Delineating the Role of Direct and Indirect Photolysis on Trichloroacetaldehyde (TCAL) and Dichloroacetonitrile (DCAN) in Water Degradation by Ultraviolet Irradiation

Abstract

Haloacetaldehydes (HALs) and haloacetonitriles (HANs) are groups of carcinogenic disinfection by products (DBPs) present in water supplies, of which trichloroacetaldehyde (TCAL) and dichloroacetonitrile (DCAN) are frequently detected. The efficiency of ultraviolet (UV) irradiation processes in the removal of DBPs depends strongly on the contribution of direct and indirect photolysis. Significant gaps exist in research regarding kinetics of photodegradation in multi-solute systems. Therefore, in this study the efficiency of vacuum UV (VUV) and UV-C processes was tested on batch photodegradation with synthetic waters containing either TCAL or DCAN and bi-solutes. A radical scavenger test was performed to determine the presence of OH radicals. The VUV (185 + 254 nm) degraded TCAL and DCAN more effectively than UV-C (254 nm), achieving absolute elimination after 30 min (>99.9%, 113 mW/cm²) for TCAL, but only an 84% reduction in DCAN after 120 min of irradiation at fluence of >450 mW/cm². The experimental results demonstrate that the main mechanism in TCAL reduction was indirect photolysis, but for DCAN it was direct photolysis by VUV photolysis. When indirect photolysis dominated, HALs and HANs in the mixture competed for OH radicals under VUV photolysis. A degradation pathway study indicated that TCAL was degraded and transformed to formic acid, while DCAN was dechlorinated by OH radicals. Overall, this study confirms that the VUV process is more effective than UV-C in photodegrading carbonaceous DBPs.

1. Introduction

Disinfection by-products (DBPs) are formed by reactions between organic compounds and disinfectants. Nowadays, disinfection processes commonly use chlorination because of its high efficiency, low cost, and continuing effectiveness during distribution due to residual chlorine in the water. However, a disadvantage of chlorination is the formation of halogenated DBPs, as indicated by numerous reports regarding the occurrence of DBPs in chlorinated water supplies around the world [1,2]. DBPs can be divided into two groups: carbonaceous-DBPs (C-DBPs), such as trihalomethanes (THMs), haloacetic acids (HAAs), halopropanones, and haloacetaldehydes (HALs); and nitrogenous-DBPs (N-DBPs), such as nitrosamines, halonitromethanes, and haloacetonitriles (HANs) [3,4]. Nowadays, more than 800 species of DBPs were reported to be found in drinking water [5,6,7,8]. Despite presenting in low levels in drinking water or tap water, DBPs show adverse human health impacts from long-term exposure, such as carcinogenic, genotoxic, mutagenic, and cytotoxic effects [8]. HALs are a C-DBP group that raises increasing concern regarding its toxicity due to a high cytotoxicity index [9]. HANs have been determined to be the highest proportion of the N-DBP group in chlorine DBPs in drinking water [10]. The World Health Organization (WHO) reports that some HANs are type 2A, 2B, and C carcinogens [11]. HALs and HANs have many derivative species—10 and seven species, respectively. Both groups are unregulated in drinking water in many countries in the world. Therefore, it is crucial for waterworks to efficiently eliminate unregulated DBPs, such as HALs and HANs, to ensure safe drinking water.

The average concentrations observed in drinking water distribution systems (DWDSs) and swimming pool waters (SPWs) of the most frequently detected HALs and HANs, namely trichloroacetaldehyde (TCAL) (or chloral hydrate (CH)) and dichloroacetonitrile (DCAN), are about 3.67 for TCAL and 3.08 for DCAN in DWDS, and 1536 for TCAL and 43 for DCAN in SPW in µg/L in the USA [12]. In Canadian drinking water systems, the total concentration of HALs was found to be about 3.5–28.5 µg/L, and for the major species of HALs, TCAL, it was about 12.0 µg/L [1]. The highest concentration of TCAL was 1.7 to 46 µg/L and the highest concentration of DCAN was 0.6 to 24 µg/L in drinking water [13,14]. The concentration ranges of HALs and HANs are regulated at about 1–200 µg/L in some countries. Among the unregulated DBPs in drinking water, TCAL and DCAN were found in the highest concentrations [14].

