Online material

	Fig. 1 Chandra image of the central 25″ of NGC 315. Jet spectrum is extracted from an ellipse with a semi-major axis of 11.3″ and a semi-minor axis of 5.6″. Core emission is extracted from a circle centered on the source with a radius comprising 99% of the PSF (~2.7″). The rest of the medium inside a 25″ circle is considered diffuse emission. See Appendix B for more details.
Open with DEXTER

	Fig. 2 Chandra image of the central 25″ of NGC 2787. A point-like source south-east of the central LINER is present with a luminosity comparable to the core luminosity. Another six point-like sources, marked in white, are present in the field. See Appendix B for more details.
Open with DEXTER

	Fig. 3 Chandra image of the central 25″ of NGC 3226. Two sources are present in the field marked source 1 and source 2. See Appendix B for more details.
Open with DEXTER

	Fig. 4 Chandra image of the central 25″ of NGC 3998. Only one source, much fainter than the core, is present in the field. See Appendix B for more details. The horizontal bright line corresponds to the readout streak events.
Open with DEXTER

	Fig. 6 Light curves and hardness ratios of the LINER 1s observed with XMM-Newton, all binned with a 1 ks time bin-size.
Open with DEXTER

	Fig. 7 Light curves and hardness ratios of the LINER 1s observed with Chandra with a long exposure time, all binned with a 1 ks time bin-size.
Open with DEXTER

	Fig. 10 Data and best fit model of the spectra of the LINER 1s in our sample with a relatively long exposure time. Residuals of every fit are given in terms of sigma.
Open with DEXTER

Appendix A: Notes on individual sources

NGC 266. Terashima & Wilson (2003) studied the 2 ks snapshot observation made with Chandra and discussed here. They model the X-ray spectrum with an absorbed power-law and derive a hydrogen column density and a photon index of  <0.82  ×  10²² cm^-2 and 1.4, respectively. They find a 2−10 keV flux of 1.6  ×  10^-13 erg s^-1 cm^-2. The values of the photon index and the 2−10 keV flux are with a good agreement with the value we report here (within the error bars), but no additional absorption is added to our model.

NGC 315. Worrall & Birkinshaw (1994) first suggested the presence of an active galactic nucleus at the center of NGC 315 using ROSAT data. This assumption was later confirmed by Matsumoto et al. (2001) when studying a 37 ks ASCA observation (see also Terashima et al. 2002). The authors fit the hard 2 − 10 keV spectrum with a power-law and find a photon index of  ~2 and a luminosity of 3.1  ×  10⁴¹ erg s^-1. A resolved X-ray jet emission was first reported by Worrall et al. (2003) when studying the Chandra snapshot observation (obs.ID: 855). The fit to the jet emission with a power-law gave a photon index of 2.5  ±  0.7 and a 3.5  ×  10⁴⁰ erg s^-1 luminosity. The authors fit the unresolved core emission with a moderately absorbed power-law with an intrinsic hydrogen column density of  ~5  ×  10²¹ cm^-2 and a photon index of 1.4  ±  0.4. Worrall et al. (2007) found similar results when studying the longer Chandra observation reporting a harder core spectrum than the jet with photon indicies of  ~1.6 and  ~2.2 respectively. Moreover, Croston et al. (2008), after solar flare cleaning, studied the only XMM-Newton observation and fit the 60″ core spectrum with a combination of a mekal and a power-law. Our analysis of the two Chandra observations and the XMM-Newton one gives similar results to all of the above studies with a simultaneous fit to the different extracted spectra where a combination of a mekal (kT ≈ 0.5 keV) and a mildly absorbed (N_h ≈ 10²² cm^-2) power-law (Γ between 1.5 and 2) were used. The increase in the power-law slope from 1.5  ±  0.1 during the Chandra observation to  during the XMM-Newton one is accompanied by a decrease in the 2 − 10 keV flux from 9.8  ×  10^-13 to 4.6  ×  10^-13 erg s^-1. This behavior (see Sect. 5.4) is typical of X-ray emission originating in a RIAF structure which is believed to be the accretion mechanism responsible for the bulk of energy from radio to X-rays in NGC 315 (Wu et al. 2007).

