Water Air Soil Pollut (2017) 228: 229 DOI 10.1007/s11270-017-3418-1

Formation of Aldehydes and Carboxylic Acids in Humic Acid Ozonation Xin Zhong & Chongwei Cui & Shuili Yu

Received: 3 March 2017 / Accepted: 24 May 2017 / Published online: 2 June 2017 # Springer International Publishing Switzerland 2017

Abstract The purpose of this study was to determine the different kinds and concentrations of intermediates, and investigate on the effects of contact time and ozone (O3) doses on the removal of humic acid (HA), which is served as the main disinfection by-product (DBP) precursor. Based on that, the knowledge gap of DBPs generated was made up. The results showed that HA was the major precursor material for aldehydes and carboxylic acids. The concentrations of aldehydes increased as contact time and O3 doses, and reached up maximum at 2~10 min but approached a plateau at the higher O3 doses. The concentrations of formic and acetic acids increased as contact time and O3 doses. However, aromatic acids, including protocatechuic, 3hydroxybenzoic, and benzoic acids, declined rapidly at longer reaction time and higher O3 doses. It was worth mentioning that aromatic acids had been rarely reported. Besides, a possible formation pathway was proposed: X. Zhong : C. Cui (*) School of Municipal and Environmental Engineering, Harbin Institute of Technology, Harbin 150090, China e-mail: [email protected] X. Zhong e-mail: [email protected] S. Yu School of Environmental Science and Engineering, Tongji University, Shanghai 200433, China

Present Address: C. Cui No.73, Huanghe Road, Nangang District, Harbin City, Heilongjiang Province, China

(a) HA was degraded into fulvic acid (FA)-like compounds; (b) FA-like compounds were further converted into aromatic acids; (c) aromatic acids were transformed into low-molecular-weight organic matters; (d) chlorine reacted with aldehydes and/or carboxylic acids by addition, hydrolysis, and decarbonylation reactions, leading to DBP formation. Furthermore, not only HA were the main DBPs precursors, but also the oxidation intermediates of HA could be the DBPs precursors, and they gave a certain amount of DBPs. Consequently, aldehydes and carboxylic acids should be under control in drinking water treatment plants. Keywords Ozone . Carboxylic acids . Aldehydes . Chloroform . Dichloroacetic acid

1 Introduction Natural organic matter (NOM) is present in all surface, ground, and soil waters. NOM consists of a range of compounds, from largely aliphatic to highly colored and aromatic, as well as highly charged to uncharged, having also a wide variety of chemical compositions and molecular sizes (Matilainen and Sillanpää, 2010). Humic acid (HA), a major component of NOM, is a heterogeneous macromolecule widely dispersed in natural waters, which display significant resistance to biodegradation (Miao and Tao 2008). The structure of HA may consist of a skeleton of alkyl/aromatic units cross-linked by a variety of functional groups such as carboxylic, phenolic, and alcoholic hydroxyls, as well as ketone and

229 Page 2 of 11

quinine groups, and they are very resistant to biodegradation (Gaffney et al. 1996; Leenheer and Croué 2003). HA, being more highly aromatic, becomes insoluble when the carboxylate groups are protonated at low pH values. This structure allows the humic materials to function as surfactants, with the ability to bind both hydrophobic and hydrophilic materials. This function makes humic and fulvic materials effective agents in transporting both organic and inorganic contaminants in the environment. HAs are precursors of disinfection by-products (DBPs) such as trihalomethane (THMs) and haloacetic acids (HAAs) that are formed after water chlorination, which presents a major problem in drinking water treatment (Miao and Tao 2008; Ratpukdi et al. 2010; Zhao et al. 2013). Ozonation has been used as an effective disinfectant and oxidant in drinking water and wastewater treatment (Zimmermann et al. 2011). Ozone (O3) has many benefits such as microbial disinfection, coagulation aid, and taste and odor control. However, several classes of by-products form in the process. These by-products include bromate and various organics such as aldehydes, ketones, and carboxylic acids (Huang et al. 2005; Wert et al. 2008; Hu et al. 2016; Papageorgiou et al. 2017). These carbonyl compounds and low-molecular-weight DBPs (<300 Da) formed upon the NOM ozonation that is present in drinking water and wastewater (Papageorgiou et al. 2014; Liu et al. 2015). The formation of ozonation by-products depends on variables such as contact time, O3 dose,.OH radical, temperature, pH, and organic matter present in the water (Tripathi et al. 2011; Samadi et al. 2015). It has been reported that carbonyl compounds represent about 30% of the ozonation by-products (Richardson 2003). These ozonation by-products are relatively volatile, polar, and reactive as well as have potential adverse health problems. For example, formaldehyde, acetaldehyde, glyoxal, and methyl glyoxal are considered as posing potential risk and are classified as prioritized emerging DBPs in drinking water regarding their potential health impact (Richadrson et al. 2007; McGwin et al. 2010). Up to now, however, only formaldehyde and chloral hydrate have been stipulated. WHO (World Health Organization) and National Drinking Water Standards (GB5749-2006) recommended guideline of 900 mg/L for formaldehyde and 10 μg/L for chloral hydrate in drinking water. Besides, carbonyl

