PHOTOIONIZATION MODELS FOR THE SEMI-FORBIDDEN C III] 1909 EMISSION IN STAR-FORMING GALAXIES

The increasing neutrality of the intergalactic medium at z > 6 suppresses Ly alpha emission, and spectroscopic confirmation of galaxy redshifts requires the. detection of alternative ultraviolet lines. The strong [C III] lambda 1907 + C III] lambda 1909 doublet frequently observed in low-metallicity, actively star-forming galaxies is a promising emission feature. We present CLOUDY photoionization model predictions for C III] equivalent widths (EWs) and line ratios as a function of starburst age, metallicity, and ionization parameter. Our models include a range of C/O abundances, dust content, and gas density. We also examine the effects of varying the nebular geometry and optical depth. Only the stellar models that incorporate binary interaction effects reproduce the highest observed C III] EWs. The spectral energy distributions from the binary stellar population models also generate observable C III] over a longer timescale relative to single-star models. We show that diagnostics using C III] and nebular He II lambda 1640 can separate star-forming regions from shock-ionized gas. We also find that density-bounded systems should exhibit weaker C III] EWs at a given ionization parameter, and C III] EWs could, therefore, select candidate Lyman continuum-leaking systems. In almost all models, C III] is the next strongest line at < 2700 angstrom after Lya, and C III] reaches detectable levels for a wide range of conditions at low metallicity. C III] may therefore serve as an important diagnostic for characterizing galaxies at z > 6.