2-Methyl-4-chlorophenol (MCP) is the main transformation intermediate of the herbicide MCPA in surface waters and it is more toxic than its parent compound. MCP is produced from MCPA by both direct photolysis and •OH reaction. The latter process has higher yield of MCP from MCPA: 0.5 vs. 0.3 for the direct photolysis. Our model results show that the formation rate of MCP would be higher in waters that contain low organic matter and high nitrate and nitrite. Such conditions are favourable to MCPA direct photolysis and •OH-induced transformation, which are both inhibited by organic matter, while •OH formation is enhanced by nitrate and nitrite. Good agreement is obtained between model predictions and field data of MCPA transformation in the Rhône river delta (Southern France). The field data also suggest that MCP undergoes slightly faster transformation than MCPA in that environment.
The occurrence of xenobiotic compounds depends on emission by human activities and on the transformation that these molecules undergo in the environmental compartments. Sunlight-driven processes play a considerable role in the attenuation of xenobiotics in surface waters, and they are comparatively more important when biodegradation reactions are slow (Boreen et al., 2003). Photochemical reactions in surface waters can be broadly divided into direct photolysis and indirect phototransformation. The direct photolysis is the transformation of a compound that absorbs solar radiation; it proceeds via bond breaking or ionisation, for example. In some cases radiation absorption is followed by inter-system crossing (ISC) to an excited triplet state, which may undergo transformation reactions directly or upon interaction with other molecules, including the solvent. The probability of an absorbed photon to induce transformation is measured by the photolysis quantum yield (Braslavsky, 2007).
Indirect phototransformation involves reaction with transient species, produced upon absorption of solar radiation by photoactive molecules called photosensitisers. The main photosensitisers in surface waters are chromophoric dissolved organic matter (CDOM) and the ions nitrite and nitrate (Al Housari et al., 2010). All these species can produce •OH radicals upon radiation absorption. In the case of CDOM, part of the •OH photoproduction could be accounted for by photogeneration of H2O2 followed by photolysis or the Fenton reaction (Richard et al., 2007; Mostofa and Sakugawa, 2009). Irradiated CDOM also yields excited triplet states (3CDOM*) and singlet oxygen (1O2). Oxidation of HCO−3 and CO2−3 by •OH (and to a lesser extent by 3CDOM*) yields the carbonate radical (Canonica et al., 2005). In summary, the transients •OH, 3CDOM*, 1O2 and CO3−• are key players of indirect photochemistry in surface waters.
The direct or indirect photochemical transformation yields intermediates that are often less harmful than the parent compound, but which in some cases can have higher environmental and/or health impact. Therefore, a complete assessment of the environmental attenuation of xenobiotics requires taking into account their transformation intermediates (Buth et al., 2009).
MCPA is a widely used phenoxy herbicide, active against broad-leaf weeds. It is an acid with pKa = 3.1, thus its anionic form usually prevails in surface waters (Iglesias et al., 2010). The main phototransformation pathways of MCPA in the environment are the reaction with •OH (second-order rate constant of 6.6 × 109 M−1 s−1; Benitez et al., 2004) and the direct photolysis. The limited absorption of solar radiation by MCPA is compensated for by the elevated photolysis quantum yield ΦMCPA, which is around 0.5 for the anionic form (Zertal et al., 2001). It has recently been found that ΦMCPA is decreased by dissolved organic compounds. MCPA photolysis proceeds via formation of the triplet state (3MCPA*) and subsequent reactions. However, 3MCPA* can oxidise dissolved organic compounds and produce MCPA−•, which can give back MCPA in the presence of oxygen (see reactions 1–4; S is a dissolved organic substrate). Reactions (3) and (4) would decrease the observed transformation rate of MCPA (Vione et al., 2010).
The main phototransformation intermediate of MCPA is MCP, which is formed by MCPA direct photolysis and reaction with •OH (Zertal et al., 2001). MCP has been detected in rice fields in the Rhône delta (S. France), after application of MCPA as post-emergent herbicide (Chiron et al., 2009). Both MCP and MCPA are ATPase inhibitors, but MCP is significantly more effective (Duchnowicz et al., 2005) and its formation might account for the observed increase of the toxicity of MCPA solutions upon irradiation (Zertal et al., 2001). As a potentially harmful intermediate of MCPA photochemical transformation, MCP should be considered as a secondary pollutant of environmental concern.
