Bioresource Technology xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Influences of dissolved organic matter characteristics on trihalomethanes formation during chlorine disinfection of membrane bioreactor effluents Defang Ma, Bo Peng, Yuhang Zhang, Baoyu Gao ⇑, Yan Wang, Qinyan Yue, Qian Li Shandong Key Laboratory of Water Pollution Control and Resource Reuse, School of Environmental Science and Engineering, Shandong University, Ji’nan 250100, China
h i g h l i g h t s MBR-effluent DOM was fractionated through UF and XAD-8 resin adsorption. DOM fractions were characterized by 3DEEM. Macromolecular and hydrophobic organics were the primary THMs precursors. Aromatic moieties and humic acid-like contributed largely to the formation of THMs. Low-MW and hydrophilic DOM were more susceptible to produce bromo-THMs.
a r t i c l e
i n f o
Article history: Received 13 December 2013 Received in revised form 23 February 2014 Accepted 26 February 2014 Available online xxxx Keywords: Dissolved organic matter Fractionation Chlorination THMs 3DEEM
a b s t r a c t Dissolved organic matter (DOM) in MBR-treated municipal wastewater intended for reuse was fractionated through ultrafiltration and XAD-8 resin adsorption and characterized by fluorescence spectroscopy. To probe the influences of DOM characteristics on trihalomethanes (THMs) formation reactivity during chlorination, THMs yield and speciation of DOM fractions was investigated. It was found that chlorine reactivity of DOM decreased with the decrease of molecular weight (MW), and MW > 30 kDa fractions produced over 55% of total THMs in chlorinated MBR effluent. Hydrophobic organics had much higher THMs formation reactivity than hydrophilic substances. Particularly, hydrophobic acids exhibited the highest chlorine reactivity and contributed up to 71% of total THMs formation. Meanwhile, low-MW and hydrophilic DOM were susceptible to produce bromine-containing THMs. Of the fluorescent DOM in MBR effluent, aromatic moieties and humic acid-like had higher chlorine reactivity. Conclusively, macromolecular and hydrophobic organics containing aromatic moieties and humic acidlike must be removed to reduce THMs formation. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Membrane bioreactor (MBR) technology has been used for municipal wastewater treatment and reclamation (Francy et al., 2012) to relieve global water scarcity (Chen et al., 2012). Stable and high quality reclaimed water can be produced from municipal wastewater by an MBR system (Hirani et al., 2010). Chlorine disinfection is generally required to prevent the potential transmission of pathogenic microorganisms in reclaimed water (USEPA, 2004). However, reactions between chlorine and dissolved organic matter (DOM) in MBR effluents create genotoxic and cytotoxic chlorine
⇑ Corresponding author. Tel.: +86 531 88366771; fax: +86 531 88364513. E-mail addresses: [email protected], [email protected] (B. Gao).
disinfection by-products (DBPs), such as trihalomethanes (THMs) (Chellam and Krasner, 2001). Comerton et al. (2005) observed that total THMs (TTHM) concentration in chlorinated municipal wastewater treated by an MBR was up to 689 lg/L, which was high enough to be concerned for human health security. DOM in MBR effluents is a complex mixture of various organic materials. It contains initial persistent organic matters (e.g. recalcitrant natural organic matter (NOM) from drinking water and synthetic organic chemicals added during anthropogenic use) in the raw municipal wastewater and refractory soluble microbial products (SMP) derived during biological processes in the bioreactor. Particularly, SMP has been found to be the most significant component of DOM in MBR effluents (Barker and Stuckey, 1999; Jarusutthirak and Amy, 2006), and is the primary source of THMs precursors (Ma et al., 2013).
