Abstract

This study evaluates the effect of photochemical advanced oxidation processes (AOPs) (O3/UV, H2O2/UV and O3/H2O2/UV) on the formation potential (FP) of emerging disinfection by-products including nitrogenous by-products (N-DBPs) and haloketones (HKs) in groundwater from part of the Pannonian Basin (AP Vojvodina, Republic of Serbia). Among the N-DBPs, the haloacetonitrile (HAN) precursor contents were 9.83 ± 0.59 μg/L while precursors of halonitromethanes, particularly trichloronitromethane (TCNM) were not detected. Similarly, precursors of HKs as carbonaceous DBPs were also not detected in raw water. Ozonation alone and the H2O2/UV process with a lower UV dose maximally decomposed HAN precursors (about 70%) while during O3-based AOPs, HANFP varied significantly. Application of UV photolysis and H2O2/UV processes with increasing UV dose doubled the HANFP. Ozone alone, O3/UV and H2O2/UV slightly increased HK formation potential, particularly 1,1-dichloro-2-propanone FP (0.93 ± 0.21 to 2.01 ± 0.37 μg/L). None of the investigated treatments influenced the formation of TCNM precursors. The effect of the applied treatments on bromide incorporation was most evident for HANs.

ABBREVIATIONS

     
  • AOPs

    advanced oxidation processes

  •  
  • BCAN

    bromochloroacetonitrile

  •  
  • C-DBPs

    carbonaceous disinfection by-products

  •  
  • DBAN

    dibromoacetonitrile

  •  
  • DBPs

    disinfection by-products

  •  
  • DCAN

    dichloroacetonitrile

  •  
  • DCP

    1,1-dichloro-2-propanone

  •  
  • DOC

    dissolved organic carbon

  •  
  • DON

    dissolved organic nitrogen

  •  
  • FP

    formation potential

  •  
  • HAAs

    haloacetic acids

  •  
  • HANs

    haloacetonitriles

  •  
  • HKs

    haloketones

  •  
  • HNMs

    halonitromethanes

  •  
  • N-DBPs

    nitrogenous disinfection by-products

  •  
  • NOM

    natural organic matter

  •  
  • RW

    raw water

  •  
  • TCAN

    trichloroacetonitrile

  •  
  • TCP

    1,1,1-trichloropropanone

  •  
  • THMs

    trihalomethanes

  •  
  • TOC

    total organic carbon

INTRODUCTION

Natural organic matter (NOM) is ubiquitously present in all water types including drinking water sources and has been recognized as containing the main precursors of disinfection by-products (DBPs) such as trihalomethanes (THMs) since the mid-1970s (Agbaba et al. 2012). It is important to note that among them, unregulated DBPs are considered as emerging DBPs (Yang & Zhang 2016). Therefore, more recently increasing attention has been paid to the nitrogenous (N)-DBPs detected in drinking water or chlorinated natural waters simulated in laboratory conditions, including haloacetonitriles (HANs), halonitromethanes (HNMs) and nitrosamines (Bond et al. 2011). Although the emerging N-DBPs have been detected at significantly lower concentrations compared with the carbonaceous (C)-DBPs such as THMs and haloacetic acids (HAAs), they are of emerging importance due to their cytotoxicity or genotoxicity, which is orders of magnitude higher than the regulated THMs and HAAs (Chiang et al. 2010; Zeng et al. 2016). The nitrogen involved in N-DBP formation can derive either from organic precursors such as dissolved organic nitrogen (DON), which reacts with the disinfectant (e.g. chlorine, chloramines) or from the disinfectant itself such as in the case of chloroamination (which is of high importance for N-nitrosodimethylamine formation). Free amino acids are amongst the DON components considered as significant precursors to N-DBPs (WHO 2011; Chu et al. 2016). Bond et al. (2011) in their review paper indicate that among the three major groups of N-DBPs, haloacetonitriles were detected at the highest concentrations, in the range of 1–36 μg/L. Measured concentrations are related to the water characteristics, bromine content, influence of wastewater, etc. In drinking water, the concentration of N-nitrosodimethylamine is typically observed in the low ng/L range while concentrations of chloropicrin (trichloronitromethane, TCNM) in drinking water are usually less than 5 μg/L (WHO 2011).

