w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 2 2 0 1 – 2 2 0 8

Avai lab le a t www.sc iencedi rec t .com

journa l homepage : www.e lsev ie r . com/ loca te /wat res

Controversies about the occurrence of chloral hydrate indrinking water

Agata Dabrowska*, Jacek Nawrocki

Adam Mickiewicz University, Faculty of Chemistry, Department of Water Treatment Technology, 60-613 Poznan, ul. Drzymały 24,

Wielkopolska, Poland

a r t i c l e i n f o

Article history:

Received 13 March 2008

Received in revised form

15 February 2009

Accepted 18 February 2009

Published online 26 February 2009

Keywords:

Chlorine

Disinfection by-products

Chloral hydrate

Trihalomethanes

Aldehyde

* Corresponding author. Tel.: þ48 618293432;E-mail address: agatadab@amu.edu.pl (A

0043-1354/$ – see front matter ª 2009 Elsevidoi:10.1016/j.watres.2009.02.022

a b s t r a c t

Besides trihalomethanes (THMs) and haloacetic acids (HAAs), chloral hydrate (CH) is the

next most prevalent disinfection by-product (DBP) in drinking water, formed as a result of

the reaction between chlorine and natural organic matter (NOM). Chloral hydrate (tri-

chloroacetaldehyde) should be limited in drinking water because of its adverse health

effect. The controversies concerning the appearance of CH in disinfected water found in

literature are discussed in the present paper. According to some authors the CH yield

during chlorination of water depends only on TOC. However, there are other data available

that do not confirm this relationship. Another fact requiring clarification is the dependence

of CH formation on pH. In the present study, CH formation is analysed in different types of

water disinfected with different doses of chlorine. Formation of CH is correlated with the

dose of Cl2 and the contact time. The formation of chloral hydrate takes place as long as

chlorine is available in the water. Total organic carbon (TOC) is not considered the main

factor influencing the production of chloral hydrate in water treated with Cl2 as the

production depends also on the nature of NOM. Higher levels of CH are observed at alkaline

conditions (pH> 7). A significant correlation (R2> 0.9) between the concentrations of

chloral hydrate and chloroform has been observed. The preozonation increases signifi-

cantly the chloral hydrate formation potential in the water treated. Biofiltration process

does not remove all of CH precursors and its efficiency depends strongly on the contact

time. Chloral hydrate was analyzed by gas chromatography with electron capture detector

with the detection limit 0.1 mg L�1.

ª 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Chlorination process is the most widely used method of

drinking water disinfection. However, this method is respon-

sible for conversion of nontoxic organic compounds naturally

occurring in waters into toxic, mutagenic organohalogens,

undesired disinfection by-products (DBPs). Beside trihalo-

methanes and haloacetic acids, chloral hydrate has been

fax: þ48 618293400.. Dabrowska).er Ltd. All rights reserved

reported as one of the main genotoxic and carcinogenic

compounds (Koudjonou and LeBel, 2006; Richardson et al.,

2007; Meric, 2007). Chloral hydrate is the next (to THMs and

HAAs) most prevalent disinfection by-product in drinking

water. According to the World Health Organisation (WHO)

guideline, chloral hydrate should be limited in drinking water

because of its adverse health effect and the current MCL is set

at 10 mg L�1 (WHO, 2004). Chloral hydrate is not known to

.

Nomenclature

HAAs haloacetic acids

BAF biologically active filter

BDOC biodegradable dissolved organic carbon

CH chloral hydrate

DBPs disinfection by-products

ECD electron capture detector

GC gas chromatography

I.D. internal diameter

MCL maximum contaminant level

MTBE methyl-t-butyl ether

NOM natural organic matter

PFBOA O-(2,3,4,5,6-pentafluorobenzyl)hydroxylamine

RSD relative standard deviation

THMs trihalomethanes

TOC total organic carbon

WHO World Health Organisation

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 2 2 0 1 – 2 2 0 82202

occur as a natural product, so the major route of exposure to

chloral hydrate is from drinking water. Chloral hydrate was

synthesized in 1832 and has been used widely since the end of

the 19th century as a sedative and hypnotic drug in human

and veterinary medicine (EPA, 2000). Although the use of this

compound is related to a significant risk of mutagenesis and

carcinogenicity, it is still in therapeutic use. Fatalities have

been observed in the case of CH overdoses or of the combi-

nation of CH with large doses of alcohol (Meyer et al., 1995;

Gaulier et al., 2001). Chloral hydrate has been previously

studied, but the appearance of chloral hydrate in drinking

water is not well established and arouses many controversies.

