OPACOS: OVRO POST-AGB CO (1–0) EMISSION SURVEY. I. DATA AND DERIVED NEBULAR PARAMETERS

This Mira variable (Groenewegen et al. 1999) appears to be point-like in our HST survey of candidate PPNs (Sahai et al. 2007a). It shows a strong 9.7 μm silicon dust feature (Heske et al. 1990) and SiO and OH maser emission (Nyman et al. 1998; Omont et al. 1993, and references therein), indicative of O-rich chemistry. The double-peaked OH maser line suggests a molecular shell expanding at about 9–10 km s⁻¹ (Heske et al. 1990; Engels & Jiménez-Esteban 2007). No spectral type of the central star is available in the literature: Based on variability and O-rich chemistry, it is likely to be a late M type.

¹²CO (J = 1–0) single-dish spectra have been previously reported (Loup et al. 1993; Groenewegen et al. 1999). Our maps show a point-like ¹²CO (J = 1–0) emission source with a line profile centered at $\mbox{$V_{\rm sys}^{\rm LSR}$}=-36.5 \pm 0.7$ km s⁻¹ and full width similar to that spanned by the OH maser, ∼16 km s⁻¹ (Figures 3.1 and 4). The ¹²CO profile and total line flux are comparable (within the errors) to the single-dish spectrum, therefore major flux losses are not present in our interferometric data.

Taking into account the galactic coordinates and radial velocity of IRAS 03206+6521 (l = 13797, b = +0726, and $\mbox{$V_{\rm sys}^{\rm LSR}$}=-36.5$ km s⁻¹), we derive a value of the kinematic distance of d_k = 3.6 kpc. For this object, there is an estimate of the OH maser phase-lag distance, d = 3.4 kpc (Herman et al. 1986), which we adopt in this paper. (There is another estimate of the distance obtained by comparing the linear and angular sizes of the OH shell, d = 6.4 ± 3.6 kpc. This geometric distance is, however, quite uncertain because the linear size of the OH shell is indirectly estimated from a model and the angular size of the envelope has large error bars (04 ± 015); Chapman et al. 1984.) The bolometric flux of this object is F_bol ∼ 1250 L_☉ kpc⁻², from which we derive a value of d_L = 2.2 kpc. For d ∼ 3 kpc, the visual ISM extinction toward the source is A_V ∼ 0.7 mag and, therefore, the intrinsic (dereddened) bolometric flux is F_bol ∼ 1280 L_☉ kpc⁻², which implies a total luminosity of 1.2 × 10⁴ L_☉ at d = 3 kpc. This is at the high end of the range of typical luminosities for pAGB objects.

This semiregular variable appears to be a bright point-like source in our HST survey of candidate PPNs (Sahai et al. 2007a). Its spectral type is ∼G8 Ia (G8-M2 as per the Variable Stars Catalogue). It has also been classified as a yellow supergiant (Mv = −7.2 to −8 mag, which is consistent with luminosity class I), so there is some possibility that its progenitor was a massive (≳8 M_☉) star. Based on the presence of silicate dust features in the mid-IR and its classification as 2.SE7 by Sloan et al. (2003), this is an O-rich source. No SiO or OH maser emission has been detected so far (Nyman et al. 1998; Omont et al. 1993).

No CO emission detection has been reported for this object before: ¹²CO (J = 2–1) emission was unsuccessfully searched for by Omont et al. (1993). Our observations show a ¹²CO (J = 1–0) emission line with sharp edges centered at V_LSR ∼ +25 km s⁻¹ and rather affected by ISM contamination (Figures 3.2 and 4). The relatively large expansion velocity inferred from the CO line core (FWHM = FWZI ∼ 25 km s⁻¹) supports a high-mass nature in this case. The velocity-channel maps show structured emitting regions; some of these structures are partially due to the presence of absorption and emission by intervening or nearby interstellar clumps. The velocity-integrated CO map shows an extended source with a beam-deconvolved half-maximum size of 75 × 39 and roughly oriented along the EW direction. Our data indicate a velocity gradient approximately in the perpendicular direction: The blue- and red-wing emission clumps are offset by ±1'' to the north and south, respectively (Figure 3.2, bottom-right panel). The elongation and velocity gradient found could indicate an equatorially dense expanding torus, however, the origin of the velocity gradient and asymmetry observed cannot be firmly established as predominantly circumstellar (but, rather, may be an artifact due to the strong ISM contamination) and, therefore, we consider this result very tentative.

The distance to IRAS 18055−1833 is extremely uncertain. Given the low galactic longitude and latitude of this source (l = 116, b = 07), the values of the near and far kinematic distance are unreliable due to non-circular motions close to the galactic center. The bolometric flux of this object is F_bol ∼ 5400 L_☉ kpc⁻², from which we derive a value of d_L = 1.05 kpc. For d ∼ 1 kpc, the visual ISM extinction toward the source is A_V = 0.5 mag and, therefore, the intrinsic (dereddened) bolometric flux is $\mbox{$F_{\rm bol}$}= L \sim 6400$ L_☉ kpc⁻². For a larger value of d, e.g., d_k(near) = 3.3 kpc, we expect A_V = 2 mag and F_bol ∼ 1.1 × 10⁴ L_☉ kpc⁻², implying a rather high luminosity of ≈10⁵ L_☉ at such a distance, which is more typical of a massive supergiant star. Here we adopt d = 2 kpc, which is intermediate to d_L and d_k(near).

This non-variable OH/IR star (van der Veen et al. 1989) appears as a point-like source in our HST survey of candidate PPNs (Sahai et al. 2007a; Siódmiak et al. 2008) and in 8 μm subarcsecond-resolution images (Lagadec et al. 2011). SiO, H₂O, and OH masers have been detected, which, together with the presence of 9.7 μm silicate absorption, indicates an O-rich chemistry (Nakashima & Deguchi 2007; Omont et al. 1993; Deguchi et al. 2007, and references therein). The stellar spectral type is unknown, but since CO bandheads appear in absorption in the NIR spectrum of this object implying a photospheric temperature of ∼3000–5000 K (Oudmaijer et al. 1995), it may be G5-K0 (assuming it is a supergiant). The double-peaked OH maser emission profile is consistent with shell expansion at ∼15 km s⁻¹ (Engels & Jiménez-Esteban 2007).

No CO emission has been previously reported for this object. Our data (Figures 3.3 and 4) show a ¹²CO (J = 1–0) emission line centered at $\mbox{$V_{\rm sys}^{\rm LSR}$}=0 \pm 1$ km s⁻¹, and with a peculiar triple-peaked profile: There is a relatively intense central emission component with a full width similar to that spanned by the OH masers; at both red and blue sides of the line core (approximately at $\mbox{$V_{\rm sys}^{\rm LSR}$}\pm 22$ km s⁻¹), there are two narrow emission features indicative of faster motions. The redshifted narrow feature is significantly altered by strong ISM contamination, especially in the range V_LSR = [+22:+27] km s⁻¹. A similar triple-peaked profile is observed in the ¹³CO (J = 1–0) single-dish spectrum of CRL 2688 by Quintana-Lacaci et al. (2007). As shown by those authors, a bipolar wind with a low inclination (with respect to the plane of the sky) leads to wings like those observed in IRAS 18135−1456, that is, narrow and detached from the central line core.

Our velocity-integrated ¹²CO map, excluding channels with ISM contamination, has a deconvolved half-maximum size of ∼31 × 18 with its major axis roughly oriented along the EW direction. The centroids of the red- and blue-velocity-integrated maps at V_LSR = [+14.3:+11.7] and [−14.3:−11.7] km s⁻¹, i.e., adjacent to the line core, are offset to the south and north, respectively (see Figure 3.3, bottom-right panel). The elongation of the velocity-integrated intensity map orthogonal to the NS velocity gradient is consistent with the bulk of the low-velocity emission arising in an expanding equatorial torus. The rms noise in our maps and ISM contamination of the circumstellar profile does not enable us to reliably locate the emission from the weak, broad wings of the profile (beyond ±14.3 km s⁻¹) and, therefore, the presence of a fast bipolar flow, suggested by the triple-peaked profile, cannot be confirmed.

We detect 2.6 mm continuum emission from a point-like source with a total flux ∼11 mJy, consistent with thermal ≈100 K dust emission according to the slope of the SED from the IR to the millimeter-wavelength spectral region.

The distance to IRAS 18135−1456 is uncertain. Given its low galactic longitude (l = 157), the values of the near and far kinematic distance are unreliable due to non-circular motions close to the galactic center. We adopt here d = d_L = 2.5 kpc, taking into account the bolometric flux of this object F_bol ∼ 980 L_☉ kpc⁻². For such a distance the visual ISM extinction is A_V = 1.5–2.0 mag and, thus, the intrinsic (dereddened) bolometric flux is F_bol ∼ 1020 L_☉ kpc⁻², which implies an insignificantly smaller value of the distance, d_L ∼ 2.4 kpc.

This OH/IR star appears as a point-like object in our optical HST survey of candidate PPNs (Sahai et al. 2007a). The central star has a spectrum consistent with an F 7 classification and it is surrounded by a compact Hα-emitting region (Sánchez Contreras et al. 2008). It shows strong OH maser emission with a double-peaked profile in the LSR range [+164.6:+187.8] km s⁻¹; no SiO maser emission has been detected (van Langevelde et al. 1990; Nyman et al. 1998; Engels & Jiménez-Esteban 2007).

