[NicolasM]

Can someone explain why the nitrogen isn't sp2 hybridized, but it is sp3? Wouldn't amines be more stable if it was sp2 hybridized and the lone pair was in a p orbital?

[opsomath]

Why would it be more stable in that case?

Using basic VSEPR logic, you can see that sp2 hybridization would put the hydrogens in a triangle around the center nitrogen. Then, the lone pair would be sticking up out of the middle, making the angle between the lone pair and the hydrogens 90° with a big empty space on the other side.

Naturally, the electron clouds in the bonds and lone pair will repel each other and the bonds will be "pushed" to the other side of the nitrogen atom until the point where they repel each other to an amount that matches the force from the nitrogen lone pair.

[gritch]

VSEPR theory is fine but hybridization isn't a great way to look at the geometry of more complicated molecules. The fact that ammonia has a distorted tetrahedral geometry instead of a planar geometry has more to do with electronic rather than electrostatic effects, as such it's better to explain it shape using molecular orbital theory.

NH₃ has a total of 8 valance electrons. When combining the orbitals of the central nitrogen atom and the 3 hydrogen atoms it is energetically favorable to form a distorted tetrahedral geometry. If however we reduce the number of valance electrons (say using BH₃ with 6 electrons) the trigonal planar geometry becomes energetically more favorable.

I'm having a hard time explaining it so here are some sources to look at:

http://www.dartmouth.edu/~chem64/64%20pdf%20files/PS3A.pdf
Nice correlation diagrams of AH₃ in both the planar and tetrahedral geometries.

http://www.pci.tu-bs.de/aggericke/PC4e/Kap_V/Walsh_AH3.htm
A Walsh diagram detailing the change of energies of the molecule orbital as a function of molecular geometry. To read it: consider each line to represent a molecular orbital. Each orbital can hold 2 electrons and you can fill in the orbitals according to the aufbau principle. To determine the energy of the molecule at a given geometry simple sum the energies (the y-coordinate) of all electrons. There's no real scale so it's difficult to be quantitative but it should help to imagine the total energy qualitatively.

Probably more than what you were looking for there but I don't like hybridization theory it's much better to try and go into molecular orbital theory instead.

[opsomath]

Great explanation, I agree that MO theory is the way to go there. I was just trying to keep it simple.

[NicolasM]

Thank you a lot

[magician4]

... 'cept that sp³-geometry is no good way to describe the situation esp. with NH₃

all that is said so far is valid and relevant for example for some alkylamines (i.e urotropine and thatlike) : they're sp³ indeed, as this structure is forced by the ringsystem(s)

however, in reality the "average" geometry of the very NH₃ molecule (and a lot of simple amines, too) is more like sp² , as all the atoms are "vibrating " ( in fact, the nitrogen is also tunneling part of it's pathway) and inverting 'round that planar geometry

http://en.wikipedia.org/wiki/Nitrogen_inversion


as a consequence, for example you can't have enantiomer-separation from molecules like  methyl-ethyl-amine (as you should have expected from a naive analysis, if the molecule was sp³ indeed)


regards

Ingo