In the present study, the characteristics of the combined ultraviolet (UV) and ozone disinfection process were investigated from kinetic and mechanistic viewpoints employing Escherichia coli (E. coli) as an indicator microorganism. Compared to individual unit processes, the combined UV/O3 tests produced excess hydroxyl radicals (HO•) and yielded synergistic inactivation of E. coli in the initial phase of reaction. The presence of O3 during UV exposure caused the destruction of cell structure, and then repressed bacteria regrowth after treatment. Moreover, the formation of malondialdehyde (MDA) showed that the improved generation of intermediate HO• via ozone photolysis accelerated the decomposition of bacterial cell surfaces, which was further confirmed by the leakage of intracellular potassium ions (K+). The results suggested that the synergistic bactericidal effect of combined UV/O3 owed mainly to the enhanced destruction of bacterial cell structure.

INTRODUCTION

During recent decades several technologies have been applied for water disinfection to prevent the spread of diseases caused by waterborne pathogens. Traditional biocidal compounds, such as chlorinated agents and ozone, are efficient for inactivation of a broad range of microorganisms, but can produce disinfection byproducts (DBPs) that have a negative impact on the aquatic environment (Lian et al. 2014; Lyon et al. 2014). Ultraviolet irradiation in the UV-C spectral region has been considered an excellent disinfection technology for not forming dangerous or malodorous halogenated compounds in low UV dose (Sisti et al. 2014). Nevertheless, some species of microorganisms have shown high resistance to UV-C irradiation, and many bacteria have regrowth potential after UV treatment (Guo et al. 2012).

Recently, combining UV radiation with ozone (UV/O3) has been proposed as a promising disinfection technology with high effectiveness while being environmentally friendly (Bustos et al. 2010). Several researchers have reported the synergistic bactericidal effect of the UV/O3 process (Tawabini et al. 2013), and evidence has also shown that the presence of UV radiation during ozone exposure can reduce the formation of toxic byproducts (Zhao et al. 2013). However, the disinfection efficiency of the UV/O3 process in the literature has been significantly influenced by the species of microorganisms (Liu et al. 2014), and some authors have stated that hastened decomposition of ozone under UV radiation could diminish its efficacy as a disinfectant (Venosa et al. 1984).

Compared to the large database on kinetics, the microbial inactivation mechanism of the UV/O3 process is relatively poorly established. Generally, direct impairment in intracellular functions is considered to be the primary reason for bacteria growth inhibition after UV radiation treatment (Beck et al. 2014), while cell death by ozone is primarily related to cell surface damage (Cho et al. 2010). Short wavelength UV light can directly attack DNA to form pyrimidine dimers, which can prohibit the replication of DNA and cause the inactivation of microorganisms (Poepping et al. 2014). In the case of UV/O3 disinfection, not only UV radiation, but also molecule ozone and intermediate reactive oxygen species (ROS) can attack bacterial cells. However, few systematic studies have examined the multiple bactericidal mechanism of the UV/O3 process.

The objective of this study was to evaluate the disinfection efficacy of the combined UV/O3 process, and to elucidate the bacteria inactivation mechanism. The roles of UV, ozone and HO• in the UV/O3 process were estimated, and the destruction of cell structure was examined.

MATERIALS AND METHODS

Preparation of bacteria culture

The Escherichia coli (strain DH5α) bacteria was pre-cultured aerobically in Luria-Bertani (LB) nutrient broth based on procedures reported elsewhere (Wu et al. 2011). Briefly, after incubation at 37°C for 20 h, the bacterial cells were harvested from culture by centrifugation, and were then washed three times using saline water (0.9%, w/w). The sediment was resuspended in ultrapure water to give a cell concentration of approximately 108 colony-forming units per millilitre (CFU/mL) for tests.

Disinfection experiments

Experiments were conducted in a poly(methyl methacrylate) reactor (Figure 1), with an inner diameter of 100 mm, height of 300 mm, and working volume of 1.5 L. A low-pressure UV lamp (UV-C light at 253.7 nm) was fixed in a quartz tube (30 mm in diameter and 330 mm in height), which was placed at the reactor center. For experiments with ozone, 1.5 L/min ozonized gas generated from dry oxygen by a laboratory ozone generator (COM-AD-01, Anseros, Germany) was sparged into the reactor through a porous plate situated at the reactor bottom, while equal compressed air was used in experiments with UV alone. UV fluence rate on the outer surface of the quartz tube was fixed at 16 W/m2, and ozone concentration in the supplied gas was 8.4 mg/L. To estimate the bactericidal effect of HO• in the combined UV/O3 system, 10 mmol/L tert-butyl alcohol (t-BuOH) was added as an HO• scavenger. All materials used were autoclaved at 121°C for 20 min before tests.
Figure 1

Schematic diagram of equipment for testing.

