ww.sciencedirect.com

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 4

Available online at w

journal homepage: www.elsevier .com/locate/watres

Oxidation ofMicrocystis aeruginosa and Anabaena flos-aquae byozone: Impacts on cell integrity and chlorination by-productformation

Lucila A. Coral a,*, Arash Zamyadi a, Benoit Barbeau a,**, Fatima J. Bassetti c,Flavio R. Lapolli b, Michele Prevost a,***a Ecole Polytechnique de Montreal, Civil, Mineral and Mining Engineering Department, P.O. Box 6079, Station Centre-ville, Montreal,

Quebec, Canada H3C 3A7bUniversidade Federal de Santa Catarina, Sanitary and Environmental Engineering Department, Water Reuse Laboratory,

P.O. Box 476, Florianopolis, Santa Catarina 88040-970, BrazilcChemistry and Biology Department, Universidade Tecnologica Federal do Parana, Sete de Setembro Avenue 3165, Curitiba,

Parana 80230-901, Brazil

a r t i c l e i n f o

Article history:

Received 29 October 2012

Received in revised form

4 March 2013

Accepted 7 March 2013

Available online 18 March 2013

Keywords:

Cyanobacteria

Ozonation

Cell integrity

Dissolved organic carbon

Chlorination by-products

* Corresponding author. Universidade FederLaboratory, P.O. Box 476, Florianopolis, Sant** Corresponding author.*** Corresponding author.

E-mail addresses: lucila_coral@yahoo.coBarbeau), bassetti@utfpr.edu.br (F.J. Bassetti0043-1354/$ e see front matter ª 2013 Elsevhttp://dx.doi.org/10.1016/j.watres.2013.03.012

a b s t r a c t

Pre-ozonation of cyanobacterial (CB) cells in raw water and inter-ozonation of settled water

can cause CB cell damage. However, there is limited information about the level of lysis or

changes in cell properties after ozonation, release of intracellular compounds and their

contribution to the formation of disinfection by-products (DBPs). This study aims to: (1)

assess the extent of the pre-ozonation effects on CB cell properties; (2) determine the CT

(ozone concentration � detention time) values required for complete loss of cell viability;

and (3) study the DBPs formation associated with the pre-ozonation of cyanobacterial cells

in laboratorial suspensions. To these ends, both Microcystis aeruginosa and Anabaena flos-

aquae suspensions were prepared at concentrations of 250,000 cells mL�1 and

1,500,000 cells mL�1 and were subjected to ozone dosages of 0.5, 2.0 and 4.0 mg L�1 at pH 6

and pH 8. A quick and complete loss of viability was achieved for both CB species after

exposure (CT) to ozone of <0.2 mg min L�1, although no significant decrease in total cell

numbers was observed. Maximum dissolved organic carbon (DOC) releases of 0.96 mg L�1

and 1.63 mg L�1 were measured after ozonation of 250,000 cells mL�1 of M. aeruginosa and

A. flos-aquae, respectively. DOC release was found to be pH and ozone dose dependent.

Ozonation of CB cells increased formation of trihalomethanes (THM) and haloacetic acids

(HAA), mainly for suspensions of A. flos-aquae at pH 8 (by 174% and 65% for THM and HAA

respectively). Utilities considering using ozone for oxidising CB cells should weigh out the

benefit of CB control with the potential increased formation of chlorinated DBPs.

ª 2013 Elsevier Ltd. All rights reserved.

al de Santa Catarina, Sanitary and Environmental Engineering Department, Water Reusea Catarina 88040-970, Brazil. Tel.: þ55 48 3721 7750.

m.br (L.A. Coral), arash.zamyadi@polymtl.ca (A. Zamyadi), benoit.barbeau@polymtl.ca (B.), frl@ens.ufsc.br (F.R. Lapolli), michele.prevost@polymtl.ca (M. Prevost).ier Ltd. All rights reserved.

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 42984

1. Introduction 2012b). Recent studies in the literature demonstrate the

The presence of cyanobacteria is ubiquitous in the environ-

ment. Cyanobacteria has been reported inwater sources, such

as lakes and reservoirs (Oberholster et al., 2006; Carvalho

et al., 2008; Somek et al., 2008; Graham et al., 2010), where

they may cause drastic changes in turbidity, pH, dissolved

oxygen and taste and odour. These variations in water quality

can negatively impact treatment processes used for drinking

water supply. Consequently, cyanobacterial blooms pose a

great challenge for drinking water production (Merel et al.,

2010; Zamyadi et al., 2012a).

Increases in cell concentrations and metabolic substances

produced by cyanobacteria, referred to as algogenic organic

matter (AOM), are responsible for a series of disturbances in

water treatment processes (Henderson et al., 2008b). AOM

includes extracellular organic matter (EOM) and intracellular

organic matter (IOM). EOM results from the metabolic activity

of cells during exponential and stationary growth phases

whereas IOM can be release in the aquatic environment

following cell lysis. The latter can be induced by aging of the

algae population (Pivokonsky et al., 2006; Henderson et al.,

2008a) and by water treatment processes such as oxidation

(Plummer and Edzwald, 2002; Huang et al., 2009).

Although conventional treatment processes (coagulation/

flocculation, sedimentation/flotation and sand filtration) are

effective in removing intact cyanobacteria cells (Chow et al.,

1999; Teixeira and Rosa, 2006), the massive presence of algae

and cyanobacteria may reduce cell-removal efficiency

(Zamyadi et al., 2012b). Pre-oxidation, especially with chlorine,

ozone or potassium permanganate, has often been used as a

pre-treatment ahead of the conventional process. These stra-

tegies may increase the overall removal of cyanobacteria cells

by conventional treatment (Widrig et al., 1996; Plummer and

Edzwald, 2002; Chen et al., 2009; Ma et al., 2012a). However,

the use of oxidants can also cause damage to cyanobacteria

cells and promote the release of IOM, including intracellular

metabolites such as MIB (2-methyl isoborneol), geosmin or

cyanotoxins (Daly et al., 2007; Lin et al., 2009; Miao and Tao,

2009; Zamyadi et al., 2010, 2012a; Ma et al., 2012b). The release

of IOM increases dissolved organic carbon (DOC) which may

negatively impact the coagulationprocess (Maet al., 2012b) and

contribute to DBPs formation (Plummer and Edzwald, 2001;

Nguyen et al., 2005; Hong et al., 2008; Huang et al., 2009; Fang

et al., 2010a,b; Li et al., 2012; Zamyadi et al., 2012a).

Previous studies have shown that pre-ozonation, similarly

to other oxidants, may cause changes in cell surface charac-

teristics, increases in DOC concentrations and the release of

other intracellular metabolites (Plummer and Edzwald, 2001,

2002;Hoegeret al., 2002;Huanget al., 2008;MiaoandTao, 2009).

However, information about the extent of ozone damage to

various cyanobacterial cells under variable oxidation condi-

tions (i.e., contact time, ozone exposure (CT), and pH) remains

limited. Such information is necessary to evaluate whether

pre-ozonation is a viable option for controlling cyanobacterial

blooms. This is especially important given recent evidence

showing that CBs cells can accumulate in various processes

within treatment plants (especially clarifiers) and that CB

species influence the efficacy of coagulation (Zamyadi et al.,

effectiveness of ozone in different applications aiming at the

removal of algae and cyanobacteria (Cheng et al., 2010; Li et al.,

2011; Wu et al., 2012). For this reason, it is interesting to eval-

uate the ozone damages in cellular structure of these micro-

organisms and the cell response to the oxidant application.

The overall objective of this study is to assess the oxidation

kinetics of cyanobacterial cells by ozone and the contribution

of algal organicmatter to DBPs formation under various ozone

operational conditions. More specifically, this research (1) de-

velops a kinetic model to describe the reactivity of Microcystis

aeruginosa andAnabaena flos-aquaewith ozone, (2) assesses the

impact of ozonation on cell characteristics (morphology and

membrane integrity) and the release of IOM for increasing

ozone exposures (CT), and (3) evaluates the release of THMand

HAA precursors after pre-ozonation of cyanobacterial cells.