Water Air Soil Pollut (2017) 228: 229

compounds may also be responsible for the undesirable odor in the treated waters because of their very low odor detection thresholds, e.g., 4 μg/L for acetaldehyde (Dąbrowska and Nawrocki 2013). In addition, carbonyl compounds usually served as carbon source for bacteria, potentially causing re-growth problems in distribution systems, especially in the cases of with long residence times. Therefore, monitoring of carbonyl compounds is important and beneficial for the optimization of operations of water/wastewater treatment facilities and achievement of biologically stable quality of treated waters. Insomuch as, these compounds are directly related to public health, it became evident that there is need to study the formation of by-products from the reaction between O3 and NOM. Additionally, O3 can be rapidly consumed under typical drinking water conditions. It cannot be used to maintain a disinfectant residual throughout the distribution system, and most ozonation treatment scenarios include either chlorine (Cl2) or chloramine as a final disinfectant (Hua and Reckhow 2013). Hence, it needs more attention that the impact of O3 on the formation of DBPs during subsequent chlorination. There is incomplete information about NOM oxidation intermediates. The main identified oxidation intermediates are aldehydes, formic, acetic, and oxalic acids as well as several ketoacids (Nawrocki et al. 2003). As far as aromatic acids, they are rarely identified in water as well as their influence on DBPs, owing to lack of proper analytical methods. Aromatic acids should be transformed into low-polar, and stable derivates apply to gas chromatography (GC) detection. In order to avoid the complicated derivatization procedure, our team employed solid phase extraction-ultra high-performance liquid chromatography (SPE-UPLC) to detect aromatic acids (Zhong et al. 2016). Moreover, our previous studies showed that carbonyl compounds were widespread in raw water and each process unit of DWTPs using O3 disinfectant (Zhong et al. 2017a, b). In this study, the influence of pre-ozonation on the structural transformation of HAs molecules as well as the formation of DBPs was systematically investigated. On the basis of the SPE-UPLC method of independent establishment, our research not only provides information about oxidation intermediates and DBPs in ozonation-chlorination processes but also has broad applicability for optimizing water treatment by reducing DBP formation.

Water Air Soil Pollut (2017) 228: 229

Page 3 of 11 229

2 Materials and Methods 2.1 Materials Standard of HA was obtained from Sigma-Aldrich (Shanghai, China). Its molecular formula was C 9 H 9 NO 6 , and its relative molecular mass was 227.1699. The analytical standard containing 15 carbonyl compounds (formaldehyde, acetaldehyde, propanal, butanal, pentanal, hexanal, cyclohexanone, crotonaldehyde, heptanal, octanal, benzaldehyde, nonanal, decanal, glyoxal, and methyl glyoxal) and the derivatization agent O-(2,3,4,5,6-pentafluoro-benzyl) hydroxylamine (PFBOA) were purchased from AccStandard (New Haven, USA). HPLC grade n-hexane (Anpel, Shanghai, China) was used as a solvent for the liquid-liquid extraction. Organic-free water was provided by the Synergy UV-Ultrapure Water System (Millipore, Molsheim, France). Standards of the carboxylic acids (>95% purity), f or m i c , a c et i c , f u m a r i c , p r o t o ca t ec h u i c , 3 hydroxybenzoic, benzoic acids, and standard of dichloroacetic acid (DCAA) were purchased from J&K (Beijing, China). Stock solutions of the individual acids (10 g/L) were prepared in purified water. All these solutions were stored at 4 °C. Potassium dihydrogenphosphate, hydrochloric acid, sodium hydroxide, sodium hypochlorite, and orthophosphoric acid were analytical reagent and supplied by Shanghai (Shanghai, China). LiChrolut EN (particle size 40–120 μm) was purchased from Merck (Darmstadt, Germany). Silica-reverse phase sorbent with octadecyl functional groups (Supelclean ENVI-18) was supplied by Supelco (Bellefonte, PA, USA). 2.2 Experimental Procedures The HA solution was prepared by dissolving 1 g HAs in 1000 mL of 0.01 mol/L NaOH solution with vigorous

