ghanbari et al. (1982a).

Distilled water was passed through two successive Barnstead

deionizing units (Barnstead, Division of Sybron Corporation,

Boston, MA) and then through a glass column containing Porapak

Q (Supelco, Inc., Bellefonte, PA). The water was analyzed by

iodometric titration (APHA, 1985) to determine the presence

of any available chlorine. Water used for the preparation of

all the reagents and substrates was prepared this way.



Generation of Aqueous Chlorine

Commercial chlorine is contaminated with a variety of

organochlorine compounds including chloroform, phosgene,

carbon tetrachloride, methylene chloride, etc. (Christman,

1983). To ensure that no such contaminants were present in

the chlorine sample used in this study, a modification of the

methods of Ghanbari et al. (1983) was used for the generation

of aqueous chlorine. Chlorine gas was prepared by dropwise

addition of 3 N HC1 to KMnO4 (Fisher Scientific, Fair Lawn,

NJ) in a closed system under vacuum. The liberated gas was









77

trapped in ice-cold chlorine demand-free water or 0.1 M sodium

phosphate buffer, pH 7 as desired. Under these generation

conditions, the gas was efficiently trapped mainly as

hypochlorous acid. This was detected by a drop in the pH of

the solution from 7.0 to 2.5-3 due to the liberation of

hydrogen ions. The purity and available chlorine

concentrations of the stock solutions were determined by

iodometric titration (APHA, 1985). Aqueous chlorine solutions

were diluted when necessary and used immediately after

preparation.



Generation of Aqueous Chlorine Dioxide

Stock solutions of aqueous chlorine dioxide were prepared

by slow addition of 2 N H2SO4 to 6 g sodium chlorite (Eastman

Kodak Co., Rochester, NY) in 6 mL water in a closed system

(APHA, 1985). The chlorine dioxide gas generated from the

reaction was passed through a sodium chlorite clean-up column

to remove any contaminants such as chlorine. The gas was then

collected as a greenish-yellow solution in ice-cold chlorine

demand-free water or 0.1 M sodium phosphate buffer as desired.

As before, the concentrations of chlorine dioxide solutions

were determined iodometrically. To prevent losses due to

volatilization and/or breakdown, chlorine dioxide solutions

were always prepared fresh, diluted when necessary, and used

immediately after preparation. All chlorine and chlorine








78

dioxide stock solutions were prepared under yellow lights to

prevent photodecomposition.



Iodometric Titration

lodometric titration was used to determine available

chlorine and chlorine dioxide concentrations in all the stock

solutions as well as chlorine residuals remaining after the

reactions. The method involves liberation of iodine from

potassium iodide (KI) by chlorine. The liberated iodine is

then titrated with sodium thiosulfate (Na2S203) to a starch end

point (APHA 1985). Excess KI acidified with acetic acid to

pH 3-4 was mixed with aliquots of the stock chlorine or

chlorine dioxide solutions in glass-stoppered flasks at room

temperature. In the case of chlorine dioxide solutions, the

chlorine dioxide-KI solutions were kept in the dark for about

5 minutes prior to titration. The mixtures were then titrated

under yellow lights with 0.01 N Na2S203 solution prepared by

diluting from standardized stock solutions, until the yellow

color of the liberated iodine was almost clear. A few drops

of a 0.5% starch indicator solution were added and titration

was continued to a colorless end point. A blank titration

using only chlorine demand-free water was also performed to

correct for interference from oxidizing and reducing

substances and to correct for the concentration of iodine

bound starch at the end point (APHA, 1985).








79

The total available chlorine or chlorine dioxide was

calculated as follows:

mg Cl or ClO2 as C12/L (ppm) = (A B) x N x 35450
mL sample


where: A = Volume of Na2S203 used for the sample,

B = Volume of Na2S203 used for the blank,

N = Normality of the Na2S203 used.



