Astron. Astrophys. 335, 691-702 (1998)

1. Introduction

The relative elemental abundances change significantly from the photosphere to the corona and the solar wind: elements with a first ionization potential (FIP) below 10 eV (low-FIP elements) are enriched compared to those with a FIP higher than 10 eV (high-FIP elements). The factor of enhancement is typically of the order of 4 (2) in the slow (fast) solar wind (see Fig. 1). This fractionation , also called FIP-effect , is found not only from the photosphere to the solar wind, but also from the photosphere to the corona, e.g. in polar plumes. It was already in the Sixties that a change in the abundances from the photosphere to the corona was recognized by Pottasch (1963). Much observational work on both, spectroscopic based diagnostics and in-situ measurements in the solar wind, has been done in the meantime.

Fig. 1. Measured fractionation in the solar wind in relation to oxygen.

The FIP-effect is present not only on the sun, but can also be found with stars. In a series of papers Drake et al. (1995, 1996, 1997) looked for the fractionation in three stars: -Centauri showing a similar FIP-effect as the sun, -Eridani with a smaller fractionation and Procyon, where no FIP-effect could be found. This leads to the question why some stars show a fractionation and others do not, which will be briefly discussed in the conclusions.

It is of importance to understand the abundance variations from the solar or stellar surface to the corona/wind for a number of reasons. The abundances are crucial for the diagnosis of the coronal spectra such as emission measure analysis and density or temperature diagnostics using line ratio techniques. The densities of the minor species are also important for the thermodynamics of the corona, because they dominate the radiative losses at temperatures above K (e.g. Cook et al. 1989). Furthermore the chemical composition of the solar (stellar) wind is of importance for mapping back the source region of the wind - only those parts of the chromosphere/corona can contribute a significant fraction of the mass flux to the wind that have abundances which are similar to those in the wind (see e.g. Geiss et al. 1995).

Results of fractionation models might be used to determine atmospheric parameters; e.g. in the present model the velocities of the slow and fast wind in the chromosphere can be calculated by comparing the fractionation in the wind with the model results for the chromosphere. But it may also be possible (in future models) to get information on the heating mechanisms or the magnetic topology by using the elemental abundances. Thus the theoretical understanding of the fractionation may lead to new diagnostic techniques for the chromosphere and the corona.

Many attempts have been made to understand the fractionation theoretically. See Hénoux & Somov (1992), Meyer (1993), Hénoux (1995) or von Steiger et al. (1997) for an overview. It is now widely accepted that the ion-neutral-separation is the most relevant process leading to the fractionation, located in the chromosphere, at temperatures below K (see Sect. 2.5): the step between the two plateaus is at 10 eV, (see Fig. 1) which corresponds to the energy of the Lyman- line (1215 Å), the most prominent line in the UV. As Lyman- can ionize the low- but not the high-FIPs, the photoionization and thus the separation of the ions from the neutrals must be expected to play an important role. Various scenarios were presented, but the model of von Steiger & Geiss (1989) was the first which allowed a detailed comparison of its predictions with the measurements of the fractionation in the solar energetic particles (SEP) and in the slow solar wind.

Recently Vauclair (1996) presented an interesting preliminary model with a completely different idea: horizontal magnetic field emerging from the photosphere is lifting the ions but not the neutrals. This would enrich those elements in the corona that are easy to ionize: the low-FIP elements. The main problem of this model is that it is restricted to regions with predominately horizontal magnetic fields. Thus one cannot expect an explanation for the polar plumes or the fast wind coming out of the coronal funnels.

The present paper will follow the "tradition" of the von Steiger & Geiss (1989) model. Their main idea was to combine the effects of photoionization and diffusion to describe an ion-neutral-separation. In their model the pressure gradient drives an initially neutral mixture across ambient magnetic structures. Marsch et al. (1995) applied the same process for a stationary flow along the magnetic field lines. They found a very simple formula describing the fractionation in the slow wind: the fractionation is proportional to the square root of the quotients of the respective ionization times and collision frequencies in the neutral phase. Following the philosophy of the Marsch et al. (1995) model, Peter (1996) presented an analytical study leading to a velocity-dependence of the fractionation. This enabled the explanation of the fractionation not only in the slow, but also in the fast wind within the same model.

In the present paper the analysis of Peter (1996) will be extended to explain a greater variety of fractionation phenomena summarized in Sect. 2. Some simple ideas to understand the fractionation mechanisms are presented in Sect. 3. A more detailed discussion and a comparison to measurements of the velocity-dependence as found by Peter (1996) will be presented in Sect. 5.1, as it was not done in the original paper. Applying his results to the heavy noble gases their fractionation can be understood qualitatively (Sect. 5.2). For a description of the absolute fractionation or the strong enrichments of magnesium in polar plumes numerical trace gas models have to be applied (Sect. 5.3and 5.4). The respective numerical calculations for the hydrogen background have been done by Peter & Marsch (1998). In Sect. 6the role the of diffusion for the fractionation will be discussed in more detail. Finally Sect. 7summarizes the main results of this paper.

© European Southern Observatory (ESO) 1998

Online publication: June 18, 1998
helpdesk.link@springer.de