http://dx.doi.org/10.1016/j.biortech.2014.02.126 0960-8524/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Ma, D., et al. Influences of dissolved organic matter characteristics on trihalomethanes formation during chlorine disinfection of membrane bioreactor effluents. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.02.126
2
D. Ma et al. / Bioresource Technology xxx (2014) xxx–xxx
Therefore, fractionating DOM into more homogenous components according to the physical and chemical properties is an effective method to facilitate understanding its characteristics. Resin adsorption chromatography (RAC) and ultrafiltration (UF) methods have been widely used for the fractionation of organic matters (Kitis et al., 2002). XAD-8 resins are the most widely used resins which concentrate and fractionate DOM into operationally defined components (e.g., hydrophobic and hydrophilic fractions) based on its chemical properties (i.e. their affinities to XAD-8 resins and their back-elution efficiencies) (Leenheer, 1981). The UF separation method fractionates DOM depending on its MW. Such fractionation methods can achieve high DOM recovery and primely preserve the properties of the source waters, e.g. the specific ultraviolet absorbance at 254 nm (SUVA) and the reactivity with chlorine (Kitis et al., 2002). RAC method was used to isolate DOM in secondary-treated municipal wastewater into hydrophobic bases (HoB), hydrophobic acids (HoA) and hydrophobic neutrals (HoN), hydrophilic bases (HiB), hydrophilic acids (HiA) and hydrophilic neutrals (HiN) to investigate their related disinfection by-products formation potential (Zhang et al., 2009). So far, however, the studies on the fractionation of MBR-effluent DOM have mainly focused on their membrane fouling behaviors in MBR operation (Liang et al., 2007). How physical and chemical properties of DOM in the MBR-treated municipal wastewater influence the formation of THMs during chlorine disinfection have been little reported. Therefore, the primary objective of this research was to probe the influences of DOM physicochemical characteristics on THMs formation in MBR-treated municipal wastewater during chlorine disinfection by evaluating THMs formation and speciation of fractions generated by XAD-8 resin adsorption or UF separation. DOM fractions obtained from resin adsorption or UF were also characterized using three-dimensional excitation and emission matrix (3DEEM) fluorescence spectroscopy to investigate the effects of source and chemical composition of DOM on THMs formation. The results of this study will provide some fundamental information about which classes of DOM in MBR-treated municipal wastewater must be removed to reduce the THMs formation during chlorine disinfection to ensure a safe water reuse. 2. Methods 2.1. MBR and operation conditions A laboratory-scale submerged MBR (Ma et al., 2013) was used to treat municipal wastewater in this study. The municipal wastewater was pumped from the detritus chamber of a local sewage treatment plant (Jinan, China) and was pre-treated with a mesh sieve (0.2 mm) before being delivered to the MBR tank. Water quality characteristics of the pre-treated municipal wastewater are listed in Table 1. The MBR was operated under 20 ± 1 °C with hydraulic retention time (HRT) of 12 h and sludge retention time (SRT) of 60 days. The dissolved oxygen (DO) concentration and
Table 1 Water quality parameters of influent and effluent samples in the MBR system.
BOD5 (mg/L) COD (mg/L) DOC (mg/L) UV254 (cm1) NH+4 (mg/L) Br (mg/L)
Pre-treated influent
MBR effluent
98.01 ± 19.23 169.75 ± 12.17 18.605 ± 2.058 0.293 ± 0.078 40.24 ± 7.32 –a
2.16 ± 0.19 15.36 ± 2.65 4.714 ± 0.322 0.119 ± 0.023 0.42 ± 0.01 0.85 ± 0.09
a Br in the pre-treated municipal wastewater was not measured; BOD5, five-day biological oxygen demand; COD, chemical oxygen demand; DOC, dissolved organic carbon; UV254, UV absorbance at 254 nm; NH+4, ammonia nitrogen.
