Control of halophenol formation in seawater during chlorination using UV/TiO₂ pre-treatment

The use of seawater as a source of water to cope with freshwater shortages has been a common practice in many areas, especially in coastal regions. Seawater may be applied for industrial cooling and heating purposes, and/or serves as an alternative source for drinking water (Kim et al. 2015; Manasfi et al. 2019). However, one of primary issues associated with the uses of seawater has been biofouling, which may reduce heat-transfer coefficients during cooling and heating processes, or clog the reverse osmosis membranes that are commonly applied for seawater desalination (Kim et al. 2015; Manasfi et al. 2019). Chlorination is the most widely used anti-biofouling technique for seawater to prevent bacterial and biofilm growth (Khalanski & Jenner 2012; Kim et al. 2015). The unintended consequences during seawater chlorination are the formation of a vast array of disinfection by-products (DBPs), resulting from reactions of chlorine with organic and inorganic compounds, which are a cause for concern for the impact on the environment and human health. Being low in total organic carbon levels, it is expected that DBP formation would be lower in seawater than in municipal wastewater, however, the high levels of bromide (50,000 to 80,000 μg/L) present in seawater likely enhance the formation of brominated DBPs during chlorination, which are more cytotoxic than their chlorinated analogues (Kim et al. 2015). Therefore, it is critical to control DBP formation in seawater applications.

The trihalomethanes (THMs) and haloacetic acids (HAAs) are the most abundant and frequently identified DBPs in chlorinated seawater (Kim et al. 2015). Other DBPs, though detected at lower levels, such as bromophenols, have been reported to be more toxic than the former two classes (Liu & Zhang 2014). Furthermore, the low taste and odor threshold of bromophenols (in the range of μg/L to ng/L) makes the problem more prominent in phenol- and bromide (Br⁻)-containing seawater than in drinking water (Agus & Sedlak 2010). In addition to bromophenols, chloro- and chlorobromo-phenols may also be formed in chlorinated seawater (Acero et al. 2005). Few studies have investigated pretreatment technologies to reduce halophenol (HP) formation during chlorination. It has been demonstrated that pre-ozonation may effectively degrade HP precursors, while the potential for the generation of other DBPs with bromide could impede its use for seawater (Ding et al. 2018).

As an environmentally friendly advanced oxidation process, photocatalytic oxidation is based on using light of wavelength near UV radiation to photoexcite semiconductor catalysts in the presence of oxygen to attack organic contaminants in water. Among various photocatalysts, titanium dioxide (TiO₂) has been most widely used in water treatment, due to its high photocatalytic activity and low cytotoxicity (Sillanpää et al. 2018). UV/TiO₂ is able to degrade water contaminants such as bacteria (Rincón & Pulgarin 2004), taste and odor compounds (Fotiou et al. 2016), as well as natural organic matter (NOM) (Gora et al. 2018), which is a primary contributor of DBP precursors. It has been reported that reduction in the total THM formation potential and total HAA formation potential by using UV/TiO₂ could be in the range of 20% and 90%, respectively (Kent et al. 2011). So far, little information is known on the efficacy of UV/TiO₂ in control of halophenol formation during seawater chlorination. The significance of this research study is to identify the effectiveness of UV/TiO₂ in reducing DBP formation by analyzing the mechanisms during the process, which indicates the use of UV/TiO₂ pretreatment for control of HP formation in seawater during chlorination.

The objectives of this study are for (a) analysis of the changes of the characteristics of dissolved organic matter in seawater by UV/TiO₂, (b) optimization of the effectiveness of UV/TiO₂ for HP precursor degradation, and (c) identification of the species and degradation of HPs formed during seawater chlorination by UV/TiO₂ pretreatment.

The seawater samples were collected from Bohai Bay in Tianjin, China, immediately followed by filtration by glass fiber filters (0.22 μm), and were transported and stored at 4 °C before use. A final concentration of 1 mg/L phenol was added to the seawater sample for the following experiments. The characteristics of the seawater sample with phenol and the original seawater sample are shown in Table 1 and Table S1 in the Supplementary Material. The methodologies for the characterization of the seawater sample in Table 1 and the preparation of the phenol standard sample are described in Ding et al. (2018).

