We introduce vacuum ultraviolet (VUV) photolysis at 172 nm as a more efficient process for bromate and chlorate removal than conventional ultraviolet (UV) photolysis at 254 nm. We discuss the effects of pH and coexisting salts on VUV photolysis. In experiments at various pH levels, the VUV photolysis performance decreased in the alkaline region for pH above 7. Furthermore, nitrate and inorganic carbon compromised the performance, probably owing to the generation of active oxidant species (such as hydroxyl radicals and carbonate radicals) during the photolysis. However, photochemical experiments conducted in pure water revealed that VUV photolysis is 4.1-fold more energy efficient than UV photolysis using a low-pressure mercury lamp. Although the performance of VUV photolysis in tap water was about 40% less efficient than that in pure water, mainly owing to the presence of bicarbonate ions, the performance was still higher than that of UV photolysis in pure water. However, VUV photolysis may not be feasible in tap water with a high concentration of nitrate ions, because nitrate produces a stronger inhibition effect than other salts.

INTRODUCTION

Bromate () and chlorate () ions are regulated disinfection byproducts in water supplies. Bromate is a possible carcinogen to humans (Group 2B by International Agency for Research on Cancer 1999), and can lead to the oxidative damage of red blood cells and can interfere with the thyroid function (World Health Organization 2011). These halogenated oxyanions are produced by the reactions between disinfectants and precursor compounds. Hypochlorite in a feedstock solution can oxidize the bromide ion (Br) as a contaminant into , and is known as a decomposition product of hypochlorite (Weinberg et al. 2003). Chlorine dioxide in tap water also generates as a decomposition product (Korn et al. 2002). Accordingly, disinfectants usually contain a certain amount of and as contaminants (Bolyard & Fair 1992; Weinberg et al. 2003). Furthermore, introduction of advanced water purification technologies such as ozonation and advanced oxidation into water purification processes also accelerates and formation via oxidation of Br and chloride (Cl) ions (von Gunten 2003). Thus, and contamination of potable water is a serious threat to human health.

Various technologies for and removal, such as chemical reduction (Westerhoff 2003), electrochemical reduction (Kishimoto et al. 2012), biological remediation (Hijnen et al. 1999; Nerenberg & Rittmann 2004), activated carbon techniques (Gonce & Voudrias 1994; Asami et al. 1999; Huang & Cheng 2008), and ultraviolet (UV) photolysis (Phillip et al. 2006) have been discussed. Rapid removal of these oxyanions was demonstrated by using chemical reduction and electrochemical reduction methods. However, chemical reduction requires additional chemical inputs, and electrochemical reduction has technical difficulties related to scaling up the method. Biological remediation also requires the addition of chemicals such as hydrogen gas (Nerenberg & Rittmann 2004) and/or organic compounds (Hijnen et al. 1999) to serve as electron donors. In the activated carbon techniques, catalytic reduction by activated carbon is responsible for removal (Huang & Cheng 2008), but no catalytic reduction of has been observed (Gonce & Voudrias 1994). Furthermore, Asami et al. (1999) reported that the transition of granular activated carbon to biological activated carbon apparently reduced the removal rate. Phillip et al. (2006) reported that UV photolysis successfully reduced to Br, which was enhanced by vacuum ultraviolet (VUV) light at 185 nm, emitted from a low-pressure mercury lamp. Because UV photolysis does not require the addition of any chemicals and has already been used for the disinfection of potable water, it may be a promising technology for and removal.

Here, we focus on the VUV photolysis technique for and removal. Although the VUV light is less effective for microorganism disinfection than the UV light at 254 nm (Wang et al. 2010), the photolysis of and is expected to be enhanced by the VUV light's high photon energy, as has been reported by Phillip et al. (2006). We also discuss the effects of pH and coexisting substances on VUV photolysis. Furthermore, conventional UV photolysis using a low-pressure mercury lamp has also been conducted for comparison, because low-pressure mercury lamps are conventionally used for the disinfection of potable water.

