High-ionization coronal lines trace quasar-like activity in recently quenched galaxies at high redshift

F. Valentino, K. Ito, M. Farcy, F. Fontanot, C. Lagos, G. De Lucia, M. Hirschmann, G. Brammer, V. Kokorev, M. Hamadouche, P. Zhu, G. Scarpe, A. Pensabene, K. E. Whitaker, W. M. Baker, P. Araya-Araya, J. Antwi-Danso, D. Ceverino, A. L. Faisst, S. Fujimoto, S. Gillman, O. Ilbert, C. K. Jespersen, T. Kakimoto, M. Kubo, A. W. S. Man, G. Magdis, M. Onodera, R. Shimakawa, M. Tanaka, S. Toft, L. Xie, J. R. Weaver, P. -F. Wu

First listed 2026-06-01 | Last updated 2026-05-28

Abstract

We report the detection of the high-ionization line [NeV]$λ$3427 in the JWST/NIRSpec archival spectra of 6 massive quenched galaxies at $z \sim 1.5-4.5$, identified from a parent sample of 87 systems. With an ionization potential of approximately 97 eV, [NeV] can only be produced by strong nuclear activity in these massive systems, providing a clean and unambiguous tracer of highly accreting supermassive black holes uncontaminated by residual star formation. For 4 of the 6 [NeV]-detected systems, we detect broad H$α$ emission ($\mathrm{FWHM} \gtrsim 4000$ km s$^{-1}$), yielding black hole masses of $M_{\rm BH} = 10^{8.5-9.5}\,M_\odot$, consistent with local scaling relations with stellar mass and velocity dispersion. The [NeV] luminosities imply quasar-like bolometric outputs ($L_{\rm bol} = 10^{45-46}$ erg s$^{-1}$) and Eddington ratios of $λ_{\rm Edd} \approx 10$-$50$%, with black hole accretion rates of a few $M_\odot$ yr$^{-1}$ that match or exceed the residual star formation rates in the most extreme cases. The strongest [NeV] emitters are preferentially found in the youngest post-starburst systems ($D_n4000 \lesssim 1.3$), while old quenched galaxies are systematically devoid of such activity, a trend independently reproduced by theoretical models. These results reveal that intense, radiatively efficient SMBH growth can persist several hundred Myr after the main quenching epoch, with duty cycles of approximately 100-200 Myr. They also underscore the importance of very high accretion episodes and rates in the theoretical models that seek to reproduce the earliest quenched galaxies in the universe.

Short digest

This paper searches JWST/NIRSpec archival spectra of 87 massive quenched galaxies at z≈1.5-4.5 and finds [NeV] λ3427 in six, arguing that this coronal line cleanly isolates luminous SMBH accretion in systems where residual star formation is not a viable explanation. Four of the six [NeV] hosts also show broad Hα with FWHM ≳ 4000 km s^-1, implying MBH ≈ 10^8.5-10^9.5 Msun, while the [NeV] luminosities indicate quasar-like power outputs of roughly 10^45-10^46 erg s^-1 and Eddington ratios of about 10-50%. The strongest [NeV] emitters sit in the youngest post-starburst galaxies with Dn4000 ≲ 1.3, whereas older quenched systems lack comparable activity, pointing to intense radiatively efficient black-hole growth that can continue for several hundred Myr after the main quenching event. That timing matters because it links early quenching to delayed but still powerful SMBH growth, and provides a concrete empirical target for models that aim to produce the first massive passive galaxies.

Key figures to inspect

• Figure 1. Use this as the setup figure because it defines the parent sample, the subset with [NeV] coverage, and where the six detections sit in stellar mass and redshift relative to the broader quenched-galaxy population. The overplotted broad- and intermediate-line classifications also make clear from the start that the coronal-line detections are tied to the most AGN-like systems rather than being a generic feature of all quenched galaxies.
• Figure 2. This is the core observational evidence figure: it shows the NIRSpec spectra, the pPXF decomposition into stellar and gas components, and the zoomed [NeV] windows for the detected sources. It lets readers see directly that the headline claim rests on identifiable coronal-line detections in individual quenched galaxies, not only on catalog-level measurements.
• Figure 6. This figure should be highlighted because it connects the [NeV] and broad-Hα detections to the paper’s physical interpretation. The left panel places the broad-line systems on the MBH-Mstar relation, while the right panel shows MBH versus Eddington ratio and demonstrates that these sources occupy the regime of massive black holes accreting at substantial fractions of Eddington, consistent with quasar-like activity after quenching.
• Figure 8. Include this later synthesis figure because it directly compares the observed quenched galaxies with AGN activity to predictions from Shark, Gaea, and Black Dawn/Mistral. It is the cleanest visual summary of the paper’s argument that matching the observed stellar masses, black-hole masses, MBH/Mstar ratios, and Eddington ratios requires models to produce very strong post-quenching accretion episodes.
• Figure 9. This is the best figure for the timing claim in the abstract. By showing Eddington ratio versus time since quenching across multiple models and AGN luminosity thresholds, it makes the duty-cycle interpretation concrete and illustrates why the authors conclude that luminous SMBH growth can persist for roughly 100-200 Myr, and in some cases several hundred Myr, after the main quenching phase.