Enhanced transformation of aquatic organic compounds by long-lived photooxidants (LLPO) produced from dissolved organic matter

Dissolved organic matter (DOM) plays a crucial role in the photochemical transformation of organic contaminants in natural aquatic systems. The present study focuses on the characterization of a specific effect previously observed for electron-rich phenols, consisting in an acceleration of the DOM-photosensitized transformation of target compounds at low concentrations (< 1 µM). This effect was hypothesized to be caused by DOM-derived "long-lived" photooxidants (LLPO). Pseudo-first-order rate constants for the transformation of several phenols, anilines, sulfonamide antibiotics and phenylureas photosensitized by Suwannee River fulvic acid were determined under steady-state irradiation using the UVA and visible wavelengths from a medium-pressure mercury lamp. A significant enhancement (by a factor of 2.4 - 16) of the first-order transformation rate constant of various electron-rich target compounds was observed for an initial concentration of 0.1 μM compared to 5 μM . This effect points to a relevant reactivity of these compounds with LLPO. For phenols and anilines the enhancement effect occurred only above certain standard one-electron oxidation potentials. From these data series the standard one-electron reduction potential of LLPO was estimated to be in the range of 1.0 - 1.3 V versus the standard hydrogen electrode. LLPO are proposed to mainly consist of phenoxyl radicals formed by photooxidation of electron-poor phenolic moieties of the DOM. The plausibility of this hypothesis was successfully tested by studying the photosensitized transformation kinetics of 3,4-dimethoxyphenol in aqueous solutions containing a model photosensitizer (2-acetonaphthone) and a model electron-poor phenol (4-cyanophenol) as DOM surrogates.


S2
Text S1. Estimation of the lifetime of LLPO Two deactivation pathways for LLPO are considered. 1) A hypothetical unimolecular decay of LLPO, with first-order rate constant LLPO d,0 .
The pseudo-first-order rate constant attributed to this reaction is the product of the second- value of 3 × 10 9 M ˗1 s ˗1 for electron-rich phenols based on a review of available experimental second-order rate constants for organic peroxyl radicals (Neta et al. 1990) and phenoxyl radicals (Neta and Grodkowski 2005).
The overall deactivation of LLPO in the presence of TC can be described by the pseudo-first- This lifetime is in a similar range as estimated in a previous study (> 100 µs) (Canonica and Hoigné 1995).

Text S2. Estimation of the lifetime of 3 CDOM*
We refer here to the methodology used in a previous review article ( Rosario-Ortiz and Canonica 2016) and Scheme 1 in the main paper. Accordingly, the pseudo-first-order deactivation rate constant of 3 CDOM* in the presence of a target compound (TC) can be defined as: = 5 × 10 4 s -1 ). Schmitt et al. (2017)  = (3.8 − 8.3) × 10 4 s -1 ). Zepp et al. (1985) proposed pressure are expected to fall in the range of 1.6 − 6.3 µs. a All compounds were analyzed on a reverse-phase C18 column (COSMOSIL 5C18-MS-II packed column, pore size 120 Å, particle size 5 µm, internal diameter 3.0 mm, length 100 mm) with a column oven temperature of 25 °C (Agilent system) or 30 °C (Dionex system), a flow rate of 0.5 mL min -1 , and an injection volume of 100 µL. The employed measuring ranges for phenols, anilines and 2-acetonaphthone were 0.5 − 5 µM with the photodiode array detector, and 0.01 − 0.1 µM with the fluorescence detector, the measuring range for trolox, phenylureas and sulfonamides was 0.01 − 5.0 µM with the photodiode array detector. Standard deviations of measured concentrations were typically < 5 %. b Buffer composition: 10 mM H3PO4 in ultrapure water, pH=2.1. c In the presence of 2-acetonaphthone and 4-cyanophenol, the method described in note d was used, and the retention time

Text S5. Stability of 2-acetonaphthone and 4-cyanophenol during irradiation experiments of chemical model system solutions
The concentrations of the photosensitizer 2-acetonaphthone and of the electron-poor phenol 4-cyanophenol were also monitored during the irradiation experiments performed to study the transformation kinetics of 3,4-dimethoxyphenol (DMOP) in the chemical model system solutions. Figure S1 illustrates the typical evolution of concentration of DMOP, 2-acetonaphthone and 4-cyanophenol as a function of irradiation time for the highest (5.0 µM, Fig. S1a) and lowest (0.1 µM, Fig. S1b) initial concentrations of DMOP. In these two specific cases, after the whole irradiation period, the concentration of 2-acetonaphthone decreased by 7% and 1%, respectively, and the concentration of 4-cyanophenol decreased by 2% and 3%, respectively, compared to their initial concentration. These minor reductions, which are mostly even smaller than the analytical precision in the quantification of concentrations (≈ 5%), confirm the stability of 2-acetonaphthone and 4-cyanophenol under the employed experimental conditions. Therefore, no corrections were applied to the determination of the pseudo-firstorder rate constants k obs for the transformation of DMOP. The reactions considered for the modelling and explained in detail in the main paper, Section 3.4., are compiled in Table S8, together with details about the used rate constants.
Particular attention had to be paid to the determination of the first-order rate constant for the excitation of the photosensitizer 2-acetonaphthone (2-AN) to yield its excited triplet state Substituting eq. (S13) into eq. (S12) and solving for 2-AN* To calculate 2-AN* 3 f using eq. (S14), the values of the following three rate constants are needed: : The first-order deactivation rate constant of 3 2-AN* in aerated aqueous solution (see eq. (4) in the main paper) was determined as 6.44 × 10 5 s ˗1 in a previous study using laser flash photolysis (Canonica et al. 2000).

3)
2-AN* 3 ,DMOP r : This rate constant (corresponding to the reaction of eq. (5) in the main paper) was set equal to the second-order rate constant for the quenching of 3 2-AN* by DMOP (3.1 × 10 9 M ˗1 s ˗1 ) determined in a previous study by laser flash photolysis (Canonica et al. 2000). The validity of this approximation is supported by the high radical yields (approaching 1.0) determined for the quenching of the excited triplet states of aromatic ketones by phenoxides in water-acetonitrile solutions (Das and Bhattacharyya 1981).
The second-order rate constant for the reaction of 3 2-AN* with the 4-cyanophenoxide ion (4-CN-PhO − ) leading to the formation of the 4-cyanophenoxyl radical (4-CN-PhO ⦁ , eq. (6) in the main paper) was estimated as 2-AN* 3 ,4-CN-PhO − r = 1 ×10 9 M -1 s -1 based on the quenching rate constant determined for aqueous solutions of 4-cyanophenol at pH 6.0 (Canonica et al. 2000) and the high radical yields (approaching 1.0) determined for the quenching of the excited triplet states of aromatic ketones by phenoxides in water-acetonitrile solutions (Das and Bhattacharyya 1981). Note that the reaction of 3 2-AN* with undissociated 4-cyanophenol was neglected in the kinetic model because it is ≤ 10 7 M -1 s -1 (Canonica et al. 2000).
The second-order rate constants for the reactions of 4-CN-PhO ⦁ with DMOP (eq. (7) in the main paper) and its oxidation products DMOPox (eq. (8) in the main paper) are unknown. They were estimated as 2 × 10 9 M -1 s -1 , i.e. the maximum values measured for the reactions of various phenoxyl radicals with different phenoxides (Steenken and Neta 1979).