Synergistic disinfection and removal of biofilms by a sequential two-step treatment with ozone followed by hydrogen peroxide

w a t e r r e s e a r c h 6 4 ( 2 0 1 4 ) 9 4 e1 0 1

Available online at www.sciencedirect.com

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

Synergistic disinfection and removal of biofilms by a sequential two-step treatment with ozone followed by hydrogen peroxide Mariko Tachikawa*, Kenzo Yamanaka School of Pharmacy, Nihon University, 7-7-1 Narashinodai, Funabashi, Chiba 274-8555, Japan

article info

abstract

Article history:

Synergistic disinfection and removal of biofilms by ozone (O3) water in combination with

Received 18 February 2014

hydrogen peroxide (H2O2) solution was studied by determining disinfection rates and

Received in revised form

observing changes of the biofilm structure in situ by confocal laser scanning microscopy

14 May 2014

(CLSM) using an established biofilm of Pseudomonas fluorescence. The sequential treatment

Accepted 29 June 2014

with O3, 1.0e1.7 mg/L, followed by H2O2, 0.8e1.1%, showed synergistic disinfection effects,

Available online 9 July 2014

while the reversed treatment, first H2O2 followed by O3, showed only an additive effect. The decrease of synergistic disinfection effect by addition of methanol (CH3OH), a scavenger of

Keywords:

hydroxyl radical (
OH), into the H2O2 solution suggested generation of hydroxyl radicals on

Synergistic disinfection effects

or in the biofilm by the sequential treatment with O3 followed by H2O2. The primary

Sequential treatment

treatment with O3 increased disinfection rates of H2O2 in the secondary treatment, and the

Ozone

increase of O3 concentration enhanced the rates. The cold temperature of O3 water (14  C

Hydrogen peroxide

and 8  C) increased the synergistic effect, suggesting the increase of O3 adsorption and

Biofilm

hydroxyl radical generation in the biofilm. CLSM observation showed that the sequential

Pseudomonas fluorescens

treatment, first with O3 followed by H2O2, loosened the cell connections and thinned the extracellular polysaccharides (EPS) in the biofilm. The hydroxyl radical generation in the biofilm may affect the EPS and biofilm structure and may induce effective disinfection with H2O2. This sequential treatment method may suggest a new practical procedure for disinfection and removal of biofilms by inorganic oxidants such as O3 and H2O2. © 2014 Elsevier Ltd. All rights reserved.

1.

Introduction

The formation of bacterial biofilms and their removal by disinfection has become an important subject in water quality management of swimming pools, food processing lines, industrial water systems, etc. Since microorganisms in a biofilm are protected by matrices of extracellular polymeric substances and are more tolerant to antibiotics and biocides than

* Corresponding author. Tel.: þ81 47 465 5846; fax: þ81 47 465 6077. E-mail address: [email protected] (M. Tachikawa). http://dx.doi.org/10.1016/j.watres.2014.06.047 0043-1354/© 2014 Elsevier Ltd. All rights reserved.

planktons, the importance of tests using a biofilm system has been pointed out for evaluation of disinfection efficacy as a biofilm disinfectant. Among the oxidants used for disinfection, O3 has been given attention because of its excellent oxidation power. In the previous study (Tachikawa et al., 2009), we evaluated the disinfection efficacy of O3 water on biofilms established from ubiquitous bacteria, Pseudomonas fluorescens and Pseudomonas aeruginosa. The results indicated O3 water to be effective for biofilm disinfection; however, it required much

w a t e r r e s e a r c h 6 4 ( 2 0 1 4 ) 9 4 e1 0 1

higher concentrations of O3 than those for planktonic cells (EPA, 1999; Viera et al., 1999) and showed tail-off disinfection curves suggesting the formation of physical hindrance to O3 penetration in biofilms. In order to use O3 water for disinfection of biofilms, means to raise its efficiency are anticipated. In the field of water treatment, the advanced oxidation process (AOP) has been utilized to increase O3 decomposition by adding H2O2 to generate a highly reactive hydroxyl radical (
OH) (EPA, 1999). Though radicals show synergistic oxidation effects on organic solutes in water, it is known that the microbicidal activity of the radicals is greatly affected by the O3 dose, H2O2/O3 ratio, contact time, source water quality and type of microorganisms tested (Wolfe et al., 1989; Labatiuk et al., 1994). The very short life time of free radicals made it difficult to apply the AOP system directly to biofilm disinfection and removal. However, it may be considered that, if free radical generation occurs on or in biofilms, AOP may give effects on the structure of biofilms and on its disinfection efficacy. Therefore, the sequential use of O3 and H2O2 for disinfection and removal of biofilms was studied by determination of disinfection rates and by observation of structural changes of biofilms by confocal laser scanning microscopy (CLSM) using an established biofilm of P. fluorescens (Tachikawa et al., 2005).

