Removal efficiency and possible pathway of odor compounds (2-methylisoborneol and geosmin) by ozonation

Separation and Purification Technology 117 (2013) 53–58

Contents lists available at SciVerse ScienceDirect

Separation and Purification Technology journal homepage: www.elsevier.com/locate/seppur

Removal efficiency and possible pathway of odor compounds (2-methylisoborneol and geosmin) by ozonation Baoling Yuan a,⇑, Dongmei Xu a, Fei Li a, Ming-Lai Fu b,⇑ a b

Institute of Municipal and Environmental Engineering, College of Civil Engineering, Huaqiao University, Xiamen, Fujian 361021, PR China Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen, Fujian 361021, PR China

a r t i c l e

i n f o

Article history: Received 4 December 2012 Received in revised form 11 April 2013 Accepted 18 April 2013 Available online 26 April 2013 Keywords: Ozone MIB GSM Lyngbya kuetzingii Degradation pathway

a b s t r a c t 2-Methylisoborneol (MIB) and geosmin (GSM) are taste and odor compounds produced as secondary metabolites by some cyanobacteria and actinomycetes, and thus they can be present in some drinking water sources. The removal efficiency, intermediate by-products, and degradation pathway of MIB and GSM in synthetic water by ozonation were studied. The results show that ozone is efficient in removing MIB and GSM from an aqueous solution, depending on pH and the initial MIB and GSM concentration. Ozonation of algal suspension was also studied and the removal efficiency of GSM mainly produced by Lyngbya kuetzingii can reach 99.91% although the ozonation could damage the algal cells and release the intracellular organic compounds. The degradation by-products of MIB or GSM were identified by gas chromatography–mass spectrometry and dehydration and open ring compounds are the main byproducts. Possible degradation pathways for the ozonation of MIB and GSM were proposed. Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction In recent years, with the aggravation of eutrophication in surface waters, cyanobacterial blooms occur frequently [1,2]. 2-methylisoborneol (2-MIB) and geosmin (1,2,7,7-tetramethyl-2norborneol) (GSM) are semi-volatile compounds, which are usually produced as the secondary metabolites by some actinomycetes, planktonic and epiphytic cyanobacteria [3–6]. These two compounds impart earthy–musty taste and odor to water at very low odor thresholds concentration of less than 15 ng L 1 [7,8]. In China, the threshold odor concentrations of these odor compounds in drinking water are 10 ng L 1 according to the China Standards for Drinking Water Quality (GB5749-2006). Conventional water treatment processes (pre-chlorination, coagulation, sedimentation, and filtration) cannot remove these two odor compounds efficiently [9–11]. These secondary metabolites can serve as precursors to form disinfection by-products (DBPs) during pre-chlorination [12,13], producing potential health risks [14–16]. Only the adsorption on activated carbon has been applied successfully to reduce the concentration below the threshold odor concentration [17]. In the past few years, the research on the removal of taste and odor compounds has been greatly focused on oxidative techniques and advanced oxidation processes (AOPs) [18–21]. In advanced drinking water treatment practices, ozonation is the process most commonly employed to remove or decom⇑ Corresponding authors. Tel.: +86 592 616 2780; fax: +86 592 6190762. E-mail addresses: [email protected] (B. Yuan), [email protected] (M.-L. Fu). 1383-5866/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.seppur.2013.04.029

pose these taste and odor compounds (T&Os) as well as precursors of triharomethane because of its strong oxidation potential [22– 24]. Westerhoff et al. [24] conducted batch ozonation experiments to study the effect of ozone oxidation parameters on the removal of MIB and GSM. Liang et al. [25] found that pH is a significant factor affecting oxidation of MIB by ozonation and the presence of natural organic matters did not have a significant effect on ozonation of MIB and GSM. Removal of MIB and GSM by ozonation is dependent on reaction parameters such as pH, ozone dosage, reaction time, water quality parameters, temperature and initial concentrations. This paper not only investigates the correlation between ozonation and water quality parameters, but also presents a detailed investigation of MIB and GSM degradation in an attempt to gain more insight into the underlying reaction mechanisms by ozonation. The intermediate by-products were particularly focused and the possible pathway during the ozonation reaction was proposed. 2. Materials and methods 2.1. Algae culturing and preparation 2.1.1. Algal culturing Planktonic blue–green algae Lyngbya kuetzingii (FACHB 388) was obtained from the Institute of the Freshwater Algae Culture Collection of the Institute of Hydrobiology (FACHB-Collection), Chinese Academy of Sciences. L. kuetzingii was cultivated in batch cultures in BG-11 medium under a cool-white fluorescent lamp with