mixing for 24 h, readjusting pH to 7, and filtering with 0.45 μm cellulose acetate filters. Working solution at 10 mg/L concentration was obtained by water dilution. All ozonation of HA experiments was operated on 500-mL SIMAX bottles fitted with a magnetic stirring bar. O3 was generated from pure oxygen (≥99.2% purity) by COM-AD-01 O3 generator (4 g/h, Anseros, Germany) and transferred immediately into the ultrapure water using a diffuser placed at the bottom of the reactor. The O3 concentration of the stock solution (20 mg/L) was determined using the direct UV absorbance method at 258 nm with a molar absorptivity of 2950/M/cm. The ozonation reactions were terminated by sodium nitrite (J&K, Beijing, China), and 10 μL sodium nitrite solution (7 g/L) was added to each 100 mL O3 reaction solution (Hammes et al. 2006; Gunten and Ramseier 2009). The ozonated samples were stored at 4 °C for no more than 24 h before the chlorination experiments. The experimental matrix encompassed the following ranges of conditions: (a) time = 1–30 min, O3 dose = 1.0 mg/L, pH of 7; and (b) time = 10 min, O3 dose = 0.5~5 mg/L, pH of 7. Samples were collected for subsequent analyses of aldehydes, short-chain acids, and aromatic acids. Dissolved organic carbon (DOC) concentration and UV absorbance at 254 nm (UV254) were analyzed. Specific UV absorbance (SUVA) was calculated as UV254 divided by DOC. The properties of ozonated HA waters are summarized in Tables 1 and 2. Chlorination was conducted on ozonated samples using 100-mL chlorine-free bottles. Cl 2 doses (10 ~ 13 mg/L) were chosen to ensure that a substantial Cl2 residual was present after incubation for 24 h so that formation reactions would not be Cl2 limited. After being dosed with Cl2, samples were stored headspacefree at pH of 7, 25 ± 1 °C in the dark for 24 h. Sodium thiosulfate was used to quench the residual Cl2. Samples were collected headspace-free in 40-mL glass vials with polypropylene screw caps and Teflon-lined septa for subsequent analyses of THMs and HAAs.

Table 1 The properties of ozonated HA waters at a different time (O3 dose = 1 mg/L, pH = 7) Parameters

Time 1 min

2 min

5 min

10 min

20 min

30 min

DOC(mg/L)

2.673

2.643

2.609

2.571

2.562

2.553

UV254 (/cm)

0.229

0.225

0.223

0.218

0.216

0.215

SUVA (L/m/mg)

8.567

8.513

8.547

8.479

8.431

8.421

229 Page 4 of 11

Water Air Soil Pollut (2017) 228: 229

Table 2 The properties of ozonated HA waters at different O3 dose (t = 10 min, pH = 7) Parameters

O3 dose 0.5 mg/L

1 mg/L

2 mg/L

3 mg/L

4 mg/L

5 mg/L

DOC (mg/L)

2.642

2.571

2.332

2.181

2.015

1.923

UV254 (/cm)

0.247

0.218

0.194

0.164

0.143

0.126

SUVA (L/m/mg)

9.349

8.479

8.319

7.519

7.09

6.552

2.3 Analytic Methods DOC concentrations were measured using a total organic carbon analyzer (OI, Aurora1030) according to Standard Method 5310B (APHA 1998). UV254 was measured by a UV-visible spectrophotometer (Shimadzu, UV-1800). The EPA 556 method and EPA 552.2 method were individually employed for the determination of aldehydes and DCAA in water samples (EPA 1995, 1998). Aldehydes and DCAA were analyzed by 7890B gas chromatography (Agilent Technologies, Palo Alto, USA) equipped with an electron capture detector (ECD). For the separation, a DB-5MS fused silica capillary column (30 m × 0.25 mm I.D. × 0.25 μm film thickness, Agilent Technologies, Bellefonte, PA, USA) was used. Helium (1 mL/min) was the carrier gas and nitrogen (30 mL/min) the detector make-up gas. The oven temperature program for the aldehydes was as follows: 50 °C for 1 min, programmed at 4 °C/min to 220 °C, programmed at 20 °C/min to 250 °C, and held at 250 °C for 10 min. The oven temperature program for DCAA was as follows: 35 °C for 10 min, programmed at 2 °C/min to 40 °C, programmed at 5 °C/min to 75 °C and held at 75 °C for 15 min, programmed at 40 °C/min to 100 °C and held at 100 °C for 15 min, and programmed at 40 °C/min to 135 °C. Chloroform was measured by the purge and trap gas chromatographic method using 4660-7890B-5077A gas chromatography equipped with mass spectrometer (Agilent Technologies, Palo Alto, USA). The operating conditions were the following: EI mass spectra were obtained at 70 eV electron energy with ion source at 250 °C. The magnetic analyzer was scanned from 35 to 200 m/z. Carrier gas: helium, flow rate: 1 mL/min, initial temperature: 30 °C for 10 min, programming rate: 7 °C/min up to 72 °C and hold at 72 °C for 1 min, and programmed at 40 °C/min to 220 °C and hold at 220 °C for 1 min.