Although it is common knowledge that chloral hydrate is

formed in the reaction between natural organic matter and

chlorine (Richardson, 1998; Barrott, 2004) the mechanisms of

this reaction is not so straightforward because of the complex

character of NOM. There are many different precursors of CH.

Trehy et al. (1986) report that amino acids are potential

precursors of chloral hydrate and suggest that there are

different precursors of chloroform and chloral hydrate.

According to Ueno et al. (1995), nitrogen compounds and

amino acids produced chloral hydrate in the chlorination

process. Krasner et al. (1993), Weinberg et al. (1993), LeBel et al.

(1997) and LeBel and Benoit (2000) consider low molecular

weight aldehydes formed during preozonation process as

responsible for the efficient production of CH after chlorina-

tion. Koudjonou et al. (2008) identified chloral hydrate in most

Canadian drinking water that had been treated with a chlori-

nated disinfectant and the levels of CH were generally higher

in systems when ozone was used as the primary disinfectant.

The studies of Koudjonou et al. (2008) indicated increased

formation of CH with pH and temperature along with

concurrent decrease of acetaldehyde. Barrott (2004) has

reported the effect of pH value on CH formation and suggested

that chloral hydrate decomposes at alkaline pH to form

chloroform.

In this paper, we discuss the controversies connected with

the occurrence of chloral hydrate in drinking water and the

influence of water treatment conditions on its formation. The

effect of chlorine dose, time of reaction, TOC content, pH

value and preozonation with biofiltration process on chloral

hydrate content in drinking water is discussed against the

results of other authors. We have found a significant corre-

lation between chloral hydrate and chloroform in the dis-

infected water, the knowledge of this relationship could be

helpful when concentration of CH is not monitored regularly

in the water-supply system.

2. Materials and methods

2.1. Water samples

Various water samples were collected for the study:

� raw surface water from three rivers from the Poznan area:

Warta (pH¼ 7.25; TOC¼ 7.9 mg L�1; 0.40 mg NH3 L�1),

Cybina (pH¼ 7.04; TOC¼ 12.6 mg L�1; 0.69 mg NH3 L�1) and

Bogdanka (pH¼ 7.35; TOC¼ 5.6 mg L�1; 0.40 mg NH3 L�1)

� raw surface water from a lake in Poznan: Malta (pH¼ 7.10;

TOC¼ 12.9 mg L�1; 0.49 mg NH3 L�1)

� infiltrated water after iron and manganese removal

(pH¼ 6.80; TOC¼ 5.1 mg L�1; 0.20 mg NH3 L�1)

� ozonated natural surface water

� Millipore organic free water spiked with humic acid

� Millipore organic free water spiked with acetaldehyde

� tap waters from different cities in Poland: Wroc1aw, Ples-

zew, Torun, Gdansk, Mogilno

� P1, P2 – tap waters, collected from two points of the water-

supply system in Poznan.

Water samples were stored for three days at 5 �C.

2.2. Chemicals and materials

O-(2,3,4,5,6-pentafluorobenzyl)hydroxylamine (PFBOA) and

analytical standards were purchased from Aldrich–Chemie

(Steinheim, Germany) and BDH (Pool, UK; AnalaR and GPR

grades). PFBOA was prepared gravimetrically as an aqueous

solution in organic free water, while acetaldehyde solution was

prepared in methanol (J.T. Baker, Germany). Hexane (J.T. Baker,

Holland) and tert-butyl methyl ether dried over molecular sieve

(Sigma–Aldrich, Switzerland) were used as solvents for extrac-

tion. Humic acid (Aldrich, Germany) was used as an aqueous

solution in Millipore organic free water. Sodium chloride

(CHEMPUR, Poland) was dried in a muffle furnace at 350 �C for

45 min prior to use. Chlorine was dosed as hypochlorite NaOCl

and was obtained directly from the Water Treatment Plant at

Mosina Water Intake. The chlorine concentration was

measured with the iodometric method. Sodium sulphite (Fluka,

Germany) was used as dechlorinating agent.

2.3. Determination of analytes

The following analytes were determined in water samples:

chloral hydrate, acetaldehyde and trihalomethanes

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 2 2 0 1 – 2 2 0 8 2203

(chloroform, bromodichloromethane, dibromochloro-

methane, bromoform).

