w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/watres
Formation of bromate in sulfate radical based oxidation: Mechanistic aspects and suppression by dissolved organic matter Holger V. Lutze a,c,*, Rani Bakkour a, Nils Kerlin a, Clemens von Sonntag a,b, Torsten C. Schmidt a,c,d a
University Duisburg-Essen, Instrumental Analytical Chemistry, Universita¨tsstr. 5, D-45141 Essen, Germany Max-Planck-Institut fu¨r Strahlenchemie, Stiftstr. 34-36, Mu¨lheim an der Ruhr, Germany c IWW Water Centre, Moritzstr. 26, D-45476 Mu¨lheim an der Ruhr, Germany d University Duisburg-Essen, Centre for Water and Environmental Research (ZWU), Universita¨tsstr. 2, D-45141 Essen, Germany b
article info
abstract
Article history:
Sulfate radical based oxidation is discussed being a potential alternative to hydroxyl radical
Received 25 November 2013
based oxidation for pollutant control in water treatment. However, formation of undesired
Received in revised form
by-products, has hardly been addressed in the current literature, which is an issue in other
29 December 2013
oxidative processes such as bromate formation in ozonation of bromide containing water
Accepted 1 January 2014
(US-EPA and EU drinking water standard of bromate: 10 mg L1). Sulfate radicals react fast
Available online 9 January 2014
with bromide (k ¼ 3.5 109 M1 s1) which could also yield bromate as final product. The mechanism of bromate formation in aqueous solution in presence of sulfate radicals has
Keywords:
been investigated in the present paper. Further experiments were performed in presence of
Sulfate radicals
humic acids and in surface water for investigating the relevance of bromate formation in
Ozone
context of pollutant control. The formation of bromate by sulfate radicals resembles the
Bromate
well described mechanism of the hydroxyl radical based bromate formation. In both cases
Water treatment
hypobromous acid is a requisite intermediate. In presence of organic matter formation of
Pollutant degradation
bromate is effectively suppressed. That can be explained by formation of superoxide formed in the reaction of sulfate radicals plus aromatic moieties of organic matter, since superoxide reduces hypobromous acid yielding bromine atoms and bromide. Hence formation of bromate can be neglected in sulfate radical based oxidation at typical conditions of water treatment. ª 2014 Published by Elsevier Ltd.
1.
Introduction
Oxidative water treatment based on highly reactive hydroxyl radicals (OH) is referred to as advanced oxidation processes (AOP) and can be used for degrading recalcitrant pollutants
such as pesticides, X-ray contrast media and fuel additives (e.g., MTBE) (von Gunten, 2003a). OH can be generated in various ways e.g., by photolysis of hydrogen peroxide (UV/ H2O2) (Legrini et al., 1993) or in ozonation (von Gunten, 2003a). Beside OH, sulfate radicals ðSO4 Þ are frequently investigated
* Corresponding author. University Duisburg-Essen, Instrumental Analytical Chemistry, Universita¨tsstr. 5, D-45141 Essen, Germany. Tel.: þ49 201 183 6791. E-mail address: [email protected] (H.V. Lutze). 0043-1354/$ e see front matter ª 2014 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.watres.2014.01.001
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
as potential alternative oxidants for water treatment (Anipsitakis and Dionysiou, 2003, 2004; Anipsitakis et al., 2006; Hori et al., 2004, 2008, 2005; Kutsuna and Hori, 2007; Liang et al., 2008, 2007; Manoj et al., 2007) and have already been applied in ground water remediation (Siegrist et al., 2011). SO4 can be 2 formed in various ways using S2 O8 as a radical precursor. One possibility is its photolysis by UVC-radiation (UV/S2 O2 8 ) which is in analogy to UV/H2O2. A major drawback in oxidative water treatment is the formation of undesired by-products. Ozone based processes, for instance, can be limited by the formation of bromate ðBrO 3 Þ a potential carcinogen (US-EPA and EU drinking water standard: 10 mg L1) arising from the oxidation of Br (von Gunten, 2003b). Thereby, formation of BrO 3 can be driven by O3 and/or OH. Corresponding mechanisms have already been discussed in detail (Haag and Hoigne´, 1983; von Gunten, 2003b; von Gunten and Hoigne, 1994; von Gunten and Oliveras, 1998). For reactions in which solely O3 or OH are involved, hypobromous acid (HOBr) is a requisite intermediate. At typical conditions of water treatment HOBr can effectively be reduced by hydrogen peroxide (H2O2) yielding Br (von Gunten and Oliveras, 1997). This instance is used for mitigating BrO 3 formation in ozone applications by addition of H2O2 (von Gunten, 2003b). In most OH based processes H2O2 is used as a radical source (e.g., UV/H2O2), thus preventing BrO 3 formation. UV/TiO2 is known to not require H2O2 to form OH. However, studies indicated that BrO 3 is not formed in this process (Tercero Espinoza and Frimmel, 2008). Also gamma radiolysis which is discussed being a potential water treatment option for both, disinfection (de Souza et al., 2011) and pollutant degradation (Dessouki et al., 1999; Getoff, 2002; Tahri et al., 2010) might oxidize Br yielding BrO 3 . In analogy to OH the reaction of SO4 plus Br (k ¼ 3.5 109 M1 s1 (Redpath and Willson, 1975)) yield BrO 3 (Fang and Shang, 2012). In the recent work of Fang and Shang (2012) an empirical model has been established, which was used for describing the formation of HOBr/OBr and BrO 3 . This approach has been extended in our work by a mechanistic discussion. The present study provides a reaction mechanism based on data available in the literature for developing a kinetic model of SO4 driven formation of BrO3 . This model has been used to describe the behavior of which was experiHOBr and BrO 3 in the oxidation of Br mentally determined at various conditions. Furthermore, the potential of BrO 3 formation in natural matrices has been investigated and contrasted to the oxidation strength available for pollutant control.
2.
Methods
All chemicals were commercially available and used as received. Acetonitrile (99.9%) Sigma Aldrich, atrazine (97.4%) Riedelde Hae¨n, 4-chlorobenzoic acid (pCBA) (99%) Aldrich, hydrochloric acid (37% in water, p.a.) Merck, hydrogen peroxide (30%) Sigma Aldrich, methanol (p.a.) SigmaeAldrich, 4-nitrobenzoic acid (pNBA) (97.4%) Sigma Aldrich, oxygen (99.9%) Liquid Air, phosphoric acid (85%) Merck, potassium bromate (99.5%) Fluka, potassium chloride (99.5%) Riedel-de Hae¨n, pure water has been prepared by treating deionized water with a pure lap ultra instrument (Elga) (electrical resistance 18.6 MU), sodium
371
bicarbonate (99.5%) KMF optichem, sodium bromate (99%) Fluka, sodium carbonate (99.8%) Riedel-de Hae¨n, sodium hydroxide (99.9%, p.a.) VWR, sodium peroxodisulfate (p.a.) SigmaeAldrich, sulfuric acid (95e97%) Applichem International, Suwannee River NOM (reverse osmosis concentrate) International Humic Acid Society, Uridine (99%) Sigma. Sulfate radicals were generated by photolysis of peroxodisulfate (UV/S2 O2 8 ) in a merry-go-round apparatus equipped with a low pressure mercury lamp. This radiation source emits monochromatic light at 254 nm (Heraeus Noble Light GPH303T5L/4, 15 W (185 nm band suppressed)). The fluence rate has been determined by uridine actinometry according to von Sonntag and Schuchmann (1992). Solutions were buffered with phosphate. Even though SO4 reacts with HPO2 4 with a considerable rate (k ¼ 1.2 106 M1 s1) (H2 PO 4 is nearly inert (k < 7 104 M1 s1)) the reactions under study are faster by several orders of magnitudes (k(SO4 plus 9 1 1 Br ) ¼ 3.5 10 M s ). This allows addition of phosphate buffer in excess over bromide (e.g., factor 100), which is necessary for keeping pH constant (experimental details can be found in the caption of corresponding figures). pH-adjustments have been done by addition of sulfuric acid or sodium hydroxide, respectively. Methanol was added to the samples (1 M in the sample) for scavenging low levels of SO4 which may be formed during storage time by thermolysis of S2 O2 8 . BrO3 and Br were analyzed by ion chromatography (Metrohm 883 basic) equipped with a conductivity detector coupled with ion suppression (anion separation column with quaternary ammonium groups: Metrosep A Supp 4 e 250/4.0 mm, particle size 2 9 mm; eluent HCO 3 (1.7 mM), CO3 (1.8 mM) mixed with acetonitrile (30% (v/v)); flow: 1 mL min1; retention times: Br: 3.6 min, BrO 3 : 4.4 min). For determining BrO3 and Br at concentrations in the mM range a different IC system has been used coupled with ion- and subsequent CO2-suppression (Metrohm, 881 Compact IC plus), equipped with a high capacity anion separation column (Metrosep A Supp 5e250/4.0, particle size 5 mm) which was necessary for separation of Cl and BrO 3 (Eluent: 3.2 mM Na2CO3 and 1.0 mM NaHCO3, flow: 0.7 mL min1; retention times: BrO 3 : 8.0 min, Cl : 8.7 min, in River Ruhr Br: 13.0 min). For determining Br and BrO 3 water an IC-ICP-MS system has been used (200 Series from Perkin Elmer, Eluent: 10 mM NaOH, flow: 1.8 mL min1). Atrazine, 4-chlorobenzoic acid (pCBA) and 4-nitrobenzoic acid (pNBA) have been determined by HPLC with UV-detection (Shimadzu) (C18 reversed phase separation column: Bischoff, NUCLEOSIL 100, 250/4.0 mm, particle size: 5.0 mm). As eluent a gradient of methanol/water has been used (gradient program (methanol content (v/v)): 0e3 min: gradient 20e50%, 3e25 min: gradient 50e75%; 25e30 min: gradient 75e20%, 30e38 min: isocratic 20%; flow 0.6 mL min1; retention times/measured wave length: pNBA: 20.6 min/262 nm; atrazine: 25.8 min/ 234 nm, pCBA: 26.5 min/234 nm). HOBr has been determined by UV-absorption as OBr at 329 nm (ε(OBr) ¼ 332 M1 cm1 (Troy and Margerum, 1991)) by adjusting the solution to pH 11. Model calculations have been performed by using the software tool Kintecus (Ianni, 2008) (quantum yields for radical formation and molar absorption coefficients of peroxodisulfate can be obtained from Mark et al., 1990). Experiments in presence of humic acids where performed for simulating bromate formation in a real water during UV/
372
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
S2 O2 8 . Beside Br , probe compounds have been added for two purposes:
1. As model compound for simulating the degradation of pollutants, atrazine has been added (k(atrazine þ SO4 ) ¼ 3 109 M1 s1 (Manoj et al., 2007)) 2. As indication for OH which may be formed in the reaction of SO4 with Cl (McElroy, 1990). For this purpose pNBA was chosen because it is inert in presence of SO4 (k(SO4 þ pNBA) 106 M1 s1 (Neta et al., 1977)) and UV radiation at present experimental conditions. However, it reacts fast with OH (k(OH þ pNBA) ¼ 2.6 109 M1 s1 (Buxton et al., 1988)). Thus, it is a sensitive indicator for the presence of OH.