¹²CO (J = 1–0) and (J = 2–1) single-dish spectra have been reported by Winnberg et al. (1991): The emission is centered at V_LSR = +176 km s⁻¹ and its profile is consistent with expansion at ∼12 km s⁻¹. Our data show a weak, spatially unresolved ¹²CO (J = 1–0) emission source around V_LSR = +176 km s⁻¹ (Figures 3.4 and 4). Our measurement of the line flux (Table 3) is smaller than that obtained from single-dish observations (∼0.5 ± 0.25 Jy). The flux discrepancy may be attributed, at least partially, to the large flux error bars in both data sets (of up to 50%). Although we cannot rule it out completely, we think our data, with relatively low resolution (HPBW ∼ 5''), are unlikely to be affected by significant interferometric flux losses, which would imply the presence of a CO envelope with an angular diameter ≳18''. The presence of such a large envelope is improbable because, as for most PPNs, the outer radius of the CO envelope is unlikely to be larger than few × 10¹⁷ cm (Neri et al. 1998), implying an upper limit to the angular radius of the CO-emitting nebula of less than ∼1'' given the probable large distance to this object (see the next paragraph).

As discussed by Winnberg et al. (1991), considering its galactic coordinates (l = 1852, b = +141), the radial velocity of IRAS 18167−1209, $\mbox{$V_{\rm sys}^{\rm LSR}$}=+176$ km s⁻¹, is close to the maximum allowable velocity for circular motions and, therefore, it is probably located near the tangent point, i.e., d_k ∼ 7–8 kpc. Since the bolometric flux of this object is F_bol ∼ 100 L_☉ kpc⁻², we derive a very similar value for d_L = 7.7 kpc. For d ∼ 7–8 kpc, the visual ISM extinction is A_V = 1.5–3.5 mag and the intrinsic (dereddened) bolometric flux is F_bol ∼ 100–140 L_☉ kpc⁻², which consistently implies a similar value for d_L of 6.5–7.6 kpc. No reliable OH maser phase-lag distance has been determined in this case because of its weak variability. We adopt a distance to IRAS 18167−1209 of d ∼ 7 kpc.

This young O-rich PPN has been imaged in our optical/HST and NIR/adaptive optics (AO) surveys of candidate PPNs (Sahai et al. 2007a; Sánchez Contreras et al. 2006b). It appears as a bipolar object with searchlight beams and arcs, and has been studied in detail by Sánchez Contreras et al. (2007). The spectral type of the central star is earlier than ∼K 5 (Le Bertre et al. 1989). The progressive disappearance of H₂O maser emission is consistent with a recent drop of the mass-loss rate from ≈10⁻⁵ M_☉ yr⁻¹ to well below ≈10⁻⁷ M_☉ yr⁻¹ (Engels 2002). The non-detection of SiO masers in IRAS 18276−1431 is also consistent with its pAGB nature (Nyman et al. 1998). Strong OH maser emission is observed in a dense, equatorial region elongated in the direction of P.A. = 110°, i.e., approximately perpendicular to the bipolar lobes (Bains et al. 2003). The bipolar lobes are marginally resolved in subarcsecond-resolution 8–12 μm images by Lagadec et al. (2011).

Single-dish ¹²CO (J = 2–1) emission has been previously reported by Heske et al. (1990). The ¹²CO (J = 1–0) transition was observed by Knapp et al. (1989) and Likkel et al. (1991) but undetected with rms ∼0.2 Jy. Data from this survey, therefore, represent the first detection of ¹²CO (J = 1–0) emission from this target. Our CO maps from OPACOS (Figure 3.5) were presented and analyzed prior to this paper in Sánchez Contreras et al. (2007); here, we briefly summarize these results. We find an extended CO envelope with a deconvolved half-maximum size of 5'' × 3''; this size similar to that of the halo observed in the NIR (Sánchez Contreras et al. 2007). The long axis of the CO envelope appears to be roughly oriented along the EW direction (after beam deconvolution of the maps), that is, similarly to the equatorially dense structure probed by OH maser emission and visible as a dark waist separating the lobes in the optical and NIR images. Therefore, a significant part of the CO emission probably arises in an equatorially dense toroidal structure. We note, however, that the envelope parameters deduced from our CO maps are relatively uncertain due to the ISM contamination of the circumstellar profile and moderate angular resolution.

Given the flux of the ¹²CO (J = 1–0) line derived from our data (∼1.5 Jy), this line should have been detected in Knapp et al. (1989) and Likkel et al. (1991) single-dish studies. We do not know the reason for this inconsistency, but large errors in the absolute flux calibration, pointing, and/or strong ISM contamination in the single-dish data could explain the flux discrepancy. In any case, flux losses are unlikely to affect our data. We detect ¹³CO (J = 1–0) emission from this source for the first time. The profile, although noisy, is consistent with a double-peaked shape centered at a few km s⁻¹ redward from the ¹²CO line (Figure 5).

A compact source of 2.6 mm continuum emission is detected toward IRAS 18276−1431. As discussed in detail by Sánchez Contreras et al. (2007), a population of big dust grains (radius ≳4 mm) with temperatures ∼20–150 K can explain the observed millimeter and submillimeter flux.

OH maser phase-lag measurements indicate a distance to IRAS 18276−1431 in the range d = 2–5.4 kpc (Bowers et al. 1983; Herman & Habing 1985). The near kinematic distance is d_k = 4.5 kpc (Le Bertre et al. 1989). In this paper, we adopt an intermediate value of d = 3 kpc, which is also close to the value of d_L derived from the bolometric flux of this object, F_bol ∼ 870 L_☉ kpc⁻². For d ∼ 3 kpc, the visual ISM extinction is A_V = 1.0–2.0 mag and, therefore, the intrinsic (dereddened) bolometric flux is F_bol ∼ 900–915 L_☉ kpc⁻², which consistently implies a similar value of d_L ∼ 2.6 kpc.

This is an O-rich Mira-type variable near the tip of the AGB phase. It was observed but not detected in our HST optical survey of candidate PPNs (Sahai et al. 2007a). The source has very bright Two Micron All Sky Survey (2MASS) and Midcourse Space Experiment (MSX) counterparts. It shows strong 9.7 and 18 μm silicate features and water ice far-IR bands (Sylvester et al. 1999). It has a spectral type of M (as per SIMBAD database). OH and SiO masers are detected. The OH maser emission arises in an elongated shell of radius ∼25 with its major axis oriented along P.A. = 95° (Herman et al. 1985; Bowers & Johnston 1990; Etoka & Diamond 2010). SiO maser emission, mapped with the Very Long Baseline Array, reveals a rather chaotic, asymmetric ring-like distribution with a characteristic radius ∼001 and maser plumes roughly oriented along the NS direction (Cotton et al. 2008).

Single-dish spectra of ¹²CO (J = 1–0) and higher-J transitions have been previously reported by, e.g., Heske et al. (1990) and Justtanont et al. (1996): Strong galactic contamination of the single-dish ¹²CO (J = 1–0) profile is noticed. Interferometric ¹²CO (J = 1–0) maps with 98 × 7'' resolution also exist (Fong et al. 2002). We find ¹²CO emission over a full velocity range of ∼24 km s⁻¹, similar to that spanned by the OH masers (Figures 3.6 and 4). The ¹²CO profile has a hint of asymmetry, the blue side is weaker, maybe due to self-absorption or absorption by intervening ISM components. The upper limit to the ¹²CO (J = 1–0) line flux published by Heske et al. (1990) from single-dish observations, ≲0.5 Jy, is consistent with our measurement, ∼0.6 Jy, within 20%–30% absolute flux errors. This indicates that there are no significant flux losses in our interferometric data. The ¹²CO (J = 1–0) flux measured by Fong et al. (2002) with a ∼9'' beam is also ∼0.6 Jy.

The object appears extended in our ¹²CO (J = 1–0) maps (Figure 3.6): The deconvolved, half-maximum size of the velocity-integrated CO map is 43 × 38, with its long axis oriented roughly along the EW direction, i.e., parallel to the major axis of the OH shell and dusty environment probed by mid-IR emission (Bowers & Johnston 1990; Chesneau et al. 2005). This result is consistent with CO probing an equatorially dense toroidal structure. The deconvolved, half-maximum size of the CO envelope measured in our data is smaller than that measured by Fong et al. (2002), ∼85 × 55, in their lower (HPBW ∼ 9'') resolution maps. This discrepancy is unlikely to result from interferometric flux losses or the sensitivity limit in our maps given the good agreement between the line flux measured by us and that obtained from single-dish data and the interferometric maps by Fong et al. (2002) themselves (see the previous paragraph). The different size of the CO envelope derived from maps with different angular resolution may reflect the two distinct mass-loss regimes presumably undergone by this object (see Justtanont et al. 1996).

