Implicit solvent electrolyte models

[longemen3000]

There are a lot of electrolyte models that are salt-based instead of ion-based. Before, we had some problems on that part, but i have some new insights:

• for ion-based models, we don't actually care of the salts (except inside the methods)
• for salt-based models, we don't actually care of the ions (except inside the methods)
• We only care about a correct stoichiometric matrix, that has to have a rank of nions-1, for salt-based models, we can check that invariant at construction time, for ion-based models, we can construct an arbitrary salt basis of the size we want.
• if we add the electroneutrality constraint to the stoichiometric matrix, we actually have a matrix of size (nions,nions). Furthermore, neutral components are just diagonal elements, so we can create a matrix of size (nc,nc) that returns a vector of size nc. this matrix can be used to transform between salt-based compositions and ion-based compositions relatively cheaply.
• the neutral model inside an Explicit-solvent model is in Explicit-solvent basis. the neutral model inside a Implicit-solvent model should be in a Implicit-solvent basis

[simonmb]

Yes, but what do you do when you have n ions?

[longemen3000]

it is something that needs to be check in a case-by-case basis. if a model was not designed with that in mind, it will not work even under transformations. My main objective is to allow mix-and-match of ion and salt-based models. For a ionic liquid (in an electrolyte framework). the salt-based model should be capable of calculating bulk properties.