NGC 2681. One of the two observations performed with Chandra has been already reported in Satyapal et al. (2005). The authors fit the 0.5−8 keV spectrum with a combination of a thermal component with kT ≈ 0.7 keV and a power-law with a photon index Γ ≈ 1.6. No intrinsic absorption was required. The same observation was treated in González-Martín et al. (2009) and same results were derived after fitting the spectrum with a mekal and a power-law. González-Martín et al. (2009) derived a 2−10 keV flux of  ~2  ×  10^-13 erg s^-1 cm^-2. We fit the spectrum of the two Chandra observations of NGC 2681 simultaneously with a combination of a thermal component and an absorbed power-law. We found similar results to that derived in the previous works (Γ ≈ 1.5 and kT ≈ 0.6 keV) with a 2−10 keV flux of 3  ×  10^-13 erg s^-1 cm^-2.

NGC 2787. Ho et al. (2001), after studying a Chandra snapshot, gave this source a class III X-ray morphology, showing hard X-ray nucleus embedded in diffuse emission. Terashima & Wilson (2003) derived a 2 − 10 keV flux of about 3  ×  10^-14 erg s^-1 cm^-2, after assuming a photon index of 2 and a Galactic absorption. González-Martín et al. (2009) analyzed both Chandra and XMM-Newton long observations and fit the Chandra spectrum with a power-law with a rather soft Γ of 2.3 and fit the XMM-Newton spectrum with a combination of two power-laws and a thermal component with a power-law absorption of  ~10²² cm^-2. We found a good fit for both Chandra and XMM-Newton spectra simultaneously with a single absorbed power-law and found little absorption of  ~2  ×  10²¹ cm^-2 and a soft power-law photon index, Γ = 2.4, and a 2 − 10 keV flux of 4  ×  10^-14 erg s^-1 cm^-2.

NGC 3226. George et al. (2001) fit the 0.5 − 10 keV spectrum extracted from the long Chandra observation with an absorbed (N_H ≈ 5  ×  10²¹ cm^-2) power-law (Γ ≈ 1.9). Terashima & Wilson (2003) fit the spectrum of the 2.5 ks snapshot Chandra observation with a moderately absorbed power-law with N_H ≈ 10²² cm^-2 and Γ ≈ 2.2. Gondoin et al. (2004) fit the data of the 35 ks XMM-Newton observation with a partial covering absorber to a bremsstrahlung and found that the X-ray emitting region, with a temperature kT ≈ 0.9 keV is 90% covered by an absorber with N_H ≈ 5  ×  10²¹ cm^-2. Binder et al. (2009) studied the  ~100 ks XMM-Newton observation and fit the spectrum with a partially covered power-law with Γ ≈ 1.9, a covering fraction of 90%, and an intrinsic hydrogen column density of 10²¹ cm^-2. We fit the spectra of both XMM-Newton and the long Chandra observations simultaneously with an absorbed power-law and found an intrinsic column density varying between the observations from  ~3  ×  10²¹ cm^-2 to  ~9  ×  10²¹ cm^-2 and a mean photon index Γ  =  1.9.

NGC 3718. Satyapal et al. (2005) studied the snapshot Chandra observation and fit the spectrum with an absorbed (N_H ≈ 10²² cm^-2) power-law (Γ ≈ 1.5) in excellent agreement with our fit results to the same observation. We studied two XMM-Newton observations of NGC 3718, being in the field of view of the observations of the heavily absorbed Seyfert 2 galaxy UGC 6527. We fit the spectra with an absorbed power-law and found an intrinsic hydrogen column density similar to the one derived for the Chandra observation of  ≈ 10²² cm^-2 but a somewhat softer power-law with Γ ≈ 1.8. This softening is accompanied with a 2 − 10 keV flux decrease from 3.3  ×  10^-12 erg s^-1 cm^-2 to 1.6  ×  10^-12 erg s^-1 cm^-2.

NGC 3998. Ptak et al. (2004) studied the 10 ks XMM-Newton observation and fit the spectrum with a slightly absorbed (N_H ≈ 10²⁰ cm^-2) power-law (Γ ≈ 1.9). Same results were found for observations made with BeppoSAX (Pellegrini et al. 2000) and ASCA (Ptak et al. 1999). González-Martín et al. (2009) fit the Chandra spectrum with a combination of two absorbed power-laws and a thermal component. We fit the XMM-Newton and the Chandra spectra simultaneously with a mildly absorbed (N_H ≈ 10²⁰ cm^-2) power-law and found a varying Γ from 1.8 to 2.1 respectively, occurring with a flux decrease from 1.1  ×  10^-11 erg s^-1 cm^-2 to 6.5  ×  10^-12 erg s^-1 cm^-2.

