A new upper bound on the specific free energy of dilute Bose gases
Giulia Basti, Chiara Boccato, Serena Cenatiempo, Andreas Deuchert
TL;DR
The paper proves a sharp finite-temperature upper bound for the specific free energy of a dilute 3D Bose gas by constructing a sophisticated correlated trial state in the bosonic Fock space. The authors combine a grand-canonical uncorrelated baseline with a carefully controlled, unitary-like correlation dressing that implements Bijl–Dingle–Jastrow-type pair correlations, including precise momentum-space cutoffs and a truncated exponential dressing to maintain tractable entropy. They derive explicit bounds for kinetic and interaction energies, isolating the Lee–Huang–Yang–type correction term proportional to $4\pi \mathfrak{a} (2\varrho^2 - [\varrho - \varrho_c(\beta)]_+^2)$ and bounding the remainder with powers of the diluteness $(\varrho \mathfrak{a}^3)$, achieving a rate $(\varrho \mathfrak{a}^3)^{1/7}$ in the error. This work extends previous upper-bound results to a broader class of repulsive potentials, provides a transparent energy-entropy balance, and offers a framework adaptable to extensions in finite volumes and temperature regimes.
Abstract
We prove an upper bound for the free energy (per unit volume) of the dilute Bose gas in the thermodynamic limit, showing that the free energy at density $ρ$ and inverse temperature $β$ differs from that of the non-interacting system by the correction term $4 π\frak{a} (2 ρ^2 - [ρ- ρ_{\textsf{c}}(β)]^2_+ )$. Here, $\frak{a}$ denotes the scattering length of the interaction potential, $ρ_{\textsf{c}}(β)$ the critical density for Bose-Einstein condensation of the non-interacting gas and $[\cdot]_+=\max\{0,\cdot\}$. This result was previously established by Yin in [37]. Our proof applies to a broader class of interaction potentials, yields a better rate, and we believe it has potential for further extensions.