54

B. Yuan et al. / Separation and Purification Technology 117 (2013) 53–58

the light intensity 2000 lx in a light/dark (L/D) 12 h/12 h cycles at an ambient temperature of 25 ± 1 °C and the algal solutions were shaken for several times every day for avoiding the aggregation. Culture was harvested in the log growth phase after 15 days. 2.1.2. Preparation of the algae suspensions Algae cell concentration in suspensions solution was determined by the concentration of chlorophyll a (Chl-a). Stock culture of L. kuetzingii (2–8  1011 cells L 1) was collected by centrifugation at 5000 rpm for 10 min. The supernatant was used for the measurement of extracellular GSM. The cells separated during the centrifugation were washed three times and then re-suspended in deionized water. The cells were then subjected to three freeze/thawing cycles, and then ultrasonic treatments (5–5 s pulse, 300 W, 10 min), before filtered through 0.45 lm cellulose acetate membranes. The organic matters in the filtrate were referred as intracellular GSM. 2.2. Ozonation Ozone was generated by a corona discharge ozone generator (Model CF-YG5, China). A batch method was used for the ozonation of synthetic waters (containing MIB and GSM) or algal suspensions. The volume of Millli-Q purified water sample was 500 mL and an ozone solution (4.19 mg L 1) was obtained by bubbling ozone gas into Millli-Q purified water at 4 °C constantly. MIB and GSM standard mixture was added into the vessel with the continued stirring. The ozone gas was kept entering into the vessel for 30 min. Ozonation was quenched with 0.1 mol L 1 sodium thiosulfate solution. The samples were prepared for analysis. All ozonation experiments were conducted at 25 ± 1 °C and pH of 7.3. Each experiment was performed in triplicate. O3 concentrations were measured using the standard methods for drinking water (GB/T 57502006). Fig. 1 shows a schematic diagram of the experimental procedure for the decomposition of MIB and GSM. 2.3. Analytical methods 2.3.1. MIB and GSM exaction in synthetic water by Purge-Trap (PT) MIB and GSM were purchased from Sigma–Aldrich (USA) as solutions of 100 mg L 1 in methanol. Internal standards of (+/ )d5-geosmin (d5-GSM) and ( )-2-methylisoborneol-d3 (d3-MIB) were also obtained from Sigma–Aldrich (USA). Five milligrams of d3-MIB and d5-GSM were dissolved in methanol to make a mixed stock solution at an approximate concentration of 500 mg L 1 for each. The synthetic water containing MIB and GSM was kept at 4 °C after filtering through a membrane of 0.45 lm and then determined within 14 days. PT was performed by Eclipse 4660 Purge and Trap Sample Concentrator, with 4551A autosampler (OI Analytical Company, USA), a #10 trap (OI Analytical Company, USA), and a 25 mL purge tube. An OI-Analytical Eclipse 4660 with Te-

