Capturing In Situ Glyphosate (De)sorption Kinetics in Floodplain Aquifer Sediment Columns: Geophysical Measurements and Reactive Transport Modeling

• Published in: Environmental science & technology 2022-09, Vol.56 (18), p.12955-12964
• Main Authors: Mellage, Adrian, Dörrich, Manuel, Haderlein, Stefan B.
• Format: Article
• Language: English
• Subjects: Adsorption; Aquifers; Contaminants; Contaminants in Aquatic and Terrestrial Environments; Floodplains; Geologic Sediments; Geophysical methods; Glycine - analogs & derivatives; Glyphosate; Groundwater; Herbicides; Induced polarization; Kinetics; Model matching; Organic contaminants; Pollution monitoring; Retention; Saturated flow; Sediment pollution; Sediments; Soil; Soil contamination; Sorption; Spatial distribution
• ISSN: 0013-936X; 1520-5851
• DOI: 10.1021/acs.est.1c07119

Glyphosate, an ionizable organic herbicide, is frequently detected in soils and groundwater globally despite its strong retention via sorption. Understanding its apparent mobility hinges on our ability to quantify its system-specific sorption behavior, hindered by its affinity to adsorb onto sediments, yielding very low aqueous concentrations. Here, we present findings from a saturated flow-through column experiment in which we monitored glyphosate sorption onto a natural calcareous aquifer sediment, using the noninvasive geophysical method spectral induced polarization (SIP). Our kinetic sorption reactive transport model predicted the strong nonlinear reversible retention of glyphosate and reproduced the spatial profile of retained glyphosate in the sediment, with a measured maximum of 0.06 mg g–1. The strong contribution of sorption to pore fluid conduction masked the expected variations in imaginary conductivity, σ″. However, time constants derived from a Cole–Cole model matched the timing and spatial distribution of model-predicted sorbed concentration changes, increasing from 0.8 × 10–3 to 1.7 × 10–3 s with an increase in sorbed glyphosate of 0.1 mg g–1. Thus, glyphosate sorption modified the surface charging properties of the sediment proportional to the solid-bound concentrations. Our findings link SIP signal variations to sorption dynamics and provide a framework for improved monitoring of charged organic contaminants in natural sediments.