MATERIAL AND METHODS

Photochemical reactor

The photochemical reactor was composed of a glass beaker with a volume of 3 L and an UV lamp with a length of 120 mm, which was installed at the center of the beaker. Two UV lamp types were used. One was an excited xenon dimer lamp (excimer lamp; 172/120Z, Heraeus, Germany), which consumed 20 W of electric power and mainly emitted VUV light at the wavelength of 172 nm. The total irradiance at 172 nm, measured by using a VUV sensor (VUV-S172, Ushio, Japan), was 5.7 W. The second type was a low-pressure mercury UV lamp (UVL20PH-6, SEN Lights, Japan), which consumed 20 W of electric power and mainly emitted 254 nm of UV light. The VUV light at 185 nm was cut off by a water jacket around the UV lamp. The total irradiance at 254 nm measured by using a UV sensor (UVD-S254, Ushio, Japan) was 4.3 W. Two liters of or contaminated water was treated in this reactor for 60 minutes. The contaminated water was continuously mixed by using a magnetic stirrer (CT-3, As One, Japan) during the photolysis.

Contaminated water

Potassium salts and deionized water were used for preparing or contaminated water. The initial concentration was 1.00 mg/L. The solution pH was adjusted by adding sulfuric acid or sodium hydroxide. The effects of sulfate, phosphate, ammonium, bicarbonate, and nitrate were determined by adding the corresponding sodium or sulfate salts at the final concentration of 0.500 mM. VUV photolysis was also applied to tap water by adding potassium bromate at the final concentration of .

Chemical analysis

The concentrations of and other anions were determined by using ion chromatography coupled with a post-column system (DX-500 with BRS-500, Dionex, USA), but the post-column system was not used when the concentration was above 100 μg/L. Analytical conditions were as follows. Column: Dionex IonPac AS12A with a suppressor (Dionex ASRS-ULTRA 4 mm); mobile phase: aqueous solution with 2.7 mM Na2CO3 and 0.3 mM NaHCO3; flow rate: 1.0 mL/min; sample injection volume: 100 μL; oven temperature: 40 °C. The reaction conditions in the post-column were as follows. Reactant A: 1.5 M KBr and 1.0 M H2SO4; reactant B: 1.2 mM NaNO2; flow rate: 0.4 mL/min for reactant A and 0.2 mL/min for reactant B; reaction temperature 40 °C; detection: absorbance at 268 nm. The concentrations of and other anions were also determined by using ion chromatography under the same conditions as those for determining . A pH meter (B-212, Horiba, Japan) was used for measuring pH. Bicarbonate concentration was determined in pursuance of 4500-CO2 D in Standard Methods (2012).

RESULTS AND DISCUSSION

Effect of pH on VUV photolysis

Figure 1 shows the temporal changes in and Br concentrations during VUV photolysis at pH 7. The concentration decreased with time, and the corresponding increase in Br was observed. Thus, was successfully reduced into Br by VUV photolysis. Because exhibits a strong absorption spectrum in the VUV range (Peldszus et al. 2004), direct photolysis was concluded to be the dominant mechanism for VUV photolysis, as has been pointed out by Jung et al. (2014). Figure 2 shows the and degradation curves during VUV photolysis for various pH levels on a semilogarithmic axis. The data clearly demonstrate that VUV photolysis of both oxyanions followed the pseudo-first-order reaction kinetics, because the following linear relation was established under the pseudo-first-order reaction kinetics: 
formula
1
where C is the contaminant concentration, C0 is the initial concentration of the contaminant, t is the elapsed time [min], and k is the rate constant of degradation [min−1]. These kinetics were similar to the decomposition kinetics under UV light from low-pressure mercury lamps (Phillip et al. 2006). Figure 3 summarizes the pseudo-first-order degradation rate constants of and at various pH levels, estimated from the slopes of the corresponding degradation curves plotted using a semilogarithmic axis. The degradation rate constants decreased for pH values above 7. This tendency was observed for both and , although the degradation rate of was higher than that of . The acid dissociation constant of bromic acid (HBrO3) was reported to be 73.9 ± 3.0 mol/L (Kamble & Nandibewoor 1996). Accordingly, the main chemical species for all pH values in this research was . Therefore, the reduced degradation rate constant of at alkaline pH was not owing to the different reactivity associated with different chemical species. Although the acid dissociation constant of chloric acid (HClO3) has not been reported, the above discussion will apply to HClO3, owing to the similar chemical characteristics of HBrO3 and HClO3. Hayon & McGarvey (1967) reported that hydroxyl radicals (•OH) were produced by VUV photolysis of hydroxide ions (OH). The •OH can oxidize Br and Cl and their oxyanions, and finally produce and (von Gunten & Oliveras 1998; von Gunten 2003). Therefore, the regeneration of and by •OH from the photolysis of OH was inferred to be responsible for inhibiting the and removal under alkaline pH.
Figure 1

