Surprisingly large uncertainties in temperature extraction from thermal fits to hadron yield data at LHC

The conventional hadron-resonance gas (HRG) model with the Particle Data Group (PDG) hadron input, full chemical equilibrium, and the hadron type dependent eigenvolume interactions is employed to fit the hadron mid-rapidity yield data of ALICE Collaboration for the most central Pb+Pb collisions. For the case of point-like hadrons the well-known fit result $T = 154 \pm 2$ MeV is reproduced. However, the situation changes if hadrons have different eigenvolumes. In the case when all mesons are point-like while all baryons have an effective hard-core radius of 0.3 fm the $\chi^2$ temperature dependence of the $\chi^2$ has a broad minimum in the temperature range of $155-210$ MeV, with fit quality comparable to the $T \sim 155$ MeV minimum in the point-particle case. Very similar result is obtained when only baryon-baryon eigenvolume interactions are considered, with eigenvolume parameter taken from previous fit to ground state of nuclear matter. Finally, when we apply the eigenvolume corrections with mass-proportional eigenvolume $v_i \sim m_i$, fixed to particular proton hard-core radius $r_p$, we observe a second minimum in the temperature dependence of the $\chi^2$, located at the significantly higher temperatures. For instance, at $r_p = 0.5$ fm the fit quality is better than in the point-particle HRG case in a very wide temperature range of $170-320$ MeV, which gives an uncertainty in the temperature determination from the fit to the data of 150 MeV. These results show that thermal fits to the heavy-ion hadron yield data are very sensitive to the modeling of the short-range repulsion eigenvolume between hadrons, and that chemical freeze-out temperature can be extracted from the LHC hadron yield data only with sizable uncertainty.