Stronger disinfection techniques are required to inactivate Bacillus subtilis spores as surrogate microorganisms for Cryptosporidium parvum oocysts. In this study, the effects of UV and persulfate separately and also in combination were investigated on B. subtilis spore inactivation. Central composite design and response surface methodology were used to optimize target microorganism reduction. Contact time, initial pH, and persulfate dosage were considered as input experimental variables. Based on the design of the experiments, first and second order response surface models have been developed to correspond to the output response of B. subtilis spore reduction. It can be concluded that microbial reduction by UV alone was more effective than persulfate, while the combined UV/persulfate process demonstrated the highest log reduction (4.1) under the following optimal conditions: 60 min contact time, pH = 7.8, and persulfate dosage of 30 mM. On the other hand, the optimal condition for UV treatment was a contact time of 60 min at a pH of 5.0, which led to a 3.19 log spore inactivation. Consequently, the UV/persulfate system can be introduced as an alternative disinfectant for the inactivation of B. subtilis spores.

INTRODUCTION

Disinfection is a very crucial step in water treatment processes for destroying pathogenic microorganisms (Post et al. 2011). Traditionally, chlorination has been the most commonly adopted disinfection process for the treatment of drinking water. However, with the appearance of chlorine-resistant pathogenic microorganisms such as Giardia lamblia cysts and Cryptosporidium parvum oocysts, many conventional drinking water treatment plants that use free chlorine are faced with problems in reaching the required target goals for microbial inactivation (Jung et al. 2008). Since the direct measurement of C. parvum or G. lamblia is difficult, time-consuming and expensive, Bacillus subtilis spores were selected as an indicator (Facile et al. 2000; Driedger et al. 2001).

Bacillus subtilis is an aerobic, non-pathogenic, gram-positive soil bacterium that is easy to count and identify in laboratory conditions. Under certain conditions (e.g., starvation), spores are formed that are highly resistant to disinfection (Driedger et al. 2001; Radziminski et al. 2002). It has been reported that the Bacillus spores are 10–75 times more resistant than its vegetative form (Blatchley et al. 2005).

Due to several disadvantages associated with chlorination, current studies have extensively investigated alternative disinfection processes such as ozone, UV irradiation, chlorine dioxide, chloramines, plasma discharge and electrochemical processes as well as the sequential or combined processes for the removal of resistant microorganisms (Radziminski et al. 2002; Larson & Mariñas 2003; Cho et al. 2006; Jung et al. 2008, 2010; Mezule et al. 2014). Today most of the advanced oxidation process studies in water and wastewater disinfection have focused on hydroxyl radicals (OH·); however, only a few studies discuss other radical species. For example, sulfate radicals (SO4•−) have a significantly high reduction potential of 2.5–3.1 V, and that of OH· is 1.9–2.9 V (Lin et al. 2011; Gao et al. 2012; Wordofa 2014). Moreover, SO4•− also exhibits a nonselective oxidation pattern, and is capable of quickly decomposing most of the organic pollutants and biological toxins in water (Lin et al. 2011; Gao et al. 2012; Tan et al. 2013; Xie et al. 2015). SO4•− is produced via activation of peroxydisulfate (S2O82−, commonly known as persulfate) with UV, catalysts (e.g., Mn2+, Fe2+), heat and high pH (Roshani & Leitner 2011; Gao et al. 2012; Tan et al. 2013; Lutze et al. 2015; Xie et al. 2015). Among these methods, UV-based oxidation has a high potential destruction rate due to the production of more powerful components such as sulfate radicals (Tan et al. 2013). Nevertheless, there is insufficient information about the application of sulfate radicals in water disinfection processes and especially for inactivating B. subtilis spores.

Response surface methodology (RSM) is a compilation of statistical and mathematical techniques valuable for developing, improving, and optimizing processes (Carley et al. 2004). Prior to applying the RSM, an experimental design was implemented to determine which experiments should be carried out in the experimental region being studied (Bezerra et al. 2008). RSM was used to model and optimize spore reduction (Huang et al. 2007). In this study, a central composite design (CCD) was used as a novel approach to optimize the log reduction of B. subtilis spores by the UV/persulfate process. Contact time, pH, and concentration of persulfate solution were studied as key parameters affecting disinfection in this investigation.

