Dielectronic recombination and autoionization yields in weak static electric fields

Dielectronic recombination (DR) of ions and electrons occurs via intermediate doubly excited Rydberg states converging to an excited state of the ion, and the total energy-integrated rate is the sum of the rates through the n, l Rydberg states. To make stringent comparisons between calculated and observed rates, it is useful to resolve the contributions to DR of different n and l states. It is possible to separate experimentally the n states due to their different energies, but it is not possible to separate the nearly degenerate high-l states. We propose that it should be possible to measure the contribution to DR of energetically unresolved high-l states by measuring the recombination rate as a function of electric field. As the field is raised, progressively lower l states are converted to Stark states. Autoionization rates increase as l is decreased, and when the field is raised to the point that an l state with an autoionization rate in excess of the radiative decay rate is added to the manifold of Stark states, the DR rate will exhibit an observable increase. To test this proposal we have made measurements of autoionization yields of the autoionizing Ba 6p(1/2)nk and 6p(3/2)nk Stark states. The autoionization yield is complementary to DR, and the measurements indicate that the proposed approach should work well. Although it is not surprising that this approach works for excited ion states that are isotropic, such as the Ba+ 6p(1/2) state, it is less obvious that it should work for an anisotropic ion state, such as the Ba+ 6p(3/2) state. In the latter case there are four quantum defects for each l state instead of one, and the Stark effect of the Ba 6p(3/2)nk Stark states is far more complex than that of the Ba 6p(1/2)nk Stark states. Calculations of the Stark effect reveal that, while there are four times as many 6p(3/2)nk levels as 6p(1/2)nk levels, many of the interactions of the 6p(3/2)nk levels are negligible, and in practice the problem is no more complicated than the 6p(1/2)nk Stark problem.