The three aromatic organic acids and one aliphatic carboxylic acid (fumaric, protocatechuic, 3hydroxybenzoic and benzoic acid) were analyzed by SPE-UPLC (Zhong et al. 2016). This method has been described in detail in Zhong et al. (2017a, b). The four aliphatic carboxylic acids were analyzed by ICS-2100 Ion chromatography (IC, Thermo Scientific, USA) equipped with a Dinoex IonPac AS-19 capillary column (0.4 m × 250 mm) and a Dinoex IonPac AS-19 guard column (0.4 m × 50 mm). The mobile phase was produced by a Dinoex RFIC-EG eluent generator at flow of 1 mL/min with following concentrations: 0– 10 min 10 mM KOH, 10–42 min linear ramp to 52 mM KOH, 42–45 min linear ramp to 70 mM KOH, and 45–50 min 10 mM KOH. 2.4 Quality Control The method of aldehydes detection limits varied from 0.2 to 4.0 μg/L. The method of eight carboxylic acids detection limits varied from 0.8~5.0 μg/L. The recovery rates of each compound were greater than 80%. Experiments were carried out in duplicate. Relative standard deviation of the two measurements was generally below 15%. Calibration curves prepared for each compound were linear (R2 > 0.980). The SPSS software (IBM SPSS, version 17) was used for statistical analysis of data.

3 Results and Discussion 3.1 The Properties of Ozonated HA Waters at Different Conditions It was clear from Tables 1 and 2 that O3 was quite effective for the removal of DOC and UV254 absorbing compounds (e.g. having double bonds and aromatic structure), especially for UV254, resulting in SUVA

Water Air Soil Pollut (2017) 228: 229

values decreasing. The SUVA values reduced to 8.421 L/m/mg and 6.552 L/m/mg, respectively. SUVA is the ratio of UV254 to DOC, which has been widely employed as an indicator of DBP precursors and humic content of the NOM (Mao et al. 2014). Generally, water with high SUVA value is rich in hydrophobic and high molecular organic matter. After 10 min of ozonation reaction (O3 dose 5 mg/L), SUVA removal efficiency was about 50%. These results indicated that ozonation altered the HA properties from hydrophobic matters into hydrophilic organic substances of smaller molecular weights. In addition, HA could be decomposed directly or indirectly by O3, and get converted to non-humic substances (hydrophilic fractions). What were the specific small molecular substances generated would be discussed in detail in the following section. As oxidation proceeds, some of HA or fulvic acid (FA) may be expected to form benzoic compounds. 3.2 Carbonyl Compounds Formed from HA Ozonation at Different Conditions 3.2.1 Aldehydes The classes and formation yields of aldehydes under different contact time and O 3 dose are shown in Figs. 1 and 2, respectively. The predominant aldehydes formed during HA ozonation were formaldehyde, acetaldehyde, propanal, butaldehyde, glyoxal, and methyl glyoxal. Benzaldehyde was not discussed Fig. 1 Aldehyde formation during HA ozonation at a different contact time (time 1– 30 min, O3 dose 1.0 mg/L, pH = 7)

