Do White Dwarfs Sample Water-Rich Planetary Material?

TL;DR

The study uses oxygen abundances in a broad sample of polluted white dwarfs to statistically constrain the water content of accreted exoplanetary material, accounting for accretion-phase effects with Markov-chain Monte Carlo modeling. By comparing mass-buildup, mass-settling, and steady-state scenarios, the authors derive a population-wide distribution of water mass fractions with a peak near $20$–$30\%$ and find that a majority of systems are consistent with non-zero water under realistic parent-body compositions. Their results indicate white dwarfs sample a range from dry to water-rich material, with a median water fraction around $25\%$, and reveal systematic differences between H-dominated and He-dominated atmospheres. The paper also reports new abundances for three H-dominated WDs, including one confirmed water-rich object, and discusses methodological reasons for discrepancies with previous analyses.

Abstract

Polluted white dwarfs offer a unique way to directly probe the compositions of exoplanetary bodies. We examine the water content of accreted material using the oxygen abundances of 51 highly polluted white dwarfs. Within this sample, we present new abundances for three H-dominated atmosphere white dwarfs that showed promise for accreting water-rich material. Throughout, we explore the impact of the observed phase and lifetime of accretion disks on the inferred elemental abundances of the parent bodies that pollute each white dwarf. Our results indicate that white dwarfs sample a range of dry to water-rich material, with median uncertainties in water mass fractions of $\approx$15\%. Amongst the He-dominated white dwarfs, 35/39 water abundances are consistent with corresponding H abundances. While for any individual white dwarf it may be ambiguous as to whether or not water is present in the accreted parent body, when considered as a population the prevalence of water-rich bodies is statistically robust. The population as a whole has a median water mass fraction of $\approx$25\%, and enforcing chondritic parent body compositions, we find that 31/51 WDs are likely to have non-zero water concentrations. This conclusion is different from a similar previous analysis of white dwarf pollution and we discuss reasons why this might be the case. Pollution in H-dominated white dwarfs continues to be more water-poor than in their He-dominated cousins, although the sample size of H-dominated white dwarfs remains small and the two samples still suffer a disjunction in the range of host star temperatures being probed.