[mana]

hi all
for calculation of covalent radius in a 2 atomic molecule, when the two atoms are the same, it is easy and by dividing to 2 you can find the covalent radius, but when we have a bond between two different atoms, you have to subtract the atomic covalent of one of these atoms from the bond length, and in this way, the number of covalent radius for an atom can be different, for example for hydrogen in H-F and H-I. this means the covalent radius is not closely related to the atomic radius, am I right?

[Borek]

Even atomic radius is not precisely defined. Atom doesn't abruptly end like a ball, electron density goes exponentially down as the distance from a nucleus grows, at some arbitrary selected point we say "atom ends here". As the cutoff value is arbitrary, so is the atomic radius.

The other approach is to not treat atom in isolation, but in a crystal structure of the pure element, and call the "atomic radius" half the distance between atoms in such a crystal. Sadly, it is ambiguous as well - first, atoms in crystals do bond, so it is kind of a covalent radius again, second - what to do with elements that have different possible allotropes, with different crystal structures, yielding different atomic radii?

Take home lesson is to not worry too much about these radii. They do show some trends and they do help with some rule of thumb predictions, but treating them too seriously is a road to perdition

[Orcio_87]

What about gradient of electron density along the axis between the two atoms ?

There is always a point in the middle where electron density is largest, and then - decreases in any direction.

It can be taken as the border between the two atoms.

Drawback is that in polarized bonds electron density is shifted towards more electronegative atom, so it will give results that the more electronegative atom is smaller - for example nitrogen is smaller than hydrogen in NH₃.