Figure 1

Schematic diagram of equipment for testing.

All experiments were performed at room temperature (20 ± 2°C) and pH 7.1. Tests under each experimental condition were conducted in triplicate with their mean and standard deviations being reported. At fixed time intervals, 10.0 mL bacterial suspension was harvested from the reactor and dosed with 0.2 mL 0.1 mol/L Na2S2O3 to stop the inactivation reaction.

Cell viability

A series of 10-fold dilutions was performed and 0.1 mL of each dilution was plated on LB agar plates. All plates were incubated at 37°C for 24 h, and the numbers of colonies on the plates were counted.

To determine the E. coli photoreactivation after treatment by various processes, samples were transferred into several 50 mL glass flasks, and then were exposed to white fluorescent light with light intensity of 1.1 × 104 lx. At each time interval, a 0.1 mL sample was transferred and used immediately for cell viability assays.

Atomic force microscopy

Bacterial cells after various treatments were characterized directly by atomic force microscopy (AFM). Briefly, a 0.2 mL specimen was dropped on a glass slide, followed by air-drying in a clean bench. Specimens were observed and photographed under an AFM system (BioScope, Veeco, Germany).

Physicochemical analysis

UV-C fluence rate was determined by a UV irradiance meter (Photoelectric Instrument Factory of Beijing Normal University, China). Ozone concentration was measured by the iodometric method (APHA 1998). The temperature and pH were measured with a Hach sensION378 meter (USA). The HO• concentration was assayed using the p-chlorobenzoic acid (pCBA) method (Cho et al. 2004). The quantity of malondialdehyde (MDA) was measured using the thiobarbituric acid (TBA) method (Esterbauer & Cheeseman 1990). The K+ concentration was determined by an inductively coupled plasma optical emission spectrometer (ICP-OES, Optima 5300DV, Perkin-Elmer, Inc., USA).

RESULTS AND DISCUSSION

Inactivation tests

The inactivation of E. coli using the combined UV/O3 process was compared with that in each individual unit process as illustrated in Figure 2(a). Both UV irradiation and ozonation led to fast disinfection, and the detected inactivation efficiencies were about 1.8 log and 1.0 log at 10 s treatments, respectively. Synergistic inactivation of bacterial cells was obtained in the combined UV/O3 process, and about 3.1 log reduction was detected. The inactivation of E. coli by the UV/O3 process with the presence of t-BuOH is also presented in Figure 2(a) as a reference, and 2.6 log reduction was obtained within 10 s. Therefore, in the initial phase of the reaction, the calculated percentages of UV, molecule O3 and HO• in E. coli inactivation were 58.1%, 25.8% and 16.1%, respectively. In the subsequent treatments, combined rather than individual treatments consistently yielded the highest levels of mortality in bacterial cells. At 80 s treatment, the UV, O3, ‘UV/O3 + tert-butyl alcohol’ and UV/O3 tests inactivated 6.3 log, 6.0 log, 7.2 log and 7.6 log of cells, respectively. These observations agree with the results in a previous report (Fang et al. 2014).
Figure 2

(a) The inactivation of E. coli using different processes and (b) photoreactivation of E. coli after treatment (UV intensity = 16 W/m2; sparging gas flow rate = 1.5 L/min; inlet O3 concentration = 8.4 mg/L; white fluorescent light = 1.1 × 104 lx).

Figure 2

(a) The inactivation of E. coli using different processes and (b) photoreactivation of E. coli after treatment (UV intensity = 16 W/m2; sparging gas flow rate = 1.5 L/min; inlet O3 concentration = 8.4 mg/L; white fluorescent light = 1.1 × 104 lx).