pH of the sludge mixture was maintained at 2.0–5.0 mg/L and 6.0–8.0, respectively. Activated sludge obtained from aeration basin of a local sewage treatment plant (Ji’nan, China) was used to seed the MBR. After seeding, a period of 30 days was provided for the activated sludge microbes to acclimate the MBR conditions with no sludge wastage. Another 60 days were provided for the MBR system to reach stabilization, during which period, SRT was maintained at 60 days. When the MBR kept stable for 20 days, DOM fractionation, characterization and chlorination experiments for MBR effluents were performed. The membrane-filtered effluents used for DOM fractionation and chlorination experiments were taken from the MBR system at 20, 25, 30, 35 and 40 days after steady state (i.e., 110, 115, 120, 125 and 130 days after the addition of seed sludge). 2.2. DOM fractionation 2.2.1. UF fractionation A MinimTM II Tangential Flow Filtration (TFF) system (PALL, USA) described in our previous report (Ma et al., 2013) was used for the UF fractionation of MBR effluents. A series of OmegaTM membranes with molecular weight cut-off of 1, 5, 10, 30 and 100 kDa were used in this study to fractionate the MBR-effluent DOM into six size fractions. A blank sample with ultrapure water for each membrane module was collected before the application of sample. 2.2.2. XAD-8 resin fractionation DOM was fractionated into hydrophilic substances (HiS), HoA, HoB and HoN by using nonionic Amberlite XAD-8 resins. In this study, 130 mL XAD-8 resin was required for the fractionation process as determined by the following equation: 0
V E ¼ 2V 0 ð1 þ k Þ
ð1Þ
where VE is the volume of sample that is processed through the XAD-8 resin column (VE = 8 L), V0 is the void volume of the column (60% of the bed volume), k0 is the column capacity factor (k0 = 50). XAD-8 resins were prepared and cleaned as described by Leenheer (1981). The resins required (20–60 mesh) were separated from the large beads and fines, and stored in 0.1 mol/L NaOH solution for 24 h. Then, the resins were cleaned by sequential 24-h Soxhlet extractions with acetone and hexane to remove organic resin contaminants. The XAD-8 resins were rinsed with methanol until the effluent is free of hexane, and then rinsed with ultrapure water until the DOC of the effluent is less than 0.5 mg/L. 130 mL (wet volume) of cleaned XAD-8 resin was packed into a glass column and rinsed with 0.1 mol/L NaOH, 0.1 mol/L HCl and ultrapure water just before application of the sample. A blank sample was collected after the final rinse with ultrapure water. The fractionation of DOM was performed following the procedure modified from Leenheer (1981). The XAD-8 resin fractionation procedure can be seen from Supplemental Fig. S1. 8 L water sample followed by 2.5 bed volumes of ultrapure water was pumped through the cleaned XAD-8 resin column at a rate of 5 bed volumes per hour (i.e., 0.65 L/h). This effluent was named as DOM1. The HoB was backflush eluted with 2 bed volumes of 0.1 mol/L HCl. DOM1 was acidified to pH 2 with HCl and recycled through the XAD-8 column at the rate of 0.65 L/h. The portion that passed through the resin column collected along with an additional 130 mL of 0.01 mol/L HCl used to rinse the XAD-8 was denoted the HiS fraction. HoA was desorbed by backflush elution with 0.5 bed volume of 0.1 mol/L NaOH followed by 1.5 bed volumes of ultrapure water. After HoA was desorbed, the XAD-8 resin was freeze-dried and then soxhlet-extracted with methanol to obtain the HoN fraction. The excess methanol was removed by
Please cite this article in press as: Ma, D., et al. Influences of dissolved organic matter characteristics on trihalomethanes formation during chlorine disinfection of membrane bioreactor effluents. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.02.126
3
D. Ma et al. / Bioresource Technology xxx (2014) xxx–xxx Table 2 DOM recoveries during UF and XAD-8 fractionation.a Fractionation processes
Fractions
Volume (L)
DOC (mg/L)
UV254 (L/mg m)
DOC Recovery
UV254 Recovery
UF
Source water >100 kD 30–100 kDa 10–30 kDa 5–10 kDa 1–5 kDa 100 kDa fraction contributed about 52% of the total DOC (Ma et al., 2013). The MW 10 kDa) had more humic acid-like components with high UV-absorbance as proved by the 3DEEM fluorescence spectroscopy in Section 3.4. SUVA of XAD-8 fractions were sequenced in the order from large to small as: HoA > HoB > HiS > HoN. HoA exhibited the highest SUVA (2.741 ± 0.110 L/mg m) even higher than that of the MBR effluent, which was consistent with previous studies (Kitis et al., 2002; Zhang et al., 2009). SUVA of HoB (2.445 ± 0.144 L/mg m) was much higher than that of HoN (1.484 ± 0.154 L/mg m), which was distinct with other reports (Chen et al., 2003; Zhang et al., 2009). The larger SUVA of HoA and HoB fractions was due to the presence of high-aromaticity components having more UVabsorbance, e.g. aromatic protein, which was proved by 3DEEM fluorescence spectroscopy in Section 3.4.
