Conclusions
In this work a new approach to fit results from LLDP analysis has been proposed. Instead of the usual polynomial fitting equations without physical meaning, a log-normal pore size distribution was assumed for the membrane, requiring only three fitting parameters. Comparing the expected behavior of the model with the actual results, the log-normally shaped distribution is a reasonable choice.
The accuracy of the fitting approach has been tested with several polymeric membranes having a large range of pores, according to the nominal MWCO values. Results of the fitting are reasonably good and are in accordance with most of the parameters arising from traditional LLDP analysis.
One of the advantages of the model is that it can be applied using few experimental points which would reduce significantly LLDP procedure times. In addition, it has been shown that acquiring information especially at low and high pressures or just employing the first point, one in the middle and one in the end (indicated by the IF and ItF labels, respectively), the model leads to reliable membrane information.
It must not be forgotten that a good fitting is only possible if good data is collected. Whatever is the procedure for measuring or fitting data, LLDP experiments are not and they will never be so easy to perform as GLDP ones, where only 5 min are enough to get a good pore size distribution.
On the contrary LLDP experiments need longer time (never less than 1 h), so any approach aimed to reduce the number of experimental points, such as that considered in this work, requires a reliable data fitting for a possible extensive or commercial use of the technique.