He did the first theoretical study, but his calculated exhaust velocity was too high, because he didn't take dissociation into account.
I read a recent paper on LOX/LH2 combustion, claiming there were 27 reversable chemical reactions you had to take into account to calculate accurately -- H2, O2, HO, H2O, H2O2, OH-, and other molecular and atomic species are present in the combustion chamber -- and they weren't taking account of ionization which must be relevant too. Sigh...nothing is ever simple.
Don, I failed to take this in to account when I assumed that a two to one molar ratio was optimum for exhaust velocity. The interesting thing is that adding more Hydrogen does not lower the Temperature as much as one would expect, if all reactions went to compleation. As a result the performance penalty in lowering the temperature, is (to a point) out weighed by the performance enhancement of lowering the molecular weight of the exhaust.