MATERIALS AND METHODS

The materials used in this study included sodium persulfate (Merck, No. 106609), NaOH (Merck, No. 105587), H2SO4 (Merck, No. 100731), sodium thiosulfate (Merck, No. 106512) and NaCl (Merck, No. 106406). Necessary stock solutions were made from the above-mentioned neat materials and double distilled water. All glassware was washed using tap water and rinsed with distilled water. All of the instruments were sterilized by autoclaving at 121 °C for 15 min prior to use.

Spore production and sample preparation

Bacillus subtilis (ATCC 6633) vegetative cells were cultured in tryptic soy agar (TSA, Merck No. 105458) medium and incubated for 24 h at 35 °C. In vitro production of B. subtilis spores was conducted by cultivating the vegetative cells at 37 °C on the antibiotic assay medium 1 (Himedia, M003) supplemented by 0.3 g L−1 MnSO4 for 5–7 days (Radziminski et al. 2002). Subsequently, the spore suspension was washed from the top of the medium by saline water and transferred to 10 mL sterile centrifuge tubes. The suspension was centrifuged for 15 min at 4,000 g and stored in a 4 °C refrigerator until use.

For sample preparation, 4 mL of the spore suspension was poured into 500 mL of sterilized drinking water. For further inactivation of vegetative cells, the samples were heated up to 80 °C and kept for 12 min at this temperature (Larson & Mariñas 2003). The turbidity of this suspension was between 25 and 32 Nephelometric Turbidity Units (NTU), which led to the production of 106–108 spores/mL.

Experimental design

The CCD was employed to investigate the process of B. subtilis inactivation using UV, persulfate, and the combination of UV/persulfate. The RSM was used to determine the combined effects of contact time (X1), pH (X2), and persulfate dosage (X3) on log inactivation of B. subtilis (Y) as the output response. The study was divided into three different phases including the UV, persulfate, and combined UV/persulfate phases. In the first phase, 17 experiments were designed to determine the UV radiation effect on B. subtilis inactivation. In the second phase, a design considering the three above-mentioned controllable variables was carried out. An orthogonal CCD in two cube and star blocks was designed. A total of 26 experiments were carried out consisting of 12 center points, 23 = 8 design points, and 2 × 3 = 6 axial points. The design of the third phase was similar to that of the second phase except for including a constant UV irradiation in the last phase. Bacillus subtilis spore removal served (Y) as the dependent variable of the process. The actual and coded levels of the design are shown in Table 1. The design, mathematical modeling, and optimization of this work were implemented by the RSM package (Lenth 2009). There are two important models in the RSM: first order and second order model. A general explanation of the first order model and the quadratic model is based on Equations (1) and (2), respectively (Carley et al. 2004; Khuri & Mukhopadhyay 2010). 
formula
1
 
formula
2
where y is the response; Xi and Xj are the variables; β0 is the constant coefficient; βi, βj, βij are the interaction coefficients of linear, quadratic, and second order terms, respectively; k is the number of studied parameters; and is the system error.
Table 1

Coded and real values of the Independent variables used

VariablesSymbolRange and levels
− 101
Contact time (min) X1 16.1 32.5 48.8 60 
pH X2 6.2 9.7 11 
Concentration of persulfate (mM) X3 8.4 16.5 24.5 30 
VariablesSymbolRange and levels
− 101
Contact time (min) X1 16.1 32.5 48.8 60 
pH X2 6.2 9.7 11 
Concentration of persulfate (mM) X3 8.4 16.5 24.5 30 

The effect of each factor was evaluated by analysis of variance (ANOVA) (95 and 99%). The model permitted the evaluation of the effects of linear, quadratic and interactive terms of the independent variables on the dependent variable. The response surface and contour plots were drawn to illustrate the main and interactive effects of the independent variables on disinfection efficiency. The optimum values of the selected variables were obtained by solving the regression equation and by analyzing the response surface plot.