95

H2O2 concentrations. For preparation of O3 water (0.45e1.72 mg/L), tap water distributed by the Funabashi municipal water supply (pH 7.4) was dechlorinated by passing it through an activated carbon column and then led to an O3 water generator (AOD-TH, Ai Electronic Ind. Co. Ltd., Japan). An O3 water generator was available for continuous supply of O3 water at a constant concentration. Since O3 water was too unstable to maintain a constant concentration during treatment, we used a flow-through system for O3 treatment as we used in the previous study (Tachikawa et al., 2009). Hydrogen peroxide solutions (0.92 mg/L and 0.13e1.1%) and a CH3OH solution (0.3 M, 0.96%) were prepared by dilution with sterilized Milli Q water in order to avoid the effects of chlorine and other solute in water. The pH of 1% H2O2 solution was 5.4. Concentrations of residual O3 in the test water were determined by the indigo colorimetric method (Standard Methods, 1998b). Since the concentrations of O3 in the water were higher than 0.3 mg/L, an accurate volume (10e20 mL) of fresh O3 water was added into a volumetric flask containing indigo reagent II and the resulting mixture was diluted to 100 mL with Milli Q water. The optical absorbance of the solution was measured at 600 nm in 1-cm cells, and the concentration of O3 was calculated according to the equation described below, mg O3 =L ¼ ð100  DAÞ=ð0:42  b  VÞ

2.

Materials and methods

2.1.

Bacterial strains and biofilm preparation

P. fluorescens (JCM no. 2779) was obtained from the Japan Collection of Microorganisms (JCM) in Riken Bioresource Center, Tsukuba, Japan. Biofilms of P. fluorescens were grown on clean and sterile small glass slides (14  26 mm) according to our established method as described elsewhere (Tachikawa et al., 2005). The glass slides (20 plates) were placed in a glass culture dish (i.d. 145 mm) holding 150 mL of growth medium containing 1% of glucose and phosphate and a small amount of minerals (LeChevallier et al., 1988). The dish was inoculated with a culture of P. fluorescens, and incubated at 28  C with continuous slow stirring with magnetic stirrers for 2 nights. The number of viable cells in the biofilm formed on the glass slide was determined by colony counting on tryptone glucose yeast agar (Standard Methods, 1998a) following ultrasonic dispersion and serial dilution. In the present study, the mean cell density of the biofilm was 1.1 ± 0.5  109 cfu/slide.

2.2.

Reagents

Hydrogen peroxide (30%) and methanol (CH3OH, special grade) were obtained from Wako Pure Chemical Industry, Osaka, Japan. Fluorescence dyes, Alexa Fluor® 633 conjugate concanavalin A (AlexaFluor® 633-ConA) and LIVE/DEAD BacLightTM, were obtained from Invitrogen™ Corp. CA. Other reagents used were all of the reagent grade.

2.3.

Water

Milli Q® ultrapure water was used for preparation and dilution of the reagent solutions for the determination of ozone and

where DA ¼ difference in absorbance between the sample and blank, b ¼ path length of cell in cm (1 cm), and V ¼ volume of the sample in mL. H2O2 in the test water was determined by a modified iodometric method for chlorine (Standard Methods, 1998c). One to 10 mL of test water was added into the flask with a ground stopper containing 10 mL of Milli Q water, 5 mL of 10% H2SO4 and 5 mL of 10% KI solution. The mixed solution was kept standing for 20 min away from direct sunlight. After a dilution with 20 mL Milli Q water, the solution was titrated with 0.1N Na2S2O3 solution using starch solution as an indicator until the blue color disappeared. One mL of 0.1N Na2S2O3 solution (factor 1.000) is equivalent to 1.701 mg H2O2.

2.4. Disinfection of biofilms by the sequential treatment with O3 and H2O2 Effects of the sequential treatment with O3 and H2O2 on biofilms were evaluated as follows. The biofilms established on a glass slide were immersed in sterilized water twice to remove planktonic cells. Then the biofilms were treated with O3 water (1.3 mg/L) in a flow system and with H2O2 water (1.1%) in a static system for indicated times (10 or 120 s) in the following combinations: (A), treated only with H2O2; (B), treated only with O3; (C), first with H2O2 followed by O3; (D), first with O3 followed by H2O2. For elucidation of the mechanism of the synergistic effect, the following experiments were done: I, the biofilms were treated sequentially with O3 (1.7 mg/L) and then with a H2O2 solution (0.75%) containing CH3OH (0.96%) as a scavenger of hydroxyl radicals for indicated times; II, the biofilms were sequentially treated first with various concentrations of O3 (0.45e1.5 mg/L) and then with H2O2 (0.13e1.05%) for indicated times. After the treatments, the biofilms were

