Element Segregation in Stellar Outer Layers

The fundamental importance of the diffusion of chemical elements in stars was recognized by the pioneers of stellar structure computations. The historic paper by Eddington (19 16) (On the Radiative Equilibrium of the Stars), which gave the first hint about stellar structure, was closely followed by Chapman's ( 1917 c) paper Convection and Diffusion in Giant Stars . This paper was a first attempt to determine the relative importance of convection and diffusion within a star. In his computations of diffusion processes, Chapman introduced the effects of the pressure and temperature gradients . He then believed that thermal diffusion opposed gravitational diffusion in stars. Eddington (1926) pointed out (and Chapman confirmed it later) that this was not the case for ionized particles repelling each other with an inverse square law: the heavy particles then diffuse towards the hotter regions so that thermal diffusion acts in the same direction as gravitational diffusion. Chapman ne­ glected the effect of the electric field, and he assumed that radiation pressure acted proportional to the masses of the particles, so that it had no selective effect. He found that no heavy elements should be left in the atmospheres, which was obviously not the case. His conclusion was that either radiation pressure acted selectively on the various chemical elements (but he had no means to prove this at the time), or mixing by convection was efficient enough to prevent diffusion. A discussion about selective radiation pressure appeared in Eddington (1926, Ch. 1 0). He explained how the radiation force is much stronger on heavy atoms than on light ones because of line absorption: contrary