Constraints on neutron star superfluidity from the cooling neutron star in Cassiopeia A using all Chandra ACIS-S observations

Peter S. Shternin, Dmitry D. Ofengeim, Craig O. Heinke, Wynn C. G. Ho

Submitted on 4 November 2022


Analysis of Chandra observations of the neutron star (NS) in the centre of the Cassiopeia A supernova remnant taken in the subarray (FAINT) mode of the ACIS detector performed by Posselt and collaborators revealed, after inclusion of the most recent (May 2020) observations, a significant decrease of the source surface temperature from 2006 to 2020. The obtained cooling rate is consistent with those obtained from analysis of the 20002019 data taken in the GRADED mode of the ACIS detector, which is potentially more strongly affected by instrumental effects. We performed a joint spectral analysis using all ACIS data to constrain the NS parameters and cooling rate. We constrain the mass of the Cassiopeia A NS at M=1.55±0.25 M, and its radius at R=13.5±1.5 km. The surface temperature cooling rate is found to be 2.2±0.3 per cent in 10 years if the absorbing hydrogen column density is allowed to vary and 1.6±0.2 per cent in 10 years if it is fixed. The observed cooling can be explained by enhanced neutrino emission from the superfluid NS interior due to Cooper Pair Formation (CPF) process. Based on analysis of all ACIS data, we constrain the maximal critical temperature of triplet neutron pairing within the NS core at (49.5)×108 K. In accordance with previous studies, the required effective strength of the CPF neutrino emission is at least a factor of 2 higher than existing microscopic calculations suggest.


Comment: 20 pages, 17 figures. Accepted for publication in MNRAS

Subject: Astrophysics - High Energy Astrophysical Phenomena


Surface temperature evolution for modes with fixed NS parameters and variable $N_{\mathrm{H}}$ (left) or fixed $N_{\mathrm{H}}$ (right). The employed values of the NS parameters are given in the top right corner in each panel. There, $N_{\mathrm{H},22}\equiv N_{\mathrm{H}}/(10^{22}~\mathrm{cm}^{-2})$. Horizontal axes show time in modified Julian days with respect to MJD 55500. Filled dots, open dots, filled diamonds and open diamonds correspond to the FAINT mode data, FAINT mode data without pile-up, GRADED mode data and GRADED mode data fitted without introducing the calibration factor $A$, respectively. Filled strips show 68 per cent prediction intervals for the regression model obtained by fitting equation~(\ref{eq:T-t_linear}) to all data (filled dots and diamonds). Regression slope $s$ estimates and best-fit $\chi^2$ values are indicated at each panel for regression using all data and the FAINT and GRADED data points alone. Lower panels show the standardised residuals around the best-fit regression curve for all data. To guide the eye, the distribution of residuals is compared to the standard normal distribution shown by the red solid curve.