An X-Ray Spectral Study of the Origin of Reflection Features in Bare Seyfert 1 Galaxy ESO 511–G030

To investigate the excess emission in $6\mbox{--}7\,\mathrm{keV}$, we first introduced a narrow Gaussian to the absorbed power-law model. An energy-independent multiplicative factor was used to account for the relative normalization between different instruments of Suzaku. Assuming a bare nature of the source we did not include any neutral or partially ionized absorber model to the set of models used here. In XSPEC the model reads as constant× tbabs× (po+ zgauss). The best-fit statistics for this set of models is $\,{\chi }^{2}/\mathrm{dof}=1587/1414$. The line energy is consistent with $6.4\,\mathrm{keV}$ for all three observations and the equivalent width ranges between $62\,\mathrm{eV}$ to $94\,\mathrm{eV}$. We found the source frame line width $\sigma =0.08\mbox{--}0.14\,\mathrm{keV}$ between observations, which indicates the presence of a broad Fe emission line in the source spectra. The broad Gaussian if replaced with a narrow Gaussian line ($\sigma =0.01\,\mathrm{keV}$) provides a relatively poor fit statistics to the data sets ($\,{\chi }^{2}/\mathrm{dof}=1603/1417$). Next, we replaced the broad Gaussian line with a more realistic broad Fe line profile diskline, (Fabian et al. 1989), which resulted in a similar fit statistics ($\,{\chi }^{2}/\mathrm{dof}=1587/1414$). In the diskline model, the inner and outer radius of the disk was fixed at ${r}_{\mathrm{in}}=6{r}_{{\rm{g}}}$ and ${r}_{\mathrm{out}}=400{r}_{{\rm{g}}}$, respectively, for all three observations. The power-law dependence of emissivity (β), the line energy and the model normalizations are made free for obs 1 and obs2. The inclination angle is tied between observations. The best-fit line energy is consistent between observations with a value of 6.42 keV. The high value of emissivity profile (q = 0.9–9.1) indicates a large amount of coronal radiation intercepts the disk.

Alternatively, a partial covering absorber model (neutral or ionized) may mimic the red wing of the broad Fe emission line in the broadband X-ray spectrum. The partial covering model with neutral or ionized absorption (zpcfabs or zxipcf) when multiplied with the absorbed (galactic) power-law model resulted in a poor statistics for all the three data sets (${\chi }^{2}/\mathrm{dof}=1868/1421$ and ${\chi }^{2}/\mathrm{dof}=1873/1422$ for zpcfabs or zxipcf, respectively). See Table 2 for the comparison of a different set of models used to fit the Fe line emission. Our analysis indicates the presence of a broad Fe Kα emission line in the X-ray spectra of all three observations. Next, we study the full X-ray band to further characterize the reflection features.

Table 2.  The Comparison of Different Models Used to Fit the Excess Around $6\,\mathrm{keV}$ for the Joint Analysis of Two Suzaku and XMM-Newton Observations of ESO 511–G030

Models	Obs-1	Obs-2	Obs-3	Simultaneous Fit
Used	${\chi }^{2}/\mathrm{dof}$	${\chi }^{2}/\mathrm{dof}$	${\chi }^{2}/\mathrm{dof}$	${\chi }^{2}/\mathrm{dof}$
model 1 + zxipcf	209/109	1091/883	567/437	1868/1420
model 1 + narrow Gaussian	145/105	945/881	513/433	1603/1417
model 1 + broad Gaussian	131/104	942/880	510/433	1587/1414
model 1 + diskline	133/104	943/879	512/433	1587/1414
model 1 + relline	137/105	946/880	516/433	1599/1411

Note. Notes: Model 1 is above $3\,\mathrm{keV}$ power-law fit modified by the Galactic absorption. The normalization parameter of all the models were made free for each observations.

Download table as:  ASCIITypeset image

We initially analyzed the broadband X-ray data with a phenomenological set of models to test the presence and strength of reflection features in the source spectra. The baseline phenomenological model used here includes a neutral galactic absorption (tbabs), a multiple blackbody component to model the soft excess (diskbb, Mitsuda et al. 1984) and the coronal emission described by a power law. In addition, we used a diskline model to describe excess emission at around $6.4\,\mathrm{keV}$. An energy-independent multiplicative factor was introduced to account for the relative normalization of different XIS and PIN instruments. In XSPEC notation, the phenomenological model reads as constant× tbabs× (powerlaw+diskbb+diskline). Addition of a ztbabs model did not improve the fit statistics and we conclude that all three observations are free from any intrinsic neutral absorption of the host galaxy. We included the pexrav model to check for the reflection component from the neutral medium which contributes to the Compton hump. The XMM-Newton EPIC-pn data ($\lt 10\,\mathrm{keV}$) is insufficient to constrain the pexrav reflection parameters and hence tied with the Suzaku observation (obs2) with longer exposure. We found a significant improvement in the fit statistics upon the addition of pexrav $(\,{\rm{\Delta }}{\chi }^{2}\sim 72$ for 4 dof) indicating the presence of a neutral reflection component in the hard X-ray band above $10\,\mathrm{keV}$. The phenomenological model provides a satisfactory description to the two Suzaku as well as XMM-Newton spectra. The best-fit parameter values obtained using the phenomenological models are marginally consistent between observations and quoted in Table 3. We also quoted the improvement in statistics ($\,{\rm{\Delta }}{\chi }^{2}/\mathrm{dof}$) for each spectrum upon the addition of different model components to determine the statistical significance of the model (see Table 3). The line energy of the Fe emission is consistent with $6.4\,\mathrm{keV}$ for all three observations. Our results indicate that a prominent soft excess and a broad Fe K_α emission line is present in all the observations that are usually considered as signs of relativistic reflection from the disk.