Temporal changes in bromate and bromide concentrations during VUV photolysis at pH 7.

Figure 1

Temporal changes in bromate and bromide concentrations during VUV photolysis at pH 7.

Figure 2

Temporal changes in bromate and chlorate concentrations during VUV photolysis at various pH levels.

Figure 2

Temporal changes in bromate and chlorate concentrations during VUV photolysis at various pH levels.

Figure 3

Pseudo-first-order degradation constants of VUV photolysis at various pH levels. The rate constants are based on the common logarithm. Error bars denote the 95% confidence intervals.

Figure 3

Pseudo-first-order degradation constants of VUV photolysis at various pH levels. The rate constants are based on the common logarithm. Error bars denote the 95% confidence intervals.

Effect of coexisting salts on VUV photolysis

Figure 4 shows the degradation rate constants obtained when various salts coexisted with the and ions at pH 7. Figure 4 clearly shows that sodium ions, sulfates, and phosphates did not affect the degradation rates of and , because the degradation rate constant in the presence of sodium sulfate (Na2SO4) or sodium dihydrogen phosphate (NaH2PO4) did not differ from that obtained in the control experiment. Consequently, the inhibition of photolysis in alkaline pH was not caused by sodium ions that were used for pH control.

Figure 4

Pseudo-first-order degradation rate constants in VUV photolysis obtained in the presence of various salts at pH 7. The rate constants are based on the common logarithm. No salts were added in control experiments. Error bars denote the 95% confidence intervals.

Figure 4

Pseudo-first-order degradation rate constants in VUV photolysis obtained in the presence of various salts at pH 7. The rate constants are based on the common logarithm. No salts were added in control experiments. Error bars denote the 95% confidence intervals.

Contrary to Na2SO4 and NaH2PO4, the presence of sodium nitrate (NaNO3) and sodium bicarbonate (NaHCO3) significantly inhibited the photolysis of and . It is known that photolysis of nitrates yields •OH (Mark et al. 1996; Mack & Bolton 1999), which can oxidize Br and Cl and their oxyanions into and (von Gunten & Oliveras 1998; von Gunten 2003). Therefore, the regeneration of and by •OH from the photolysis of nitrate ions was inferred to contribute to inhibiting the and removal in the presence of nitrate. Photolysis of carbonate and bicarbonate ions yields carbonate radicals (Hayon & McGarvey 1967; Behar et al. 1970), which can enhance formation by accelerating the oxidation of hypobromite ions into hypobromite radicals (von Gunten 2003). Therefore, carbonate radicals produced by the VUV photolysis of bicarbonate ions were inferred to cause the decrease in the degradation rate. It was also inferred that the carbonate radicals inhibited removal in a similar manner.

Comparison of VUV photolysis with UV photolysis

Figure 5 shows changes in concentrations during VUV and UV photolyses. Both lamps successfully degraded . However, the degradation rate constant in VUV photolysis was 4.1 times higher than that in UV photolysis, although the electrical input to both lamps was the same. Consequently, VUV photolysis was potentially more energy efficient than the conventional UV photolysis, judged on the basis of both the electrical input and total irradiance.

