Home | english  | Impressum | Sitemap | KIT

Estrogenic micropollutant removal by nanofiltration membranes. Part B Model development

Estrogenic micropollutant removal by nanofiltration membranes. Part B Model development
Autor:

Semião, A. / Fourcher, M. / Schäfer, A. (2012)

Quelle:

Journal of Membrane Science 431 (2013), 257-266

Datum: 2012

Semião, A. / Fourcher, M. / Schäfer, A. (2012): „Estrogenic micropollutant removal by nanofiltration membranes. Part B Model development“. In: Journal of Membrane Science 431 (2013), 257-266

Abstract

PDF
ONLINE
  Download Poster Web

 

Removal of estrone (E1) and estradiol (E2) by nanofiltration membranes at neutral pH is carried out both by size exclusion and adsorption. Size exclusion is dependent on the solute to pore radius ratio and the hormone-membrane affinity. It has been shown that the higher the affinity between the trace contaminant and the membrane active layer, the more will partition and penetrate inside the membrane and in consequence permeate. Adsorption, on the other hand is dependent on the hormone concentration, the adsorption isotherm constant (i.e., proportion between the equilibrium concentration and the mass adsorbed) and membrane area available for sorption.

To establish the transport mechanisms involved in the removal of trace contaminants by NF membranes, it is essential to discern between the different contributions of each of these effects independently. In reality, NF membranes have different pore radius, internal surface area and affinity with the contaminants, of which most of these are difficult to determine.

This paper developed for the first time a model that describes the transport of hormones E1 and E2 through the NF membrane pores, in this case the NF 270 membrane, by taking transient adsorption into account. The different mechanisms of size exclusion, adsorption and membrane affinity were considered by modifying the hydrodynamicmodel.

The model was numerically solved and allowed to obtain the concentration profiles along the NF270 membrane pore as a function of time. The model was validated by integrating these profiles and determining the total mass adsorbed on the membrane in transient regime that was then compared to the experimentaldata.

Despite E1 adsorbing 20% more mass than E2 in static mode for the NF 270 membrane (i.e., no pressure applied), E1 adsorbs twice as much as E2 under the same cross-flow conditions. The much higher adsorption obtained for E1 in filtration mode can be explained by a higher partitioning inside the membrane, compared to E2. A higher partitioning increases the concentration of E1 inside the membrane pore, and as a consequence, causes higher adsorption.

The model developed allowed therefore to clearly distinguish between the contributions of the different mechanisms involved in the removal of adsorbing hormones through NF membranes, i.e., concentration polarisation, solute-membrane affinity, size exclusion, adsorption, diffusion and convection.

Although much debate exists on whether solute transport in NF pores is carried out solely by diffusion or by diffusion–convection, this model showed that transport by convection and diffusion described well the transport of adsorbing hormones by NF membranes, where convection especially contributed to the hormone transport for pressures above 11 bar.