Acetaldehyde analysis involved the processes of derivati-

zation and chromatographic separation – the method earlier

improved by Sclimenti and co-workers (Sclimenti et al., 1990)

and optimised in our laboratory. The reaction of carbonyl

compound with PFBOA was used to obtain oximes that, after

extraction with organic solvent, were analysed by gas chro-

matography (GC), using Fisons Serie 8000 GC equipped with63Ni electron capture detector (ECD) with nitrogen (99.9999%)

as a make-up gas and helium (99.999%) was used as carrier

gas. A Rtx-5MS (Restek) fused silica capillary column

(30 m� 0.25 mm I.D.� 0.25 mm film) was employed for anal-

ysis. Injection volume 0.5 ml was used into ‘‘on-column’’

system. The analysis was performed with the temperature

programme starting from 80 �C for 4 min, then increased to

240 �C at 5 �C min�1, then raised to 270 �C at 20 �C min�1. The

CSW (Chromatography Station for Windows, Version 1.7 Eval,

Built, Czech Republic) system was used to collect and process

the chromatographic data. Quantification was carried out by

means of an external standard calibration curve. The details

of the method are described elsewhere (Dabrowska et al.,

2003, 2005; Nawrocki et al., 2002).

There are many chromatographic techniques for identifi-

cation of chloral hydrate (Brunzzoniti et al., 2001; Delinsky

et al., 2005; Golfinopoulos and Nikolaou, 2005; Andrzejewski

et al., 2007). In the present work we use the method 551.1

recommended by USEPA (EPA, 1995) and optimised in our

laboratory. THMs and chloral hydrate were analysed in the

following way: 20 ml of collected water was extracted for

2 min with 2 ml of tert-butyl methyl ether. 4 g of NaCl were

added to each sample. The extracts were analysed by the

same GC/ECD system, but the analysis was performed with

the temperature programme starting from 35 �C for 10 min,

then increased to 290 �C at the rate 20 �C min�1. Calibration

curves were made by plotting concentration versus peak

area. The detection limit for chloroform was 0.2 mg L�1

and 0.1 mg L�1 for bromodichloromethane, dibromochloro-

methane, bromoform and chloral hydrate. Average relative

standard deviation (RSD) was about 10% for each of the

monitored trihalomethanes and 8% for CH.

0.4

0.6

1.9

3.0

5.1

8.0

9.3

121086420

1

2

4

6

8

10

12

TO

C [m

gC

L-1]

Chloral hydrate [µg.L-1

]

A

Fig. 1 – Chloral hydrate content depending on total organic carbo

A: homogeneous matrix (Millipore organic free water spiked wit

2.4. TOC measurement

Total organic carbon was measured by means of AURORA

Model 1030 (I.O. Analytical) using the persulphate/100 �C wet

oxidation method.

2.5. Ozonation process

Ozone was generated from pure oxygen by means of an

AZCOZON generator (Crystal, Canada) for laboratory

purposes.

The residual ozone concentration was determined using

the indigo method according to Standard Methods (Standard

Methods, 1980).

2.6. Biofiltration process

2 m high and 4 cm I.D. filter filled by granular activated carbon

F300 (FILRASORB 300, Chemviron Carbon, Belgium) was

applied as biologically active filter (BAF). The mean particle

diameter of F300 was 1.6 mm and surface area (N2 BET

method) was 950 m2 g�1.

2.7. Production of chloral hydrate

Production of chloral hydrate in the samples of different

natural waters used in experiments was established after

30 min reaction time on the basis of the increase in the CH

concentration after the reaction with 0.2 mg Cl2 per 1 mg TOC.

High purity, organic free water obtained with a Millipore

Simplicity Ultrapure Water System was used for blanks

control.

3. Results and discussion

3.1. Correlation between chloral hydrate andTOC content

The formation of trihalomethanes as organic by-products of

chlorination process is well established (Richardson, 1998).

1.1

11.4

3.8

26.7

11.3

9.5

302520151050

5.1

5.6

7.9

9.1

12.6

12.9

TO

C [m

gC

L-1]

Chloral hydrate [µg.L-1

]

Malta

Cybina

Warta

Bogdanka

Preozonated water

Infiltrated water

B

n (chlorine dose 0.2 mg Cl2/mg Corg, contact time 30 min) in

h humic acid), B: complex matrix (different natural waters).

Table 1 – Production of chloral hydrate for differentnatural waters (chlorine dose 0.2 mg Cl2 per 1 mg Corg,contact time 30 min).

