Approximate Hofstadter- and Kapit-Mueller-like parent Hamiltonians for Laughlin states on fractals

TL;DR

This paper develops simpler, approximate Hamiltonians for fractional quantum Hall states on fractal lattices, achieving high overlaps with model states using only onsite and two-site interactions, facilitating analysis of topological properties.

Contribution

It introduces an inverse method to construct approximate parent Hamiltonians on fractals with only local interactions, improving simplicity over exact nonlocal models.

Findings

High overlap between ground states and model states with up to third neighbor hopping.

Nearly perfect overlaps achieved by increasing maximum hopping distance.

High overlap with quasihole wavefunctions, especially for nonlocal Hamiltonians.

Abstract

Recently, it was shown that fractional quantum Hall states can be defined on fractal lattices. Proposed exact parent Hamiltonians for these states are nonlocal and contain three-site terms. In this work, we look for simpler, approximate parent Hamiltonians for bosonic Laughlin states at half filling, which contain only onsite potentials and two-site hopping with the interaction generated implicitly by hardcore constraints (as in the Hofstadter and Kapit-Mueller models on periodic lattices). We use an ``inverse method'' to determine such Hamiltonians on finite-generation Sierpi\'{n}ski carpet and triangle lattices. The ground states of some of the resulting models display relatively high overlap with the model states if up to third neighbor hopping terms are considered, and by increasing the maximum hopping distance one can achieve nearly perfect overlaps. When the number of particles is reduced and additional potentials are introduced to trap quasiholes, the overlap with a model quasihole wavefunction is also high in some cases, especially for the nonlocal Hamiltonians. We also study how the small system size affects the braiding properties for the model quasihole wavefunctions and perform analogous computations for Hamiltonian models.