Considering ion charge when deducing reactions

[samblau]

We currently ignore the charge of a metal ion when deducing reactions. This can be a problem for lithium systems when Li0 is present in the network, which allows you to take a Li+ starting molecule and reduce it to Li0 and then combine that Li0 with an EC0 to obtain Li+EC-, since only the total molecular charge is checked. But by just removing the Li0 from the network, this can mostly be avoided. However, for magnesium systems the problem becomes more prevalent because Mg2+ and Mg1+ are both legitimate species. In principle, we could track the metal ion charge via partial charges or Critic charges and then disallow reactions in which the metal charge is changing in order to avoid reactions like e.g. Li0 + EC0 -> Li+EC-. However, for the Mg system, maybe we don't want to disallow those types of reactions because internal charge transfer can and does happen, e.g. Mg+TFSI- -> CF3 + Mg2+TFSI-(without CF3). I don't know the solution, but I wanted to open this issue so that we don't forget the underlying problem.