If they are due to the random unsymmetrical distribution of electrons around a molecule and thus creating a temporary dipole, why don't you get a (random) distribution that causes molecules to repel. Which would then cancel the intermolecular attraction?
This is a very good question, Andrew. You are missing one thing, and that is that the temporary dipoles in neighboring molecules are not independent. Instead, temporary dipoles induce temporary dipoles in neighboring molecules in such a way that an attractive interaction results.
Think of it this way: consider three nonpolar molecules. In my diagrams, the crosses are the nuclei and the circles represent the electron clouds.
While this does not happen sequentially, and indeed is likely to be destroyed by electron motions in the next moment, the overall result is a situation in which attractive forces slightly outnumber the repulsive ones.
Another, more complicated, way to think of this is as follows:Remember that dispersion forces are weak, and they fall off very rapidly with distance. Because of this, I usually compare dispersion forces to friction, because they are related to molecular surface area. I usually compare interactions between nonpolar molecules to those between pieces of cooked pasta. It's harder to separate large pasta shells than small ones, and much harder to separate cooked spaghetti strands from each other than it is to separate individual pieces of cooked macaroni.Dispersion forces, strictly speaking, are due to the attraction of the atomic nuclei of one molecule for the electrons of another molecule.
Any charged object (like an atomic nucleus) creates an electric field that extends to infinity. The field falls off as the square of the distance, but it never goes all the way to zero. Even though there are other electrons in between, "masking" the nucleus by providing a repulsive force to balance the nucleus' attraction, electrons in a neighboring molecule still feel a weak attractive force from that nucleus. This is, to a certain extent, completely independent of the distribution of the electrons belonging to the nucleus in question!
You might want to think about why it is that the electrons in an atom can't completely mask the nucleus of that atom from the electrons in a molecule close by.
Some of this answer may be found in P.W. Atkins' Physical Chemistry. Some of it is my own opinion, bolstered by my knowledge of physics.
| Bluffton College |