The purpose of this article is to carry out a model assessment of MCP formation from MCPA in natural waters. A model approach is adapted based on the known data of MCPA phototransformation and on water chemical composition and depth. Model predictions are compared with the available field data.
Methods including the photochemical model
The field data of MCPA and MCP used in this work are reported in Chiron et al. (2009) and are referred to Rhône delta lagoons (S. France). The transformation of anionic MCPA in surface waters would mainly take place via direct photolysis and reaction with •OH (Vione et al., 2010) and can be quantitatively assessed by means of a photochemical model (Maddigapu et al., 2011). Model application to MCPA yields its pseudo-first order transformation rate constants (kMCPA,•OH and kMCPA,DP for •OH reaction and direct photolysis, respectively) as a function of water chemical composition and column depth. Input data are the bimolecular reaction rate constant between MCPA and •OH (6.6 × 109 M−1 s−1; Benitez et al., 2004) and ΦMCPA. The latter depends on the organic matter content of the solution (expressed as dissolved organic carbon [DOC] or non-purgeable organic carbon [NPOC]), as follows (Vione et al., 2010):
MCP is formed from MCPA by both direct photolysis and •OH reaction. The respective yields are ηMCP,DP = 0.3 and ηMCP,•OH = 0.5 (Zertal et al., 2001). The pseudo-first order formation rate constant of MCP is kMCP = ηMCP,•OH kMCPA,•OH + ηMCP,DP kMCPA,DP and the overall yield is ηMCP = kMCP (kMCPA)−1, where kMCPA = kMCPA,•OH + kMCPA,DP. The values of kMCP and ηMCP depend on water chemical composition and depth.
The time unit adopted by the model is the SSD (Summer Sunny Day), equivalent to a fair-weather 15 July at 45°N (Maddigapu et al., 2011). The model can also give the steady-state concentrations of the reactive transients (•OH, CO3−•, 1O2, 3CDOM*) under definite irradiation conditions (e.g. 22 W m−2 UV irradiance) (Minella et al., 2011). For MCPA, [•OH] would be the most significant concentration because the reactions with CO3−•, 1O2 and 3CDOM* are less important than that with •OH (Vione et al., 2010).
Results and Discussion
The concentration of the herbicide MCPA (CMCPA) was reduced to one third of its initial value in 15 days in a Rhône delta lagoon, having the following features: average water depth d = 1 m, 51 μM nitrate, 3 μM nitrite, 2.1 mM bicarbonate, 26 μM carbonate and 4.5 mg C L−1 NPOC (Chiron et al., 2009). Under the hypothesis of a first-order degradation kinetics of MCPA, one obtains a rate constant k′ = 0.07 day−1. By comparison, the model predicts kMCPA = 0.06 SSD−1, with good agreement if 1 day = 1 SSD, which is reasonable because field data were collected in June-July. The model also foresees that about 55% of MCPA transformation would be accounted for by reaction with •OH and the remaining 45% by direct photolysis. Two sample cases of the model results for MCP formation are now reported.
Figure 1a reports the modelled kMCP vs. nitrite and NPOC. The other parameters are the same as for the Rhône delta lagoon described above. Dissolved organic matter (DOM and its chromophoric fraction CDOM), quantified as NPOC, would considerably inhibit MCP formation by decreasing both kMCPA,•OH and kMCPA,DP. In fact, DOM scavenges •OH and decreases ΦMCPA (Equation (5)). CDOM competes for irradiance with nitrate, nitrite and MCPA, inhibiting both •OH formation and MCPA direct photolysis.
The trend of [•OH] under the conditions of Figure 1a and with a constant 22 W m−2 UV irradiance is reported in Figure 1b. The [•OH] trend reflects the fact that nitrite is an important •OH source and organic matter a key •OH scavenger. kMCP (Figure 1a) and [•OH] (Figure 1b) have similar trends because the •OH reaction is an important pathway under the reported conditions ([•OH] = (5–20) × 10−16 M). Figure 1b shows a residual [•OH] without NO−2. Indeed, •OH is also produced by CDOM (partly via H2O2 formation as already explained; Richard et al., 2007; Mostofa and Sakugawa, 2009) and, at a lesser extent, by nitrate.