We find a velocity gradient in the nebula along the NS direction: The redshifted (blueshifted) emission is systematically offset toward the north (south). Although we cannot rule out such a gradient being an artifact produced by ISM absorption of the circumstellar profile, the remarkable symmetry of the contours of the velocity map around the CO emission peak favors a circumstellar origin. In fact, an expanding torus with its symmetry axis projected along the NS direction and tilted with respect to the line of sight in such a way that the back (receding) and front (approaching) sides of the torus are offset to the north and south, respectively, could satisfactorily explain both the EW elongation of the CO envelope and the NS velocity gradient observed. Alternatively, the velocity gradient may indicate the presence of a compact bipolar outflow running along the NS direction (i.e., orthogonally to the torus), co-linear with the asymmetric distribution of the SiO maser plumes. Higher-angular resolution CO mapping is needed to unambiguously characterize the structure and kinematics of the molecular envelope in this compact object.

We report the detection of weak ¹³CO (J = 1–0) emission in the range V_LSR = [+15:+35] km s⁻¹ (Figure 5).

We adopt a distance to this object of d = 1.1 kpc, which is the average of two independent estimates of the OH maser phase-lag distance (d = 1.4 and 0.9 kpc, respectively; van Langevelde et al. 1990; Bowers & Johnston 1990). The bolometric flux of this object is F_bol ∼ 6300 L_☉ kpc⁻², which yields a value for the luminosity distance of d_L ∼ 1 kpc. For d ∼ 1 kpc, the visual ISM extinction is A_V = 0.1–1.0 mag (Hakkila et al. 1997) and, therefore, the intrinsic (dereddened) bolometric flux is not significantly higher, F_bol ∼ 6300–6600 L_☉ kpc⁻².

This OH/IR object was imaged in our optical/HST and NIR/AO surveys of candidate PPNs and shows a central bipolar shape (oriented along P.A. ∼ −60°) surrounded by an extended halo traced out to a radius of ∼3'' with centrosymmetric arc-like features (Sahai et al. 2007a; Sánchez Contreras et al. 2006b). One, possibly two, secondary (minor) lobes with collimated shapes (oriented along P.A. ∼ 75°–80°) emanate from the central region.⁶ The central star has been assigned an M 1I spectral type (Suárez et al. 2006). A double-peaked OH maser emission profile indicates a dense molecular shell expanding with V_exp ∼ 14 km s⁻¹ (Engels & Jiménez-Esteban 2007). No SiO or H₂O maser emission has been reported (Gómez et al. 1990).

No CO detection has been previously reported for this object. Our data show ¹²CO (J = 1–0) emission with a slightly asymmetric profile over a total velocity range a little larger than the OH maser line (Figures 3.7 and 4). The CO emission arises in an elongated envelope with a deconvolved half-maximum size of ∼35 × 12 and roughly oriented along the NS direction. We identify a velocity gradient along the same axis: The redshifted and blueshifted emission regions are systematically displaced toward the north and south, respectively. The similar orientation of the long axis of the CO envelope and the direction along which the velocity gradient is observed suggest that the CO emission probes a bipolar molecular outflow. This outflow may be tracing a (third) bipolar ejection distinct from those probed by the optical HST images. ¹³CO (J = 1–0) emission is also detected (Figure 5). The envelope traced by this transition, which is also extended, has a deconvolved half-maximum size of ∼56 × 42 and similar orientation (within the errors) to the ¹²CO envelope.

The near and far kinematic distances to IRAS 18420−0512 taking into account its galactic coordinates and radial velocity (l = 2757, b = −085, $\mbox{$V_{\rm sys}^{\rm LSR}$}=+105$ km s⁻¹) are d_k(near) = 6 and d_k(far) = 9 kpc. As discussed by de Jong (1983) and Herman et al. (1985), since the radial velocity is close to the maximum velocity allowed in that direction, IRAS 18420−0512 could be at the tangent point (d ∼ 7.5 kpc). The bolometric flux of this object is F_bol ∼ 160 L_☉ kpc⁻² from which we derive d_L = 6.1 kpc. At such a distance (6–7 kpc), the visual ISM extinction in the direction to IRAS 18420−0512 is A_V = 2–4 mag and, thus, the dereddened bolometric flux is F_bol ∼ 190–230 L_☉ kpc⁻², which yields a consistent value for the distance of d_L = 5–5.6 kpc. We adopt a value for the distance to this object of d = 6 kpc.

This OH/IR star shows H₂O maser emission spread over a very wide velocity range (≳200 km s⁻¹) and likely belongs to the class of "water-fountain" PPNs (Deguchi et al. 2007, and references therein). Although no 2MASS counterpart could be found for this object, it has a bright MSX match and it is also identified at wavelengths 3.5–5.8 μm in the Spitzer/GLIMPSE survey (Deguchi et al. 2007). The rising mid-IR spectrum toward longer wavelengths points to a thick dust envelope, which is spatially unresolved in 8–12 μm subarcsecond-resolution images (Lagadec et al. 2011). In addition to the very fast outflow probed by the H₂O masers, the double-peaked OH maser emission profile (with peaks at V_LSR = +110 and +140 km s⁻¹) indicates a dense shell expanding at V_exp ∼ 15 km s⁻¹ (Engels & Jiménez-Esteban 2007). The optical counterpart to IRAS 18460−0151, which is expected to be very faint, cannot be unambiguously identified in our optical/HST images,⁷ which show several faint, point-like objects near the MSX and OH maser source.

No CO detection has been previously reported toward this object. Our data show ¹²CO (J = 1–0) emission over a total velocity range similar to that of the OH maser line (Figures 3.8 and 4). There is strong ISM contamination of the maps near the source but at velocities outside the circumstellar profile (at V_LSR ∼ [+103:+108] and ∼90 km s⁻¹). The CO envelope is marginally extended with a deconvolved half-maximum size of 21 × 05 and oriented along P.A. ∼ 50°. We find a tentative velocity gradient along the NS direction: The redshifted and blueshifted emission are systematically displaced toward the north and south, respectively. Since this source is only marginally resolved, the errors in the envelope parameters derived, in particular the CO envelope orientation and asymmetry, are expected to be larger than the average. Therefore, the relative orientation of the CO envelope and the tentative gradient found is uncertain and we cannot elucidate the structure(s) responsible for the asymmetries observed.

As discussed, e.g., by de Jong (1983) and Deguchi et al. (2007), the radial velocity of IRAS 18460−0151 ($\mbox{$V_{\rm sys}^{\rm LSR}$}=+126$ km s⁻¹) is close to the maximum velocity allowed in that direction (l = 3101, b = −022) and, therefore, this object is probably at the tangent point, implying a distance of d_k ∼ 7 kpc, which we adopt. The bolometric flux of this object is F_bol ∼ 600 L_☉ kpc⁻². At a distance of ∼7 kpc, the visual ISM extinction in the direction to IRAS 18420−0512 is A_V ∼ 3.5 mag and, therefore, the dereddened bolometric flux is F_bol ∼ 690 L_☉ kpc⁻². This implies a total luminosity of the source L = 3 × 10⁴ L_☉ at d = 7 kpc, which is at the high end of the typical luminosity range for pAGB objects. The luminosity distance for this object is d_L = 3 kpc.

This Mira-type variable star does not have an optical counterpart down to a limit of ∼20 mag in the F photographic band of The Guide Star Catalog, Version 2.3.2 (GSC2.3). A very deep 9.7 μm silicate feature indicates an optically thick envelope with O-rich chemistry (Heske et al. 1990). OH, H₂O, and SiO masers have been detected (Omont et al. 1993; Deguchi et al. 2007; Nyman et al. 1998). The H₂O maser profile varies with time in a unique manner, exhibiting a double-peaked profile similar to the 1612 MHz OH maser line during the bright phase of the star and a single line close to the stellar radial velocity while the star was dim (Engels et al. 1997).

¹²CO (J = 2–1) emission was reported by Heske et al. (1990) with a profile severely affected by ISM contamination. Heske et al. (1990) and Neri et al. (1998) published a marginal detection of ¹²CO (J = 1–0) with the IRAM 30 m radio telescope. Our data (Figures 3.9 and 4) show circumstellar CO emission over a total velocity range slightly larger than the OH maser profile, V_LSR ∼ [0:+43] km s⁻¹. There are ISM absorption and emission clumps near IRAS 18560+0638, which results in a triple-peaked CO profile with very sharp edges. Due to ISM contamination, it is very difficult to obtain an accurate value for the CO flux (the value in Table 3 is probably a lower limit). The ¹²CO (J = 1–0) intensity at the line peak from single-dish data (references above), 0.8 ± 0.4 Jy, and from OPACOS (0.45 Jy) are comparable taking into account the mentioned flux uncertainties. Therefore, and considering also the relatively large beam in our maps of IRAS 18560+0638, it is unlikely that significant flux losses affect our interferometric data.

The velocity-integrated CO map, using only channels with weak or no ISM contamination, as well as the individual velocity-channel maps indicates an extended CO envelope, with a deconvolved half-maximum size of ∼3''–5'' and oriented roughly along P.A. ≈ −30°. A velocity gradient is found along this axis: The redshifted (blueshifted) emission is displaced toward the north (south). We detect ¹³CO (J = 1–0) emission for the first time in this object. The line is centered at V_LSR = +20 km s⁻¹ with a full velocity extent comparable to that of the ¹²CO line (Figure 5). Our velocity-integrated ¹³CO map also suggests an extended envelope with a deconvolved half-maximum size of 52 × 25 with its long axis oriented similarly to the ¹²CO envelope. The alignment (within errors) of the velocity gradient and the long axis of the ¹²CO and ¹³CO envelope suggest a bipolar outflow in that direction. However, due to the strong ISM contamination of the ¹²CO maps and also because of the moderate S/N and angular resolution of the ¹³CO (J = 1–0) data set, we consider this result tentative.