Reactions of Aqueous Chlorine or Chlorine Dioxide with
L-Tryptophan

Preliminary reactions of aqueous chlorine

In preliminary experiments, 0.1 M L-tryptophan (Sigma

Chemical Co., St. Louis, MO) solutions were prepared in 0.1

M sodium phosphate buffer at pH 7.0 with slight warming to

ensure that all the amino acid was dissolved. Two molar

ratios of aqueous chlorine to amino acid, 1:1 and 3:1, and a

reaction time of 24 hours was used to ensure that the reaction

was complete. This study used aqueous chlorine alone with pH

adjustment of the reaction mixture after 24 h to obtain

acidic, neutral and basic fractions. Each fraction was

extracted separately first with hexane, then with methylene

chloride and lastly with ethyl ether.

Reactions of aqueous chlorine and chlorine dioxide

In the next series of experiments, three molar ratios of

L-tryptophan to aqueous chlorine or chlorine dioxide, 1:1, 1:3

and 1:7 (M:M), were chosen. Tryptophan solutions were








80

prepared in 0.1 M sodium phosphate buffer, pH 7.0, as before

and aqueous chlorine and chlorine dioxide solutions were

prepared in chlorine demand-free water. For the 1:1 and 1:3

reactions, 14 mM tryptophan solutions (approximately 2.86 g

tryptophan in 1 L sodium phosphate buffer) were reacted with

14 and 42 mM chlorinating agents (total volume 1 L),

respectively. The 1:7 reactions were carried out using a 10

mM amino acid solution (2.04 g tryptophan in 1 L sodium

phosphate buffer) and a 70 mM chlorine or chlorine dioxide

(total volume 1 L). The amino acid concentration used for the

latter reactions were limited by the amount of chlorinating

agent generated by the above procedures. All the reactions

were carried out using equal volumes of reagent and substrate

(2 L final volume), therefore; the final concentrations of

chlorinating agents and amino acid were halved. Reactions

were carried out in the dark in glass jars at ambient

temperature for 24 hours with continuous mixing on a Corning

stirrer. Procedural blanks containing only amino acid,

aqueous chlorine or chlorine dioxide in sodium phosphate

buffer were used as controls.

After reaction, the samples were filtered using a 0.45

gm membrane filter (Gelman Sciences, Inc., Ann Arbor, MI) to

remove any precipitate present. Aliquots of the reaction

mixture were titrated iodometrically to determine if any

available chlorine remained. The samples were then split in

half. The pH of one-half was readjusted to 2.5 with 85%









81

phosphoric acid and saved for resin adsorption, while the

other fraction was kept at neutral pH for liquid-liquid

extraction using ethyl ether.



Chlorine Consumption

Chlorine concentrations before and after chlorination

were measured iodometrically (APHA, 1985). Any excess

chlorine or chlorine dioxide remaining after the 24-hour

reaction was quenched using sodium sulfite (Fisher Scientific,

Fair Lawn, NJ) in order to avoid contamination with elemental

sulfur produced from dechlorination with sodium thiosulfate

(Trehy et al., 1986).



Concentration Techniques

Different techniques, mainly liquid-liquid extraction,

Amberlite XAD adsorption and column chromatography were used

to concentrate the reaction mixtures for mutagenicity and

other assays.

Liquid-liquid extraction

Preliminary studies. In preliminary liquid-liquid extraction

studies, the active ingredients in the aqueous chlorination

reaction mixtures were extracted using organic solvents of

different polarity. Ethyl ether and hexane (Fisher

Scientific) were reagent A.C.S. and methylene chloride (Fisher

Scientific) was HPLC grade. The reaction mixtures containing

aqueous chlorine and tryptophan at neutral pH were extracted









82

three times by mixing vigorously on a stirrer with equal

volumes of hexane, followed by methylene chloride and then

ethyl ether. Each extraction took approximately 2 hours and

represented the neutral fraction. After these extractions,

the aqueous mixture was divided into two portions. The pH of

one fraction was adjusted to 1.2 using concentrated HC1 and

that of the other half was changed to pH 11.7 using 10 N

sodium hydroxide. These fractions were each extracted with

the three organic solvents as before to obtain acidic and

alkaline fractions, respectively, as shown in Fig. 5.