n¼
½CHCl2 Br þ 2½CHClBr2 þ 3½CHBr3 TTHM
ð4Þ
where the concentrations are on a mole basis. The results showed that fractions with MW larger than 1 kDa exhibited the similar n value of about 0.2 mol/mol, which was lower than that of the MBR effluent (n = 0.24). While the CHBr2Cl. For the CHBr2Cl. In contrast, CHBr2Cl was dominant for the HiS fraction. HiS had a rather higher bromine incorporation factor (n = 1.27) than hydrophobic fractions. Very small n values (less than 0.1) were found in the HoA, HoB and HoN fractions. These results suggested that compared with hydrophobic components, hydrophilic DOM was more easily to combine with bromine.
Please cite this article in press as: Ma, D., et al. Influences of dissolved organic matter characteristics on trihalomethanes formation during chlorine disinfection of membrane bioreactor effluents. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.02.126
6
D. Ma et al. / Bioresource Technology xxx (2014) xxx–xxx
3.4. Effect of source and chemical compositions of DOM on THMs formation As mentioned above, MBR-effluent DOM is a heterogeneous mixture of recalcitrant NOM residual in drinking water, synthetic organic chemicals added during anthropogenic use and refractory SMP generated through microbial metabolism. Therefore, it is very necessary to trace the source of DOM and investigate the influences of its source and chemical compositions on THMs formation. The fluorescence index (FI, the ratio of emission intensity at 450 nm to that at 500 nm with an excitation of 370 nm), has been used to distinguish between DOM derived from microbial versus terrestrial sources (Fellman et al., 2009; Murphy et al., 2011). It has been proved that FI of terrestrial DOM is less than 1.4, whereas FI of microbial DOM is larger than 1.9 (Goldman et al., 2012). FI of the MBR effluent was 2.11, indicating that microbial products were the dominated DOM in MBR effluents. The FI of UF and XAD-8 fractions were larger than 1.9, with the exception of HoA fraction. The fluorescence index of HoA was 1.66, suggesting the presence of terrigenous humus in MBR effluents. 3DEEM fluorescence spectroscopy was introduced to distinguish and quantify source material and chemical compositions of DOM for UF and XAD-8 fractions (Fellman et al., 2009). The specific peak and intensity in EEM spectra represent the specific type and concentration of DOM. Interpretation of the EEM spectra for each fraction was done as described by (Chen et al., 2003). Fig. 1 presents the EEM spectra of the UF fractions. The excitation and emission maxima for the UF fractions varied from sample
to sample but common trends are observed. UF fractions had three types of fluorescent substances including fulvic-acid like, soluble microbial by-product-like and humic acid-like. All of the six UF fractions had an EEM center of fulvic-acid like and soluble microbial by-product-like except that the 10–30 kDa fraction exhibited an EEM shoulder of soluble microbial by-product-like. The high MW fractions (MW > 10 kDa) had a fluorophore of humic-acid like which exhibited high UV absorbance at 254 nm. On the contrary, fractions with MW less than 10 kDa had an EEM shoulder of humic-acid like. Fluorescence intensity of humic acid-like reduced with the decrease of MW, which was consistent with the variation of specific THMFP. On the contrary, fluorescence intensity of soluble microbial by-product-like generally increased with the decrease of MW. And for fractions with MW lower than 10 kDa, soluble microbial by-product-like exhibited the highest fluorescence intensity. Clear relationship between fluorescence intensity of fulvic-acid like and THMs formation reactivity was not found. The results indicated that THMs formation was mostly attributed to humic acid-like rather than fulvic acids or tyrosin & protein-like. In addition, the
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Influences of dissolved organic matter characteristics on trihalomethanes formation during chlorine disinfection of membrane bioreactor effluents Defang Ma, Bo Peng, Yuhang Zhang, Baoyu Gao ⇑, Yan Wang, Qinyan Yue, Qian Li Shandong Key Laboratory of Water Pollution Control and Resource Reuse, School of Environmental Science and Engineering, Shandong University, Ji’nan 250100, China
h i g h l i g h t s MBR-effluent DOM was fractionated through UF and XAD-8 resin adsorption. DOM fractions were characterized by 3DEEM. Macromolecular and hydrophobic organics were the primary THMs precursors. Aromatic moieties and humic acid-like contributed largely to the formation of THMs. Low-MW and hydrophilic DOM were more susceptible to produce bromo-THMs.
