1 Introduction

The Stark widths which are measured in accessible and well-known light sources, besides to improve the knowledge about spectral line broadening phenomena, allow to diagnose remote light sources of great interest to astrophysics.

Oxygen plays an important role in astrophysical, technical and laboratory plasmas; nevertheless the amount of available experimental information on the Stark broadening for the species OII continues to be rather scarce. On the other hand, the different theoretical and semiempirical approaches show some discrepancies and need to be compared with experimental measurements.

This work furnishes experimental Stark widths of 32 different OII spectral lines, in the range 400-470 nm, belonging to $\rm 3s-3p$, $\rm 3s'-3p'$, $\rm 3p-3d$, $\rm 3p'-3d'$, $\rm 3p'''-3d'''$ and $3{\rm d}-4{\rm f}$ transition arrays. For 16 of these lines no previous measured data have been found in the literature. The Stark widths have been measured in relatively large ranges of electron temperature and density. Standard diagnostic techniques, carefully checked, have been used to get these plasma parameters. Double and independent methods have been applied for each parameter, Starting with one wavelength interferometry using the 457.9 nm and also 514.0 nm transitions of an argon ion laser and H$_\alpha$ Stark broadening to determine electron density. Second, the intensities of 14 OII selected lines, and also 13 NeII lines emitted by the same plasma, help determine the electron temperature by means of different Boltzmann-plots. The measured profiles have been corrected for Doppler and instrumental broadening. Self-absorption has been proved to be completely negligible.

In the ranges of electron density and temperature of this work the measured Stark widths verify a good linear dependence on plasma electron density. After taking into account the experimental uncertainties, this work results are in good agreement with previous measures like those of Platisa et al. (1975); Djenize et al. (1991, 1998) and with theoretical works by Griem (1974); Hey & Breger (1980) and Dimitrijevic (1982). The data dispersion and discrepancies around 20% between the different works do not allow establish a clear dependence on temperature.