The distance to IRAS 18560+0638 is estimated to be d = 1–2 kpc from OH maser phase-lag measurements (van Langevelde et al. 1990). This value is in excellent agreement with our estimate of the near kinematic distance, d_k = 1.4 kpc (l = 3971, b = +149, and V_LSR = +20 km s⁻¹), which we adopt in this paper. At a distance of ∼1–2 kpc, the visual ISM extinction in the direction to IRAS 18560+0638 is A_V ∼ 0.1–1.4 mag and, therefore, the dereddened bolometric flux is F_bol ∼ 9050–10⁴ L_☉ kpc⁻². At d = 1.4 kpc the total luminosity is L ∼ 1.8 × 10⁴ L_☉.

This object is a young O-rich PPN as deduced from the detailed multiwavelength study by Sahai et al. (2005), which included our ¹²CO (J = 1–0) maps from OPACOS. HST optical images show a multipolar nebula of size ∼37 × 23, with at least six lobes emanating from the central source. A central compact region shows Hα emission with very broad wings (FWZI ∼ 2400 km s⁻¹) and a P Cygni like profile, indicating the presence of a ≈100 km s⁻¹ outflow (Sánchez Contreras et al. 2008). The central star, which is obscured by a dusty equatorial structure in the optical images, has a spectral type G 0-5 (Sahai et al. 2005). OH maser emission suggests the presence of a dense expanding envelope at V_exp = 13–14 km s⁻¹ (Sevenster 2002). H₂O maser emission is not detected (Suárez et al. 2007).

¹²CO (J = 1–0) emission was first detected in this object as part of OPACOS and reported by us in Sahai et al. (2005). Likkel et al. (1991) sought ¹²CO (J = 1–0) with the 30 m IRAM telescope but emission was not detected down to a 1σ level of ∼0.13 Jy. According to the peak line intensity measured by us (∼0.7 Jy), the line should have been detected by these authors, however, pointing errors in the single-dish observations, or flux calibration uncertainties and/or ISM contamination of the profile in both data sets may be the origin of the observed discrepancy. In spite of these uncertainties and, since we measure more flux than the rms of the single-dish data, we conclude that flux losses are improbable.

Our data (Figures 3.10 and 4) show a relatively broad ¹²CO (J = 1–0) profile centered at V_LSR = +50 km s⁻¹ and affected by ISM contamination in several channels (e.g., at V_LSR = +37.7 and +50.7 km s⁻¹). The velocity-integrated CO map is spatially unresolved, however, the emission from the individual channels is centered at slightly different offsets that lie within a ∼25-long region oriented along P.A. ∼ +45°, i.e., well aligned with the optical lobes. Therefore, the CO nebula probed by our maps is marginally extended, with a value of θ_1/2 ≈ 1''. The velocity gradient found (see Figure 3.10, bottom-right panel) is consistent with the presence of a molecular bipolar outflow.

Here we revise our value for the distance to IRAS 19024+004 as estimated in Sahai et al. (2005). We adopt d = 10 kpc, which is the far d_k (l = 3521, b = −0265, and V_LSR = +50 km s⁻¹). The bolometric flux is F_bol = 260 L_☉ kpc⁻², which would imply a very low luminosity of L ∼ 3500 L_☉ for the near d_k = 3.7 kpc. For the far d_k the luminosity, L ∼ 2.6 × 10⁴ L_☉, is at the high end of the typical range for pAGB objects and in excellent agreement with the independent estimate of the luminosity obtained from the strength of the luminosity-dependent O i 7771–5 Å IR triplet (Sánchez Contreras et al. 2008). The compactness of the optical and CO nebula also supports a large distance to this object.⁸ At a distance of ∼10 kpc, a value of A_V ∼ 2.2 mag is derived using the algorithm by Hakkila et al. (1997), which leads to a dereddened F_bol = 270 L_☉ and a total luminosity of L ∼ 2.7 × 10⁴ L_☉.

Ground-based optical imaging and spectroscopy indicates a PN nature for this IRAS source (Kerber et al. 1996). HST optical images show a roughly elliptical, limb-brightened nebula of angular dimensions ∼11'' × 8'' with its long axis oriented at P.A. ∼ 15° (Sahai et al. 2011b). A faint point-like source is observed at the nebula center, which is very likely the central star. Radio emission maps and IR images from the Spitzer/GLIMPSE survey (Urquhart et al. 2009) show a nebular morphology similar to that in the optical.

Our maps show several CO emission clumps in the velocity range V_LSR = [+51:+66] km s⁻¹ (Figure 3.11). None of the clumps coincide with the location of the central star of IRAS 19234+1627 or the 2.6 mm continuum source and, therefore, the detected CO emission is likely to have an interstellar origin.

Intense continuum emission at 2.6 mm is detected, F_2.6 mm = 70 mJy. The continuum source is extended and has a deconvolved half-maximum size of ∼55 × 47 (Table 5). The orientation of the 2.6 mm continuum envelope is the same as that of the optical, IR, and radio nebulae. The origin of the millimeter-wave continuum emission is probably free–free emission from the ionized nebula.

Since the V^LSR_sys of IRAS 19234+1627 is unknown, we are unable to estimate d_k. From the bolometric flux of the source, F_bol = 60 L_☉ kpc⁻², we derive a large value of the luminosity distance of d_L = 10 kpc. At a distance of ∼10 kpc, the visual ISM extinction is A_V < 6.5 mag, which leads to a dereddened F_bol < 70 L_☉ and a lower limit to the distance of d_L > 9.2 kpc. Here we adopt d = 9.5 kpc.

This is a very young O-rich PN characterized by a bipolar radio continuum emission morphology (Aaquist 1993; Miranda et al. 2001). The characteristic S-shape morphology of the radio lobes can be successfully reproduced by a precessing jet, evolving in a dense circumstellar medium (Velázquez et al. 2007). Our optical/HST and NIR/AO images show bipolar nebula with a well-defined point-symmetric structure (Sahai et al. 2007a; Sánchez Contreras et al. 2006b). Ground-based optical images and R−I color map are consistent with the presence of a dense structure in the equatorial plane of the nebula (Miranda et al. 2000). Both OH and H₂O maser emission are observed, suggesting that IRAS 19255+2123 departed from the PPN phase only a few decades ago (Miranda et al. 2001; Tafoya et al. 2007). The H₂O maser emission is found in the equatorial regions and at the tips of the lobes (Miranda et al. 2001); SiO maser emission is not detected (Jewell et al. 1991). From the presence of He ii emission in the optical spectrum, a temperature of T_eff ⩾ 60,000 K has been inferred for the central star (de Gregorio-Monsalvo et al. 2004; Miranda et al. 2000).

Single-dish ¹²CO (J = 2–1) observations have been reported by Tafoya et al. (2007), who detected an intense, narrow ISM feature at V_LSR ∼ +10 km s⁻¹ plus a weak, broad component observed over a velocity range of V_LSR ∼ [0:+40] km s⁻¹ that is most likely associated with IRAS 19255+2123. A tentative (3σ) detection of circumstellar ¹²CO (J = 1–0) emission is also reported by these authors.

Our data confirm the presence of circumstellar ¹²CO (J = 1–0) emission from this object with a relatively broad profile, over the velocity range V_LSR = [+10.9:+47.3] km s⁻¹ (Figures 3.12 and 4). At V_LSR = +10.9 km s⁻¹, we identify a ¹²CO emission clump east of IRAS 19255+2123 that is most likely the main contributor to the narrow ISM emission component found at this velocity in previous single-dish spectra. This ISM feature is less intense in our interferometric data than in single-dish spectra, suggesting that the extended emission from the ISM cloud is partially filtered out by the interferometer. The flux of the broad, circumstellar component of the CO profile measured by us and by Tafoya et al. (2007) from their single-dish spectra is in agreement, which implies that there are no interferometric flux losses of the circumstellar CO emission in our data.

Our velocity-integrated CO map shows an extended source with a deconvolved size of 84 × 39 and with its long axis roughly perpendicular to the optical lobes.⁹ The redshifted (blueshifted) emission arises from the west (east) side of the equatorially extended structure: The absolute angular offset between these two emission components, both also elongated along the EW direction, is ΔR.A. ∼ 2''. The observed elongation and velocity gradient may point to elongated outflows running along a direction perpendicular to the long axis of the radio and optical lobes. The presence of multi-directed and nearly orthogonal molecular flows are observed in other PPNs (e.g., CRL 2688; Cox et al. 2000). A large, rotating equatorial structure can be ruled out as an alternative interpretation of the velocity gradient observed in IRAS 19255+2123 since in that case the inferred mass for the central star would be unrealistically high (over ≈100 M_☉).

