1 Département des sciences de la Terre et de l'atmosphère, Université du Québec à Montréal, Montréal, QC H2X 1Y4, Canada; Geotop Research Centre, Montréal, QC H2X 3Y7, Canada. Electronic address: matias-lv@hotmail.com.
2 Département de chimie, Université du Québec à Montréal, Montréal, QC H3C 3P8, Canada. Electronic address: ohlund.leanne@uqam.ca.
3 Département de chimie, Université du Québec à Montréal, Montréal, QC H3C 3P8, Canada. Electronic address: sleno.lekha@uqam.ca.
4 Geotop Research Centre, Montréal, QC H2X 3Y7, Canada; Department of Chemistry and Biochemistry, Concordia University, Montréal, QC H4B 1R6, Canada. Electronic address: yves.gelinas@concordia.ca.
5 Laboratoire Chimie Environnement, Aix-Marseille Université, 13331 Marseille, France. Electronic address: patrick.hohener@univ-amu.fr.
6 Département des sciences de la Terre et de l'atmosphère, Université du Québec à Montréal, Montréal, QC H2X 1Y4, Canada; Geotop Research Centre, Montréal, QC H2X 3Y7, Canada. Electronic address: ponsin.violaine@uqam.ca.

Many processes can contribute to the attenuation of the frequently detected and toxic herbicides atrazine and metolachlor in surface water, including photodegradation. Multi-element compound-specific isotope analysis has the potential to decipher between these different degradation pathways as Cl is a promising tool for both pathway identification and a sensitive indicator of degradation for both atrazine and metolachlor. In this study, photodegradation experiments of atrazine and metolachlor were conducted under simulated sunlight in buffered solutions (direct photodegradation) and with nitrate (indirect photodegradation by OH radicals) to determine kinetics, transformation products and isotope fractionation for C, N and for the first time Cl. For metolachlor, the C-Cl dual isotope slope (?_C/Cl = 0.46 ± 0.19) is identical to previously reported values for hydrolysis and biodegradation in soils, suggesting the same reaction mechanism (C-Cl bond breakage by SN₂ nucleophilic substitution). For atrazine, both direct and indirect photodegradation resulted in a pronounced inverse isotope effect for chlorine (e_Cl = 6.9 ± 3.3 ‰, and e_Cl = 2.3 ± 1.2 ‰, respectively), leading to characteristic dual isotope slopes (?_C/Cl = -0.49 ± 0.17 and ?_C/Cl = -0.31 ± 0.10, respectively). These values are distinct from those previously reported for abiotic hydrolysis, biotic hydrolysis and oxidative dealkylation which are all relevant processes in surface water, opening the path for pathway identification in future field studies.