Experimental procedure

Disinfection experiments for the three candidate disinfection processes (persulfate alone, UV alone and combined UV/persulfate process) were performed at different contact times (5–60 min) and pH ranges of 5–11. It should be noted that a contact time exceeding 60 min is not convenient in the disinfection process. In addition, the range of the initial pH was chosen according to previous literatures (Lin et al. 2011; Tan et al. 2013; Wordofa 2014; Xie et al. 2015). The persulfate concentration was also varied from 3 to 30 mM, and the minimum persulfate dosage was chosen based on a study of the effects of sulfate radicals on the reduction of E. coli (Wordofa 2014), while the maximum amount was selected according to the fact that the spores are more resistant microorganisms.

In order to estimate the enhancement resulting from the combined process, inactivation efficiencies were compared with those of individual processes. Both the UV and UV/persulfate processes were performed using a low pressure UV lamp, which emitted monochromatic UV radiation at 254 nm. The UV intensity was measured as 112 μW/cm2 at the distance of 125 mm between the lamp and water surface. The dose can be calculated by multiplying the mentioned intensity by the contact times and is presented as mW.s/cm2.

Fifty milliliters of synthetically prepared sample was poured into a sterile petri dish (90 mm diameter), and the solution's pH was adjusted to the desired value using a few drops of H2SO4 or NaOH. In the UV experiments, the sample was put under a UV lamp and gently stirred throughout the exposure period. In the case of the persulfate and UV/persulfate procedures, the persulfate stock solution was prepared in such a way that only 1 mL was required to provide the defined persulfate concentration. The sample was gently stirred throughout the contact time. At the end of each contact time, residual persulfate was neutralized by adding 1 mL of sodium thiosulfate based on a mole to mole ratio. At the end of each exposure, 10 mL of the disinfected sample was collected for microbial analysis.

Microbiological analysis

Viability was determined by the membrane filtration technique using nitrocellulose filters (Sartorius Stedim, No. 11406) with a 47 mm diameter and a nominal pore size of 0.45 μm (Larson & Mariñas 2003). For each sample, a dilution series was made by transferring 10 mL of sample to a 100 mL dilution bottle containing 90 mL of 9% saline water. After filtration, the filters were placed in a petri dish containing 5–6 mL of nutrient agar (Merck, No. 1.05450) and 0.015 mg L−1 of trypan blue (Titrachem, No. 512). Colonies were counted after incubation at 37 °C for 24 h (Larson & Mariñas 2003).

RESULTS AND DISCUSSION

CCD and regression models

In this study, the effects of contact time, pH, and persulfate dosage were investigated on B. subtilis spore inactivation separately, and also in combination with UV and persulfate. The experiments were performed using a CCD, and the corresponding results for the UV and UV/persulfate processes are shown in Tables 2 and 3, respectively. Multiple regression analysis of three experimental phases is presented in Table 4. Multiple R2, adjusted R2, and p-value were used to assess the model adequacy. Both contact time and pH were significant in the UV process alone, while no interaction was observed between the variables (Table 4). It should be mentioned that terms which could show interactions were omitted from the final model to optimize the model, since these were not statistically meaningful.

Table 2

Arrangement of the CCD for the two independent variables used in the UV process

Run no.Exposure time (X1)pH (X2)Log inactivation (Y)
32.5 1.448158 
13 10 0.062148 
32.5 1.477121 
13 5.8 0.029963 
52 10 1.833669 
52 5.8 2.435729 
32.5 1.462398 
32.5 1.492361 
32.5 1.246672 
10 32.5 11 1.176091 
11 
12 32.5 1.477121 
13 32.5 1.455932 
14 32.5 1.778151 
15 60 3.259637 
16 32.5 1.641431 
17 32.5 1.632023 
Run no.Exposure time (X1)pH (X2)Log inactivation (Y)
32.5 1.448158 
13 10 0.062148 
32.5 1.477121 
13 5.8 0.029963 
52 10 1.833669 
52 5.8 2.435729 
32.5 1.462398 
32.5 1.492361 
32.5 1.246672 
10 32.5 11 1.176091 
11 
12 32.5 1.477121 
13 32.5 1.455932 
14 32.5 1.778151 
15 60 3.259637 
16 32.5 1.641431 
17 32.5 1.632023 
Table 3

Arrangement of the CCD for the three independent variables in the UV/persulfate process