Page 5 of 11 229

here because its concentrations were below or close to the corresponding analytical detection limits in all ozonation conditions. By increasing ozonation contact time, an increase of aldehydes was observed up to 2~10 min (shown in Fig. 1). Analogously, other researchers also reported maximum formation of aldehydes at 4~8 min (Nawrocki et al. 2003; Dąbrowska et al. 2005; Papageorgiou et al. 2017). The increase in aldehydes can be attributed to the conversion of high-molecular-weight compounds to those of lower-molecular-weight by O3 oxidation. The prolonged ozonation resulted in a decrease of the formation of carbonyl compounds. For one thing, aldehydes are easily oxidized to corresponding carboxylic acids or other organic acids by O3,.OH or other radical; for another, ozonation cleaves unsaturated aliphatic chain, opens aromatic rings, and removes or oxidizes alkyl groups to aldehydes. Besides, aldehydes oxidation rates were usually lower than their formation rates, leading to their accumulation in the treated water (Zhong et al. 2017a, b). Aldehyde production increased with O3 doses but approached a plateau at the higher O3 doses (shown in Fig. 2), especially for glyoxal and methyl glyoxal. Nawrocki et al. (2003) reported that the formation of formaldehyde and acetaldehyde depended on the O3 dose, whereas other aldehydes such as glyoxal and methyl glyoxal were not influenced by this. Moreover, it was known that glyoxal and methyl glyoxal were more difficult to remove (biological activated

229 Page 6 of 11

Water Air Soil Pollut (2017) 228: 229

Fig. 2 Aldehyde formation during HA ozonation at a different O3 dose (time = 10 min, O3 dose 0.5~5 mg/L, pH = 7)

carbon) than other aldehydes (Samadi et al. 2015). Previous studies found that aldehydes and carboxylic acids were ubiquitous in raw water of Taihu Lake as well as the each process unit of drinking water treatment plants (Zhong et al. 2017a, b). Therefore, it is necessary to take some measures to effectively control the aldehyde formation. 3.2.2 Carboxylic Acids The classes and formation yields of carboxylic acids under different contact time and O3 dose are shown Fig. 3 Carboxylic acid formation during HA ozonation at a different contact time (time 1– 30 min, O3 dose 1.0 mg/L, pH = 7)

in Figs. 3 and 4, respectively. Protocatechuic, 3hydroxybenzoic, benzoic, fumaric, acetic, and formic acids were detected in ozonated HAs. This is in accordance with previous studies, where O3 can react with NOM, leading to the formation of monocarboxylic, dicarboxylic, and aromatic acids (Huang et al. 2005; Zhang et al. 2008; Sonntag and Gunten 2012). Very few aromatic organic acids had been previously published. Due to the fact that intermediates were frequently more reactive than the initial substrate, the yields of intermediates were no accumulation in the reaction mixture. Besides,

Water Air Soil Pollut (2017) 228: 229

Page 7 of 11 229

Fig. 4 Carboxylic acid formation during HA ozonation at a different O3 dose (time = 10 min, O3 dose 0.5~5 mg/L, pH = 7)

carboxylic acids were formed at much greater concentrations than aldehydes. The concentrations of formic and acetic acids increased monotonically with increasing O3 doses or reaction times, reaching 234.89 and 245.81 μg/L. Although formic acid reacted with O3 at second-order rate constant of about 140/M/s (Gunten 2003), it was the most abundant carboxylic acid in the treated water. It was worth noting that protocatechuic, 3hydroxybenzoic, and benzoic acids had relatively initial concentration. Subsequently, the levels of their concentration declined rapidly at longer reaction time (shown in Fig. 3) and higher O3 doses (shown in Fig. 4), especially hydroxybenzoic acids. This was possible because that O3 is a strong oxidizer with selectivity. It was generally known that O3 direct oxidation was effective for degradation of organic compounds bearing active groups, such as –OH, −NH2, double bond (Hu et al. 2016). The existence of active groups was beneficial for O3 to electrophilic attacking aromatic ring. According to Staehelin and Hoigne (1985), a part of NOM reacted rapidly with O3 to produce benzoic compounds and the presence of some of these acids may contribute to accelerate the decomposition of O3. For example, phthalic acid has been identified as both an initiator and promoter of.OH chain reactions (Huang et al. 2005). Besides, our pervious finding that the order of the reactivity of O3 with aromatic acids was protocatechuic acid > 3hydroxybenzoic acid > benzoic acid > phthalic acid.

3.3 DBPs Formed from HA Ozonation-Chlorination at Different Conditions Figures 5 and 6 respectively illustrated the effects of contact time and O3 dose on DBP formation during subsequent chlorination. As shown in Figs. 5 and 6, direct chlorination disinfection, different contact time, and O3 dose chlorination disinfection generated a number of chloroform and DCAA, chlorination condition: pH of 7, 25 ± 1 °C in the dark for 24 h. After 30 min of ozonation reaction, the removal rates of chloroform and DCAA were 45.15 and 45.81%, respectively. At higher O3 dose, the removal rates of chloroform and DCAA were 71.01 and 63.11%, respectively. This phenomenon was contrary to previous studies about fulvic acid (FA) ozonation (Zhong et al. 2017a, b). The results of FA ozonation shown that O3 reacts with FA having SUVA values <2, resulting in THM and HAA formation increasing. These results were consistent with Hua and Reckhow (2013). According to Hua and Reckhow (2013), ozonation generally reduced the THM and HAA formation by more than 30% for raw waters with medium to high SUVA values (>2). However, O 3 was not effective in destroying THM and HAA precursors in waters with low SUVA values (<2), and it may actually increase the THM and HAA formation potentials of these waters under certain conditions. Therefore, it can be indicated that O3 reacted with aromatic-rich hydrophobic HAs having SUVA values >2 (shown in Tables 1