The UV/TiO₂ treatment experiment was carried out in a collimated beam reaction apparatus equipped with a 500 W mercury lamp. The lamp emitted UV light at nearly 254 nm and was mounted horizontally above a 250 mm long × 70 mm i.d. stainless steel collimating tube in the collimated beam apparatus. The photo fluence was measured with a UV fluence meter with a detection range of 0.1–199.9 × 10³ μW/cm² (UV-A, Photoelectric Instrument Factory of Beijing Normal University, China). The reaction samples were placed on an adjusted platform mounted over a stirrer beneath the open end of the tube. The average irradiation intensity at the reaction platform was 1.47 mW/cm², which was measured at the pure water condition without any suspended solids. A different concentration of nanoscale TiO₂ (99.8%, Huaweiruike, China), of 0.1, 0.3, 0.5, 0.7, 1.0, 2.0, 3.0, 4.0, 5.0, and 10.0 g/L, was added into each reaction sample, respectively, and stirred uniformly in the dark at 700 rpm for 30 min, immediately followed by UV exposure. For each UV exposure, a 100 mL seawater sample with TiO₂ suspension was exposed to UV for up to 120 min. The stirring speed was 700 rpm. Samples were withdrawn at 2, 4, 6, 8, 10, 15, 20, 25, 30, 45, 60, 75, 90, 105, and 120 min, and immediately filtered by glass fiber filters (0.22 μm) for UV₂₅₄ analysis and EEM identification. Quality assurance and quality control (QA/QC) is presented in the Supplementary Material.

Raw and irradiated seawater samples with phenol were chlorinated by 2.5 mg Cl₂/L free chlorine up to 30 min. Samples were withdrawn at 2, 4, 6, 8, 10, 15, 20, 25, and 30 min for HP analysis. Detailed procedures are described in Ding et al. (2018).

The HPs in the seawater samples were subjected to solid phase extraction and derivatization with a silylation reagent (BSTFA + TMCS, 99:1, Supelco, USA), followed by analysis by gas chromatography–mass spectrometry (GC/MS) (DB-5 ms, 30 m × 0.25 mm × 0.25 μm; GCMS-QP2010 Plus, Shimadzu, Japan). Detailed procedures were described in Ding et al. (2018). A total of 11 species of HPs were analyzed: 2-chlorophenol (2-CP), 3-chlorophenol (3-CP), 4-chlorophenol (4-CP), 2,4-dichlorophenol (2,4-DCP), 2,5-dichlorophenol (2,5-DCP), 2,4,6-trichlorophenol (2,4,6-TCP), 2-bromophenol (2-BP), 4-bromophenol (4-BP), 2,4-dibromophenol (2,4-DBP), 2,6-dibromophenol (2,6-DBP), and 2,4,6-tribromophenol (2,4,6-TBP).

The UV₂₅₄ of seawater samples treated by UV/TiO₂ is presented in Figure 1 (tabular data shown in Table S2, Supplementary Material). It is shown that as UV exposure time extended, UV₂₅₄ of seawater samples without the addition of TiO₂ slightly increased without great variation. For those samples treated by UV/TiO₂, UV₂₅₄ rapidly increased until UV exposure of up to 20 min, and then it gradually decreased as UV exposure time extended to 120 min. UV₂₅₄ is an indicator of typical DOM with more than three conjugated double bonds (Noethe et al. 2009). Under lower UV doses (UV exposure less than 20 min), the free radicals excited by photocatalysis may not completely oxidize NOM, but may cause a shifting in the organics towards smaller and less aromatic moieties (Mayer et al. 2014), while by extending UV exposure time, the mineralization of organic matter can be readily achieved with oxidants (Liu et al. 2008; Mayer et al. 2014). It is observed that the optimal TiO₂ for removal of UV₂₅₄ was 4.0 g/L (Figure 1). As TiO₂ dose increased up to the optimum of 4.0 g/L, the number of active sites on the photocatalysts increased, which enhanced the generation of hydroxyl and superoxide radicals (Akpan & Hameed 2009). However, once the concentration of the photocatalyst increased beyond the optimal value, the increased turbidity caused by excess photocatalysts limited the illumination, thus the photocatalytic degradation efficiency reduced accordingly (Lathasree et al. 2004; Sun et al. 2008). In addition, the high concentration beyond the optimum might have resulted in agglomeration of the photocatalyst, which inhibited the active sites for photon absorption and led to a lower photocatalytic activity (Huang et al. 2008).