a r t i c l e
i n f o
Article history: Received 13 December 2013 Received in revised form 23 February 2014 Accepted 26 February 2014 Available online xxxx Keywords: Dissolved organic matter Fractionation Chlorination THMs 3DEEM
a b s t r a c t Dissolved organic matter (DOM) in MBR-treated municipal wastewater intended for reuse was fractionated through ultrafiltration and XAD-8 resin adsorption and characterized by fluorescence spectroscopy. To probe the influences of DOM characteristics on trihalomethanes (THMs) formation reactivity during chlorination, THMs yield and speciation of DOM fractions was investigated. It was found that chlorine reactivity of DOM decreased with the decrease of molecular weight (MW), and MW > 30 kDa fractions produced over 55% of total THMs in chlorinated MBR effluent. Hydrophobic organics had much higher THMs formation reactivity than hydrophilic substances. Particularly, hydrophobic acids exhibited the highest chlorine reactivity and contributed up to 71% of total THMs formation. Meanwhile, low-MW and hydrophilic DOM were susceptible to produce bromine-containing THMs. Of the fluorescent DOM in MBR effluent, aromatic moieties and humic acid-like had higher chlorine reactivity. Conclusively, macromolecular and hydrophobic organics containing aromatic moieties and humic acidlike must be removed to reduce THMs formation. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Membrane bioreactor (MBR) technology has been used for municipal wastewater treatment and reclamation (Francy et al., 2012) to relieve global water scarcity (Chen et al., 2012). Stable and high quality reclaimed water can be produced from municipal wastewater by an MBR system (Hirani et al., 2010). Chlorine disinfection is generally required to prevent the potential transmission of pathogenic microorganisms in reclaimed water (USEPA, 2004). However, reactions between chlorine and dissolved organic matter (DOM) in MBR effluents create genotoxic and cytotoxic chlorine
⇑ Corresponding author. Tel.: +86 531 88366771; fax: +86 531 88364513. E-mail addresses: [email protected], [email protected] (B. Gao).
disinfection by-products (DBPs), such as trihalomethanes (THMs) (Chellam and Krasner, 2001). Comerton et al. (2005) observed that total THMs (TTHM) concentration in chlorinated municipal wastewater treated by an MBR was up to 689 lg/L, which was high enough to be concerned for human health security. DOM in MBR effluents is a complex mixture of various organic materials. It contains initial persistent organic matters (e.g. recalcitrant natural organic matter (NOM) from drinking water and synthetic organic chemicals added during anthropogenic use) in the raw municipal wastewater and refractory soluble microbial products (SMP) derived during biological processes in the bioreactor. Particularly, SMP has been found to be the most significant component of DOM in MBR effluents (Barker and Stuckey, 1999; Jarusutthirak and Amy, 2006), and is the primary source of THMs precursors (Ma et al., 2013).