We detect for the first time ¹³CO (J = 1–0) emission from IRAS 19255+2123 (Figure 5). Our maps show a point-like source centered near the redshifted ¹²CO (J = 1–0) clump. The ¹³CO line peaks around V_LSR = +30km s⁻¹, which is consistent with a double-peaked profile with the blue side of the line weaker than the red one. The full width of the line is uncertain due to limited S/N in our maps and ISM contamination. We believe circumstellar emission is confirmed in the LSR velocity range [+12.6:+34.4] km s⁻¹.

Intense 2.6 mm continuum emission is detected from a point-like region near the blueshifted ¹²CO (J = 1–0) emission clump (see Figure 3.12). The origin of the millimeter-wave continuum emission is probably free–free emission from a compact, ionized structure.

We are intrigued by the separation (|ΔR.A.| ∼ 2'', equivalent to ∼8000 AU at d = 4 kpc, see the next paragraph) of the redshifted ¹²CO and ¹³CO (J = 1–0) emission clumps relative to the continuum source; the latter is well aligned with the blueshifted ¹²CO (J = 1–0) clump and the center of the optical nebula. It cannot be ruled out that this separation is pointing to a different nearby CO-emission source (not necessarily physically associated with IRAS 19255+2123) with slightly different radial velocity. Alternatively, if a bipolar flow is the cause of the elongated CO red and blue clumps (confirmation by higher-angular resolution is needed) then the observed distribution would imply that such an outflow is not symmetric with respect to the central source, i.e., the blue and red clumps may represent two distinct bipolar flows emerging from the nebular core with a different inclination with respect to the line of sight. In particular, the blue lobe may be directed nearly along the line of sight, resulting in a very small lobe length projected in the sky plane, whereas the red lobe may run along a direction closer to the plane of the sky.

There are several distance estimates for this object based on statistical methods with values of d in the range ∼4–6.6 kpc (Cahn et al. 1992; Zhang 1995; Phillips 2004). Taking into account the galactic coordinates and radial velocity of IRAS 19255+2123 (l = 561, b = +021, and $\mbox{$V_{\rm sys}^{\rm LSR}$}=+23$ km s⁻¹) we derive values for the near and far d_k of 1.8 and 7.5 kpc, respectively. (Our values differ from those obtained by Aaquist (1993), d_k = 0.6 and 8.9 kpc, because of the different V^LSR_sys adopted by these authors, V^LSR_sys = +10 km s⁻¹.) The bolometric flux of IRAS 19255+2123 is F_bol = 350 L_☉ kpc⁻², which would imply an unrealistically low luminosity for the near d_k. The total visual ISM extinction across the Galaxy toward this object is A_V = 7 mag. In particular, at d ∼ 4–6 kpc a smaller value of A_V ∼ 1.5–2 mag is expected. This leads to an intrinsic (dereddened) bolometric flux of F_bol ∼ 360 ≲ 410 L_☉ kpc⁻², implying d_L ∼ 4 kpc, a value that we adopt for d. This value is consistent with a recent trigonometric parallax measurement (of H₂O masers) by Tafoya et al. (2011).

This object is a young O-rich PPN imaged in our optical/HST and NIR/AO surveys of candidate PPNs (Sahai et al. 2007a; Sánchez Contreras et al. 2006b). These images show a bipolar nebula (lobe length ∼09) surrounded by an elliptical halo that can be traced out to a radius of 25. The central star is probably hot (B-type?); no Hα or other emission lines have been detected so far (Sánchez Contreras et al. 2008). The presence of molecular gas is evidenced by OH maser emission from a spherically symmetric shell with radius ∼05 expanding at V_exp ∼ 13.5 km s⁻¹ (Chapman et al. 1984; Te Lintel Hekkert et al. 1989; Engels & Jiménez-Esteban 2007).

We report a first-time detection of ¹²CO (J = 1–0) emission toward this object. Our data (Figures 3.13 and 4) show CO emission over a total velocity range similar to that spanned by the OH maser line (∼[−4:+25] km s⁻¹ LSR). There are strong ISM absorption/emission features near IRAS 19292+1806 over the range V_LSR = [+45:+65] km s⁻¹ and also around V_LSR = −6.5 km s⁻¹. Our velocity-integrated CO map shows an extended source with a deconvolved half-maximum size of 24 × 17 and oriented along P.A. = −45°, that is, along the direction of the optical lobes. We also find ¹³CO (J = 1–0) emission toward IRAS 19292+1806 with a relatively broad profile, FWZI ∼ 24 km s⁻¹ centered at V_LSR ∼ 12 km s⁻¹ (Figure 5). The velocity-integrated ¹³CO map shows an extended source with a deconvolved half-maximum size of ∼4'' × 2'' and oriented similarly to the ¹²CO envelope. In this case, both the ¹²CO and ¹³CO envelopes are marginally resolved, therefore, the derived values for the envelope size and orientation are uncertain.

The near and far kinematic distance to IRAS 19292+1806 are d_k = 0.8 and d_k = 9 kpc, respectively (l = 536, b = −024, V^LSR_sys = 11 km s⁻¹). The bolometric flux is F_bol = 260 L_☉ kpc⁻², which would imply an unrealistically low luminosity for the near d_k. The luminosity distance is d_L = 4.8 kpc. At such a distance, the visual ISM extinction toward this object is A_V ∼ 2.8 mag and, therefore, the intrinsic (dereddened) bolometric flux is F_bol ∼ 280 L_☉ kpc⁻², implying d_L ∼ 4.6 kpc. We adopt d = 5 kpc in this work, although a larger distance of d_k(far)∼9 kpc is possible, yielding an acceptable value for the total luminosity of ≈2 × 10⁴ L_☉.

This is a yPN candidate (e.g., Hrivnak et al. 2000; Lowe & Gledhill 2007, and references therein). Optical HST images show a bipolar nebula with roughly cylindrical ∼2''-long lobes surrounded by a tenuous round halo that can be traced out to a radius of ∼4'' (Sahai et al. 2007a). The central star, visible through the nebula, has a B0-1 I spectral type (Sánchez Contreras et al. 2008). Hα emission with a P Cygni profile and broad (FWZI ∼ 2600 km s⁻¹) wings is detected toward the nebula core, indicative of ongoing pAGB mass loss (Sánchez Contreras et al. 2008). This object has a mixed C+O chemistry since it shows both carbon- and oxygen-rich dust features (Hrivnak et al. 2000; Hodge et al. 2004; Cerrigone et al. 2009). Emission from several NIR transitions of H₂ indicate the presence of a ∼2''-sized molecular component (Kelly & Hrivnak 2005). No OH and H₂O maser emission has been detected (Likkel et al. 1991; Suárez et al. 2007).

¹²CO (J = 1–0) emission was unsuccessfully searched for by Likkel et al. (1991). We have detected ¹²CO (J = 1–0) emission toward IRAS 19306+1407 for the first time. The line is clearly visible in two adjacent channels at V_LSR = 96.5 and 99.1 km s⁻¹ (Figures 3.14 and 4). The good match of the systemic velocity derived from CO, V_LSR ∼ +98 km s⁻¹, and from the optical line absorption spectrum of the star, V^LSR_sys = 95–100 km s⁻¹ (Sánchez Contreras et al. 2008), favors a circumstellar origin for the CO emission in spite of its narrow profile. We observe weak, broader emission wings (at a 3σ level) from the source position over the LSR velocity range [+91.3:+106.9] km s⁻¹. The observed line peak intensity is comparable to the rms noise level in the single-dish data by Likkel et al. (1991), implying that there are no flux losses in our interferometric data. There is also weak emission toward the source at much larger velocities, V_LSR = [+190:+203] km s⁻¹. At these velocities there are several emission clumps in the field of view that probably represent dense ISM clouds, so it is likely that this central emission component is interstellar.

The velocity-integrated ¹²CO (J = 1–0) map is consistent with a point-like source, however, careful examination of the individual velocity-channel maps suggests a velocity gradient along the NS direction, with redshifted (blueshifted) emission arising from the north (south) side of the envelope. The good alignment of the axis along which the velocity gradient is observed with the optical/NIR lobes suggests the presence of a bipolar outflow, but we caution that this is a tentative result given the resolution of our data compared with the angular dimensions of the optical nebula (see above). ¹³CO (J = 1–0) emission is marginally detected toward IRAS 19306+1407 at V_LSR = +107 km s⁻¹.

The observed radial velocity of IRAS 19306+1407 (V^LSR_sys = +98 km s⁻¹) is close to the maximum value possible in that direction (l = 5030, b = −0248), so this source is probably at the tangent point, i.e., at d_k = 5.5 kpc, a value which we adopt for d. The bolometric flux is F_bol = 300 L_☉ kpc⁻². Adopting a visual ISM extinction of A_V = 4.2 mag, estimated by Lowe & Gledhill (2007), the dereddened bolometric flux is F_bol = 380 L_☉ kpc⁻², which yields a total luminosity of ≈10⁴ L_☉ at 5.5 kpc. This value lies within the range of expected luminosities for pAGB objects and is in excellent agreement with the independent estimate of L obtained from the strength of the O i 7771–5 Å IR triplet (Sánchez Contreras et al. 2008).

