Chemical concepts are century-old. One can define them in the context of quantum mechanics. But are these definitions consistent with our expectations, or is reality more complex than can be covered by them?
In order to support the latter viewpoint, we discuss simple examples that questions the notion of an atom in a molecule, of atomic shells, \dots

The technique used is by defining the region of the space for which the probability to have a given number of electrons is maximal. To give a trivial example, one finds a region of space around a nucleus with charge $Z$,  where the probability of having number of electrons equal to $Z$ is maximal, and take this as a reasonable definition of an atom. Probably many people would agree that in an ionic crystal, one better uses regions of space with a number of electrons larger or smaller than $Z$.