The process for eliminating HALs and HANs involves attempting to remove their precursor prior to disinfection, for instance, removing natural organic matter (NOM) through adsorption processes, ozonation, and membrane filtration. However, a significant short-coming of some NOM removal processes is a greater increase in DBP formation than from disinfection alone. For example, in four drinking water treatment plants (DWTPs) in China, applying pre-ozonation resulted in an increase in HALs formation of 1.66–1.63 times compared to chlorination of raw and treated water [15]. Moreover, the DBPs exist due to the interaction of remaining NOM with free chlorine in the water treatment process. Hence, the choice of a disinfection process, such as using chlorine or monochloramine, is one of the promising ways to avoid DBP formation. When comparing chlorine and monochloramine, it was determined that monochloramine reduced the concentration of THMs concentration by 10 times, but in turn DCAN formation was increased by five times [16]. Therefore, it is imperative the DBPs removal process follows after formation. It includes adsorption by activated carbon, high-pressure membrane filtration, oxidation process, and advanced oxidation processes (AOPs). One of the most popular processes for DBPs removal is AOPs because it can mineralize pollutants (convert pollutants to carbon dioxide and water), minimizing potential burden on other posttreatments and the environment.

AOPs are a set of chemical treatment processes designed to remove trace organic pollutants (and sometimes inorganic pollutants) in water and wastewater by oxidation through reactions between the pollutants and hydroxyl radicals (•OH). AOPs show good efficiency including high reactivity, powerful oxidation, and harmless characters of reactive species. AOPs involve many processes, such as the Fenton Process, ozonation, photolysis, and combined processes. Ultraviolet (UV) irradiation processes are usually considered to be relatively effective, sustainable, and environmentally friendly techniques [17]. UV photolysis has many advantages such as easy operation and suitability for low-concentration pollutants. The main mechanism of ultraviolet degradation consists of direct and indirect photolysis. Direct photolysis involves irradiation of a pollutant with ultraviolet light. Indirect photolysis consists of two processes: ultraviolet irradiation of water induces the formation of hydroxyl and hydrogen radicals, and hydroxyl radicals then oxidize the pollutant.

The UV wavelength affects degradation of pollutants because UV-induced degradation will occur under suitable conditions while reaching sufficient energy to break the bonds of pollutants [18]. The electromagnetic energy of VUV (100–200 nm) is greater than UV-C (200–280 nm), UV-B (280–315 nm), and UV-A (315–400 nm). Electromagnetic energy will increase as the emission wavelengths decrease [19]. Removal efficiency of UV-C and VUV during degradation of imipramine (IMI) reached 64.7% and 97%, respectively, and direct photolysis was identified as a major degradation mechanism [17]. Solar irradiation can remove tribromoacetaldehyde (TBAN) and DCAN by 38% and 46%, respectively [3].

Degradation of DCAN, trichloroacetonitrile (TCAN) and dibromoacetonitrile (DBAN) by UV-C lamp in 24 h has been determined to be about 29%, 84% and 94%, respectively [20]. Applying different wavelengths of VUV and UV-C while irradiating water has resulted in different degradation performance levels, and VUV photolysis is more effective in removing HANs than UV-C [21]. For the degradation of TCAL in water by UV photolysis, TCAL was decreased by 67% with a 12-watt UV-C lamp over 8 h [22]. Nonetheless, significant gaps exist in research regarding kinetics of photodegradation in multi-solute systems.

This study focuses on processes for degrading HALs and HANs in water. Based on existing research, it was anticipated that applying UV-C or VUV will reduce levels of TCALs and DCANs with different efficiencies, depending on the combined presence of TCALs and DCANs, and whether direct or indirect photolysis is involved. Degradation pathways of TCALs and DCANs were studied in the course of the experiments. In terms of analytical method, experiments were conducted in a closed cylindrical reactor, mixtures were analyzed using gas chromatography, and the degradation mechanism was determined from a radical capture procedure. The experimental process is detailed in the following section.