NGC 4143. Terashima & Wilson (2003) fit the Chandra snapshot observation with an absorbed power-law with N_H < 10²¹ cm^-2 and Γ ≈ 1.7. We fit the same snapshot observation with a power-law without a requirement of an intrinsic absorption and found a similar power-law photon index within the error bars, Γ ≈ 1.9. We fit the XMM-Newton observation with an absorbed power-law and found a N_H = 6  ×  10²⁰ cm^-2 and a Γ  ≈  2.2.

NGC 4203. A power-law fit to the ASCA spectrum resulted in a Γ ≈ 1.8 (Iyomoto et al. 1998). Ho et al. (2001) gave NGC 4203 a class I X-ray morphology showing dominant X-ray nucleus. We find that the 40 ks Chandra spectrum is well fitted with a simple power-law affected by Galactic absorption with Γ ≈ 2.3, softer than the result reported for ASCA, most likely due to contamination from X-ray sources in the ASCA extraction region of 1′.

NGC 4278. See Y10.

NGC 4750. Dudik et al. (2005), and according to the only Chandra snapshot observation, gave this source a morphological X-ray type II, exhibiting multiple, hard off-nuclear point sources of comparable brightness to the nuclear source. We fit the spectrum of this same observation with an absorbed power-law with N_H < 3  ×  10²¹ cm^-2 and Γ  =  1.8.

NGC 4772. We fit the spectrum of the only, Chandra snapshot, observation with an absorbed (N_H ≈ 5  ×  10²¹ cm^-2) power-law (Γ  ≈  1.7).

NGC 5005. Terashima et al. (2002) fit the ASCA spectrum with a combination of an absorbed (N_H < 9  ×  10²¹ cm^-2) power-law (Γ ≈ 1) and a thermal component (kT ≈ 0.8). Dudik et al. (2005) gave NGC 5005 a morphological X-ray type III, showing a hard nuclear point source embedded in diffuse emission. González-Martín et al. (2009) fit the XMM-Newton spectrum with a combination of a thermal component with kT ≈ 0.3 keV and an absorbed (N_H ≈ 6  ×  10²¹ cm^-2) power-law (Γ ≈ 1.5). We find a hotter thermal component when fitting the same data set with kT ≈ 0.6 keV and a mildly absorbed power-law with N_H ≈ 10²¹ cm^-2 and Γ  ≈  1.7.

Appendix B: Surrounding sources of the centers of galaxies observed with Chandra

In this Appendix, we report the spectral analysis of resolved and/or unresolved off-nucleus sources detected with the Chandra telescope, but blended within the central LINER in the XMM-Newton extraction region (Sect. 3.2). Figures 1 − 4 show the surrounding sources of these LINER 1s detected with Chandra within a 25″-radius circle. The surrounding medium of NGC 4278 is already reported in Y10.

NGC 315. This source is the only source in our sample that shows a resolved X-ray jet. We extracted from the longer Chandra observation the spectrum of the jet from an ellipse with a semi-major axis of about 11.3″ and semi-minor axis of 5.6″. The base of the ellipse extends down to the 1.1  ×  99% PSF of the central source. We fit the spectrum with a combination of an absorbed power-law and a thermal mekal component and found a good fit with a reduced χ² of 0.9 for 42 d.o.f. We find a hydrogen column density upper limit of 2  ×  10²¹ cm^-2 and a photon index . The thermal component had a temperature of  keV. We find a corrected 0.5 − 10 keV flux for the jet emission of (1.2 ± 0.1)  ×  10^-13 erg s^-1 cm^-2, which corresponds to a 0.5 − 10 keV luminosity of 6  ×  10⁴⁰ erg s^-1, corresponding to 8% of the nuclear flux. The power-law emission contributes to 95% to the total emission of the jet. For the diffuse emission, we extracted the spectrum from an annulus with inner circle delimited by 1.1  ×  99% PSF of the central source and an outer radius of 25″ excluding the jet extraction region. The same model fit to the jet gave a good fit with a reduced χ² of 1.3 for 61 d.o.f. We find an intrinsic absorption to the powerlaw component   ×  10²¹ cm^-2 and a photon index Γ = 1.7  ±  0.8, possibly representing emission from unresolved X-ray binaries. The thermal component has a temperature similar to the one derived for the jet emission with kT = 0.6  ±  0.2 keV. We found a corrected 0.5 − 10 keV flux of   ×  10^-13 erg s^-1 cm^-2 with the power-law contributing only to 30% of the total emission. This corresponds to a luminosity of  ~10⁴¹ erg s^-1 which is  ~14% of the total core luminosity.