nax-Silica Gel-Charcoal sorbent was used. PT was programmed as follows: a 25 mL of water sample (containing 5 g NaCl) with internal standards was drawn by a sample loop of autosampler and transferred to the purge tube. Target compounds were purged from the sample by high-purity nitrogen, with a flow rate of 40 mL min 1 at 60 °C for 13 min, and trapped on Tenax-Silica Gel-Charcoal sorbent. Subsequently, the trap was heated to 200 °C and the trapped components were desorbed by helium for 0.5 min, and then transferred directly to the GC system. Meanwhile, the sampling needle, loop and purge tube were washed with HPLC water three times, and the trap was baked at 240 °C for 10 min. These processes are enough to clean the purge system. 2.3.2. MIB and GSM extraction in algal suspension by Solid Phase Micro-Extraction (SPME) One hundred and fifty milliliters of water samples were placed into a 250 mL vial containing a magnetic stirrer. After addition of 45 g of NaCl and 150 lL of mixed internal standard stock solution (500 mg L 1), the vial was sealed with a silicon-Teflon septum cap. The sealed vial was placed in a water-bath at 60 °C for 30 min and stirred at 800 rpm of the stirring rate. Odor organic compounds derived from algae or formed during ozonation in solution were extracted by using SPME manual devices (Supelco, USA). After pre-heating, 50/30 lm Divinylbenzene/Carboxen/Polydimethylsiloxane (DVB/CAR/PDMS) fiber was used to penetrate the septum and the fiber extended into the headspace for extraction. After exposures for 40 min, the fiber was immediately inserted into the GC injection port for desorption. 2.3.3. Detection of MIB, GSM and by-products by GC/MS method The injector temperature was set at 280 °C in the split mode (split ratio: 20:1) for PT. The injection temperature was set at 230 °C in the splitless mode and desorption time was 3 min for SPME. Analyses were carried out with an Agilent 6890 series GC system coupled with a 5973i series mass spectral detector. Highpurity helium (99.9999%) with a constant flow rate of 1 mL min 1 was used as the carrier gas. The oven temperature was programmed from an initial temperature of 50 °C held for 2 min, then ramped to 160 °C at 5 °C min 1, finally ramped to 280 °C at 20 °C min 1 and held for 8 min, and the total run time was 38 min. The GC–MS transfer line temperature was maintained at 300 °C. The electron impact (EI) ion source of the mass spectrometer and the quadrupole temperature was set at 230 °C and 150 °C, respectively. The EI ionization mode was used with an electron energy of 70 eV. The mass spectra were obtained at a mass-to-charge ratio scan ranging from 45 to 200 amu to determine appropriate masses for selected ion monitoring (SIM). Mass spectral quantitative ions were m/z 107, 138, 112, and 114 for MIB, d3-MIB, GSM and d5-GSM, respectively [26]. Identification of by-products by ozonation was also carried out by mass-spectra. 3. Results and discussion 3.1. Ozonation of MIB and GSM

Air

Ozone generator

Reactor

KI solution

Fig. 1. Experimental setup of ozonation of MIB and GSM.

The ozonation of MIB and GSM were conducted at 4.19 mg L 1 of ozone concentration for 20 min of reaction time. Concentrations of MIB and GSM were determined using PT-GC/MS analyses of the solutions at different reaction times. Control experiments confirm that the concentrations of MIB and GSM remain constant throughout the experimental process without ultrasonic irradiation (no loss due to evaporation or adsorption). Upon ozonation, MIB and GSM (100 ng L 1) are readily degraded and the removal rates within 5 min are more than 60%, as showed in Fig. 2. After ozonation for 20 min, over 90% of both compounds were degraded. As Waterhoff

55

B. Yuan et al. / Separation and Purification Technology 117 (2013) 53–58

(3) The process of ozonation of organic matter falls into two possible pathways, the direct reaction, which is slow and selective, and the indirect reaction, which is radical reaction. On the acidic condition, ozone oxidizes organic compounds in molecular morphology, belonging to direct reaction. But considerable OH radicals can be produced by ozone with the increase of pH and reaction can be regarded as radical reaction. It is well known that OH can react with organic compounds in non-selective mode of with high reaction rate constant (107–109 L (mol s) 1) [24,25].

MIB GSM

80

60

40

Consequently, the removal efficiency of MIB and GSM in neutral and alkaline conditions is better than that in acidic conditions. The pH range of 7–8 is best for the decomposition of MIB and GSM because the pH of surface waters is in this range and no additional pH adjustment is necessary to achieve the maximum removal efficiency.

20

0

0

5

10

15

20

25

30

Time (min) Fig. 2. Effect of reaction time on the removal rate ([MIB]0 = [GSM]0 = 100 ng L 4.19 mg L 1 ozone; pH = 7.3).

