Characterizing the phase diagram of finite-size dipolar Bose-Hubbard systems

TL;DR

This study maps the phase diagram of finite-size dipolar Bose-Hubbard systems using advanced numerical methods, highlighting the impact of boundary effects on observable order parameters and proposing machine learning for phase identification.

Contribution

It provides a detailed analysis of finite-size effects on phase detection in dipolar Bose-Hubbard models and introduces machine learning techniques for phase classification.

Findings

Occupation imbalance is more robust than structure factor for small systems.

Supersolid order detection is highly sensitive to boundary effects.

Density measurements combined with machine learning can distinguish phases.

Abstract

We use state-of-the-art density matrix renormalization group calculations in the canonical ensemble to determine the phase diagram of the dipolar Bose-Hubbard model on a finite cylinder. We consider several observables that are accessible in typical optical lattice setups and assess how well these quantities perform as order parameters. We find that, especially for small systems, the occupation imbalance is less susceptible to boundary effects than the structure factor in uncovering the presence of a periodic density modulation. By analysing the non-local correlations, we find that the appearance of supersolid order is very sensitive to boundary effects, which may render it difficult to observe in quantum gas lattice experiments with a few tens of particles. Finally, we show how density measurements readily obtainable on a quantum gas microscope allow distinguishing between superfluid and solid phases using unsupervised machine-learning techniques.