The [Ne V] AGN Diagnostic in SDSS-IV/eBOSS: An Anticorrelation Between Ionization State and AGN Luminosity in Low-Redshift Coronal-Line Galaxies

Abstract: Traditional narrow-line diagnostics fail at high redshift, where low metallicity drives star-forming galaxies and Active Galactic Nuclei (AGN) into overlapping regions of the Baldwin-Phillips-Terlevich (BPT) diagram. Coronal lines, with ionization potentials exceeding 100 eV, offer a robust alternative: they cannot be produced by normal stellar populations, arise almost exclusively from AGN or fast shocks, and are insensitive to abundance evolution. Among these, the [Ne V] doublet is the brightest optical coronal line tracer of AGN activity. Existing coronal line catalogs contain only about 1750 objects, too few to statistically assess the completeness and purity of [Ne V]-selected AGN samples. We identify 33,817 galaxies with [Ne V] S/N > 5 at z = 0.147-1.12 in the Sloan Digital Sky Survey IV extended Baryon Oscillations Spectroscopic Survey (SDSS-IV/eBOSS), exceeding the combined literature total by more than an order of magnitude. [Ne V] preferentially selects rapidly accreting, relatively unobscured AGN; only 41.5% of BPT AGN show detectable emission. By stacking spectra that lack [Ne V] on a grid of black hole mass and [O III] luminosity, we recover [Ne V] in the majority of bins, implying that non-detections reflect limited survey sensitivity rather than a deficit of coronal emission. We find [Ne V]/[Ne III] is only weakly correlated with [O III]/[O II]. Galaxies with high [Ne V]/[Ne III] for their [O III]/[O II] ratio have lower Eddington ratios and are more likely classified as LIERs or Composites. The [Ne V]/[O III] ratio shows a strong anticorrelation with [O III] luminosity, with the most extreme coronal-line strengths found among the lowest-luminosity AGN, in analogy with the accretion states of stellar-mass black holes. Comparison with a sample of z=2-9 [Ne V]-detected galaxies suggests some, but not all, high-redshift AGN follow the same relations.