Helium and heavier elements in hot white dwarf atmospheres

Establishing the relationships between those groups of white dwarfs whose compositions are dominated by hydrogen or helium is crucial to understanding the evolution of the population and the physical processes that determine this. Of particular importance are accurate measurements of photospheric composition which can tell us how the helium and heavy element content of white dwarfs may change as the stars cool. Photometric observations leave several major questions unanswered.

Is there any trace helium in the atmospheres of hot H-rich white dwarfs?

Do hot white dwarfs undergo continuing mass-loss, affecting the photospheric composition?

What is the underlying structure of the white dwarf photosphere?

The availability of spectroscopic information from EUVE has provided us with an opportunity of searching directly for helium in the atmospheres of H-rich white dwarfs in the crucial region of the HeII Lyman absorption series between 304 Angstroms and the limit at 228 Angstroms. Those stars studied can be divided into two distinct groups - those with and without heavy element opacity. EUV spectra of the latter category can be described completely by a synthetic spectrum calculated from an H+He model atmosphere together with a suitable model of the interstellar medium. If heavy elements are present, the stars can only be properly studied using complex line-blanketed models with a comprehensive selection of opacity sources.

Approximately a dozen EUVE spectra of stars without heavy element opacity have been carefully examined for spectroscopic traces of helium. In general, no helium was detected (see Barstow et al., 1997, White Dwarfs , eds. Isern et al, Kluwer, pp.237-243) and only upper limits to its abundance could be obtained. Only two stars show clear evidence of photospheric helium, REJ0720-317 and GD50. However, both stars are apparently the result of binary evolution and cannot be considered representative of the general population.

Carrying out similar studies with stars containing significant quantities of heavy elements has proved more difficult as it has only recently been possible to reconcile the predicted shape and flux level of the EUV spectra of the most extreme high-opacity stars such as G191-B2B, Feige 24, REJ2214-491 and REJ0623-371, with the observations (e.g. Barstow et al. 1996a, ApJ, 473 , 1089-1093 ). Barstow et al indicated that the problem in modelling the EUV data may arise from not including a sufficiently large number of Fe lines in the calculations and demonstrated that increasing the number of lines incorporated did yield a better match to the EUV data. We have been able to compute non-LTE model atmospheres for G191-B2B, incorporating nearly 10 million Fe and Ni lines. Using heavy element abundances estimated from the far UV data we have been able to match closely the overall shape and absolute flux level of the EUV spectrum for the first time. However, this good agreement is only possible if a significant quantity of HeII is included to suppress the flux below 228 Angstroms . This additional opacity may be interstellar/circumstellar material or could reside in the photosphere of the star.

Indeed, the many lines of heavy elements present in the EUV spectrum might easily be blended with members of the HeII Lyman series, masking the signature of photospheric He. Nevertheless, the result is an important breakthrough in understanding the EUV spectra of these extreme objects. All similar stars require the additional He opacity at more or less the same level. Proving whether or not the opacity really exists, and if so, finding its location was the goal of our rocket-borne high resolution EUV spectrometer, J-PEX.

More recently, it has been shown that photospheric heavy elements may not be distributed homogeneously (by depth) and that more complex stratified structures must be considered if the overall spectral distribution across the soft X-ray, EUV and far-UV bands is to be reconciled with the models (Barstow et al 1999, MNRAS, vol 307, p884; Dreizler and Wolff 1999, A&A, vol 348, p189). Important progress has also been made in incorporating radiative levitation and diffusion self-consistently into the atmosphere calculations (Dreizler and Wolff 1999; Schuh et al 2001, in the Proc.12th European White Dwarf Workshop, ASP Conference series vol 226, p79.). The need for a He contribution is reduced in these stratified models but not eliminated.