96

w a t e r r e s e a r c h 6 4 ( 2 0 1 4 ) 9 4 e1 0 1

2.5. Effect of low temperatures of O3 water on the synergistic disinfection In preliminary experiments, freshly prepared O3 water (100 mL, at 25  C) in a glass beaker was chilled in cold water at 6  C and changes of water temperature and O3 concentration with time were monitored. The water temperature and concentration of O3 dropped from 25  C to 8  C and from 1.52 mg/L to 1.28 mg/L, respectively, in 7.5 min of cooling. The biofilms were treated first with O3 water at 25, 14 and 8  C for 30 s, and then with H2O2 (1.03%) at 20  C for 120 s. The concentration of O3 at each temperature was determined by the indigo method simultaneously. The numbers of viable cells in the biofilms treated were determined as described above.

2.6.

CLSM observations

Biofilms formed on the glass slide were treated with O3 and H2O2 for 30 s each in turn; (a), first with H2O2 followed by O3; (b), first with O3 followed H2O2. After the treatment, the biofilms on the glass slide were soaked in sterilized saline water containing Na2S2O3 to neutralize residual oxidants and rinse off planktonic cells. For preparation of dye solutions, a 10 mM phosphate buffer solution at pH 7.4 was used. Fluorescencestaining of the biofilms was performed in the following two steps. First, the extracellular poly saccharide (EPS) in the biofilm was stained by gently dropping 0.1 mL of AlexaFluor® 633ConA solution (1 mg/mL) on the biofilm surface and incubated for 30 min in the dark at room temperature, and then the dye was rinsed off by soaking in sterilized saline water. Second, the cells in the biofilm were stained by soaking and incubating in 3 mL of phosphate buffer solution containing 9 mL of LIVE/DEAD BacLightTM, an equal volume mixture of SYTO 9 and propidium iodide solutions supplied in the kit, for 15 min in the dark at room temperature. The intact cells were stained green and the damaged cells were stained red. One-tenth mL of water was dropped on a cover glass (25  50 mm), and then the slide of the stained biofilms was placed upside down. Photomicrographs were taken at a magnification of x 100 with an oil immersion lens under a confocal laser scanning microscope (CLSM, LSM 510, Carl Zeiss). Computer image micrographs of the vertical section of the biofilms were obtained by using the Z-stack function of the LSM 510 (Tachikawa et al., 2005).

2.7.

Statistical analysis

A regression line was obtained from the simple linear leastsquare regression analysis and statistical differences

between groups were calculated by Welch's t-test (p* < 0.05), by using the Excel-add-in computer program, StatceL2 (Yanai, 2009), respectively.

3.

Results

3.1. Synergistic disinfection effects by sequential treatments In our previous studies, the survival ratio of P. fluorescens biofilms after the single treatment with 1 mg/L O3 for 2 min was 0.05 (Tachikawa et al., 2009), and that with 1% H2O2 solution at 24  C for 5 min was 0.09 (data not published), respectively. Taking these moderate but consistent disinfection effects under these experimental conditions in consideration, a combination of treatment times, 2 min and 10 s, was adopted for the detection of a synergistic effect in the sequential treatments. The survival ratios of P. fluorescens in biofilms obtained after a single treatment with H2O2 for 10 s (A) or O3 for 2 min (B), and after sequential two-step treatments with H2O2 and O3, (C and D), are illustrated in Fig. 1. When the latter sequential treatments are compared, the treatment C, first with H2O2 followed by O3, showed a simple additive disinfection effect of A plus B, whereas the treatment D, first with O3 followed by H2O2, showed a synergistic effect, i.e., a ca. 10 times lower survival ratio than that of C. When the exposure times were interchanged, i.e. O3 for 10 s and H2O2 for 2 min, the survival ratios after combination treatment showed a similar tendency, although their survival ratios were generally higher than those in Fig. 1. These results indicate that the order of the disinfectants, i.e. first with O3 followed by H2O2, is essential for a synergistic disinfection effect to be observed.

10

Biofilms of P. fluorescens Survival ratio

transferred into a sterilized flask containing a 0.04 mM Na2S2O3 solution for neutralizing the residual O3 and H2O2. The number of viable cells as colony forming units (CFU) in each biofilm was determined by colony-counting after ultrasonic dispersion and serial dilution. Survival ratios were obtained by division of the number of CFU in biofilms treated (Nt) by those in intact (control) biofilms (Nc). From a plot of survival ratios, on a logarithmic scale, against exposure times, a line fitted to the data points was obtained by using simple linear least-square regression analysis. The slope of the line may provide a disinfection rate.