229 Page 8 of 11

Water Air Soil Pollut (2017) 228: 229

Fig. 5 DBP formation from HAs ozonation-chlorination at a different O3 contact time (O3 contact time 1–30 min, O3 dose 1.0 mg/L, pH = 7)

and 2) and converted it to carboxylate-rich hydrophilic fraction. This change in HA structure led to reduced chloroform and DCAA formation. Hua and Reckhow (2013) investigated DBP formation from various NOM fractions. Results showed that hydrophobic NOM more easily generates THMs, whereas the hydrophilic NOM more easily produced DCAA. Conversely, Chiang et al. (2009) showed that the hydrophobic fraction is not always the primary source of THM precursors, with

Fig. 6 DBP formation from HAs ozonation-chlorination at a different O3 dose (time = 10 min, O3 dose 0.5~5 mg/L, pH = 7)

the hydrophilic fraction bearing the highest amount of THM precursor of all the NOM fractions. Additionally, the mechanism of THM formation involves the different species and moieties which include aldehydes and ketones, which react with chlorine to yield THMs (Yang et al. 2015). Thus, not only HAs were the main DBPs precursors, but also the oxidation intermediates of HA could be the DBPs precursors, and they gave a certain amount of DBPs. Consequently, aromatic acids and low-molecular-

Water Air Soil Pollut (2017) 228: 229

weight aldehydes and carboxylic acids should be under control in drinking water treatment plants. 3.4 The Possible Pathway of DBP Formation from HA Ozonation-Chlorination In a recent study, Wang et al. (2017) studied the transformation of dissolved organic matter (DOM) during ozone-based oxidation processes. The obtained results showed that HA was first converted into FA, and then the majority of these intermediates were further converted to hydrophilic fraction. Meanwhile, in this study and our previous studies, both confirmed the presence of aromatic acids. Besides, He et al. (2015) confirmed that the degradation of humic substances led to the form of dimethyl phthalate in the hydrophilic fractions after Fenton oxidation process.

Page 9 of 11 229

Therefore, based on the previous and present studies results, as well as combined with the results of scholars, a possible formation pathway (shown in Fig. 7) was proposed that HA was first degraded into FA-like compounds, and then FA-like compounds were further converted into aromatic acids (hydrophilic fractions). Subsequently, the direct degradation of aromatic acids into organic substances of smaller molecular weights or inorganic carbon dioxide contributed to the mineralization of organic carbon. Lastly, chlorination reaction involved an electrophilic addition of a chlorine atom to the α-carbon of an enolizable carbonyl compound, subsequent halogenation at the α-carbon, and hydrolysis yield chloroform and DACC. With β-dicarbonyl species, the electron-withdrawing effect of both carbonyls makes the hydrogen groups attached to the α-carbon more

Fig. 7 The possible pathway of DBP formation from HA ozonation-chlorination

229 Page 10 of 11

acidic. Therefore, in the process of chlorine disinfection, formaldehyde, acetaldehyde, glyoxal, methyl glyoxal, aromatic acids, and short-chain acids can be replaced by chlorine atoms and generated DBPs along with decarbonylation reactions.

4 Conclusions This work determined the species and the yields of the intermediates during ozonated HA, as well as DBP formation in ozonation combined with chlorination disinfection. The obtained conclusions were the following: 1. HA was the major precursor material for aldehydes and carboxylic acids. The major intermediates resulting from the ozonation of HA in this study were aldehydes, short-chain acids, and aromatic acids. The concentrations of aldehydes increased as contact time and O3 doses, and reached up maximum at 2~10 min but approached a plateau at the higher O3 doses. The concentrations of formic and acetic acids increased as contact time and O3 doses. However, aromatic acids declined rapidly at longer reaction time and higher O3 doses. It was worth mentioning that aromatic acids include protocatechuic, 3-hydroxybenzoic, and benzoic acids, which had been rarely reported. 2. The possible formation pathway was proposed that HA was first degraded into FA-like compounds, and then FA-like compounds were further converted into aromatic acids. Subsequently, the direct degradation of aromatic acids into low-molecular-weight organic matters. Lastly, chlorine reacted with aldehydes and/or carboxylic acids by addition, hydrolysis, and decarbonylation reactions, leading to DBP formation. 3. Not only HA was the main DBP precursor, but also the oxidation intermediates of HA could be the DBP precursors, and they gave a certain amount of DBPs. Consequently, aromatic acids and low-molecularweight aldehydes and carboxylic acids should be under control in drinking water treatment plants.