Table 3.  The Best-fit Parameters of the Baseline Phenomenological Models for the XMM-Newton and Suzaku Observations of ESO 511–G030

Models	Parameter	obs1	obs2	obs3
Gal. abs.	${N}_{{\rm{H}}}(\times {10}^{20}\,\,{\mathrm{cm}}^{-2})$	4.34 (f)	4.34 (f)	4.34 (f)
power law	Γ	${1.77}_{-0.02}^{+0.03}$	${1.80}_{-0.01}^{+0.01}$	${1.96}_{-0.01}^{+0.01}$
	norm (10−3)	${5.13}_{-0.01}^{+0.01}$	${5.21}_{-0.01}^{+0.01}$	${9.1}_{-0.46}^{+0.41}$
diskbb (1)	Tin(keV)	${0.36}_{-0.02}^{+0.01}$	${0.23}_{-0.01}^{+0.01}$	0.23 (t)
	norm (103)	${16}_{-3}^{+2}$	${2}_{-1}^{+1}$	2 (t)
diskbb (2)	Tin (keV)	${0.11}_{-0.01}^{+0.01}$	${0.07}_{-0.01}^{+0.01}$	${0.08}_{-0.01}^{+0.01}$
	norm (103)	${4.9}_{-0.4}^{+0.4}$	${97.5}_{-3.4}^{+1.4}$	${87.8}_{-0.6}^{+2.6}$
diskline	E( keV)	${6.31}_{-0.03}^{+0.03}$	${6.28}_{-0.02}^{+0.01}$	${6.28}_{-0.03}^{+0.03}$
	β	$-{0.8}_{-0.7}^{+9.1}$	${9.9}_{-10.3}^{+0.1}$	9.9 (t)
	${r}_{\mathrm{in}}({r}_{{\rm{g}}})$	6 (f)	6 (f)	6 (f)
	Incl in(°)	${10}_{-7}^{+4}$	10 (t)	10 (t)
	norm (10−5)	${1.75}_{-0.45}^{+0.48}$	${1.54}_{-0.16}^{+0.15}$	${1.55}_{-0.33}^{+0.33}$
	a $\,{\rm{\Delta }}{\chi }^{2}/\mathrm{dof}$	178/3	373/3	96/3
Pexrav B	R	−0.56 (t)	$-{0.56}_{-0.12}^{+0.21}$	$-{0.74}_{-0.15}^{+0.13}$
	a $\,{\rm{\Delta }}{\chi }^{2}/\mathrm{dof}$	12/1	34/2	26/1
$\mathrm{reduced}{\chi }^{2}/\mathrm{dof}$		1.28/168	1.13/1699	1.11/1108

Note.

^aThe $\,{\rm{\Delta }}{\chi }^{2}$ improvement in statistics upon addition of the corresponding discrete component. (f) indicates a frozen parameter and (t) indicates parameters are tied between observations.

Download table as:  ASCIITypeset image

However, the origin of soft excess is still debated and apart from the relativistic reflection from disk, intrinsic disk Comptonization has also been found to be a viable physical model in other Seyfert 1 galaxies. In our work, we used two sets of physical models to describe both the soft excess and the Fe emission line in the Suzaku and XMM-Newton X-ray spectra of ESO 511–G030. In XSPEC, our first set of physical models reads as constant × tbabs × (relxill+MyTorus). Here, the relxill model, version 1.3.7, (García et al. 2014) describes the soft X-ray excess, the power-law continuum, the broad Fe Kα emission line and the reflection of primary hard X-ray photons off an ionized accretion disk; on the other hand, the Compton reflection of hard X-ray photons off cold, neutral material is modeled with MYTorus. The MYTorus model comprised three components: first, the torus-absorbed primary power law; second, the scattered emission (MYTorusS) due to the reflection of primary hard X-ray photons from the torus; and third, the iron FeK_α and Kβ lines (MYTorusL), which are assumed to arise due to the reflection by the torus. ESO 511–G030 is a type-1 AGN, hence we do not expect any obscuration due to torus along the line of sight. We have used only MYTorusS and the MYTorusL component in our spectral fitting and tied their column densities together. The column density of MYTorus gives a value in an equatorial direction of a dusty torus, so a Compton-thick value is expected even in type 1 sources (Yaqoob et al. 2016; Laha et al. 2019; Ghosh & Laha 2020). When made free, the MYTorus column density pegged at ${10}^{25}{\mathrm{cm}}^{-2}$ and we got a lower limit of $2.9\times {10}^{24}{\mathrm{cm}}^{-2}$. Hence, we froze the MYTorus column density to a value of ${10}^{25}{\mathrm{cm}}^{-2}$, assuming that a Compton-thick torus is responsible for this emission. We also fixed the inclination angle between the torus polar axis and the observer's line of sight to 45 degrees for both sets of physical models. The photon index of the MYTorus model was tied with that of the photon index of the primary continuum of the relxill model. In the relxill model, the primary hard X-ray emission from corona illuminates the accretion disk and produces fluorescence emission lines. These fluorescence lines get blurred and distorted due to extreme gravity near the central supermassive black hole and produce the soft excess. The transition between Newtonian and relativistic geometry is marked by a breaking radius r_br. In our work, the emissivity index of reflection from the disk outside r_br (q2) is fixed at 3, as in Newtonian geometry, for a point source, the emissivity at a large radius out from the source has a form r⁻³. The emissivity index inside r_br falls under the relativistic high-gravity regime and calculations (Dabrowski & Lasenby 2001; Miniutti et al. 2003; Wilkins & Fabian 2011) suggest a very steeply falling profile in the inner regions of the disk. Hence, we made the emissivity index of the inner part (q1) of the accretion disk inside r_br, free to vary in between $3\lt q1\lt 10$.