Figure 5

Changes in bromate concentrations during VUV and UV photolyses at pH 7. The numerical values correspond to the pseudo-first-order degradation rate constants, calculated based on the common logarithm.

Figure 5

Changes in bromate concentrations during VUV and UV photolyses at pH 7. The numerical values correspond to the pseudo-first-order degradation rate constants, calculated based on the common logarithm.

VUV photolysis in tap water

VUV photolysis of in tap water is compared with that in pure water in Figure 6. The degradation rate constant of VUV photolysis in tap water was about 40% smaller than that in pure water. Tap water contained 0.52 mM of bicarbonate and only 0.011 mM of nitrate. When 0.500 mM of bicarbonate or nitrate was added, the degradation rate constant of decreased from 0.0241 to 0.0148 min−1 or 0.0064 min−1, respectively, as shown in Figure 4. Assuming that the reduction of the degradation rate is proportional to the concentrations of bicarbonate and nitrate, the bicarbonate and nitrate ions in tap water were expected to slow the degradation rates by 0.0097 min−1 and 0.00039 min−1, respectively. Consequently, the degradation rate in tap water was estimated to be 0.140 min−1, which was nearly equal to the observed rate (0.141 min−1). Thus, it was inferred that the reduction of the degradation rate in our experiments was mainly caused by the presence of bicarbonate ions in tap water, whereas nitrate ions made a relatively weak contribution. However, if tap water contains higher concentrations of nitrate ions, and control by using VUV photolysis might not be feasible, because nitrate ions produce stronger inhibition effects than other salts.

Figure 6

Changes in bromate concentrations during VUV photolysis in pure water and tap water at pH 7. The numerical values correspond to the pseudo-first-order degradation rate constants, calculated based on the common logarithm. The original tap water contained the following: F: 0.06 mg/L, Cl: 13.4 mg/L, Br: 0.3 mg/L, : 0.66 mg/L, : 11.9 mg/L, orthophosphate: 0.01 mg-P/L, : not detected, residual chlorine: 0.4 mg/L, and total alkalinity: 26 mg-CaCO3/L.

Figure 6

Changes in bromate concentrations during VUV photolysis in pure water and tap water at pH 7. The numerical values correspond to the pseudo-first-order degradation rate constants, calculated based on the common logarithm. The original tap water contained the following: F: 0.06 mg/L, Cl: 13.4 mg/L, Br: 0.3 mg/L, : 0.66 mg/L, : 11.9 mg/L, orthophosphate: 0.01 mg-P/L, : not detected, residual chlorine: 0.4 mg/L, and total alkalinity: 26 mg-CaCO3/L.

CONCLUSIONS

The efficacy of VUV photolysis at 172 nm for and removal was determined. VUV photolysis successfully degraded and into Br and Cl, respectively. and removal followed the pseudo-first-order reaction kinetics on and concentrations, respectively. The degradation rate constant was nearly the same across the pH values in neutral and acidic ranges, although the rate was reduced at alkaline pH. The presence of nitrate and bicarbonate ions also yielded lower and degradation rates. The negative effects of pH and nitrate and bicarbonate ions on the degradation rates were attributed to the regeneration of and by active oxidant species produced via the photolysis of coexisting ions, namely •OH that was produced by the photolysis of nitrate or OH, and carbonate radicals that were produced by the photolysis of bicarbonate ions.

The comparison of VUV photolysis with the conventional UV photolysis at 254 nm revealed that the energy efficiency of VUV photolysis on the basis of electrical input to the lamp was 4.1 times higher than that of UV photolysis in pure water. Although the degradation rate constant in tap water obtained by using VUV photolysis was about 40% lower than that in pure water, mainly owing to the presence of bicarbonate ions and insignificantly owing to the presence of nitrate ions, the degradation rate constant observed in tap water was still higher than that obtained for UV photolysis in pure water. However, if tap water contains a high concentration of nitrate, the and control by using VUV photolysis might not be feasible, because nitrate ions produce stronger inhibition effects than other salts.

