Bond angle

i don't think there's simple way to "calculate" the bond angle. even with quantum mechanics, there're even some special cases that bonds aren't straight, e.g. cyclopropane.
the orbitals in the C-C bonds are not strictly head-on overlapping; they're described as "banana bonds".


however, there's a simple way to "estimate" the bond angle, that is, by ESPER theory.
also the fact that "lone pair-bond pair repulsion > bond pair-bond pair repulsion" also takes into account.

for ammonia, ideal bond angle is 109.5* (degree); however, since the lone pair on N poses greater repulsion to the bond pair, the N-H bonds are "squeezed" together, reducing the bond angle.

water is another example: two lone pairs, even greater repulsion exerted on bonds, further squeezing bonds together.


there're cases that behavior of bonding atoms comes into play, eg PH3 has bond angle of ~90*. detailed reasons is complicated; go to wikipedia for details.


NO2 has non-octet structure, but VSEPR still works. repulsion exerted by single-electron-group is weaker than a bond-pair, so N-O bonds are less-squeezed. O-N-O bond angle is greater than 120*.

cases for other compounds are similar.
first find the principle shape for numbered groups of electrons (# electron pairs);
then arrange groups and lone pairs accordingly;
then estimate effect of lone-pairs on bond-pairs.


2011-07-03 01:09:45 補充：
in ideal cases like methane, bond angles are 109.5* and can be calculated with geometry, given that there's no bond stretching, vibration etc.
then, how many ideal cases are there, in our real world?

我諗"己式庚辛"未必岩得晒

109.5個隻係有得計的
用pyth.thm.

不過無錯既係 有好多都唔係關計既事
bond angle只係repulsion做出尼既形狀only.

要用 Quantum mechanics 計的