During our simultaneous analysis of joint XMM-Newton and Suzaku data sets, the parameters that are unlikely to change within human timescale, e.g., the black hole spin and the inclination angle, were tied between all observations. Apart from the power-law photon index, normalization, and the reflection fraction (R), all other parameters were tied between the two Suzaku observations. The MYTorus model requires a hard X-ray spectrum to constrain its parameters. Due to the absence of $\gt 10\,\mathrm{keV}$ spectrum in XMM-Newton observation, the MYTorus parameters were tied with best-fit parameter values of obs2. This set of physical models produced a satisfactory fit with fit statistics ${\chi }^{2}/\mathrm{dof}=3431/2981$, however, we observed some excess in the residual at ∼0.5 and $0.9\,\mathrm{keV}$. Addition of two narrow Gaussian line profiles to the model yielded an improved fit (${\chi }^{2}/\mathrm{dof}=3345/2976$). We estimated the high-energy cutoff of the primary power-law component using the Suzaku observation with the longest exposure (obs1). We found a lower limit to the high-energy cutoff with ${E}_{{\rm{c}}}\gt 297\,\mathrm{keV}$. We tied this value for all other observations. A rapidly spinning black hole (>0.98) is required by the model to fit the X-ray spectra. We carried out a test to find out if the black hole is indeed maximally spinning. First, we froze the spin parameter to zero and kept the inner radius (${r}_{\mathrm{in}}$) of the relxill model fixed to the inner circular stable orbit for a non-rotating black hole (${r}_{\mathrm{in}}=6{r}_{{\rm{g}}}$) and fitted the data sets. Next, we froze spin to 0.998 and fixed the ${r}_{\mathrm{in}}$ value to that of a maximally spinning black hole (${r}_{\mathrm{in}}=1.24{r}_{{\rm{g}}}$). We found that the maximally spinning scenario provides a significantly better fit statistics ($\,{\rm{\Delta }}{\chi }^{2}=501$). Following this result, we fixed the inner radius of the disk at ${r}_{\mathrm{in}}=1.24{r}_{{\rm{g}}}$ throughout the rest of the fit and allowed the spin parameter to vary freely. We found the iron abundance of the material in the accretion disk to be significantly lower than the solar value (${A}_{\mathrm{Fe}}\lt 0.7$) for the two Suzaku observations. The best-fit reflection fraction (R) ranges between R = 0.5–5.3 for the three observations. The best-fitting model parameters along with the best-fitting statistics for each observation are quoted in Table 4.

Table 4.  The Best-fitting Parameters When We Modeled the Full X-Ray Energy Band of XMM-Newton and Suzaku Observations of ESO 511–G030 with the Ionized Reflection Model, relxill (Model 1), and the Thermal Comptonization Model, optxagnf (Model 2)