In the preliminary studies, high mutagenic activity was

detected in the ethyl ether extracts of the reaction products.

Therefore, in the latter part of the study when three ratios

of aqueous chlorine and chlorine dioxide were compared, only

ethyl ether was used for the liquid-liquid extractions.

Mixtures containing different molar ratios of aqueous chlorine

or chlorine dioxide with tryptophan were extracted vigorously

with mixing on a stirrer with the aid of a magnetic stirrer

for approximately 4 hours, and then separated with a

separatory funnel. The aqueous phase was discarded and the

ethyl ether extract was dried by rotary evaporation and saved

for further analysis.

Evaluation of product distribution in the ethyl ether

extracts. A solution containing radiolabeled '4C-tryptophan

of known activity was prepared as described below. The

reaction products were extracted with ethyl ether for 2 hours.













0.1 M Aqueous chlorine + 0.1 M tryptophan
in 0.1 M sodium phosphate buffer (pH 7.0)



250C, 24 h


Filter


Precipitate


Aqueous phase



Extract 3x with
-hexane
-methylene chloride
-ethyl ether


Aqueous pha


Adjust to pH 1.2



Extract 3x with
land


Adjust to pH 11.7



hexane, methylene chloride
ethyl ether


\V
-> Concentrate by rotary evaporation
in vacuo



Assay for mutagenic activity


Figure 5. Schematic presentation of sample preparation
for mutagenicity assessment in the preliminary
studies


fl 3 n n =a


y**ac=


__








84

The ethyl ether layer was separated from the aqueous layer and

partially dried to a known volume. The percentage

distribution of radioactivity in the aqueous solution and

ether extracts was measured by dissolving aliquots in 15 mL

Scintiverse II cocktail and counted on a Beckman Instruments

liquid scintillation spectrometer as described below.


Amberlite XAD polymeric resin adsorption

Evaluation of Amberlite XAD procedure for concentrating the

reaction products. Amberlite XAD-2, a nonpolar polystyrene

adsorbent and XAD-8, an acrylic ester of intermediate

polarity, were obtained from Rohm and Haas (Philadelphia, PA).

The resins were prepared prior to use by continuous soxhlet

extraction with acetone for 5 hours followed by further

washing with methanol for an additional 5 hours. The clean

resins were stored separately in methanol until needed. Prior

to use, the resins were slurry packed in methanol in a glass

column, 52 cm by 1.9 cm i.d. to form a bed height of 30 cm.

The resins were then rinsed with 3-5 bed volumes of distilled

water to ensure that all the methanol was removed.

A 7 mM tryptophan solution in 0.1 M sodium phosphate

buffer (pH 7.0) was prepared by dissolving 0.5 g of the amino

acid in 350 mL buffer. Radiolabeled L-tryptophan (side chain-

3-14C) with a specific activity of 53.76 mCi/mmole was obtained

from New England Nuclear, Inc. (Boston, MA). Radiolabeled

tryptophan (0.1 mCi in 0.1 mL phosphate buffer) was mixed with








85

the tryptophan solution and the total radioactivity of the

solution was measured. Equimolar concentrations of either

aqueous chlorine or chlorine dioxide were added to separate

reaction flasks and allowed to react for 4 hours in the dark

at room temperature. The reaction mixtures were filtered

separately using a 0.45 pm membrane filter and the pH and

radioactivity of the solutions were measured. The solutions

were then passed through a glass column (52 cm x 1.9 cm i.d.)

containing either XAD-2 alone or in combination with XAD-8

(1:1 mixture) at a bed height of 30 cm.

