The matching algorithm would fail for any electrode scale that had string in potentials.as.SHE()$conc.string that was not "saturated" (which happened to be the alkali metals). Fixed this issue by making the matching step-wise, and adding a few more checks for zero rows returned.
Goofed and declared dependency on photoec in previous commit. Fixed it by removing it and the non-exported function that used it.
Note to self: this package should not import any of my other packages.
Previous behaviour returned NAs for values that were already vs SHE. That issue now fixed. The fix included added SHE (zero at all temperatures) to as.SHE.data.