Component	Parameter	obs1		obs2		obs3
		Model 1	Model 2	Model 1	Model 2	Model 1	Model 2
Gal. abs.	${N}_{{\rm{H}}}({10}^{20}{{cm}}^{-2})$	4.34 (f)	4.34 (f)	4.34(f)	4.34(f)	4.34 (f)	4.34 (f)
relxill	${{\rm{A}}}_{\mathrm{Fe}}$	${0.7}_{-0.1}^{+0.2}$	⋯	<0.6	⋯	0.5(t)	⋯
	$\mathrm{log}\xi (\mathrm{erg}\,\mathrm{cm}\,{{\rm{s}}}^{-1})$	${2.29}_{-0.06}^{+0.02}$	⋯	${3.20}_{-0.01}^{+0.01}$	⋯	3.20 (t)	⋯
	Γ	${2.05}_{-0.01}^{+0.01}$	⋯	${1.74}_{-0.01}^{+0.01}$	⋯	${1.88}_{-0.01}^{+0.01}$	⋯
	${E}_{\mathrm{cut}}(\,\mathrm{keV})$	>297	⋯	388(t)	⋯	388(t)	⋯
	${n}_{\mathrm{rel}}{({10}^{-5})}^{a}$	${8.72}_{-0.15}^{+0.08}$	⋯	${11.24}_{-0.38}^{+0.16}$	⋯	${2.69}_{-0.12}^{+2.31}$	⋯
	q1	${7.3}_{-0.3}^{+0.3}$	⋯	${4.6}_{-1.7}^{+0.3}$	⋯	${8.2}_{-0.9}^{+0.2}$	⋯
	a	>0.98	⋯	0.99 (t)	⋯	0.99 (t)	⋯
	$R({reflfrac})$	${1.2}_{-0.1}^{+0.1}$	⋯	${0.5}_{-0.3}^{+0.1}$	⋯	${5.3}_{-0.3}^{+0.3}$	⋯
	${R}_{\mathrm{in}}({r}_{{\rm{g}}})$	1.24(f)	⋯	1.24(f)	⋯	1.24(f)	⋯
	${R}_{\mathrm{br}}({r}_{{\rm{g}}})$	${4.9}_{-0.3}^{+0.4}$	⋯	${4.3}_{-0.1}^{+0.2}$	⋯	${3.6}_{-0.3}^{+0.2}$	⋯
	${R}_{\mathrm{out}}({r}_{{\rm{g}}})$	400 (f)	⋯	400(f)	⋯	400(f)	⋯
	$i({degree})$	${27}_{-1}^{+3}$	⋯	27(t)	⋯	27 (t)	⋯
MYTorusL	$i({degree})$	45 (t)	45 (t)	45(f)	45(f)	45 (t)	45 (t)
	norm (10−3)	7.37(t)	8.18(t)	${7.37}_{-0.37}^{+0.40}$	${8.18}_{-0.04}^{+0.06}$	7.37 (t)	8.18(t)
MYTorusS	NH(${10}^{24}{\mathrm{cm}}^{-2}$)	10.0(t)	10.0 (t)	10(f)	10(f)	10.0(t)	10.0 (t)
	norm (10−3)	3.20(t)	6.05(t)	${3.20}_{-0.71}^{+1.18}$	${6.05}_{-0.39}^{+0.58}$	3.20(t)	6.05(t)
optxagnf	${M}_{{BH}}^{b}$	⋯	4.57(f)	⋯	4.57(f)	⋯	4.57(f)
	$d\,(\mathrm{Mpc})$	⋯	95 (f)	⋯	95(f)	⋯	95(f)
	$(\tfrac{L}{{L}_{E}})$	⋯	${0.004}_{-0.001}^{+0.001}$	⋯	${0.008}_{-0.001}^{+0.001}$	⋯	${0.008}_{-0.001}^{+0.001}$
	${{kT}}_{e}(\,\mathrm{keV})$	⋯	${0.39}_{-0.02}^{+0.12}$	⋯	${0.10}_{-0.04}^{+0.15}$	⋯	${0.54}_{-0.10}^{+0.02}$
	τ	⋯	${12.9}_{-1.9}^{+1.7}$	⋯	>11.8	⋯	${7.9}_{-2.5}^{+3.0}$(t)
	${r}_{\mathrm{cor}}({r}_{{\rm{g}}})$	⋯	${11.8}_{-3.3}^{+4.7}$	⋯	${5.2}_{-0.9}^{+2.7}$	⋯	5.2(t)
	a	⋯	>0.78	⋯	0.99(t)	⋯	0.99(t)
	fpl	⋯	${0.56}_{-0.05}^{+0.03}$	⋯	${0.41}_{-0.04}^{+0.08}$	⋯	${0.59}_{-0.07}^{+0.02}$
	Γ	⋯	${1.74}_{-0.02}^{+0.01}$	⋯	${1.79}_{-0.02}^{+0.02}$	⋯	${1.86}_{-0.03}^{+0.01}$
	$\mathrm{reduced}{\chi }^{2}/\mathrm{dof}$	1.37/170	1.18/171	1.13/1703	1.14/1706	1.05/1110	1.09/1112

Note. Model 1 = tbabs × (relxill +MyTorus); Model 2 = tbabs × (optxagnf +MyTorus). The MYTorus parameters for obs1 and obs3 are tied with obs2, which is the longest Suzaku observation (see Table 1). (f) indicates a frozen parameter. (*) indicates parameters are not constrained. (a) n_rel represent normalization for the model relxill (b) in units of ${10}^{8}{M}_{\odot }$.

Download table as:  ASCIITypeset image

We have also tested other flavors of relxill, such as relxillD and relxilllP. The model relxillD allows a higher density for the accretion disk (between $\mathrm{log}N/{\mathrm{cm}}^{-3}=15.3$ to $\mathrm{log}N/{\mathrm{cm}}^{-3}\,=18$) and the model relxilllP assumes a lamp post geometry and determines the variation in the position of the hard X-ray emitter. Both these models produced a relatively poor statistics (${\chi }^{2}/\mathrm{dof}=3434/2973$ for relxillD and ${\chi }^{2}/\mathrm{dof}=3469/2976$ for relxilllP, respectively) compared to relxill, for all the observations and no significant variation either in the disk density or in the position of the hard X-ray emitter was found.