ACKNOWLEDGEMENT

This work was performed under JSPS KAKENHI No. 26340064, funded by the Japan Society for the Promotion of Science.

REFERENCES

REFERENCES
Asami
M.
Aizawa
T.
Morioka
T.
Nishijima
W.
Tabata
A.
Magara
Y.
1999
Bromate removal during transition from new granular activated carbon (GAC) to biological activated carbon (BAC)
.
Water Research
33
,
2797
2804
.
Behar
D.
Czapski
G.
Duchovny
I.
1970
Carbonate radical in flash photolysis and pulse radiolysis of aqueous carbonate solutions
.
The Journal of Physical Chemistry
74
,
2206
2210
.
Bolyard
M.
Fair
P. S.
1992
Occurrence of chlorate in hypochlorite solutions used for drinking water disinfection
.
Environmental Science and Technology
26
,
1663
1665
.
Hijnen
W. A. M.
Jong
R.
van der Kooij
D.
1999
Bromate removal in a denitrifying bioreactor used in water treatment
.
Water Research
33
,
1049
1053
.
Huang
W.-J.
Cheng
Y.-L.
2008
Effects of characteristics of activated carbon on removal of bromate
.
Separation and Purification Technology
59
,
101
107
.
International Agency for Research on Cancer
1999
IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 73
,
International Agency for Research on Cancer
,
Lyon
,
France
.
Kamble
D. L.
Nandibewoor
S. T.
1996
Kinetics of vanadium(V) catalyzed oxidation of gallic acid by potassium bromate in acid medium
.
International Journal of Chemical Kinetics
28
,
673
679
.
Kishimoto
N.
Kishimoto
A.
Nakayama
A.
2012
Rapid removal of bromate ion from water streams with an electrolytic flow cell
.
Journal of Water Supply: Research and Technology – AQUA
61
,
103
110
.
Mack
J.
Bolton
J. R.
1999
Photochemistry of nitrite and nitrate in aqueous solution: a review
.
Journal of Photochemistry and Photobiology A: Chemistry
128
,
1
13
.
Mark
G.
Korth
H.-G.
Schunchmann
H.-P.
von Sonntag
C.
1996
The photochemistry of aqueous nitrate ion revisited
.
Journal of Photochemistry and Photobiology A: Chemistry
101
,
89
103
.
Nerenberg
R.
Rittmann
B. E.
2004
Hydrogen-based, hollow-fiber membrane biofilm reactor for reduction of perchlorate and other oxidized contaminants
.
Water Science and Technology
49
(
11–12
),
223
230
.
Peldszus
S.
Andrews
S. A.
Souza
R.
Smith
F.
Douglas
I.
Bolton
J.
Huck
P. M.
2004
Effect of medium-pressure UV irradiation on bromate concentrations in drinking water, a pilot-scale study
.
Water Research
38
,
211
217
.
Standard Methods for the Examination of Water and Wastewater
2012
22nd edn
.
American Public Health Association/American Water Works Association/Water Environment Federation
,
Washington, DC
,
USA
.
von Gunten
U.
Oliveras
Y.
1998
Advanced oxidation of bromide-containing waters: bromate formation mechanisms
.
Environmental Science and Technology
32
,
63
70
.
Weinberg
H. S.
Delcomyn
C. A.
Unnam
V.
2003
Bromate in chlorinated drinking waters: occurrence and implications for future regulation
.
Environmental Science and Technology
37
,
3104
3110
.
Westerhoff
P.
2003
Reduction of nitrate, bromate, and chlorate by zero valent iron (Fe0)
.
Journal of Environmental Engineering
129
,
10
16
.
World Health Organization
2011
Guidelines for Drinking Water Quality
,
4th edn
.
World Health Organization
,
Geneva
,
Switzerland
.