1

;

et al. [24] reported, ozonation of MIB and GSM in aqueous media often follows second-order kinetics and hydroxyl radical dominated the oxidant when compared to ozone alone. Furthermore, GSM shows better reaction kinetics when compared to MIB. GSM, with lower aqueous solubility and higher Kow, is more hydrophobic than MIB, so GSM is more volatile and more easily transferred from bulk liquid to gas phase, and subsequently reacts with ozone [20,21]. 3.2. Effect of pH on the removal of MIB and GSM in synthetic water The removal efficiency of MIB and GSM under the observed pH values is showed in Table 1. From Table 1, the removal efficiency of these two compounds increases gradually with the increase of pH. The removal efficiency is higher than 90% when pH is 7.3 and no MIB and GSM can be detected at pH 9.1. The effect of pH on the ozonation process is reflected on the following aspects; (1) The oxidation ability of ozone highly depends on the pH. Oxygen atoms of ozone molecules have strong affinity for electrons or protons, to show strong oxidation. It is well accepted that decomposition rate of ozone in water increases with the increase of pH. When pH increases by one unit, the decomposition of ozone would speed up almost three times, resulting in more high reactivity OH groups [24]. Therefore, the removal rate of odor compounds increases with the increase of pH. (2) The absorption rate of ozone in alkaline conditions is better than that in acidic conditions [27]. Increased absorption rate promotes the diffusion of ozone from the interphase into the liquid phase, so that more ozone participates in oxidation reaction, thereby enhancing the removal rate.

3.3. Effect of initial concentration of MIB and GSM in synthetic water Three different initial concentrations of 100, 200, 500 ng L 1 of MIB and GSM mixture were used to investigate the removal rate of MIB and GSM by ozonation as displayed in Fig. 3. From Fig. 3, the removal rates of GSM at three concentration levels are all higher than those of MIB, which is consistent with Ho et al.’s report [22]. The reasons might be due to the lower solubility and higher partition coefficient of GSM. However, the removal rate tends to decline with the increase of initial concentration of MIB and GSM. For MIB, the removal rate decreases from 93.6% at 100 ng L 1 of MIB and GSM mixture to 66.4% at 500 ng L 1 of MIB and GSM mixture. For GSM, the removal rate decreases from 97.9% at 100 ng L 1 of MIB and GSM mixture to 72.6% at 500 ng L 1 of MIB and GSM mixture. 3.4. Removal of GSM in algal suspension by ozonation The effect of ozonation on the removal of GSM in algal suspension of L. kuetzingii is displayed in Table 2. From Table 2, it can be seen that the contents of GSM vary in the different algal suspension. 548 ng L 1 of GSM is in the extracellular extraction and 1.568  104 ng L 1 of GSM is in the intracellular extraction. 1.623  104 ng L 1 of GSM in the algal suspension almost equals to the total amount of GSM in extracellular and intracellular extraction. The main content of GSM is from the intracellular

100

Odor removal rate (%)

Odor removal rate (%)

100

MIB GSM

80

60

40

20 Table 1 Removal rate of odor compounds under different pH conditions ([MIB]0 = [GSM]0 = 100 ng L 1; 4.19 mg L 1 ozone; 20 min of reaction time; pH = 7.3). (%)

pH = 5.4 (%)

pH = 7.3 (%)

pH = 9.1 (%)

MIB GSM

47.1 88.5

93.6 97.9

100 100

0

100

200

500

Odor concentration (ng/L) Fig. 3. Effect of initial concentration of MIB and GSM mixture on removal rate (4.19 mg L 1 ozone; 20 min of reaction time; pH 7.3).

56

B. Yuan et al. / Separation and Purification Technology 117 (2013) 53–58

Table 2 Variation of Chl-a concentration and removal efficiency of GSM in L. kuetzingii (4.19 mg L Chl-a concentration (mg m

Before ozonation After ozonation Removal efficiency (%)

1

ozone; 20 min of reaction time; pH = 7.3).