Water samples TOC[mg L�1]

Production ofCH [mg mg�1]

Infiltrated water (after iron and

manganese removal)

5.1 0.2

Surface water from Bogdanka river 5.6 2.1

Surface water from Warta river 7.9 0.5

Preozonated water 9.1 2.9

Surface water from Cybina river 12.6 0.9

Surface water from Lake Malta 12.9 0.7

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 2 2 0 1 – 2 2 0 82204

The mechanism of chloral hydrate formation is similar to the

mechanism of THMs formation. As follows from earlier

experience of many authors (LeBel et al., 1997; Richardson,

1998; Barrott, 2004) the more NOM in the raw water, the higher

the concentration of chloral hydrate can be expected. We

prepared water samples with different doses of organic

matter, using aqueous solution of humic acid in the experi-

ment as the chloral hydrate precursor. The quantity of organic

matter in the water samples was characterized by the

concentration of organic carbon (Corg) defined as TOC. The

TOC concentration in the samples was 1, 2, 4, 6, 8, 10 and

12 mg Corg L�1 respectively. All the spiked waters were sub-

jected to chlorination process with the dose of Cl2 0.2 mg per

1 mg Corg and after 30 min of the reaction the chloral hydrate

concentration has been analyzed. The results are shown in

Fig. 1A: the higher the initial TOC concentration in a water

sample the greater the increase in chloral hydrate formation.

These results are in agreement with the data described by

other authors. In another series of experiments, several types

of natural waters were subjected to the same chlorination

(0.2 mg Cl2/mg TOC) and the results were quite different as

can be seen in Fig. 1B. The concentration of total organic

carbon ranged from 5.1 to 12.9 mg L�1 while the concentra-

tions of chloral hydrate were found in the range from

1.1 mg L�1 to 11.4 mg L�1 and even 26.7 mg L�1 (for the sample

treated with ozone before chlorination). It is seen that the

chloral hydrate concentration has been changing irregularly

and there is no correlation between chloral hydrate concen-

tration and TOC content. The nitrification as one potential

cause for the decrease of ammonia during storage, occurred in

all examined samples and significantly affected the ammonia

content. Generally, the decrease in ammonia after three days

Dose of chlorine [mgCl2/mgC

org]

0

5

10

15

20

Cybina Malta WartaSurface waters

Ch

lo

ral h

yd

rate

[µg

.L

-1]

0.1 0.2 0.5A

Fig. 2 – Chloral hydrate formation at various chlorine doses and

tc [ 72 h.

of storage at 5 �C was about 67% (thus the average concen-

tration was about 0.1 NH3 mg L�1) and ammonia did not play

significant role in competitions with NOM for the reaction

with chlorine. Although the level of organic carbon was the

highest in the surface water from Lake Malta (12.9 mg L�1),

after chlorination process the highest CH concentration was

observed in the surface water from the Bogdanka river, with

a much lower TOC value (5.6 mg L�1). The production of

chloral hydrate calculated for various samples used in

experiments is presented in Table 1. It can be concluded that

the effect of NOM nature on chloral hydrate formation is very

important and CH production depends more on the quality of

the precursors than on their quantity. However, when the

same type of organic matter is considered, then higher

concentration of the matter produces higher concentration of

chloral hydrate.

3.2. Effect of chlorine dose and reaction timeon CH concentration

The influence of the chlorine dose on the chlorinated by-

products formation is significant. Three doses of chlorine

(calculated per 1 mg Corg) were applied: 0.1, 0.2 and 0.5 mg and

three samples of surface waters were investigated, from: the

Warta river, the Cybina river and Lake Malta. The concentra-

tion of chloral hydrate was analysed after 0.5 h and 72 h of the

reaction. The results are shown in Fig. 2A and B. The increase

in the chlorine dose evidently leads to higher concentrations

of CH in all types of water samples studied, particularly in the

water from the Cybina river and Lake Malta.

It is generally known that the concentration of the chlori-

nated by-products increases with the duration of the reaction.

However, in the samples from the Cybina river and Lake Malta

we noted a small decrease in the chloral hydrate concentra-

tion after 72 h. This phenomenon occurs when small doses of

chlorine are used and it is in agreement with the suggestion of

Barrott (2004) that chloral hydrate can disappear with time in

disinfected water. In our study the chloral hydrate formation

was not continued after the disinfectant had been completely

consumed and the decay of CH in time was observed.