Inactivation may be temporary or permanent, depending upon the type and dose of disinfectant used (Magbanua et al. 2006). As shown in Figure 2(b), photoreactivation of E. coli occurred in samples of the UV-only process, and the count of reactivation bacteria cells was positive with exposure time, especially in the initial phase. After exposure to white fluorescent light for 1 h, bacteria counts were approximately ten times higher than that obtained right after UV radiation. Conversely, extended bacterial mortality was observed after reactions with the presence of ozone, suggesting that the UV/O3 process could overcome the limitation of bacteria regrowth in the UV-only process.

Several studies have shown that HO• is mainly responsible for pollutant degradation in the UV/O3 process (Illés et al. 2014). Thus, the significance of HO• for inactivation of target bacteria was investigated with qualitative analysis and quantitative computation. As shown in Figure 3, aqueous ozone concentration increased with aeration time. The pCBA (an HO• probe compound) degradation rate in the combined UV/O3 system was faster than that in the ozone-only system, corresponding with the results of ozone decomposition. The steady-state concentrations of HO• calculated from the slope of the curves in Figure 3 were 2.6 × 10−13 mol/L and 5.6 × 10−13 mol/L for the O3 and UV/O3 processes, respectively.
Figure 3

Time-dependent decay of pCBA under UV/O3 irradiation and ozonation alone. The insert shows the ozone concentration changes with aeration time (UV intensity = 16 W/m2; sparging gas flow rate = 1.5 L/min; inlet O3 concentration = 8.4 mg/L).

Figure 3

Time-dependent decay of pCBA under UV/O3 irradiation and ozonation alone. The insert shows the ozone concentration changes with aeration time (UV intensity = 16 W/m2; sparging gas flow rate = 1.5 L/min; inlet O3 concentration = 8.4 mg/L).

In the current system, HO• could be produced via the reaction between ozone and OH (Equations (1) to (4)) (Magbanua et al. 2006). Under UV irradiation, photolysis of O3 catalyzed the production of HO• (Equations (5) and (6)) (von Gunten 2003): 
formula
1
 
formula
2
 
formula
3
 
formula
4
 
formula
5
 
formula
6
It should be noted that the present purpose is different from the degradation of organic compounds, which relies primarily on the generation of HO• for its removal effect. The above observations implied that UV and ozone were mainly responsible for inducing the inactivation of E. coli in the combined tests, and the produced intermediate HO• could accelerate bacteria inactivation. While DNA damage caused by UV-C could be self-repaired in some bacteria (Shang et al. 2009), the decomposition of bacterial cell structure induced by ozone may be irreparable. The simultaneous use of UV irradiation and ozone can not only cause cell DNA damage, but also destroy cell structure, which is a possible reason for the repression of bacteria regrowth after UV/O3 treatment. Furthermore, the produced intermediate HO• are a much stronger oxidant than O3 (von Gunten 2003). The destruction of bacteria cell structure might be expedited in the UV/O3 process, thereby accelerating the killing of bacteria. To verify these hypotheses, the changes of cell structure in different disinfection processes were examined, as described and discussed in the following.

Destruction of cell structure checked by AFM

To investigate the bactericidal mechanism of UV/O3, the morphologies of E. coli cells from different bactericidal experiments were characterized by AFM (Figure 4). It can be seen that no obvious surface damage on cells was observed after 60 s UV radiation treatment (Figures 4(a) and (b)). In contrast, the morphology of the E. coli cells had changed greatly after 30 s O3 treatment (Figures 4(c) and (d)), with some rumples and small holes being shown in the cell surface. With the UV/O3 combined treatment (Figures 4(e) and (f)), the damage was more severe than with O3 acting alone. The cells were not intact anymore, and parts of the cell wall and cell membrane disappeared, leading to an obvious leakage of the interior components.
Figure 4

AFM images of E. coli: UV-treated for 60 s ((a) and (b)), O3-treated for 30 s ((c) and (d)), UV/O3-treated for 30 s ((e) and (f)). Images of individual bacteria are shown in (a), (c) and (e) and their corresponding zoomed-in surface structures are shown respectively in (b), (d) and (f).

Figure 4

AFM images of E. coli: UV-treated for 60 s ((a) and (b)), O3-treated for 30 s ((c) and (d)), UV/O3-treated for 30 s ((e) and (f)). Images of individual bacteria are shown in (a), (c) and (e) and their corresponding zoomed-in surface structures are shown respectively in (b), (d) and (f).