Figure 2 shows the trends of kMCP and [•OH] as a function of nitrite and NPOC. The adopted values of d, bicarbonate and carbonate are the same as in Figure 1, but nitrate is lower (1 μM) and nitrite is varied in a range of lower concentrations. Under such conditions, most of •OH would be produced by CDOM. Figure 2b shows that [•OH] increases with NPOC at low nitrite and decreases at high nitrite. In this system organic matter would be both an important source and scavenger of •OH, but at low nitrite the •OH production by organic matter would prevail over the scavenging, while the opposite case would occur at high nitrite. This accounts for the opposite trend of [•OH] vs. NPOC in the two cases.
Comparing Figures 2a and b, one sees rather different trends of kMCP and [•OH]. The main reason is that here ([•OH] = (1–6) × 10−17 M) the direct photolysis would strongly prevail over •OH reaction as MCPA transformation pathway, and organic matter would always inhibit MCPA photolysis.
Figure 3 reports the yield ηMCP = kMCP (kMCPA)−1, for the conditions depicted in Figure 1,(3a) and Figure 2,(3b). Because ηMCP,DP = 0.3 and ηMCP,•OH = 0.5, ηMCP is expected to vary in the 0.3–0.5 range. It would reach the lower limit when the direct photolysis is the main phototransformation pathway, the upper limit when •OH reaction prevails. Intermediate values mean that the two processes are both important. Figure 3 shows that ηMCP does not approach 0.5, implying that the direct MCPA photolysis is never negligible. Nitrite enhances ηMCP because of the increasing importance of the •OH reaction; ηMCP is poorly dependent of NPOC because organic matter inhibits both •OH reaction and direct photolysis.
In the Rhône delta lagoon (water data already reported), in 15 days the MCPA concentration (CMCPA) decreased from 1.2 to 0.4 nM, MCP concentration (CMCP) from 0.84 to 0.35 nM (Chiron et al., 2009). The model predicts ηMCP = 0.41. It is possible to elaborate a kinetic equation system for the rates of MCPA and MCP. The formation rate of MCP (d(CMCP)/dt) would be a fraction ηMCP of the MCPA transformation rate, d(CMCPA)/dt. MCP would also undergo degradation with rate constant k″:
From Equation (6) one gets , where C°MCPA is the initial MCPA concentration. From the field data (C°MCPA = 1.2 nM, CMCPA = 0.4 nM, t = 15 days) one gets k′ = 0.07 day−1. By substituting into (7) and by solving the differential equation one gets , where C°MCP = 0.84 nM is the initial MCP concentration. With C°MCPA = 1.2 nM, CMCP = 0.35 nM, ηMCP = 0.41, k′ = 0.07 day−1 and t = 15 days one gets k″ = 0.1 day−1. MCP would thus undergo faster degradation in the Rhône delta lagoon than the parent MCPA (k″ > k′). This is reasonable because, for instance, the reaction rate constant between MCP and •OH is higher than that of MCPA (Bojanowska-Czajka et al., 2007).
In sunlit surface waters, the formation of MCP from MCPA would mainly take place upon direct photolysis and reaction with •OH. The MCP yield from MCPA in the latter process (0.5) is higher than for the direct photolysis (0.3). By means of a model approach it is shown that the pseudo-first order formation rate constant of MCP from MCPA, kMCP, is higher at low NPOC values and at elevated concentrations of nitrate and, most notably, nitrite. Indeed, dissolved organic matter inhibits MCPA transformation: CDOM competes for irradiance with MCPA, nitrate and nitrite; DOM decreases ΦMCPA and consumes •OH that is produced by CDOM, nitrite and nitrate. The MCP yield from MCPA (ηMCP = kMCP (kMCPA)−1) is usually included in the 0.3–0.4 range. Very good agreement is obtained between model predictions and field data of MCPA phototransformation in the Rhône delta waters (Southern France). Field data also suggest that MCP transformation is a bit faster compared to the parent MCPA (k″ = 0.1 day−1 vs. k′ = 0.07 day−1).
Financial support by PNRA – Progetto Antartide is gratefully acknowledged. The Ph.D. grant of Elisa De Laurentiis is provided by Progetto Lagrange – Fondazione CRT, Torino, Italy. Davide Vione acknowledges financial support from Università di Torino-EU Accelerating Grants, project TO_Call2_2012_0047 (Impact of radiation on the dynamics of dissolved organic matter in aquatic ecosystems-DOMNAMICS).