2. Materials and Methods

2.1. Reagent and Chemicals

Herein, TCAL is represented by the HALs group and DCANs are represented by the HANs group. TCAL (>99.0%, TCI, Tokyo, Japan), DCAN (99.0%, Sigma-Aldrich, MO, USA) and Acetone (HPLC grade, Macron fine chemicals, PA, USA) were used to prepare synthetic water. No buffer solution was added. Sodium sulfate anhydrous (crystal, Carlo erba reagents, Val de Reuil Cedex, France) and tert-Butyl methyl ether (MtBE, HPLC grade, 99.7%, Loba chemie, Mumbai, India) were used for liquid–liquid extraction by US EPA method 551.1. Tert-Butanol (TBA, 99.5%, Penreac, Darmstadt, Germany) was used in the radical capture experiment to study the degradation mechanism. DCAN stock solution was prepared in acetone. Synthetic water samples consisting of deionized water were prepared (resistivity > 18.2 MΩ·cm), and mixtures of TCAL or DCAN were prepared, each with an initial concentration of 100 ppb in water.

2.2. Experimental Setup

The experiments were done in a closed cylindrical reactor (capacity: 8 L), as shown in Figure 1. The reactor contained a quartz tube (22 × 25 × 365 mm) with a 14-watt low-pressure VUV (LP-VUV) lamp, which was composed of VUV emitter at 185 nm (10% of UV-C emitting photon flux at 254 nm, Heraeus, model GPH287T5VH/4) and a low-pressure UV-C (LP-UV-C, 254 nm, Heraeus, model GPH287T5L) lamp. Actual intensity of the low-pressure UV (LP-UV) lamps was determined with ferrioxalate actinometry [23,24]. The photon flux for VUV (185 + 254 nm) and UV-C (254 nm) was calculated as 0.364 and 0.063 mW/cm², respectively. The reactor was made from stainless steel with an agitator to vigorously mix the solution over the period of reaction. Exactly eight liters of synthetic water were added into the reactor before turning on the UV lamp. Samples of 100 mL were collected at 0, 5, 15, 30, 45, 60, 75, 90, 105, and 120 min during irradiation. The final volume was always greater than 85% of the initial volume. All experiments were conducted in triplicate.

2.3. Analytical Method

TCAL and DCAN concentrations were analyzed using a modified US EPA method 551.1, liquid–liquid extraction (LLE), and a gas chromatograph equipped with an electron capture detector (GC/ECD, Agilent GC7890, CA, USA) [25]. The LLE used 30 mL of sample aliquot with 2 mL of MtBE and 5 g of sodium sulfate anhydrous to improve the phase separation. The degradation intermediates of HALs and HANs resulting from VUV were identified by GC/mass spectrometer (GC/MS, Agilent GC7890B CA, USA).

The gas chromatography (GC/ECD, Agilent GC7890, CA, USA) condition and parameters were set as follows. The velocity of helium-carrier gas was established at 25 cm/s and nitrogen was used for make-up gas. The fused silica capillary column, GC capillary column (Rtx-1) (0.32 mmID × 30 m × 1 µm crossbond^® 100% dimethyl polysiloxane coating, RESTEK, PA, USA), was used in the analysis of TCAL and DCAN. The injection and ECD detector temperatures were set at 200 °C and 300 °C (split mode). The initial programmed temperature of the column oven was set at 30 °C for 6.5 min. The limit of detection (LOD) for both DBPs was determined as 0.05 µg/L and the percentage of recovery from liquid–liquid extraction for TCAL was 89.53% and for DCAN it was 91.28% [11].

2.4. Radical Capture

The degradation mechanism was determined from TBA (scavenger adding) for •OH capture. As employed by Maniero et al. (2008), 10 mM of TBA was added to the initial solution (prior to irradiation) [25]. The removal efficiency and kinetic rate constants were compared with TBA and without TBA [22].

