Cosmic Duets I. High-spatial resolution spectroscopy of dual and lensed AGN with MUSE

TL;DR

This study addresses the census and characterization of dual and lensed AGN at sub-arcsecond separations by combining Gaia GMP pre-selection with AO-assisted MUSE spectroscopy. The authors present the first-year results for 30 GMP-selected targets, classifying 6 dual AGN, 13 lensed AGN (including 3 quadruple lenses), and 11 AGN+star alignments, with redshifts up to $z\sim3.2$ and separations down to $\sim0.2''$. The methodology—spectral decomposition, IFU-based extraction, and microlensing-aware classification via MmD—yields redshifts, emission-line diagnostics, absorption-line inventories (intrinsic and intervening NALs), and detailed lens properties. The findings demonstrate the efficiency of GMP pre-selection with MUSE in uncovering tightly coupled SMBH systems and provide a foundation for statistical studies of dual AGN incidence and precise strong-lensing mass models in the full program of ~150 targets, with broad implications for SMBH growth, galaxy evolution, and gravitational-wave progenitors.

Abstract

We present the first-year results of the MUSE Large Program "Cosmic Duets", aimed at obtaining adaptive-optics assisted MUSE observations with an angular resolution of 0.1"-0.2", providing integral-field spectroscopy of sub-arcsec separation dual and lensed active galactic nuclei (AGN) candidates. These observations reveal previously unexplored properties of dual and lensed systems, key to understanding galaxy evolution, black hole mergers, and strong-lensing modeling. Targets were selected using the Gaia Multi-Peak (GMP) technique, which identifies pairs of point-like sources with separations below 0.8" in the Gaia catalog. MUSE spatially resolved spectroscopy provides redshifts, ionization diagnostics, and absorption systems along the line of sight. We report results for 30 GMP-selected targets at z=0.5-3.5. All systems show at least two spatially resolved components. Nineteen objects are confirmed as AGN multiplets, including 6 dual AGN, 10 doubly-lensed quasars, and 3 quadruply-lensed systems, while the remaining 11 correspond to alignments with foreground stars. Among spectroscopically confirmed dual AGN in the literature, 24 pairs have separations below 7 kpc, and our sample accounts for 25% of them. We study dual and lensed AGN distributions as a function of redshift, magnitude, and projected separation, and find that bright systems (J<16.5) are dominated by lensed quasars, whereas the fraction of dual AGN increases at fainter magnitudes. This first-year sample demonstrates the high efficiency of GMP pre-selection combined with MUSE spectroscopy. The full program, targeting 150 systems, will enable statistical studies of dual AGN incidence and detailed constraints on mass distribution in strong-lensing galaxies.

Figures (14)

• Figure 1: Distribution of dual (green) and lensed (blue) AGN as a function of projected separation (top panel) and redshift (bottom panel). The sample, observed with MUSE–NFM, includes sources presented in this work and by Scialpi24. For quads, the Einstein radius was used to characterize the separation.
• Figure 2: White-light images of the double AGN systems (lensed or dual) in unit of 10$^{-15}$ erg/s/cm$^{-2}$. The target name and the separation between the two components are indicated in each panel. The spectra are extracted from the circular apertures shown on the maps, with purple and blue apertures corresponding to the A (brightest) and B (faintest) components, respectively. Dual systems are marked with the flag D, lensed systems with L. For systems identified as gravitational lenses, if an additional lensed image is detected, it is shown inside the red aperture and labeled as component G.
• Figure 3: White-light images of the three quadruple AGN systems. The target name and the value of the Einstein radius ($R_e$) is indicated in each panel. The spectra are extracted from the colored circular apertures. The colors of the apertures follow the order of brightness: A (purple, brightest), B (blue), C (orange), and D (green, faintest). For J0957, we also detect the lensing source (E) in red.
• Figure 4: Final MUSE spectra of the dual AGN systems, with target names and redshifts indicated in each panel. The spectra were extracted from the circular apertures shown in Figure \ref{['fig:cubes_AGN']} (same color-coding), corrected for Galactic extinction, normalized to Gaia magnitudes, and shifted to the rest frame. For clarity, the spectra of the primary AGN (component A) have been re-normalized to facilitate comparison with those of component B; the applied renormalization factor is indicated in each panel. Vertical dotted purple lines mark the expected positions of prominent emission lines at redshift of AGN A. The gap around $\lambda_{\rm obs} = 6000~\AA$ (observed frame) corresponds to the New Generation Controllers (NGC) used in AO observations.
• Figure 5: Final MUSE spectra of the double lensed AGN systems, with target names and redshifts indicated in each panel. The spectra were extracted from the circular apertures shown in the white-light images of Figure \ref{['fig:cubes_AGN']} (same color-coding), corrected for Galactic extinction, normalized to Gaia magnitudes, and shifted to the rest frame. Vertical dotted purple lines mark the expected positions of prominent emission lines at redshift of AGN A. The gap around $\lambda_{\rm obs} = 6000~\AA$ corresponds to the NGC used in AO observations.