induce a cascade of reactions (Table 1) yielding HOBr and BrO 3 as main products. The HOBr yield displays a distinctive dependency on the pH-value, which has also been observed in the reaction of OH (von Gunten and Oliveras, 1998). The speciation of HOBr/OBr (pKa 8.8e9 (Haag and Hoigne´, 1983; von Gunten and Hoigne, 1994)), largely affects the rates of its consecutive degradation by Br which predominantly reacts with hypobromite (OBr)
These compounds have been added in low concentrations (0.5 mM) for assuring that their contribution in scavenging of SO4 /OH is below 10%. In this situation most radicals react with the main water constituents (e.g., DOC), which is typical in oxidative water treatment. As another matrix for simulating real water treatment conditions River Ruhr water (RWW, Mu¨lheim a.d. Ruhr) has been used. The water has been filtered with cellulose acetate/ cellulose nitrate filters with a pore size of 0.45 mm (Whatman). If not noted otherwise, the concentration of HCO 3 has been adjusted as follows. The original amount of HCO 3 present in the River water has been removed by acidification (pH 2) and purging with oxygen ( 5 min). Then the pH has been readjusted to pH 7.2 and defined amounts of HCO 3 were added to achieve the desired concentration. Ozonation experiments were performed as follows. An aqueous O3 stock solution was prepared by purging an O3 enriched gas through ice cooled pure water. At stationary conditions a steady state concentration of ozone of 1 mM O3 could be achieved. This solution was used to dose specific amounts of O3 to samples under study. Therefore glass syringes with stainless steel cannulas where used. For maintaining a constant O3 concentration in the O3 stock solution, continuous purging with the ozone enriched gas was necessary. O3 was generated by an O3 generator purchased from BMT Messtechnik Berlin (Philaqua 802x) with oxygen as feed gas.
3.
Results and discussion
3.1.
Oxidation of bromide
The reaction of SO4 with Br leads to the formation of BrO 3 in aqueous solution (Fig. 1). The product pattern is in good agreement with the study of Fang and Shang (2012) in that they observed a complete turnover of Br via HOBr/OBr to BrO 3 . Furthermore, Fang and Shang reported HOBr/OBr evolving to a peak concentration corresponding to 25% of the initial Br concentration (at pH 7), which matches well the value of the present study (20%, Fig. 1b). This supports the experimental results of both studies. The chemistry of bromine is well documented in the literature and sufficient reactions including their kinetics are reported for describing the oxidation of Br by SO4 as will be explained below. Bromine atoms formed as primary products (Reaction 1) may
Fig. 1 e Product formation in the reaction of SO4L with BrL. Reaction conditions: [BrL]: 1 mM, [S2 O2L 8 ]: 10 mM, [phosphate]: 100 mM, T: 25 C, a) Initial pH: 8.0e8.1 (gradual decrease to pH 7.6 within a reaction time of 60 min), inset: bromate formation in presence of 1 mM H2O2, dotted-dashed line (inset) model calculation of H2O2 consumption; b) Initial pH 7.0 (gradual decrease to pH 6.8 within a reaction time of 90 min); circles: BrL, triangles: HOBr/OBrL; stars: BrOL 3 , solid lines model calculations, dashed line model of HOBr including the reaction of Br plus HOBr (k: 5 3 107 ML1 sL1) (not shown in a), see below).
373
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
Table 1 e Reactions occurring in UV/S2 O2L of bromide-containing aqueous solutions. 8 No.
Reaction
kforward ; kbackward ; or pKa
Ref.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Br- þ SO4 /Br þ SO2 4 Br þ Br- %Br2 Br2 þ Br2 /Br- þ Br3 Br3 #Br- þ Br2 Br2 þ H2 O#HOBr þ Br- þ Hþ HOBr#BrO- þ Hþ BrO þ Br2 /BrO þ 2Br BrO- þ Br /BrO þ Br þ 2BrO þ H2 O/BrO- þ BrO 2 þ 2H - þ BrO BrO þ BrO /BrO 2 2 BrO þ Br /Br- þ BrO þ BrO-
3.5 109 M1 s1 1 1010 M1 s1, 1 105 s1 1.8 109 M1 s1 1 107 s1, 1.5 109 M1 s1 97 s1, 1.6 1010 M2 s1 pKa: 8.8 8 0.7 107 M1 s1 4.1 109 M1 s1 4.9 1 109 M1 s1 3.4 0.7 108 M1 s1 8.0 0.8 107 M1 s1 1.4 109 M1 s1, 7.4 107 s1 4.2 107 M1 s1
(Redpath and Willson, 1975) (Zehavi and Rabani, 1972) (Neta et al., 1988) (Beckwith et al., 1996) (Beckwith et al., 1996) (Haag and Hoigne´, 1983) (Buxton and Dainton, 1968) (Kla¨ning and Wolff, 1985) (Buxton and Dainton, 1968) (Buxton and Dainton, 1968) (Buxton and Dainton, 1968) (Buxton and Dainton, 1968) (Buxton and Dainton, 1968)
2
2
2BrO2 #Br2 O4 þ 2BrO2 þ H2 O/BrO 2 þ BrO3 þ 2H
(von Gunten and Oliveras, 1998). That explains the higher yields in HOBr/OBr with decreasing pH. In analogy to mechanistic considerations in OH driven formation of BrO 3 the present mechanism has been modeled with equations shown in Table 1 (modeled data: lines in Fig. 1). For determining the reaction rates in reactions involving HOBr a pKa of 8.8 has been used, which was suggested for solutions with an ionic strength of > 0.15 M (Haag and Hoigne´, 1983) that suits our experimental conditions (ionic strength > 0.250 M). The good agreement of the model with the experimental results indicates a similar reaction mechanism compared to that proposed by von Gunten and Oliveras for OH-based BrO 3 formation (von Gunten and Oliveras, 1998). However, the model does not fit well to the experimental results of BrO 3 formation at pH 6 (data not shown). At pH 6 the formation of HOBr/OBr is largely overestimated, indicating that oxidative species may also react with HOBr, which has not been taken into account in the model calculations (Table 1). A good agreement with the HOBr concentration pattern can be achieved by taking a reaction of the bromine atom with HOBr into account assuming a reaction rate of 5 107 M1 s1. This rate constant is around 2 orders of magnitude smaller compared with the reaction rate of Br plus OBr. This seems to be a realistic value which has also been reported for other acid base couples in their reaction with Br such as HCN (k ¼ 1 107 M1 s1) and CN (k ¼ 8 108 M1 s1) (NIST). Furthermore, this reaction does not change the other model calculations for higher pH values significantly (dashed line in Fig. 1b), because the reaction Br plus OBr is largely dominating (note that at pH 8 this effect must be even weaker (calculations not shown in Fig. 1a)). A contribution of SO4 to oxidation of HOBr is not likely to happen. Included in the model, this reaction affects the outcome only if a very high rate constant is assumed for both species OBr and HOBr (k(SO4 þ HOBr/OBr)) ¼ 1010 M1 s1), since it has to compete with the reaction SO4 plus Br (Reaction 1). Such a rate constant would be much higher than that for the reaction with Br (3.5 109 M1 s1 (Redpath and Willson, 1975)), which is most unlikely. In the current model the formation of HOBr/BrO has been considered to be a requisite intermediate. This has been confirmed by experiments in presence of H2O2, which reduces HOBr/OBr to give rise to Br and thus, suppressing the
formation of BrO 3 . This effect is shown in Fig. 1a (inset). Upon addition of 1 mM H2O2 before irradiation, BrO 3 formation is delayed by ca. 20 min. This has been modeled by including the reduction of HOBr/OBr via H2O2 in the model (note that the reaction rate is pH-dependent (k ¼ 1.96 105 M1 s1 at pH 8) (von Gunten and Oliveras, 1997)). Taking this rate constant into account the delay in BrO 3 formation can be described by the model (Fig. 1a (inset)). Both, experimental data and the model outcome revealed that the presence of H2O2 effectively suppresses the Br turnover (note, that the photochemical degradation of H2O2 is largely overwhelmed by its reaction with OBr/HOBr). This is a strong evidence for HOBr being a requisite intermediate. To this end experiments were conducted in absence of typical constituents in natural waters such as DOC. The effect of a natural water matrix has also been investigated in the present study. In a first step the effect of DOC on SO4 driven BrO3 formation was investigated. Therefore, DOC and Br concentrations were varied for achieving different pro portions of SO4 reacting with bromide ([DOC] ¼ 1, 3, and 10 mg L1, [Br] ¼ 1 mM). The ratios of [Br]/[DOC] range from 0.1 to 0.5 mM L mg1, covering the range found in natural waters such as river Ruhr ([DOC] ¼ 1.9 mg L1, [Br] ¼ 0.54 mM (see caption of Fig. 2), Br/DOC ¼ 0.28 mM L mg1). Within a reaction time sufficient for degrading >90% atrazine neither Br degradation nor BrO 3 formation has been observed (½S2 O2 8 ¼ 0.5 mM, reaction time ¼ 2, 6 and 25 min for [DOC] ¼ 1, 3 and 10 mg L1). Even an increase of Br concentration in presence of 10 mg L1 DOC from 1 mM to 50 mM did not reveal any formation of BrO 3 within a reaction time of 50 min, even though the proportion of SO4 reacting with Br increased largely. Note that in presence of 50 mM Br atrazine degradation is mainly driven by photolysis (Fig. 2). In 1990 McElroy postulated the formation of OH upon oxidation of Cl to Cl by SO4 (McElroy, 1990). The humic acids used in the present study contain small amounts of Cl, which cannot be avoided without changing the chemical composition of the organic matter. Furthermore, Cl can be introduced by several means such as pH-measurements and contaminated reagents. IC measurements revealed Cl concentrations of 9.1e11.5 mM in experiments in presence of 9e10 mg L1 DOC and a Cl content of 1.2 0.3 mM per mg L1 DOC as average value for all experiments in presence of humic acids. For
374
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
Fig. 2 e Degradation of atrazine (squares) and pNBA L (triangles) in UV/S2 O2L 8 in presence of humic acids and Br ; L [Br ] [ 1 mM (open symbols), 50 mM (filled symbols), stars: direct photo degradation of atrazine, [DOC] [ 9.8e10.0 mg LL1, [phosphate] [ 2.5 mM, [ClL] [ 9.1e11.5 mM, pH: 7.1e7.2, [S2O2L 8 ] = 0.5 mM.