2. Experimental procedures

2.1. Cell culture and harvesting

A toxic strain ofM. aeruginosa (CPCC 299) and a nontoxic strain

of A. flos-aquae (CPCC 64), which were obtained from the Ca-

nadian Phycological Culture Centre, were cultured in ASM-1

(Gorham et al., 1964) and BG-11 (Rippka et al., 1979) inorganic

growth media, respectively. The cultures were incubated at

26� 1 �Cwith sufficient aerationunder 12h of light-dark cycles

at a light intensity of 70 mmol s�1 m�2. Algae cells in the sta-

tionary phase were separated by filtration (1.2 mm and 10 mm),

re-suspended in borate buffer at pH 7 and enumerated by 400�microscopy, using a Sedgewick-Rafter counting chamber.

2.2. Cell characteristics: lysis, integrity and morphology

To determine cell lysis, cell integrity and cell wall morphology,

three types of microscope techniques were used: (1) cell lysis

wasassessedbyevaluating thedecline in total cell countsusing

optical microscopy in a Sedgewick-Rafter counting chamber

after preservation in a lugol solution (duplicate enumeration/

sample), (2) cell integrity was determined using the BacLight�test (Boulos et al., 1999; Stocks, 2004; Berney et al., 2007), and (3)

cell wall morphology was qualitatively evaluated using a

scanning electronmicroscope (SEM, JSM-7600TFE, JEOL, Japan).

The cell volume for each cyanobacteria was estimated

from cell measures by optical microscopy considering diam-

eter of M. aeruginosa or the width and length of A. flos-aquae.

Measurements revealed an average diameter of 4.5 mm for M.

aeruginosa (approximately 48 mm3 of cell volume) and an

average width and length of 4 and 6.5 mm for A. flos-aquae,

respectively (approximately 75 mm3 of cell volume). Measure-

ments were performed on 30 cells were repeated on several

occasions to yield the average values.

2.3. Ozone experiments

For experiments investigating ozone decay and its impact on

cyanobacterial viability, stockozonesolutions (60e70mgO3 L�1)

were prepared by diffusing gaseous ozone that was produced

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 4 2985

with a bench-top ozone generator (TG-10, Ozone Solutions, Inc,

USA) through a glass flask containing ultrapure water chilled at

4 �C. The concentration of the ozone stock solution and the

dissolved ozone residual in samplesweredeterminedaccording

to the standard colorimetricmethod 4500-O3 (APHA, 2005) using

indigo trisulfonate ( 3600nm ¼ 20,000 M�1 cm�1). Samples were

analysed at 600 nmwith a Varian spectrophotometer (Cary 100,

Varian) in a 1-cm or 2-cm quartz cell. Ozone performance was

evaluated for both M. aeruginosa and A. flos-aquae using three

different applied doses (0.5, 2.0 or 4.0 mg L�1), two initial cell

densities (250,000 and 1,500,000 cell mL�1) and two pH condi-

tions (6 or 8). This experimental design yielded a total of 24

ozonationassays, towhich6additional assays, in theabsenceof

algae cells, were added as controls. Previous experiments with

CB cells without ozone at two pH conditions (6 and 8) showed

there was no difference on cell viability. All experimental pro-

cedures were performed at room temperature of 20 �C.The pH values were adjusted using a borate buffer (ionic

strength: 0.05 M). An ozone decay curve was monitored in a

true-batch reactor for each of the conditions described above

(ozone dose, cell density and pH). The applied ozone doses

were administered by injecting an aliquot of the ozone stock

solution via a syringe into a continuously stirred 2-L glass

reactor containing the sample (initial sample volume between

500 and 1200 mL) and equipped with a floating Teflon lid to

prevent ozone degassing. Ozone residuals were measured

over a 10 min period by collecting 5 mL samples that were

dispensed into 20 mL of indigo (0.02, 1 or 3%) solution. In

parallel experiments, samples were collected for DOC mea-

surements, cell viability and cell counts at predetermined

contact times (0.5, 1.5, 3, 5 and 10min). The samples (sampling

volume of 150 mL) were immediately quenched with sodium

bisulfite (NaHSO3 e 2.2 mg/mg O3) to stop the ozone reaction.

To quantify �OH radical formation, a parallel reactor was al-

ways spiked with 200 mg L�1 (1.28 mM) of para-chlorobenzoic

acid (pCBA). A sub-sample of 1.5 mL was filtered using

0.45 mm pore filter (Millex-HV) and analysed using HPLC. The

column characteristics, mobile phase and detection were the

same as the column characteristics, mobile phase and

detection described by Vincent et al. (2010).

2.4. DOC quantification

DOC samples were filtered through pre-rinsed 0.45 mm hy-

drophilic polyethersulfone membranes (Supor�-450, PALL,

USA). DOC analysis measurements were conducted on a

5310C laboratory total organic carbon analyser (Sievers In-

struments, Inc., GE, USA). Extracellular DOC samples were

obtained by filtration immediately after ozone exposure. For

the measurements of total DOC values (extracellular and

intracellular DOC), the samples were subjected to three cy-

cles of freezing and thawing to achieve a complete disruption

of cells and the release of intracellular substances (Moreno

et al., 2004; Qian et al., 2010) Preliminary microscopic obser-

vations of disrupted cells confirmed the efficiency of the

freeze-thaw procedure in releasing cell-bound toxins. DOC

release during ozonation results were normalised by dividing

the DOC values by the ultimate filterable DOC released (DOC

extracellular), which was observed at the highest pH and

ozone dosage.

2.5. Formation of DBPs precursors

The potential for pre-ozonation to increase DBPs formation

was evaluated by chlorinating raw and pre-ozonated water

samples. An amount of free chlorine that was sufficient to

achieve a free chlorine residual of 1.0 � 0.2 mg L�1 after a 24-

hr chlorine contact time at 22 �C was added to the samples.

At the end of the incubation period, samples for THM and

HAA were quenched with excess of ammonium sulphate.

Measurements of free chlorine residuals were performed

using the DPD colorimetric method 4500C1-G (APHA, 2005).

Samples were analysed for four THMs (chloroform, bromo-

form, bromodichloromethane, dibromochloromethane)

referred as total THM (TTHM) and six HAA (monochloroacetic

acid, dichloroacetic acid, trichloroacetic acid, monobromo-

acetic acid, dibromoacetic acid and bromochloroacetic acid)

referred as HAA6. THMs were extracted by liquid/liquid

extraction with hexane as described in Standard Methods

6232B (APHA, 1998). HAA compounds were extracted by

liquid/liquid extraction with methyl tert-butyl ether (MtBE)

followed by derivation with acidic methanol in accordance

with USEPA Method 552.2 (USEPA, 1995). All DPBs were ana-

lysed on a gas chromatograph (CP-3800, Varian) equipped

with an electron capture detector (GC/ECD) (Column charac-

teristics: capillary column RTX-5MS with 30 m � 0.25 mm ID

and 0.25 mm film thickness).

2.6. CT calculations: molecular and free radicalexposures

For each specific contact time, the molecular ozone exposure

(CT value e ozone concentration � detention time),

expressed as mg min L�1, was calculated by numerical inte-

gration using the trapezoidal method, which calculates the

area under the decay curve. To validate the estimated nu-

merical CT, they were compared to modelled CT values,

which were obtained by assuming a pseudo-first-order decay

kinetic (k0) in (Eq. (1))

CTO3 ¼ C0

k0 ½1� expð�k0$tÞ� (1)

where C0 ¼ initial ozone residual (mg L�1) determined from

the exponential fit of the relationship between the ozone re-

sidual and time (min); and k0 ¼ the ozone pseudo-first-order

decay constant (min�1).�OH radical exposure (CT�

OH) was estimated using the

reaction-rate constant for �OH radical (kOH,pCBA ¼ 5.2

� 109 M�1s�1) and a second-order kinetic such that

dðpCBAÞdt

¼ �kOH$pCBA$½pCBA�$½$OH� (2)

where kOH.pCBA is the pCBA rate constant; [pCBA] is the pCBA

concentration; and [�OH] is the free hydroxyl radical

concentration.

