It took me some time to realize that basic conceptual thing. Hope you won’t make the following kind of stupid mistake:

(- i\hbar\nabla-q\vec{A})^2=-\hbar^2\nabla^2+i\hbar q(\vec{A}\cdot\nabla+\nabla\cdot\vec{A})+q^2\vec{A}^2

Instead, it should be:

(-i\hbar\nabla-q\vec{A})^2=-\hbar^2\nabla^2+i\hbar q(\vec{A}\cdot\nabla+2\nabla\cdot\vec{A})+q^2\vec{A}^2

The point is that we should always treat \nabla as an operator rather than a vector, i.e., when we are deriving a formula involving \nabla, we should always keep a function f at the end of each step !!!

Advertisement