This is a strong IRAS source originally proposed to be a PPN candidate by Kwok et al. (1987). Optical HST images show a ∼3'' × 2'' bipolar nebula, which corroborates its pAGB status (Ueta et al. 2000, also see our Figure 3.15). In these images, we also find an elliptical extended halo (R_halo ∼ 3'') with a brightness distribution suggestive of a mass-loss rate increasing with time. Recent subarcsecond-resolution 8–12 μm images are consistent with an elliptical detached shell along P.A. ∼ 11°, i.e., roughly aligned with the optical/NIR lobes (Lagadec et al. 2011). We have recently discovered Hα emission with ≳600 km s⁻¹ broad wings from the stellar vicinity, signaling ongoing fast pAGB ejections (Sánchez Contreras et al. 2008). The line absorption spectrum of the central star is consistent with a B 3-6I classification (Sánchez Contreras et al. 2008). The chemistry of the envelope is not certain, but there is some evidence for the presence of the 9.7 μm silicate feature in absorption and, therefore, for an O-rich chemistry (Lawrence et al. 1990). Near IR H₂ line emission has been reported showing the presence of shocked molecular gas (Kelly & Hrivnak 2005).

The previous search for ¹²CO (J = 2–1) emission toward this object by Hu et al. (1994) resulted in no detection. ¹²CO (J = 1–0) emission is detected for the first time in this work. Our maps reveal a very broad, triangular-shaped line with a full extent of ≳160 km s⁻¹ centered at $\mbox{$V_{\rm sys}^{\rm LSR}$}\sim -37$ km s⁻¹ (Figures 3.15 and 4). Recently, one of us (C.S.C.) has obtained a ¹²CO (J = 2–1) single-dish spectrum (that will be published elsewhere) showing a similar wing-dominated profile with an even larger full velocity extent (FWZI ∼ 300 km s⁻¹). The absence of a prominent narrow core both in the interferometric and single-dish profiles (otherwise common in PPNs—see Section 1) rules out interferometric flux losses of a presumably extended, slow nebular component as the cause of the atypical wing-dominated profile. The CO envelope appears to be point-like in our velocity-integrated ¹²CO maps (with an HPBW = 98 × 7''). ¹³CO (J = 1–0) emission is tentatively detected (at a 3σ level) toward the CO-emission source around V_LSR = −31 km s⁻¹ (Figure 5).

We adopt a distance to this object of d = d_k = 11 kpc (for l = 5970, b = +096, and V^LSR_sys = −37 km s⁻¹). The bolometric flux is F_bol = 650 L_☉ kpc⁻², which would yield d_L = 3 kpc. At ∼11 kpc, the total luminosity of this source, L ∼ 7.8 × 10⁴ L_☉, is consistent with the strength of the O i 7771–5 Å IR triplet in this source, which independently suggest a luminosity of up to L ≈ 10⁵ L_☉ (Sánchez Contreras et al. 2008). The ISM extinction is expected to be high in the direction to IRAS 19374+2359. Adopting a visual ISM extinction of A_V = 3.5 ± 1.5 mag, the dereddened bolometric flux is F_bol = 740 ± 40 L_☉ kpc⁻², which at 11 kpc yields a total luminosity of ∼9 × 10⁴ L_☉.

This is an O-rich PPN that has been extensively studied by us using multiwavelength data. Optical HST images reveal a quadrupolar nebula of size about ∼10'' × 5'', with two pairs of elongated lobes emanating from the center; a faint, surface-brightness-limited, diffuse halo surrounds the lobes (Sahai et al. 2007a, 2007b). The central star, which is directly visible through the nebula, has a line absorption spectrum consistent with an F 3Ib classification (Klochkova et al. 2002; Sánchez Contreras et al. 2008). The Hα absorption stellar profile is partially filled by Hα emission indicative of current pAGB mass loss (Sánchez Contreras et al. 2008). Emission features due to crystalline silicates are identified in the Infrared Space Observatory Short Wavelength Spectrometer spectra of the star, which support an O-rich chemistry (Sarkar & Sahai 2006).

Extensive study of ¹²CO (J = 2–1) emission with 2'' resolution was presented by Sánchez Contreras et al. (2006a). These data probe two distinct molecular components: a slowly expanding, 44 × 44-sized shell and a fast bipolar outflow (∼75 long) oriented along the longer and more tenuous pair of the optical lobes (P.A. ∼ 80°). Our ¹²CO (J = 1–0) maps (Figure 3.16) are consistent with the nebular morphology and kinematics deduced from the higher-angular resolution J = 2–1 maps: The bulk of the emission arises in an extended shell expanding at V_exp ∼ 15 km s⁻¹, which is probably the remnant AGB CSE; weaker emission is seen from a fast bipolar outflow, which presents a clear velocity gradient: the ¹²CO emission from the east (west) lobe is redshifted (blueshifted) with respect to V^LSR_sys. This nebular morphology and kinematics are also confirmed by the more recent ∼1''-resolution CO maps by Castro-Carrizo et al. (2010). By comparing the ¹²CO (J = 1–0) fluxes obtained from OPACOS and from single-dish data (Sánchez Contreras et al. 2006a) we conclude that no significant flux losses affect our interferometric maps.

A compact source of continuum emission at 2.6 mm is detected. The observed flux, ∼3.8 mJy, is in excess of the value expected from the spectral slope of the SED in the far IR range. As concluded from the detailed analysis/modeling of the SED of this object, the 2.6 mm continuum flux together with previous submillimeter-continuum flux measurements imply a substantial mass of large (radius ≳1 mm), cold (∼30 K) dust grains (Sarkar & Sahai 2006; Sahai et al. 2007b). We note that Castro-Carrizo et al. (2010) do not detect continuum emission at 2.6 mm in their ∼1''-resolution interferometric maps (with an rms limit of 0.4 mJy). Such a discrepancy with our data suggests that the dust continuum emission is extended (possibly associated with the optical diffuse halo) and is partially filtered out in Castro-Carrizo et al.'s interferometric data.

Here we revise our value for the distance to IRAS 19475+ 3119 previously estimated in Sánchez Contreras et al. (2006a). Taking into account its radial velocity (V^LSR_sys = +18 km s⁻¹) and galactic coordinates (l = 6716, b = +273), this source is probably located at the tangent point and, therefore, d_k = 3.5 kpc. At this distance, the visual ISM extinction is A_V ∼ 1.3 mag, which leads to a dereddened F_bol = 660 L_☉ kpc⁻² and, therefore, to a value for d_L of 3.0 kpc. We adopt d = 3.5 kpc.

This is a PPN candidate with a C-rich envelope around an O-rich M6  giant (e.g., Bujarrabal et al. 1994; Groenewegen et al. 1996; Speck et al. 2009, and references therein). HST optical images show a point-like source coincident with the MSX6C G067.3478+01.0186 and 2MASS J19564844+3044026 sources.

Single-dish ¹²CO (J = 1–0) and J = 2–1 spectra of this source have been previously reported (Likkel et al. 1991; Bujarrabal et al. 2001). We find ¹²CO (J = 1–0) emission over a total velocity range of ∼44 km s⁻¹ centered at V_LSR = +5.5 km s⁻¹ (Figures 3.17 and 4). The blue side of the circumstellar profile is affected by strong ISM absorption, at V_LSR = −20 km s⁻¹. The total ¹²CO (J = 1–0) flux measured by us is consistent with that obtained from single-dish spectra (references above) and interferometric data (Neri et al. 1998; Castro-Carrizo et al. 2010), and, therefore, no flux losses affect our maps.

Our velocity-integrated ¹²CO map shows an extended structure with a deconvolved half-maximum size of 63 × 48, comparable to that derived by Neri et al. (1998), elongated roughly along P.A. = −11°. The individual velocity-channel maps reveal a velocity gradient nearly along the same direction: The intensity peaks lie on a ∼15-long vector roughly oriented along P.A. ∼ −20° with the blue and redshifted emission arising from the south and north side of the nebula, respectively. Given the large S/N and relatively small beam in our maps, the offset error bars are expected to be ≲01. This velocity gradient suggests the presence of a compact bipolar flow at the center of the source carving out the interior of the outer, more extended envelope. Castro-Carrizo et al. (2010) find that the emission centroids for many of the velocity channels in their interferometric CO maps are displaced by a few fractions of an arcsecond toward the NE, however, they do not make a conclusion about the nature of this displacement. ¹³CO (J = 1–0) emission is observed in our data with a flat-topped profile and similar velocity extent to the ¹²CO line (Figure 5). Our velocity-integrated ¹³CO map also shows an extended source with size and orientation similar to that derived from ¹²CO within errors.

Continuum emission at 2.6 mm is detected toward the source with flux ∼6 mJy. Our map suggests an extended structure with a deconvolved half-maximum size of ∼18 × 13 and with its long axis roughly perpendicular to the CO envelope. The 2.6 mm flux is in excess of that expected by extrapolating the spectral slope of the SED in the far IR (∼60–140 μm), which is mainly produced by thermal emission by a ∼150 K dust component. An additional population of cooler, big grains could account for the observed mm flux excess in IRAS 19548+3035, as happens for other PPNs (Sahai et al. 2006; Sánchez Contreras et al. 2008). The morphology and orientation of the continuum-emitting region suggests that this cold dust component (or part of it) may be located in a thick, dusty disk or waist perpendicular to the molecular bipolar flow. The 2.6 mm continuum flux measured by Castro-Carrizo et al. (2010) in their ∼1''-resolution maps is <1.3 mJy, that is, smaller than that measured by us with a larger beam, which suggests that part of the dust emission arises in an extended component partially filtered out in Castro-Carrizo et al.'s data.

