Wettability (with PVP) !!!
The new bottle of PVP (polyvinylpyrrolidone) sitting on the shelf was calling my name this morning. I apologize that this post is long, but its pretty exciting – I think we have half-solved the wettability problem.
According to MSDS sheets, PVP is soluble in water and alcohols. Since drying water on the membrane is a poor idea because of the surface tension, I dissolved it in methanol. I created stock solutions of 0.1, 0.01 and 0.001 g/mL. I placed an 8 uL drop on the membrane surface, which spread very evenly due to the low surface tension. The higher concentrations resulted in about 1/3 to 1/2 of the slits breaking upon drying. Under Brian’s microscope, you could see that the membrane fractured, but did not peel back all the way as usual. In one case, there was still 1/3 of the membrane extended from each wall and about 1/3 peeled back in the middle – clearly the polymer was doing something, but there was too much. 0.001 mg/mL appeared to do the trick – it resulted in a change in optical appearance, but nearly 100% membrane yield after drying. I only treated the flat side of the membrane.
After drying for >20 minutes, I assembled the membranes into SepCons. I used Chris’ air pressure device to apply a constant positive pressure with 500 uL of water behind the membrane. In membranes without pinholes, a wet film appeared on the bottom (PVP-treated) side within a few minutes at 2.5-3.1 psi. After about 30 minutes, 100 and 240 uL passed in two different samples. I now tried these samples in the centrifuge with only water on the top side. Water flowed for a few minutes, but then stopped. I hypothesized that the water and/or the polymer was being drawn off the membrane surface due to the high g-force. I placed these samples back in the air pressure device and very little if any water passed, suggesting that the polymer was gone.I fabricated a few new samples with PVP treatment and re-did the air pressure tests, but pulled off the water droplet as it formed, it was apparent that flow slowed down, suggesting that I was pulling the polymer off the membrane.
I hypothesized that there may be an effect of the hydrophobicity of the membrane. Remember that track-etched memrbanes are hydrophobic before being treated with PVP. Is this a requirement for good adherence? I tested this idea with ozone treated and untreated membranes. From my small sample size, it did appear that ozone treatment before PVP treatment gave poor results.
In terms of the SepCon – the current device is still not acceptable with this treatment. However – with the bucket that Chris and I are developing, this PVP treatment will allow the bucket to fill and maintain water contact on the bottom side without the user having to pipette clean buffer on the bottom side before use. It also enables a constant pressure device, where the filtrate is not pulled away from the membrane, but where one side begins dry (but PVP treated).
Remaining questions:
1. What is the best concentration – 0.001g/mL works well, but this method is variable because it depends on how much dries on the surface
2. How does the PVP attach? Is it charge based? Is hydrophobicity important or just charge? Can we covalently attach it? Can a similar covalent attachment give the same effect? PEG?
3. How does it affect pore size and permeability to molecules?
It’s curious that the flow slows down in the pressure cell when you remove the droplet, since both sides of the membrane should still be wet. As long as it stays wet, the PVP should not be needed. I wonder if the 1mg/mL concentration is still a bit high, potentially caking on the membrane. The PVP layer is probably much thicker than the membrane even at the lowest concentration. If you apply a uL, that’s 1 ug of PVP. The membrane is a few ng, so even accounting for the area, this is probably still an excess of PVP.
It would be interesting to see if 0.1 mg/mL or 0.01 mg/mL also work…
Tom’s made up a few 1mg/mL samples for me to look at under EM. I wonder if the beam will burn up the polymer though… worth a shot.