Acknowledgements This work was performed with the financial support from the National Water Pollution Control and Major Projects of Science and Technology Management Item No. 2012ZX07403001. The authors gratefully acknowledge the assistance from Wujiang Hua-Yan Water Co. Ltd.

Water Air Soil Pollut (2017) 228: 229

References Chiang, P., Chang, E. E., Chang, P., & Huang, C. (2009). Effects of pre-ozonation on the removal of THM precursors by coagulation. Science of the Total Environment, 407, 5735–5742. Dąbrowska, A., Hordern, B. K., & Nawrocki, J. (2005). Aldehydes formation during water disinfection by ozonation and chlorination process. Global NEST, 7, 61–71. Dąbrowska, A., & Nawrocki, J. (2013). Aldehyde concentrations in wet deposition and river waters. Science of the Total Environment, 452-453, 1–9. EFA (U.S. Environmental protection Agency) (1998). Method 556: determinantion of carbonyl compounds in drinking water by pentafluorobenzylhydroxylamine derivatization and capillary gas chromatography with electron capture detection. EFA (U.S. Environmental protection Agency) (1995). Method 552.2: determinantion of haloacetic acids and dalapon in drinking water by liquid-liquid extraction, derivatization and gas chromatography with electron capture detection. Gaffney, J. S., Marley, N. A., & Clark, S. B. (1996). Humic and fulvic acids and organic colloidal materials in the environment (pp. 2–17). Chicago: American Chemical Society national meeting. Gunten, U. V. (2003). Ozonation of drinking water: part I. Oxidation kinetics and product formation. Water Research, 37, 1443–1467. Gunten, U. V., & Ramseier, M. K. (2009). Mechanisms of phenol ozonation-kinetics of formation of primary and secondary reaction products. Ozone: Science and Engineering, 31, 201–215. Hammes, F., Salhi, E., Köster, O., Kaiser, H., Egli, T., & von Gunten, U. (2006). Mechanistic and kinetic evaluation of organic disinfection by-product and assimilable organic carbon (AOC) formation during the ozonation of drinking water. Water Research, 40, 2275–2286. He, R., Tian, B., Zhang, Q., & Zhang, H. (2015). Effect of Fenton oxidation on biodegradability, biotoxicity and dissolved organic matter distribution of concentrated landfill leachate derived from a membrane process. Waste Management, 38, 232–239. Hu, R., Zhang, L., & Hu, J. (2016). Study on the kinetics and transformation products of salicylic acid in water via ozonation. Chemosphere, 153, 394–404. Hua, G., & Reckhow, D. A. (2013). Effect of pre-ozonation on the formation and speciation of DBPs. Water Research, 47, 4322–4330. Huang, W., Fang, G., & Wang, C. (2005). The determination and fate of disinfection by-products from ozonation of polluted raw water. Science of the Total Environment, 345, 261–272. McGwin Jr., G., Lienert, J., & Kennedy Jr., J. I. (2010). Formaldehyde exposure and asthma in children: a systematic review. Environmental Health Perspectives, 118, 3845–3852. Leenheer, J. A., & Croué, J. P. (2003). Peer reviewed: characterizing aquatic dissolved organic matter. Environmental Science & Technology, 37, 18A–26A. Liu, C., Tang, X., Kim, J., & Korshin, G. V. (2015). Formation of aldehydes and carboxylic acids in ozonated surface water and wastewater: a clear relationship with fluorescence changes. Chemosphere, 125, 182–190.

