One of the comments mention deionized water. That's something I've been wanting to experiment with, to see if you can realistically keep water dielectric over time as an immersion cooling medium.<p>Given that commercial applications would rather stick to the super-expensive Fluorinert I suspect it's a pretty hard problem to solve.