Cell structure damage was an important cause of bacteria inactivation. The above AFM investigation indicated that ozone played a critical role in bacteria cell disruption, and the presence of UV radiation during ozonation expedited the decomposition of cell structure. This may have been mainly due to the improved generation of HO• in the UV/O3 process. On the other hand, cell damage induced by UVC radiation during UV/O3 disinfection might depress cell antioxidant capacity (Santos et al. 2013), thereby enhancing the bactericidal effects of ozone and HO•.

Evidence of cell membrane damage

The mechanism of cell inactivation was further investigated by examining the formation of MDA, which is interpreted as a lipid peroxidation product formed from the oxidation of the cell membrane. As shown in Figure 5(a), UV irradiation minimally damaged the cell membrane, with only 0.42 μmol/L MDA detected after 180 s. For disinfection with ozone alone, the profile of MDA formation was similar to that obtained for the combined UV/O3 process, except that the latter resulted in the oxidation of the cell membrane more rapidly. For instance, with an exposure time of 30 s, the MDA produced in the O3 and UV/O3 processes was 2.4 μmol/L and 3.0 μmol/L, respectively. A subsequent decline in the MDA concentration was obtained in reactions with ozone, indicating that the formed MDA was further oxidized.
Figure 5

Concentrations of (a) MDA and (b) leaked K+ change with reaction time using different processes (UV intensity = 16 W/m2; sparging gas flow rate = 1.5 L/min; inlet O3 concentration = 8.4 mg/L).

Figure 5

Concentrations of (a) MDA and (b) leaked K+ change with reaction time using different processes (UV intensity = 16 W/m2; sparging gas flow rate = 1.5 L/min; inlet O3 concentration = 8.4 mg/L).

The cell membrane plays a vital role in providing a barrier of selective permeability for bacteria. The leakage of K+, which exists universally in bacteria and plays an important role in protein synthesis and regulation of polysome content (Ren et al. 2009), was utilized to reflect the permeability of the cell membrane. Similar to the results of MDA formation (Figure 5(a)), the curves described in Figure 5(b) suggest that UV radiation was ineffective for K+ leakage, while ozone caused the leaked K+ concentration to increase significantly. The K+ leakage rate in the UV/O3 process was faster than that in the ozone-only test, with approximately 1.05 mg/L and 0.84 mg/L K+ being detected after 10 s, respectively. The maximum K+ concentration in the ozone-only or UV/O3 tests was about 1.23 mg/L at 180 s treatment.

The above observations agree with the results of AFM (Figure 4), that the combination of ozone with UV radiation expedited the destruction of bacterial cell structure, and then accelerated the leakage of intracellular components. The gradual leakage of intracellular substances from impaired cells might also induce the inactivation of bacteria after O3 and UV/O3 exposure, which was probably a reason for the extended bacterial mortality that is depicted in Figure 2(b).

CONCLUSIONS

This work revealed that the application of ozone in conjunction with UV radiation could hasten and enhance the inactivation of E. coli cells. UV and ozone were mainly responsible for E. coli inactivation in the coupled process, and the improved disinfection efficiency could have partly resulted from the abundant production of HO•, which expedited the destruction of bacterial cell structures. Restated, not only can the combined treatment synergistically inactivate harmful cells in water, but also the limitations of using ozone and UV separately can be reduced. It is concluded that UV/O3 is an efficient technology for water disinfection, and the multiple forms of damage caused by ozone, UV and HO• result in enhanced inactivation efficiency.

ACKNOWLEDGEMENTS

We are grateful for grants from Natural Science Foundation of Jiangsu Province (No. BK20,130,835), China Postdoctoral Science Foundation (No. 2013M541,600), Fundamental Research Funds for the Central Universities of Hohai University (No. 2013B13,020,026), and A Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions.