Subsequently, �OH exposure is defined as:

CTOH ¼Z

$OH$dt ¼Ln

� ½pCBA�½pCBA�o

�kpCBA

(3)

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 42986

2.7. Kinetic modelling

Ozone decay was modelled as a two-step process: immediate

demand followed by an n-th order kinetic. The immediate

ozone demand characterises the amount of ozone consumed

in the first stage of the oxidation process (measured after a

few seconds) (Park et al., 2001). In this study, the immediate

ozone demand was evaluated as the ozone consumed after

15 s of contact time (our earliest measurement). A step-

forward general regression model was developed to predict

ozone demand as a function of ozone dose, pH, organism type

and DOC. Ozone decay after immediate demand was

modelled using two alternative strategies. Initially, pseudo-

first-order kinetics were fitted successively to each individ-

ual decay curve (Eq. (4)) as

CCo

¼ expð�k$tÞ (4)

where C and C0 are the final and initial concentrations of the

target compound (mg L�1), respectively; k is the pseudo-first-

order reaction rate constant (min�1); and t is the time (min).

As an alternative approach, a generalised empirical kinetic

rate law (Eq. (5)) was developed as followed:

dO3

dt¼ �k$ðO3Þn$

�OH��m$ðDOC TÞp (5)

where O3 is the ozone residual; t is time, OH� is the hydroxide

concentration, DOC_T is the total concentration of organic

carbon; and n, m, p and k are fitting parameters.

Assuming that pH and DOC_T are constant during ozona-

tion, Eq. (5) can be integrated into Eq. (6). Because no mea-

surement is available to represent the reactivity of DOC

according to time, DOC_T is assumed to be constant to

simplify the calculations. In addition, DOC represents a series

of organic compounds rather than a single solute.

O3 ¼"

1

ðO3Þn�10

þ k� t� �OH��m � ðDOC TÞp

#1�n

(6)

Equation (6) was fitted separately for the ozone assays

that were conducted for each of the two organisms as well

as for the entire dataset (i.e., combined data for both

organisms).

3. Results and discussion

3.1. DOC release

Ozonation of CB cell suspensions led to a substantial increase

in DOC. For example, Fig. 1 shows increases in DOC

(0.1e1.5 mg C L�1) for M. aeruginosa and A. flos-aquae after

ozonation using 0.5, 2.0 or 4.0mg L�1 of ozone for a cell density

of 250,000 cells mL�1. The large amount of extracellular DOC

immediately after ozone exposure (0.5 min) can be related to

two mechanisms: a rapid reaction between ozone and cell

biomass, with a direct formation of DOC; and/or an indirect

reaction from cell lysis, leading the release of intracellular

compounds (Muller et al., 2003; Hammes et al., 2007; Ramseier

et al., 2011).

As shown in Fig. 1, higher ozone doses increased the pro-

portion of DOC release. After 10 min of contact time, an in-

crease from 0.39 mg L�1 of DOC with 0.5 mg O3 L�1 to

0.96 mg L�1 after 4 mg O3 L�1, and from 0.67 mg L�1 of DOC

with 0.5 mg O3 L�1e1.58 mg L�1 after 4 mg O3 L�1 were

observed forM. aeruginosa andA. flos-aquae, respectively. Miao

and Tao (2009) investigated DOC changes following ozone

oxidation of M. aeruginosa cultures (undisclosed cell concen-

trations). They observed that the release of cell associated

DOC increased from 0.34 to 2.56 mg L�1 with increasing ozone

dosages from 0 to 5 mg O3 L�1.

A greater release of DOC while ozonating A. flos-aquae can

be linked to cell size, which resulted in a greater release of

intracellular substances or a greater quantity of cell biomass

available for oxidation. Also, it can be observed on some oc-

casions (cf.M. aeruginosa at pH 6), that increasing the dosage of

ozone from 0.5 to 2mg L�1 did not result in an increase in DOC

concentration. This may indicate that the internal contents of

cells would have already been completely released at the

lowest ozone dosage. Thus, the greater concentration of DOC

obtained with the higher ozone dosage (4 mg L�1) would be

related to the oxidation of the organic constituents in the cell

structure. Similarly, for A. flos-aquae at pH 8, the increase in

ozone dose from 2 to 4 mg L�1 resulted in no increase in DOC

concentration, suggesting that all the organic material of cell

would have been completely released or oxidized with an

ozone dose of 2 mg O3 L�1.

The increase in extracellular DOC can also be correlated

with the pH values. Our results indicate that the amount of

DOC released at pH 8 was greater than at pH 6. Fig. 2 presents

the normalised DOC release (expressed in % of the maximal

DOC released) for both organisms as a function of molecular

ozone exposure (CTO3) and/or hydroxyl radical exposures

(CT�OH). In Fig. 2a, there is no obvious difference in the nor-

malised DOC release between M. aeruginosa and A. flos-aquae.

The higher DOC release, as observed in Fig. 1, was essentially

caused by the higher biomass load of the larger A. flos-aquae.

However, for both organisms, there is an important difference

caused by the pH. DOC release at pH 6 plateaued at 30e50% of

the ultimate DOC release observed at pH 8. Interestingly,

plotting CT release with respect to CT�OH (Fig. 2b) provided a

better description of the entire dataset. We infer that free

radical plays an important role in the oxidation of algae cells.

CT�OH is correlated with CTO3; higher CTO3 also leads to a

higher CT�OH. However, it also accounts for the effect of pH

which explains its superiority over CTO3 in predicting DOC

release. Huang et al. (2008) also observed an increase in DOC

when pHwas increased from 5.5 to 9.0. The authors suggested

that this increase was caused by the increased liberation of

intracellular organic matter and cell-wall polysaccharides

during cell ozonation and that the action of pH on DOC release

was regardless of ozone dosages. At alkaline pH, the higher

concentration of hydroxide ion (OH�) promotes ozone

decomposition and the formation of hydroxyl free radicals

(�OH) (Widrig et al., 1996; Von Gunten, 2003). According to

Widrig et al. (1996), �OH radicals oxidation mechanisms are

more prone to degrade saturated aliphatic molecules, which

are largely present in algal-derived organic substances. Thus,

higher concentrations of �OH radicals due to alkaline pH or

higher O3 doses may lead to increased cell damage or

0.0

0.3

0.6

0.9

1.2

1.5

0.0 0.5 1.5 3.0 5.0 10.0

Ext

race

llula

r D

OC

(mg

L-1

)

Time (min)

(a)0.0

0.5

1.0

1.5

2.0

2.5

0.0 0.5 1.5 3.0 5.0 10.0

Ext

race

llula

r D

OC

(mg

L-1

)

Time (min)

(b)

Fig. 1 e DOC release for (a) M. aeruginosa and (b) A. flos-aquae for a density of 250,000 cells mL L1; ozone dose of 0.5, 2.0 and

4.0 mg LL1 and pH of 6 and 8 (MIC: M. aeruginosa, ANA: A. flos-aquae).

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 4 2987

reactions with organic material in the cell and, consequently,

a greater amount of extracellular DOC.

3.2. Impact of CB cells and DOC on ozone decay

A total of 30 ozone decay curves were generated, including 24

reactors with algae and 6 reactors without cells, which served

as controls. Immediate ozone demands are presented first. In

a second step, the ozone decay kinetic was characterized

using a classical pseudo-first-order model. Finally, a general-

isation of ozone kinetic data was accomplished by the devel-

opment of an n-th order model.

3.2.1. Immediate ozone demandImmediate ozone demands varied from 0.00 to 0.82 mg L�1;

higher demands were observed at higher ozone dosages, pH

and DOC (Table 1). Interestingly, when all other variables were

equivalent, immediate ozone demand was not statistically

different ( p < 0.05) for M. aeruginosa and A. flos-aquae.

Consequently, it was possible to predict immediate ozone for

both organismswith a high correlation (R2¼ 0.90) using Eq. (7):

DO

C R

elea

se (

%)

= D

elta

DO

C/D

OC

_U

0 2 4 6 8 10 12 14 16 18 20 22 24

0%

20%

40%

60%

80%

100%A. flos-aquae - pH 6M. aeruginosa

A. flos-aquae - pH 8 M. aeruginosa - pH 8

CT_O3 (mg min L-1)

(a)

Fig. 2 e Normalised DOC release: (a) as a funct

D0 ¼ aþ b�O3Doseþ c� pHþ d�DOC�O3Dose (7)

whereD0 is the immediate ozone demand; and a, b, c and d are

regression coefficients.

