The literature on ab initio calculations of chemical shift in NMR experiments usually provide Lamb's and Ramsey's formulae as the solution. Yet the expressions explicitly depend on vector potential gauge choise. Some authors claim that the theory is intrinsically gauge invariant, giving no proper proof yet. Can the solution be expressed in an explicitly gauge invariant form? Some literature on this.
NOTE: I know about London orbitals theory, yet it's still not fully understood by me, how it can deliver a type of expression I want. Some suggestion on where this is clarified are very welcome.