For IRAS 19548+3035, we derive values for the near and kinematic distance of d_k = 0.5 and 6 kpc (l = 06735, b = +0102, V^LSR_sys = +5.5 km s⁻¹). The bolometric flux is F_bol = 1800 L_☉ kpc⁻², which would imply an unrealistically low luminosity of L ∼ 500 L_☉ for the near d_k. For the far d_k the luminosity, L ∼ 6.5 × 10⁴ L_☉, is at the high end of the expected range for pAGB objects. We derive a value of d_L = 1.8 kpc. The ISM extinction in the direction to this object is A_V < 8.5 mag and, therefore, the intrinsic (dereddened) luminosity is ≲3000 L_☉ kpc⁻², implying a lower limit to the distance of d_L > 1.4 kpc. Here we adopt d = 4 kpc, which is intermediate to d_L and d_k(far). At such a distance, A_V ∼ 2.1 mag and the dereddened F_bol ∼ 2070 L_☉ kpc⁻², implying a total luminosity ∼3 × 10⁴ L_☉.

This is an optically thick OH/IR envelope around a central object with a controversial evolutionary status: It has been suggested to be a hypergiant/supergiant star (Lewis 2001) and an evolved post-main-sequence star located near the tip of the AGB (Zuckerman & Lo 1987; Likkel et al. 1991). The optical and NIR counterparts are rather faint (R = 20.23 mag and 2MASS K = 14.3 mag); no HST optical image is available. The presence of a very deep 9.7 μm silicate feature in the IR spectrum as well as OH and H₂O maser emission indicates O-rich chemistry (e.g., Galt et al. 1989; Likkel 1989). Surprisingly, a number of far IR features have been identified recently indicative of a C-rich chemistry, including HCN, C₂H₂, and C₆H₆ transitions (García-Hernández et al. 2012, in preparation). Based on this, we classify IRAS 19566+3423 as an object with a mixed C+O chemistry. The OH maser has a peculiar twin-double-peaked profile suggestive of an expanding concentric shell structure (Galt et al. 1989). Unlike most other OH maser sources, the OH lines in IRAS 19566+3423 show strong circular polarization and time variability. The OH emission velocity range is also variable, reaching values as high as >70 km s⁻¹ (Lewis 2001).

We report the first detection of ¹²CO (J = 1–0) emission toward this peculiar object, coincident with the bright MSX6C G070.7766+02.6836 and 2MASS J19583227+3431337 sources. Our data (Figures 3.18 and 4) show ¹²CO emission centered at V_LSR = −40 km s⁻¹ and with a relatively narrow core (FWHM = 13 km s⁻¹) plus tentative broad wings. By smoothing our maps every three channels, the wings can be traced over a total velocity range V_LSR = [−21:−68] km s⁻¹ (i.e., FWZI = 47 km s⁻¹). The CO-emitting region is spatially unresolved in our maps. Likkel et al. (1991) reported a ¹²CO (J = 1–0) non-detection toward this source within the noise of their single-dish spectrum, rms ∼0.6 Jy. Since the observed line peak intensity in our interferometric data, ∼0.27 Jy, is smaller than the rms noise in the single-dish spectrum, we cannot evaluate possible interferometric flux losses. Given the probable large distance to IRAS 19566+3423 (see below), significant flux losses of an extended nebular component are unlikely to affect our maps.

For IRAS 19566+3423, we estimate a kinematic distance of d_k ∼ 9–10 kpc for a radial velocity of V^LSR_sys = −40 km s⁻¹ and galactic coordinates (l,b) = (7078, +0268). The bolometric flux is F_bol = 800 L_☉ kpc⁻², which would imply a high, but acceptable, luminosity of L ∼ [6.5–8] × 10⁴ L_☉ at 9–10 kpc. Adopting L ∼ 6000 L_☉, we derive a value of d_L = 2.7 kpc. The ISM extinction in the direction to this object is A_V < 4.9 mag and, therefore, the intrinsic (dereddened) bolometric flux is F_bol < 930 L_☉ kpc⁻², implying a lower limit to the distance of d_L > 2.5 kpc. At d ∼ 9 kpc (adopted here), the visual ISM extinction is A_V ∼ 2.2 mag and the dereddened F_bol and luminosity are ∼850 L_☉ kpc⁻² and 7 × 10⁴ L_☉, respectively.

This object is a C-rich PPN (e.g., Hrivnak et al. 2000, 2010; Gledhill et al. 2002, and references therein). HST images show an elongated (elliptical or bipolar) nebula with an angular size of ∼2'' × 1'' and oriented along P.A. = −75° (Sahai et al. 2007a). This compact nebula is surrounded by a larger, fainter round halo (traced out to a radius of 5'') with a non-uniform radial brightness distribution suggestive of two distinct episodes of time-variable mass loss. The central star, which is variable with a pulsational period of ∼150 days, has a mean spectral type G8 Ia (Hrivnak et al. 2010; Klochkova & Kipper 2006). NIR imaging polarimetry by Gledhill et al. (2001) suggests that multiple scattering is occurring in the core region, and is consistent with a geometry in which the star is surrounded by an obscuring disc or torus. Mid-IR images with ∼1''-resolution show a compact marginally resolved dust component (Meixner et al. 1999).

Circumstellar ¹²CO (J = 2–1) emission has been found by Likkel et al. (1991), who also report a tentative (∼1.4σ) detection of the J = 1–0 transition with the 30 m IRAM telescope. Both CO single-dish profiles appear strongly affected by ISM absorption at the line center, at V_LSR ∼ 13–14 km s⁻¹. Our data show intense ¹²CO (J = 1–0) emission centered at V^LSR_sys = +12.4 km s⁻¹ over a full velocity range of ∼28 km s⁻¹; ISM contamination is present in our velocity-channel map at V_LSR = 14.3 km s⁻¹ (Figures 3.19 and 4). The line profile observed is consistent with the presence of weak-broad emission wings adjacent to the intense-narrow core, which suggests two distinct kinematic components (see Section 4.2.1).

The circumstellar CO emission arises in an extended region with a deconvolved, half-maximum size 37 × 11 and oriented along P.A. ∼ 11°. This is consistent with the presence of an equatorial torus-like structure roughly perpendicular to the optical lobes. We observe a velocity gradient approximately along the EW direction, with the blue- and red-wing emission arising at the west and east side of the nebula, respectively. In principle, the gradient observed could be due to (1) expansive motions in a tilted torus and (2) a bipolar molecular outflow.

We believe (2) is the most likely explanation in this source because: (1) The angular distance between the redshifted and blueshifted CO emission clumps (∼2'') is similar to the distance between the tips of the east and west optical lobes but larger than the minor axis of the torus possibly identified (θ_b ∼ 11). Moreover, (2) if the gradient is due to the bipolar outflow then the implied tilt is consistent with that inferred from the different brightness and colors of the optical lobes: The east lobe is fainter and redder and, therefore, more likely to be located behind the front side of the torus. (This is making the reasonable assumption that the lobes run along the symmetry axis of the torus.) Higher-angular resolution maps are needed to robustly identify and characterize the different molecular components (torus and bipolar outflow) tentatively present in this object.

There is a significant difference between the value of V^LSR_sys derived from the CO profile and that deduced from the metallic absorption line spectrum of the central star, which is V^LSR_sys(star) = 3.1 ± 2 km s⁻¹ (Klochkova & Kipper 2006). This discrepancy could indicate the presence of a binary system at the core of the nebula, with V^LSR_sys(CO) = +12.4 km s⁻¹ representing the velocity of the mass center, and V^LSR_sys(CO)−V^LSR_sys(star) ∼9 km s⁻¹ being a lower limit to the amplitude of the radial velocity curve of the primary star.

A compact 2.6 mm continuum source is detected toward this object with a total flux of 6 mJy. Continuum emission at 850 μm and 1.2 mm has been previously reported with a spectral slope consistent with thermal dust emission in that range (Gledhill et al. 2002; Buemi et al. 2007). Our measurement of the 2.6 mm flux is in clear excess with respect to the expected emission from such a dust component. This excess could be due to an additional contribution by free–free emission from a dense, compact ionized structure. Since the central star of IRAS 20000+3239 has a low temperature, T_eff ∼ 5000 K, and therefore is unable to produce a significant ionizing UV radiation, the 2.6 mm flux excess (if attributable to free–free emission) could point to a binary system with a hot companion and/or an ionized accretion disk. Alternatively, the 2.6 mm continuum excess could be due to cold dust; in this case, the inferred temperature of the grains would have to be rather low (probably as low as 4–5 K) to explain the observed flux. Detailed modeling of the SED as well as continuum flux measurements in the cm-wavelength range are need to obtain a firm conclusion on the nature of the 2.6 mm continuum emission in this PPN.