From the group of halogenated acetonitriles, guideline values for drinking water are set only for dibromoacetonitrile (DBAN) and dichloroacetonitrile (DCAN) (70 and 20 μg/L, respectively). Guideline values for bromochloroacetonitrile (BCAN), trichloroacetonitrile (TCAN) and TCNM have not been established due to the fact that available data are still inadequate to permit derivation of health-based guideline values (WHO 2011). Maximum permissible concentrations for haloacetonitriles are set for DBAN, DCAN and TCAN (100 μg/L, 90 μg/L and 1 μg/L) by Serbian national regulation (Official Gazette SRJ 42/98 and 44/99). Haloketones (HKs), which belong to the important unregulated volatile fraction, can be formed in ozone and chlorine treated water, including chloramines and chlorine dioxide treatment, or in water treated by combination of ozone and chlorine agents (Serrano et al. 2014).

The best way to control DBP formation is effective control and removal of their precursors before disinfection. It has been found that N-DBP precursors are hard to remove by conventional coagulation/sedimentation and filtration processes due to the poor removal of low molecular weight DON (Chu et al. 2016; Yimeng et al. 2017). Recently, a large number of drinking water treatment studies have investigated the use of advanced oxidation processes (AOPs) in order to control NOM content and minimize DBP formation. Recent studies have demonstrated that ozone pre-oxidation can decrease HAN formation, but can also increase HNM formation. Additionally, in the presence of hydroxyl radicals, formed during AOPs, greater formation of all DBP groups has been observed (De Vera et al. 2015). Molnar et al. (2012) showed that the application of ozonation and catalytic ozonation at pH 6 resulted in an increase in HAN formation potential compared with raw water. Chiang et al. (2010) indicated that application of ozonation prior to chlorination can cause a negative effect on TCNM and HK reduction. Chu et al. (2016) reported that UV did not have significant influence on HAN formation during subsequent chlorination, while UV/H2O2 AOPs could alter DBP formation. Srithep & Phattarapattamawong (2017) suggested that among the photo-based AOPs, UV/H2O2/O3 was the most effective process for removing HAN precursors, followed by UV/H2O2 and UV/O3.

Due to the different treatment conditions and variety in raw water characteristics, data about emerging DBPs, their associated health side-effects, formation and behaviour during different treatment stages is still relatively limited. Therefore, the aim of this study was to investigate the impact of photochemical AOPs using ozone, UV light and hydrogen peroxide (O3/UV, H2O2/UV and O3/H2O2/UV) on the formation potential (FP) of N-DBPs and HKs in water rich in NOM. Bromide incorporation into DBPs was also considered.

EXPERIMENT AND METHODS

Groundwater samples

The investigated groundwater originated from Maglić (Autonomous Province (AP) of Vojvodina, Republic of Serbia) and was sampled from a 110-m-deep well. AP Vojvodina is a northern region of the Republic of Serbia, geographically belonging to the southern part of the Pannonian Basin, together with Hungary, Romania, Slovakia and Slavonia in Croatia. In this area the quality of groundwater is highly affected by the presence of NOM and varies from aquifer to aquifer and also depends on the location. The investigated groundwater characteristics were as follows: pH 8.01 ± 0.10; total alkalinity 753 ± 24 mg CaCO3/L; turbidity 1.82 ± 0.09 NTU; electrical conductivity 1,407 ± 103 μS/cm; TOC 5.82 ± 0.39 mg C/L; total iron 0.24 ± 0.04 mg/L, manganese 0.074 ± 0.005 mg/L; bromide 623 ± 21 μg/L; chloride 19.2 ± 5.1 mg/L; nitrate <0.005 mg/L as N; nitrite <0.002 mg/L as N.

Chemicals

The analytical standard mix contained HANs (BCAN, DBAN, DCAN, TCAN), TCNM and HKs (1,1-dichloro-2-propanone – DCP and 1,1,1-trichloropropanone – TCP) in acetone in concentrations of 2,000 μg/mL were purchased from Supelco. The internal standard used was 4-bromofluorobenzene (2,000 μg/mL in methanol, Supelco). All solvents (J. T. Baker) were for organic residue analysis. Hydrogen peroxide (30% w/w, reagent grade) was purchased from POCH S.A.