Run no.Contact time (X1)pH (X2)Persulfate concentration (X3)Log inactivation (Y)
ActualPredicted
16 24.5   
49 24.5 2.30103 2.277813 
16 10 8.5 0.330993 0.784313 
32.5 16.5 1.315753 1.118263 
32.5 16.5 0.90309 1.118263 
16 10 24.5 1.346787 1.577113 
32.5 16.5 0.989276 1.118263 
32.5 16.5 0.929962 1.118263 
16 8.5 0.167491 0.867913 
10 49 10 8.5 1.69897 1.401413 
11 32.5 16.5 1.610834 1.118263 
12 49 10 24.5 1.69897 2.194213 
13 49 8.5 1.802719 1.485013 
14 32.5 16.5 1.022276 1.118263 
15 32.5 16.5 1.014723 1.118263 
16 32.5 16.5 1.875061 1.118263 
17 32.5 30 4.051153 3.58235 
18 32.5 2.363178 2.2445 
19 32.5 16.5 0.937852 0.571288 
20 32.5 16.5 1.034762 1.118263 
21 32.5 16.5 0.967815 1.118263 
22 32.5 16.5 1.034762 0.691363 
23 60 16.5 0.937852 1.632513 
24 32.5 16.5 1.210853 1.118263 
25 32.5 11 16.5 0.967815 0.565963 
26 16.5 0.69897 0.604013 
Run no.Contact time (X1)pH (X2)Persulfate concentration (X3)Log inactivation (Y)
ActualPredicted
16 24.5   
49 24.5 2.30103 2.277813 
16 10 8.5 0.330993 0.784313 
32.5 16.5 1.315753 1.118263 
32.5 16.5 0.90309 1.118263 
16 10 24.5 1.346787 1.577113 
32.5 16.5 0.989276 1.118263 
32.5 16.5 0.929962 1.118263 
16 8.5 0.167491 0.867913 
10 49 10 8.5 1.69897 1.401413 
11 32.5 16.5 1.610834 1.118263 
12 49 10 24.5 1.69897 2.194213 
13 49 8.5 1.802719 1.485013 
14 32.5 16.5 1.022276 1.118263 
15 32.5 16.5 1.014723 1.118263 
16 32.5 16.5 1.875061 1.118263 
17 32.5 30 4.051153 3.58235 
18 32.5 2.363178 2.2445 
19 32.5 16.5 0.937852 0.571288 
20 32.5 16.5 1.034762 1.118263 
21 32.5 16.5 0.967815 1.118263 
22 32.5 16.5 1.034762 0.691363 
23 60 16.5 0.937852 1.632513 
24 32.5 16.5 1.210853 1.118263 
25 32.5 11 16.5 0.967815 0.565963 
26 16.5 0.69897 0.604013 
Table 4

Regression coefficients and their significance

Model termCoefficient estimateStandard errort valuep(>|t|)
UV 
 Intercept 0.216537 0.358525 0.604 0.5555 
 Time 0.056403 0.004398 12.824 3.97e-09*** 
 pH −0.080641 0.040558 −1.988 0.0667 
R2: 0.9233, Adjusted R2: 0.9123, F-statistic: 84.21 on 2 and 14 DF, p-value: 1.568e-08 
Persulfate 
 Intercept 0.047522 0.446634 0.106 0.916 
 pH 0.041435 0.046474 0.892 0.382 
 Time −0.003002 0.004892 −0.614 0.546 
 Per. concentration −0.002442 0.011100 −0.220 0.828 
R2: 0.05254, Adjusted R2: −0.07666, F-statistic: 0.4067 on 3 and 22 DF, p-value: 0.7497 
Persulfate/UV 
 Intercept 1.118091 0.112403 9.9471 0.459e-09*** 
 Time 0.514252 0.172432 2.9824 0.007362** 
 pH −0.062135 0.178168 −0.3487 0.730928 
 Per. concentration 0.667458 0.191492 3.4856 0.002332** 
 I (pH^2) −0.489495 0. 273205 −1.7917 0.088330 
 I (Per. concentration^2) 1.794601 0.291788 6.15045 0.215e-06*** 
F-statistic: 12.39, on 5 and 20 DF p-value: 1.456e-05 
Model termCoefficient estimateStandard errort valuep(>|t|)
UV 
 Intercept 0.216537 0.358525 0.604 0.5555 
 Time 0.056403 0.004398 12.824 3.97e-09*** 
 pH −0.080641 0.040558 −1.988 0.0667 
R2: 0.9233, Adjusted R2: 0.9123, F-statistic: 84.21 on 2 and 14 DF, p-value: 1.568e-08 
Persulfate 
 Intercept 0.047522 0.446634 0.106 0.916 
 pH 0.041435 0.046474 0.892 0.382 
 Time −0.003002 0.004892 −0.614 0.546 
 Per. concentration −0.002442 0.011100 −0.220 0.828 
R2: 0.05254, Adjusted R2: −0.07666, F-statistic: 0.4067 on 3 and 22 DF, p-value: 0.7497 
Persulfate/UV 
 Intercept 1.118091 0.112403 9.9471 0.459e-09*** 
 Time 0.514252 0.172432 2.9824 0.007362** 
 pH −0.062135 0.178168 −0.3487 0.730928 
 Per. concentration 0.667458 0.191492 3.4856 0.002332** 
 I (pH^2) −0.489495 0. 273205 −1.7917 0.088330 
 I (Per. concentration^2) 1.794601 0.291788 6.15045 0.215e-06*** 
F-statistic: 12.39, on 5 and 20 DF p-value: 1.456e-05 

