Constraining the X-ray reflection in low accretion rate AGN using XMM-Newton, NuSTAR and Swift

An interesting feature in active galactic nuclei (AGN) accreting at low rate is the weakness of the reflection features in their X-ray spectra, which can result from the gradual disappearance of the torus with decreasing accretion rates. It has been suggested that low luminosity AGN (LLAGN) would have a different reflector configuration compared with high luminosity AGN, either covering a smaller fraction of the sky or simply having less material. Additionally, we note that the determination of the spectral index ($\Gamma$) and the cut-off energy of the primary power-law emission is affected by the inclusion of reflection models, showing the importance of using them to study the accretion mechanism, especially in the case of the LLAGN that have previously shown a high dispersion on the relation between $\Gamma$ and the accretion rate. Our purpose is to constrain the geometry and column density of the reflector in a sample of LLAGN covering a broad X-ray range of energy combining data from XMM-Newton + NuSTAR + Swift of a hard X-ray-flux limited sample of 17 LLAGN from BASS/DR2 with accretion rates $\lambda_{Edd}$=L$_{\rm Bol}$/L$_{\rm Edd}$<10$^{-3}$. We fit all spectra using the reflection model for torus (borus02) and accretion disk (Xillver) reflectors. We found a tentative correlation between the torus column density and the accretion rate, LLAGN shows a lower column density compared with the high-luminosity objects. We also confirm the relation between $\Gamma$ and $\lambda_{Edd}$, with a smaller scatter than previously reported, thanks to the inclusion of high-energy data and the reflection models. Our results are consistent with a break at $\lambda_{Edd}\sim10^{-3}$, suggestive of a different accretion mechanism compared with higher accretion AGN.