NICER observations of the evidence of Poynting-Robertson drag and disk reflection during type I X-ray bursts from 4U 1636$-$536

TL;DR

This study analyzes NICER data of two X-ray bursts from 4U 1636-536, providing evidence for Poynting-Robertson drag and disk reflection effects influencing burst spectra and fluxes.

Contribution

It presents the first clear observational evidence of both Poynting-Robertson drag and disk reflection during type I X-ray bursts from a neutron star binary.

Findings

Both Poynting-Robertson drag and disk reflection are observed during the bursts.

Reflection contributes approximately 20-30% to the total burst emission.

Peak fluxes are consistent with or slightly above Eddington limits when reflection is considered.

Abstract

Type I X-ray bursts are the result of an unstable thermonuclear burning of accreting matter on the neutron star (NS) surface. The quick release of energetic X-ray photons during such bursts interacts with the surrounding accretion disk, which raises the accretion rate due to Poynting-Robertson drag and, thus, a fraction of the burst emission is reflected. We analyzed two photospheric radius expansion bursts in the NS low-mass X-ray binary 4U 1636--536 that took place in 2017, using data from Neutron star Interior Composition Explorer. The time-resolved burst spectra showed clear deviations from a blackbody model. The spectral fitting can be significantly improved by introducing either the enhanced persistent emission (the $f_a$ model) or the reflection from the accretion disk (the \texttt{relxillNS} model). The $f_a$ model provides a higher blackbody temperature and higher burst flux compared with the \texttt{relxillNS} model. The peak fluxes of two bursts from the $f_a$ model, $4.36\times10^{-8}~\mathrm{erg~cm^{-2}~s^{-1}}$ and $9.10\times10^{-8}~\mathrm{erg~cm^{-2}~s^{-1}}$, are slightly higher than the Eddington limits of mixed hydrogen-helium and pure helium bursts from previous observations, respectively. When the disk reflections have been taken into account simultaneously, the peak fluxes are lower to match the preferred values. We find evidence to support the finding that both the Poynting-Robertson drag and disk reflection have been appeared during these two X-ray bursts. Moreover, the disk reflection may contribute $\sim20-30\%$ of the total burst emissions.