I think the problem is in the boundary conditions... the issue is that if you use and infinite sheet or a cylinder in the equations you always take cyclic boundary condition.
If this guys are right then the mass of the quasi-particles in a crystal depends on its topology... this is a major thing...