However, for larger doses of chlorine the reaction of Cl2 with

organic matter takes place as long as chlorine is available in

the water. Comparing the results presented in Fig. 2A and B,

we can see that the CH concentration in the Warta river

sample increased several times within 72 h, when 0.5 mg/

Dose of chlorine [mgCl2/mgC

org]

05

10152025303540

Cybina Malta WartaSurface waters

Ch

lo

ral h

yd

rate

[µg

.L

-1]

B 0.1 0.2 0.5

contact times A: contact time tc [ 0.5 h B: contact time

Contact time

05

101520253035

pH=4 pH=5 pH=7 pH=8 pH=9

Ch

lo

ral h

yd

rate

[µg

.L

-1]

t=0.5h t=1h t=24h

0

2

4

6

8

10

12

pH=5 pH=7 pH=8 pH=9

Ch

lo

ral h

yd

rate

[µg

.L

-1]

A B

Contact time

t=0.5h t=1h t=24h

Fig. 3 – Chloral hydrate formation at various pH values and contact times, A: surface natural water (Bogdanka river);

B: Millipore organic free water spiked with acetaldehyde.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 2 2 0 1 – 2 2 0 8 2205

mg Corg chlorine dose was used. Thus an uncontrolled

increase in the chloral hydrate concentration in the distribu-

tion system may be expected, when the CH precursors are not

sufficiently removed from the water treated and chlorine is

available in the system.

6

8

10

ate [µg

.L

-1]

30

40

Ch

lo

ro

fo

r

CH in P1 CH in P2CHCl3 in P1CHCl3 in P2

3.3. Effect of pH on CH formation

The influence of pH on chloral hydrate formation in dis-

infected water is discussed in literature (Barrott, 2004; Koud-

jonou et al., 2008). Miller and Uden (1983) found that CH

formation was dependent on pH with a steady increase in

concentration from pH 4 up to a neutral pH at which its

concentration began to decrease. Our results do not confirm

this relationship. Two kinds of water samples: the natural

surface water from the Bogdanka river and organic free water

spiked with acetaldehyde were tested in a wide range of pH,

from 4 to 9. The samples were chlorinated with the constant

dose of chlorine 0.2 mg mg�1 Corg. The CH formation was

controlled after 0.5 h, 1 h and 24 h of reaction. In the natural

water (Fig. 3A) the chloral hydrate concentration increased

steadily with increasing pH, and the maximum CH concen-

tration was identified at pH 8. However, in the water

contaminated with acetaldehyde, the highest CH concentra-

tion was found at pH 9 (see Fig. 3B). Additionally, regardless of

pH value, the CH concentration was even higher when the

contact time with the disinfectant was extended (compare

Fig. 3A and B). It is evident that pH is an important parameter

in CH formation, and generally higher pH means higher CH

formation. As can be seen in Fig. 3, this correlation is strongly

sample dependent.

0

2

4

30.03

.07.

2.04.0

7.

13.04

.07.

17.04

.07.

24.04

.07.

27.04

.07.

Date of sample collection

Ch

lo

ral h

yd

r

0

10

20 m [µ

g.L

-1]

Fig. 4 – Comparison of chloral hydrate and chloroform

formation in chlorinated tap water (P1 and P2 – sites of

sample collection, situated near to and far from the Water

Plant respectively).

3.4. Correlation between chloral hydrate andTHMs formation

In the 30 years since the THMs were identified as DBPs in

drinking water, significant research efforts have been directed

toward understanding the DBP formation, occurrence, and its

health effects (Nieuwenhuijsen et al., 2000; Richardson et al.,

2007). As the generation of THMs upon water chlorination is

well established (Richardson, 1998) and widely monitored in

drinking water, we tried to find the relationship between

THMs and CH formation. The THM concentrations in tap

water were compared with chloral hydrate concentrations

found in the same samples. For one month the chlorinated tap

water in Poznan was collected from two independent sites of

the water-supply system (signified as P1 and P2, situated near

to and far from the Water Plant respectively). Then trihalo-

methanes and chloral hydrate were determined and we have

compared the variation in the trihalomethanes concentration

with that of CH concentration. The best correlation was found

between chloroform and chloral hydrate. It appeared that the

variation curves drawn for chloroform concentration are

highly correlated with those for chloral hydrate, indepen-

dently of the site of sample collection (see Fig. 4).

According to Barrott (2004) the formation potential for

chloroform appears to be different from that for chloral

hydrate, indicating that there are different precursors of these

chlorination by-products. However, taking into consideration

the results presented above, a close relationship between the

occurrence of chloroform and chloral hydrate is observed.