An assessment of the ANOVA table for this regression (data

not shown) revealed that the interaction between DOC and

ozone dose was the most important predictor in the equation.

This interaction reflects the basic fact that an increase in

ozone dosage produces large immediate demands in DOC-rich

waters. As for pH, its increase led to higher immediate de-

mands, which is a direct consequence of the increased free

radicals activity (as will be shown later).

3.2.2. Pseudo-first-order kinetic modelFig. 3 presents ozone decay for the various ozonation condi-

tions that were investigated. For each individual ozone decay

curve, a pseudo-first-order model was fitted by linear regres-

sion. Table 2 summarises the regression parameters derived

from this analysis. Although a pseudo-first order model

generally provided a good fit to the experimental data (typical

R2 > 0.95), some conditions clearly exhibited two kinetic

DO

C R

elea

se (

%)

= D

elta

DO

C/D

OC

_U

0,0E

-01

2,0E

-07

4,0E

-07

6,0E

-07

8,0E

-07

1,0E

-06

1,2E

-06

1,4E

-06

1,6E

-06

0%

20%

40%

60%

80%

100%

A. flos-aquae - pH 6M. aeruginosa

A. flos-aquae - pH 8M. aeruginosa - pH 8

Release = 0.72 x (1-exp(-1.11E07*CT_OH))

CT_OH (mg min L-1)

(b)

ion of CTO3 and (b) as a function of CT�OH.

Table 1 e Immediate ozone demands. R2 [ 0.90.

pH 6 pH 8

Ozonedose(mg L�1)

Immediatedemand(mg L�1)

Ozonedose

(mg L�1)

Immediatedemand(mg L�1)

0.5 0.00 0.5 0.26

2.0 0.22 2.0 0.50

4.0 0.50 4.0 0.82

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 42988

regimes (e.g., Fig. 3c). Half-life values varied from as low as

0.21 min to as high as 64 min (Table 2). As expected, higher pH

and initial cell densities led to lower half-lifes.

Under identical ozonation conditions, higher ozone decay

were always observed for A. flos-aquae than for M. aeruginosa.

This effect may be due to the fact that the latter algae is larger

(6.5 mm in length � 4 mm in width) than the former algae

(4.5 mm in diameter). Because the initial cell density was

equivalent for both organisms, A. flos-aquae brought a higher

organic load. On average, DOC concentrations were 90%

higher during assays performed with A. flos-aquae. To better

distinguish the impacts of pH, organic carbon concentration

and ozone residual on ozone decay kinetics, a generalised

empirical rate law was fitted to the entire dataset. This will be

the topic of the next section.

3.2.3. N-th order modelEquation (7) was adjusted for the four datasets described in

Table 3: the ultrapure water dataset, M. aeruginosa dataset, A.

0.0

1.5

3.0

4.5

0.0 2.0 4.0 6.0 8.0 10.0

Ozo

ne

Res

idua

l (m

g L

-1)

Time (min)

250,000 cells mL-1

pH 6

0.0

1.5

3.0

4.5

0.0 2.0 4.0 6.0 8.0 10.0

Ozo

ne R

esid

ual (

mg

L-1

)

Time (min)

1,500,000 cells mL-1

pH 6

Fig. 3 e Ozone decay curves in Milli-Q water without CB cells a

(ANA).

flos-aquae dataset and the combination ofM. aeruginosa and A.

flos-aquae into one dataset. The models provided a good fit to

the data (R2 ¼ 0.87e0.92). Ozone decomposition in ultrapure

water proved to be second-order with respect to dissolved

ozone, whereas reaction order for the hydroxide concentra-

tion was 0.53. Mizuno et al. (2007) developed a similar kinetic

model to describe ozone decay in ultrapure water. They also

observed second-order kinetics with respect to dissolved

ozone. However, the order of reaction for hydroxide (0.73) was

slightly higher than that obtained in this study, which is most

likely due to the fact that the Mizumo et al. model was

developed for a wider range of pH (4.0e7.8).

In the presence of algal organic matter (AOM), ozone decay

is always second-order for O3 and almost first-order for DOC.

As for pH, the orders of reaction (0.18e0.34) were lower than in

ultrapure waters, which indicates that the impact of pH on O3

decay was reduced in the presence of AOM. The reactivity of

M. aeruginosa (k ¼ 37.6 M(2.4) (mg C/L)�1.2 min�1) was also

higher than the reactivity forA. flos-aquae (k¼ 2.46M(2.3) (mg C/

L)�1.0 min�1). Such results indicates that the lower half-life

values observed while oxidizing A. flos-aquae were caused by

increased DOC concentrations rather than increased DOC

reactivity. As discussed earlier, higher DOC was promoted by

the larger biovolume of A. flos-aquae as compared with M.

aeruginosa.

To directly compare ozone reactivity of both algae, the

predictive models for immediate ozone demand and ozone

decay were used to predict ozone decay profiles for a common

ozonation condition (DOC_T ¼ 2 mg L�1, dose ¼ 2 mg L�1) and

0.0

1.5

3.0

4.5

0.0 2.0 4.0 6.0 8.0 10.0

Ozo

ne R

esid

ual (

mg

L-1

)

Time (min)

250,000 cells mL-1

pH 8

0.0

1.5

3.0

4.5

0.0 2.0 4.0 6.0 8.0 10.0

Ozo

ne R

esid

ual (

mg

L-1

)

Time (min)

1,500,000 cells mL-1

pH 8

nd in the presence of M. aeruginosa (MIC) and A. flos-aquae

Table 2 e First-order apparent rate constants (k) for ozone decay.a

CB specie pH O3 dose (mg L�1) 250,000 cells mL�1 1,500,000 cells mL�1

k (s�1) Half-life (min) k (s�1) Half-life (min)

No cell 6 0.5 3.52 � 10�4 (0.95) 32.9 3.52 � 10�4 (0.95) 33.0

2.0 1.82 � 10�4 (0.78) 63.6 1.82 � 10�4 (0.78) 64.0

4.0 3.17 � 10�4 (0.98) 36.4 3.17 � 10�4 (0.98) 36.0

8 0.5 2.02 � 10�3 (0.99) 5.7 2.02 � 10�3 (0.99) 5.7

2.0 2.38 � 10�3 (0.99) 4.9 2.38 � 10�3 (0.99) 4.9

4.0 2.87 � 10�3 (0.97) 4.0 2.87 � 10�3 (0.97) 4.0

M. aeruginosa 6 0.5 6.57 � 10�3 (0.97) 1.8 9.85 � 10�3 (0.96) 1.2

2.0 1.53 � 10�3 (0.94) 7.5 6.30 � 10�3 (0.97) 1.8

4.0 8.17 � 10�4 (0.99) 14.2 2.10 � 10�3 (0.97) 5.5

8 0.5 1.57 � 10�2 (0.96) 0.7 1.36 � 10�2 (0.98) 0.8

2.0 1.18 � 10�2 (0.99) 1.0 2.18 � 10�2 (0.99) 0.5

4.0 8.65 � 10�3 (0.98) 1.3 2.74 � 10�2 (0.99) 0.4

A. flos-aquae 6 0.5 1.48 � 10�2 (0.99) 0.8 4.56 � 10�2 (0.96) 0.2

2.0 4.68 � 10�3 (0.98) 2.5 3.19 � 10�2 (0.99) 0.4

4.0 1.75 � 10�3 (0.99) 6.6 2.04 � 10�2 (0.99) 0.6

8 0.5 1.71 � 10�2 (0.95) 0.7 5.52 � 10�2 (0.96) 0.2

2.0 1.56 � 10�2 (0.99) 0.7 3.81 � 10�2 (0.99) 0.3

4.0 1.30 � 10�2 (0.99) 0.9 3.91 � 10�2 (0.99) 0.3

a Correlation coefficients (R2) presented in parentheses.

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 4 2989

variable pH conditions (Fig. 4). Although ozone kinetics of A.

flos-aquae are slightly faster than the ozone kinetics for M.

aeruginosa, pooling the data for both organisms provides good

predictions of ozone decay under equivalent ozonation con-

ditions (pH, applied dose and DOC_T), as illustrated in Fig. 4.