**p<0.01, ***p<0.001.

The difference between R2 and R2adj, less than 0.2, as well as a lower p-value (1.56 × 10−8), suggests that the model is reliable. All of the linear coefficients in the persulfate process were not significant (Table 4). A lower value of model R2 and significant difference of more than 0.2 between R2 and R2adj shows that the model is not significant and acceptable; moreover, the higher p-value of 0.749 implies that there is a weak correlation between the variables and the response produced by the model.

The results of the reduced quadratic model for the UV/persulfate process are also shown in Table 4. All linear coefficients, except pH, and two quadratic terms (X22 and X32), were significant (p < 0.05), while no significant interaction was found (p > 0.05). In addition, the model R2 is promising and it is very consistent with R2adj (Table 5). Statistical testing of the model was done in the form of ANOVA analysis, which is required to test the significance and adequacy of the model. A summary of ANOVA results for the selected quadratic model for the UV/persulfate process is provided in Table 5. The higher F value of 19.64 and the lower probability value of p < <0.05 indicates that the model is highly significant. The model also indicates a statistically insignificant lack of fit.

Table 5

ANOVA for the developed RSM model for the UV/persulfate process

SourceR2Adjusted R2DfSum sqMean sqF valuePr > F
Model 0.756 0.695 6.9288 3.4644 19.6445 >0.00001 
Residuals – – 20 3.5271 0.1764 – – 
Lack of fit – – 10 2.5461 0.2546 2.5955 0.074224 
Pure error – – 10 0.9810 0.0981 – – 
SourceR2Adjusted R2DfSum sqMean sqF valuePr > F
Model 0.756 0.695 6.9288 3.4644 19.6445 >0.00001 
Residuals – – 20 3.5271 0.1764 – – 
Lack of fit – – 10 2.5461 0.2546 2.5955 0.074224 
Pure error – – 10 0.9810 0.0981 – – 

Response surface and contour plotting

The graphical representations of the regression models, called the response surfaces and the contour plots, were obtained using the R software (Figures 1 and 2). In the case of the UV process, it is clear that the spore reduction increased in a pH range of 5 and 11 from less than 0.5 to 3 log in a 5 to 60 min duration (Figure 1). The results obtained for inactivating B. subtilis spores (less than 1 log) with persulfate alone in conditions used in this study (i.e., time: 5–60 min, pH: 5–11, and persulfate concentration 3–30 mM) were not promising.
Figure 1

Contour plot of the effect of pH and exposure time on log inactivation of Bacillus subtilis spores by UV irradiation.

Figure 1

Contour plot of the effect of pH and exposure time on log inactivation of Bacillus subtilis spores by UV irradiation.

Figure 2

Contour plot of log inactivation of Bacillus subtilis spores by a combined UV/persulfate process. Shows the effect of (a) persulfate concentration and pH, (b) persulfate concentration and contact time, (c) pH and contact time, and their mutual interactions on B. subtilis spore inactivation.