3

)

GSM concentration (ng L

66.0 5.4 91.8

1

)

Algal suspension

Extracellular

Intracellular

1623 13.9 99.9

548.0 13.9 97.5

1568 246.5 98.4

Table 3 Possible by-products of MIB after ozonation. CAS

Retention time (min)

Compounds

027538-47-2

8.712

2-Methylenebornane

072540-93-3

9.664

2-Methyl-2-Bornene

007787-20-4

14.829

Camphor

000106-67-2

11.410

2-Ethyl-4-Methyl-1-Pentanol

000104-76-7

11.410

2-Ethyl-1-Hexanol

000112-31-2

15.257

Decanal

GSM. After ozonation of algal suspension, only 246.5 ng L 1 of intracellular GSM was detected. Compared with the content of 1.568  104 ng L 1 in intracellular extraction, the ozonation damages the algal cells and releases the intracellular organic compounds, in which GSM is also released. Fortunately, the released GSM would be further removed by ozonation and 99.9% of GSM removal rate was obtained within 30 min at the ozone dosage of 4.19 mg L 1. Ozonation of the algal suspensions also resulted in the great decrease of Chl-a, which represent the reduction of the living algae cells. 91.8% of Chl-a could be removed within 30 min at the ozone dosage of 4.19 mg L 1. The Chl-a deduction also verifies the damage of algae cell after ozonation. Therefore, ozonation could degrade GSM very effectively although the algal cells were damaged and the intracellular organic compounds were released after ozonation. 3.5. Degradation pathways of MIB and GSM in synthetic water by ozonation A few researches have been carried out for the ozonation of odor compounds until now. Saito et al. [28] reported the microbiological degradation of GSM. In this process, four biodegradation

Structure

by-products were identified by gas chromatography, among which a dehydration product and an oxidation product were identified by mass spectroscopic. Song and O’Shea [20] reported the degradation of MIB and GSM by ultrasonic irradiation. Dehydration and open ring compounds are the main by-products identified by GC/MS. Qi et al. [29] identified camphor was a primary degradation product, which was further oxidized to form other degradation intermediates, such as aldehydes, ketones, and carboxylic acids after the ozonation of MIB. Degradation by-products of MIB were identified by PT-GC/MS and listed in Table 3. GC–MS analyses indicate the formation of two directly dehydrated by-products: 2-methyl-2-bornene (2) and 2-methylenebornane (3). Molecular elimination of H2O plays a key role during oxidizing degradation of MIB. Camphor (4) is also an observed degradation product, which can be explained by elimination of methane from MIB. With respect to bond scission, dissociation of C–C bond is more likely occurred than C–O or C– H bonds based on relative homolytic-bond dissociation energies. MIB, a tertiary alcohol, is susceptible to C–C dissociation and subsequent skeletal rearrangement. Rearrangement and C–C homolytic-bond scissions lead to the ring opening pathways and then 2-ethyl-4-methyl-1-pentanol (5) and 2-ethyl-1-hexanol (6) were

57

B. Yuan et al. / Separation and Purification Technology 117 (2013) 53–58

OH

CH3

3 -H 2O

CH3

-H 2O

8

-H 2O

2

-H 2O

1 Bond scission

Bond scission -CH 3

Bond scission

9

6

5

Bond scission

4 -H 2O

Rearragament

Bond scission

10

Rearragament

-H 2O

Bond scission

Bond scission

11

7 Fig. 4. Possible degradation pathway of MIB degradation ([MIB]0 = 100 ng L 4.19 mg L 1 ozone; 20 min of reaction time; pH = 7.3).

1

;

13

12

Fig. 5. Possible degradation pathway of GSM degradation ([GSM]0 = 100 ng L 4.19 mg L 1 ozone; 20 min of reaction time; pH = 7.3).

