Transport of graphene oxide in the capillary fringe: Insights from sandbox experiments and numerical simulation

Abstract

Studying contaminant transport in the capillary fringe (CF), a crucial part of the vadose zone, offers insights into the mechanisms controlling pollution in soils and groundwater aquifers. This paper investigated contaminant transport in the CF by continuously injecting a conservative tracer (NaCl) and graphene oxide nanoparticle (GONP), an adsorptive contaminant, into a sandbox. After entering the CF from the unsaturated zone, both NaCl and GONP underwent lateral transport. The breakthrough curves (BTCs) for NaCl and GONP were derived from water samples collected at predetermined sampling holes. Subsequently, contaminant transport in the CF was modeled using a one-dimensional–two-dimensional (1D-2D) coupled hydrodynamic model. This model incorporated lateral dispersivity (αL = 1.198 cm) and longitudinal dispersivity (αT = 0.286 cm), calculated using a point-by-point method. The hydrodynamic dispersion coefficients obtained were then applied to the Brooks and Corey (BC) and the van Genuchten (VG) parametric models. The BC model more accurately simulated the NaCl migration compared to the VG model, leading to its application in simulating GONP transport in the CF. However, the simulated BTCs for GONP showed a lag behind the measured data, especially at high ionic strengths. This discrepancy was attributed to the variable adsorption partition coefficient of GONP under different ionic conditions. During the experiment, GONP adsorption onto the porous media's surface altered the capillary dynamics, notably increasing capillary rise height, decreasing seepage velocity, and reducing GONP dispersion. Therefore, it is necessary to consider the adsorption capacity of the contaminants in order to accurately assess their transport within the vadose zone.

Keywords

• Capillary fringe
• Transport
• Sandbox experiment
• Numerical simulation