EUROPEAN SOUTHERN OBSERVATORY VERY LARGE TELESCOPE OPTICAL SPECTROSCOPY OF BL LACERTAE OBJECTS. III. AN EXTENSION OF THE SAMPLE

The optical spectrum of a BLL can be described, in a first approximation, as the superposition of two components. The first one is a non-thermal continuum emitted by the active nucleus. The second is the contribution from the stars and the ISM of the BLL host galaxy. Extensive studies in the past (e.g., Urry et al. 2000) have shown that these galaxies are usually giant ellipticals, the optical magnitudes of which follow a narrow Gaussian distribution centered around M^host_R = −22.9 ± 0.5 (Sbarufatti et al. 2005a). In most cases the host galaxy signature was not detected in our spectrum because it was too faint with respect to the nuclear component. For these objects, we performed a fit of the optical continuum with a simple power-law model (F_λ ∝ λ^−α). In two cases however (RBS 1752 and RBS 1915) the host galaxy spectral features were detected, and we performed a fit to the spectrum using a two-component model (see Figure 2): a power-law plus an elliptical galaxy spectrum as in the template given in Kinney et al. (1996; also see S06 for details on the fitting procedure). The results from our fits are reported in Table 2, where for each object we give the power-law index, the object class (High-Energy Peaked BLL, HBL; Low-Energy Peaked BLL, LBL; see Padovani & Giommi 1995 for a definition), the apparent magnitude, and the extinction coefficient. For RBS 1752 and RBS 1915, we also report the absolute magnitudes of the host galaxies corrected for the aperture effects, which were M_R = −23.3 and −22.4, respectively, in agreement with the expected distribution.

Zoom In Zoom Out Reset image size

Since emission and absorption lines in a BLL spectrum can be very faint, their detection can be a difficult task. Using the same technique presented in S06, we estimated the minimum detectable equivalent width, EW_min, for each spectrum, and considered all features with equivalent width (EW) above this threshold as line candidates which were then carefully inspected for identification or rejection. EW_min values for each spectrum are given in Table 1, while line identifications, full width at half-maximum (FWHM), and EW are reported in Table 3. The continuum and line properties confirmed the BL Lac classification for 12 objects. In Section 3.3 we report notes on the individual sources.

3.2.1. Redshift Lower Limits

Most of the confirmed BLLs in our sample show featureless spectra, despite the high S/N reached using VLT. As extensively discussed in S06 (in Section 4.2.4), it is possible to estimate a lower limit to the redshift of such sources, knowing the EW_min and the nucleus apparent magnitude, and exploiting the assumption that BL Lac hosts can be considered as candles (Sbarufatti et al. 2005a). We remark here that Equation (1) in S06 contains a typographical error, which was noted also by Finke et al. (2008). The correct expression is

$$\begin{equation} \mathrm{EW}_{\rm obs}=\frac{(1+z) \times \mathrm{EW}_0}{1+\rho / A(z)}, \end{equation} \tag{ 1 }$$

where EW_obs is the observed equivalent width, EW₀ is the equivalent width of the feature in the host galaxy template (Kinney et al. 1996), ρ is the nucleus-to-host flux ratio, z is the redshift, and A(z) is the aperture correction.

We applied this procedure to all featureless spectra in the sample, obtaining the lower limits reported in Section 3.3.

PKS 0019+058. This radio-selected source (Condon & Jauncey 1974) was first classified as a BLL by Fricke et al. (1983), based on a featureless optical spectrum. We observed this object in two epochs separated by about a month, noticing an optical variability of 0.7 mag (R band) and an evolution of the spectral index from 0.65 to 0.76. The reality of the optical variability is fully confirmed by the R-band images exposed prior to the spectra which enable direct photometry. The optical variability was known, both in magnitude (V = 19.2 and V > 21 in Fricke et al. 1983; Abraham et al. 1991, respectively) and in spectral index (α_ν = 0.8 and 0.94 in Fricke et al. 1983; Chen et al. 2005). We find no intrinsic feature with EW > EW_min in any of the two spectra. The Na i λ5891 absorption from the Galaxy ISM is detected in both spectra with EW = 0.43 Å (July 12) and 0.66 Å (August 8). A marginally significant excess around 5600 Å is present in both observations, but it is most probably spurious because of residuals left after the subtraction of a nearby atmospheric emission line. The most stringent redshift lower limit, obtained from the 2006 August spectrum, is z > 0.64.

