Removal of 2-MIB and geosmin using UV/persulfate: Contributions of hydroxyl and sulfate radicals

Highlights

• Degradation of 2-MIB and geosmin by UV/persulfate was evaluated for the first time.

• A model was set up to study the kinetics in UV/persulfate process for the first time.

• Impacts of different conditions including pHs and dosages of persulfate were discussed.

• Different scavengers including NOM and bicarbonate ions were discussed.

Abstract

2-methylisoborneol (2-MIB) and geosmin are two odor-causing compounds that are difficult to remove and the cause of many consumer complaints. In this study, we assessed the degradation of 2-MIB and geosmin using a UV/persulfate process for the first time. The results showed that both 2-MIB and geosmin could be degraded effectively using this process. The process was modeled based on steady-state assumption with respect to the odor-causing compounds and either hydroxyl or sulfate radicals. The second order rate constants for 2-MIB and geosmin reacting with the sulfate radical (${\text{SO}}_4^{-}$) were estimated to be (4.2 ± 0.6) × 10⁸ M⁻¹ s⁻¹ and (7.6 ± 0.6) × 10⁸ M⁻¹ s⁻¹ respectively at a pH of 7.0. The contributions of the hydroxyl radical (OH) to 2-MIB and geosmin degradation were 3.5 times and 2.0 times higher, respectively, than the contribution from ${\text{SO}}_4^{-}$ in Milli-Q water with 2 mM phosphate buffer at pH 7.0. The pseudo-first-order rate constants ($k_{\text{o}}^{\text{s}}$) of both 2-MIB and geosmin increased with increasing dosages of persulfate. Although pH did not affect the degradation of 2-MIB and geosmin directly, different scavenging effects of hydrogen phosphate and dihydrogen phosphate resulted in higher values of $k_{\text{o}}^{\text{s}}$ for both 2-MIB and geosmin in acidic condition. Bicarbonate and natural organic matter (NOM) inhibited the degradation of both 2-MIB and geosmin dramatically through consuming OH and ${\text{SO}}_4^{-}$ and were likely to be the main radical scavengers in natural waters when using UV/persulfate process to control 2-MIB and geosmin.

Introduction

Cyanobacteria or blue-green algae can produce 2-methylisoborneol (2-MIB) and geosmin which result in many complaints of taste and odor in drinking water from consumer (Watson et al., 2008, Srinivasan and Sorial, 2011). The concentration of the two odor-causing compounds can even reach more than 1000 ng/L during algae bloom (Mizuno et al., 2011). The guidelines of either 2-MIB or geosmin for drinking water in China, South Korea and Japan are set at 10 ng/L (MOH and SAC, 2006, Koester, 2011). Due to their extremely low odor threshold concentrations (several ng/L), conventional water treatment processes using coagulation, sedimentation, sand filtration and chlorination are ineffective in removing these compounds (Srinivasan and Sorial, 2011, Antonopoulou et al., 2014). As a result, additional eliminating processes such as oxidation, adsorption or membrane filtration are generally required for such taste and odor problems (Srinivasan and Sorial, 2011, Agus et al., 2011). Sedlak and co-researchers suggest that reverse osmosis membrane, ozonation followed by biological activated carbon and advanced oxidation technologies (AOTs) such as UV combined with hydrogen peroxide (UV/H₂O₂) could remove odors from water effectively (Agus et al., 2011). AOTs which typically involve strategies for generating hydroxyl radicals (OH), due to high redox potential (1.9–2.7 V), are efficient in degrading such pollutants (Buxton et al., 1988). Examples of such AOTs including UV/H₂O₂, vacuum UV, O₃/H₂O₂, electrochemical oxidation and ultrasonic irradiation, all of which have been shown to degrade 2-MIB and geosmin effectively (Jo et al., 2011, Kutschera et al., 2009, Mizuno et al., 2011, Li et al., 2010, Song and O'Shea, 2007, Antonopoulou et al., 2014).

There has been recent interest in the generation of the sulfate radical (${\text{SO}}_4^{-}$) in AOTs due to its high redox potential of 2.5–3.1 V which is comparable to OH (Neta et al., 1988). However, ${\text{SO}}_4^{-}$ is more selective than OH while still reacting rapidly with many organic substrates (Neta et al., 1988). ${\text{SO}}_4^{-}$ can be generated from the activation of peroxydisulfate (S₂O₈²⁻) or peroxymonosulfate (HS₂O₅⁻) by using UV, heat, base, or transition metals (Lau et al., 2007, Guan et al., 2011, Waldemer et al., 2007, Furman et al., 2010, Zou et al., 2013, Zhang et al., 2013). Numerous studies have demonstrated that ${\text{SO}}_4^{-}$ can degrade endocrine disrupting compounds, chlorinated compounds, drugs, perfuluorinated compounds and algal toxins (Lau et al., 2007, Guan et al., 2011, Waldemer et al., 2007, Guan et al., 2013, Gao et al., 2012, Hori et al., 2005, Antoniou et al., 2010).

Persulfate irradiated by UV was proposed to generate ${\text{SO}}_4^{-}$ through reaction 1 in Table 1 with a quantum yield of 1.4 mol Es⁻¹ (254 nm) (Mark et al., 1990). In this process, ${\text{SO}}_4^{-}$ are produced with activation of persulfate without the production of OH (Mark et al., 1990). However, OH can also be generated when ${\text{SO}}_4^{-}$ reacts with water at a rate constant of 8.3 M⁻¹ s⁻¹ (reaction 4) (Yu et al., 2004), or with OH⁻ at a rate constant of 6.5 × 10⁷ M⁻¹ s⁻¹ under alkaline conditions (reaction 3) (Neta et al., 1988). Thus, ${\text{SO}}_4^{-}$ generation may be an attractive AOTs strategy for removing 2-MIB and geosmin. However, to date the degradation efficiency and rates by ${\text{SO}}_4^{-}$ have not been reported for 2-MIB and geosmin.

The objectives of this study were (1) to investigate the effects of water chemistry and kinetics of the removal of 2-MIB and geosmin by UV (254 nm)/persulfate; (2) to determine the reaction rates between 2-MIB (and geosmin) and ${\text{SO}}_4^{-}$; (3) and to model the removal of 2-MIB and geosmin in this process with water quality that is likely to be encountered in water treatment.