Figure 2

Contour plot of log inactivation of Bacillus subtilis spores by a combined UV/persulfate process. Shows the effect of (a) persulfate concentration and pH, (b) persulfate concentration and contact time, (c) pH and contact time, and their mutual interactions on B. subtilis spore inactivation.

The results (Table 4 and Figure 1) show that the most important factor affecting disinfection efficiency in the UV alone process was exposure time, and the reduction pattern was in good consistency with the results reported in the literature (Cho et al. 2006). According to Cho et al. (Cho et al. 2006), pH did not affect spore inactivation in the UV254 process, and a 2 log reduction was achieved at both pH 5.6 and 8.2. Moreover, Mamane et al. (Mamane et al. 2007) did not obtain any spore inactivation at higher wavelengths (>295 nm).

The next step was to investigate the effect of activated persulfate by UV irradiation on B. subtilis spore inactivation. The perspective plot illustrates how the microbial reduction varied with the independent variables in the combined UV/persulfate process (Figure 2). Bacillus subtilis spore reduction as a function of persulfate concentration and pH at a fixed time (31.38 min) is shown in Figure 2(a). A higher spore reduction was achieved when the persulfate concentration increased from 25 to 30 mM and the pH varied from 7 to 9. Figure 2(b) shows the effect of persulfate concentration and contact time at a fixed pH of 8. The reduction of B. subtilis spores in the persulfate dosage range of 23 to 30 mM increased from 2 to 4 log as a function of increased time. The effect of pH and time sliced at 16.5 mM persulfate dosage are displayed in Figure 2(c). The reduction of B. subtilis spores increased with time at a slightly alkaline pH (i.e., 7–9), reaching a 1.5 log reduction after 30–60 minutes of contact time.

The persulfate dosage can significantly affect disinfection efficiency in this process because higher concentrations lead to more sulfate radical generation. Our results, in line with other studies (Tan et al. 2013; Wordofa 2014), suggested that oxidation potential can be promoted in higher pH conditions and higher persulfate dosage. However, some reports have shown that sulfate radical generation and their efficacy were not affected by pH (Lin et al. 2011; Xie et al. 2015). According to the results, a maximum disinfection performance was achieved in the combined UV/persulfate process in a pH range of 7 to 9. Tan et al. (Tan et al. 2013) attributed this effect to increased activation energy in the UV/persulfate system at a high pH and suggested two associated reasons. First, activated persulfate was augmented at alkaline pH (Equation (3)), and second, in the presence of hydroxyl ions (OH) some of the peroxydisulfate converts to hydroxyl radicals (Equation (4)) (Driedger et al. 2001). Generally, higher pH, contact time, and persulfate dosage result in a higher oxidation potential (Lin et al. 2011; Wordofa 2014; Xie et al. 2015). 
formula
3
 
formula
4

Process optimization

Optimization formulas were obtained (Equations (5)–(7)) according to the regression models (Table 4) for UV and persulfate alone and the combined disinfection process. 
formula
5
 
formula
6
 
formula
7
where Y1, Y2, Y3 represent log inactivation of B. subtilis spores by the UV, persulfate and UV/persulfate processes, respectively; X1 is the contact time; X2 is the pH of the solution; and X3 is the persulfate concentration in the sample. The above equations were solved using Solver add-Ins in Microsoft Excel 2013, and the optimization results for the three processes are presented in Table 6. Based on these results, the maximum inactivation of B. subtilis spores was predicted at 4.1 log after 60 min of treatment by the combined UV/persulfate process at pH 7.8 and 30 mM persulfate dosage. At this optimization point, the UV dose can be calculated by multiplying the lamp intensity (112 μW/cm2) by optimum contact times (60 min), which is equal to 6.72 mW.s/cm2. According to the results shown in Figures 1 and 2, and the results predicted in the optimized conditions, it is clear that the predicted amounts are consistent with those obtained in the laboratory. Although a higher amount of reduction out of the boundary of the space that was studied is possible, it does not seem to be practically feasible, since the conventional disinfection period is about 30–60 min. Therefore, the optimum reduction level was limited to values which are achievable by changing the independent variables as defined in the study.
Table 6

