Towards flavour diffusion coefficient and electrical conductivity without ultraviolet contamination

TL;DR

This paper refines lattice measurements of the thermal vector-current correlator by removing vacuum contributions, enabling model-independent analysis of the flavour diffusion coefficient and electrical conductivity with results smaller than previous estimates.

Contribution

It introduces a method to subtract vacuum contributions from lattice data, reducing short-distance divergences and allowing more accurate, model-independent estimates of transport coefficients.

Findings

Estimated flavour-diffusion coefficient: 2πT D ≥ 0.8

Electrical conductivity estimates are significantly smaller than previous results

The approach may have broader applicability despite current systematic errors

Abstract

By subtracting from a recent lattice measurement of the thermal vector-current correlator the known 5-loop vacuum contribution, we demonstrate that the remainder is small and shows no visible short-distance divergence. It can therefore in principle be subjected to model-independent analytic continuation. Testing a particular implementation, we obtain estimates for the flavour-diffusion coefficient (2 pi T D \gsim 0.8) and electrical conductivity which are significantly smaller than previous results. Although systematic errors remain beyond control at present, some aspects of our approach could be of a wider applicability.