testing if OH affect the reaction system the degradation of pNBA was observed, that readily react with OH (k ¼ 2.6 109 M1 s1 (Buxton et al., 1988)), but barely reacts with SO4 (k < 106 M1 s1 (Neta et al., 1977)). A slow degradation of pNBA can be observed which indicates a rather small impact of OH (and/or Cl) on the present reaction system. However, their contribution on the transformation of probe compounds or Br is negligible as will be explained below. OH Since atrazine also reacts readily with (k ¼ 3 109 M1 s1 (Acero et al., 2000; Tauber and von Sonntag, 2000)) a similar degradation rate of pNBA and atrazine would indicate a strong contribution of OH in the present reaction system. The degradation rate of atrazine is the sum of SO4 driven degradation and direct photolysis. The first order
degradation rates in direct photolysis and UV/S2 O2 were 8 determined to be k(photolysis of atrazine) ¼ 0.0005 s1 and 1 k(atrazine in UV/S2 O2 8 ) ¼ 0.0019 s . Subtracting the rate of direct photo-degradation from degradation in UV/S2 O2 8 yields the degradation rate driven by SO4 (k(atrazine degradation by SO4 ) ¼ 0.0014 s1). pNBA was not significantly degraded by direct UV-radiation, thus the degradation rate is based on its 1 reaction with OH only (k (pNBA in UV/S2 O2 8 ) ¼ 0.00012 s ). Since the radical driven oxidation of atrazine is z12 times faster than pNBA degradation, the influence of OH on the reaction system under study seem to be negligible. Analogous experiments were performed in river Ruhr water, which is used as raw water for drinking water production. The Br content of this water was 40e50 mg L1 (0.51e0.63 mM), which may already lead to an exceedance of the drinking water standard for bromate in ozonation under certain treatment conditions (von Gunten, 2003b). River Ruhr water also contains Cl in the mM range (0.78 0.02 mM) that leads to a large proportion of SO4 reacting with Cl (>90%) yielding OH (McElroy, 1990) (see above). This does not necessarily prevent BrO 3 formation because BrO3 can also be formed in the reactions with OH as mentioned above (von Gunten and Oliveras, 1998). However, BrO 3 has not been in River Ruhr water. Variation of the formed in UV/S2 O2 8 HCO 3 concentration (0e2 mM) also did not lead to a detectable formation of BrO 3 . This is illustrated in Fig. 3a showing the concentration of Br in different photochemical processes. It can be observed that the Br concentrations in are near the Br level of the original River Ruhr UV/S2 O2 8 water (red line) and resembles that of UV/H2O2 and UV which (the are both known to not form BrO 3 in presence of Br differences in case of UV and UV/H2O2 at 0 mM HCO 3 are probably related to an analytical issues). This suggests, that no turnover of Br took place, while the oxidation power achieved in UV/S2 O2 8 suffices for >95% degradation of atrazine and up to 95% degradation of pNBA (Fig. 3b) (note that pNBA probably is degraded by OH formed in the reaction of
Fig. 3 e a) Concentration of BrL in River Ruhr water at different alkalinities in photochemical treatment; UV/S2 O2L (circles), 8 UV/H2O2 (triangles), UV (crosses), vertical lines indicate initial conditions, the following points represent subsequent samples after different reaction times given at the x-axis; red line: concentration of BrL in untreated River Ruhr water, horizontal black lines: upper and lower range of the BrL concentration in untreated River Ruhr water; b) Degradation of organic compounds in the experiments of a); gray area [ atrazine, black [ pCBA; a) and b) numbers indicate the 2L concentration of HCOL 3 in mM; shaded area [ original River Ruhr water (no adjustment of alkalinity); [S2 O8 ]0 [ [H2O2]0 [ L 0.5 mM, [phosphate] [ 1 mM; T 25 C, pH 7.2, [DOC] [ 1.9 mg/L, [HCO3 ] [ 0e2 mM.
375
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
Fig. 4 e a) Turnover of BrL (open circles) to BrOL 3 (closed circles) after different ozone doses, the black line indicates the sum ; b) Degradation of atrazine (triangles) and pNBA (crosses) vs. BrOL of BrL and BrOL 3 3 formation in river Ruhr water. Ozone L ] [ 1.57 mM, [Br ] [ 43 ± 0.3 mg LL1. doses are noted in the figure; pH 7.2, [DOC] [ 1.9 mg LL1, [HCOL 3
SO4 plus Cl, see above). Note, that UV/S2 O2 applied on 8 River Ruhr results in a much larger pNBA degradation compared to experiments in presence of humic acids (Fig. 2). This can be explained by a much larger concentration of Cl (see above). For comparing these experimental results with ozonation, ozone was dosed to the unmodified river Ruhr water. As shown in Fig. 4 ozonation leads to a quantitative turnover of Br to BrO 3 (complete mass balance).
Br
(16)
HOBr þ O2 /O2 þ OH þ Br2
(17) (18) (3) (19)
BrO þ O2 !O2 þ 2OH þ Br2 Br2 þ O2 /O2 þ Br þ Br Br2 þ Br2 /Br þ Br3 Br3 þ O2 /O2 þ Br þ Br2
Br ; H2 O
competition kinetics. The Br2 yield depends on the concentration of Br, however, the alternative reaction Br plus Br yielding Br2 gives rise to Br2 and HOBr (Reaction 5) which again might react with O2 (Reaction 16). A similar reaction pathway might also apply for Reaction 17. However, Reaction 17 is much slower than Reaction 16 and can probably be neglected at pH values below the pKa value of HOBr/OBr ((pKa 8.8e9 (Haag and Hoigne´, 1983; von Gunten and Hoigne, 1994) (note that the pH-value of River Ruhr water and humic acids solutions was set to 7.1e7.2). k ¼ 3.5 109 M1 s1
However, a dose of 3 mg L1 O3 is needed for achieving the same extent of pNBA and atrazine degradation as achieved in 1 BrO UV/S2 O2 8 . At that dose 16 0.1 mg L 3 is formed which is largely exceeding the EU and USEPA drinking water standard of 10 mg L1 (von Gunten, 2003b) (Fig. 4). 2 The lack of BrO 3 formation in UV/S2O8 can be explained by the formation of superoxide anion ðO2 Þ that has been reported to occur in reactions of SO4 and S2 O2 8 with different substrates (Furman et al., 2010; Peyton, 1993; Siegrist et al., 2011). Its reaction rate constants with inorganic bromine species are fast and can reduce several intermediates such as HOBr/BrO‒ (Reactions 16e19). It has been suggested that the reduction of HOBr by O2 (Reaction 16) yields HOBr and O2 as primary products and HOBr in turn dissociates into Br plus OH (Schwarz and Bielski, 1986). Br2 may then be formed in a consecutive reaction of Br plus Br (Reaction 2 (see above)). The rate of Reaction 16 determined by Schwarz and Bielski is based on the observed forma tion of Br2 and thus, related to the overall process of Br2 formation (Schwarz and Bielski, 1986). However, that reaction rate fairly agrees with a rate constant published by Sutton and Downes for the reaction of O2 plus HOBr yielding OH, O2 and Br (k ¼ 9.5 108 M1 s1 (Sutton and Downes (1972)) determined by
k k k k
< ¼ ¼ ¼
2 108 M1 s1 1.7 108 M1 s1 1.8 109 M1 s1 1.5 109 M1 s1
(Schwarz and Bielski, 1986) (Schwarz and Bielski, 1986) (Wagner and Strehlow, 1987) (Neta et al., 1988) (Schwarz and Bielski, 1986)
Finally Br is reformed by reduction of Br2 (Reaction 18) and (Reaction 19) (note that Br3 species might be formed in Reaction 3). The reducing agent O2 is known to be formed in the re action of OH with benzene (Pan et al., 1993)). The reaction of SO4 with benzene leads to the formation of a radical cation which subsequently reacts with water to give rise to a hydroxycyclohexadienyl radical (Norman et al., 1970). That radical is also formed in the primary attack of OH on benzene (addition). The subsequent steps are the same for both, OH and SO4 . Thus, a similar yield of O2 can be expected in the reactions of SO4 with aromatic compounds. The constant concentration of Br in UV/S2 O2 8 of River Ruhr water (Fig. 3a) corroborates a reductive pathway resulting in a reformation of Br (Reactions 16e19). Any other sink of Br, Br2 or HOBr/OBr such as bromination of phenolic moieties of the DOC would result in a deficient Br balance. Fang and Shang (2012) also reported a suppression of BrO 3 formation in UV/S2 O2 8 by NOM (Shatin Water Treatment Works in Hong Kong and Suwannee River NOM) (Fang and Shang, 2012). However, their explanation that this suppression is due to scavenging of reactive bromine species by NOM, has been ruled
Br3
376
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
out in the present study by the complete Br balance (Fig. 3) as explained above. The authors further compared the formation of BrO 3 with the degradation of pCBA and observed a suppres 1 sion of BrO 3 formation at a high Br to NOM ratio of 10 mM L mg (absolute concentration of Br ¼ 20 mM), while pCBA was largely degraded (Fang and Shang, 2012). However, at even higher Br to NOM ratios, a turnover of Br to BrO 3 was observed (at 80% degradation of pCBA following turnovers of Br to BrO 3 can be derived from the study of Fang and Shang: 5% turnover at a Br to NOM ratio of 40 mM L mg1 and z25% turnover at a Br to NOM ratio of 200 mM L mg1; The initial Br concentration was kept constant at 20 mM while NOM dosages were varied) (Fang and Shang, 2012). It has to be noted that the very large concentrations of both pCBA (20 mM) and Br (20 mM) Fang and Shang applied, could largely diminishes the fraction of SO4 reacting with NOM (when applied in the concentration range of 0.1e1 mg L1), while Br and pCBA comprise the main reactants. Despite this small fraction of the reaction NOM plus SO4 , BrO 3 formation is effectively suppressed (Fang and Shang, 2012) confirming that the inhibition of BrO 3 is very sensitive towards NOM. It has to be pointed out, that pCBA does probably not interfere by yielding O2 in its reaction with SO4 , since the re action of pCBA plus SO4 led to no distinctive suppression of BrO 3 in absence of NOM (Fang and Shang, 2012). In ozonation the situation is somewhat different, because at typical conditions of water purification, ozone exposures are much higher compared with radical exposures in any radical based processes. Furthermore, O3 competes with the bromine intermediates for O2 with a fast reaction (k ¼ 1.6 109 M1 s1 (Bu¨hler et al., 1984)) yielding OH (Bu¨hler et al., 1984) and BrO 3 can also be formed in a reaction without HOBr/OBr being a requisite intermediate (von Gunten, 2003b). These factors favor the formation of BrO 3 in ozone based processes.
4.
Conclusion
The present work has shown that SO4 oxidizes Br to BrO 3 in a mechanism which is analogous to OH based oxidation. However, the formation of BrO 3 is effectively suppressed in SO4 -based water treatment by NOM and is not to be expected in a hypothetical application of UV/S2 O2 8 in oxidative water treatment. This also applies for processes in which OH are formed with another precursor than H2O2 (e.g., gamma radiolysis or UV/TiO2, (von Sonntag, 2008)). Furthermore, the results demonstrate that main water matrix constituents such as NOM are not only diminishing oxidation strength by consumption of oxidants but can also affect the mechanism of reactions. This underlines the necessity that experiments performed at well-defined conditions (e.g., in pure water matrix) have to be complemented by studying the influences of natural matrices.