1

0.1

0.01

0.001

0.0001

Control

(A)

(B)

H2O2 10 s

O3 2 min

(C)

(D)

H2O2 O3 O3 H2O2 10 s 2min 2 min 10 s

Fig. 1 e Disinfection efficacy of H2O2 and O3 in their single, (A) and (B), and sequential two-step, (C) and (D), treatments of the biofilms of P. fluorescens. Biofilms were treated with H2O2 or O3, or both for indicated times. The concentrations of H2O2 and O3 were 1.1% and 1.3 mg/L, respectively. Survival ratios are expressed as means ± SD (n ¼ 3). The dotted line indicates an estimated additive survival ratio of H2O2 and O3, obtained by multiplying the survival ratio of H2O2 (A) by that of O3 (B); 0.27 £ 0.03 ¼ 0.0081.

97

w a t e r r e s e a r c h 6 4 ( 2 0 1 4 ) 9 4 e1 0 1

3.2. Decrease of the synergistic effect by addition of CH3OH into the H2O2 solution The possibility of hydroxyl radical generation in the sequential treatment was studied by addition of CH3OH, a scavenger of hydroxyl radicals, into the H2O2 solution, and the results are shown in Fig. 2. In a preliminary test, the concentration of H2O2 solution (1.05%, 0.31 M) was not reduced by the addition of CH3OH (0.96%, 0.3 M) (data not shown). The single treatment of biofilms with 0.96% CH3OH solution for 60 s had little microbicidal effect. The sequential treatment, first with O3 water (1.7 mg/L) for 30 s followed by 0.96% H2O2 solution containing 0.3 M CH3OH for 60 s, significantly increased the survival ratio from 0.03 to 0.08, although it did not restore the estimated additive survival ratio of the two independent single treatments (0.144; 0.36  0.40). Hydroxyl radical might work as a strong oxidant exactly where it is formed because of its short life. In the sequential treatment, hydroxyl radical may be formed first on the surface of biofilms on the moment when H2O2 of the second treatment encounters with O3 of the first treatment. Thereafter, as H2O2 penetrates into the biofilm, hydroxyl radical might be formed in the biofilm where H2O2 and O3 meet. Methanol also penetrates into the biofilm and will work as a scavenger. These results suggest that O3 remaining on or in the biofilms may induce synergistic effect with H2O2 by hydroxyl radical generation.

Fig. 3 e Effect of O3 pretreatment on the disinfection rate of H2O2 in the biofilms. (a), Biofilms were treated with 1.13% H2O2 for 120 s (B, ); (b), biofilms were treated first with 1.1 (C) or 1.4 (-) mg/L O3 for 30 s followed by 1.13% H2O2 for 30e90 s. The survival ratios are expressed as means ± SD (n ¼ 3). A line fitted to the data points was obtained by using simple linear least-square regression analysis. Line (a), y ¼ e¡0.03x (R2 ¼ 0.9177); line (b), y ¼ 0.355e¡0.053x (R2 ¼ 0.9000).

▫

3.3. Influence of O3 and H2O2 concentrations on the synergistic effect The effect of the primary treatment with O3 on the disinfection rate of the secondary treatment with H2O2 was studied. As illustrated in Fig. 3, the bacterial survival ratios of the

Survival ratio

1

0.1

n=2

n=2

n=2

n=2

n=5

n=4

0.01

Fig. 2 e Decrease of synergistic effect in the sequential treatments by addition of CH3OH, a
OH scavenger, into H2O2 solution. The concentrations and lengths of time of treatment with O3, H2O2 and CH3OH were 1.7 mg/L for 30 s, 0.75% for 60 s and 0.96% for 60 s, respectively. Survival ratios are expressed as means (n ¼ 2) or means ± SD (n ¼ 4 or 5). The dotted line indicates an estimated additive survival ratio obtained by multiplying the survival ratio of O3 by that of H2O2; 0.36 x 0.4 ¼ 0.144. Asterisk indicates statistical significance of difference between groups by Welch's t-test (P < 0.05).