Water treatment

UV photolysis and UV AOPs were performed in a stainless steel photochemical reactor with a quartz reaction vessel at the reactor centre equipped with a low pressure mercury lamp (Philips TUV 16 W) (more details are given in Molnar et al. 2015). The raw water (RW) was ozonated in a 2 L glass column (85 mm diameter) and ozone was generated electrochemically by an Argentox ozone generator. For the O3/UV treatment, ozonated water was immediately subject to the UV treatment in the photochemical reactor. For the H2O2/UV treatment, hydrogen peroxide solution was added to the water prior to UV irradiation. For the O3/H2O2/UV treatment, hydrogen peroxide solution was added to ozonated water which was then subject to UV light. A more detailed procedure for the applied treatments is given in Agbaba et al. (2016). All experiments were carried out in duplicate and the results reported as average values with standard deviation. Treatment lines were as follows: O3 – ozonation (0.5 mg O3/mg dissolved organic carbon (DOC)); UV1 – photolysis (600 mJ/cm2); UV2 – photolysis (3,000 mJ/cm2); O3/UV1 (0.5 mg O3/mg DOC; 600 mJ/cm2); O3/UV2 (0.5 mg O3/mg DOC; 3,000 mJ/cm2); O3/H2O2/UV1 (0.5 mg O3/mg DOC; 10 mg H2O2/L; 600 mJ/cm2); O3/H2O2/UV2 (0.5 mg O3/mg DOC; 10 mg H2O2/L; 3,000 mJ/cm2); H2O2/UV1 (10 mg H2O2/L; 600 mJ/cm2); H2O2/UV2 (10 mg H2O2/L; 3,000 mJ/cm2).

Analytical methods

HANFP, TCNMFP and HKFP were determined according to the standard method for measurement of THMFP (APHA 2012). Emerging DBPs were determined by liquid–liquid extraction using methyl tert-butyl ether followed by analysis by GC/μECD (Agilent 6890N), according to USEPA method 551.1. The specific emerging DBP FPs were calculated as the ratios between the DBP FPs and the total organic carbon (TOC) content presented in our previous study (Agbaba et al. 2016).

Bromide and chloride anion content was determined using a Dionex ICS-3000 Ion Chromatography System according to DIONEX Application Note 154. The pH and turbidity measurements were carried out using the InoLab pH/ION 735 and Hanna HI 93703 portable instruments. Electrical conductivity of water was measured with the aid of a Hanna model HI 933000 field conductometer. Water alkalinity was measured according to the standard methods (APHA 2012). Iron and manganese concentrations were determined by graphite furnace atomic absorption spectroscopy (Perkin Elmer AAnalyst 700) according to USEPA Method 7010. Inorganic nitrogen substances (nitrite and nitrate) were analysed according to the SRPS ISO 7890-3:1994 and SRPS EN 26777:2009 methods.

RESULTS AND DISCUSSION

Effects of photochemical AOPs on the emerging DBP formation potential

Our previous work has shown that the investigated groundwater is rich in mostly hydrophobic natural organic matter which is characterized by a high content of fulvic acids (75%) relative to the 25% hydrophilic acid and non-acidic structures. Total NOM content according to the TOC value and UV absorbance at 254 nm was 5.82 ± 0.39 mg C/L and 0.231 ± 0.05 cm−1, respectively. The NOM present is highly reactive towards forming C-DBPs such as THMs and HAAs (305 ± 44 μg/L THMFP and 348 ± 53 μg/L HAAFP) (Agbaba et al. 2016). This paper presents a continuation of research in which the impact of photochemical AOPs on the emerging DBP formation is addressed. Control experiments using ozone and photolysis alone were also conducted. When it comes to the N-DBPs, HAN precursor contents (9.83 ± 0.59 μg/L HANFP) were significantly lower than the C-DBPs. The dominant HAN compound formed during chlorination of raw water is DCAN (6.76 ± 0.35 μg/L DCANFP), followed by DBAN (3.07 ± 0.24 μg/L DBANFP), while precursors of TCAN and BCAN were not detected. TCNM precursors, as representative of HNMs, were also not detected (<0.400 μg/L). It can be supposed that amino acids, the main constituents of DON in drinking water sources, are responsible for the formation of HANs. Other authors have reported similar contents of HAN precursors in chlorinated water samples (Krasner et al. 2006; Zeng et al. 2016). Precursors of HKs, particularly TCP and DCP were also not detected in raw water samples (<0.300 μg/L).