3. Results and Discussion

3.1. Degradation of HALs and HANs by UV-C and VUV Irradiation

Figure 2 shows the degradation of TCAL by VUV and UV-C. The half-life of TCAL during irradiation was approximately 15 min with VUV while complete degradation of TCAL was attained by VUV irradiation after 30 min. It is worth noting that the remaining TCAL concentration in the VUV-irradiated water was less than the detection limit (LOD = 0.05 µg/L) at 30 min. In contrast, UV-C reduced TCAL concentration by only 18% after 120 min irradiation at a fluence rate of 0.063 mW/cm². A recent study reported that the half-life of TCAL under UV-C using a LP-UV lamp was circa 5 h at a fluence rate of 6.37 mW/cm² [22]. Although TCAL can be readily degraded under UV-C, its degradation is governed by direct and indirect photolysis. Hence, the relative contributions of the reactions are solely dependent on the solution pH, UV lamp intensity, and its initial concentration. At elevated solution pH (i.e., alkaline pH), relatively more •OH formation occurred due to enhanced hydrolysis reactions, and thus the rate of hydrolysis became comparable to that of photolysis [8].

3.1.1. Degradation of TCAL Under VUV or UV-C Irradiation

The VUV wavelength ranged from 100 to 200 nm, typically showing a peak at 185 nm. While irradiated, water molecules absorbed VUV intensively at 185 nm to generate •OH (E⁰ = 1.8–2.7 V), which has been reported as a dominant radical species. VUV generates •OH more than UV-C [26]. It is worth noting that VUV at 185 nm and UV-C at 254 nm can be integrated into a VUV/UV-C system. This combination is proved to be more effective than either one alone for micropollutant degradation since both direct and indirect photolysis are responsible for micropollutant abatement [21,27]. In this study, VUV (i.e., VUV/UV-C) was apparently more effective in the TCAL reduction than UV-C. As presented in the inset of Figure 2 (Appendix A.1), based on the same fluence of circa 450 mW/cm², the TCAL concentration was reduced by approximately 18% by UV-C and 89% by VUV, demonstrating a relatively high efficiency of VUV irradiation in reducing TCAL. In fact, due to direct photolysis of water by VUV 185 nm irradiation, the VUV process has the most •OH generation pathways compared to the UV-C alone [17]. This is consistent with higher photon flux obtained from ferrioxalate actinometry for VUV.

In terms of reduction rate (time depending reduction), the kinetics constant rate of TCAL reduction (k_{TCAL_VUV} and k_{TCAL_UV-C}) is shown in

Table 1. The kinetic constant rate of TCAL and DCAN degradation by VUV and UV-C photolysis.

DBPs	Apparent First-Order Constants Rate
	kVUV (min−1)	R2	kUV-C (min−1)	R2
TCAL	0.1482	0.9495	0.0012	0.9484
DCAN	0.0147	0.9769	0.0007	0.9881

. The kinetic reduction data from both VUV and UV-C obeys the first-order reaction with the R-square of 0.9495 and 0.9484, respectively. k_{TCAL_VUV} was about two orders of magnitude greater than k_{TCAL_UV-C}, indicating that the VUV lamp emitted a higher energy that degraded TCAL more than the UV-C lamp. This finding can be explained by the electromagnetic radiation concept in that energy carried by irradiation increases with decreasing wavelengths. VUV delivers more energy for breaking bonds and more photons that produce •OH in water than UV-C [18,19,28]. Therefore, considering energy efficiency, the appropriate UV treatment to eliminate TCAL is using VUV irradiation in comparison with UV-C. However, further exploration of the main degradation mechanisms of TCAL by photolysis and oxidation from reactive species like •OH is needed.

3.1.2. Degradation of DCAN Under VUV or UV-C Irradiation

Figure 3 shows the effect of the VUV lamp and UV-C lamp on the reduction in DCAN, one of most frequently detected halogenated nitrile (–C≡N) compounds. It is known that nitrogenous DBPs are more susceptible to solar irradiation than carbonaceous DBPs, while compounds with a high degree of halogenation are less stable than less-halogenated species [3]. It is then anticipated that DCAN is less susceptible than TCAL to photodegradation by irradiation based on its lesser degree of halogenation. However, the presence of nitrile moiety in DCAN might make it more susceptible to photolysis than a compound containing a carbonyl moiety like TCAL. The pattern of DCAN photodegradation is apparently different from that of TCAL. Over 120 min of irradiation from the UV-C lamp, DCAN concentrations showed a reduction (by 8%) while the VUV lamp offered greater efficiency in reducing DCAN (by 84%). This finding shows consistency in the effect of electromagnetic radiation and the fact that DCAN absorbs the incident UV photon at the specified wavelength to cause an electronically excited molecule. The result confirms previous studies of the effect of UV-C wavelength radiation. A VUV lamp degraded organic pollutants better than UV-C [17,20,21,28]. In terms of kinetics, the kinetics constant rate of DCAN (k_{DCAN_VUV} and k_{DCAN_UV-C}) are shown in Table 1. Reactions involving TCAL or DCAN with VUV irradiation were first-order reactions.

