1 Introduction

The principal means of determining elemental abundances in nebular plasmas has, until recently, been from the measurement of collisionally excited optical forbidden lines. The emissivities of these lines are very sensitive to the electron temperature at the temperatures typical of photoionized nebulae. An alternative method of determining abundance is to ratio the intensities of recombination lines with those of hydrogen. Such ratios are only weakly dependent on temperature. Abundances of C, N and O derived from recombination lines are, however, larger than those derived from forbidden lines by factors ranging from 2 to 10 (e.g. Liu et al. 1995). The origin of these differences is at present unexplained. The prerequisite for determination of recombination line abundances is reliable recombination coefficients for atomic ions. In this paper we present new recombination coefficients for the formation of lines of Neii.

Recombination coefficients for Neii ions have been given by Péquignot et al. (1991) considering only radiative recombination. Nussbaumer & Storey (1987) tabulated dielectronic recombination coefficients for Neii obtained from a model in which resonance states are represented by bound-state wave functions. We follow the approach of Storey (1994) who, for recombination to Oii, used a unified approach to the treatment of radiative and dielectronic recombination by calculating recombination coefficients directly from photoionization cross-sections for each initial state. A new calculation of photoionization cross-sections is carried out using the ab initio methods developed for the Opacity Project (Seaton 1987; Berrington et al. 1985) and the Iron Project (Hummer et al. 1993), hereafter referred to only as OP methods. These calculations employ the R-matrix formulation of the close-coupling method, and the resulting cross-sections are expected to be of high quality. The existing photoionization data for Neii states deposited in the Opacity Project database (Cunto et al. 1993) is inadequate for our purposes, as explained more fully below. Transition probabilities for all low-lying bound states are also calculated using the same method, so that the bound-bound and bound-free radiative data used here are expected to be significantly more accurate than those used by previous authors.

The effects of finite electron density are also incorporated, using the methods described by Hummer & Storey (1987) for hydrogenic ions, but the treatment is not complete and the results are only applicable to plasmas of relatively low electron density. The range of validity is discussed more fully in Sect. 4.1. The process of high temperature dielectronic recombination originally described by Burgess (1964) is not included in the present calculations, so the results are only appropriate for relatively low electron temperatures, $T \leq 27\,000$ K (Nussbaumer & Storey 1983). The temperature and density range are nonetheless sufficient for the analysis of the spectra of nebular objects; planetary nebulae, Hii regions and nova shells.