The influence of UV photolysis, ozone oxidation and AOPs involving O3, UV and H2O2 (O3/UV, H2O2/UV and O3/H2O2/UV) on the HAN precursor content is presented in Table 1. Treatment by ozone alone decreased HANFP by 71% whereby the precursors of DCAN were dominantly reduced. During O3-based AOPs (O3/UV and O3/H2O2/UV) HANFP varied significantly (3.49 ± 0.14–10.1 ± 0.82 μg/L). Among the ozone-based AOPs, only the O3/UV process (at the lower UV dose) was able to oxidize HAN precursors, reducing the HANFP by 64% compared with the raw water. However, it is obvious that the addition of H2O2 and UV light did not have a further positive influence on decreasing HANFP compared with the ozone oxidation alone. The H2O2/UV process with a lower UV dose decomposed HAN precursors by 65% similar to ozone alone. During the UV photolysis and UV/H2O2 processes, increasing the UV dose doubled the HANFP. In all the treatments, of the haloacetonitriles, only DCAN and DBAN precursors were detected in water. None of the investigated treatments, UV photolysis, ozone oxidation or AOPs, had any effect on the formation of TCNM precursors (<0.400 μg/L). Additionally, it is worth noting that the HANFP levels determined, which were used to assess the level of HANs in a worst-case scenario, were below Serbian national regulation and WHO guideline values.

Table 1

Emerging DBP FPs (actual and normalized to the measured TOC) in raw and treated water samples

Water treatment DCANFP DBANFP HANFPa TCNMFP HKFPb 
μg/L 
RW 6.76 ± 0.35 3.07 ± 0.24 9.83 ± 0.59 <0.400 <0.300 
O3 0.670 ± 0.20 2.18 ± 0.51 2.85 ± 0.56 <0.400 0.600 ± 0.15 
UV1 6.76 ± 1.78 1.46 ± 0.47 8.22 ± 1.05 <0.400 <0.300 
UV2 12.1 ± 1.63 5.06 ± 1.34 17.2 ± 0.28 <0.400 <0.300 
O3/UV1 0.940 ± 0.18 2.55 ± 0.33 3.49 ± 0.14 <0.400 0.600 ± 0.12 
O3/UV2 6.90 ± 1.07 1.58 ± 0.27 8.48 ± 0.81 <0.400 <0.300 
O3/H2O2/UV1 8.21 ± 0.37 1.85 ± 0.25 10.1 ± 0.82 <0.400 <0.300 
O3/H2O2/UV2 4.65 ± 0.64 2.44 ± 0.81 7.09 ± 1.10 <0.400 <0.300 
H2O2/UV1 1.08 ± 0.27 2.39 ± 0.54 3.47 ± 0.66 <0.400 2.01 ± 0.37 
H2O2/UV2 13.6 ± 2.08 8.61 ± 1.94 22.2 ± 0.14 <0.400 0.93 ± 0.21 
μg/mg TOC 
RW 1.16 ± 0.10 0.528 ± 0.06 1.69 ± 0.15 nc nc 
O3 0.124 ± 0.04 0.403 ± 0.11 0.527 ± 0.12 nc 0.110 ± 0.02 
UV1 1.27 ± 0.37 0.272 ± 0.08 1.54 ± 0.24 nc nc 
UV2 2.18 ± 0.34 0.903 ± 0.22 3.08 ± 0.11 nc nc 
O3/UV1 0.185 ± 0.03 0.506 ± 0.09 0.691 ± 0.06 nc 0.118 ± 0.02 
O3/UV2 1.36 ± 0.16 0.312 ± 0.06 1.67 ± 0.10 nc nc 
O3/H2O2/UV1 1.55 ± 0.02 0.350 ± 0.06 1.90 ± 0.10 nc nc 
O3/H2O2/UV2 0.891 ± 0.10 0.471 ± 0.17 1.36 ± 0.25 nc nc 
H2O2/UV1 0.206 ± 0.05 0.455 ± 0.09 0.661 ± 0.11 nc 0.381 ± 0.06 
H2O2/UV2 2.76 ± 0.52 1.74 ± 0.33 4.50 ± 0.18 nc 0.188 ± 0.04 
Water treatment DCANFP DBANFP HANFPa TCNMFP HKFPb 
μg/L 
RW 6.76 ± 0.35 3.07 ± 0.24 9.83 ± 0.59 <0.400 <0.300 
O3 0.670 ± 0.20 2.18 ± 0.51 2.85 ± 0.56 <0.400 0.600 ± 0.15 
UV1 6.76 ± 1.78 1.46 ± 0.47 8.22 ± 1.05 <0.400 <0.300 
UV2 12.1 ± 1.63 5.06 ± 1.34 17.2 ± 0.28 <0.400 <0.300 
O3/UV1 0.940 ± 0.18 2.55 ± 0.33 3.49 ± 0.14 <0.400 0.600 ± 0.12 
O3/UV2 6.90 ± 1.07 1.58 ± 0.27 8.48 ± 0.81 <0.400 <0.300 
O3/H2O2/UV1 8.21 ± 0.37 1.85 ± 0.25 10.1 ± 0.82 <0.400 <0.300 
O3/H2O2/UV2 4.65 ± 0.64 2.44 ± 0.81 7.09 ± 1.10 <0.400 <0.300 
H2O2/UV1 1.08 ± 0.27 2.39 ± 0.54 3.47 ± 0.66 <0.400 2.01 ± 0.37 
H2O2/UV2 13.6 ± 2.08 8.61 ± 1.94 22.2 ± 0.14 <0.400 0.93 ± 0.21 
μg/mg TOC 
RW 1.16 ± 0.10 0.528 ± 0.06 1.69 ± 0.15 nc nc 
O3 0.124 ± 0.04 0.403 ± 0.11 0.527 ± 0.12 nc 0.110 ± 0.02 
UV1 1.27 ± 0.37 0.272 ± 0.08 1.54 ± 0.24 nc nc 
UV2 2.18 ± 0.34 0.903 ± 0.22 3.08 ± 0.11 nc nc 
O3/UV1 0.185 ± 0.03 0.506 ± 0.09 0.691 ± 0.06 nc 0.118 ± 0.02 
O3/UV2 1.36 ± 0.16 0.312 ± 0.06 1.67 ± 0.10 nc nc 
O3/H2O2/UV1 1.55 ± 0.02 0.350 ± 0.06 1.90 ± 0.10 nc nc 
O3/H2O2/UV2 0.891 ± 0.10 0.471 ± 0.17 1.36 ± 0.25 nc nc 
H2O2/UV1 0.206 ± 0.05 0.455 ± 0.09 0.661 ± 0.11 nc 0.381 ± 0.06 
H2O2/UV2 2.76 ± 0.52 1.74 ± 0.33 4.50 ± 0.18 nc 0.188 ± 0.04 

aTCANFP < 0.700 μg/L; BCANFP < 0.300 μg/L; bTCPFP < 0.300 μg/L; nc – not calculated (corresponding DBP FPs were below the method detection limit).

De Vera et al. (2015) also demonstrated that application of ozone with a similar dose (0.75 mg/mg TOC) can decrease total HAN formation in water with high TOC content. Mitch et al. (2009) demonstrated that ozone can have extreme effects on HNM formation when followed by chlorination, reporting an increase in median chloropicrin occurrence of 226%, which was not observed in our study. According to Chu et al. (2016), UV photolysis alone at a dose up to 585 mJ/cm2 followed by chlorination for selected amino acids did not significantly alter HAN formation, while UV/H2O2 may alter DBP formation. Srithep & Phattarapattamawong (2017) showed that in water with lower DOC content than in our study, UV/H2O2/O3 was the most effective process for removing HAN precursors, followed by UV/H2O2 and UV/O3. The better control of HAN precursors in our study could be attributed to the different characteristics of the NOM and DON, which are specific for every water body.

Concerning HKFP, application of the ozone and O3/UV processes led to the generation of DCP precursors (0.600 μg/L). HKs can be formed from NOM structure, as a consequence of ozone oxidation, which may explain the obtained results. The DCP formed could be further oxidized to TCP, which was not observed in our study. Additionally, no relationship between UV dose and H2O2 addition in ozone-based AOPs was observed. Enhanced formation of HK precursors was observed as a consequence of UV/H2O2 process application (0.93 ± 0.21 to 2.01 ± 0.37 μg/L DCPFP) indicating that NOM oxidation favoured by free radicals can also promote HK formation. UV photolysis alone has no influence on the formation of HK precursors. None of the investigated treatments had any influence on TCP formation (Table 1). Generally low values for HKFP are somewhat expected because haloketones are important intermediate DBPs which are rapidly formed as soon as the water is chlorinated and may be finally converted to THMs or HAAs via hydrolysis or reactions with residual chlorine (Fu et al. 2017). The highest formation of HKs was observed in UV/H2O2-treated water with higher UV dose, where in our previous study the highest THMFP was also observed (Agbaba et al. 2016).