biofilms treated only with H2O2 (closed symbols) and those of the biofilms treated with O3 followed by H2O2 (open symbols) were plotted and a line fitted to each plotting group was obtained. The slopes of the fitted lines corresponded to the disinfection rates of 0.03 and 0.053 s1, respectively. After the O3 pretreatment, the disinfection rate of H2O2 was enhanced nearly twice. Then, effects of the concentrations of O3 and H2O2 in the sequential treatment were examined in several combinations of O3 and H2O2 concentrations. As shown in Fig. 4, at the higher concentrations of H2O2 (0.85 and 1.05%), the disinfection rates increased as the concentration of O3 in the pretreatment increased (0.45e1.5 mg/L). Although the H2O2 concentration of 4(e) (0.85%) was a little lower than that of 4(d) (1.05%), the disinfection efficacy of 4(e) increased and was larger than that of 4(d). In the present study, the concentration ratio of H2O2 to O3 is more than 5000; thus it was considered that there was no difference in O3 scavenging ability at the concentration of H2O2 between 0.85 and 1%. Therefore, the increased disinfection efficacy of Fig. 4(e) might be due to the higher concentration of O3 (1.5 mg/L versus 1.1 mg/L) in Fig. 4(d). At the lower concentration of H2O2 (0.13%), the disinfection rate (the slope of the fitted line) was not affected by the O3 concentration. We did a sequential twostep treatment at similar concentrations of O3 and H2O2, i.e., 2.4 mg/L and 0.9 mg/L for 30 s each, respectively. Though the treatment time (30 s) for H2O2 was shorter than those in the other experiments shown in Fig. 4, the survival ratios of the biofilms by the single treatment with the O3 water and by the sequential two-step treatment with the O3 water followed by the H2O2 solution were 0.19 ± 0.07 and 0.13 ± 0.03 (each, n ¼ 3), respectively. No synergistic disinfection effect was observed

98

w a t e r r e s e a r c h 6 4 ( 2 0 1 4 ) 9 4 e1 0 1

the sequential treatment, that is, the cold temperature of O3 water might increase the stability of O3 in the biofilms and increase the amount of hydroxyl radicals generated.

3.5. In situ observation of the synergistic effect on biofilms by CLSM

Fig. 4 e Effects of O3 and H2O2 concentrations on their disinfection efficacy in their sequential treatments of the biofilms. The biofilms were treated first with O3, 0.45e1.50 mg/L for 30 s followed by H2O2, 0.13 or 1.05% for 120e150 s. The combinations of concentrations of O3 and H2O2 are indicated respectively in the figure window. Survival ratios are expressed as means ± SD (n ¼ 3).

at similar concentrations of O3 and H2O2. These results indicate that the concentration level of H2O2 around 1% may be required for the synergistic effect.

3.4. Effect of cold temperature of O3 water on the synergistic disinfection Since it was reported that changes in temperature of O3 water affected both O3 decomposition (Takahashi and Katsuki, 1982) and inactivation of Cryptosporidium parvum (Li et al., 2001a; b, Rennecker et al., 1999), effects of low temperatures of O3 water on the sequential treatment were studied. The biofilms were treated first with O3 water at temperatures of 25, 14, and 8  C and then treated with H2O2 at 20  C. The concentrations of O3 in the test water ranged 1.52, 1.36 and 1.28 mg/L at 25, 14 and 8  C, respectively. As summarized in Table 1, the synergistic disinfection effect increased as the O3 water became colder, in spite of the fact that O3 concentration in the test water became lower along with the descending temperature. These results suggested that there was another disinfection mechanism in

Table 1 e Influence of the temperature of O3 water on disinfection efficacy in the sequential treatment of biofilms of P. fluorescens with O3 and H2O2. O3 water

Exp. Ia Exp. IIb Exp. IIIc a,b,c

Temperature ( C)

Concentration (mg/L)

Survival ratiod (Nt/Nc)

25 14 8

1.52 1.36 1.28

0.0026 ± 0.0017 0.0004 ± 0.0002 0.0001 ± 0.0001

The biofilms were treated first with O3 at the indicated concentration and temperature for 30 s, and then with 1.03% H2O2 at 20  C for 120 s. d Survival ratios are expressed as mean ± SD (n ¼ 3). The biofilms treated with neither O3 nor H2O2 were taken as control.

The structures of biofilms after the sequential treatments with O3 and H2O2 were observed by CLSM. Micrographs of the vertical sections of biofilms are shown in Fig. 5. The matrices of the biofilms, EPS and cells, were clearly defined by computer image analysis after fluorescent staining. In Fig. 5(a), the biofilm was treated first with H2O2 followed by O3; in Fig. 5(b), first with O3 followed by H2O2. EPS residues in the biofilms are shown in grayish white (the top row). Bacteria having intact cell membranes were stained green (the third row from top), whereas those having damaged membranes were stained red (the second row from top). When these computer images of the biofilms (a) and (b) are compared, the EPS residue adhered to the slide glass of biofilm (a) was shown to be white, though that in biofilm (b) was shown to be hazy and light-colored gray. The light color density might suggest the promoted EPS exfoliation in biofilm (b). The damaged cells exfoliated from the matrices of the biofilm (b) and its number was larger than that of the biofilm (a). As a result, cell binding in the biofilm (b) seemed to have become looser than in the biofilm (a). The increase of the exfoliation of EPS and cell damage observed in the biofilm (b) might have been brought about as the result of the synergistic disinfection effect.

4.

