NGC 4395: Evidence Against the Starburst Hypothesis for Seyfert 1 Nuclei and QSOsa

In a very provocative series of papers, Terlevich et al. (1992, and references therein) argue that the properties of active galactic nuclei (AGNs) can be explained by vigorous bursts of star formation in the high-metallicity nuclei of galaxies, especially those of early Hubble type. Specifically, Terlevich & Melnick show that the ultraviolet spectrum of a cluster of massive stars can resemble a power law of index ɑ = -1.5 (where f_v ∝ v^ɑ), as in classical AGNs. The ionizing continuum will therefore cause nearby gas to produce an emission-line spectrum similar to that
\nof Type 2 Seyfert nuclei and low-ionization nuclear emission-line regions (LINERs). The starburst model is extended to include Seyfert 1 galaxies and radio-quiet QSOs
\nby considering the effect of Type II (hydrogen-rich) supernovae evolving in a dense (n ≈ 10^7 cm-^3) circumstellar medium. These objects become strongly radiative while still expanding at velocities of several thousand km s^(-1). Empirical evidence for this hypothesis is provided by the observed spectra of some Type II supernovae, which do indeed superficially resemble those of Seyfert 1 nuclei Only one supernova per year suffices for an AGN with M_B ≈ -21 mag. A range of stellar masses and
\nevolutionary phases is required, however, to explain the multiwavelength spectra of AGNs. For example, the high-excitation spectrum of the narrow-line region is
\nproduced by clouds of gas irradiated by hot, evolved Wolf-Rayet stars in the cluster. These stars could account for the smooth ultraviolet/optical continuum, and the
\nassociated red supergiants would produce the near-infrared Ca II absorption triplet often observed at least in Seyfert 2 nuclei. Heckman and Filippenko discuss the strengths and weaknesses of the starburst hypothesis for AGNs. Their general conclusion is that the scenario has severe
\nproblems for AGNs that have radio jets, show rapid X-ray variability, or emit a substantial fraction of their power at hard X-ray and gamma-ray energies. On the other hand, it might explain some Seyfert 2 nuclei and a subset of LINERs
\nand perhaps those QSOs and Seyfert 1 nuclei lacking the above properties.