Schrodinger equation with magnetic vector potential

In many papers dealing with the Schrodinger equation with magnetic potential $$u_t=i(\nabla+iA(t,x))^2u$$ the authors say that this equation can be studied with Kato's methods for abstract evolution equations. Is there someone who can suggest some reference in which this approach is used?

• I guess you mean $u_t = i(\nabla+iA)^2 u$ (otherwise the equation is not self-adjoint). – Carlo Beenakker Jan 7 '15 at 21:17
• Do you mean operator splitting? – Steve Huntsman Jan 7 '15 at 22:15
• No I mean methods for the study of abstract evolution equation (like stable families of generator of continuous semigroups) – Sam Jan 7 '15 at 22:26
• What exactly are you interested in? The operator on the RHS is skew-adjoint (provided you have "good" boundary conditions and/or reasonable magnetic potential), hence it generates a unitary group on $L^2$ by Stone's Theorem. What is usually referred to as "Kato's theory" is a collection of much deeper results on admissible scalar potentials. – Delio Mugnolo Jan 8 '15 at 8:43
• Quite frequently (though I cannot guarantee it) in this context the reference to "Kato's methods" refers to the developments centered around his two papers ams.org/mathscinet-getitem?mr=279626 and ams.org/mathscinet-getitem?mr=326483 The theory is strong enough that oftentimes authors just refers to it as a blackbox guaranteeing the existence of "evolution" for the linear operator. But you can probably find something interesting if you look at MathSciNet references to those two papers. – Willie Wong Jan 8 '15 at 8:53