Mobility of zero-valent iron nanoparticles in porous media - A study using X-ray computed micro-tomography

In this contribution we present the results of our study on the progressive porous media clogging induced by deposition of zero-valent iron nanoparticles (nZVI). We study this specific particle since it has proven to deliver effective degradation of chlorinated hydrocarbons when injected in sub-surface layers contaminated by these organic-based toxic substances. The technology is known as nanoremediation and is an emerging technology with great potential for in-situ remediation of contaminated aquifers. For the degradation to occur the nZVI particles need to be injected (in form of aqueous suspensions) into the contaminant bearing sediments1. nZVI nanoparticles are highly reactive and excellent degraders of the chlorinated hydrocarbon contaminants by reduction reactions. While the nanoremediation concept is proven to be successful at laboratory, pilot, and field scales, there are a number of limitations to the technique, including the mobility of nZVI suspensions in sub-surface sediment layers. Particle agglomeration and retention in porous media restricts efficient delivery of nZVI nanoparticles to the contaminated zones. The two primary challenges of working with nanoparticles in porous media: (i) particle aggregation and (ii) particle retention causing clogging1. These challenges are, mainly, dealt with by controlling the size and surface charge of nanoparticles, the chemistry of fluids (ionic strength and pH), and the sediment surface charge. Given their small sizes (<100nm) compared to typical pore-throat sizes of naturally occurring sediment layer (≥10s μm), nanoparticles should cause no clogging issues, if suspended properly. However, nanoparticles form aggregates (~μm size) which reduces mobility, surface area, and reactivity. Darcy-scale modelling approaches commonly used to describe nanoparticle transport in porous media are based on a modified advection-dispersion equation that incorporates exchanges with the porous matrix, namely deposition and release (DR-ADE). The DR-ADE formulations (e.g. clean bed deposition, straining, ripening, blocking) are derived assuming a range of pore-scale processes that, in common laboratory tests (column tests), cannot be investigated and verified in details, but only inferred from large-scale observations. Here we used the experimental results obtained at pore-scale (using X-ray computed micro-tomography2) to obtain darcyscale data, averaged on our sample, which were then fitted with a DR-ADE, similar to column test results, using the software MNMs. MNMs are the particle transport models developed at the Polytechnic University of Turin (Politecnico di Torino). This software (freely available) is the only coupled flow and particle transport model currently available. It models porous media clogging through a range of mechanisms as outlined above. By combining the fitting results and the detailed information on real pore-scale processes occurring in our sample, we verified the validity of the attachment/detachment kinetic models typically implemented in the DR-ADE. Our data is a sequence of 3D images which were captured before, during, and after injection of nZVI suspensions in a sand column. Analysing these images shows how accurate and inclusive the previously identified clogging mechanisms included in the MNMs are.