3.3. Effect of ozone on the integrity of M. aeruginosaand A. flos-aquae cells

The impact of ozonation on cellular morphology and cell

integrity was verified by SEM and BacLight� methods,

respectively. No loss of cell integrity was observed due to the

direct addition of cells in buffered water. However, as shown

in Fig. 5aed, an immediate loss of integrity after exposure to

ozone was observed for both tested species. All cells turned

red/orange, even for the lowest contact time (30 � 5 s) and

ozone dosage of 0.5 mg L�1. These results indicate immediate

cell membrane damage at CT values �0.2 mg min L�1 for all

ozonation experiments. While carbohydrates and fatty acids

react only slightly with ozone, amino acids, proteins and

nucleic acids all react very quickly with this oxidant (Bablon

et al., 1991). Both DNA and RNA can be degraded by ozone

Table 3 e Estimated parameters to determine the behaviour of

Estimatedparameters

Ultrapure waterR2 ¼ 0.92

O3 n 2.00

Rate constant k 131.50

pH effect m 0.53

DOC_T effect p e

MIC: M. aeruginosa, ANA: A. flos-aquae.

a M(2.4) (mg C/L)�1.2 min�1.

b M(2.3) (mg C/L)�1.0 min�1.

c M(2.3) (mg C/L)�1.0 min�1.

attack (molecular ozone or free radicals), even with the cell

remaining intact (Theruvathu et al., 2001; Cataldo, 2005). With

the increase in ozone dosages and contact time, propidium

iodide binding was progressively lost as the target binding

sites were oxidised.

Like other oxidants agents, ozone first attacks the cell wall

and plasma membrane glycoproteins, glycolipids or certain

amino acids. Rapid cell death after ozonation can be attributed

to changes in cellular permeability followed by cell lysis

(Greene et al., 1993; Morato et al., 2003). However, due to the

high oxidation potential of ozone and because ozone further

reacted with the nuclear material, lysis is not the primary

inactivation mechanism; rather it is a consequence of high

applied oxidant dosages (Greene et al., 1993). The cell counts

performed during this study indicated that no lysis of A. flos-

aquae occurred for all the ozonation conditions investigated.

No lysis was observed at pH 8 for M. aeruginosa; lysis was only

significant at pH 6 for the two highest ozone dosages. A total of

32% and 41% of cells were lysed when ozone dosages

increased from 2mg L�1 and 4mg L�1, respectively. The higher

ozone residual concentrations at pH 6 may have favoured cell

lysis, although a more visible disruption of external cell

ozone decay.

MICR2 ¼ 0.90

ANAR2 ¼ 0.87

MIC þ ANAR2 ¼ 0.88

2.02 2.08 2.04

37.6a 2.46b 13.85c

0.34 0.18 0.28

1.19 1.03 1.03

0.0

0.5

1.0

1.5

2.0

2.5

0.0 2.0 4.0 6.0 8.0 10.0

Ozo

ne R

esid

ual (

mg

L-1

)

Time (min)

Milli-Q_pH6 Milli-Q_pH8 MIC_pH6

MIC_pH8 ANA_pH6 ANA_pH8

MIC+ANA_pH6 MIC+ANA_pH8

Fig. 4 e Predicted ozone decay curves for Milli-Q water

without cells and cells of M. aeruginosa, A. flos-aquae and

M. aeruginosa D A. flos-aquae at pH 6 and 8 (ozone dose:

2 mg O3 LL1; DOC: 2 mg LL1; MIC: M. aeruginosa, ANA: A.

flos-aquae).

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 42990

morphology was observed at pH 8 by SEM. We suggest that�OH radicals have larger impacts on the cell. However, the

higher concentration of extracellular DOC at pH 8 also acts as

scavengers of �OH radicals, reducing the availability of �OH

radicals to attack the target, as observed by Onstad et al. (2007)

that verified a reduced efficiency of MC-LR lysis due to high

concentration of DOC. In Fig. 5, SEM microphotographs are

provided for conditions before (Fig. 5e and g) and after (Fig. 5f

and h) exposure to ozone. Although there was no cell lysis

verified for M. aeruginosa after 10 min of contact time and

Fig. 5 e Impact of ozonation on cell viability by BacLight� (aed

and (b) after ozonation (pH 6 with 0.5 mg O3 LL1 e 30 s of conta

ozonation (d) (pH 6 with 0.5 mg O3 LL1 e 30 s of contact time); (e)

4.0 mg O3 LL1 e 10 min of contact time); (g) A. flos-aquae cells be

contact time).

4.0 mg L�1 of ozone at pH 8 (Fig. 5f), a visible alteration in cell

wall morphology was observed. In the other ozonation con-

ditions, changes in cell wall morphology were less notable.

After only 30 s of contact time with 2 mg L�1 of ozone, A. flos-

aquae cells appeared more impacted by ozone than M. aeru-

ginosa with visible changes (Fig. 5h) being noted in cell wall

morphology. However, no reduction in cell number was

observed.

Miao and Tao (2009) also evaluated the effects of ozone on

M. aeruginosa cells morphology. Doses of 1, 3 and 5 mg L�1 of

ozone were applied to a cyanobacteria suspension. Removals

up to 91% of Chl-a were obtained for an ozone dose of 5mg L�1

within 60 min of contact time. The authors verified that ozone

dose in the range of 1e5 mg L�1 resulted in cell wall alter-

ations, with complete cell lysis after treatment with 3 mg L�1

of ozone, an opposed to the results observed in this study for a

similar ozone dosage (4 mg O3 L�1). Lin et al. (2009) also

observed a greater susceptibility of the Anabaena genus to

oxidation by chlorine as compared with Microcystis. In com-

parison with chlorine experiments, ozone exhibited a more

effective impact on the integrity of CB cells. For example,

Zamyadi et al. (2012a) verified the loss of integrity for several

cyanobacteria species after the chlorination experiments at 2

and 5 mg Cl2 L�1 using higher contact times. A CT higher than

31 mg min L�1 was needed to disrupt the integrity (measured

with the FDAmethod) of more than 99% ofM. aeruginosa cells,

whereas a CT of approximately 8 mg min L�1 had an equiva-

lent impact on Anabaena circinalis, Cylindrospermopsis racibor-

skii, and Anabaena issatsckenka cells. The same behaviour was

observed by Daly et al. (2007) while chlorinating M. aeruginosa

cells. Under the conditions evaluated in this study, a similar

ozone dose (4 mg O3 L�1) was needed to lyse only 41% of M.

aeruginosa cells, which implies that ozone is more effective

than chlorine to render CB cells non viable at an equivalent

) and SEM (eeh). (a) M. aeruginosa cells before ozonation

ct time); (c) A. flos-aquae cells before ozonation and after

M. aeruginosa cells before and after ozonation (f) (pH 8 with

fore and after ozonation (h) (pH 6 with 2 mg O3 LL1 e 30 s of

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 4 2991

oxidant exposure (CT). The results from this study are

consistent with the results in the literature: release of DOC

and internal metabolites precede cell lysis (Plummer and

Edzwald, 2001; Miao and Tao, 2009). Our results indicate that

very low ozone exposure may be sufficient to induce cell

damage and loss of integrity.

3.4. Formation of disinfection by-products from CB cells

Several studies have quantitatively evaluated the reactivity of

algal and cyanobacteria cells and their metabolites as pre-

cursors of chlorinated DBPs (Plummer and Edzwald, 2001;

Hong et al., 2008; El-Aty et al., 2009; Huang et al., 2009;

Zamyadi et al., 2012a). Pre-ozonation of cyanobacteria cells

release intracellular compounds which may contribute to the

pool of chlorinated DBPs precursors (assuming that post-

chlorination is used as a final treatment). To evaluate their

reactivity, ozonated samples spiked with the lower concen-

tration of algae cells (250,000 cell mL�1) were chlorinated and

subjected to a THM and HAA6 analyses.

Our results indicated that the predominant chlorinated

DBPs formed were chloroform (CHCl3) and dichloroacetic acid

(DCAA) forM. aeruginosa, and CHCl3, DCAA and trichloroacetic

acid (TCAA) for A. flos-aquae (Supplementary Data e Table SD-

1). Fig. 6 presents the total THM (TTHM) and HAA (THAA)

concentrations measured for increasing ozone dosages.

