Can there be molecular structures according to QM?

[Log]

In molecular orbital theory, only the electrons are considered and from that we draw conclusions about molecular shape and such. I was thinking just now, what about the nuclei?

If we apply QM to the nuclei as well, wouldn't that mean that we can't know where the atoms are? or is the probability density very concentrated due to the relative size of the atom so that the location is more or less well defined?

For if the location of the nuclei isn't well defined, then how can we be certain of the molecular shape?

- confused undergraduate

 

[f95toli]

First of all, the nuclei is extremely heavy so is more or less "classical"

Secondly, "shape" is a bit of a vague term. Remember that while it is meaningless to talk about the position of an electron in an atom, we can still do calculations and e.g. plot the shape of the orbitals.

Don't make the misstake of assuming that Heisenbergs uncertaintly principle in any way prevents us from doing extremely exact calculations using QM; the HUP comes into play when we do measurements , but it is not (directly) involved if we e.g. want to calculate the binding energy of a molecule.

 

[DrDu]

That's a very interesting question.
The classic discussion of that topic is:
Hans Primas; Chemistry, Quantum mechanics and Reductionism: Perspectives in Theoretical Chemistry;Springer, 1983
The point is that the concept of molecular structure arises from the Born-Oppenheimer approximation, which is, like any approximation, not exact.
In fact there are many molecules which are not in one definite molecular structure, e.g. ammonia in the gas phase. The ammonia molecules are in a superposition of different nuclear structures and due to the relatively low mass of the hydrogen atoms the corresponding energy splitting between the different superpositions can be observed and forms in fact the basis for the ammonia maser which preceeded the development of lasers.