ELEMENT FRACTIONATION BY DIFFUSION IN THE SOLAR CHROMOSPHERE

A new mechanism to explain the observed first ionization potential (FI P) fractionation of coronal and solar wind element abundances is propo sed. By the FIP fractionation, low-FIP (< 10 eV) elements are enriched in the solar corona and solar wind relative to the photosphere. This effect has been located earlier to take place in the chromosphere, at densities of N similar or equal to 10(16) - 10(18) m(-3) and a tempera ture of T similar or equal to 10(4) K, where a large fraction of the g as is still neutral. We discuss a new mechanism for the FIP fractionat ion in the form of a stationary diffusion model. It is based on a weak ly stratified chromospheric layer of constant density of the element h ydrogen and constant temperature. This layer is permeated everywhere b y ionizing photons and contains a homogeneous vertical magnetic field. Otherwise, our model does not invoke any particular geometry or speci al set up of the system. It is thus founded solely on robust and well understood atomic collisional physics. Technically, a boundary value p roblem of four coupled differential equations is solved for each chemi cal element, i.e. a continuity equation and a momentum equation for bo th atoms and singly ionized particles. By splitting the system into a main gas (hydrogen) and trace gases (16 elements from He to Xe), an an alytical solution for the former can be found. This then serves as a b ackground for the numerical integration of each trace gas system, for which we consider collisions between its atoms and ions with the main gas, i.e. protons and hydrogen. Boundary conditions are such that the gas is neutral at the bottom of the slab and fully ionized at its top, as a result of irradiation by the solar coronal EUV. Starting with a uniform density at the bottom of the layer, we find that, after a few hydrogen diffusion lengths, each minor species asymptotically approach es a constant density. The ratios of these density values to some refe rence trace element reproduce the observed FIP fractionation pattern o f heavy elements remarkably well. The step between low-FIP and high-FI P element abundances is about a factor of 5, and He is somewhat deplet ed relative to the high-FIP elements, in agreement with the observatio ns. The model fractionation pattern proves to be remarkably stable aga inst changes in the external parameters (within reasonable chromospher ic values), particularly N and T.