Higher concentrations of THM were measured when

compared with HAA. A study from Hong et al. (2008) indicated

that THM and HAA formation is directly related to cyano-

bacteria biochemical cellular composition. The formation of

THM tends to be higher than the formation of HAAwhen there

is a greater proportion of lipids in the cell composition. The

formation of HAA, in turn, is related to the amount of proteins.

According to Hong et al. (2008) the amount of proteins is

proportionally higher in M. aeruginosa than in Anabaena sp.

cells, but Anabaena sp. cells present a higher amount of lipids

than M. aeruginosa cells. This finding could explain the higher

A. flos-aquae THM formation observed in this study.

Higher ozone dosages resulted in higher DBPs formation,

except for M. aeruginosa at pH 6 for which no THM could be

detected even after 4 mg L�1 of ozone. Higher THM and HAA

0.0

0.5

1.0

1.5

2.0

0

10

20

30

40

50

60

70

No ozone

0.5 mg/L

O3

2.0 mg/L

O3

4.0 mg/L

O3

No ozone

0.5 mg/L

O3

2.0 mg/L

O3

4.0 mg/L

O3

Extracellular D

OC

(mg L

-1)

DB

P c

onc.

(µg

L-1

) THM

HAA

Extracellular DOC

pH 6 pH 8

No ozone

0.5 mg O

3L

-1

2.0 mg O

3L

-1

4.0 mg O

3L

-1

0.5 mg O

3L

-1

2.0 mg O

3L

-1

4.0 mg O

3L

-1

No ozone

(a)

Fig. 6 e TTHM and HAA6 formation following ozonation of (a) M

of contact time with 0.5, 2.0, and 4.0 mg O3 LL1 for 250,000 cell

1.0 ± 0.2 mg LL1 after a 24-hr of contact time at 22 �C).

formation were observed for A. flos-aquae, a conclusion that is

consistent with the higher organic carbon extracellular after

ozonationdue to their larger cellular structure.Althoughhigher

DOC generally led to a higher DBPs formation, DBPs reactivity

was both pH and organism-dependant. The following

maximumyieldswere obtained forM. aeruginosa at pH 8: 11 mg/

mgC (TTHM¼ 38mg L�1) and4 mg/mgC (HAA6¼ 14mg L�1). ForA.

flos-aquae, maximum yields of 16 mg/mg C (TTHM ¼ 63 mg L�1)

and 6 mg/mgC (HAA6¼ 24 mg L�1) were calculated for TTHMand

THAA at pH 8, respectively. Ratios of TCAA/DCAA for A. flos-

aquae ranged from0.14 to 0.21at pH6and from0.30 to0.48atpH

8, with lower ratios being achieved for higher ozone doses.

These results indicate that the DOC produced by A. flos-aquae

was more prone to DCAA than TCAA formation (Nguyen et al.,

2005; El-Aty et al., 2009), which can be related to the lower level

of aromatic carbon compared to the amino acid found in CB

EOM (Hong et al., 2008, 2009). Due to the higher proportion of

proteinaceous material in EOM, the dominance of DCAA is

intensified (Huang et al., 2009). As noted in Table SD-1

(Supplementary Data), increasing ozone dose resulted in a

slight increase of DCAA and decrease of TCAA at pH 8. This

behaviour is most likely related to changes in hydrophobicity

following their oxidation by ozone which increases the hydro-

philic fraction (�Swietlik et al., 2004) more prone to DCAA for-

mation. Hong et al. (2013) observed in their study that

hydrophobic organic molecule fractions tended to be more

important as precursors of THM and TCAA, while the hydro-

philic fractions resulted in higher concentrations of DCAA.

The influence of pH on THM formation is the result of two

mechanisms. Inonemechanism, ozonationat ahigher pH led to

the release of increasedDOC (as shown in Fig. 1). In addition, it is

well-known that a higher pH increases free chlorine reactivity

with natural organic matter. In a study that assesses the for-

mation of DBPs from carbohydrates, Navalon et al. (2008)

observed that basic pH values favour all chlorine reactivity

pathways. Higher pH values favours THM formation (Liang and

Singer, 2003; Navalon et al., 2008; Fang et al., 2010a) but can

reduce the formation of TCAA (Fang et al., 2010a). DCAA for-

mation is normally independent of pH (Liang and Singer, 2003).

This explainswhy the impact of basicpHonHAAformationwas

less than the impact of pH on THM formation. In that case,

0.0

0.5

1.0

1.5

2.0

0

10

20

30

40

50

60

70

No ozone

0.5 mg/L

O3

2.0 mg/L

O3

4.0 mg/L

O3

No ozone

0.5 mg/L

O3

2.0 mg/L

O3

4.0 mg/L

O3

Extracellular D

OC

(mg L

-1)

DB

P c

onc.

(µg

L-1

)

THM

HAA

ExtracellularDOC

pH 6 pH 8

No ozone

0.5 mg O

3L

-1

2.0 mg O

3L

-1

4.0 mg O

3L

-1

0.5 mg O

3L

-1

2.0 mg O

3L

-1

4.0 mg O

3L

-1

No ozone

(b)

. aeruginosa (b) and A. flos-aquae at pH 6 and 8 using 10 min

s mLL1 (chlorination conditions: chlorine residual of

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 42992

higherHAAformationatpH8canbeconsideredadirect resultof

the increase in the availability of DBPs (DOC and other intracel-

lular compounds) precursors under this ozonation condition.

A comparative evaluationof thepotential formationofDBPs

from EOM and CB cells was not evaluated in this study. How-

ever, some of the observations on EOM were made on the

accumulated by products during long term cell culture which

raises the representativity of these findings in a real bloom

(2012a).Other results obtainedbyLiuet al. (2011) inchlorination

studies with M. aeruginosa indicate that only 22% of total THM

formedwas derived from EOM in amixture including cells and

EOM. In a study with M. aeruginosa cells, Fang et al. (2010b)

evaluated the contribution of EOM, IOM and cells to the DBPs

formation. They observed that EOM formed lower concentra-

tions of chloroform and DCAA than IOM and CB cells, which

presented a similar contribution to the final total concentration

ofDBPs formed.All thesefindings confirmthat pre-oxidationof

CB cells can impact significantly the formation of chlorination

by-products in certain conditions by increasing the precursor’s

pool, especially when cell concentrations are elevated. How-

ever, in many cases, background DBPs precursors level have

been shown to be the main factor affecting the overall forma-

tionofDBPs innaturalwatersandthekey factor limiting theuse

of pre-chlorination (2012a). In the case of ozonation, treatment

conditions can enhance the reactivity and DBPs yield of cell-

bound and cell-released precursors. When considering pre-

ozonation, the contribution of CB cells to the pool of THM and

HAA precursor should be considered carefully, especially if no

additional treatment is capable of removing the released DBPs

precursors after pre-ozonation.

4. Conclusion

A low-ozone dose (0.5 mg L�1) was sufficient to cause a com-

plete loss of integrity ofM. aeruginosa and A. flos-aquae cells, for

both cell densities and pH values investigated. Although no

significant cell lysis was observed for most ozonation condi-

tions, the damage on cell structure after ozonation resulted in

an increase inDOC.HigherDOCreleasewasobserved forA.flos-

aquae, an observationwhichwas attributed to their differences

in biovolume. DOC release increased with ozone dose and pH.

The greater DOC release at pH 8 suggests that the attack of free

radicals is an important contributor to this process. Kinetic

analysis indicated that presence of CB cells in water can result

in a considerable immediate ozone demand and this behaviour

is dependent on ozone dose, DOC concentration and pH, but

was independent of the type of CB species tested. Ozone decay

was influenced by CB species and pH values, with the fastest

ozone decay observed for A. flos-aquae cells at pH 8.

DOC release from CB cells was confirmed to contribute to

DBPs formation. The ozone dose, pH and CB species were the

most important predictors of DBPs formation. A greater

amount of THMandHAAwasquantified forA. flos-aquae and at

pH 8, which may be related to the greater amount of DOC

release and the cell composition of this algae. When higher

ozone doses were applied, greater concentrations of THM and

HAA were quantified.

Pre-ozonation is an attractive solution for CB impacted

source waters. It can prevent CB cell accumulation within the

plant (clarifiers and filters) thus reducing the risk of toxin

release during treatment and facilitating the disposal of

sludge. However, it is important to take into consideration the

effect of pre-ozonation on DOC andDBPs formationwhich can

be significant under certain ozonation conditions (high pH,

high CB cell concentrations). Thus, the benefits of CBs control

must be weighed against potential significant increases in

DBPs formation considering the water quality at a given site.

