Adsorption Quantification VIII
Here is the full updated Adsorption Quantification figure:
We were concerned that cellulose appeared to adsorb more than PES when the literature clearly states that it shouldn’t due to its hydrophilicity. We considered that this same hydrophilicity might be part of the problem with our method. Macroscopically, it appears that the protein solution applied to cellulose wets it and potentially soaks into the material. This is not true for PES, so we considered that the BSA may be contacting more of the suface in the case of cellulose.
We then soaked both membranes overnight and applied protein only to the top side. We found that in this case there appeared to be less protein adsorbing to the cellulose and about the same amount to PES. This might be because there’s no wicking to draw the protein further into the cellulose membrane if the whole thing is wet.
If we look only at the prewet samples this is the figure we get:
Cellulose is on the same scale as pnc-Si and glass while PES is higher adsorbing. If we buy the whole hydrophobicity arguement for adsorption this may be reasonable, as pnc-Si and glass are fairly hydrophilic too.
We are considering a second method to back up the SDS-PAGE densitometry that I’ve been doing here. We think that while radiolabeling may be a difficult process with a lot of red tape, but it would enable us to correlate between our membrane and polymeric membranes. The idea here would be to adsorb 14C or 3H labeled BSA to probe with a scintillation counter.
(Jim the change in these figures since the last time you saw them is due to a correction in the frontal surface area for the prewet samples.)


I’m pretty comfortable with the last figure. Given that the polymer membranes are ~50% porous, the amount of available surface area for binding is higher than the frontal area wheres with our membranes and glass the frontal area is a very good estimate of the total binding area. So this just might be right. Still it would be great to back up the data with a second method since we will probably cite this study until the end of time. The radio-labeling idea is difficult to execute because of regulatory hurdles, but it is direct and the only other method we could think of that allows measurement on all sample types.
If anyone else has an idea for a second methodology, please share.