For the second set of physical models, we use intrinsic thermal Comptonization from a warm corona (optxagnf). In the optxagnf model, thermal Comptonization of disk photons by a warm (${\rm{T}}\sim 0.5\mbox{--}1\,\mathrm{keV}$) and optically thick ($\tau \sim 10\mbox{--}20$) corona (Magdziarz et al. 1998; Done et al. 2012), that lies in the inner part ($10\mbox{--}20{{\rm{r}}}_{{\rm{g}}}$) of the accretion disk, produces the soft excess. The gravitational energy released in the accretion process fuels the disk emission in UV, the soft X-ray excess and the power-law emission. The model normalization flux is determined by the source Eddington rate, the black hole mass, the black hole spin and the luminosity distance. Hence in our analysis, we froze this value at unity. The hard excess emission, in this set of models, arises exclusively due to neutral Compton reflection from the torus. In XSPEC notation, the model reads as (constant × tbabs × (optxagnf + MYTorus)). We used a black hole mass of $\sim 4.57\times {10}^{8}{M}_{\odot }$, adopted from Ponti et al. (2012), who derived it using X-ray variability of the source. In our simultaneous fit of Suzaku and XMM-Newton, apart from the black hole spin, all other optxagnf model parameters were made free to vary. On the other hand, for the two Suzaku observations, except for the Eddington rate, photon index and the parameter f_pl, that determines the fraction of powerlaw emitted as soft excess, all other parameters were tied between observations. This set of physical models produced a poor fit statistics (${\chi }^{2}/\mathrm{dof}=3411/2992$) compared to our first set of models, however, we observed similar excess emission in the residual at $\sim 0.5\,\mathrm{keV}$ and $\sim 0.9\,\mathrm{keV}$. Addition of two Gaussian line profile to the set of models provided similar fit statistics (${\chi }^{2}/\mathrm{dof}=3351/2987$) compared to the reflection model relxill. All three observations show sub-Eddington accretion rate (∼0.004–0.008) and infer a maximally spinning black hole (>0.98). Table 5 includes the best-fitting parameters obtained using this model. Figure 2 shows both the best-fitting physical models used here and their respective residuals. Next, we included the OM data from the XMM-Newton and tried to get a better constrain on the optxagnf model parameters.

Table 5.  The Best-fitting Parameters When We Modeled the Optical-UV to Hard X-Ray Energy Band of XMM-Newton and Suzaku Observations of ESO 511–G030 with the Ionized Reflection Model, relxill (Model 1), and the Thermal Comptonization Model, optxagnf (Model 2)

