It's a dumb question maybe, but I'm not sure how to proceed.
Suppose you have two different models which generate molecules (in the form of coordinates of atoms). The molecules generated by models are not the same.
Next, you have a method to evaluate energy of a molecule (given atomic coordinates, it outputs a number). I can also optimize atomic coordinates with this method.
The question is, how to compare these two generative models in terms of energy?
My guess is that I can run minimization for each configuration and can evaluate dE = E_final - E_initial. But how can I compare/aggregate dE between different molecules? My guess is that to have a crude estimate one can divide these quantities by the total charge of the nuclei in the molecule (which equals to the number of electrons for neutral molecules). Is this reasonable or better ways exist?
That is an ancient and very standard type of assay. You simply need a calorimeter. All of the math is useless without a measurement, those equations are ideal, not real.
In a calorimeter you put perhaps a hundred milligrams in [a very old one] and the instrument does the rest.
I have synthesized two different compounds which is having a different number of atoms count, but similar type of atoms (i.e C, H, N, Ag).
- Badges
- Science topic