Acknowledgements

This study was financially supported by Natural Sciences and

Engineering Research Council of Canada (NSERC) Industrial

Chair on Drinking Water at Ecole Polytechnique de Montreal,

Brazilian Coordination for the Perfection of Higher Level

Personnel (CAPES) and BrazilianNational Council for Scientific

and Technological Development (CNPq). The assistance of

Jacinthe Mailly, Julie Philibert, Marcelin Fosting, Yves Fon-

taine, and Mireille Blais is gratefully acknowledged. The In-

dustrial Chair includes the City of Montreal, John-Meunier Inc.

and the City of Laval.

Appendix A. Supplementary data

Supplementary data related to this article can be found at

http://dx.doi.org/10.1016/j.watres.2013.03.012.

r e f e r e n c e s

American Public Health Association (APHA), 1998. StandardMethods for the Examination of Water and Wastewater,twentieth ed. American PublicHealth Association (AWWA) andthe Water Environment Federation, Washington, DC, USA.

American Public Health Association (APHA), 2005. StandardMethods for the Examination of Water and Wastewater,twentyfirst ed. American Public Health Association (AWWA)and the Water Environment Federation, Washington, DC,USA.

Bablon, G., Bellamy, W.D., Bourbigot, M.-M., Daniel, F.B., Dore, M.,Erb, F., Gordon, G., Langlais, B., Laplanche, A., Legube, B.,Martin, G., Masschelein, W.J., Pacey, G., Reckhow, D.A.,Ventresque, C., 1991. In: Langlais, B., Reckhow, D.A.,Brink, D.R. (Eds.), Ozone in Water Treatment: Application andEngineering. Lewis Publishers, Inc., Chelsea, Michigan,pp. 11e132.

Berney, M., Hammes, F., Bosshard, F., Weilenmann, H.-U., Egli, T.,2007. Assessment and interpretation of bacterial viability byusing the LIVE/DEAD BacLight kit in combination with flowcytometry. Applied and Environmental Microbiology 73 (10),3283e3290.

Boulos, L., Prevost, M., Barbeau, B., Coallier, J., Desjardins, R., 1999.LIVE/DEAD� BacLight�: application of a new rapid stainingmethod for direct enumeration of viable and total bacteria indrinking water. Journal of Microbiological Methods 37, 77e86.

Carvalho, L.R., Pipole, F., Werner, V.R., Laughinghouse IV, H.D.,Camargo, A.C.M., Rangel, M., Konno, K., Sant’Anna, C.L., 2008.A toxic cyanobacterial bloom in an urban coastal lake, RioGrande do Sul state, Southern Brazil. Brazilian Journal ofMicrobiology 39, 761e769.

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 4 2993

Cataldo, F., 2005. Ozone degradation of ribonucleic acid (RNA).Polymer Degradation and Stability 89, 275e281.

Chen, J.-J., Yeh, H.-H., Tseng, I.-C., 2009. Effect of ozone andpermanganate on algae coagulation removal e pilot andbench scale tests. Chemosphere 74, 840e846.

Cheng, Y.-L., Juang, Y.-C., Liao, G.-Y., Ho, S.-H., Yeh, K.-L.,Chen, C.-Y., Chang, J.-C., Liu, J.-C., Lee, D.-J., 2010. Dispersedozone flotation of Chlorella vulgaris. Bioresource Technology101, 9092e9096.

Chow, C.W.K., Drikas, M., House, J., Burch, M.D.,Velzeboer, R.M.A., 1999. The impact of conventional watertreatment processes on cells of the cyanobacteriumMicrocystisaeruginosa. Water Research 33 (15), 3253e3262.

Daly, R.I., Ho, L., Brookes, J.D., 2007. Effect of chlorination onMicrocystis aeruginosa cell integrity and subsequentmicrocystin release and degradation. Environmental Scienceand Technology 41, 4447e4453.

El-Aty, A.M.A., Ibrahim, M.B.M., El-Dib, M.A., Radwan, E.K., 2009.Influence of chlorine on algae as precursors fortrihalomethane and haloacetic acid production. WorldApplied Sciences Journal 6 (9), 1215e1220.

Fang, J., Ma, J., Yang, X., Shang, C., 2010a. Formation ofcarbonaceous and nitrogenous disinfection by-products fromthe chlorination of Microcystis aeruginosa. Water Research2010, 1934e1940.

Fang, J., Yang, X., Ma, J., Shang, C., Zhao, Q., 2010b.Characterization of algal organic matter and formation ofDBPs chlor(am)ination. Water Research 44, 5897e5906.

Gorham, P.R., Mclachlan, J., Hammer, U.T., Kim, W.K., 1964.Isolation and culture of toxic strains of Anabaena flos-aquae(Lyngb.) de Bred. International Association of Theoretical andApplied Limnology 15, 796e804.

Graham, J.L., Loftin, K.A., Meyer, M.T., Ziegler, A.C., 2010.Cyanotoxin mixtures and taste-and-odor compounds incyanobacterial blooms from the Midwestern United States.Environmental Science and Technology 44 (19), 7361e7368.

Greene, A.K., Few, B.K., Serafini, J.C., 1993. A comparison ofozonation and chlorination for the disinfection of stainlesssteel surfaces. Journal of Dairy Science 76, 3617e3620.

Hammes, F.,Meylan, S., Salhi, E., Koster,O., Egli, T., VonGunten,U.,2007. Formation of assimilable organic carbon (AOC) andspecific natural organic matter (NOM) fraction duringozonation of phytoplankton. Water Research 41, 1447e1454.

Henderson, R., Parsons, S.A., Jefferson, B., 2008a. The impact ofalgal properties and pre-oxidation on solid-liquid separationof algae. Water Research 42, 1827e1845.

Henderson, R.K., Baker, A., Parsons, S.A., Jefferson, B., 2008b.Characterisation of algogenic organic matter extracted fromcyanobacteria, green algae and diatoms. Water Research 42,3435e3445.

Hoeger, S.J., Dietrich, D.R., Hitzfeld, B.C., 2002. Effect of ozonationon the removal of cyanobacterial toxins during drinking watertreatment. Environmental Health Perspectives 110 (11),1127e1132.

Hong, H.C., Mazumder, A., Wong, M.H., Liang, Y., 2008. Yield oftrihalomethanes and haloacetic acids upon chlorinating algalcells, and its prediction via algal cellular. Water Research 42,4941e4948.

Hong, H.C., Wong, M.H., Liang, Y., 2009. Amino acids asprecursors of trihalomethane and haloacetic acid formationduring chlorination. Archives of EnvironmentalContamination and Toxicology 56, 638e645.

Hong, H.C., Huang, F.Q., Wang, F.Y., Ding, L.X., Lin, H.J., Liang, Y.,2013. Properties of sediment NOM collected from a drinkingwater reservoir in South China, and its association with THMsand HAAs formation. Journal of Hydrology 476, 274e279.

Huang, W.-J., Cheng, Y.-L., Cheng, B.-L., 2008. Ozonationbyproducts and determination of extracellular release in

freshwater algae and cyanobacteria. EnvironmentalEngineering Science 25 (2), 139e151.

Huang, J., Graham, N., Templeton, M.R., Zhang, Y., Collins, C.,Nieuwenhuijsen, M., 2009. A comparison of the role of twoblue-green algae in THM and HAA formation. Water Research43, 3009e3018.

Li, H., Yao, C., Dong, X., Dong, W., Fan, Z., 2011. Effect of pH oninactivation of Microcystis aeruginosa by ozonation air insequencing batch reactor. Journal of Chemical TechnologyBiotechnology 86, 468e471.

Li, L., Gao, N., Deng, Y., Yao, J., Zhang, K., 2012. Characterizationof intracellular & extracellular algae organic matters (AOM) ofMicrocystis aeruginosa and formation of AOM-associateddisinfection byproducts and odor & taste compounds. WaterResearch 46, 1233e1240.

Liang, L., Singer, P.C., 2003. Factors influencing the formation andrelative distribution of haloacetic acids and trihalomethanesin drinking water. Environmental Science and Technology 37,2920e2928.

