In principle, the logarithm of a partition function and its derivatives can be used to rigorously determine the nature of phase transitions. However, this approach is often ineffective in numerical calculations in practice, since resolving the singularity of a logarithmic partition function requires extremely high-precision data,especially when distinguishing weakly first-order phase transitions and deconfined quantum critical points. Based on the finite-size scaling hypothesis, we propose that the singular contribution to the logarithmic partition function in quantum phase transitions can be extracted by eliminating the analytic terms. This is achieved by computing the Renyi thermal entropy (RTE) and its derivative (DRTE). We have derived the scaling relations for RTE and DRTE, which successfully determine the phase transition points and critical exponents through data collapse in various quantum many-body systems. Notably, beyond their utility in locating quantum critical points, we find that DRTE is remarkably effective in detecting weakly first-order phase transitions, such as in JQ models, which were initially thought to be continuous deconfined quantum phase transitions. The discontinuity of these transitions is challenging to detect using traditional local physical quantities. This work provides a new and powerful paradigm for studying quantum many-body phase transitions.
Wang et al. (Thu,) studied this question.