[isears]

Are there any rules for when to expand valence shells in lewis structures? The rule of thumb I've been going with is to go ahead and expand valence shells if it means formal charges can be evened out and the expanded atom is in period > 4 (or Phosphorus, in some cases). However, I've been getting a lot of exercises wrong when I check my answers in the book I'm working from.

For example: OSbCl₃

My lewis structure for this molecule is as follows:


As you can see, Sb's valence shell is expanded, but all formal charges work out to 0 -- very stable & happy molecule

However, the book's model looks like this:


Is there a better rule for valence shell expansion in lewis structures? Or perhaps the character of the bond between Sb and O is more ionic in nature and the lewis structure confusion is just a natural result of the lewis structure not being a good model for ionic bonding.