Water Air Soil Pollut (2017) 228: 229 Mao, Y., Wang, X., Yang, H., Wang, H., & Xie, Y. F. (2014). Effects of ozonation on disinfection byproduct formation and speciation during subsequent chlorination. Chemosphere, 117, 515–520. Matilainen, A., & Sillanpää, M. (2010). Removal of natural organic matter from drinking water by advanced oxidation processes. Chemosphere, 80, 351–365. Miao, H., & Tao, W. (2008). Ozonation of humic acid in water. Journal of Chemical Technology and Biotechnology, 83, 336–344. Nawrocki, J., Swietlik, J., Raczyk-Stanisawiak, U., Dabrowska, A., Bilozor, S., & Ilecki, W. (2003). Influence of ozonation conditions on aldehyde and carboxylic acid formation. Ozone: Science & Engineering, 25, 53–63. Papageorgiou, A., Stylianou, S. K., Kaffes, P., Zouboulis, A. I., & Vouts, D. (2017). Effects of ozonation pretreatment on natural organic matter and wastewater derived organic matterpossible implications on the formation of ozonation by-products. Chemosphere, 170, 33–40. Papageorgiou, A., Voutsa, D., & Papadakis, N. (2014). Occurrence and fate of ozonation by-products at a full-scale drinking water treatment plant. Science of the Total Environment, 481, 392–400. Ratpukdi, T., Siripattanakul, S., & Khan, E. (2010). Mineralization and biodegradability enhancement of natural organic matter by ozonee VUV in comparison with ozone, VUV, ozone-UV, and UV: effects of pH and ozone dose. Water Research, 44, 3531–3543. Richadrson, S., Plewa, M., Wagner, E., Schoeny, R., & Demarini, D. (2007). Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfection by-products in drinking water: a review and roadmap for research. Mutation Research, 636, 178–242. Richardson, S. (2003). Disinfection by-products and other emerging contaminants in drinking water. TrAC Trends in Analytical Chem, 22, 666–684. Samadi, M. T., Azarian, G., Seifipour, F., Huang, C., Yang, X., & Poormohammadi, A. (2015). The formation of aldehydes and ketones ozonation by-products and their variation through general water treatment plant in Hamadan, Iran. Global NEST, 17, 682–691. Sonntag, C.V., Gunten, U.V. (2012). Chemistry of ozone in water and wastewater treatment. IWA Publishing.

Page 11 of 11 229 Staehelin, J., & Hoigne, J. (1985). Decomposition of ozone in water in the presence of organic solutes acting as promoters and inhibitors of radical chain reactions. Environmental Science & Technology, 19, 1206–1213. Tripathi, S., Pathak, V., Tripathi, D. M., & Tripathi, B. D. (2011). Application of ozone based treatments of secondary effluents. Bioresource Technology, 102, 2481–2486. Wang, H., Li, X., Hao, Z., Sun, Y., Wang, Y., Li, W., Tsang, Y.F. (2017). Transformation of dissolved organic matter in concentrated leachate from nanofiltration during ozone-based oxidation processes (O3,O3/H2O2 and O3/UV). Journal of Environmental Management, 191, 244–251. Wert, E. C., Neemann, J. J., Rexing, D. J., & Zegers, R. E. (2008). Biofiltration for removal of BOM and residual ammonia following control of bromate formation. Water Research, 42, 372–378. Yang, X., Gan, W., Zhang, X., Huang, H., & Sharma, V. K. (2015). Effect of pH on the formation of disinfection byproducts in ferrate (VI) pre-oxidation and subsequent chlorination. Separation and Purification Technology, 156, 980–986. Zhang, T., Lu, J., Ma, J., & Qiang, Z. (2008). Comparative study of ozonation and synthetic goethite-catalyzed ozonation of individual NOM fractions isolated and fractionated from a filtered river water. Water Research, 42, 1563–1570. Zhao, G., Lu, X., Zhou, Y., & Gu, Q. (2013). Simultaneous humic acid removal and bromate control by O3 and UV/O3 processes. Chemical Engineering Journal, 232, 74–80. Zhong, X., Cui, C., & Yu, S. (2016). Determination of benzenoid acids in water by solid-phase extraction/ultra-high performance liquid chromatography. China Water & Wastewater, 32, 107–112. Zhong, X., Cui, C., & Yu, S. (2017a). The determination and fate of disinfection by-products from ozonation-chlorination of fulvic acid. Environmental Science and Pollution Research, 24, 6472–6480. Zhong, X., Cui, C., & Yu, S. (2017b). Seasonal evaluation of disinfection by-products throughout two full-scale drinking water treatment plants. Chemosphere, 179, 290–297. Zimmermann, S. G., Wittenwiler, M., Hollender, J., Krauss, M., Ort, C., Siegrist, H., & Gunten, U. V. (2011). Kinetic assessment and modeling of an ozonation step for full-scale municipal wastewater treatment: micropollutant oxidation, byproduct formation and disinfection. Water Research, 45, 605–617.