Additionally, a correlation between the chloral hydrate and

chloroform formation was noted in all chlorinated surface

water samples examined. Evidently chloral hydrate formation

accompanies chloroform formation. As we can see in Fig. 5,

the average correlation coefficient is very high i.e. R2¼ 0.9447.

According to the results in Fig. 5, when the concentration of

chloroform is higher than ca. 38 mg L�1 a concentration of

0% 10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0min

5min

15min

30min

45min

Tim

e o

f b

io

filtratio

n

Fig. 7 – Extent of chloral hydrate precursor depending on

biofiltration time.

R2 = 0.9447

20

25

30

35

40

45

50

2 4 6 8 10 12 14 16Chloral hydrate [µg.L

-1]

Ch

lo

ro

fo

rm

[µg

.L

-1]

Fig. 5 – Correlation between chloral hydrate and

chloroform formation (average data for natural surface

waters).

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 2 2 0 1 – 2 2 0 82206

chloral hydrate higher than 10 mg L�1 can be expected, i.e.

higher than the MCL recommended by WHO guideline.

Methyl-t-butyl ether (MTBE) is recommended as the

primary extraction solvent in Method 551.1 (EPA , 1995) since it

effectively extracts all of the chlorination disinfection by-

products. However, due to safety concerns associated with

MTBE, pentane is offered as an optional extraction solvent by

some laboratories, particularly when THMs are identified.

Pentane is an appropriate solvent for all chlorination by-

products except for chloral hydrate. When pentane is used as

an optional solvent for THMs, CH is not extracted and in

practice chloral hydrate is not found in the sample. On the

basis of the data on chloroform it is however possible to assess

the approximate concentration of chloral hydrate. When the

treatment plant has problems with the excessive formation of

chloroform, it seems probable that chloral hydrate will be also

created in excessive amounts in the water treated.

3.5. Effect of ozonation and biofiltration process onCH presence

Ozone is the most powerful oxidant used for drinking water

treatment. The oxidation of NOM with ozone increases the

biodegradable dissolved organic carbon (BDOC) content in

water while subsequent biofiltration of water through bio-

logically active filters should remove the biodegradable

0

10

20

30

40

50

60

70

80

Cl2=0.2mg/mgC O3=1.7mg/mgCCl2=0.2mg/mgC

Cl2=0.5mg/mgC O3=1.7mg/mgCCl2=0.5mg/mgC

C

on

cen

tratio

n [µ

g.L

-1]

Chloral hydrateTHMs

Fig. 6 – Chloral hydrate formation after preozonation

process, natural surface water from Bogdanka river

(TOC [ 5.6 mg C LL1, pH [ 7.3 and contact time 30 min).

compounds. The most common groups of ozonation by-

products are aldehydes and carboxylic acids (Nawrocki

et al., 2003; Dabrowska et al., 2005). Ozone produces much

smaller amounts of mutagenic by-products than chlorine or

chlorine dioxide. It has been observed by us that preozo-

nation process used in water treatment contributes to the

decrease in THM precursors, but at the same time, a signif-

icant rise in CH concentration is noted. When NOM oxida-

tion with ozone was applied, acetaldehyde as an ozonation

by-product appeared in water and in consequence, after

chlorination dose 0.2 mg mg�1 Corg, the level of chloral

hydrate concentration was twice as high (see results in

Fig. 6) as with the same water before ozonation. It is evident

that the application of preozonation before chlorination

increases chloral hydrate formation potential, because

acetaldehyde is formed acting as a chloral hydrate

precursor. However, in the water treatment train ozonation

is usually followed by biofiltration that removes the

majority of biodegradable by-products. The biofiltration

process contributes to the removal of acetaldehyde (i.e. the

main precursor of chloral hydrate), but its effectiveness

strongly depends on contact time. As follows from Fig. 7,

the amount of CH precursors significantly decreases after

15–30 min of biofiltration. Biologically active filters (BAF) do

not remove all CH precursors.

02468

101214161820

13.04

.2006

21.04

.2006

24.11

.2006

11.12

.2006

30.03

.2007

2.04

.2007

13.04

.2007

date of sample collection

ch

lo

ral h

yd

rate

[µg

.L

-1]

Fig. 8 – Chloral hydrate content monitored in tap water in

Poznan.

