Voltage EO performance
I dug out some of my old data and found this comparison for 3x and 6x window flow rates at different voltages (minimal repeats, error bars not included):
This shows that as voltage goes up, the flow rate increases for each of the active areas shown here. Obviously if we were to publish this we’d probably want data for the 1x and 9x chips as well. Also the lack of repeats is problematic. If I include this data on the active area vs. flow rate graph, here’s what I get:
Looks like a good trend, though it’s unfortunately low on data. I don’t have too many more chips left to repeat this. And since all of this was done with devices formatted for circular chips, we may need to update to a square system, which has its own problems i.e. o-ring sealing, before we can flesh this out enough.
Jess, if all you need is a few wafers to finish this up, let me know what you need and we can get you some circular chips. Might be easier than re-designing a new setup.
I like the data the way they are. Of course, more is better but the trends seem clear.
One thing the flow vs V figure begs is a discussion of the lack of flow with low voltage. This will be a distraction if we are trying to sell this as an ultralow voltage pumps (mV).
The low flow at low (applied) voltage is a limitation in the setup. If she plotted effective voltage vs. flow I think pnc-Si would compare very favorably, right?
The effective voltage scales with the applied voltage, it’s just in the mV range. So as you drop the applied voltage, you’re also dropping the effective voltage. If you do the normalization, we’re still higher than other materials at those lower voltages.
Have you done an I-V curve for your system? I doubt this system is linear at low voltage, as some ion flow is needed to generate any voltage across the 15nm thick membrane. Of course, this effect may be invisible, given the dominance of the bulk series resistance.
Any idea what the equivalent thermal energy is for the ions in solution? It would not surprise me if sub-mV potential cannot supply enough energy to overcome thermal motion, so some threshold would be expected.
We need to figure out a way to get the electrodes closer together….
I-V curves have all been done in a previous post.
Bulk resistance is so high we don’t see membrane contribution. It would be hard to see effects at low voltage.
Right now I calculate the potential across the membrane to be about 10 mV. If this calculation is correct, it seems that millivolt potentials are enough to move ions within the membrane region. I’m not sure what the threshold would be.
Maryna is working on using her metal coated membranes for electroosmosis. We’re not completely sure if this will work yet and there are some challenges right now with making a good device.
Sorry, I meant a more sensitive I-V curve for low voltage. This is such a common measurement that there are tons of tools designed to measure very low currents and voltages, and you generally sweep from a negative voltage, though zero, and to a positive voltage. Doing this often allows sub-threshold effects to be measured. Dave may know if any good systems are available at UR, or JP could help you with a system at RIT. RIT typical has better electrical measurement capability than UR. We actually bought an expensive pre-amp for the Keithley 4200 system at RIT when I was a grad student, so technically should have intermittent access to that tool. They don’t like wet things in their labs, but if you package up your cell, and just need to connect wire, it should be easy.
You may also consider Capacitence-voltage measurements, as these may also tease out some data on what’s actually happening at the membrane. This is also vary common, so somewhere at UR or RIT, there may be useful systems.