REFERENCES

REFERENCES
APHA, AWWA, WEF
1998
Standard Methods for the Examination of Water and Wastewater
, 20th edition.
American Public Health Association
,
Washington, DC
,
USA
.
Beck
S. E.
Rodriguez
R. A.
Linden
K. G.
Hargy
T. M.
Larason
T. C.
Wright
H. B.
2014
Wavelength dependent UV inactivation and DNA damage of adenovirus as measured by cell culture infectivity and long range quantitative PCR
.
Environmental Science & Technology
48
(
1
),
591
598
.
Bustos
Y. A.
Vaca
M.
Lopez
R.
Torres
L. G.
2010
Disinfection of a wastewater flow treated by advanced primary treatment using O3, UV and O3/UV combinations
.
Journal of Environmental Science and Health. Part A, Toxic/Hazardous Substances & Environmental Engineering
45
(
13
),
1715
1719
.
Cho
M.
Kim
J.
Kim
J. Y.
Yoon
J.
Kim
J. H.
2010
Mechanisms of Escherichia coli inactivation by several disinfectants
.
Water Research
44
(
11
),
3410
3418
.
Fang
J. Y.
Liu
H. L.
Shang
C. I.
Zeng
M. Z.
Ni
M. L.
Liu
W.
2014
E. coli and bacteriophage MS2 disinfection by UV, ozone and the combined UV and ozone processes
.
Frontiers of Environmental Science & Engineering
8
(
4
),
547
552
.
Illés
E.
Szabó
E.
Takács
E.
Wojnárovits
L.
Dombi
A.
Gajda-Schrantz
K.
2014
Ketoprofen removal by O3 and O3/UV processes: kinetics, transformation products and ecotoxicity
.
Science of the Total Environment
472
,
178
184
.
Lian
L.
Yue
E.
Li
J.
Blatchley
E. R.
2014
Volatile disinfection byproducts resulting from chlorination of uric acid: implications for swimming pools
.
Environmental Science & Technology
48
(
6
),
3210
3217
.
Liu
J.
Zhou
L.
Chen
J. H.
Mao
W.
Li
W. J.
Hu
W.
Wang
S. Y.
Wang
C. M.
2014
Role of ozone in UV-C disinfection, demonstrated by comparison between wild-type and mutant conidia of Aspergillus niger
.
Photochemistry and Photobiology
90
(
3
),
615
621
.
Lyon
B. A.
Farre
M. J.
De Vera
G. A.
Keller
J.
Roux
A.
Weinberg
H. S.
Gernjak
W.
2014
Organic matter removal and disinfection byproduct management in South East Queensland's drinking water
.
Water Science & Technology: Water Supply
14
(
4
),
681
689
.
Magbanua
B. S.
Jr.
Savant
G.
Truax
D. D.
2006
Combined ozone and ultraviolet inactivation of Escherichia coli
.
Journal of Environmental Science and Health. Part A, Toxic/Hazardous Substances & Environmental Engineering
41
(
6
),
1043
1055
.
Poepping
C.
Beck
S. E.
Wright
H.
Linden
K. G.
2014
Evaluation of DNA damage reversal during medium-pressure UV disinfection
.
Water Research
56
,
181
189
.
Ren
J.
Wang
W.
Sun
S.
Zhang
L.
Chang
J.
2009
Enhanced photocatalytic activity of Bi2WO6 loaded with Ag nanoparticles under visible light irradiation
.
Applied Catalysis B: Environmental
92
(
1–2
),
50
55
.
Santos
A. L.
Oliveira
V.
Baptista
I.
Henriques
I.
Gomes
N. C.
Almeida
A.
Correia
A.
Cunha
A.
2013
Wavelength dependence of biological damage induced by UV radiation on bacteria
.
Arch. Microbiol
195
(
1
),
63
74
.
Shang
C.
Cheung
L. M.
Ho
C.-M.
Zeng
M.
2009
Repression of photoreactivation and dark repair of coliform bacteria by TiO2-modified UV-C disinfection
.
Applied Catalysis B: Environmental
89
(
3–4
),
536
542
.
Sisti
M.
Pieretti
B.
De Santi
M.
Brandi
G.
2014
Inactivation of pathogenic dermatophytes by ultraviolet irradiation in swimming pool thermal water
.
International Journal of Environmental Health Research
24
(
5
),
412
417
.
Venosa
A. D.
Petrasek
A. C.
Brown
D.
Sparks
H. L.
Allen
D. M.
1984
Disinfection of secondary effluent with ozone/UV
.
Journal of the Water Pollution Control Federation
56
(
2
),
137
142
.
Wu
D.
You
H.
Jin
D.
Li
X.
2011
Enhanced inactivation of Escherichia coli with Ag-coated TiO2 thin film under UV-C irradiation
.
Journal of Photochemistry and Photobiology A: Chemistry
217
(
1
),
177
183
.
Zhao
G.
Lu
X.
Zhou
Y.
Gu
Q.
2013
Simultaneous humic acid removal and bromate control by O3 and UV/O3 processes
.
Chemical Engineering Journal
232
,
74
80