Optimization results of B. subtilis spore reduction by UV, persulfate and combined UV/persulfate processes

ProcessContact time (min)pHPersulfate dosage (mM)Log inactivation
UV 60 – 3.19 
Persulfate 11 0.48 
UV/Persulfate 60 7.8 30 4.1 
ProcessContact time (min)pHPersulfate dosage (mM)Log inactivation
UV 60 – 3.19 
Persulfate 11 0.48 
UV/Persulfate 60 7.8 30 4.1 

Based on the optimal conditions, the disinfection potential of sulfate radicals in comparison with hydroxyl radicals generated by O3/UV (Jung et al. 2008) and H2O2/UV (Cho et al. 2011) processes is 1.6 and 0.8 log higher in conditions similar to those used in this study, respectively. On the other hand, Wordofa (Wordofa 2014) reported 3.5 log inactivation of E. coli at lower persulfate concentrations, and longer contact times than those used in this research; however, it should be considered that spores are much hardier than bacteria.

Scanning electron microscopy images

Figure 3 shows scanning electron microscopy (SEM, ZEISS, DSM960A, Germany) images of the extent of spore coat damage of B. subtilis spores under optimized conditions conducted in the laboratory. As shown in Figure 3(a), the unexposed B. subtilis spores were ellipsoidal, and the main spore structure (i.e., inner and outer coat layers) was clear. Figure 3(b) shows the effects of UV irradiation on the spore's structure as pits in the spore's coat as well as transformed and amorphous morphology. A comparison of spores treated with persulfate (Figure 3(c)) and control spores (Figure 3(a)) revealed that the spore coats were unaffected, and no distinct differences were observed. Analysis by SEM image showed damaged and ruptured B. subtilis spore coats and cores after treatment by combined UV/persulfate (Figure 3(d)).
Figure 3

Scanning electron microscope (SEM) images of Bacillus subtilis spores under optimal conditions in different processes. (a) Unexposed, (b) UV alone (t = 60 s, pH = 5), (c) persulfate (t = 5 s, pH = 11, C = 3 mM), (d) UV/persulfate (t = 60 s, pH = 7.8, C = 30 mM).

Figure 3

Scanning electron microscope (SEM) images of Bacillus subtilis spores under optimal conditions in different processes. (a) Unexposed, (b) UV alone (t = 60 s, pH = 5), (c) persulfate (t = 5 s, pH = 11, C = 3 mM), (d) UV/persulfate (t = 60 s, pH = 7.8, C = 30 mM).

Ultrastructural analysis of SEM images shows that B. subtilis spores were resistant to persulfate treatment but that UV irradiation was associated with damaged spore coat layers and that combined UV/persulfate could disrupt the spore coat and core with the highest efficiency. In fact, sulfate radicals were able to enter spores, and cellular compositions diffused into the outer perimeter of spores. This SEM image demonstrates a significant reduction in size and number and also amorphous morphologies of the spores after the combined UV/persulfate process.

In addition, Riesenman and Nicholson (Riesenman & Nicholson 2000) reported that the spore coat, particularly the inner coat layer, could not resist 254 nm UVC radiation. The destructive mechanisms underlying the disrupted spore coat and core were not the goal of this study and require further analysis.

CONCLUSIONS

Bacillus subtilis spores were used as target microorganisms that are resistant to conventional disinfectants. The effects of three disinfectants including UV, persulfate, and combined UV/persulfate were investigated for their efficacy in the inactivation of B. subtilis spores. In the case of each individual process, UV irradiation reduced spores up to 3.19 log in certain conditions (i.e., pH = 8 and 60 min contact time), while persulfate did not achieve a significant level of spore inactivation at higher applied dosages. Among the tested disinfectants, the combined UV/persulfate process demonstrated the greatest effect on spore inactivation with a 4.1 log reduction. In this process, SO4•− radicals generated from the persulfate activation by UV irradiation showed an enhancement of inactivated B. subtilis spores with close to a 1 log reduction. Although the results achieved in the UV/persulfate process are promising for spore inactivation, additional studies are required to evaluate other aspects such as the various UV intensity, process cost, and bench scale use of this process.