Table 4 Possible by-products of GSM after ozonation. CAS

Retention time (min)

Compounds

024145-89-9

16.307

cis-1,4-Dimethyladamantane

000702-79-4

16.307

1,3-Dimethyl-Adamantane

056362-87-9

16.025

cis-1-Ethylideneoctahydro-7a-Methyl-1H-Indene

015869-89-3

11.415

2,5-Dimethyl-Octane

000106-35-4

5.915

3-Heptanone

Structure

1

;

58

B. Yuan et al. / Separation and Purification Technology 117 (2013) 53–58

observed. These intermediate by-products (4, 5, 6) are further oxidized after bond scission and dehydration to form final degradation product-decanal (7). A possible degradation pathway for the ozonation of MIB is proposed in Fig. 4. Degradation by-products of GSM were identified by PT-GC/MS and listed in Table 4. Cis-1,4-dimethyl-adamantane (9) and 1,3-dimethyl-adamantane (10) are the intermediate by-products of GSM dehydration. Cis-1-ethylideneoctahydro-7a-methyl-1H-Indene (11) is the observed intermediate via rearrangement of intermediates (9 and 10). Ring opening reaction and electrophilic substitution reaction can form 2,5-dimethyl-octane (12) and subsequent dehydrogenation yields final product – 3-heptanone (13). A possible degradation pathway for the ozonation of GSM is proposed in Fig. 5. 4. Conclusions The results of the investigations presented here confirm that ozonation is efficient to remove MIB and GSM in the synthetic water, and GSM in algal suspension. The removal efficiency of MIB and GSM in synthetic water increases with the increase of pH and initial concentration of MIB and GSM. The removal rate of GSM in synthetic water is higher than that of MIB due to higher Henry’s law constant. The ozonation of L. kuetzingii result in the damage of algae cell and the release of cellular cytoplasm and intracellular GSM. However, further ozonation can also lead to 91.8% of the algae removal and 99.9% of GSM removal rate within 30 min at the ozone dosage of 4.19 mg L 1. The degradation by-products of MIB or GSM in synthetic water were identified by gas chromatography–mass spectrometry and dehydration, and open ring compounds are the main by-products. Two possible degradation pathways for the ozonation of MIB and GSM were proposed. Acknowledgements The authors greatly appreciated the financial supports from the Natural Science Foundation of China (Grant No. 51178117), the Xiamen Natural Science Funds for Distinguished Young Scholar (No. 3502Z20126009), the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry and the Open Fund of Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences (No. KLUEH201102). References [1] G. Izaguirre, W.D. Taylor, Geosmin and MIB events in a new reservoir in southern California, Water Sci. Technol. 55 (2007) 9–14. [2] A. Peter, O. Koester, A. Schildknecht, U. von Gunten, Occurrence of dissolved and particle-bound taste and odor compounds in Swiss lake waters, Water Res. 43 (2009) 2191–2200. [3] G. Izaguirre, W.D. Taylor, J. Pasek, Off-flavor problems in two reservoirs, associated with planktonic Pseudanabaena species, Water Sci. Technol. 40 (1999) 85–90.