Table 2 – Chloral hydrate concentration in differentwater-supply system.

l.p. Place of sample collection Chloral hydrate [mg L�1]

1 Wroc1aw 4.0

2 Pleszew 2.5

3 Toruna 8.9

4 Gdansk 0.8

5 Mogilno 0.4

a Preozonation process used before chlorination.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 2 2 0 1 – 2 2 0 8 2207

3.6. Identification chloral hydrate in tap water

Chloral hydrate was monitored periodically in chlorinated tap

water in Poznan and occasionally in other water-supply

systems in Poland, where chlorine is used for final disinfec-

tion. The results are presented in Fig. 8 and in Table 2.

Generally, chloral hydrate concentration does not exceed MCL

in tap waters, but CH was identified in each of the samples

delivered. As shown in Fig. 8, in the tap water in Poznan the

concentration of CH higher than MCL was noted twice over

the period studied reaching 12 mg L�1 and 17 mg L�1on the 13th

and 21st of April, respectively. Williams et al. (1997), LeBel and

Benoit (2000) identified chloral hydrate in tap water in Canada

in summer at a level above 20 mg L�1. LeBel and Koudjonou

et al. (2008) found a relationship between the amount of

chloral hydrate and the water temperature. They suggested

that in summer higher chloral hydrate concentrations are

more probable.

4. Conclusions

� The CH production depends to a higher degree on the

quality of NOM than on its quantity.

� Increasing chlorine doses lead to higher concentrations of

chloral hydrate.

� The reaction of chlorine with organic matter takes place as

long as chlorine is available in the water. The concentration

of chloral hydrate may continuously increase in the water-

supply system.

� The pH influences CH formation; the higher the pH the

higher the production of chloral hydrate.

� Preozonation process increases significantly CH production

in the water treated.

� The precursor removal by BAF depends on the contact time.

Biologically active filters do not remove all CH precursors.

� There is a notable correlation between chloral hydrate and

chloroform formation.

� This correlation can be used to estimate the risk of chloral

hydrate formation in chlorinated waters on the basis of

chloroform concentration.

r e f e r e n c e s

Andrzejewski, P., Dabrowska, A., Gofinopoulos, S., 2007.Analytical techniques for DBP measurement in control ofdisinfection by-products in drinking water systems. In:

Nikolaou, A., Rizzo, L., Selcuk, H. (Eds.). Nova SciencePublishers, Inc., New York, pp. 163–209.

Barrott, L., 2004. Chloral hydrate: formation and removal bydrinking water treatment. Journal of Water Supply: Researchand Technology-AQUA 53 (6), 381–390.

Brunzzoniti, M.C., Mentasti, E., Sarzanini, C., Tarasco, E., 2001.Liquid chromatographic methods for chloral hydratedetermination. Journal of Chromatography A 920, 283–289.

Delinsky, A.D., Bruckner, J.V., Bartlett, M.G., 2005. A review ofanalytical methods for the determination of trichloroethyleneand its major metabolites chloral hydrate, trichloroacetic acidand dichloroacetic acid. Biomedical Chromatography 19,617–639.

Dabrowska, A., Swietlik, J., Nawrocki, J., 2003. Formation ofaldehydes upon ClO2 disinfection.Water Research 37, 1161–1169.

Dabrowska, A., Kasprzyk-Hordern, B., Nawrocki, J., 2005.Aldehydes formation during water disinfection by ozonationand chlorination process. Global NEST Journal 7 (1), 61–71.

EPA, 1995. Method 551.1 Determination of ChlorinationDisinfection Byproducts, Chlorinated Solvents, andHalogenated Pesticides/Herbicides in Drinking Water byLiquid–Liquid Extraction and Gas Chromatography withElectron-Capture Detection. National Exposure ResearchLaboratory Office of Research and Development U.S.Environmental Protection Agency, Cincinnati, Ohio 45268.

EPA, 2000. Toxicological Review of Chloral Hydrate, In Support ofSummary Information on the Integrated Risk InformationSystem (IRIS). EPA/635/R-00/006. EPA, Washington, pp. 1–63.

Gaulier, J.-M., Merle, G., Lacassic, E., Courtiade, B., Haglund, P.,Marquet, P., Lachatre, G., 2001. Fatal intoxications with chloralhydrate. Journal of Forensic Science 46 (6), 1507–1509.

Golfinopoulos, S.K., Nikolaou, A.D., 2005. Survey of disinfectionby-products in drinking water in Athens, Greece. Desalination176 (1–3), 13–24.