DBP FPs normalized to the measured TOC can provide a better insight into the variations in the reactivity of organic matter towards DBP formation than actual FPs (Table 1). HANFPs, when normalized by TOC content (presented in our previous study by Agbaba et al. 2016) follow the same trend. The ozone and H2O2/UV process with a lower UV dose maximally reduced the reactivity of HAN precursors up to 70%, while the highest reactivity toward HAN formation was observed in water treated by the H2O2/UV process with a high UV dose. HKFP normalized to the TOC was even lower than HANFP (0.110 ± 0.02–0.381 ± 0.06 μg/mg TOC) indicating lower reactivity of residual NOM after oxidation treatments toward HK formation relative to HAN formation.

Insight into distribution of DBPs and bromide incorporation

The distributions of individual DBP groups formed during chlorination (THMFP, HAAFP, HANFP, THNMFP and HKFP) in the raw water and after the investigated treatments are presented in Figure 1. The bromide incorporations for these groups are presented in Figure 2. In both Figures 1 and 2, previously published data (Agbaba et al. 2016) relating to THMFP and HAAFP was used for the calculations. Emerging DBP precursors were only 1.5% of the total sum of DBPs formed in DBP FPs tests in the raw water samples, increasing slightly during H2O2/UV treatment and reaching a maximum of 3.1% after the H2O2/UV process with higher UV dose. The lowest share of emerging DBPs (only 0.4%) was observed after chlorination of ozonated water. In all water samples THM and HAA were dominantly formed during subsequent chlorination (over 97% of the DBPs investigated).

Figure 1

Percent distribution of investigated DBP FPs in raw and treated water samples.

Figure 1

Percent distribution of investigated DBP FPs in raw and treated water samples.

Figure 2

Percentage of bromide incorporation for THMs, HAAs and HANs in raw and treated water samples.

Figure 2

Percentage of bromide incorporation for THMs, HAAs and HANs in raw and treated water samples.

Taking into account the high bromide ion content (623 ± 21 μg/L) of the investigated groundwater, there is an increased risk of brominated DBP formation through Br- oxidation to bromine and further reaction with organic precursors in water. The percentage of bromide ion incorporation into brominated DBPs was calculated according to the equations given by Sohn et al. (2006). It was found that the applied treatments have no significant influence on the changes in percentage of bromide incorporation into DBPs compared with the raw water (Figure 2). The influence of the applied treatments on bromide incorporation was most evident for emerging DBPs, particularly HANs, where bromide incorporation increases up to three times after H2O2/UV treatment but only with the higher UV dose. Similar results were obtained for THMs and HAAs, which led to a 1.5-times increase in bromine incorporation. It may be supposed that in the H2O2/UV system with a high UV dose and in the absence of ozone, highly reactive and unselective hydroxyl radicals are the main oxidation species. Through the contribution of photolysis and hydroxyl radical attack, NOM oxidation occurs, forming new, more hydrophilic species which are prone to further bromide incorporation. Additionally, UV photolysis using the higher UV dose of 3,000 mJ/cm2 almost doubled bromine incorporation, while other treatments generally have no further influence on changing bromine incorporation compared with the raw water.

Results were also correlated between N-DBP FPs and: (a) total organic matter according to the TOC, UV254 and SUVA and (b) specific DBP precursors: THMFP, HAAFP and the total aldehydes and carboxylic acids (Agbaba et al. 2016). There was no correlation between HANFP and the total organic matter parameters. Additionally, the absence of a correlation between N-DBPs and the C-DBPs and oxygen containing DBPs indicates the completely different nature of their precursors. The formation of HAN before and after the treatments could be mostly attributed to the transformation of DON containing combined amino acids such as oligopeptides and proteins as the main precursors of nitrogenous DBPs. N-DBP formation was significantly lower than that of C-DBPs, which was expected as DON contents are usually lower than dissolved organic carbon contents (Chu et al. 2016). The influence of wastewater effluents on the DON content and subsequent N-DBP formation may be excluded in the case of this type of water. TCNM precursors were not detected after any of the oxidation treatments, indicating a different formation mechanism for halonitromethane and haloacetonitrile in water.

