The Cooling of Old White Dwarfs in 47 Tucanae

TL;DR

This study uses deep Hubble observations and stellar modeling to analyze the late-stage cooling of white dwarfs in 47 Tucanae, focusing on core crystallization, envelope convection, and element diffusion effects.

Contribution

It compares different element diffusion treatments and envelope thicknesses in white dwarf models, providing insights into late cooling behavior and validating MESA's diffusion implementation.

Findings

Thicker hydrogen envelopes are favored in models.

Standard MESA diffusion treatment fits observed luminosity functions well.

Models with non-ideal gas effects offer alternative cooling scenarios.

Abstract

We analysed the cooling of white dwarfs in the globular cluster 47 Tucanae using deep observations from the Hubble Space Telescope that resolve the white dwarf cooling sequence to late enough cooling times that the white dwarf core has begun to crystallise and the envelope has become convectively coupled to the core. At such late cooling times, both the state of matter assumed for ions in the treatment of element diffusion and the thickness of the outer H envelope become important considerations for modelling white dwarf cooling. Using the stellar evolution software Modules for Experiments in Stellar Astrophysics (MESA), we created a suite of white dwarf cooling models for different treatments of element diffusion, as well as different values of the white dwarf mass and H envelope thickness parameters. Three different diffusion scenarios were considered: i) the standard MESA implementation, which implicitly uses an ideal gas approximation for the ions, ii) a custom modified implementation that accounts for non-ideal gas effects, and iii) no diffusion. An unbinned likelihood analysis was performed to compare these cooling models to the observations. This work both constrains the values of parameters important for modelling white dwarf cooling and tests the implementation of element diffusion in MESA to late cooling times. We find that models with thicker H envelopes are preferred and that the standard MESA diffusion treatment produces a best-fitting model that well reproduces the cumulative white dwarf luminosity functions of the observations.