From short-range to contact interactions in the 1d Bose gas (Q2189720): Difference between revisions

In the paper under review, the authors consider a system of finite bosons in dimension \(1\) with point interactions described by a Hamiltonian \(H\). They introduce a family of operators \(H_\epsilon\) (\(\epsilon>0\)) that are obtained from \(H\) by a regularization procedure of the point interactions. The main result of the paper is the norm resolvent convergence of \(H_\epsilon\) to \(H\), as \(\epsilon\) goes to \(0\). Under a stronger requirement on the regularized potentials, they also get a bound on the convergence rate. As a consequence, the time-evolution group associated to \(H_\epsilon\) provides a good approximation of the one associated to \(H\). As explained in the paper, the restriction to the dimension \(1\) avoids complicated difficulties. The proof is based on an appropriate representation of the Hamiltonian \(H_\epsilon\) and on the use of the Konno-Kuroda formula. The later provides a good expression of the difference between the resolvent of \(H_\epsilon\) and the one of the free Hamiltonian. Further important ingredients of the proof are the notion of \(\Gamma\)-convergence and an appropriate use of the Green's function of the multidimensional Laplace operator. This yields, together with the above convergence, a formula for the difference between the resolvent of \(H\) and the one of the free Hamiltonian and it can be shown that it only depends on the \(\mathrm{L}^1\)-norm of the regularized potentials. We refer to the Introduction of the paper for more details. We point out, that, in the definition of the regularized potentials, it seems that they are implicitly considered as real functions.