Quantitative spectroscopy of (pre-) white dwarfs and supersoft X-ray sources

All stars with initial masses up to 8-10 solar masses finally become white dwarfs (WDs). They are compact objects, containing roughly one solar mass within an Earth-like volume. Therefore, surface gravities are 3-4 orders of magnitude higher than for normal main-sequence stars. As a consequence, their atmospheres are (almost) “clean” in the sense that heavy elements have settled out of the atmosphere into the stellar interior. Depending on the previous evolution, the atmospheres are made of the lightest element, being either hydrogen (spectral type DA) or helium (spectral type non-DA). The origin of the two spectral sequences is still not well understood.

White dwarfs and central stars of planetary nebulae

We are therefore interested in objects which mark the transition from the immediate pre-WD stage to the beginning of the WD cooling sequence. They comprise the hottest white dwarfs and their progenitors (either central stars of planetary nebulae or subdwarf O stars). We perform quantitative spectral analyses in order to clarify their evolutionary history. We find their position in the HRD and conclude on interior nucleosynthesis and mixing processes from element abundance determinations. To this end, we perform spectroscopic observations and stellar atmosphere modeling.


Many WDs are “contaminated” by heavy elements. In hot WDs, metal “clouds” are supported by radiative levitation. In cool WDs this process cannot work. Convection can be responsible, that mixes heavy elements from the WD interior into the atmosphere. For some WDs it is believed that on-going external pollution occurs. Matter is accreted at a more or less constant rate, perhaps stemming from a debris disk that is fed by tidally disrupted asteroids. Hence, the abundance determination of trace elements in WDs allows us to investigate the physics of diffusion in stars, and to determine the composition of the accreted material, i.e., the composition of solid bodies in extrasolar systems.


Besides spectroscopic observations, we perform time-dependent photometry of (pre-) WDs with our institute’s 80 cm reflector. Often, these observations are part of world-wide network campaigns for day- or week-long uninterrupted photometry. They aim at asteroseismic investigations of pulsating stars.

Supersoft X-ray sources and novae

Accreting WDs in binary systems (= Cataclysmic Variables) can become much hotter than single white dwarfs at the beginning of their cooling phase, hence, they can be very prominent X-ray emitters. The so-called supersoft X-ray sources (SSS) are WDs accreting matter at a rate that allows steady hydrogen burning at the surface. In contrast, the nova phenomenon occurs in a low-rate accreting WD. The accreted matter is burned during recurrent thermonuclear explosions. The spectral study of SSSs and novae enables us to conclude on details of the fusion and explosion processes and related envelope mixing events.

Chandra LETG spectra showing the temporal evolution of the nova V4743 Sgr (black) compared with spectra from models (red) with T<sub>eff</sub>=720,000 K and different chemical composition, displaying the increasing C/N ratio. (From Rauch et al. 2010)

Observation of stellar spectra

The spectra that we analyse are obtained at both ground-based and space-based observatories. We utilize 4-8m class ground-based telescopes for optical observations because most of our targets are rather faint. Most observations are performed at the Calar Alto Observatory (Spain) and at the telescopes of the European Southern Observatory (Chile). Space-UV observations are also essential, because our stars are very hot (Teff=30,000-200,000 K). The chemical elements in the stellar atmospheres are highly ionized, such that the strongest spectral lines are located in the UV wavelength range. The most important facility is the Hubble Space Telescope, however, the archives of previous UV missions (ORFEUS, FUSE, IUE) are useful resources, too.


Only a few single WDs are hot enough such that soft thermal X-ray emission from their atmospheres can be detected. In contrast, SSSs and novae are among the brightest soft X-ray sources in the Sky. We are analysing spectra that are obtained with the XMM-Newton and Chandra X-ray space observatories.

Stellar atmospheres and spectral modeling

The analysis of stellar spectra is performed with stellar atmospheres and spectral synthesis codes which we have developed during the past two decades. Their essential ingredient is the physics of non-local thermodynamic equilibrium (non-LTE). In essence, the equations of radiation transfer, radiative and hydrostatic equilibrium are solved self-consistently with the non-LTE rate equations. The numerical method is based on the Accelerated Lambda Iteration. All our codes (the Tübingen Model Atmosphere Package, TMAP) and atomic input data are publicly available, or can be run via an interface provided by the German Astrophysical Virtual Observatory, GAVO).

German Astrophysical Virtual Observatory (GAVO)

In the framework of  the GAVO project, we provide spectral energy distributions (SEDs) of non-LTE model atmospheres via the so-called TheoSSA service. TheoSSA gives easy access to pre-calculated SED grids. In case that no suitable SED is available, the user is transferred to TMAW, a WWW interface to calculate SEDs using our model atmospheres codes.