I believe it's the half-life for each bond. But 10^22 is plenty high enough that essentially all bond are broken, and certainly not enough left to reconstruct the genome. (You need unbroken sections of certain minimal lengths to statistically reconstruct what was there before.)

I think you're right that it's a temperature thing.