Revealing the Condensate and Non-Condensate Distributions in the Inhomogeneous Bose-Hubbard Model

TL;DR

This paper calculates the spatial and momentum distributions of condensate and non-condensate atoms in the inhomogeneous Bose-Hubbard model, highlighting the importance of accurate distributions for experimental phase transition detection.

Contribution

It introduces the use of Quantum Monte Carlo generated distributions to improve experimental accuracy in identifying phase transitions in the Bose-Hubbard model.

Findings

Approximate distributions can lead to errors in estimating the condensate.

Strong interactions create pedestal-like structures around the condensate peak.

QMC-based distributions enhance the accuracy of experimental measurements.

Abstract

We calculate the condensate fraction and the condensate and non-condensate spatial and momentum distribution of the Bose-Hubbard model in a trap. From our results, it is evident that using approximate distributions can lead to erroneous experimental estimates of the condensate. Strong interactions cause the condensate to develop pedestal-like structures around the central peak that can be mistaken as non-condensate atoms. Near the transition temperature, the peak itself can include a significant non-condensate component. Using distributions generated from QMC simulations, experiments can map their measurements for higher accuracy in identifying phase transitions and temperature.