Discussion

The main concept of these sequential treatments has been derived from the AOP system of water treatment where O3 and H2O2 are applied together for generating hydroxyl radicals,
OH and
O2H, known as highly reactive and strong oxidants (EPA, 1999; Steahelin and Hoigne, 1982, 1985). H2O2 is a weak acid, and it partially dissociates into the hydroperoxide ion (HO 2 ). The ion accelerates O3 decomposition and hydroxyl radical formation in chain. However, direct simultaneous introduction of O3 and H2O2 in water could not yield reliable microbicidal activity on planktonic cells (Wolfe et al., 1989; Labatiuk et al., 1994). In our preliminary experiments, simultaneous addition of O3 and H2O2 into the water where an established biofilm was placed could not increase disinfection efficacy for the biofilm (data were not shown). The rate of direct oxidation with hydroxyl radicals (1012e1014 M1 sec1) is faster than O3 (105e107 M1 s1), but the concentration of hydroxyl radical is smaller than O3 under normal ozonation (EPA, 1999). The life times of hydroxyl radical, O3 and H2O2 are different: hydroxyl radical, very short (107 s); O3, relatively short (several min); and H2O2, stable, and O3 and hydroxyl radicals (
OH and HO 2) react with each other at high rate constants (106e109 M1 s1)
et al., 1985). Thus, for the effective hydroxyl radical (Hoigne formation in a disinfection process, an appropriate order of oxidant addition as well as proper oxidants' concentrations would be required. In the practical AOP system for water disinfection, the processes involve two steps: O3 dissolution and H2O2 addition. Therefore, we tried sequential treatments

w a t e r r e s e a r c h 6 4 ( 2 0 1 4 ) 9 4 e1 0 1

99

Fig. 5 e Observation of the biofilms of P. fluorescens treated with O3 and H2O2 sequentially by CLSM after staining with AlexaFlu®633-ConA, and Live/Dead BacLightTM. (a), Biofilms were first treated with H2O2 for 30 s followed by O3 for 30 s; (b), first with O3 for 30 s followed by H2O2 for 30 s. The concentrations of O3 and H2O2 were 1.2 mg/L and 1.1%, respectively. The bars represent 10 mm.

with O3 and H2O2 for biofilm disinfection, where the hydroxyl radicals might be generated on and/or inside the biofilm. In the sequential treatments of biofilms with O3 and H2O2, a synergistic disinfection effect was obtained only in the treatment first with O3 followed by H2O2 (Fig. 1). The synergistic survival ratios were ca 0.001. In the previous study (Tachikawa et al., 2009), the survival ratios of P. fluorescens biofilms after the exposure to O3 water (0.92 mg/L) for 5 min was ca. 0.01, and did not reach 0.001 levels within 20 min. The decrease of disinfection efficacy brought about by the addition of CH3OH, a radical scavenger, into the H2O2 solution might indicate involvement of hydroxyl radical generation in the synergistic effect (Fig. 2). In the AOP system for water treatment, the optimal dose ratio of H2O2 to O3 for hydroxyl radical generation was reported to be 0.4 to 0.6 by weight basis (Karimi et al., 1997). It has been reported that, since H2O2 acts as a scavenger as well as a promoter for hydroxyl radical generation, the excessive concentration ratio of H2O2 to O3 in the simultaneous feeding system could not retain an enough concentration of hydroxyl radicals for water treatment (Buxton et al., 1988). However, in the present sequential treatment, the effective concentrations of H2O2 and O3 were 1% and 1e2 mg/L, respectively: the concentration ratio of H2O2 to O3 was 5000e10000. Therefore, the synergistic effect in the present treatment could not be explained only by the formation of hydroxyl radicals in the AOP. Thus, the present sequential treatment method made it possible to yield radicals for the

synergistic disinfection effect in the high concentration ratio of H2O2 to O3. In the previous study (Tachikawa et al., 2009), the CLSM observation showed the upside cell exfoliation of biofilm after 1 min treatment with 1.2 mg/L O3. The exfoliation by O3 may assist in penetrating of H2O2. Conceivably, prior adsorption and/or diffusion of O3 into the biofilm layer might be necessary for the synergistic effect, and the O3 adsorbed may react with H2O2 to generate hydroxyl radicals in the biofilm. There would be a condition where the concentration of O3 was higher than that of H2O2 at the very beginning of the secondary treatment with H2O2. In the inverse order treatment, first with H2O2 followed by O3, the predominant concentration of H2O2 over that of O3 in the biofilm could not have generated enough hydroxyl radicals for yielding the synergistic effect. It has been reported that H2O2 exhibits bacteriostatic effects at 20e40 mg/L, only a moderate bactericidal activity (Baldry, 1983), and that H2O2 requires higher concentrations (0.3e3%) and/or high temperatures (up 80  C) for getting kill rates for bacteria, yeast and viruses (Block, 1991). The results from Figs. 3 and 4 indicate that the O3 pretreatment enhanced the microbicidal effect of H2O2 and a considerable concentration of H2O2 (>1%) was necessary for the synergistic effect. As the concentration of O3 in the pretreatment was increased, the microbicidal rate of H2O2 for biofilms increased. The disinfection rates by H2O2 seemed to be dependent on the amount of hydroxyl radicals generated by O3 and H2O2 in the biofilm. The observation of biofilms by CLSM in Fig. 5 suggests