Trial experiments were conducted to determine the best

Amberlite XAD concentration technique for the reaction

products. In one experiment, XAD-8 was placed on top of XAD-2

while in another the order was reversed. In another trial,

only XAD-2 was used. Also, the importance of sample pH on the

recovery of the reaction products was evaluated by adjusting

sample pH. The pH of the reaction products in one experiment

was adjusted to 2.5 with 85% phosphoric acid while in another,

it was maintained at 7.0. All the reaction products were

passed through the various resin columns at a flow rate of

approximately 40 mL/minute. The first pass effluent was

poured on the column. The adsorbed substances were eluted

successively with 200 mL each of methanol, ethyl ether and

acetone. The combined organic solvents were removed by rotary

evaporation under vacuum and the residues were diluted to 10

mL with chlorine demand-free water. Radioactivity of the









86

aqueous filtrates and concentrates were measured by counting

0.1 mL aliquots in 15 mL Scintiverse II cocktail (Fisher

Scientific, Fair Lawn, NJ) using a Beckman Instruments

spectrometer Model LS 2800 (Beckman Instruments, Inc.,

Fullerton, CA) with automatic external standardization.

Distribution of radioactivity in the filtrates and eluates

was used to measure the recovery of the reaction products.

Similar sample preparation was used to determine product

distribution in the liquid-liquid extractions.

Due to the high mutagenic activity in the Amberlite

XAD-2/8 acetone eluates of the 7:1 (aqueous chlorine or

chlorine dioxide:tryptophan) ratio, radioactivity distribution

in these eluates were also determined.

Sample concentration for mutagenicity assessment using

Amberlite XAD resins. Based on the results from the

preliminary studies, a resin column with Amberlite XAD-8

placed on top of XAD-2 (1:1 mixture) was used for

concentrating the reaction products. The pH of the aqueous

mixtures containing the reaction products of either chlorine

or chlorine dioxide with tryptophan was adjusted to 2.5 with

85% phosphoric acid and passed through the resins at a flow

rate of approximately 40 mL/minute. This procedure was

repeated using the first-pass effluent. Control samples

containing aqueous chlorine, chlorine dioxide or tryptophan

alone in sodium phosphate buffer were dechlorinated with

sodium sulfite and also concentrated by resin adsorption.









87

After most of the aqueous fraction had passed through the

column and the water level was barely on top of the resin,

acetone was added and allowed to flow freely through the

column until it reached the bottom. Flow was interrupted and

the acetone allowed to equilibrate with the resins for 10-15

minutes. A total of 300 mL of acetone was used to elute most

of the adsorbed substances from the column. Any color or

substances remaining on the column was further eluted

successively with 300 mL methanol followed by 300 mL ethyl

ether. All the eluates were concentrated separately by rotary

evaporation (Fig. 6). The dry samples were weighed, dissolved

in and diluted with spectrophotometric grade dimethyl

sulfoxide (DMSO) and aliquots used for the mutagenicity

studies.

Rotary evaporation

The rotary evaporator used was a Buchi Rotavapor R 110

(Brinkmann Instruments, Inc., Westbury, NY). The aqueous

reaction mixtures containing tryptophan and chlorine as well

as all the organic solvents obtained from the liquid-liquid

extractions and Amberlite XAD adsorptions were concentrated

by rotary evaporation in vacuo at 450C.

Concentrated samples were quantitatively transferred into

preweighed teflon-capped vials and the remaining traces of

organic solvents were removed under a gentle stream of

nitrogen gas. Samples were then weighed, dissolved in and

diluted with spectrophotometric grade DMSO and used for the












Tryptophan solution in sodium phosphate buffer
(pH 7.0)

+

Aqueous chlorine or chlorine dioxide
chlorine to tryptophan ratio (1:1, 3:1 or 7:1)


250C, 24 h


Dechlorinate with sodium sulfite



Filtration



Adjust to pH 2.5 Extract with ether


Amberlite XAD-2/8 adsorption


Elute with-acetone
-methanol
-ether


Concentrate separately by rotary evaporation




Assay for mutagenic activity





Figure 6. Schematic presentation of sample preparation for
mutagenicity assessment in the reactions of
aqueous chlorine or chlorine dioxide with
tryptophan




University of Florida Home Page
© 2004 - 2011 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated May 24, 2011 - Version 3.0.0 - mvs