NGC 2787. An X-ray source south-east of the nucleus of NGC 2787 at a distance less than 10″ is present. We fit the spectrum of this source with an absorbed power-law. We used the same redshift and Galactic absorption as for the NGC 2787 nucleus, assuming that this X-ray source is located in NGC 2787 and not a background quasar. The fit is acceptable with a reduced χ² of 1.0 for 19 d.o.f. We found a power-law photon index , typical of X-ray binaries in nearby galaxies (Irwin et al. 2003; Fabbiano 2006), and an upper limit on the intrinsic hydrogen column density of 2  ×  10²¹ cm^-2. We derived a 0.5 − 10 keV corrected flux of 7  ±  1  ×  10^-14 erg s^-1 cm^-2, which resulted in a luminosity of, adapting the NGC 2787 distance of 7.48 Mpc, 5  ±  1  ×  10³⁸ erg s^-1. This luminosity is close to the NGC 2787 core luminosity of  ~9  ×  10³⁸ erg s^-1. Such a source could be a luminous low mass X-ray binary (LMXB) similar to some seen in early type galaxies (Fabbiano 2006). The rest of the medium in an annulus of outer radius 25″ around NGC 2787 is formed by six other X-ray sources, much fainter than the closest one to the center. We could not perform spectral analysis on the different sources aside, so we fit the spectrum of the six sources simultaneously with a power-law only affected by Galactic absorption. We found a photon index of  ~2 for the six sources and a total corrected 0.5 − 10 keV flux of   ×  10^-15 erg s^-1 cm^-2, resulting in a 0.5 − 10 keV luminosity of   ×  10³⁷ erg s^-1, corresponding to 5% of the nucleus luminosity, when adapting the NGC 2787 distance of 7.48 Mpc.

NGC 3226. Two X-ray sources are within a  ~12″ distance from the nucleus of NGC 3226 (source 1: CXOU J102334.1+195347, source 2: CXOU J102326.7+195407). Both sources were reported in George et al. (2001). Based on the hardness ratio between the counts in the 0.3 − 2 keV band and the counts in the 2 − 10 keV band, the authors estimated the sources to have a flux of a few 10^-14 erg s^-1 cm^-2, and therefore a luminosity between a few times 10³⁸ erg s^-1 to a few times 10³⁹ erg s^-1. We fit the spectrum of both sources with an absorbed power-law, using the cash statistic due to a low number of counts. We find that source 1 and source 2 have an intrinsic absorption upper limit of 6  ×  10²² cm^-2 and 5  ×  10²¹ cm^-2, respectively. The photon index we found for source 2 is typical, within the error bars, of accreting objects with a . On the other hand, we found a harder spectrum for source 1 with a Γ = 0.2  ±  1.3. The corrected 0.5 − 10 keV flux we derive for source 1 and source 2 are  ~4  ×  10^-14 erg s^-1 cm^-2 and  ~2  ×  10^-14 erg s^-1 cm^-2, respectively. This implies, assuming the distance of NGC 3226 to both sources, a corrected 0.5 − 10 keV luminosity of 3  ×  10³⁹ erg s^-1 and 2  ×  10³⁹ erg s^-1 for source 1 and source 2 respectively. Both luminosities are well beyond the luminosity of a typical neutron star LMXB of  ~3  ×  10³⁸ erg s^-1, and hence could be BHs greater than or equal to a few solar masses, very young supernovae, or microquasars (George et al. 2001).

NGC 3998. Only one source is detected within a 25″ circle around NGC 3998 in the Chandra image. We fit the source with an absorbed power-law, using the cash statistic. We found an upper limit on the intrinsic column density of 10²¹ cm^-1 and a power-law photon index of . We derived a corrected 0.5 − 10 keV flux of 3  ±  1  ×  10^-14 erg s^-1 cm^-2, which corresponds to a luminosity of 7  ±  1  ×  10³⁸ erg s^-1, adapting the NGC 3998 distance of 14.1 Mpc. This corresponds to only 0.2% of the total core luminosity of NGC 3998 and match the luminosity of XRBs in nearby galaxies.

© ESO, 2011