http://dx.doi.org/10.1016/j.biortech.2014.02.126 0960-8524/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Ma, D., et al. Influences of dissolved organic matter characteristics on trihalomethanes formation during chlorine disinfection of membrane bioreactor effluents. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.02.126
2
D. Ma et al. / Bioresource Technology xxx (2014) xxx–xxx
Therefore, fractionating DOM into more homogenous components according to the physical and chemical properties is an effective method to facilitate understanding its characteristics. Resin adsorption chromatography (RAC) and ultrafiltration (UF) methods have been widely used for the fractionation of organic matters (Kitis et al., 2002). XAD-8 resins are the most widely used resins which concentrate and fractionate DOM into operationally defined components (e.g., hydrophobic and hydrophilic fractions) based on its chemical properties (i.e. their affinities to XAD-8 resins and their back-elution efficiencies) (Leenheer, 1981). The UF separation method fractionates DOM depending on its MW. Such fractionation methods can achieve high DOM recovery and primely preserve the properties of the source waters, e.g. the specific ultraviolet absorbance at 254 nm (SUVA) and the reactivity with chlorine (Kitis et al., 2002). RAC method was used to isolate DOM in secondary-treated municipal wastewater into hydrophobic bases (HoB), hydrophobic acids (HoA) and hydrophobic neutrals (HoN), hydrophilic bases (HiB), hydrophilic acids (HiA) and hydrophilic neutrals (HiN) to investigate their related disinfection by-products formation potential (Zhang et al., 2009). So far, however, the studies on the fractionation of MBR-effluent DOM have mainly focused on their membrane fouling behaviors in MBR operation (Liang et al., 2007). How physical and chemical properties of DOM in the MBR-treated municipal wastewater influence the formation of THMs during chlorine disinfection have been little reported. Therefore, the primary objective of this research was to probe the influences of DOM physicochemical characteristics on THMs formation in MBR-treated municipal wastewater during chlorine disinfection by evaluating THMs formation and speciation of fractions generated by XAD-8 resin adsorption or UF separation. DOM fractions obtained from resin adsorption or UF were also characterized using three-dimensional excitation and emission matrix (3DEEM) fluorescence spectroscopy to investigate the effects of source and chemical composition of DOM on THMs formation. The results of this study will provide some fundamental information about which classes of DOM in MBR-treated municipal wastewater must be removed to reduce the THMs formation during chlorine disinfection to ensure a safe water reuse. 2. Methods 2.1. MBR and operation conditions A laboratory-scale submerged MBR (Ma et al., 2013) was used to treat municipal wastewater in this study. The municipal wastewater was pumped from the detritus chamber of a local sewage treatment plant (Jinan, China) and was pre-treated with a mesh sieve (0.2 mm) before being delivered to the MBR tank. Water quality characteristics of the pre-treated municipal wastewater are listed in Table 1. The MBR was operated under 20 ± 1 °C with hydraulic retention time (HRT) of 12 h and sludge retention time (SRT) of 60 days. The dissolved oxygen (DO) concentration and
Table 1 Water quality parameters of influent and effluent samples in the MBR system.
BOD5 (mg/L) COD (mg/L) DOC (mg/L) UV254 (cm1) NH+4 (mg/L) Br (mg/L)
Pre-treated influent
MBR effluent
98.01 ± 19.23 169.75 ± 12.17 18.605 ± 2.058 0.293 ± 0.078 40.24 ± 7.32 –a
2.16 ± 0.19 15.36 ± 2.65 4.714 ± 0.322 0.119 ± 0.023 0.42 ± 0.01 0.85 ± 0.09
a Br in the pre-treated municipal wastewater was not measured; BOD5, five-day biological oxygen demand; COD, chemical oxygen demand; DOC, dissolved organic carbon; UV254, UV absorbance at 254 nm; NH+4, ammonia nitrogen.
