That is to say, I'm not sure there really is a mechanism that will make the small percentage of alcohol drive off the remaining water.
Actually there is: the two form a positive azeotropic mixture, consisting of about 88% water and 12% IPA, with a boiling point lower than either component (around 80 C in this case). This is a well known phenomenon with many binary and ternary mixtures.
Last edited: