The average field approximation for almost bosonic extended anyons (Q906934): Difference between revisions

In spacetime dimension smaller than four, particles may theoretically satisfy exchange statistics that is neither bosonic nor ferminoic but ``anyonic''. While bosonic statistics means that wave functions of identical particles are invariant under permutations of the particles, fermionic statistics means that a factor of \(-1\) enters when the permutation is odd. In contrast, anyonic statistics is governed by the braid group. In the simplest case, considered in the paper, a phase \(e^{i\pi\alpha}\) is acquired under exchange of two particles, with the bosonic and fermionic case recovered for \(\alpha=0\) and \(\alpha=1\), respectively. The authors work with a model for such particles in spacetime dimension \(2+1\) with a trapping potential. Mathematically, the particle states are modeled in a bosonic Fock space while a ``statistical gauge vector potential'' interaction term in the Hamiltonian implements the anyonic statistics. Since the Hamiltonian obtained this way is too singular, the statistical gauge vector potential is smeared by a small characteristic length scale \(R\). The interest is in an effective description of this model in the limit of large particle number \(N\to \infty\) while simultaneously the statistics becomes approximately bosonic, \(\alpha\to 0\). Such a description has previously been achieved through an ``average field approximation'' with a leading order effect of the anyonic statistics evident. The authors state: ``Our aim in this work is to justify this description rigorously.'' The main result of the paper is Theorem 1.1 which consists of a statement affirming the validity of the average field approximation under certain assumptions. These assumptions are: 1. a lower bound on the potential in terms of a power law with exponent \(s>0\), \(V(x)\geq c|x|^s -C\) where \(c\) and \(C\) may be arbitrary constants, 2. a scaling of the smearing scale \(R\) as \(R\sim N^{-\eta}\) with \(0<\eta<s/(4 (s+1))\), 3. a scaling of the statistics parameter as \(\alpha=\beta/(N-1)\) where \(\beta\) is fixed. The main body of the paper is devoted to the proof of the theorem. To this end, first bounds on the interaction part of the Hamiltonian operator are obtained. Secondly, an a priori bound on the ground state is obtained in terms of a bound on the corresponding one-body energy expectation value. Thirdly, the mean-field (large particle number) limit is taken using a strategy similar to the one in [\textit{M. Lewin} et al., Trans. Am. Math. Soc. 368, No. 9, 6131--6157 (2016; Zbl 1392.35245)].