Component	Parameter	obs1		obs2		obs3
		Model 1	Model 2	Model 1	Model 2	Model 1	Model 2
Gal. abs.	${N}_{{\rm{H}}}({10}^{20}{{cm}}^{-2})$	4.34 (f)	4.34 (f)	4.34(f)	4.34(f)	4.34 (f)	4.34 (f)
diskbb	${T}_{{in}}(\,\mathrm{keV})$	${0.004}_{-0.001}^{+0.001}$	⋯	0.004(t)	⋯	0.004(t)
	norm (1010)	${2.54}_{-1.42}^{+2.18}$	⋯	2.54(t)	⋯	2.54(t)
relxill	${A}_{\mathrm{Fe}}$	0.7 (f)	⋯	0.5 (f)	⋯	0.5 (f)	⋯
	$\mathrm{log}\xi (\,\mathrm{erg}\,\mathrm{cm}\,{{\rm{s}}}^{-1})$	2.29 (f)	⋯	3.20(f)	⋯	3.20 (f)	⋯
	Γ	2.05 (f)	⋯	1.74(f)	⋯	1.88 (f)	⋯
	${E}_{\mathrm{cut}}(\,\mathrm{keV})$	388 (f)	⋯	388(t)	⋯	397(t)	⋯
	${n}_{\mathrm{rel}}{({10}^{-5})}^{a}$	8.27(f)	⋯	11.24 (f)	⋯	2.69(f)	⋯
	q1	7.3 (f)	⋯	4.6 (f)	⋯	8.2 (f)	⋯
	a	0.99(f)	⋯	0.99 (t)	⋯	0.99 (t)	⋯
	$R({reflfrac})$	1.2(f)	⋯	0.5 (f)	⋯	5.3 (f)	⋯
	${R}_{\mathrm{in}}({r}_{{\rm{g}}})$	1.24(f)	⋯	1.24(f)	⋯	1.24(f)	⋯
	${R}_{\mathrm{br}}({r}_{{\rm{g}}})$	4.9 (f)	⋯	4.3 (f)	⋯	3.6 (f)	⋯
	${R}_{\mathrm{out}}({r}_{{\rm{g}}})$	400 (f)	⋯	400(f)	⋯	400(f)	⋯
	$i({degree})$	27 (f)	⋯	27(t)	⋯	27 (t)	⋯
MYTorusL	$i({degree})$	45 (t)	45 (t)	45 (f)	45(f)	45 (t)	45 (t)
	norm (10−3)	7.37(t)	8.18(t)	7.37(f)	8.18(f)	7.37 (t)	8.18(t)
MYTorusS	NH(${10}^{24}{{cm}}^{-2}$)	10.0 (t)	10.0 (t)	10.0(f)	10.0(f)	10.0(t)	10.0 (t)
	norm (10−3)	3.20(t)	6.05(t)	3.20 (f)	6.05(f)	3.20(t)	6.05(t)
optxagnf	${M}_{{BH}}^{b}$	⋯	4.57(f)	⋯	4.57(f)	⋯	4.57(f)
	$d\,(\mathrm{Mpc})$	⋯	95 (f)	⋯	95(f)	⋯	95(f)
	$(\tfrac{L}{{L}_{E}})$	⋯	${0.004}_{-0.001}^{+0.001}$	⋯	${0.007}_{-0.001}^{+0.001}$	⋯	${0.008}_{-0.001}^{+0.001}$
	${{kT}}_{e}(\,\mathrm{keV})$	⋯	${0.40}_{-0.02}^{+0.02}$	⋯	${0.09}_{-0.03}^{+0.10}$	⋯	${0.53}_{-0.10}^{+0.03}$
	τ	⋯	${12.7}_{-0.4}^{+0.5}$	⋯	>11.9	⋯	${7.9}_{-1.7}^{+2.0}$
	${r}_{\mathrm{cor}}({r}_{{\rm{g}}})$	⋯	${11.8}_{-1.8}^{+2.5}$	⋯	${5.9}_{-0.5}^{+1.9}$	⋯	5.9(t)
	a	⋯	>0.98	⋯	0.99(t)	⋯	0.99(t)
	fpl	⋯	${0.57}_{-0.02}^{+0.02}$	⋯	${0.41}_{-0.08}^{+0.06}$	⋯	${0.60}_{-0.06}^{+0.06}$
	Γ	⋯	${1.74}_{-0.02}^{+0.02}$	⋯	${1.79}_{-0.01}^{+0.01}$	⋯	${1.86}_{-0.01}^{+0.01}$
	${\rm{reduced}}{\chi }^{2}/{\rm{dof}}$	1.38/170	1.38/171	1.13/1704	1.14/1706	1.05/1108	1.09/1112

Note. Model 1 = tbabs × (diskbb + relxill + MyTorus); Model 2 = tbabs × (optxagnf +MyTorus). The MYTorus parameters for obs1 and obs3 are tied with obs2 which is the longest Suzaku observation (see Table 1). (f) indicates a frozen parameter. (*) indicates parameters are not constrained. (a) n_rel represent normalization for the model relxill (b) in units of ${10}^{8}{M}_{\odot }$.

Download table as:  ASCIITypeset image

The thermal Comptonization model uses disk photons to produce the power-law spectra and the soft excess. Therefore, the UV bump is required to constrain the optxagnf parameters. In this section, we re-fit the X-ray data along with the simultaneously obtained UV spectra to constrain the model parameters.

We used the optical/UV fluxes in four bands (B, W1, W2, and M2) measured with the XMM-Newton to constrain the thermal emission from the disk. The galactic extinction correction was done following (Fitzpatrick 1999) reddening law with R_v = 3.1 and is taken into account by the REDDEN model. The parameter value used here is E_(B-V) = 0.056 (Schlafly & Finkbeiner 2011). We have added 5% systematic error to the OM fluxes to account for the host galaxy contamination, the nuclear emission lines and the intrinsic reddening, which produce a considerable amount of systematic uncertainty in the measured optical-UV continuum flux. The two sets of physical models, relxill and optxagnf, were used to jointly fit the OM and EPIC-pn data of XMM-Newton along with the two Suzaku observations. The best-fitting parameter values are quoted in Table 5. The model parameters of relxill and MyTorus were frozen in the joint fit as they were determined by the X-ray energy band only and we added a diskbb model to fit the OM data. The model resulted in a fit-statistic of ${\chi }^{2}/\mathrm{dof}=3349/3001$. In the case of optxagnf model, we froze the MyTorus model parameters and allowed optxagnf model parameters to vary freely. We find that this second set of physical models yield a marginally poor fit statistics (${\chi }^{2}/\mathrm{dof}=3369/2989$) (see Table 5) compared to the ionized reflection model nevertheless both provide a satisfactory description to the broadband optical-UV to hard X-ray data.

Our simultaneous analysis of UV and X-ray data enabled us to get a better constraint on the warm corona properties, e.g., the electron temperature, the radius, and the opacity (see Table 5). Our results suggest the presence of a warm (${\mathrm{kT}}_{{\rm{e}}}=0.1\mbox{--}0.5\,\mathrm{keV}$), optically thick (τ = 8–13) and compact (r_corona = 5.9–11.8 r_g) corona and a maximally rotating ($a\gt 0.98$) black hole. The best-fitting models obtained from the joint fitting of XMM-Newton and Suzaku are shown in Figure 3. A detailed discussion of these results follows in the next section.