In terms of DBP compounds, the VUV irradiation exhibits a greater degradation of TCAL (absolutely eliminated after 30 min, >99.9%) than DCAN (reduced to 84% at 120 min irradiation). These results imply that VUV treatment offers safe drinking water with concentrations of both TCAL and DCAN below the guideline values for drinking water, which are 10 ppb and 20 ppb, respectively, from WHO guidelines [6,11]. From the quantitative structure–activity relationship (QSAR) modeling, the high degree of halogenated species was less stable than less-halogenated species (trihalogenated > dihalogenated > monohalogenated) [3]. In terms of reduction rate, the kinetics constant rate of DCAN reduction (k_{DCAN_VUV} and k_{DCAN_UV-C}) is shown in Table 1. The kinetic reduction data from both VUV and UVC obeys the first-order reaction with the R-square of 0.9769 and 0.9881, respectively. The k_{DCAN_VUV} was about 21 times greater than k_{DCAN_UV-C}. Considering susceptivity of compound structures, DCAN should be photo-degraded more rapidly than TCAL. However, the opposite trend was observed, which calls for further exploration for contributing mechanisms, apart from photolysis alone.

3.2. Mechanisms of TCAL and DCAN Degradation

To confirm the main photochemical reactions contributing to the reduction in TCAL and DCAN during VUV irradiation, the degradation of TCAL (representing HALs group from C-DBPs) and DCAN (representing HANs group from N-DBPs) was investigated in the presence and absence of radical capture reagent. In this study, TBA was added to quench •OH formation in the VUV photolysis process to confirm the contribution of •OH in DBP degradation under indirect photolysis.

Herein, an excess amount on a molar basis of TBA was added to the TCAL- and DCAN-containing water. Figure 4 presents the radical capture result while the kinetic rate constant of DCAN degradation is shown in

Table 2. The kinetic constant rate of TCAL and DCAN degradation by VUV photolysis with TBA and without TBA.

DBPs	Apparent First-Order Constants Rate
	kwithout TBA (min−1)	R2	kwith TBA (min−1)	R2
TCAL	0.1482	0.9495	0.0015	0.8962
DCAN	0.0147	0.9769	0.0105	0.9764

. Adding TBA to capture •OH substantially lowered the reduction rate of TCAL. Namely, TCAL reduction in the TBA-added sample was only 3% after 30 min, lower than that tested in the TBA-free sample (99%). After 120 min, TCAL reduction was 20.0% for the TBA-added sample. The k_{TCAL_VUV} for the TBA-added sample was about two orders of magnitude smaller than the TBA-free sample.

Based on the mechanism study, it can be inferred that the main mechanism of TCAL degradation was solely indirect photolysis through •OH, self-generated in water [21,29,30,31]. Therefore, the oxidation process with the •OH formation greatly influences the TCAL degradation. However, Gan et al. (2019) reported that the indirect oxidation of TCAL at a solution pH of 7.0 under UVC irradiation shared a nearly similar contribution, in comparison with the direct photolysis [29]. It should be noted that •OH recombination to H₂O₂ is negligible in the neutral and alkaline pH conditions [22]. The difference in the relative contribution of indirect photolysis observed in this study might be related to the UV source (i.e., UV-C vs. VUV) and initial solution pH (e.g., pH ~7.0 vs. pH > 8.0 in this study) which each differently promote the •OH generation.