Acknowledgment We thank the German Water Chemistry Society for their generous financial support of this work.
references
Acero, J.L., Stemmler, K., von Gunten, U., 2000. Degradation kinetics of atrazine and its degradation products with ozone and OH radicals: a predictive tool for drinking water treatment. Environ. Sci. Technol. 34 (4), 591e597. Anipsitakis, G.P., Dionysiou, D.D., 2003. Degradation of organic contaminants in water with sulfate radicals generated by the conjunction of peroxymonosulfate with cobalt. Environ. Sci. Technol. 37 (20), 4790e4797. Anipsitakis, G.P., Dionysiou, D.D., 2004. Radical generation by the interaction of transition metals with common oxidants. Environ. Sci. Technol. 38 (13), 3705e3712. Anipsitakis, G.P., Dionysiou, D.D., Gonzalez, M.A., 2006. Cobaltmediated activation of peroxymonosulfate and sulfate radical attack on phenolic compounds. Implications of chloride ions. Environ. Sci. Technol. 40 (3), 1000e1007. Beckwith, R.C., Wang, T.X., Margerum, D.W., 1996. Equilibrium and kinetics of bromine hydrolysis. Inorg. Chem. 35 (4), 995e1000. Bu¨hler, R.E., Staehelin, J., Hoigne´, J., 1984. Ozone decomposition in water studied by pulse radiolysis. 1. HO2/O 2 and HO3/O3 as intermediates. J. Phys. Chem. 88 (12), 2560e2564. Buxton, G.V., Dainton, F.S., 1968. Radiolysis of aqueous solutions of oxybromine compounds e spectra and reactions of BrO and BrO2. Proc. R. Soc. Lond. Ser. A-Math. Phys. Sci. 304 (1479), 427e439. Buxton, G.V., Greenstock, C.L., Helman, W.P., Ross, A.B., 1988. Critical review of rate constants for reactions of hydrated electrons, hydrogen-atoms and hydroxyl radicals (OH/O) in aqueous solution. J. Phys. Chem. Ref. Data 17 (2), 513e886. de Souza, G.S.M.B., Rodrigues, L.A., de Oliveira, W.J., Chernicharo, C.A.L., Guimaraes, M.P., Massara, C.L., Grossi, P.A., 2011. Disinfection of domestic effluents by gamma radiation: effects on the inactivation of Ascaris lumbricoides eggs. Water Res. 45 (17), 5523e5528. Dessouki, A.M., Aly, H.F., Sokker, H.H., 1999. The use of gamma radiation for removal of pesticides from waste water. Czechoslov. J. Phys. 49, 521e533. Fang, J.Y., Shang, C., 2012. Bromate formation from bromide oxidation by the UV/persulfate process. Environ. Sci. Technol. 46 (16), 8976e8983. Furman, O.S., Teel, A.L., Watts, R.J., 2010. Mechanism of base activation of persulfate. Environ. Sci. Technol. 44 (16), 6423e6428. Getoff, N., 2002. Factors influencing the efficiency of radiationinduced degradation of water pollutants. Radiat. Phys. Chem. 65 (4e5), 437e446. Haag, W.R., Hoigne´, J., 1983. Ozonation of bromide-containing waters: kinetics of formation of hypobromous acid and bromate. Environ. Sci. Technol. 17 (5), 261e267. Hori, H., Hayakawa, E., Einaga, H., Kutsuna, S., Koike, K., Ibusuki, T., Kiatagawa, H., Arakawa, R., 2004. Decomposition of environmentally persistent perfluorooctanoic acid in water by photochemical approaches. Environ. Sci. Technol. 38 (22), 6118e6124. Hori, H., Nagaoka, Y., Murayama, M., Kutsuna, S., 2008. Efficient decomposition of perfluorocarboxylic acids and alternative fluorochemical surfactants in hot water. Environ. Sci. Technol. 42 (19), 7438e7443. Hori, H., Yamamoto, A., Hayakawa, E., Taniyasu, S., Yamashita, N., Kutsuna, S., Kiatagawa, H., Arakawa, R., 2005. Efficient decomposition of environmentally persistent perfluorocarboxylic acids by use of persulfate as a photochemical oxidant. Environ. Sci. Technol. 39 (7), 2383e2388. Ianni, J.C., 2008. Kintecus Version 3.95. www.kintecus.com. Kla¨ning, U.K., Wolff, T., 1985. Laser flash-photolysis of HClO, ClO, HBrO, and BrO in aqueous-solution e reactions of Cl
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
atoms and Br atoms. Berichte Der Bunsen-Gesellschaft-Phys. Chem. Chem. Phys. 89 (3), 243e245. Kutsuna, S., Hori, H., 2007. Rate constants for aqueous-phase reactions of SO 4 with C2F5C(O)O and C3F7C(O)O at 298 K. Int. J. Chem. Kinet. 39 (5), 276e288. Legrini, O., Oliveros, E., Braun, A.M., 1993. Photochemical processes for water treatment. Chem. Rev. 93 (2), 671e698. Liang, C., Huang, C.F., Chen, Y.J., 2008. Potential for activated persulfate degradation of BTEX contamination. Water Res. 42 (15), 4091e4100. Liang, C., Wang, Z.S., Bruell, C.J., 2007. Influence of pH on persulfate oxidation of TCE at ambient temperatures. Chemosphere 66 (1), 106e113. Manoj, P., Prasanthkumar, K.P., Manoj, V.M., Aravind, U.K., Manojkumar, T.K., Aravindakumar, C.T., 2007. Oxidation of substituted triazines by sulfate radical anion ðSO4 Þ in aqueous medium: a laser flash photolysis and steady state radiolysis study. J. Phys. Org. Chem. 20 (2), 122e129. Mark, G., Schuchmann, M.N., Schuchmann, H.P., von Sonntag, C., 1990. The photolysis of potassium peroxodisulphate in aqueous solution in the presence of tert-butanol: a simple actinometer for 254 nm radiation. J. Photochem. Photobiol. A. 55 (2), 157e168. McElroy, W.J., 1990. A laser photolysis study of the reaction of SO 4 with Cl and the subsequent decay of Cl 2 in aqueous solution. J. Phys. Chem. 94 (6), 2435e2441. Neta, P., Huie, R.E., Ross, A.B., 1988. Rate constants for reactions of inorganic radicals in aqueous solution. J. Phys. Chem. Ref. Data 17 (3), 1027e1040. Neta, P., Madhavan, V., Zemel, H., Fessenden, R.W., 1977. Rate constants and mechanism of reaction of sulfate radical anion with aromatic compounds. J. Am. Chem. Soc. 99 (1), 163e164. NIST National Institute of Standards and Technology e NDRL/ NIST Solution kinetics data base on the web; kinetics.nist.gov/ solution/. Norman, R.O.C., Storey, P.M., West, P.R., 1970. Electron spin resonance studies. Part XXV. Reactions of the sulphate radical anion with organic compounds. J. Chem. Soc. B: Phys. Org., 1087e1095. Pan, X.M., Schuchmann, M.N., Von Sonntag, C., 1993. Oxidation of benzene by the OH radical. A product and pulse radiolysis study in oxygenated aqueous solution. J. Chem. Soc. Perkin Trans. 2 (3), 289e297. Peyton, G.R., 1993. The free-radical chemistry of persulfate-based total organic-carbon analyzers. Mar. Chem. 41 (1e3), 91e103. Redpath, J.L., Willson, R.L., 1975. Chain reactions and radiosensitization: model enzyme studies. Int. J. Radiat. Biol. 27 (4), 389e398. Schwarz, H.A., Bielski, B.H.J., 1986. Reactions of HO2 and O-2 with iodine and bromine and the I 2 and I atom reduction potentials. J. Phys. Chem. 90 (7), 1445e1448.
377
Sutton, H.C., Downes, M.T., 1972. Reactions of the HO2 radical in aqueous solution with bromine and related compounds. J. Chem. Soc. Faraday Trans. 1: Phys. Chem. Condens. Phases 68, 1489e1507. Siegrist, R.L., Crimi, M., Simpkin, T.J. (Eds.), 2011. In sito Chemical oxidation for groundwater remediation, New York, Heidelberg, Dordrecht, London, Springer Science þ Business Media LLC. Tahri, L., Elgarrouj, D., Zantar, S., Mouhib, M., Azmani, A., Sayah, F., 2010. Wastewater treatment using gamma irradiation: Tetouan pilot station, Morocco. Radiat. Phys. Chem. 79 (4), 424e428. Tauber, A., von Sonntag, C., 2000. Products and kinetics of the OH-radical-induced dealkylation of atrazine. Acta Hydrochim. Hydrobiol. 28 (1), 15e23. Tercero Espinoza, L.A., Frimmel, F.H., 2008. Formation of brominated products in irradiated titanium dioxide suspensions containing bromide and dissolved organic carbon. Water Res. 42 (6e7), 1778e1784. Troy, R.C., Margerum, D.W., 1991. Non-metal redox kinetics: hypobromite and hypobromous acid reactions with iodide and with sulfite and the hydrolysis of bromosulfate. Inorg. Chem. 30 (18), 3538e3543. von Gunten, U., 2003a. Ozonation of drinking water: part I. Oxidation kinetics and product formation. Water Res. 37 (7), 1443e1467. von Gunten, U., 2003b. Ozonation of drinking water: part II. Disinfection and by-product formation in presence of bromide, iodide or chlorine. Water Res. 37 (7), 1469e1487. von Gunten, U., Hoigne, J., 1994. Bromate formation during ozonation of bromide-containing waters: interaction of ozone and hydroxyl radical reactions. Environ. Sci. Technol. 28 (7), 1234e1242. von Gunten, U., Oliveras, Y., 1997. Kinetics of the reaction between hydrogen peroxide and hypobromous acid: implication on water treatment and natural systems. Water Res. 31 (4), 900e906. von Gunten, U., Oliveras, Y., 1998. Advanced oxidation of bromide-containig waters: bromate formation mechanisms. Environ. Sci. Technol. 32 (1), 63e70. von Sonntag, C., 2008. Advanced oxidation processes: mechanistic aspects. Water Sci. Technol. 58, 1015e1021. von Sonntag, C., Schuchmann, H.P., 1992. UV disinfection of drinking water and by-product formation e some basic considerations. J. Water SRT e Aqua 41 (2), 67e74. Wagner, I., Strehlow, H., 1987. On the flash-photolysis of bromide ions in aqueous-solutions. Berichte Der Bunsen-GesellschaftPhys. Chem. Chem. Phys. 91 (12), 1317e1321. Zehavi, D., Rabani, J., 1972. Oxidation of aqueous bromide ions by hydroxyl radicals. Pulse radiolytic investigation. J. Phys. Chem. 76 (3), 312e319.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/watres
Formation of bromate in sulfate radical based oxidation: Mechanistic aspects and suppression by dissolved organic matter Holger V. Lutze a,c,*, Rani Bakkour a, Nils Kerlin a, Clemens von Sonntag a,b, Torsten C. Schmidt a,c,d a
University Duisburg-Essen, Instrumental Analytical Chemistry, Universita¨tsstr. 5, D-45141 Essen, Germany Max-Planck-Institut fu¨r Strahlenchemie, Stiftstr. 34-36, Mu¨lheim an der Ruhr, Germany c IWW Water Centre, Moritzstr. 26, D-45476 Mu¨lheim an der Ruhr, Germany d University Duisburg-Essen, Centre for Water and Environmental Research (ZWU), Universita¨tsstr. 2, D-45141 Essen, Germany b
article info
abstract
Article history:
Sulfate radical based oxidation is discussed being a potential alternative to hydroxyl radical
Received 25 November 2013
based oxidation for pollutant control in water treatment. However, formation of undesired
Received in revised form
by-products, has hardly been addressed in the current literature, which is an issue in other
29 December 2013
oxidative processes such as bromate formation in ozonation of bromide containing water
Accepted 1 January 2014
(US-EPA and EU drinking water standard of bromate: 10 mg L1). Sulfate radicals react fast
Available online 9 January 2014
with bromide (k ¼ 3.5 109 M1 s1) which could also yield bromate as final product. The mechanism of bromate formation in aqueous solution in presence of sulfate radicals has
Keywords:
been investigated in the present paper. Further experiments were performed in presence of
Sulfate radicals
humic acids and in surface water for investigating the relevance of bromate formation in
Ozone
context of pollutant control. The formation of bromate by sulfate radicals resembles the
Bromate
well described mechanism of the hydroxyl radical based bromate formation. In both cases
Water treatment
hypobromous acid is a requisite intermediate. In presence of organic matter formation of
Pollutant degradation
bromate is effectively suppressed. That can be explained by formation of superoxide formed in the reaction of sulfate radicals plus aromatic moieties of organic matter, since superoxide reduces hypobromous acid yielding bromine atoms and bromide. Hence formation of bromate can be neglected in sulfate radical based oxidation at typical conditions of water treatment. ª 2014 Published by Elsevier Ltd.