[4] S. Wood, S.T. Williams, W.R. White, Microbes as a source of earthy flavours in potable water – a review (Reprinted), Int. Biodeter. Biodegr. 48 (2001) 26–40. [5] Y.-M. Chen, P. Hobson, M.D. Burch, T.-F. Lin, In situ measurement of odor compound production by benthic cyanobacteria, J. Environ. Monitor 12 (2010) 769–775. [6] L. Xu, B. Xiong, Y. Pan, J. Wang, H. Cao, W. Zhao, Relationship between concentrations of odorous compounds and biomass of phytoplankton and actinomycetes in freshwater ponds of Beijing, China, Aquacul. Int. 18 (2010) 245–254. [7] W.F. Young, H. Horth, R. Crane, T. Ogden, M. Arnott, Taste and odour threshold concentrations of potential potable water contaminants, Water Res. 30 (1996) 331–340. [8] J. Romero, F. Ventura, Occurrence of geosmin and other odorous compounds of natural origin in surface and drinking waters. A case study, Int. J. Environ. Anal. Chem. 77 (2000) 243–254. [9] D. Bruce, P. Westerhoff, A. Brawley-Chesworth, Removal of 2methylisoborneol and geosmin in surface water treatment plants in Arizona, J. Water Supply Res. T. 51 (2002) 183–197. [10] T.F. Lin, J.Y. Wong, H.P. Kao, Correlation of musty odor and 2-MIB in two drinking water treatment plants in South Taiwan, Sci. Total Environ. 289 (2002) 225–235. [11] H. Ahn, S. Chae, S. Kim, C. Wang, R.S. Summers, Efficient taste and odour removal by water treatment plants around the Han River water supply system, Water Sci. Technol. 55 (2007) 103–109. [12] J. Huang, N. Graham, M.R. Templeton, Y. Zhang, C. Collins, M. Nieuwenhuijsen, A comparison of the role of two blue–green algae in THM and HAA formation, Water Res. 43 (2009) 3009–3018. [13] J. Fang, J. Ma, X. Yang, C. Shang, Formation of carbonaceous and nitrogenous disinfection by-products from the chlorination of Microcystis aeruginosa, Water Res. 44 (2010) 1934–1940. [14] A.T. Bathija, E. Ohanian, Health risk to fetuses, infants and children from stage 1 disinfectants and disinfectant by-products (D/DBPS), Toxicol. Sci. 72 (2003) 28. [15] N.J. Ashbolt, Risk analysis of drinking water microbial contamination versus disinfection by-products (DBPs), Toxicology 198 (2004) 255–262. [16] S. Chowdhury, P. Champagne, P.J. McLellan, Uncertainty characterization approaches for risk assessment of DBPs in drinking water: a review, J. Environ. Manage. 90 (2009) 1680–1691. [17] J. Ridal, B. Brownlee, G. McKenna, N. Levac, Removal of taste and odour compounds by conventional granular activated carbon filtration, Water Qual. Res. J. Can. 36 (2001) 43–54. [18] S.W. Jung, K.H. Baek, M.J. Yu, Treatment of taste and odor material by oxidation and adsorption, Water Sci. Technol. 49 (2004) 289–295. [19] G. Park, M. Yu, J. Go, E. Kim, H. Kim, Comparison between ozone and ferrate in oxidising geosmin and 2-MIB in water, Water Sci. Technol. 55 (2007) 117–125. [20] W. Song, K.E. O’Shea, Ultrasonically induced degradation of 2methylisoborneol and geosmin, Water Res. 41 (2007) 2672–2678. [21] K. Kutschera, H. Boernick, E. Worch, Photoinitiated oxidation of geosmin and 2-methylisoborneol by irradiation with 254 nm and 185 nm UV light, Water Res. 43 (2009) 2224–2232. [22] L. Ho, G. Newcombe, J.P. Croue, Influence of the character of NOM on the ozonation of MIB and geosmin, Water Res. 36 (2002) 511–518. [23] M. Sagehashi, K. Shiraishi, H. Fujita, T. Fujii, A. Sakoda, Ozone decomposition of 2-methylisoborneol (MIB) in adsorption phase on high silica zeolites with preventing bromate formation, Water Res. 39 (2005) 2926–2934. [24] P. Westerhoff, B. Nalinakumari, P. Pei, Kinetics of MIB and geosmin oxidation during ozonation, Ozone – Sci. Eng. 28 (2006) 277–286. [25] C. Liang, D. Wang, J. Chen, L. Zhu, M. Yang, Kinetics analysis on the ozonation of MIB and geosmin, Ozone – Sci. Eng. 29 (2007) 185–189. [26] B.L. Yuan, F. Li, D.M. Xu, M.-L. Fu, Comparison of two methods for the determination of geosmin and 2-methylisoborneol in algae samples by stable isotope dilution assay through purge-and-trap or headspace solid-phase microextraction combined with GC/MS, Anal. Meth. 5 (2013) 1739–1746. [27] A. Nemes, I. Fabian, G. Gordon, The kinetics and mechanism of aqueous ozone decomposition in alkaline solution, Inorg. React. Mech. 2 (2000) 327–341. [28] A. Saito, T. Tokuyama, A. Tanaka, T. Oritani, K. Fuchigami, Microbiological degradation of ( )-geosmin, Water Res. 33 (1999) 3033–3036. [29] F. Qi, B.B. Xu, Z.L. Chen, J. Ma, D.Z. Sun, L.Q. Zhang, Efficiency and products investigations on the ozonation of 2-methylisoborneol in drinking water, Water Environ. Res. 81 (2009) 2411–2419.