Previous studies of the X-ray spectra of ESO 511–G030 have mostly revealed the presence of a relativistically broad Fe emission line. Tombesi et al. (2010); Laha et al. (2014b) studied the 2007 XMM-Newton data and found a broad Fe emission line at $6.4\,\mathrm{keV}$ with an equivalent width of $\sigma =100\pm 15\,\mathrm{eV}$. On the other hand, using the same observation, Winter et al. (2012) studied the X-ray broadband properties and detected the presence of a narrow Fe emission line ($\sigma ={39}_{-8}^{+8}$) in the source spectrum. Fukazawa et al. (2011) studied the source using two Suzaku observations and found the presence of narrow Fe K lines. In our work, we have investigated the above $3\,\mathrm{keV}$ energy band with a set of models to unambiguously determine the nature of the Fe emission line. We found the presence of a broad Fe emission line at $\sim 6.4\,\mathrm{keV}$ in all the spectra of ESO 511–G030. The best-fit Fe emission line σ ($0.08\mbox{--}0.14\,\mathrm{keV}$) and equivalent width ($62\mbox{--}94\,\mathrm{eV}$) are significantly broader than the equivalent width of typical narrow Fe emission line found in nearby Seyfert 1 galaxies. The broad Gaussian line profile and the diskline model provides a better description of the Fe line emission than other sets of models and indicates the presence of a relativistically-broadened Fe emission line from an accretion disk around a rotating black hole. Our broadband spectral modeling of ESO 511-G030 with both sets of physical models favors a rapidly spinning black hole and supports this idea. In all the observations, the Fe line centroid energy was consistent with $6.4\,\mathrm{keV}$, which indicates that the iron is neutral or in low ionization state. The best-fit ionization parameter value of the XMM-Newton observation ($\mathrm{log}\xi ={2.3}_{-0.1}^{+0.1}$) supports this idea. However, for the two Suzaku observations, we got a highly ionized accretion disk ($\mathrm{log}\xi ={3.2}_{-0.1}^{+0.1}$) compared to obs1. From Figure 1 we note that the Compton hump above $10\,\mathrm{keV}$ is relatively weak but the addition of a pexrav model to the phenomenological set of models did improve the fit statistics ($\,{\rm{\Delta }}{\chi }^{2}\sim 72$ for 4 dof) significantly. From Figure 2 we find that both neutral and ionized reflection components provide a good description of the Compton hump. With the current data quality in Suzaku, we are unable to comprehensively detect or separate out the contributions of the ionized and neutral reflection components and further deep observations of Seyfert 1s are required.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Our analysis of the full X-ray band of ESO 511–G030 revealed the presence of a prominent soft excess below $2\,\mathrm{keV}$ for all three observations. We investigated the broadband multi-epoch data in detail to identify the disk and black hole properties responsible for this excess emission. Recent studies suggest that both relativistic reflection from an ionized accretion disk and the intrinsic thermal Comptonization of disk photons can successfully describe the soft excess in Seyfert 1 s (Ghosh et al. 2016; Ghosh & Laha 2020; García et al. 2019; Waddell et al. 2019; Ehler et al. 2018). These two models assume two very different geometries and physical properties of the accretion disk and the corona (e.g., the position, temperature, and the opacity). However, it is not easy to distinguish them on statistical grounds alone and often extreme values of certain model parameters are used to favor one model over the other. In some nearby Seyfert 1 s, e.g., HE 1143-1810 (Ursini et al. 2020) and Zw 229.015 (Tripathi et al. 2019), the ionized disk reflection model was ruled out due to the relatively extreme values of the iron abundance, the inclination angle and reflection fraction, compared to other Seyfert 1 s. On the other hand, in Mrk 478 (Waddell et al. 2019), the flux variability between data sets was better described by ionized disk reflection, compared to the thermal Comptonization model.

In our work, we have studied the reflection features observed in the ESO 511–G030 spectra and got comparable fit statistics for both of these physical models. The soft excess is equally well described by both these models. The best-fit parameter values obtained for the optxagnf model were in a range detected in typical Seyfert 1 galaxies (See Table 4) although, we were unable to constrain the optical depth in obs2 (see Table 5). Similarly, for the reflection model, some model parameters require extreme values to model the soft excess and the X-ray energy band. The best-fit photon index Γ of the model relxill ranges between 1.7 and 2.0 for all three observations. The best-fit ionization parameter value of the model ($\mathrm{log}\xi \sim 2\mbox{--}3\,\mathrm{erg}\,\mathrm{cm}\,{{\rm{s}}}^{-1}$) suggests a transition between moderate to highly ionized disk between obs1 and obs2, respectively. In the reflection model relxill, the black hole spin and the inner radius are degenerate and hence fixing the inner radius to the ISCO provides a better constraint on the spin. Hence we fixed the inner radius to $1.24{r}_{{\rm{g}}}$ and obtained a maximally spinning black hole ($a\gt 0.98$). We were able to constrain the inclination angle parameter and found a best-fit value of ${27}_{-1}^{+3}$. The iron abundance of the reflecting medium (${A}_{\mathrm{Fe}}$) or the disk is marginally consistent with the solar abundance value for the XMM-Newton observation, however, when made free we got an upper limit of <0.6 for the two Suzaku observations. We also found a large variation in the best-fit value of reflection fraction (R) between the three observations, ranging between 0.5 and 5.3, although consistent with other Seyfert 1 s. This value R, determines the ratio of photons entering the disk to the ones escaping to infinity and indicates that most of the hard X-ray photons are entering the disk and this results in a highly ionized disk. These results along with the high emissivity index ranging between $4.6\mbox{--}8.2$ imply that a major part of the soft excess has originated from a region very close to the central supermassive black hole due to reflection of hard X-ray photons from a highly ionized untruncated accretion disk. Although very high values of black hole spin and the reflection fraction or lower iron abundances in the disk are occasionally reported for AGNs, a corona placed so close to the black hole implies a very extreme configuration of the accretion disk and the corona.