For evaluation of the optimal treatment conditions for the investigated groundwater, it is necessary to consider bromate formation, which is classified by IARC in group 2B as possibly carcinogenic to humans. Previous study has shown that bromate formation in concentrations above guideline values (10 μg/L) is significant in ozone-based advanced oxidation processes. Thus, despite the fact that maximal reduction of THMFP and HAAFP was achieved by the O3/H2O2/UV process, in ozone-based processes there is an increased risk of bromate occurrence. Thus, the H2O2/UV process after optimizing treatment conditions (primarily UV dose) could be a possible choice to treat such challenging groundwater and to avoid bromate formation. Application of the H2O2/UV process would also reduce the cost of ozone generation. In the Strategy of water supply and water protection in AP Vojvodina (Official Gazette APV No. 1/2010) it is indicated that this type of water is the most difficult for processing to meet relevant norms for drinking water (Official Gazette SRJ 42/98 and 44/99). The quality of raw water requires, apart from the degasification/aeration and oxidation treatment used in the first stage of treatment, additional application of coagulation/flocculation and filtration (sand filtration and/or granulated activated carbon filtration) followed by membrane separation for the final phase of treatment and disinfection. Water supply strategy also emphasizes that for such complex technology of preparation it is necessary to carry out investigations using a pilot plant in order to optimize processes which would enable savings in the investment and exploitation costs of the plant. Thus, results of the conducted investigations are valuable for further treatment line investigations. Additionally, based on the available data in the Republic of Serbia there is generally a lack of data regarding the emerging DBP formation in drinking water. Besides that, apart from ozonation, UV-based advanced oxidation processes are still not applied in large water plants in the Republic of Serbia and so the conducted investigation contributes to understanding the possibility of using these processes with their benefits and disadvantages for treatment of water with similar characteristics.

CONCLUSION

The influence of photochemical advanced oxidation processes (O3/UV, H2O2/UV and O3/H2O2/UV) on the formation potential of emerging disinfection by-products in NOM rich water was investigated. The content of HAN precursors in the raw water was significantly lower compared with the C-DBPs, representing just 1.5% compared with the total DBPs formed, while TCNM and HK precursors were not detected. Treatments with ozone and H2O2/UV with lower UV dose were shown to be the most effective in reducing HANFP with no significant influence on the changes in percentage of bromide incorporation into DBPs compared with the raw water. None of the oxidation treatments led to TCNM formation, indicating a different formation mechanism for halonitromethane and haloacetonitrile in water. In general, the absence of a correlation between the emerging DBPs with C-DBPs and the oxygen containing DBPs (previously investigated) demonstrates how the behaviour of different NOM functional groups strongly depends on the oxidation process applied, resulting in widely different reactivity towards chlorine and promoting the formation of different groups of DBPs. Considering all investigated oxidation/disinfection by-product formation and available guideline values, H2O2/UV could be a possible choice to treat such challenging groundwater and avoid bromate formation. Further pilot-scale investigations are needed to define the combination of optimal processes and their conditions.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the support of the Ministry of Education, Science and Technological Development of the Republic of Serbia (grant no. 172028).