1.
Introduction
Oxidative water treatment based on highly reactive hydroxyl radicals (OH) is referred to as advanced oxidation processes (AOP) and can be used for degrading recalcitrant pollutants
such as pesticides, X-ray contrast media and fuel additives (e.g., MTBE) (von Gunten, 2003a). OH can be generated in various ways e.g., by photolysis of hydrogen peroxide (UV/ H2O2) (Legrini et al., 1993) or in ozonation (von Gunten, 2003a). Beside OH, sulfate radicals ðSO4 Þ are frequently investigated
* Corresponding author. University Duisburg-Essen, Instrumental Analytical Chemistry, Universita¨tsstr. 5, D-45141 Essen, Germany. Tel.: þ49 201 183 6791. E-mail address: [email protected] (H.V. Lutze). 0043-1354/$ e see front matter ª 2014 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.watres.2014.01.001
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
as potential alternative oxidants for water treatment (Anipsitakis and Dionysiou, 2003, 2004; Anipsitakis et al., 2006; Hori et al., 2004, 2008, 2005; Kutsuna and Hori, 2007; Liang et al., 2008, 2007; Manoj et al., 2007) and have already been applied in ground water remediation (Siegrist et al., 2011). SO4 can be 2 formed in various ways using S2 O8 as a radical precursor. One possibility is its photolysis by UVC-radiation (UV/S2 O2 8 ) which is in analogy to UV/H2O2. A major drawback in oxidative water treatment is the formation of undesired by-products. Ozone based processes, for instance, can be limited by the formation of bromate ðBrO 3 Þ a potential carcinogen (US-EPA and EU drinking water standard: 10 mg L1) arising from the oxidation of Br (von Gunten, 2003b). Thereby, formation of BrO 3 can be driven by O3 and/or OH. Corresponding mechanisms have already been discussed in detail (Haag and Hoigne´, 1983; von Gunten, 2003b; von Gunten and Hoigne, 1994; von Gunten and Oliveras, 1998). For reactions in which solely O3 or OH are involved, hypobromous acid (HOBr) is a requisite intermediate. At typical conditions of water treatment HOBr can effectively be reduced by hydrogen peroxide (H2O2) yielding Br (von Gunten and Oliveras, 1997). This instance is used for mitigating BrO 3 formation in ozone applications by addition of H2O2 (von Gunten, 2003b). In most OH based processes H2O2 is used as a radical source (e.g., UV/H2O2), thus preventing BrO 3 formation. UV/TiO2 is known to not require H2O2 to form OH. However, studies indicated that BrO 3 is not formed in this process (Tercero Espinoza and Frimmel, 2008). Also gamma radiolysis which is discussed being a potential water treatment option for both, disinfection (de Souza et al., 2011) and pollutant degradation (Dessouki et al., 1999; Getoff, 2002; Tahri et al., 2010) might oxidize Br yielding BrO 3 . In analogy to OH the reaction of SO4 plus Br (k ¼ 3.5 109 M1 s1 (Redpath and Willson, 1975)) yield BrO 3 (Fang and Shang, 2012). In the recent work of Fang and Shang (2012) an empirical model has been established, which was used for describing the formation of HOBr/OBr and BrO 3 . This approach has been extended in our work by a mechanistic discussion. The present study provides a reaction mechanism based on data available in the literature for developing a kinetic model of SO4 driven formation of BrO3 . This model has been used to describe the behavior of which was experiHOBr and BrO 3 in the oxidation of Br mentally determined at various conditions. Furthermore, the potential of BrO 3 formation in natural matrices has been investigated and contrasted to the oxidation strength available for pollutant control.
2.
Methods
All chemicals were commercially available and used as received. Acetonitrile (99.9%) Sigma Aldrich, atrazine (97.4%) Riedelde Hae¨n, 4-chlorobenzoic acid (pCBA) (99%) Aldrich, hydrochloric acid (37% in water, p.a.) Merck, hydrogen peroxide (30%) Sigma Aldrich, methanol (p.a.) SigmaeAldrich, 4-nitrobenzoic acid (pNBA) (97.4%) Sigma Aldrich, oxygen (99.9%) Liquid Air, phosphoric acid (85%) Merck, potassium bromate (99.5%) Fluka, potassium chloride (99.5%) Riedel-de Hae¨n, pure water has been prepared by treating deionized water with a pure lap ultra instrument (Elga) (electrical resistance 18.6 MU), sodium
371
bicarbonate (99.5%) KMF optichem, sodium bromate (99%) Fluka, sodium carbonate (99.8%) Riedel-de Hae¨n, sodium hydroxide (99.9%, p.a.) VWR, sodium peroxodisulfate (p.a.) SigmaeAldrich, sulfuric acid (95e97%) Applichem International, Suwannee River NOM (reverse osmosis concentrate) International Humic Acid Society, Uridine (99%) Sigma. Sulfate radicals were generated by photolysis of peroxodisulfate (UV/S2 O2 8 ) in a merry-go-round apparatus equipped with a low pressure mercury lamp. This radiation source emits monochromatic light at 254 nm (Heraeus Noble Light GPH303T5L/4, 15 W (185 nm band suppressed)). The fluence rate has been determined by uridine actinometry according to von Sonntag and Schuchmann (1992). Solutions were buffered with phosphate. Even though SO4 reacts with HPO2 4 with a considerable rate (k ¼ 1.2 106 M1 s1) (H2 PO 4 is nearly inert (k < 7 104 M1 s1)) the reactions under study are faster by several orders of magnitudes (k(SO4 plus 9 1 1 Br ) ¼ 3.5 10 M s ). This allows addition of phosphate buffer in excess over bromide (e.g., factor 100), which is necessary for keeping pH constant (experimental details can be found in the caption of corresponding figures). pH-adjustments have been done by addition of sulfuric acid or sodium hydroxide, respectively. Methanol was added to the samples (1 M in the sample) for scavenging low levels of SO4 which may be formed during storage time by thermolysis of S2 O2 8 . BrO3 and Br were analyzed by ion chromatography (Metrohm 883 basic) equipped with a conductivity detector coupled with ion suppression (anion separation column with quaternary ammonium groups: Metrosep A Supp 4 e 250/4.0 mm, particle size 2 9 mm; eluent HCO 3 (1.7 mM), CO3 (1.8 mM) mixed with acetonitrile (30% (v/v)); flow: 1 mL min1; retention times: Br: 3.6 min, BrO 3 : 4.4 min). For determining BrO3 and Br at concentrations in the mM range a different IC system has been used coupled with ion- and subsequent CO2-suppression (Metrohm, 881 Compact IC plus), equipped with a high capacity anion separation column (Metrosep A Supp 5e250/4.0, particle size 5 mm) which was necessary for separation of Cl and BrO 3 (Eluent: 3.2 mM Na2CO3 and 1.0 mM NaHCO3, flow: 0.7 mL min1; retention times: BrO 3 : 8.0 min, Cl : 8.7 min, in River Ruhr Br: 13.0 min). For determining Br and BrO 3 water an IC-ICP-MS system has been used (200 Series from Perkin Elmer, Eluent: 10 mM NaOH, flow: 1.8 mL min1). Atrazine, 4-chlorobenzoic acid (pCBA) and 4-nitrobenzoic acid (pNBA) have been determined by HPLC with UV-detection (Shimadzu) (C18 reversed phase separation column: Bischoff, NUCLEOSIL 100, 250/4.0 mm, particle size: 5.0 mm). As eluent a gradient of methanol/water has been used (gradient program (methanol content (v/v)): 0e3 min: gradient 20e50%, 3e25 min: gradient 50e75%; 25e30 min: gradient 75e20%, 30e38 min: isocratic 20%; flow 0.6 mL min1; retention times/measured wave length: pNBA: 20.6 min/262 nm; atrazine: 25.8 min/ 234 nm, pCBA: 26.5 min/234 nm). HOBr has been determined by UV-absorption as OBr at 329 nm (ε(OBr) ¼ 332 M1 cm1 (Troy and Margerum, 1991)) by adjusting the solution to pH 11. Model calculations have been performed by using the software tool Kintecus (Ianni, 2008) (quantum yields for radical formation and molar absorption coefficients of peroxodisulfate can be obtained from Mark et al., 1990). Experiments in presence of humic acids where performed for simulating bromate formation in a real water during UV/
372
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
S2 O2 8 . Beside Br , probe compounds have been added for two purposes:
1. As model compound for simulating the degradation of pollutants, atrazine has been added (k(atrazine þ SO4 ) ¼ 3 109 M1 s1 (Manoj et al., 2007)) 2. As indication for OH which may be formed in the reaction of SO4 with Cl (McElroy, 1990). For this purpose pNBA was chosen because it is inert in presence of SO4 (k(SO4 þ pNBA) 106 M1 s1 (Neta et al., 1977)) and UV radiation at present experimental conditions. However, it reacts fast with OH (k(OH þ pNBA) ¼ 2.6 109 M1 s1 (Buxton et al., 1988)). Thus, it is a sensitive indicator for the presence of OH.