100

w a t e r r e s e a r c h 6 4 ( 2 0 1 4 ) 9 4 e1 0 1

the formation of fragile structures of biofilms after the treatments with O3 followed by H2O2. In view of these results, the hydroxyl radicals generated in the biofilm by the sequential treatment might react with biofilm constituents, such as EPS, and destroy the biofilm structure. It was reported that hydroxyl radicals degraded alginic acid (Smidsrød et al., 1965), and that biofilm removal by hydroxyl radicals was mainly due to the degradation of the extracellular biopolymer matrix rather than that of intracellular components (Christensen et al., 1990). In the biofilm disinfection with O3 water, we have observed the occurrence of diffusion resistance in the tail-off disinfection curve (Tachikawa et al., 2009; Viera et al., 1999). The radical generation in the biofilm might overcome the diffusion resistance, and then enhance H2O2 diffusion into the biofilm by the destruction of biofilm structures. It had been reported that inactivation efficacy of O3 water against Cryptosporidium parvum decreased as the temperature of O3 water became lower, and that this was mainly due to the decrease of O3 decomposition responsible for radical generation (Li et al., 2001a; b; Rennecker et al., 1999). However, in the present study, the cold temperature of O3 water increased the synergistic effect, in spite of the descending O3 concentration (Table 1). Auto-decomposition of O3 in water is dependent on the pH and temperature; in the pH range higher than 6, the auto-decomposition increased remarkably as the water temperature raised from 13 to 25  C (Takahashi and Katsuki, 1982). The cold temperature could have slowed down the decomposition of O3 adsorbed to the biofilm and made it possible to increase the formation of hydroxyl radicals in the biofilm. There are few studies on the sequential disinfection of biofilms with oxidative disinfectants except our study, while there are some studies on the synergistic sequential inactivation of protozoa, Chryptsporidium parvum, with ozone/free chlorine or ozone/monochloramine (Rennecker et al., 2000; Li et al., 2001b; Corona-Vasques et al., 2002). After the ozone pretreatment, the initial lag phases for free chlorine and monochloramine inactivation disapppeared and the rate of secondary inactivation with both free chlorine and monochloramine was enhanced. In the present study with biofilms, the EPS degradation by hydroxyl radicals generated in the sequential treatment might bring about enhancement of the secondary disinfection rate with H2O2 (Fig. 3). These results may indicate that the sequential treatment method could have a possibility of producing multiple effects of disinfection and removal of biofilms. In the practical treatment system, an O3 water generator and a H2O2 feed system will be involved. The quality of water as solvent is important and it should be dependent on the object to be disinfected. Since O3 and H2O2 decompose ultimately to oxygen and water in the end, the sequential treatments with O3 and H2O2 will be applicable for disinfection and removal of biofilms on medical apparatus, processing lines of medicine and food, and on membrane filter and pipe walls in many water systems.

5.

Conclusion

By using an established biofilm of P. fluorescens, the sequential treatment with O3 followed by H2O2 showed synergistic

disinfection efficacy. In the sequential treatment, hydroxyl radicals generated may degrade the EPS of biofilms and enhanced the disinfection rate by H2O2. The in situ observation of biofilms by CLSM visualized the effective decomposition of EPS and exfoliation of the bacteria cells in the sequential treatment with O3 followed by H2O2. These results indicate that the hydroxyl radicals generated in the biofilm are likely to be essential for disinfection and removal of biofilms, and suggest that the sequential treatment may become a useful method for disinfection of biofilms.

