Why do certain surfaces behave very differently from what theoretical calculations suggest? Scientists long assumed that the aluminum oxide surface should be highly reactive and capable of splitting water molecules. In experiments, however, this behavior is barely observed.