Koudjonou, B.K., LeBel, G.L., 2006. Halogenated acetaldehydes:analysis, stability and fate in drinking water. Chemosphere 64(5), 795–802.

Koudjonou, B., LeBel, G.L., Dabeka, L., 2008. Formation ofhalogenated acetaldehydes, and occurrence in Canadiandrinking water. Chemosphere 72 (6), 875–881.

Krasner, S.W., Sclimenti, M.J., Coffey, B.M., 1993. Testing ofbiologically active filters for removing aldehydes formedduring ozonation. Journal of American Water WorksAssociation 85 (5), 62–71.

LeBel, G.L., Benoit, F.M., Williams, D.T., 1997. A one-year survey ofhalogenated disinfection by-products in the distributionsystem of treatment plants using three different disinfectionprocesses. Chemosphere 34 (11), 2301–2317.

LeBel, G.L., Benoit, F.M., 2000. Chloral hydrate in Canadian drinkingwater. In: Proceedings of the 28th Annual Water QualityTechnology Conference, Salt Lake City, UT, September 5–9.

Meric, S., 2007. Toxicity of disinfectants and disinfection by-products: what did we learn to explore more? In: Nikolaou, A.,Rizzo, L., Selcuk, H. (Eds.), Control of Disinfection By-Productsin Drinking Water Systems. Nova Science Publishers, Inc.,New York, pp. 231–249.

Meyer, E., Van Bocxlaer, J.F., Lambert, W.E., Piette, M., DeLeenheer, A.P., 1995. Determination of chloral hydrate andmetabolites in a fatal intoxication. Journal of AnalyticalToxicology 19 (2), 124–126.

Miller, J.W., Uden, P.C., 1983. Characterization of nonvolatileaqueous chlorination products of humic substances.Environmental Science and Technology 17 (3), 150–157.

Nawrocki, J., Dabrowska, A., Borcz, A., 2002. Investigation ofcarbonyl compounds in bottled waters from Poland. WaterResearch 36, 4893–4901.

Nawrocki, J., Swietlik, J., Raczyk-Stanis1awiak, U., Dabrowska, A.,Bi1ozor, S., Ilecki, W., 2003. Influence of ozonation conditions

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 2 2 0 1 – 2 2 0 82208

on aldehyde and carboxylic acid formation. Ozone Scienceand Engineering 25, 53–62.

Nieuwenhuijsen, M.J., Toledano, M.B., Eaton, N.E., Fawell, J.,Elliott, P., 2000. Chlorination disinfection byproducts in waterand their association with adverse reproductive outcomes:a review. Occupational and Environmental Medicine 57, 73–85.

Richardson, S.D., 1998. Drinking water disinfection by-products. In:Mayers, R.A. (Ed.), Encyclopedia of Environmental Analysis andRemediation. John Wiley & Sons, New York, pp. 1398–1421.

Richardson, S.D., Plewa, M.J., Wagner, E.D., Schoeny, R.,DeMarini, D.M., 2007. Occurrence, genotoxicity, andcarcinogenicity of regulated and emerging disinfection by-products in drinking water: a review and roadmap forresearch. Mutation Research 636, 178–242.

Sclimenti, M.J., Krasner, S.W., Glaze, W.H., Weinberg, H.S., 1990.Ozone disinfection byproducts: optimization of the PFBHAderivatization method for the analysis of aldehydes. Presentedat the American Water Works Association, Water QualityTechnology Conference, 11–15 November, San Diego, CA.

1980. Standard Methods For the Examination of Water andWastewater, 15th ed.

Trehy, M.L., Yost, R.A., Miles, C., 1986. Chlorination byproducts ofamino acids in natural waters. Environmental Science andTechnology 20 (11), 1117–1122.

Ueno, H., Nakamuro, K., Moto, T., Sayato, Y., 1995. Disinfectionby-products in the chlorination of organic nitrogencompounds: possible pathways for the formation ofdisinfection by-products. Water Supply 13 (3/4), 171–176.

Weinberg, H.S., Glaze, W.H., Krasner, S.W., Sclimenti, M.J., 1993.Formation and removal of aldehydes in plant that useozonation. Journal of American Water Works Association 85(5), 72–85.

WHO, 2004. Guidelines for Drinking Water Quality. In:Recommendations, third ed., vol. 1. World HealthOrganization, Geneva.

Williams, D.T., LeBel, G.L., Benoit, F.M., 1997. Disinfections by-products in Canadian drinking water. Chemosphere 34 (2),299–316.