GC 0109+224. This source was discovered in radio observations by Davis (1971), and subsequently classified as a BLL by Owen & Muffson (1977). Strong flux and polarization variability was reported both in the radio and optical band (e.g., Katajainen et al. 2000; Ciprini et al. 2003, 2004). Optical imaging by Falomo (1996) and Nilsson et al. (2003) failed to detect the host galaxy, while the detection reported by Wright et al. (1998) is dubious, since it has not been confirmed by any subsequent observation. The lower limit on the redshift proposed by Falomo (1996) is z > 0.4. Previous optical spectroscopy by Wills & Wills (1979), Falomo et al. (1994), and Sbarufatti et al. (2006a) performed with 2–4 m class telescopes showed featureless spectra, while Healey et al. (2008) report z = 0.265 for this object based on an unpublished optical spectrum. The lower limit to the redshift as deduced by Sbarufatti et al. (2006a) was based on an EW_min = 0.43 Å that yielded z > 0.18. The considerably higher S/N spectrum that we obtained with VLT has EW_min = 0.09 Å, which in turn implies z > 0.25, which improves the previous spectroscopical limit, but it is still less stringent than the imaging limit.

RBS 0231. This X-ray-selected object (Voges et al. 1999) was classified as a BLL by Schwope et al. (2000), and as a HBL by Brinkmann et al. (2000). No previous optical spectroscopy was published. The VLT spectrum is featureless, with EW_min = 1.27 Å, which implies z > 0.41.

PKS 0823 − 223. This radio-selected BLL (Allen et al. 1982) is characterized by a number of absorption lines in the UV and optical band (e.g., Rao & Turnshek 2000; Meiring et al. 2007; Falomo 1990; Veron et al. 1990; Falomo et al. 1994). These features are consistent with the presence of a subdamped Lyα (sub-DLA) absorber at z = 0.91. In our VLT spectrum we detect the absorption features of Fe ii λλ2373.7, 2383.2, 2585.9, 2599.4, Mg ii λ2798, and Mg i λ2852 at the same redshift.

PKS 1057 − 79. This radio-selected object (Shimmins & Bolton 1981) was proposed as a counterpart for the γ-ray source 2EGS 1050 − 7650 by Tornikoski et al. (2002). No previous optical spectroscopy was published. Our VLT spectrum shows several emission lines ([O iii] λλ4959, 5007, [Ne iii] λ3868, Mg ii λ2798), at z = 0.581. Since the FWHM of the Mg ii line (see Table 3) is in excess of 1000 km s⁻¹ we propose to classify this object as a broad-line active galactic nucleus (AGN).

PKS 1145 − 676. This radio source was classified as a quasar due to its point-like appearance by White et al. (1987). The flat radio spectrum and optical variability (Beasley et al. 1997; Costa 2002) prompted a blazar classification. We detect several emission lines ([O ii] λ3727, Hβλ4861, [O iii] λλ4959, 5007, Hαλ6563, [N ii] λ6585) at z = 0.210. The observed EW are in the range 4–25 Å, pointing towards a FSRQ classification. The FWHM of the Hβ line could indicate that this source is a narrow lines object, but the fact that we are unable to measure the Hα FWHM due to the blending with the [N ii] lines permits only a type ∼1.9 classification.

OM 280. This radio-selected source (Colla et al. 1972) was classified as a BLL due to its featureless spectrum by Strittmatter et al. (1974). Subsequent optical spectroscopy by Rector & Stocke (2001) also failed to discover any intrinsic spectral feature. The host galaxy was not detected in deep Hubble Space Telescope imaging (Urry et al. 2000), implying z > 0.63 (Sbarufatti et al. 2005a). The VLT spectrum is featureless, with EW_min = 0.35 Å, which implies a redshift lower limit of 0.20, less stringent than the one from imaging.

PMN J1323–3652. This radio-selected source was classified as a candidate BLL by the WGA catalog (White et al. 2000) and the Deep X-ray Radio Blazar Survey (DXRBS, Landt et al. 2001) that also provided a featureless optical spectrum. However, our VLT optical spectrum clearly shows the absorption lines and the characteristic shape of a Galactic F-type star. This indicates a misidentification of the optical counterpart of this source.

