Rejecting proposed dense-matter equations of state with quiescent low-mass X-ray binaries

TL;DR

This study measures neutron star radii in quiescent low-mass X-ray binaries to test dense matter equations of state, rejecting some models at high confidence based on spectral analysis and causality constraints.

Contribution

It provides new radius measurements and directly tests and rejects specific dense matter equations of state using observational data.

Findings

Measured neutron star radius R_NS = 9.4 +/- 1.2 km.

Rejected two dense matter equations of state at 99% confidence.

Excluded most of the R_NS parameter space if a neutron star exceeds 2.6 solar masses.

Abstract

Neutrons stars are unique laboratories to discriminate between the various proposed equations of state of matter at and above nuclear density. One sub-class of neutron stars - those inside quiescent low-mass X-ray binaries (qLMXBs) - produce a thermal surface emission from which the neutron star radius (R_NS) can be measured, using the widely accepted observational scenario for qLMXBs, assuming unmagnetized H atmospheres. In a combined spectral analysis, this work first reproduces a previously published measurement of the \rns, assumed to be the same for all neutron stars, using a slightly expanded data set. The radius measured is R_NS = 9.4 +/-1.2 km. On the basis of spectral analysis alone, this measured value is not affected by imposing an assumption of causality in the core. However, the assumptions underlying this R_NS measurement would be falsified by the observation of any neutron star with a mass >2.6 Msun, since radii <11 km would be rejected if causality is assumed, which would exclude most of the R_NS parameter space obtained in this analysis. Finally, this work directly tests a selection of dense matter equations of states: WFF1, AP4, MPA1, PAL1, MS0, and three versions of equations of state produced through chiral effective theory. Two of those, MSO and PAL1, are rejected at the 99% confidence level, accounting for all quantifiable uncertainties, while the other cannot be excluded at >99% certainty.