The method of the weakly conjugate operator (Q4356523)

In order to show that a self-adjoint operator \(H\) has no singular spectrum, one must look for a second operator \(A\) such that the commutator \(B\equiv i[H,A]\) is positive in a suitable sense. The authors show here ``the method of the weakly conjugate operator'' as an improvement of the Kato-Putnam theorem. They treat a self-adjoint, unbounded operator \(H\) having a spectral gap, and prove the following assertions for \(H\): If \(H\) admits a weakly conjugate self-adjoint operator \(A\), such that \(e^{-itA} Be^{itA}\) is strongly \(C^1\), \(H\) has purely absolutely continuous spectrum. Here \(A\) is weakly conjugate to \(H\), if \(e^{-itA} He^{itA}\) is strongly \(C^1\), and if \(B>0\). They apply their method to the spectral theory of operators \(H=H_0 +V=h(P) +V(Q)\), \(P= -i\nabla\), and of the second-order elliptic operators. In the application they use the operator \(A_\gamma= \{\gamma (P,Q)+ \gamma (Q,P)\}/2\), given by a real, symmetric, bilinear form \(\gamma\) on \(\mathbb{R}^N\), as the weakly conjugate operator. For the entire collection see [Zbl 0868.00065].