Multi-wavelength properties of changing-state active galactic nuclei: I. the evolution of soft excess and X-ray continuum

TL;DR

This work investigates how the soft X-ray excess and X-ray continuum evolve in changing-state AGNs (CSAGNs) as their accretion rates span roughly $\lambda_{\rm Edd} \sim 10^{-4}$ to $3\times10^{-1}$. By combining over 1000 observations from Swift, XMM-Newton, NuSTAR, and Suzaku for five CSAGNs, the authors quantify key spectral components (soft excess via a warm corona, primary continuum, and reflection) and derive $Q$, $R_S$, and $\lambda_{\rm Edd}$. They find a tight, nonlinear SE–PC connection and a characteristic V-shaped $\Gamma$–$\lambda_{\rm Edd}$ relation with a break near $\log\lambda_{\rm Edd}^{\rm break} = -2.47\pm0.09$, indicating a geometry transition in the inner accretion flow; the soft excess scales positively with $\lambda_{\rm Edd}$ and vanishes below $\log\lambda_{\rm Edd} \sim -2.5$, supporting warm Comptonization as the SE origin rather than blurred reflection. SE changes outpace continuum changes during CS transitions, and while the soft excess correlates with accretion activity, reflection signatures remain weak, reinforcing the warm corona scenario and providing a real-time view of accretion physics in SMBHs.

Abstract

Changing-state active galactic nuclei (CSAGNs) exhibit rapid variability, with mass accretion rates that can change by several orders of magnitude in a few years. This provides us with a unique opportunity to study the evolution of the inner accretion flow almost in real time. Here, we used over 1000 observations to study the broadband X-ray spectra of a sample of five CSAGNs, spanning three orders of magnitude in Eddington ratio ($λ_{\rm Edd}$), using phenomenological models to trace the evolution of key spectral components. We derive several fundamental parameters, such as the photon index, soft excess strength, reflection strength, and luminosities of the soft excess and primary continuum. We find that the soft excess and primary continuum emissions show a very strong positive correlation ($p \ll 10^{-10}$), suggesting a common physical origin. The soft excess strength does not show any dependency on the reflection parameter, suggesting that in these objects the soft excess is not dominated by a blurred ionized reflection process. On the other hand, the strength of the soft excess is found to be strongly positively correlated with the Eddington ratio ($p \ll 10^{-10}$), and we find that the soft excess vanishes below $\log λ_{\rm Edd} \sim -2.5$. Moreover, we find a clear `V'-shaped relation for $Γ-λ_{\rm Edd}$, with a break at $\log λ_{\rm Edd} = -2.47 \pm 0.09$. Our findings indicate a change in the geometry of the inner accretion flow at low Eddington ratios, and that the soft excess is primarily produced via warm Comptonization.

Figures (23)

• Figure 1: Representative unfolded broadband X-ray spectra of NGC 1566 in type 1 and type 2 state, in the left and right panel, respectively. In the left panel, the orange and blue points represent the data from XMM-Newton/EPIC-PN and NuSTAR/FPMA observations, respectively. In the right panels, the black, orange, and blue points represent the Suzaku/XIS1, Suzaku/XIS0+XIS3, and Suzaku/HXD-PIN observations, respectively. The solid black, dashed-dot black, red dashed, solid magenta, and dotted green lines represent the total, continuum, soft excess, iron K-line and reprocessed emission, respectively. The middle figures of each panel show the residual while data are fitted without the 'soft-excess' components. The bottom panels show the residuals for the full models.
• Figure 2: Light-curves of all five CSAGNs in different panels. Blue circles and orange diamonds mark the type 1 and type 2 states, respectively. The vertical lines in each panel represent the optical observations, where solid blue and orange dot-dashed lines represent the type 1 and type 2 spectral states, respectively. The solid horizontal line in each panel represent the median of transition Eddington ratio $\log \lambda_{\rm Edd}^{\rm tr}$ at $-2.01 \pm 0.23$ from AJ2025cl. The dot-dashed horizontal line represent the break-Eddington ratio ($\log \lambda_{\rm Edd}^{\rm break}=-2.47\pm0.098$.)
• Figure 3: Distribution of blackbody temperature ($kT_{\rm BB}$). The blue dashed and red dash-dot lines represent the median value for type 1 and type 2 states, respectively.
• Figure 4: Variation of the soft-excess luminosity ($L_{\rm SE}^{\rm 0.5-2}$) in 0.5--2 keV energy range as a function of continuum luminosity ($L_{\rm PC}^{\rm 2-10}$) in 2--10 keV flux. The filled blue circles and orange squares represent the data from type 1 and type 2 states, respectively. The hollow blue circles and orange squares represent the upper limit from type 1 and type 2 state, respectively. The yellow stars represent the binned data points. The solid black line represents the linear best-fit. The gray regions mark the $1\sigma$ scatter. The red solid line represents the linear best-fit, considering only type 1 state. The green dashed-dot and blue dashed lines represent the linear best-fit for bare AGNs from Nandi2023 and unobscured BASS AGNs from Jana et al. (in prep).
• Figure 5: Relation between $\Gamma$ and $\lambda_{\rm Edd}$. The blue circles and orange squares represent the data from type 1 and type 2 states, respectively. The yellow stars represent the binned data points. The red lines represent the linear best-fit of the dataset with a break at $\log \lambda_{\rm Edd}=-2.47\pm 0.09$. The break-point is marked by the vertical gray dashed line. The vertical green dash-dot line represent the median of the transition Eddington ratio ($\lambda_{\rm Edd}^{\rm tr}$) for CSAGNs, which is $\log \lambda_{\rm Edd}^{\rm tr} = -2.01 \pm 0.23$. The $\lambda_{\rm Edd}^{\rm tr}$ is taken from AJ2025cl. The black dashed-dot line represents the $\Gamma-\lambda_{\rm Edd}$ relation of BASS AGNs, adopted from Trakhtenbrot2017.