The DCAN reduction during VUV in the TBA-added sample shows a similar reduction pattern as the TBA-free sample (Figure 4), while the kinetic constant rate, k_{DCAN_VUV}, for the TBA-added sample was slightly lower as compared with the TBA-free sample (~16% lowered) (Table 2), indicating a minor contribution of •OH. Therefore, the main mechanism of DCAN removal at the VUV irradiation was mainly direct photolysis, in line with previous a study of DCAN removal by VUV photolysis which reported the contribution of •OH at about 4% [21].

3.3. Competitive Effect of HALs and HANs Mixture Under VUV

For mixtures of TCAN and DCAN, the results for competitive effects are shown in Figure 5a,b, which present the degradation of each DBP in the single and mixed solutions by VUV. In the mixed solution, the DCAN removal noticeably decreased in comparison with the single solution removal. Namely, after 120 min the reduction was 84% for the single solution but decreased to 68% for the mixed solution. Photodegradation of DCAN under VUV resulted in the formation of several products, probably monochloroacetonitrile (MCAN) through dichlorination of DCAN and haloacetamides through hydrolysis reactions [21]. Therefore, the mixed solution was found to have a competitive effect between groups of by-products formed during the VUV photolysis.

On the other hand, in the TCAL mixture the removal level was most different from the TCAL single removal level. The TCAL level in the single solution was degraded about 99.0% at 30 min but TCAL in the mixed solution was degraded not over 60.0% at 120 min. Therefore, the effect of the TCAL mixture was competitive under VUV photolysis when mixing DBP groups. The result is in accord with the main mechanism (indirect photolysis) observed in this study.

The results of the investigation of the kinetics of degradation were plotted in a graph ln(C/C₀) versus times. The trendline degradation of TCAL and DCAN was linear. The first-order rate constants for degradation by VUV were degraded as shown in

Table 3. The kinetic constant rate of TCAL and DCAN degradation by VUV photolysis in mixture.

DBPs	Apparent First-Order Constants Rate
	ksingle (min−1)	R2	kmixture (min−1)	R2
TCAL	0.1482	0.9495	0.0060	0.9814
DCAN	0.0147	0.9769	0.0097	0.9773

. From these results, the kinetic constant rate of VUV confirms the effects on removal of HALs and HANs in mixtures.

3.4. Mechanism Identification

In terms of pathway mechanisms, the intermediates were formed from photodegradation and •OH oxidation mechanisms. In TCAL degradation, TCAL was degraded and transformed to formic acid. Figure 6a shows mechanisms of compound degradation induced by UV-C (254 nm) photolysis [22]. The pathway of DCAN degradation was MCAN, 2-chloropropinonitrile, and dichloroacetamide (DCAcAm). In Figure 6, the intermediates of TCAL pathways are adapted from a previous study [21]. QSAR modeling determined that halogenated functional groups cause greater degradation under UV photolysis [3]. Based on the result in this study, TCAL and DCAN degradation by VUV irradiation and •OH oxidation have been confirmed. The results show the kinetics of reactions affecting DCAN are faster than TCAL in mixed solutions. It is affected by •OH oxidation to a greater extent than VUV photolysis. The contribution ratio of indirect (OH radical) photolysis to direct photolysis was determined as TCAL 99%:1% and DCAN 29%:71% (Supplemental Text S3). Therefore, the VUV lamp can improve indirect photolysis from photolysis of water.

4. Conclusions

This study explored TCAL and DCAN degradation in water under VUV (185 + 254 nm) and UV-C (254 nm) irradiation. The reductions in TCAL and DCAN were studied in terms of removal efficiency, main mechanism (direct/indirect photolysis), competition between DBPs group, and proposed intermediates pathways. The major findings are as follow:

• VUV light was more effective than UV-C light at degrading TCAL and DCAN.
• The main mechanism of DCAN removal was direct photolysis, but TCAL removal was driven by indirect photolysis and adding TBA to quench •OH formation.
• Degradation of HALs and HANs in mixtures under VUV irradiation was determined by competition for OH radicals.
• Competition in the DBP mixture group occurred under conditions of indirect photolysis.

Future research may further explore and define the contributions of direct and indirect photolysis in DBP degradation during VUV irradiation for point-of-use applications.