references

Baldry, M.G.C., 1983. The bactericidal, fungicidal and sporicidal properties of hydrogen peroxide and peracetic acid. J. Appl. Bacteriol. 54, 417e423. Block, S., 1991. Disinfection, sterilization and preservation. In: Peroxygen Compounds, fourth ed. Lea and Febiger, Philadelphia, pp. 167e181. Buxton, G.V., Greenstock, C.L., Helman, W.P., Ross, A.B., 1988. Critical review of rate constants for reactions hydrate electrons, hydrogen atoms, and hydroxyl radicals (HO/Oe) in aqueous solution. J. Phys. Chem. Ref. Data 17, 513e886. Christensen, B.E., Trønnes, H.N., Vollan, K., Smidsrød, O., Bakke, R., 1990. Biofilm removal by low concentrations of hydrogen peroxide. Biofouling 2, 165e175. ~ as, B.J., Corona-Vasques, B., Samuelson, A., Rennecker, J.L., Marin 2002. Inactivation of Cryptosporidium parvum oocysts with ozone and free chlorine. Water Res. 36, 4053e4063. EPA Guidance Manual, April 1999. Alternative Disinfectants and Oxidants Guidance Manual. http://www.epa.gov/ogwdw/ mdbp/alternative_disinfectants_guidance. pdf#search¼'EPAGuidanceManual1999'. EPA 815-R-99e014.
, J., Bader, H., Haag, W.R., Staehelin, J., 1985. Rate constants Hoigne of reaction of ozone with organic and inorganic compounds in water e III. Water Res. 19, 993e1004. Karimi, A.A., Redman, J.A., Glaze, W.H., Stolarik, G.F., 1997. Evaluating an AOP for TCE and PCE removal. J. AWWA 89, 41e53. Labatiuk, C.W., Belosevic, M., Finch, G.R., 1994. Inactivation of Giardia muris using ozone and ozone-hydrogen peroxide. Ozone: Sci. Eng. 16, 67e78. LeChevallier, M.W., Cawthon, C.D., Lee, R.G., 1988. Inactivation of biofilm bacteria. Appl. Environ. Microbiol. 54, 2429e2499.
k, L.L., Finch, G.R., Smith, D.W., Belosevic, M., 2001a. Li, H., Gyu¨re Effect of temperature on ozone inactivation of Cryptosporidium parvum in oxidant demand-free phosphate buffer. J. Environ. Eng. 127, 456e467. Li, H., Finch, G.R., Smith, D.W., Belosevic, M., 2001b. Sequential inactivation of Cryptosporidium parvum using ozone and chlorine. Water Res. 35, 4339e4348. ~ as, B.J., Owens, J.H., Rice, E.W., 1999. Rennecker, J.L., Marin Inactivation of Cryptosporidium parvum oocysts with ozone. Water Res. 33, 2481e2488. ~ as, B.J., 2000. Rennecker, J.L., Driedger, A.M., Rubin, S.A., Marin Synergy in sequential inactivation of Cryptosporidium parvum with ozone/free chlorine and ozone/monochloramine. Water Res. 34, 4121e4130. Standard Methods for the Examination of Water and Wastewater, twentieth ed., 1998. American Public Health Association/ American Water Works Association/Water Environment Federation, Washington DC, USA. pp. 9e34-36. Standard Methods for the Examination of Water and Wastewater, twentieth ed., 1998. American Public Health Association/ American Water Works Association/Water Environment Federation, Washington DC, USA. pp. 4e137e139.

w a t e r r e s e a r c h 6 4 ( 2 0 1 4 ) 9 4 e1 0 1

Standard Methods for the Examination of Water and Wastewater, twentieth ed., 1998. American Public Health Association/ American Water Works Association/Water Environment Federation, Washington DC, USA. pp. 4e55e56. Smidsrød, O., Haug, A., Larsen, B., 1965. Kinetic studies on the degradation of alginic acid by hydrogen peroxide in the presence of iron salts. Acta Chem. Scand. 19, 143e152. Steahelin, J., Hoigne, J., 1982. Decomposition of ozone in water: rate of initiation by hydroxide ions and hydrogen peroxide. Environ. Sci. Technol. 16, 676e681. Steahelin, J., Hoigne, J., 1985. Decomposition of ozone in water in the presence of organic solutes acting as promoters and inhibitors of radical chain reactions. Environ. Sci. Technol. 19, 1206e1213. Tachikawa, M., Yamanaka, K., Nakamuro, K., 2009. Studies on the disinfection and removal of biofilms by ozone water using an artificial microbial biofilm system. Ozone: Sci. Eng. 31, 3e9.

101

Tachikawa, M., Tezuka, M., Morita, M., Isogai, K., Okada, S., 2005. Evaluation of some halogen biocides using a microbial system. Water Res. 39, 4126e4132. Takahashi, N., Katsuki, O., 1982. Effect of autodecomposition of ozone on dissolution characteristics in water. I. Yosui to Haisui 24, 534e539. Viera, M.R., Guiamet, P.S., De Mele, M.F.L., Videla, H.A., 1999. Biocidal action of ozone against planktonic and sessile Pseudomonas fluorescens. Biofouling 14, 131e141. Wolfe, R.L., Stewart, M.C., Liang, S., McGuire, M.J., 1989. Disinfection of model indicator organisms in a drinking water pilot by using peroxone. Appl. Environ. Microbiol. 55, 2230e2241. Yanai, H., 2009. 4 Steps Excel Statistics with Statcel. OMS Press, Japan.