Lin, T.-F., Chang, D.-W., Lien, S.-K., Tseng, Y.-S., Chiu, Y.-T.,Wang, Y.-S., 2009. Effect of chlorination on the cell integrity oftwo noxious cyanobacteria and their releases of odorants.Journal of Water Supply: Research and Technology e AQUA 58(8), 539e551.

Liu, Z., Cui, F., Ma, H., Fan, Z., Zhao, Z., 2011. The role ofnitrobenzene on the yield of trihalomethane formationpotential in aqueous solutions with Microcystis aeruginosa.Water Research 45, 6489e6495.

Ma, M., Liu, R., Liu, H., Qu, J., 2012a. Effect of moderate pre-oxidation on the removal of Microcystis aeruginosa by KMnO4-Fe(II) process: significance of the in-situ formed Fe (III). WaterResearch 46, 73e81.

Ma, M., Liu, R., Liu, H., Qu, J., Jefferson, W., 2012b. Effects andmechanisms of pre-chlorination on Microcystis aeruginosaremoval by alum coagulation: significance of the releaseintracellular organic matter. Separation and PurificationsTechnology 86, 19e25.

Merel, S., Clement, M., Thomas, O., 2010. State of the art oncyanotoxins in water and their behaviour towards chlorine.Toxicon 55, 677e691.

Miao, H., Tao, W., 2009. The mechanisms of ozonation oncyanobacteria and its toxins removal. Separation andPurifications Technology 66, 187e193.

Mizuno, T., Tsuno, H., Yamada, H., 2007. Development of ozoneself-decomposition model for engineering design. Ozone:Science and Engineering 29, 55e63.

Morato, J., Mir, J., Codony, F., Mas, J., Ribas, F., 2003. In: Mara, D.,Horan, N. (Eds.), Water andWastewater Microbiology. Elsevier,Great Britain, pp. 657e693.

Moreno, I.M., Pereira, P., Franca, S., Camean, A., 2004. Toxiccyanobacteria strains isolated from blooms in the GuadianaRiver (southwestern Spain). Biological Research 37, 405e417.

Muller, K.C., Forster, R., Gammeter, S., Hambsch, B., 2003.Influence of ozonated cyanobacteria on bacterial growth inrapid sand filters. Journal of Water Supply: Research andTechnology e AQUA 52 (5), 333e340.

Navalon, S., Alvaro, M., Garcia, H., 2008. Carbohydrates astrihalomethanes precursors. Influence of pH and th resence ofCl� and Br� on trihalomethane formation potential. WaterResearch 42, 3990e4000.

Nguyen, M.-L., Westerhoff, P., Baker, L., Hu, Q., Esparza-Soto, M.,Sommerfeld, M., 2005. Characteristics and reactivity of algae-produced dissolved organic carbon-ASCE. Journal ofEnvironmental Engineering 131, 1574e1582.

Oberholster, P.J., Botha, A.-M., Cloete, T.E., 2006. Toxiccyanobacterial blooms in a shallow, artificially mixed urbanlake in Colorado, USA. Lakes & Reservoirs and Management11, 111e123.

wat e r r e s e a r c h 4 7 ( 2 0 1 3 ) 2 9 8 3e2 9 9 42994

Onstad, G.D., Strauch, S., Meriluoto, J., Codd, G.A., Von Gunten, U.,2007. Selective oxidation of key functional groups incyanotoxins during drinking water ozonation. EnvironmentalScience and Technology 41, 4397e4404.

Park, H.-S., Hwang, T.-M., Kang, J.-W., Choi, H., Oh, H.-J., 2001.Characterization of raw water for the ozone applicationmeasuring ozone consumption rate. Water Research 35 (11),2607e2614.

Pivokonsky, M., Kloucek, O., Pivokonska, L., 2006. Evaluation ofthe production, composition and aluminum and ironcomplexation of algogenic organic matter. Water Research 40,3045e3052.

Plummer, J.D., Edzwald, J.K., 2001. Effect of ozone on algae asprecursors for trihalomethane and haloacetic acid production.Environmental Science and Technology 35, 3661e3668.

Plummer, J.D., Edzwald, J.K., 2002. Effects of chlorine and ozoneon algal cell properties and removal of algae by coagulation.Journal of Water Supply: Research and Technology e AQUA51, 307e318.

Qian, H., Yu, S., Sun, Z., Xie, X., Liu, W., Fu, Z., 2010. Effects ofcopper sulfate, hydrogen peroxide and N-phenyl-2-naphthylamine on oxidative stress and the expression ofgenes involved photosynthesis and microcystin disposition inMicrocystis aeruginosa. Aquatic Toxicology 99, 405e412.

Ramseier, M.K., Peter, A., Traber, J., Von Gunten, U., 2011.Formation of assimilable organic carbon during oxidation ofnatural waters with ozone, chlorine dioxide, chlorine,permanganate, and ferrate. Water Research 45 (5), 2002e2010.

Rippka, R., Deruelles, J., Waterbury, J.B., Herdman, M.,Stanier, R.Y., 1979. Generic assignments, strain histories andproperties of pure cultures of cyanobacteria. Journal ofGeneral Microbiology 111, 1e61.

Somek, H., Ustao�glu, M.R., M., Y., 2008. A case report: algal bloomof Microcystis aeruginosa in a drinking-water body, E�girdir lake,Turkey. Turkish Journal of Fisheries and Aquatic Sciences 8,177e179.

Stocks, S.M., 2004. Mechanism and use of the commerciallyavailable viability stain, BacLight. Cytometry Part A 61, 189e195.

Teixeira, M.R., Rosa, M.J., 2006. Comparing dissolved air flotationand conventional sedimentation to remove cyanobacterialcells of Microcystis aeruginosa part I: the key operating

conditions. Separation and Purifications Technology 52,84e94.

Theruvathu, J.A., Flyunt, R., Aravindakumar, C.T., vonSonntag, C., 2001. Rate constants of ozone reactions with DNA,its constituents and related compounds. Journal of theChemical Society, Perkin Transactions 2, 269e274.

USEPA, 1995. Method 552.2-determination of Haloacetic Acidsand Dalapon in Drinking Water by Liquid-liquid Extraction,Derivatization and Gas Chromatography with ElectronCapture Detection. U.S. Environmental Protection Agency,Cincinnati, Ohio, pp. 1e32.

Vincent, S., Kotbi, A., Barbeau, B., 2010. Predicting hydroxylradical activity and trace contaminants removal in ozonatedwater. Ozone: Science and Engineering 32, 244e251.

Von Gunten, U., 2003. Ozonation of drinking water: part I.Oxidation kinetics and product formation. Water Research 37,1443e1467.

Widrig, D.L., Gray, K.A., McAuliffe, K.S., 1996. Removal of algal-derived organic material by preozonation and coagulation:monitoring changes in organic quality by pyrolysis-GC-MS.Water Research 30 (11), 2621e2632.

Wu, Z., Shen, H., Ondruschka, B., Zhang, Y., Wang, W.,Bremner, D.H., 2012. Removal of blue-green algae using thehybrid method of hydrodynamic cavitation and ozonation.Journal of Hazardous Materials 235-236, 152e158.

Zamyadi, A., Ho, L., Newcombe, G., Daly, R.I., Burch, M., Baker, P.,Prevost, M., 2010. Release and oxidation of cell-boundsaxitoxins during chlorination of Anabaena circinalis cells.Environmental Science and Technology 44, 9055e9061.

Zamyadi, A., Ho, L., Newcombe, G., Bustamante, H., Prevost, M.,2012a. Fate of toxic cyanobacterial cells and disinfection by-products formation after chlorination. Water Research 46,1524e1535.

Zamyadi, A., MacLeod, S.L., Fan, Y., McQuaid, N., Dorner, S.,Sauve, S., Prevost, M., 2012b. Toxic cyanobacterialbreakthrough and accumulation in a drinking water plant: amonitoring and treatment challenge. Water Research 46,1511e1523.

�Swietlik, J., Dabrowska, A., Raczyk-Stanisławiak, U., Nawrocki, J.,2004. Reactivity of natural organic matter fractions withchlorine dioxide and ozone. Water Research 38, 547e558.