pH of the sludge mixture was maintained at 2.0–5.0 mg/L and 6.0–8.0, respectively. Activated sludge obtained from aeration basin of a local sewage treatment plant (Ji’nan, China) was used to seed the MBR. After seeding, a period of 30 days was provided for the activated sludge microbes to acclimate the MBR conditions with no sludge wastage. Another 60 days were provided for the MBR system to reach stabilization, during which period, SRT was maintained at 60 days. When the MBR kept stable for 20 days, DOM fractionation, characterization and chlorination experiments for MBR effluents were performed. The membrane-filtered effluents used for DOM fractionation and chlorination experiments were taken from the MBR system at 20, 25, 30, 35 and 40 days after steady state (i.e., 110, 115, 120, 125 and 130 days after the addition of seed sludge). 2.2. DOM fractionation 2.2.1. UF fractionation A MinimTM II Tangential Flow Filtration (TFF) system (PALL, USA) described in our previous report (Ma et al., 2013) was used for the UF fractionation of MBR effluents. A series of OmegaTM membranes with molecular weight cut-off of 1, 5, 10, 30 and 100 kDa were used in this study to fractionate the MBR-effluent DOM into six size fractions. A blank sample with ultrapure water for each membrane module was collected before the application of sample. 2.2.2. XAD-8 resin fractionation DOM was fractionated into hydrophilic substances (HiS), HoA, HoB and HoN by using nonionic Amberlite XAD-8 resins. In this study, 130 mL XAD-8 resin was required for the fractionation process as determined by the following equation: 0
V E ¼ 2V 0 ð1 þ k Þ
ð1Þ
where VE is the volume of sample that is processed through the XAD-8 resin column (VE = 8 L), V0 is the void volume of the column (60% of the bed volume), k0 is the column capacity factor (k0 = 50). XAD-8 resins were prepared and cleaned as described by Leenheer (1981). The resins required (20–60 mesh) were separated from the large beads and fines, and stored in 0.1 mol/L NaOH solution for 24 h. Then, the resins were cleaned by sequential 24-h Soxhlet extractions with acetone and hexane to remove organic resin contaminants. The XAD-8 resins were rinsed with methanol until the effluent is free of hexane, and then rinsed with ultrapure water until the DOC of the effluent is less than 0.5 mg/L. 130 mL (wet volume) of cleaned XAD-8 resin was packed into a glass column and rinsed with 0.1 mol/L NaOH, 0.1 mol/L HCl and ultrapure water just before application of the sample. A blank sample was collected after the final rinse with ultrapure water. The fractionation of DOM was performed following the procedure modified from Leenheer (1981). The XAD-8 resin fractionation procedure can be seen from Supplemental Fig. S1. 8 L water sample followed by 2.5 bed volumes of ultrapure water was pumped through the cleaned XAD-8 resin column at a rate of 5 bed volumes per hour (i.e., 0.65 L/h). This effluent was named as DOM1. The HoB was backflush eluted with 2 bed volumes of 0.1 mol/L HCl. DOM1 was acidified to pH 2 with HCl and recycled through the XAD-8 column at the rate of 0.65 L/h. The portion that passed through the resin column collected along with an additional 130 mL of 0.01 mol/L HCl used to rinse the XAD-8 was denoted the HiS fraction. HoA was desorbed by backflush elution with 0.5 bed volume of 0.1 mol/L NaOH followed by 1.5 bed volumes of ultrapure water. After HoA was desorbed, the XAD-8 resin was freeze-dried and then soxhlet-extracted with methanol to obtain the HoN fraction. The excess methanol was removed by
Please cite this article in press as: Ma, D., et al. Influences of dissolved organic matter characteristics on trihalomethanes formation during chlorine disinfection of membrane bioreactor effluents. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.02.126
3
D. Ma et al. / Bioresource Technology xxx (2014) xxx–xxx Table 2 DOM recoveries during UF and XAD-8 fractionation.a Fractionation processes
Fractions
Volume (L)
DOC (mg/L)
UV254 (L/mg m)
DOC Recovery
UV254 Recovery
UF
Source water >100 kD 30–100 kDa 10–30 kDa 5–10 kDa 1–5 kDa 100 kDa fraction contributed about 52% of the total DOC (Ma et al., 2013). The MW 10 kDa) had more humic acid-like components with high UV-absorbance as proved by the 3DEEM fluorescence spectroscopy in Section 3.4. SUVA of XAD-8 fractions were sequenced in the order from large to small as: HoA > HoB > HiS > HoN. HoA exhibited the highest SUVA (2.741 ± 0.110 L/mg m) even higher than that of the MBR effluent, which was consistent with previous studies (Kitis et al., 2002; Zhang et al., 2009). SUVA of HoB (2.445 ± 0.144 L/mg m) was much higher than that of HoN (1.484 ± 0.154 L/mg m), which was distinct with other reports (Chen et al., 2003; Zhang et al., 2009). The larger SUVA of HoA and HoB fractions was due to the presence of high-aromaticity components having more UVabsorbance, e.g. aromatic protein, which was proved by 3DEEM fluorescence spectroscopy in Section 3.4.
n¼
½CHCl2 Br þ 2½CHClBr2 þ 3½CHBr3 TTHM
ð4Þ
where the concentrations are on a mole basis. The results showed that fractions with MW larger than 1 kDa exhibited the similar n value of about 0.2 mol/mol, which was lower than that of the MBR effluent (n = 0.24). While the CHBr2Cl. For the CHBr2Cl. In contrast, CHBr2Cl was dominant for the HiS fraction. HiS had a rather higher bromine incorporation factor (n = 1.27) than hydrophobic fractions. Very small n values (less than 0.1) were found in the HoA, HoB and HoN fractions. These results suggested that compared with hydrophobic components, hydrophilic DOM was more easily to combine with bromine.
Please cite this article in press as: Ma, D., et al. Influences of dissolved organic matter characteristics on trihalomethanes formation during chlorine disinfection of membrane bioreactor effluents. Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.02.126
6
D. Ma et al. / Bioresource Technology xxx (2014) xxx–xxx
3.4. Effect of source and chemical compositions of DOM on THMs formation As mentioned above, MBR-effluent DOM is a heterogeneous mixture of recalcitrant NOM residual in drinking water, synthetic organic chemicals added during anthropogenic use and refractory SMP generated through microbial metabolism. Therefore, it is very necessary to trace the source of DOM and investigate the influences of its source and chemical compositions on THMs formation. The fluorescence index (FI, the ratio of emission intensity at 450 nm to that at 500 nm with an excitation of 370 nm), has been used to distinguish between DOM derived from microbial versus terrestrial sources (Fellman et al., 2009; Murphy et al., 2011). It has been proved that FI of terrestrial DOM is less than 1.4, whereas FI of microbial DOM is larger than 1.9 (Goldman et al., 2012). FI of the MBR effluent was 2.11, indicating that microbial products were the dominated DOM in MBR effluents. The FI of UF and XAD-8 fractions were larger than 1.9, with the exception of HoA fraction. The fluorescence index of HoA was 1.66, suggesting the presence of terrigenous humus in MBR effluents. 3DEEM fluorescence spectroscopy was introduced to distinguish and quantify source material and chemical compositions of DOM for UF and XAD-8 fractions (Fellman et al., 2009). The specific peak and intensity in EEM spectra represent the specific type and concentration of DOM. Interpretation of the EEM spectra for each fraction was done as described by (Chen et al., 2003). Fig. 1 presents the EEM spectra of the UF fractions. The excitation and emission maxima for the UF fractions varied from sample
to sample but common trends are observed. UF fractions had three types of fluorescent substances including fulvic-acid like, soluble microbial by-product-like and humic acid-like. All of the six UF fractions had an EEM center of fulvic-acid like and soluble microbial by-product-like except that the 10–30 kDa fraction exhibited an EEM shoulder of soluble microbial by-product-like. The high MW fractions (MW > 10 kDa) had a fluorophore of humic-acid like which exhibited high UV absorbance at 254 nm. On the contrary, fractions with MW less than 10 kDa had an EEM shoulder of humic-acid like. Fluorescence intensity of humic acid-like reduced with the decrease of MW, which was consistent with the variation of specific THMFP. On the contrary, fluorescence intensity of soluble microbial by-product-like generally increased with the decrease of MW. And for fractions with MW lower than 10 kDa, soluble microbial by-product-like exhibited the highest fluorescence intensity. Clear relationship between fluorescence intensity of fulvic-acid like and THMs formation reactivity was not found. The results indicated that THMs formation was mostly attributed to humic acid-like rather than fulvic acids or tyrosin & protein-like. In addition, the
