Why is Flow so LOW????

I think the consistently low flow rates are a real bottleneck for all our fluid based experiments and must be overcome. On the production side, we are really pushing the limits of our capabilities to produce the current membrane geometry, which I consider to be crazy big at 0.1 x 2 mm!! Considering the 15 nm thickness, talking about mm’s is incredible (actually should be published as a feat in materials fabrication). I don’t think it makes sense to to scale up this area, as we will take a huge hit in yield, unless we can fix our current yield issues (sputtering in a low particle environment, getting a megasonics cleaner, maybe the new sputter tool will help, etc.).

I think it would be useful to step back and look at our flow rates in the context of the Tong theory (probably the best we have). If you take the outer samples of wafer 310, looking at Dave’s histogram, I think we could conservatively approximate this membrane by saying that it is 1% porosity with 30 nm pores. (Doing an exact calculation using the histogram is trivial – has anyone done this? If someone sends me the data, I’ll just do it.) With this CONSERVATIVE approximation, I calculate a flow rate of 0.4 mL/min/PSI/cm^2. This is >4X what Jess has been measuring, and this difference is likely much larger.

What’s going on here? Is the theory wrong? Are our measurement incorrect? Is there a factor that I am not compensating for?

Is it possible that we are only filling 1/5 of the pores with water? Have we given up on any oxidation or wetting chemistries? I know our strategy has been to push ahead to show interesting separations and test cutoffs, but if we could approach the theoretical flow rates, it seems like all our experiments would become easier. Was the wet-wet format been tested with oxidized/treated material? I remember that oxidation made a significant difference in diffusion studies (for the Nature paper and in general), and therefore would be critical in convective flow. Shigeru has also shown that without oxidation, our membranes are much less permeable, and he is in a wet-wet geometry.

No matter what the root cause, we either need to explain why our flow rates are so poor, or fix whatever our current problem is. I think we need to take a step back and re-think this problem.