Bayesian evidence for two peaks in the sound speed in cold dense QCD

TL;DR

This paper presents evidence for two peaks in the sound speed within cold dense QCD matter, supported by neutron star observations, indicating complex non-perturbative effects in the matter's behavior.

Contribution

It introduces a model-independent framework to identify multiple peaks in the sound speed of dense QCD matter using multimessenger neutron star data.

Findings

Evidence for two peaks in the sound speed with Bayes factor 5.1.

The second peak likely indicates non-perturbative effects like color superconductivity.

The trough between peaks is inferred to be very shallow, below 0.05.

Abstract

I show that in addition to the well-known peak inside massive neutron stars, the sound speed in cold dense QCD matter likely exhibits another peak above neutron star densities before it asymptotes to $c_s=\sqrt{C_s}=\sqrt{1/3}$. Based on the framework reported in arxiv:2408.16738, this approach does not rely on any assumption about the ultra-dense matter not realized in nature. Current multimessenger observation of neutron stars favors the two-peak scenario with a Bayes factor $5.1_{-0.7}^{+0.9}$, where the uncertainties are systematics due to models of neutron star inner cores. This evidence grows to $27_{-8}^{+11}$ if the $2.6M_\odot$ component of GW190814 is a neutron star. The trough in $C_s$ separating the two peaks is inferred to be below $0.05^{+0.04}_{-0.02}$ (at the $50\%$ level) if $2.6M_\odot$ neutron stars exist. The second peak above $1/3$ beyond baryon chemical potential $\mu_B=1.6-1.9$ GeV most likely signals non-perturbative effects in cold quark matter, for instance color superconductivity.