Super-Eddington winds from Type I X-ray bursts

TL;DR

This study uses hydrodynamic simulations to analyze super-Eddington winds from Type I X-ray bursts, revealing how nuclear burning ashes are ejected and influence burst spectra and photospheric expansion.

Contribution

First time multi-zone time-dependent simulations of super-Eddington winds from neutron star bursts, showing detailed composition evolution and wind dynamics.

Findings

Approximately 0.2% of accreted matter is ejected in the wind.

Heavy elements with A>40 dominate the wind composition after 1 second.

Ejected ashes likely cause observed spectral photoionization edges.

Abstract

We present hydrodynamic simulations of spherically symmetric super-Eddington winds from radius-expansion type I X-ray bursts. Previous studies assumed a steady-state wind and treated the mass-loss rate as a free parameter. Using MESA, we follow the multi-zone time-dependent burning, the convective and radiative heating of the atmosphere during the burst rise, and the launch and evolution of the optically thick radiation-driven wind as the photosphere expands outward to radii $r_{\rm ph} \gtrsim 100\text{ km}$. We focus on neutron stars (NSs) accreting pure helium and study bursts over a range of ignition depths. We find that the wind ejects $\approx 0.2\%$ of the accreted layer, nearly independent of ignition depth. This implies that $\approx 30\%$ of the nuclear energy release is used to unbind matter from the NS surface. We show that ashes of nuclear burning are ejected in the wind and dominate the wind composition for bursts that ignite at column depths $\gtrsim 10^9\text{ g cm}^{-2}$. The ejecta are composed primarily of elements with mass numbers $A> 40$, which we find should imprint photoionization edges on the burst spectra. Evidence of heavy-element edges has been reported in the spectra of strong, radius-expansion bursts. We find that after $\approx 1\text{ s}$ the wind composition transitions from mostly light elements ($^4$He and $^{12}$C), which sit at the top of the atmosphere, to mostly heavy elements ($A>40$), which sit deeper down. This may explain why the photospheric radii of all superexpansion bursts show a transition after $\approx 1\text{ s}$ from a superexpansion ($r_{\rm ph}>10^3\text{ km}$) to a moderate expansion ($r_{\rm ph}\sim 50\text{ km}$).