PSR J0614-3329: A NICER case for Strange Quark Stars

TL;DR

This study confronts the question of whether strange quark stars (SQSs) are favored by NICER-derived mass–radius data, focusing on PSR J0614-3329’s low-radius measurement. A comprehensive Bayesian hypothesis-ranking framework compares a wide array of physically motivated nucleonic EoSs against several strange quark matter models, using NICER data (and, in extended cases, HESS) to compute Bayes factors and posterior probabilities. The results indicate substantial to strong evidence for SQSs over nucleonic neutron stars, with the evidence strengthening as additional NICER data (and HESS) are included, and with SQS7 (a pQCD-inspired case) being ruled out. The work underscores the potential presence of QM in compact stars, motivates including SQSs in analyses of astrophysical data, and points to future observations and missions needed to decisively settle the QM vs NS debate. **Key methodological elements include KDE-based likelihoods over the M–R plane, uniform priors on mass, and Bayes factors across 26 NS EoSs vs 7 SQS EoSs.**

Abstract

Precise measurements of neutron star masses and radii by the NICER mission impose important constraints on the nuclear equation of state. The most recent NICER measurement of PSR J0614-3329 reported an equatorial radius of $R_{eq} = 10.29^{+1.01}_{-0.86}$ km for a mass of $M = 1.44^{+0.06}_{-0.07} M_{\odot}$. Considering all the NICER measurements to date, we find substantial evidence using Bayesian hypothesis ranking for strange quark stars over physically motivated models of neutron stars compatible with this low radius. This provides a strong case for quark matter in neutron stars and also for the possible existence of strange quark stars, a consequence of the Bodmer-Witten hypothesis, suggesting that they could be considered among the population of compact stars during analyses of astrophysical data. Using a wide sample of equations of state, we report the nucleonic equations of state that best fit current observations and rule out one model of strange quark matter.

Figures (3)

• Figure 1: The mass-radius curves are displayed for 28 NS (dash-dotted, solid, and dotted lines) and 7 SQS (dashed line) EoSs used in this work (see text for details). The thin silver lines are all the EoSs in the LALSuite set. Various 68$\%$ (dark) and 95$\%$ (faint) mass-radius filled contours are overlaid: the orange patch corresponds to the HESS J1731-347 Doroshenko2022 observation, the purple patch to PSR J0740+6620 Salmi2024, the red and blue patches to PSR J0437-4715 Choudhury2024, and J0030+0451 Vinciguerra2024, respectively, while the green patch corresponds to the latest observation of PSR J0614-3329 Mauviard2025. The yellow and green bands correspond to mass measurements of the two heaviest pulsars known, $M=2.08^{+0.07}_{-0.07}$ of PSR J0740+6620 Fonseca2021 and $M=2.01^{+0.04}_{-0.04}$ of PSR J0348+0432 Antoniadis2013, shown only for comparison.
• Figure 2: The bands of Bayes factor $\log_{10}{(B^{SQS}_{NS}})$ formed by SQS1-6 against each NS model EoS. The red, green, and blue bands correspond to Case A, B, and C, respectively (see text). The upper, middle, and lower curves in each band mark the highest, mean, and the lowest Bayes factor from SQS1-6. The curves with green triangles can be considered as lower bounds for each case. The diamonds on the left are the median values for each band considered as the benchmark in this study. The factor is interpreted as follows jeffreys1939theory: (i) $\log_{10}{B} \ge 2$ means decisive evidence in favor of SQS (ii) $1 \le \log_{10}{B} <2$ means strong evidence (iii) $1/2 < \log_{10}{B} < 1$ means substantial evidence and (iv) $\log_{10}{B}\le 1/2$ means insubstantial evidence; the horizontal dotted lines mark these boundaries. The region where substantial evidence exists is highlighted in yellow. *QHC19-D is a hybrid EoS model and is not considered in the calculations.
• Figure 3: The probability of SQSs (see Sec. \ref{['sec:bayesian_analysis']} for the calculation) to be the correct model given the NICER data and given the sample is shown by markers for each case connected by a black line. The purple line with an alternate axis is the corresponding Bayes equivalent. The yellow region highlights the region that can be considered a substantial evidence $1/2 < \log_{10}B <1$.