OQ 012. This radio-selected BLL (Weiler & Johnston 1980) showed a featureless optical spectrum in observations by Falomo et al. (1994). Richards et al. (2004) gave a photometric redshift estimate z = 0.475 based on data from the Sloan Digital Sky Survey (SDSS). Our VLT spectrum shows an absorption line by Galactic ISM (Na i λ5891), but no intrinsic features. The EW_min is 0.31 Å, which implies z > 0.65, inconsistent with the photometric estimate.

PMNJ 1539 − 0658. This radio source (Griffith et al. 1995) was classified as a BLL in the DXRBS (Landt et al. 2001), which also provided a featureless optical spectrum. In the VLT spectrum we detect the Na i λ5891 absorption line from the Galaxy, but no intrinsic features. The minimum detectable EW is 0.61 Å, which implies z > 0.80.

PKS 1830 − 589. This radio-selected BLL (Griffith et al. 1995; Landt et al. 2001) showed a featureless optical spectrum when observed in the DXRBS. Our VLT spectrum is also featureless (Na iλ5891 is marginally detected with EW = 0.4 Å). The minimum detectable equivalent width is EW_min = 0.46 Å, which gives a lower limit z > 0.45.

RBS 1752. This X-ray-selected BLL (Voges et al. 1999) had a tentative redshift z = 0.449 proposed in the Sedentary Survey (Giommi et al. 2005a; Piranomonte et al. 2007), based on the possible detection of the host galaxy spectral features in an ESO 3.6 m optical spectrum. The high S/N obtained using VLT allowed us to detect some weak host galaxy lines (Ca ii λλ3934, 3968, G band λ4305, and Mg i λ5175) at z = 0.449. However, the G band is possibly contaminated by the [O i] atmospheric line at 6300 Å, and the Mg i line is very close to the telluric O₂ A band. Therefore, while the absorption features reported by Piranomonte et al. (2007) are confirmed, the redshift remains tentative because of the lack of other firm absorption features. The fit to the spectrum with a power-law plus elliptical galaxy model gives M^host_R = −23.3, in good agreement with the expected distribution M^host_R = −22.9 ± 0.5.

RBS 1915. This X-ray-selected object (Voges et al. 1999) was classified as a BLL by Schwope et al. (2000). The optical spectrum reported by Chavushyan et al. (2000) was featureless. Our VLT spectrum shows faint absorption lines from the BLL host galaxy (Ca iiλλ3934, 3968, G band λ4305, and Mg i λ5175) at z = 0.243. We performed a fit to the spectrum using a power-law plus elliptical galaxy model, obtaining M^host_R = −22.4, consistent with the expected distribution (M^host_R = −22.9 ± 0.5).

TXS 2346+052. This radio-selected source (Large et al. 1981) was classified as a BLL because of its flat radio spectrum (Gorshkov et al. 2000) and the featureless optical spectrum (Chavushyan et al. 2000). Our VLT spectrum shows several emission lines (Mg ii λ2798, [O ii] λ3727, [Ne iii] λ3868, [O iii] λλ4959, 5007) at z = 0.419. The observed EW of the Mg ii and [O iii] lines (exceeding 5 Å), rules out a BL Lac classification and suggests a FSRQ nature for this source. The EW ratio between the [O ii] and [O iii] lines are untypical for an AGN, possibly indicating an ongoing star formation (as seen in PKS 2005 − 489 by Bressan et al. 2006). A measurement of the equivalent width of the Hα+[N ii] line system (which is out of the observed spectral range) could help us to clarify this issue.

1RXS J235730.1 − 171801. This X-ray-selected object (Voges et al. 1999) was classified as a BLL by Schwope et al. (2000). Our previous VLT observations (S06) gave a limit z > 0.85. The spectrum presented here has a slightly lower S/N (110, to be compared with 150 of the earlier observation, because of the different seeing conditions between the observations), that gives EW_min = 0.22 Å and z > 0.60. Ca ii λ3934 and Na i λ5891 absorptions from the Galaxy ISM are marginally detected (they were also detected in the S06 spectrum, along with several DIBs). No significant flux variations were detected between two different observation periods.