Taking into account the galactic coordinates and radial velocity (from CO) of IRAS 20000+3239 (l = 6968, b = +116, and V^LSR_sys = +12.4 km s⁻¹), we derive values for the near and far kinematic distance $\mbox{$d_{\rm k}$}$ = 2.3 and 3.6 kpc, respectively. The bolometric flux is F_bol = 630 L_☉ kpc⁻², which implies d_L = 3.1 kpc. At a distance of ∼3 kpc, The ISM extinction toward IRAS 20000+3239 is estimated to be A_V = 2.5 mag (Hrivnak et al. 2000), which yields an intrinsic (dereddened) F_bol ∼900 L_☉ kpc⁻². Here we adopt an intermediate value of d = 3 kpc, which implies a total luminosity of L = 8100 L_☉.

This object is an O-rich PPN with an F4-7 I central star that has been studied in detail by us using multiwavelength data (Sahai et al. 2003, 2006, 2007a). HST images show a bipolar nebula, an obscured central waist, and knotty linear structures along the lobes. There is also a round halo traced up to 65 from the center with a non-uniform brightness distribution suggestive of two distinct mass-loss episodes. A bright, unresolved (PSF = 133) 12.5 μm counterpart is reported by Meixner et al. (1999). This source shows OH maser emission (Zijlstra et al. 2001; Sahai et al. 2003), however, SiO or H₂O masers have not been detected (Nyman et al. 1998; Suárez et al. 2007). Intense Hα emission, with a remarkable P Cygni profile and broad wings (FWZI ∼ 2500 km s⁻¹), is found, which is indicative of shock-excited/accelerated lobes and nuclear, ongoing pAGB ejections (Sánchez Contreras et al. 2008).

Our first detection of ¹²CO (J = 1–0) emission in this object as part of the OPACOS project was reported by us in Sánchez Contreras & Sahai (2004b). Follow-up interferometric ¹²CO (J = 3–2) observations with ∼1''-resolution were presented and analyzed by Sahai et al. (2006) together with ¹²CO and ¹³CO (J = 1–0) data subsets from OPACOS (with ∼ 10''-resolution). A significant fraction of the CO emission from IRAS 22036+5306 is found to arise in a fast bipolar flow co-linear with the optical lobes (Figure 3.20). The width of the broad line wings that arise in the bipolar flow is larger in the ¹²CO (J = 3–2) data (FWZI ∼ 400 km s⁻¹) than in our discovery OPACOS data set (FWZI ∼ 120 km s⁻¹) due to the smaller spectral coverage in the latter. There is a clear velocity gradient along the bipolar outflow, the blueshifted emission arises in the bright (SW) lobe, whereas the redshifted emission arises in the faint (NE) lobe. The low-velocity emission arises in the central parts of the nebula. Our ¹³CO (J = 1–0) maps show a partially resolved structure with similar size and orientation as that probed by ¹²CO (Table 5). The total width of the ¹³CO (J = 1–0) line is smaller probably due to lower S/N in these maps (Figure 5).

A compact 2.6 mm continuum source is detected toward IRAS 22036+5306. As shown by Sahai et al. (2006), the relatively large flux observed at mm wavelengths (in excess of that expected extrapolating the fIR SED) indicates big dust grains (with radius ≳ 1 mm).

The distance to this object is very uncertain. Here we revise our value for the distance (d ∼ 2 kpc) previously adopted in Sahai et al. (2003, 2006), which was guided by keeping our determination of the total mass from the SED fitting below 10 M_☉; this mass is based on an assumed gas-to-dust mass ratio of 200, however, we cannot rule out a significantly lower value of the latter. Taking into account its galactic coordinates and radial velocity (l = 9963, b = −184, and V^LSR_sys = −42 km s⁻¹), we derive d_k = 5 kpc. The bolometric flux is F_bol ∼ 530 L_☉ kpc⁻², which implies d_L = 3.4 kpc. The visual ISM extinction in the direction to this object is A_V < 3.4 mag and, therefore, the intrinsic (dereddened) bolometric flux is F_bol < 630 L_☉ kpc⁻², implying a lower limit to the distance of d_L > 3.1 kpc. We adopt an intermediate value of d ∼ 4 kpc, which yields an intrinsic luminosity of ∼9000 L_☉. This relatively high value, in the upper end of the typical luminosity range for PPNs, is consistent with the central star being massive (≳3–4 M_☉), which is indeed independently suggested by the low ¹²C/¹³C isotopic ratio measured in this object (see Section 4.3 and Sahai et al. 2006).

This is a highly evolved O-rich AGB star with a dusty, optically thick envelope (e.g., Omont et al. 1993; Riechers et al. 2005). This object appears to be a bright point-like source in our HST survey of candidate PPNs (Sahai et al. 2007a). The central star has a spectrum consistent with an effective temperature of ∼2500 K, i.e., with an M-type spectral classification (Riechers et al. 2005, and references therein). SiO, OH, and H₂O maser emission has been detected (Omont et al. 1993, and references therein).

Different CO emission lines in the millimeter and submillimeter range have been detected toward this object (Heske et al. 1990; Neri et al. 1998; Groenewegen et al. 1999; Kemper et al. 2003; Castro-Carrizo et al. 2010). Our ¹²CO (J = 1–0) data (Figures 3.21 and 4) show emission centered at V^LSR_sys = −27 km s⁻¹ over a total velocity range, ∼40 km s⁻¹, slightly larger than that displayed by the OH maser profile (Omont et al. 1993). Strong ISM absorption and emission features are observed in our ¹²CO maps around V_LSR = −23.7 km s⁻¹ (narrow, intense emission components are also observed at similar velocities in the single-dish spectra; see references above). The ¹²CO (J = 1–0) single-dish flux is consistent with our measurement within errors, which allows us to rule out significant interferometric flux losses. Our velocity-integrated ¹²CO maps show a point-like source coincident with the intense, red MSX and 2MASS sources. The compactness of the CO envelope (<4'') is confirmed by the maps recently obtained by Castro-Carrizo et al. (2010). Weak ¹³CO (J = 1–0) emission is detected toward the source in the range V_LSR = [−13:−40] km s⁻¹ (Figure 5).

We adopt the value of the OH maser phase-lag distance, d = 2.38 ± 0.24 kpc (Herman et al. 1985; van Langevelde et al. 1990). At ∼2.4 kpc, the visual ISM extinction toward this object is A_V = 1.2 mag, which implies a dereddened bolometric flux of F_bol ∼ 2800 L_☉ kpc⁻² and a total intrinsic luminosity of ∼1.6 × 10⁴ L_☉. Taking into account the galactic coordinates and radial velocity of IRAS 22177+5936 (l = 1049, b = +241, and V^LSR_sys = −27 km s⁻¹) we derive d_k = 3.0 kpc. The value of the luminosity distance is d_L ∼ 1.5 kpc.

This is a C-rich PPN with a central semiregular pulsating star (e.g., Klochkova et al. 2010, and references therein). Optical HST images show a central round core with two small lobes protruding on the southern side and one single lobe protruding on the northern side (Sahai et al. 2007a). A prominent halo (radius traced up to 75) surrounds the central aspherical nebula. Faint thin partial shells are seen within the halo. Images at 12.5 and 18 μm show a ∼18 × 18 dusty envelope (Meixner et al. 1999). The central star, which is directly visible in the optical, is classified as F 7-9Ia based on optical spectroscopic detailed studies (e.g., Pereira & Miranda 2007; Suárez et al. 2006; Arkhipova et al. 2003; Decin et al. 1998). The C-rich chemistry of this object is indicated by the presence of 3.3 and 21 μm features and emission from a number of C-bearing molecules (e.g., Omont et al. 1993; Kwok et al. 1995; Decin et al. 1998).

Previous ¹²CO (J = 1–0) and (J = 2–1) single-dish detections have been reported (Likkel et al. 1991). Our data (Figures 3.22 and 4) show strong ¹²CO (J = 1–0) emission centered at V_LSR = −30km s⁻¹ with an intense central core and weak, broader emission wings. The good agreement of the ¹²CO (J = 1–0) line flux measured by us and previous single-dish data rules out interferometric flux losses in our maps. The CO envelope, which is spatially resolved, can be traced out to a radius of ∼5'' and has a deconvolved half-maximum size of ∼31 × 26, with its long axis oriented along the optical lobes (P.A. ∼ 15°). A velocity gradient is found along such a direction, with the redshifted (blueshifted) wing emission arising in the NE (SW) lobe of the nebula, consistent with a bipolar outflow. We report for the first time ¹³CO (J = 1–0) emission toward this object with a flat-topped profile centered around V^LSR_sys = −31 km s⁻¹ (Figure 5).

We adopt a distance to IRAS 22223+4327 of d = 4 kpc. Taking into account its galactic coordinates and radial velocity (l = 9675, b = −1156, and V^LSR_sys = −30 km s⁻¹) we estimate d_k = 4.6 kpc (see also Klochkova et al. 2010). The bolometric flux of IRAS 22223+4327 is F_bol ∼ 370 L_☉ kpc⁻², which yields d_L = 4 kpc. The ISM extinction in the direction to this object is quite low, A_V < 0.4 mag and, therefore, the intrinsic (dereddened) bolometric flux is ∼420 L_☉ kpc⁻², implying d_L ∼ 3.8 kpc.

This is a young, low excitation PN of unknown chemistry around a ∼B 0 central star (García Lario et al. 1991). Our high-angular resolution optical/HST and NIR/AO images (Sahai et al. 2007a; Sánchez Contreras et al. 2006b) show a ