[Winga]

For the pi-system of benzene, there are six MOs, pi-1 to pi-6.
As we know, each MO can occupy 2 e-.

My idea is:
For pi-1 bonding orbital, a pair of e- can be delocalized throughout the 6 carbons.
However, the other MOs have (vertical) nodal plane(s) that the e- should be restricted in the lobes.
Therefore, those e- cannot be delocalized freely all over the ring.

So, how can I use MO theory to describe this delocalization that all the 6 e- are participated in.

[plu]

http://library.tedankara.k12.tr/chemistry/vol5/Delocalised%20molecular%20orbitals%20(part%202)/o71.gif

Here is a pictoral representation of the 90% inclusion areas of the 6 benzene molecular orbitals obtained from the linear combination of the 6 p orbitals of the carbons in the benzene ring.

[movies]

Even with the nodal planes, the electrons are still delocalized.  They can tunnel from one side to the other.  Don't get too caught up in what the MOs look like, just remember that any occupied bonding orbital will represent where the electrons are, while any unoccupied anti-bonding orbital will represent where the electrons are not.  The actual phase of the orbitals plays a much bigger role when you want to make bonds than it does when you are describing the static nature of a molecule with a conjugated system.