Salt vs Neutral Form

[mattodd]

A possible reason for the poor correlation between experimental and calculated solubility seen in #1 is that many of the compounds tested are salts, and these are not being calculated correctly.

Let's test this to see if it makes a difference, @ucnvwqt.

1) Take three compounds that are salts. These will be compounds which, when you look at the structures in e.g. Datawarrior etc are shown as two compounds, one with a basic group (like an amine) and another smaller molecule with an acidic group. In this Issue, below, write their codes and show the structures.

(I'm assuming most compounds that we're looking at will be amine bases, so will have an acid counterion. If you can find an organic molecule that is a carboxylic acid, and which is given as a salt, then include that too, but these are rarer).

2) Add the SMILES strings for i) the neutral pair (when the H is on the acid, not the base), ii) the salt (when the H is on the base (i.e. it's protonated) and the acid is deprotonated, iii) just the neutral organic part and iv) just the ionised organic part.

3) Calculate the solubilities for all of these species and see which is the best match for the calculated value.

[ucnvwqt]

@mattodd for your information, only 11 compounds tested are salts so I don't think the poor correlation has anything to do with salts.

salts comparison.xlsx The attached excel file is about 3 salts compounds tested according to your suggestion in the previous comment. Have a look at it and I look forward to your comments and further instructions.

[mattodd]

@ucnvwqt interesting. So, could you prepare a more visual summary (which you'd need for your thesis in any case, so it's a good use of time) showing the structures of the molecules X, Y and Z, and the differing values for each salt type? Try to present it succinctly, as you might in a single diagram in a thesis. The key questions are: does it matter how we draw the molecules?