Phase Separation in Ultramassive White Dwarfs

Abstract

Ultramassive white dwarfs are extreme endpoints of stellar evolution. Recent findings, such as a missing multi-Gyr cooling delay for a number of ultramassive white dwarfs and a white dwarf with a quasi-Chandrasekhar mass, motivate a better understanding of their evolution. A key process still subject to important uncertainties is the crystallization of their dense cores, which are generally assumed to be constituted of $^{16}$O, $^{20}$Ne, and a mixture of several trace elements (most notably $^{23}$Na and $^{24}$Mg). In this work, we use our recently developed Clapeyron integration technique to compute accurate phase diagrams of three-component mixtures relevant to the modeling of O/Ne ultramassive white dwarfs. We show that, unlike the phase separation of $^{22}$Ne impurities in C/O cores, the phase separation of $^{23}$Na impurities in O/Ne white dwarfs cannot lead to the enrichment of their cores in $^{23}$Na via a distillation process. This severely limits the prospect of transporting large quantities of $^{23}$Na toward the center of the star, as needed in the white dwarf core collapse mechanism recently proposed by Caiazzo et al. We also show that despite representing $\approx 10\%$ of the ionic mixture, $^{23}$Na and $^{24}$Mg impurities only have a negligible impact on the O/Ne phase diagram, and the two-component O/Ne phase diagram can be safely used in white dwarf evolution codes. We provide analytic fits to our high-accuracy O/Ne phase diagram for implementation in white dwarf models.