induce a cascade of reactions (Table 1) yielding HOBr and BrO 3 as main products. The HOBr yield displays a distinctive dependency on the pH-value, which has also been observed in the reaction of OH (von Gunten and Oliveras, 1998). The speciation of HOBr/OBr (pKa 8.8e9 (Haag and Hoigne´, 1983; von Gunten and Hoigne, 1994)), largely affects the rates of its consecutive degradation by Br which predominantly reacts with hypobromite (OBr)
These compounds have been added in low concentrations (0.5 mM) for assuring that their contribution in scavenging of SO4 /OH is below 10%. In this situation most radicals react with the main water constituents (e.g., DOC), which is typical in oxidative water treatment. As another matrix for simulating real water treatment conditions River Ruhr water (RWW, Mu¨lheim a.d. Ruhr) has been used. The water has been filtered with cellulose acetate/ cellulose nitrate filters with a pore size of 0.45 mm (Whatman). If not noted otherwise, the concentration of HCO 3 has been adjusted as follows. The original amount of HCO 3 present in the River water has been removed by acidification (pH 2) and purging with oxygen ( 5 min). Then the pH has been readjusted to pH 7.2 and defined amounts of HCO 3 were added to achieve the desired concentration. Ozonation experiments were performed as follows. An aqueous O3 stock solution was prepared by purging an O3 enriched gas through ice cooled pure water. At stationary conditions a steady state concentration of ozone of 1 mM O3 could be achieved. This solution was used to dose specific amounts of O3 to samples under study. Therefore glass syringes with stainless steel cannulas where used. For maintaining a constant O3 concentration in the O3 stock solution, continuous purging with the ozone enriched gas was necessary. O3 was generated by an O3 generator purchased from BMT Messtechnik Berlin (Philaqua 802x) with oxygen as feed gas.
3.
Results and discussion
3.1.
Oxidation of bromide
The reaction of SO4 with Br leads to the formation of BrO 3 in aqueous solution (Fig. 1). The product pattern is in good agreement with the study of Fang and Shang (2012) in that they observed a complete turnover of Br via HOBr/OBr to BrO 3 . Furthermore, Fang and Shang reported HOBr/OBr evolving to a peak concentration corresponding to 25% of the initial Br concentration (at pH 7), which matches well the value of the present study (20%, Fig. 1b). This supports the experimental results of both studies. The chemistry of bromine is well documented in the literature and sufficient reactions including their kinetics are reported for describing the oxidation of Br by SO4 as will be explained below. Bromine atoms formed as primary products (Reaction 1) may
Fig. 1 e Product formation in the reaction of SO4L with BrL. Reaction conditions: [BrL]: 1 mM, [S2 O2L 8 ]: 10 mM, [phosphate]: 100 mM, T: 25 C, a) Initial pH: 8.0e8.1 (gradual decrease to pH 7.6 within a reaction time of 60 min), inset: bromate formation in presence of 1 mM H2O2, dotted-dashed line (inset) model calculation of H2O2 consumption; b) Initial pH 7.0 (gradual decrease to pH 6.8 within a reaction time of 90 min); circles: BrL, triangles: HOBr/OBrL; stars: BrOL 3 , solid lines model calculations, dashed line model of HOBr including the reaction of Br plus HOBr (k: 5 3 107 ML1 sL1) (not shown in a), see below).
373
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
Table 1 e Reactions occurring in UV/S2 O2L of bromide-containing aqueous solutions. 8 No.
Reaction
kforward ; kbackward ; or pKa
Ref.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Br- þ SO4 /Br þ SO2 4 Br þ Br- %Br2 Br2 þ Br2 /Br- þ Br3 Br3 #Br- þ Br2 Br2 þ H2 O#HOBr þ Br- þ Hþ HOBr#BrO- þ Hþ BrO þ Br2 /BrO þ 2Br BrO- þ Br /BrO þ Br þ 2BrO þ H2 O/BrO- þ BrO 2 þ 2H - þ BrO BrO þ BrO /BrO 2 2 BrO þ Br /Br- þ BrO þ BrO-
3.5 109 M1 s1 1 1010 M1 s1, 1 105 s1 1.8 109 M1 s1 1 107 s1, 1.5 109 M1 s1 97 s1, 1.6 1010 M2 s1 pKa: 8.8 8 0.7 107 M1 s1 4.1 109 M1 s1 4.9 1 109 M1 s1 3.4 0.7 108 M1 s1 8.0 0.8 107 M1 s1 1.4 109 M1 s1, 7.4 107 s1 4.2 107 M1 s1
(Redpath and Willson, 1975) (Zehavi and Rabani, 1972) (Neta et al., 1988) (Beckwith et al., 1996) (Beckwith et al., 1996) (Haag and Hoigne´, 1983) (Buxton and Dainton, 1968) (Kla¨ning and Wolff, 1985) (Buxton and Dainton, 1968) (Buxton and Dainton, 1968) (Buxton and Dainton, 1968) (Buxton and Dainton, 1968) (Buxton and Dainton, 1968)
2
2
2BrO2 #Br2 O4 þ 2BrO2 þ H2 O/BrO 2 þ BrO3 þ 2H
(von Gunten and Oliveras, 1998). That explains the higher yields in HOBr/OBr with decreasing pH. In analogy to mechanistic considerations in OH driven formation of BrO 3 the present mechanism has been modeled with equations shown in Table 1 (modeled data: lines in Fig. 1). For determining the reaction rates in reactions involving HOBr a pKa of 8.8 has been used, which was suggested for solutions with an ionic strength of > 0.15 M (Haag and Hoigne´, 1983) that suits our experimental conditions (ionic strength > 0.250 M). The good agreement of the model with the experimental results indicates a similar reaction mechanism compared to that proposed by von Gunten and Oliveras for OH-based BrO 3 formation (von Gunten and Oliveras, 1998). However, the model does not fit well to the experimental results of BrO 3 formation at pH 6 (data not shown). At pH 6 the formation of HOBr/OBr is largely overestimated, indicating that oxidative species may also react with HOBr, which has not been taken into account in the model calculations (Table 1). A good agreement with the HOBr concentration pattern can be achieved by taking a reaction of the bromine atom with HOBr into account assuming a reaction rate of 5 107 M1 s1. This rate constant is around 2 orders of magnitude smaller compared with the reaction rate of Br plus OBr. This seems to be a realistic value which has also been reported for other acid base couples in their reaction with Br such as HCN (k ¼ 1 107 M1 s1) and CN (k ¼ 8 108 M1 s1) (NIST). Furthermore, this reaction does not change the other model calculations for higher pH values significantly (dashed line in Fig. 1b), because the reaction Br plus OBr is largely dominating (note that at pH 8 this effect must be even weaker (calculations not shown in Fig. 1a)). A contribution of SO4 to oxidation of HOBr is not likely to happen. Included in the model, this reaction affects the outcome only if a very high rate constant is assumed for both species OBr and HOBr (k(SO4 þ HOBr/OBr)) ¼ 1010 M1 s1), since it has to compete with the reaction SO4 plus Br (Reaction 1). Such a rate constant would be much higher than that for the reaction with Br (3.5 109 M1 s1 (Redpath and Willson, 1975)), which is most unlikely. In the current model the formation of HOBr/BrO has been considered to be a requisite intermediate. This has been confirmed by experiments in presence of H2O2, which reduces HOBr/OBr to give rise to Br and thus, suppressing the
formation of BrO 3 . This effect is shown in Fig. 1a (inset). Upon addition of 1 mM H2O2 before irradiation, BrO 3 formation is delayed by ca. 20 min. This has been modeled by including the reduction of HOBr/OBr via H2O2 in the model (note that the reaction rate is pH-dependent (k ¼ 1.96 105 M1 s1 at pH 8) (von Gunten and Oliveras, 1997)). Taking this rate constant into account the delay in BrO 3 formation can be described by the model (Fig. 1a (inset)). Both, experimental data and the model outcome revealed that the presence of H2O2 effectively suppresses the Br turnover (note, that the photochemical degradation of H2O2 is largely overwhelmed by its reaction with OBr/HOBr). This is a strong evidence for HOBr being a requisite intermediate. To this end experiments were conducted in absence of typical constituents in natural waters such as DOC. The effect of a natural water matrix has also been investigated in the present study. In a first step the effect of DOC on SO4 driven BrO3 formation was investigated. Therefore, DOC and Br concentrations were varied for achieving different pro portions of SO4 reacting with bromide ([DOC] ¼ 1, 3, and 10 mg L1, [Br] ¼ 1 mM). The ratios of [Br]/[DOC] range from 0.1 to 0.5 mM L mg1, covering the range found in natural waters such as river Ruhr ([DOC] ¼ 1.9 mg L1, [Br] ¼ 0.54 mM (see caption of Fig. 2), Br/DOC ¼ 0.28 mM L mg1). Within a reaction time sufficient for degrading >90% atrazine neither Br degradation nor BrO 3 formation has been observed (½S2 O2 8 ¼ 0.5 mM, reaction time ¼ 2, 6 and 25 min for [DOC] ¼ 1, 3 and 10 mg L1). Even an increase of Br concentration in presence of 10 mg L1 DOC from 1 mM to 50 mM did not reveal any formation of BrO 3 within a reaction time of 50 min, even though the proportion of SO4 reacting with Br increased largely. Note that in presence of 50 mM Br atrazine degradation is mainly driven by photolysis (Fig. 2). In 1990 McElroy postulated the formation of OH upon oxidation of Cl to Cl by SO4 (McElroy, 1990). The humic acids used in the present study contain small amounts of Cl, which cannot be avoided without changing the chemical composition of the organic matter. Furthermore, Cl can be introduced by several means such as pH-measurements and contaminated reagents. IC measurements revealed Cl concentrations of 9.1e11.5 mM in experiments in presence of 9e10 mg L1 DOC and a Cl content of 1.2 0.3 mM per mg L1 DOC as average value for all experiments in presence of humic acids. For
374
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
Fig. 2 e Degradation of atrazine (squares) and pNBA L (triangles) in UV/S2 O2L 8 in presence of humic acids and Br ; L [Br ] [ 1 mM (open symbols), 50 mM (filled symbols), stars: direct photo degradation of atrazine, [DOC] [ 9.8e10.0 mg LL1, [phosphate] [ 2.5 mM, [ClL] [ 9.1e11.5 mM, pH: 7.1e7.2, [S2O2L 8 ] = 0.5 mM.
