Molecular Geometry & Bonding
what the shape decides
The index

What is refuted here

The wrong explanations this site takes apart, sorted by what kind of wrong each one is, and with the computation that settles it beside each.

Molecular structure is taught worse than almost anything else at this level. A reader arriving here has usually been told that electrons occupy orbitals, that sulfur hexafluoride borrows its d orbitals, that benzene resonates between two structures, and that spreading electrons out always lowers their energy. Being misinformed is a harder starting position than being ignorant, because each of those has to come apart before anything can be built on top of it.

So the wrong explanations are treated as content rather than as omissions. Each one below is stated in the form it is usually taught, given a verdict, and settled by something computed in the essay that tests it — not by an appeal to a better textbook.

The four verdicts are not interchangeable. Most of these claims are not simply false. The common cases are a rule carried outside the ground it was derived on, an account that stops just short of the quantity that decides, and a statement about a description that reads like a statement about a molecule — the last of which no measurement can settle, which is what makes it the most stubborn.

Simply false

The claim is not true, and something computed here says so.

748 claims

True where it was derived, applied where it was not

A rule taken outside the case it was justified in, usually because it keeps giving the right answer for a while.

120 claims

True as far as it goes, and silent on what decides

Not wrong, but stops exactly short of the quantity the question turns on.

361 claims

A property of the model, mistaken for one of the molecule

The statement describes the description. Changing basis or changing model changes it, and no measurement can settle it.

27 claims

The sharpest of them

A forbidden overlap is not small. It is zero, and the difference is the whole argument.

1s with 2px at 2.8 bohr. The two orbitals in the plane containing both nuclei, with the regions where their product is positive and negative shown faintly. The overlap integral is the signed volume of that product, and where symmetry makes the two regions mirror images it comes out exactly zero. Contours drawn: 1s at 50% of its density, |ψ| = 1.48e-1; 2px at 50% of its density, |ψ| = 3.16e-2.
Fig. 1 A 1s orbital against a 2p orbital perpendicular to the line between the nuclei. Every region where the product is positive has a mirror image where it is negative, so the integral cancels in pairs and the computed value is arithmetic noise at 10⁻¹⁷. Calling such an interaction “weak” gets the consequence right and the reason wrong, and the reason is what generalises.

The field this comes from · Every orbital, by what it encloses · All essays