REFERENCES

REFERENCES
Agbaba
J.
,
Dalmacija
B.
,
Molnar
J.
,
Krčmar
D.
,
Tubić
A.
&
Watson
M.
2012
Characterization of natural organic matter in the groundwater at selected sites in Vojvodina
.
Water and Sanitary Technology
42
(
2
),
43
52
.
Agbaba
J.
,
Molnar Jazić
J.
,
Tubić
A.
,
Watson
M.
,
Maletić
S.
,
Kragulj Isakovski
M.
&
Dalmacija
B.
2016
Oxidation of natural organic matter with processes involving O3, H2O2 and UV light: formation of oxidation and disinfection by-products
.
RSC Advances
6
(
89
),
86212
86219
.
AWWA-APHA-WEF
2012
Standard Methods for the Examination of Water and Wastewater
,
20th edn
.
American Public Health Association/American Water Works Association/Water Environment Federation
,
Washington, DC, USA
.
Bond
T.
,
Huang
J.
,
Templeton
M. R.
&
Graham
N.
2011
Occurrence and control of nitrogenous disinfection by-products in drinking water – a review
.
Water Research
45
(
15
),
4341
4354
.
De Vera
G. A.
,
Stalter
D.
,
Gernjak
W.
,
Weinberg
H. S.
,
Keller
J.
&
Farré
M. J.
2015
Towards reducing DBP formation potential of drinking water by favouring direct ozone over hydroxyl radical reactions during ozonation
.
Water Research
87
,
49
58
.
Fu
J.
,
Lee
W.-N.
,
Coleman
C.
,
Nowack
K.
,
Carter
J.
&
Huang
C.-H.
2017
Removal of disinfection byproduct (DBP) precursors in water by two-stage biofiltration treatment
.
Water Research
123
,
224
235
.
Krasner
S. W.
,
Weinberg
H. S.
,
Richardson
S. D.
,
Pastor
S. J.
,
Chinn
R.
,
Sclimenti
M. J.
,
Onstad
G. D.
&
Thruston
A. D.
2006
Occurrence of a new generation of disinfection byproducts
.
Environmental Science & Technology
40
(
23
),
7175
7185
.
Mitch
W. A.
,
Krasner
S. W.
,
Westerhoff
P.
&
Dotson
A.
2009
Occurrence and Formation of Nitrogenous Disinfection By-Products
.
Water Research Foundation, Denver
,
CO, USA
.
Molnar
J. J.
,
Agbaba
J. R.
,
Dalmacija
B. D.
,
Klašnja
M. T.
,
Dalmacija
M. B.
&
Kragulj
M. M.
2012
A comparative study of the effects of ozonation and TiO2-catalyzed ozonation on the selected chlorine disinfection by-product precursor content and structure
.
Science of the Total Environment
425
,
169
175
.
Molnar
J.
,
Agbaba
J.
,
Tubić
A.
,
Watson
M.
,
Kragulj
M.
,
Rončević
S.
&
Dalmacija
B.
2015
The effects of ultraviolet/H2O2 advanced oxidation on the content and characteristics of groundwater natural organic matter
.
Water Science & Technology: Water Supply
15
(
1
),
34
41
.
Official Gazette APV
,
No. 1/2010, Strategy of water supply and water protection in Vojvodina
.
Official Gazette SRJ 42/98 and 44/99
1998
Serbian Regulation for Drinking Water
.
Serrano
M.
,
Silva
M.
&
Gallego
M.
2014
Fast and ‘green’ method for the analytical monitoring of haloketones in treated water
.
Journal of Chromatography A
1358
,
232
239
.
Sohn
J.
,
Amy
G.
&
Yoon
Y.
2006
Bromide ion incorporation into brominated disinfection by-products
.
Water, Air, and Soil Pollution
174
,
265
277
.
SRPS EN 26777: Water quality – Determination of nitrite – Molecular absorption spectrometric method, 2009
.
SRPS ISO 7890-3: Water quality – Determination of nitrate – Part 3: Spectrometric method using sulfosalicylic acid, 1994.
USEPA
1995
Method 551.1: Determination of chlorination disinfection byproducts, chlorinated solvents, and halogenated pesticides/herbicides in drinking water, (Revision 1). US Environmental Protection Agency, Cincinnati, OH, USA
.
USEPA
2007
Method 7010: Graphite furnace atomic absorption spectrophotometry. US Environmental Protection Agency, Cincinnati. OH, USA
.
WHO
2011
Guidelines for Drinking-Water Quality
,
4th edn
.
WHO Press
,
Geneva
,
Switzerland
.
Yang
M.
&
Zhang
X.
2016
Current trends in the analysis and identification of emerging disinfection byproducts
.
Trends in Environmental Analytical Chemistry
10
,
24
34
.
Yimeng
Z.
,
Wenhai
C.
,
Ting
X.
,
Daqiang
Y.
,
Bin
X.
,
Pan
L.
&
Na
A.
2017
Impact of pre-oxidation using H2O2 and ultraviolet/H2O2 on disinfection byproducts generated from chlor(am)ination of chloramphenicol
.
Chemical Engineering Journal
317
,
112
118
.