testing if OH affect the reaction system the degradation of pNBA was observed, that readily react with OH (k ¼ 2.6 109 M1 s1 (Buxton et al., 1988)), but barely reacts with SO4 (k < 106 M1 s1 (Neta et al., 1977)). A slow degradation of pNBA can be observed which indicates a rather small impact of OH (and/or Cl) on the present reaction system. However, their contribution on the transformation of probe compounds or Br is negligible as will be explained below. OH Since atrazine also reacts readily with (k ¼ 3 109 M1 s1 (Acero et al., 2000; Tauber and von Sonntag, 2000)) a similar degradation rate of pNBA and atrazine would indicate a strong contribution of OH in the present reaction system. The degradation rate of atrazine is the sum of SO4 driven degradation and direct photolysis. The first order
degradation rates in direct photolysis and UV/S2 O2 were 8 determined to be k(photolysis of atrazine) ¼ 0.0005 s1 and 1 k(atrazine in UV/S2 O2 8 ) ¼ 0.0019 s . Subtracting the rate of direct photo-degradation from degradation in UV/S2 O2 8 yields the degradation rate driven by SO4 (k(atrazine degradation by SO4 ) ¼ 0.0014 s1). pNBA was not significantly degraded by direct UV-radiation, thus the degradation rate is based on its 1 reaction with OH only (k (pNBA in UV/S2 O2 8 ) ¼ 0.00012 s ). Since the radical driven oxidation of atrazine is z12 times faster than pNBA degradation, the influence of OH on the reaction system under study seem to be negligible. Analogous experiments were performed in river Ruhr water, which is used as raw water for drinking water production. The Br content of this water was 40e50 mg L1 (0.51e0.63 mM), which may already lead to an exceedance of the drinking water standard for bromate in ozonation under certain treatment conditions (von Gunten, 2003b). River Ruhr water also contains Cl in the mM range (0.78 0.02 mM) that leads to a large proportion of SO4 reacting with Cl (>90%) yielding OH (McElroy, 1990) (see above). This does not necessarily prevent BrO 3 formation because BrO3 can also be formed in the reactions with OH as mentioned above (von Gunten and Oliveras, 1998). However, BrO 3 has not been in River Ruhr water. Variation of the formed in UV/S2 O2 8 HCO 3 concentration (0e2 mM) also did not lead to a detectable formation of BrO 3 . This is illustrated in Fig. 3a showing the concentration of Br in different photochemical processes. It can be observed that the Br concentrations in are near the Br level of the original River Ruhr UV/S2 O2 8 water (red line) and resembles that of UV/H2O2 and UV which (the are both known to not form BrO 3 in presence of Br differences in case of UV and UV/H2O2 at 0 mM HCO 3 are probably related to an analytical issues). This suggests, that no turnover of Br took place, while the oxidation power achieved in UV/S2 O2 8 suffices for >95% degradation of atrazine and up to 95% degradation of pNBA (Fig. 3b) (note that pNBA probably is degraded by OH formed in the reaction of
Fig. 3 e a) Concentration of BrL in River Ruhr water at different alkalinities in photochemical treatment; UV/S2 O2L (circles), 8 UV/H2O2 (triangles), UV (crosses), vertical lines indicate initial conditions, the following points represent subsequent samples after different reaction times given at the x-axis; red line: concentration of BrL in untreated River Ruhr water, horizontal black lines: upper and lower range of the BrL concentration in untreated River Ruhr water; b) Degradation of organic compounds in the experiments of a); gray area [ atrazine, black [ pCBA; a) and b) numbers indicate the 2L concentration of HCOL 3 in mM; shaded area [ original River Ruhr water (no adjustment of alkalinity); [S2 O8 ]0 [ [H2O2]0 [ L 0.5 mM, [phosphate] [ 1 mM; T 25 C, pH 7.2, [DOC] [ 1.9 mg/L, [HCO3 ] [ 0e2 mM.
375
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
Fig. 4 e a) Turnover of BrL (open circles) to BrOL 3 (closed circles) after different ozone doses, the black line indicates the sum ; b) Degradation of atrazine (triangles) and pNBA (crosses) vs. BrOL of BrL and BrOL 3 3 formation in river Ruhr water. Ozone L ] [ 1.57 mM, [Br ] [ 43 ± 0.3 mg LL1. doses are noted in the figure; pH 7.2, [DOC] [ 1.9 mg LL1, [HCOL 3
SO4 plus Cl, see above). Note, that UV/S2 O2 applied on 8 River Ruhr results in a much larger pNBA degradation compared to experiments in presence of humic acids (Fig. 2). This can be explained by a much larger concentration of Cl (see above). For comparing these experimental results with ozonation, ozone was dosed to the unmodified river Ruhr water. As shown in Fig. 4 ozonation leads to a quantitative turnover of Br to BrO 3 (complete mass balance).
Br
(16)
HOBr þ O2 /O2 þ OH þ Br2
(17) (18) (3) (19)
BrO þ O2 !O2 þ 2OH þ Br2 Br2 þ O2 /O2 þ Br þ Br Br2 þ Br2 /Br þ Br3 Br3 þ O2 /O2 þ Br þ Br2
Br ; H2 O
competition kinetics. The Br2 yield depends on the concentration of Br, however, the alternative reaction Br plus Br yielding Br2 gives rise to Br2 and HOBr (Reaction 5) which again might react with O2 (Reaction 16). A similar reaction pathway might also apply for Reaction 17. However, Reaction 17 is much slower than Reaction 16 and can probably be neglected at pH values below the pKa value of HOBr/OBr ((pKa 8.8e9 (Haag and Hoigne´, 1983; von Gunten and Hoigne, 1994) (note that the pH-value of River Ruhr water and humic acids solutions was set to 7.1e7.2). k ¼ 3.5 109 M1 s1
However, a dose of 3 mg L1 O3 is needed for achieving the same extent of pNBA and atrazine degradation as achieved in 1 BrO UV/S2 O2 8 . At that dose 16 0.1 mg L 3 is formed which is largely exceeding the EU and USEPA drinking water standard of 10 mg L1 (von Gunten, 2003b) (Fig. 4). 2 The lack of BrO 3 formation in UV/S2O8 can be explained by the formation of superoxide anion ðO2 Þ that has been reported to occur in reactions of SO4 and S2 O2 8 with different substrates (Furman et al., 2010; Peyton, 1993; Siegrist et al., 2011). Its reaction rate constants with inorganic bromine species are fast and can reduce several intermediates such as HOBr/BrO‒ (Reactions 16e19). It has been suggested that the reduction of HOBr by O2 (Reaction 16) yields HOBr and O2 as primary products and HOBr in turn dissociates into Br plus OH (Schwarz and Bielski, 1986). Br2 may then be formed in a consecutive reaction of Br plus Br (Reaction 2 (see above)). The rate of Reaction 16 determined by Schwarz and Bielski is based on the observed forma tion of Br2 and thus, related to the overall process of Br2 formation (Schwarz and Bielski, 1986). However, that reaction rate fairly agrees with a rate constant published by Sutton and Downes for the reaction of O2 plus HOBr yielding OH, O2 and Br (k ¼ 9.5 108 M1 s1 (Sutton and Downes (1972)) determined by
k k k k
< ¼ ¼ ¼
2 108 M1 s1 1.7 108 M1 s1 1.8 109 M1 s1 1.5 109 M1 s1
(Schwarz and Bielski, 1986) (Schwarz and Bielski, 1986) (Wagner and Strehlow, 1987) (Neta et al., 1988) (Schwarz and Bielski, 1986)
Finally Br is reformed by reduction of Br2 (Reaction 18) and (Reaction 19) (note that Br3 species might be formed in Reaction 3). The reducing agent O2 is known to be formed in the re action of OH with benzene (Pan et al., 1993)). The reaction of SO4 with benzene leads to the formation of a radical cation which subsequently reacts with water to give rise to a hydroxycyclohexadienyl radical (Norman et al., 1970). That radical is also formed in the primary attack of OH on benzene (addition). The subsequent steps are the same for both, OH and SO4 . Thus, a similar yield of O2 can be expected in the reactions of SO4 with aromatic compounds. The constant concentration of Br in UV/S2 O2 8 of River Ruhr water (Fig. 3a) corroborates a reductive pathway resulting in a reformation of Br (Reactions 16e19). Any other sink of Br, Br2 or HOBr/OBr such as bromination of phenolic moieties of the DOC would result in a deficient Br balance. Fang and Shang (2012) also reported a suppression of BrO 3 formation in UV/S2 O2 8 by NOM (Shatin Water Treatment Works in Hong Kong and Suwannee River NOM) (Fang and Shang, 2012). However, their explanation that this suppression is due to scavenging of reactive bromine species by NOM, has been ruled
Br3
376
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
out in the present study by the complete Br balance (Fig. 3) as explained above. The authors further compared the formation of BrO 3 with the degradation of pCBA and observed a suppres 1 sion of BrO 3 formation at a high Br to NOM ratio of 10 mM L mg (absolute concentration of Br ¼ 20 mM), while pCBA was largely degraded (Fang and Shang, 2012). However, at even higher Br to NOM ratios, a turnover of Br to BrO 3 was observed (at 80% degradation of pCBA following turnovers of Br to BrO 3 can be derived from the study of Fang and Shang: 5% turnover at a Br to NOM ratio of 40 mM L mg1 and z25% turnover at a Br to NOM ratio of 200 mM L mg1; The initial Br concentration was kept constant at 20 mM while NOM dosages were varied) (Fang and Shang, 2012). It has to be noted that the very large concentrations of both pCBA (20 mM) and Br (20 mM) Fang and Shang applied, could largely diminishes the fraction of SO4 reacting with NOM (when applied in the concentration range of 0.1e1 mg L1), while Br and pCBA comprise the main reactants. Despite this small fraction of the reaction NOM plus SO4 , BrO 3 formation is effectively suppressed (Fang and Shang, 2012) confirming that the inhibition of BrO 3 is very sensitive towards NOM. It has to be pointed out, that pCBA does probably not interfere by yielding O2 in its reaction with SO4 , since the re action of pCBA plus SO4 led to no distinctive suppression of BrO 3 in absence of NOM (Fang and Shang, 2012). In ozonation the situation is somewhat different, because at typical conditions of water purification, ozone exposures are much higher compared with radical exposures in any radical based processes. Furthermore, O3 competes with the bromine intermediates for O2 with a fast reaction (k ¼ 1.6 109 M1 s1 (Bu¨hler et al., 1984)) yielding OH (Bu¨hler et al., 1984) and BrO 3 can also be formed in a reaction without HOBr/OBr being a requisite intermediate (von Gunten, 2003b). These factors favor the formation of BrO 3 in ozone based processes.
4.
Conclusion
The present work has shown that SO4 oxidizes Br to BrO 3 in a mechanism which is analogous to OH based oxidation. However, the formation of BrO 3 is effectively suppressed in SO4 -based water treatment by NOM and is not to be expected in a hypothetical application of UV/S2 O2 8 in oxidative water treatment. This also applies for processes in which OH are formed with another precursor than H2O2 (e.g., gamma radiolysis or UV/TiO2, (von Sonntag, 2008)). Furthermore, the results demonstrate that main water matrix constituents such as NOM are not only diminishing oxidation strength by consumption of oxidants but can also affect the mechanism of reactions. This underlines the necessity that experiments performed at well-defined conditions (e.g., in pure water matrix) have to be complemented by studying the influences of natural matrices.
