Comprehensive models of novae at metallicity $Z = 0.02$ and $Z = 10^{-4}$

TL;DR

This study models novae at different metallicities using MESA, revealing how metallicity and mixing influence ignition mass, ejected mass, recurrence period, and luminosity, with implications for nova evolution understanding.

Contribution

It provides comprehensive nova models at Z=0.02 and 10^{-4} including mixing effects, offering new insights into how metallicity impacts nova properties and retention efficiency.

Findings

Higher metallicity leads to smaller ignition masses and recurrence periods.

Mixing reduces accreted and ejected mass, increases maximum luminosity.

Retention efficiency decreases with metallicity, often becoming negative.

Abstract

Novae are the observational manifestations of thermonuclear runaways on the surface of accreting white dwarfs (WDs). Although novae are an ubiquitous phenomenon, their properties at low metallicity are not well understood. Using the publicly-available stellar evolution code Modules for Experiments in Stellar Astrophysics (MESA), we model the evolution of accreting carbon-oxygen WDs and consider models which accrete matter with metallicity Z=0.02 or $10^{-4}$. We consider both models without mixing and with matter enriched by CO-elements assuming that mixing occurs in the process of accretion (with mixing fraction 0.25). We present and contrast ignition mass, ejected mass, recurrence period and maximum luminosity of novae for different WD masses and accretion rates for these metallicities and mixing cases. We find that models with Z = 0.02 have ignition masses and recurrence periods smaller than models with low Z, while the ejected mass and maximum luminosity are larger. Retention efficiency during novae outbursts decreases with increasing metallicity. In our implementation, inclusion of mixing at the H/He interface reduces accreted mass, ejected mass and recurrence period as compared to the no-mixing case, while the maximum luminosity becomes larger. Retention efficiency is significantly reduced, becoming negative in most of our models. For ease of use, we provide a tabular summary of our results.