A Census of Active Galactic Nuclei Identified by the Mid-infrared [Ne v] Line at $\boldsymbol{z \leq 0.025}$

Abstract

We present a census of local active galactic nuclei (AGN) at a redshift of $z\leq0.025$ selected using the high-ionization [Ne v] $\lambda14.32\,μ$m emission line from the Infrared Database of Extragalactic Observables from Spitzer (IDEOS). We identify 103 sources with detected [Ne v] emission, which we regard as AGN within the volume. This sample represents $\sim18\%$ of the galaxy population within this redshift range, consistent with AGN fractions derived using other selection techniques. We investigate the biases and properties of this [Ne v]-selected AGN sample by comparing it with traditional AGN selection methods based on hard X-ray, optical, and mid-infrared colors. We find that our selection significantly misses AGN with underdeveloped narrow line regions (NLRs), which account for approximately half of the AGN identified by NLR-independent methods. However, approximately $\sim10\%$ of our sample are undetected in optical diagnostics, while $\sim40\%$ are missed by hard X-rays and $\sim70\%$ by infrared continuum. Notably, $\sim15\%$ of our AGN are missed by all classical methods, constituting a population of previously unidentified AGN revealed solely by the [Ne v] emission line. Based on our analysis, we show that this line can efficiently select heavily Compton-thick and host-dominated AGN systems. Our analysis also yields mean bolometric luminosities of $\log(L_{\rm bol}/{\rm erg~s^{-1}})=44.5\pm0.7$, black hole masses of $\log(M_{\rm BH}/M_{\odot})=7.3\pm0.6$, and Eddington ratio of $λ_{\rm Edd}=0.15\pm0.11$. Our sample harbors AGN with comparable luminosities but systematically lower-mass black holes accreting at higher Eddington ratios than those in the hard X-ray-selected sample. This suggests that our AGN may represent local analogs of the rapidly growing SMBH population prevalent at cosmic noon.

Figures (8)

• Figure 1: Top: Redshift distribution of the full IDEOS sample (black solid line), and [Ne v]-selected galaxies (gray shaded). The [Ne v] contributes to $\sim 11\%$ of the whole IDEOS catalog spoon22. The [Ne v] AGN sample shows a strong concentration at low redshifts while diminishing exponentially indicating some selection biases of Spitzer/IRS instrument. Bottom: Distribution of [Ne v]$\lambda$14.32 flux and luminosity against redshift. We empirically determined a flux completeness limit by identifying the maximum redshift boundary. The vertical line marks the boundary for the AGN complete sample (z = 0.025).
• Figure 2: Diagnostic diagram of $\log$([Ne v]/[Ne ii]) vs. $\log$([Ne iii]/[Ne ii]) for our [Ne v]-selected AGN sample, using empirical AGN boundaries from inami13 (gray shaded region, $\log$([Ne v]/[Ne ii]) $\geq$ -1) and melendez08 (red shaded region, $\log$([Ne iii]/[Ne ii]) $\geq$ 0). Blue circles mark sources below our luminosity threshold (Section \ref{['sample-2.1.1']}); seven of these lie inside the AGN region, confirming their AGN origin, while NGC 6328 lies just outside but close to the AGN boundary. Although many sources fall outside the AGN region, their high [Ne v] luminosities (above those of any known non-AGN source) confirm that all objects in the sample are genuine AGN, with star formation contributing to the line ratios of some objects. We also include Wolf–Rayet (WR) sources for comparison. Line ratios from tarantino24 (magenta squares) were measured in a WR nebula, while two WR galaxies from goulding09 with [Ne v] upper limits are shown as red arrows.
• Figure 3: Estimated Swift/BAT fluxes for the X-ray–undetected ($\sim$40%) AGN inferred from the [Ne v]–BAT relation as a function of redshift. The red line and shaded region denote the $5\sigma$Swift/BAT sensitivity limit. We found that out of 41 AGN undetected by Swift/BAT, 10 have expected BAT flux below the BAT sensitivity limit flux while 31/41 sources are above the limit indicating that they could have been detected by Swift/BAT. This indicates that they may be heavily Compton-thick AGN.
• Figure 4: Mid-infrared color–color diagram based on WISE photometry. The solid wedge delineates the AGN selection region defined by mateos12. Marker shapes indicate X-ray classifications, while colors denote optical classifications. Approximately $70\%$ of the [Ne v]-selected AGN lie outside the canonical WISE selection wedge and exhibit systematically bluer W1$-$W2 colors than typical AGN populations, suggesting that their mid-infrared emission is dominated by stellar processes rather than by luminous, dust-obscured AGN activity.
• Figure 5: Normalized histogram showing host galaxy inclination angle distributions for optically detected (red outline) and unidentified AGN (blue shaded). Mean inclinations are $57.8^\circ \pm 21.5$ for optical and $57.0^\circ \pm 21.4$ for non-optical AGN. A Kolmogorov–Smirnov test between the two distributions shows no significant difference ($D_{\rm KS} = 0.21$, $P_{\rm KS} = 0.75$), indicating that galaxy inclination does not affect optical detection rates.