On the other hand, the intrinsic thermal Comptonization model optxagnf combined with the neutral reflection from the torus, modeled by MyTorus, provides a similar fit statistics compared to the ionized reflection. Most of the best-fit parameter values obtained from the fitting of the full X-ray band are well constrained and are typical of other Seyfert 1 galaxies (Ghosh & Laha 2020; Porquet et al. 2018; Ursini et al. 2020). In this model, the soft excess originates due to Comptonization of thermal disk photons by a warm and optically thick corona that covers roughly $5\mbox{--}12{r}_{{\rm{g}}}$ of the inner accretion disk (Petrucci et al. 2013). However, we note that the accretion rate in optxagnf is also determined by the outer accretion flow beyond ${r}_{\mathrm{corona}}$ that contributes primarily in the optical-UV band. This explains our inability to constrain some of the model parameters e.g., the ${r}_{\mathrm{corona}}$ and the black hole spin. Hence we discuss the best-fit parameter values obtained from the simultaneous analysis of the optical-UV to hard X-ray broadband data with optxagnf as they provide us with a more clear view of the accretion disk and the black hole properties.

In the optxagnf model, the gravitational energy released in accretion, powers three distinct emission components—the UV bump, the soft excess, and the hard X-ray power law. As expected, optxagnf combined with neutral reflection from torus yielded better fit statistics (see Table 5) compared to the ionized reflection model. From the joint fits of all three observations, using the relxill+diskbb model, we find that the disk blackbody has a much colder temperature of $\mathrm{kT}={3.8}_{-1.2}^{+0.7}\,\mathrm{eV}$. This value is comparable with the inner radius temperature, $\sim 2\,\mathrm{eV}$, of the accretion disk that is accreting at a rate of ${\lambda }_{\mathrm{Edd}}=0.004$, obtained for obs1, using optxagnf model. This best-fitting accretion rate is close to our estimated value (${\lambda }_{\mathrm{Edd}}=0.002$) obtained using the bolometric luminosity of ESO 511–G030. The soft excess in XMM-Newton observation is well described as a warm (${kT}={0.4}_{-0.02}^{+0.02}\,\mathrm{keV}$) and optically thick ($\tau ={12.7}_{-0.4}^{+0.5}$) corona. These values are marginally consistent with that of obs3. In the case of obs2, we could not constrain the optical depth (τ). The measured warm-corona radius (${r}_{\mathrm{corona}}$) remains nearly consistent between observations with values of ${11.8}_{-1.8}^{+2.5}{{\rm{r}}}_{{\rm{g}}}$ and ${5.9}_{-0.5}^{+1.9}{r}_{{\rm{g}}}$ for XMM-Newton and Suzaku observations, respectively, and indicates a compact corona. We note that the optxagnf model, similar to the relxill, favors a maximally rotating black hole spin to describe the broadband spectra. We could not constrain the spin parameter in optxagnf and got a lower limit of >0.98. We argue that a highly rotating black hole is required to model the broadband data in both ionized reflection and thermal Comptonization scenarios, and the presence of a broad Fe emission line in the spectra further supports the idea of a highly spinning black hole at the center of the AGN. The f_pl parameter in the optxagnf model determines the fraction of power below the coronal radius emitted in the hard Comptonization component. This parameter value is marginally consistent between the three observations with moderate values of ∼0.6 and ∼0.4 for XMM-Newton and Suzaku observations, respectively. This implies, in obs1, around 60% of the gravitational energy released below $12{r}_{{\rm{g}}}$, is emitted as the primary power-law emission with a photon index ∼1.74 and the rest would help produce the soft excess. These values are consistent with recent studies found in other local Seyfert 1s (Ghosh & Laha 2020; Porquet et al. 2019). Hence we conclude that for ESO 511–G030, highly ionized, relativistically blurred reflection from an untruncated accretion disk and the intrinsic thermal Comptonization of disk seed photons by a warm and optically thick compact corona, both combined with the neutral reflection from the torus, provide a good description to the multi-epoch optical-UV to hard X-ray spectra with parameter values consistent with other Seyfert 1s.