By plotting EEM fluorescence spectra (Figure S1, Supplementary Material), it is shown that the fluorescence intensity of regions I and VI are the highest among the five regions, hence tyrosine-like aromatic protein and soluble microbial by-product-like organic compounds were the primary components in DOM. The volumes beneath EEM of regions I and IV of the seawater samples treated with different doses of TiO₂ and UV are presented in Figure 2. The major components of DOM degraded as UV doses increased, even without the addition of TiO₂. The addition of TiO₂ increased the normalized excitation–emission area volume (Φ_i,n) of EEM in the first place, while by UV irradiation, it drastically decreased to the level below that of the raw samples. It is noted that the application of TiO₂ greatly enhanced the effectiveness of UV exposure. By increasing TiO₂ dose from 0.5 to 1.0 g/L, the degradation of soluble microbial by-product-like organic compounds and tyrosine-like aromatic protein gradually increased. However, further increasing TiO₂ doses up to 10.0 g/L only slightly raised the efficacy of degradation. On the other hand, with TiO₂ doses higher than 1.0 g/L, Φ_i,n rapidly decreased with UV exposure extended to 30 min, followed by entering a lag phase after that. In consideration of the removal efficacy of UV₂₅₄ (shown in Figure 1, tabular data shown in Table S2) and the major components of DOM (shown in Figure 2, tabular data shown in Tables S3 and S4), the optimal and economic UV/TiO₂ treatment condition of 4.0 g/L TiO₂ with UV exposure of 30 min was selected for the chlorination experiment.

The changes of the percentage of fluorescence response (P_i,n) of DOM by the UV/TiO₂ treatment are shown in Figure 3. By treatment with 4.0 g/L TiO₂ under UV for 30 min, the primary components of tyrosine-like aromatic protein and soluble microbial by-product-like organic compounds were greatly reduced, thus P_i,n of regions I and IV decreased from 37.2% and 33.4% to 18.5% and 14.5%, respectively. UV/TiO₂ treatment was less efficient for degradation of DOM in regions II, III, and V, which were of much smaller amounts than the components in regions I and IV. This was in agreement with the findings of the raw seawater samples pretreated by ozonation in our previous study (Ding et al. 2018). UV/TiO₂ treatment in this condition achieved a reduction of EEM intensity higher than that with ozonation at 5 mg O₃/L but lower than that at 10 mg O₃/L.

The effect of treatment by 4.0 g/L TiO₂ under UV exposure for 30 min is shown in Figure 4. During chlorination, four types of HPs including 4-BP, 2,4,6-TCP, 2,4-DBP, and 2,4,6-TBP were formed in the raw seawater sample. The concentrations of 2,4,6-TCP and 2,4-DBP were constantly below 5 μg/L. The concentration of 4-BP was rapidly increased to 13.9 μg/L in the first two minutes of chlorination, and decreased to below 5 μg/L within 10 min. The concentration of 2,4,6-TBP was very much higher than the other HPs, which reached 687.4 μg/L at 4-min chlorination, and gradually decreased and fluctuated at nearly 400 μg/L at 30 min. By treatment of UV/TiO₂, only two types of HPs, 2,4-DBP and 2,4,6-TBP, were formed. UV/TiO₂ treatment did not change much the concentration of 2,4-DBP formed during chlorination, while it greatly decreased the concentration of 2,4,6-TBP, which was reduced from nearly 400 to less than 1 μg/L. The chlorine consumption of seawater samples was also reduced after UV/TiO₂ treatment (Figure S2, Supplementary Material).

It is noted that the optimum dose of the catalyst TiO₂ varied in regard to the target contaminant and the aqueous matrix. For example, the degradation of the dye Fast Green continuously increased as TiO₂ increased from 0.5 to 4.0 g/L, while the degradation of Acid Blue reached the optimum at TiO₂ of 2.0 g/L (Saquib et al. 2008). In this study, the optimal dose of TiO₂ (0.5–10.0 g/L) for HP degradation was 4.0 g/L. However, when considering industrial application, the cost of catalysts must be taken into account for a tradeoff.

In this study, UV/TiO₂ pre-treatment has been assessed to control the formation of halophenol in seawater during chlorination. TiO₂ dose and UV exposure time both affected the removal of UV₂₅₄ and DOM, and the optimal condition was determined to be 4.0 g/L TiO₂ with UV exposure of 30 min. The primary components of tyrosine-like aromatic protein and soluble microbial by-product-like organic compounds were greatly degraded under this condition. By UV/TiO₂ pretreatment, 4-BP and 2,4,6-TCP disappeared, and 2,4,6-TBP was significantly reduced from more than 400 to lower than 1 μg/L during chlorination. It is suggested that UV/TiO₂ may serve as a potential pretreatment approach for control of halophenol formation in seawater during chlorination.

This study was funded by the China Postdoctoral Science Foundation (2019M650493) and Annual Beijing Scientific Research Plan Project (KM201810011011).

The Supplementary Material for this paper is available online at https://dx.doi.org/10.2166/ws.2019.173.