Acknowledgment We thank the German Water Chemistry Society for their generous financial support of this work.
references
Acero, J.L., Stemmler, K., von Gunten, U., 2000. Degradation kinetics of atrazine and its degradation products with ozone and OH radicals: a predictive tool for drinking water treatment. Environ. Sci. Technol. 34 (4), 591e597. Anipsitakis, G.P., Dionysiou, D.D., 2003. Degradation of organic contaminants in water with sulfate radicals generated by the conjunction of peroxymonosulfate with cobalt. Environ. Sci. Technol. 37 (20), 4790e4797. Anipsitakis, G.P., Dionysiou, D.D., 2004. Radical generation by the interaction of transition metals with common oxidants. Environ. Sci. Technol. 38 (13), 3705e3712. Anipsitakis, G.P., Dionysiou, D.D., Gonzalez, M.A., 2006. Cobaltmediated activation of peroxymonosulfate and sulfate radical attack on phenolic compounds. Implications of chloride ions. Environ. Sci. Technol. 40 (3), 1000e1007. Beckwith, R.C., Wang, T.X., Margerum, D.W., 1996. Equilibrium and kinetics of bromine hydrolysis. Inorg. Chem. 35 (4), 995e1000. Bu¨hler, R.E., Staehelin, J., Hoigne´, J., 1984. Ozone decomposition in water studied by pulse radiolysis. 1. HO2/O 2 and HO3/O3 as intermediates. J. Phys. Chem. 88 (12), 2560e2564. Buxton, G.V., Dainton, F.S., 1968. Radiolysis of aqueous solutions of oxybromine compounds e spectra and reactions of BrO and BrO2. Proc. R. Soc. Lond. Ser. A-Math. Phys. Sci. 304 (1479), 427e439. Buxton, G.V., Greenstock, C.L., Helman, W.P., Ross, A.B., 1988. Critical review of rate constants for reactions of hydrated electrons, hydrogen-atoms and hydroxyl radicals (OH/O) in aqueous solution. J. Phys. Chem. Ref. Data 17 (2), 513e886. de Souza, G.S.M.B., Rodrigues, L.A., de Oliveira, W.J., Chernicharo, C.A.L., Guimaraes, M.P., Massara, C.L., Grossi, P.A., 2011. Disinfection of domestic effluents by gamma radiation: effects on the inactivation of Ascaris lumbricoides eggs. Water Res. 45 (17), 5523e5528. Dessouki, A.M., Aly, H.F., Sokker, H.H., 1999. The use of gamma radiation for removal of pesticides from waste water. Czechoslov. J. Phys. 49, 521e533. Fang, J.Y., Shang, C., 2012. Bromate formation from bromide oxidation by the UV/persulfate process. Environ. Sci. Technol. 46 (16), 8976e8983. Furman, O.S., Teel, A.L., Watts, R.J., 2010. Mechanism of base activation of persulfate. Environ. Sci. Technol. 44 (16), 6423e6428. Getoff, N., 2002. Factors influencing the efficiency of radiationinduced degradation of water pollutants. Radiat. Phys. Chem. 65 (4e5), 437e446. Haag, W.R., Hoigne´, J., 1983. Ozonation of bromide-containing waters: kinetics of formation of hypobromous acid and bromate. Environ. Sci. Technol. 17 (5), 261e267. Hori, H., Hayakawa, E., Einaga, H., Kutsuna, S., Koike, K., Ibusuki, T., Kiatagawa, H., Arakawa, R., 2004. Decomposition of environmentally persistent perfluorooctanoic acid in water by photochemical approaches. Environ. Sci. Technol. 38 (22), 6118e6124. Hori, H., Nagaoka, Y., Murayama, M., Kutsuna, S., 2008. Efficient decomposition of perfluorocarboxylic acids and alternative fluorochemical surfactants in hot water. Environ. Sci. Technol. 42 (19), 7438e7443. Hori, H., Yamamoto, A., Hayakawa, E., Taniyasu, S., Yamashita, N., Kutsuna, S., Kiatagawa, H., Arakawa, R., 2005. Efficient decomposition of environmentally persistent perfluorocarboxylic acids by use of persulfate as a photochemical oxidant. Environ. Sci. Technol. 39 (7), 2383e2388. Ianni, J.C., 2008. Kintecus Version 3.95. www.kintecus.com. Kla¨ning, U.K., Wolff, T., 1985. Laser flash-photolysis of HClO, ClO, HBrO, and BrO in aqueous-solution e reactions of Cl
w a t e r r e s e a r c h 5 3 ( 2 0 1 4 ) 3 7 0 e3 7 7
atoms and Br atoms. Berichte Der Bunsen-Gesellschaft-Phys. Chem. Chem. Phys. 89 (3), 243e245. Kutsuna, S., Hori, H., 2007. Rate constants for aqueous-phase reactions of SO 4 with C2F5C(O)O and C3F7C(O)O at 298 K. Int. J. Chem. Kinet. 39 (5), 276e288. Legrini, O., Oliveros, E., Braun, A.M., 1993. Photochemical processes for water treatment. Chem. Rev. 93 (2), 671e698. Liang, C., Huang, C.F., Chen, Y.J., 2008. Potential for activated persulfate degradation of BTEX contamination. Water Res. 42 (15), 4091e4100. Liang, C., Wang, Z.S., Bruell, C.J., 2007. Influence of pH on persulfate oxidation of TCE at ambient temperatures. Chemosphere 66 (1), 106e113. Manoj, P., Prasanthkumar, K.P., Manoj, V.M., Aravind, U.K., Manojkumar, T.K., Aravindakumar, C.T., 2007. Oxidation of substituted triazines by sulfate radical anion ðSO4 Þ in aqueous medium: a laser flash photolysis and steady state radiolysis study. J. Phys. Org. Chem. 20 (2), 122e129. Mark, G., Schuchmann, M.N., Schuchmann, H.P., von Sonntag, C., 1990. The photolysis of potassium peroxodisulphate in aqueous solution in the presence of tert-butanol: a simple actinometer for 254 nm radiation. J. Photochem. Photobiol. A. 55 (2), 157e168. McElroy, W.J., 1990. A laser photolysis study of the reaction of SO 4 with Cl and the subsequent decay of Cl 2 in aqueous solution. J. Phys. Chem. 94 (6), 2435e2441. Neta, P., Huie, R.E., Ross, A.B., 1988. Rate constants for reactions of inorganic radicals in aqueous solution. J. Phys. Chem. Ref. Data 17 (3), 1027e1040. Neta, P., Madhavan, V., Zemel, H., Fessenden, R.W., 1977. Rate constants and mechanism of reaction of sulfate radical anion with aromatic compounds. J. Am. Chem. Soc. 99 (1), 163e164. NIST National Institute of Standards and Technology e NDRL/ NIST Solution kinetics data base on the web; kinetics.nist.gov/ solution/. Norman, R.O.C., Storey, P.M., West, P.R., 1970. Electron spin resonance studies. Part XXV. Reactions of the sulphate radical anion with organic compounds. J. Chem. Soc. B: Phys. Org., 1087e1095. Pan, X.M., Schuchmann, M.N., Von Sonntag, C., 1993. Oxidation of benzene by the OH radical. A product and pulse radiolysis study in oxygenated aqueous solution. J. Chem. Soc. Perkin Trans. 2 (3), 289e297. Peyton, G.R., 1993. The free-radical chemistry of persulfate-based total organic-carbon analyzers. Mar. Chem. 41 (1e3), 91e103. Redpath, J.L., Willson, R.L., 1975. Chain reactions and radiosensitization: model enzyme studies. Int. J. Radiat. Biol. 27 (4), 389e398. Schwarz, H.A., Bielski, B.H.J., 1986. Reactions of HO2 and O-2 with iodine and bromine and the I 2 and I atom reduction potentials. J. Phys. Chem. 90 (7), 1445e1448.
377
Sutton, H.C., Downes, M.T., 1972. Reactions of the HO2 radical in aqueous solution with bromine and related compounds. J. Chem. Soc. Faraday Trans. 1: Phys. Chem. Condens. Phases 68, 1489e1507. Siegrist, R.L., Crimi, M., Simpkin, T.J. (Eds.), 2011. In sito Chemical oxidation for groundwater remediation, New York, Heidelberg, Dordrecht, London, Springer Science þ Business Media LLC. Tahri, L., Elgarrouj, D., Zantar, S., Mouhib, M., Azmani, A., Sayah, F., 2010. Wastewater treatment using gamma irradiation: Tetouan pilot station, Morocco. Radiat. Phys. Chem. 79 (4), 424e428. Tauber, A., von Sonntag, C., 2000. Products and kinetics of the OH-radical-induced dealkylation of atrazine. Acta Hydrochim. Hydrobiol. 28 (1), 15e23. Tercero Espinoza, L.A., Frimmel, F.H., 2008. Formation of brominated products in irradiated titanium dioxide suspensions containing bromide and dissolved organic carbon. Water Res. 42 (6e7), 1778e1784. Troy, R.C., Margerum, D.W., 1991. Non-metal redox kinetics: hypobromite and hypobromous acid reactions with iodide and with sulfite and the hydrolysis of bromosulfate. Inorg. Chem. 30 (18), 3538e3543. von Gunten, U., 2003a. Ozonation of drinking water: part I. Oxidation kinetics and product formation. Water Res. 37 (7), 1443e1467. von Gunten, U., 2003b. Ozonation of drinking water: part II. Disinfection and by-product formation in presence of bromide, iodide or chlorine. Water Res. 37 (7), 1469e1487. von Gunten, U., Hoigne, J., 1994. Bromate formation during ozonation of bromide-containing waters: interaction of ozone and hydroxyl radical reactions. Environ. Sci. Technol. 28 (7), 1234e1242. von Gunten, U., Oliveras, Y., 1997. Kinetics of the reaction between hydrogen peroxide and hypobromous acid: implication on water treatment and natural systems. Water Res. 31 (4), 900e906. von Gunten, U., Oliveras, Y., 1998. Advanced oxidation of bromide-containig waters: bromate formation mechanisms. Environ. Sci. Technol. 32 (1), 63e70. von Sonntag, C., 2008. Advanced oxidation processes: mechanistic aspects. Water Sci. Technol. 58, 1015e1021. von Sonntag, C., Schuchmann, H.P., 1992. UV disinfection of drinking water and by-product formation e some basic considerations. J. Water SRT e Aqua 41 (2), 67e74. Wagner, I., Strehlow, H., 1987. On the flash-photolysis of bromide ions in aqueous-solutions. Berichte Der Bunsen-GesellschaftPhys. Chem. Chem. Phys. 91 (12), 1317e1321. Zehavi, D., Rabani, J., 1972. Oxidation of aqueous bromide ions by hydroxyl radicals. Pulse radiolytic investigation. J. Phys. Chem. 76 (3), 312e319.
