Laboratory Evidence on the Mobility of mNPs

Most research has concentrated on artificial porous media, e.g. glass bead columns, in order to try and understand the basic processes [10]. In general, it has been found that a filter coefficient can be used to quantify the removal of particles during passage through a porous medium, based on the product of: a collector efficiency factor (h), representing the efficiency with which an aquifer grain (the 'collector') captures a passing NP; and a collision efficiency factor (a) representing the average number of collisions necessary before attachment occurs. Relationships exist for estimation of h, but a requires laboratory experimentation [11]; a but not h is a function of ionic strength. The most common models for h assume electrostatic repulsion to be negligible, which may or may not be appropriate. The product ha can be related to the familiar first order decay constant and used in modelling solute breakthroughs. The equilibrium concentration at any given distance, assuming that a remains constant, will be independent of the dispersivity, directly dependent on porosity, on velocity, and on grain size, and related inversely but weakly to temperature and particle density. Greater a and h values result in lower breakthrough concentrations.

However, during breakthrough conditions may change: the porous medium may 'ripen', i.e. clogging will occur and the breakthrough of particles becomes impeded, or 'blocking' may occur where the rock attachment sites fill up leading to greater breakthrough concentrations. Addition of organic solutes, either representing stabilizers form the NP source or natural organic matter, will often encourage particles to remain unattached [10]. At lower concentrations, polymers may encourage attachment / aggregation, as polymer ends attach to polymer-free parts of two particles. Under some conditions, initial breakthrough of particles may be early relative to solute breakthrough, due to pore size exclusion and/or lower diffusion rates into slower-flowing parts of the system caused by lower diffusion coefficients of NPs.

Overall, mobility will be maximized where the mNP and the rock surface are of the same charge (often the case), the ionic strength is low (fresh/ very fresh ground-water), and where pore sizes are not too small. Some experiments have been undertaken on intact rocks. Figure 16.2 shows some results of experiments on passage of 100 nm diameter silica colloids through 3-10 cm long cores of UK Permo-Triassic continental redbed sandstones (median pore size 10-40 mm). In the case of pure water (PW), the relative concentration at breakthrough is close to 100% (Fig. 16.2a) but, as the ionic strength increases towards that of fresh groundwaters (AGW) the breakthroughs become small, and often below detection limit (< ~1%). With greater lengths of time, flushing with ionic strengths equivalent to fresh groundwater results in a long slow breakthrough, suggesting a blocking mechanism (Fig. 16.2b).

S

tandard

Through core

Stan

dard

after

,il

120 140 160

60 80 100 Pore Volume

20 30 40

Pore Volumes

Fig. 16.2 Sandstone column breakthrough curves for: (a, b) SiO2 30 nm NPs

20 30 40

Pore Volumes

120 140 160

SiO2 + AGW PV zero arbitrary b

SiO2 + AGW PV zero arbitrary

200 300 Pore Volumes

200 300 Pore Volumes

Data from unpublished thesis and reports by: (a) H. Rahman;

Fig. 16.2 Sandstone column breakthrough curves for: (a, b) SiO2 30 nm NPs

In the case of 30 nm diameter Sb2O5 NPs, breakthrough also occurs in PW but soon concentration declines (Fig. 16.2c) as ripening occurs. TiO2 NPs behave similarly, but breakthrough is even more limited for Ag and Ce oxides. In all cases the attachment to the rock surfaces can be reversed to some extent by lowering the ionic strength, suggesting some secondary minimum attachment. Frequently it is found that small peaks in breakthrough concentrations occur, even in the AGW experiments, and these were dubbed 'navalanches': with no cause obvious, it may be that these are due to the dislodging of a 'key' particle behind which had been held other particles. Investigation will continue to confirm that these events are not artifacts, what their cause might be, and whether the 'hop-stop' mechanism implied does allow significant mass movement.

0 0

Post a comment