Also it should be noted that the current study was performed to evaluate the efficacy of the process; so we had to test higher spore densities to see the effects of independent variables on the log-reduction of spores. Then, the applied amount of persulfate could tend to high values of sulfate at the end of disinfection periods. Obviously, in real conditions we are not faced with such high levels of spores; so lower amounts of persulfates would be sufficient. Therefore, further investigation is required to include initial spore concentrations and lower persulfate levels as independent variables.

ACKNOWLEDGEMENTS

This study was supported by Tehran University of Medical Sciences under research fund grant No. 94-01-27-28523 to Dr Ramin Nabizadeh.

REFERENCES

REFERENCES
Bezerra
M. A.
Santelli
R. E.
Oliveira
E. P.
Villar
L. S.
Escaleira
L. A.
2008
Response surface methodology (RSM) as a tool for optimization in analytical chemistry
.
Talanta
76
(
5
),
965
977
.
Blatchley
E. R.
III
Meeusen
A.
Aronson
A. I.
Brewster
L.
2005
Inactivation of Bacillus spores by ultraviolet or gamma radiation
.
Journal of Environmental Engineering
131
(
9
),
1245
1252
.
Carley
K. M.
Kamneva
N. Y.
Reminga
J.
2004
Response surface methodology, DTIC Document
.
Driedger
A.
Staub
E.
Pinkernell
U.
Mariñas
B.
Köster
W.
von Gunten
U.
2001
Inactivation of Bacillus subtilis spores and formation of bromate during ozonation
.
Water Research
35
(
12
),
2950
2960
.
Gao
Y.-Q.
Gao
N.-Y.
Deng
Y.
Yang
Y.-Q.
Ma
Y.
2012
Ultraviolet (UV) light-activated persulfate oxidation of sulfamethazine in water
.
Chemical Engineering Journal
195
,
248
253
.
Jung
H.
Kim
D.
Gweon
B.
Moon
S.
Choe
W.
2010
Enhanced inactivation of bacterial spores by atmospheric pressure plasma with catalyst TiO2
.
Applied Catalysis B: Environmental
93
(
3
),
212
216
.
Khuri
A. I.
Mukhopadhyay
S.
2010
Response surface methodology
.
Wiley Interdisciplinary Reviews: Computational Statistics
2
(
2
),
128
149
.
Lenth
R. V.
2009
Response-surface methods in R, using rsm
.
Journal of Statistical Software
32
(
7
),
1
17
.
Mezule
L.
Reimanis
M.
Krumplevska
V.
Ozolins
J.
Juhna
T.
2014
Comparing electrochemical disinfection with chlorination for inactivation of bacterial spores in drinking water
.
Water Science & Technology: Water Supply
14
(
1
),
158
164
.
Post
G. B.
Atherholt
T. B.
Cohn
P. D.
2011
Health and aesthetic aspects of drinking water
. In:
Water Quality and Treatment: A Handbook on Drinking Water
, 6th edn (
Edzwald
J. K.
, ed.).
American Water Works Association
. McGraw-Hill, New York.
Radziminski
C.
Ballantyne
L.
Hodson
J.
Creason
R.
Andrews
R. C.
Chauret
C.
2002
Disinfection of Bacillus subtilis spores with chlorine dioxide: a bench-scale and pilot-scale study
.
Water Research
36
(
6
),
1629
1639
.
Roshani
B.
Leitner
N. K. V.
2011
Effect of persulfate on the oxidation of benzotriazole and humic acid by e-beam irradiation
.
Journal of Hazardous Materials
190
(
1
),
403
408
.
Tan
C.
Gao
N.
Deng
Y.
Zhang
Y.
Sui
M.
Deng
J.
Zhou
S.
2013
Degradation of antipyrine by UV, UV/H2O2 and UV/PS
.
Journal of Hazardous Materials
260
,
1008
1016
.
Wordofa
D. N.
2014
Application of Iron Activated Persulfate for Disinfection in Water Treatment
.
MSc Thesis
,
University of California
.
Xie
P.
Ma
J.
Liu
W.
Zou
J.
Yue
S.
Li
X.
Wiesner
M. R.
Fang
J.
2015
Removal of 2-MIB and geosmin using UV/persulfate: contributions of hydroxyl and sulfate radicals
.
Water Research
69
,
223
233
.