On the nanoparticle transport and release in layered heterogeneous porous media under transient chemical conditions

•A novel model of nanoparticle transport and release in layered media was proposed.•Heterogeneity results in remarkable concentration gradient along transverse direction.•Mass interaction of two layers was determined by dispersion across interface. The rapid increase in production and application of nanoparticles (NPs) across a wide range of commercial and industrial processes has resulted in the extensive invasion of NPs into subsurface environments. Accurately predicting the transport and release of NPs in subsurface environments is important for assessing a wide range of environmental and human health risks. In this study, a two-dimensional NP transport and release model was proposed to investigate the behavior of NPs in a layered heterogeneous medium with transient ionic strength (IS). The layered heterogeneous porous medium comprised a slow flow domain (SFD) and a fast flow domain (FFD). The transport of NPs in the two layers was described by two coupled advection–dispersion equations (ADEs). The NP deposition and release were related to the IS of the electrolyte solution (ES) in the medium and were represented by reversible and irreversible deposition terms in the ADEs. The results indicated that heterogeneity has a significant impact on the transport and release of NPs. When NPs pass through the two‐layer medium, initially part of the NPs in the FFD will migrate into the SFD. A larger amount of NPs are deposited in the SFD than the FFD owing to the higher filtration capacity of the SFD. The retained NPs in both the SFD and FFD will be released when the IS declines. The release of the retained NPs in the FFD is faster than that in the SFD because the FFD has a higher intrinsic release rate and lower IS owing to the more rapid transport of ES in the FFD. The proposed model was applied to laboratory experiments and compared to the double-region model. The proposed model fitted the experimental breakthrough curves better than the double-region model, and it is capable of fitting the release of NPs, which cannot be accomplished with the double-region model.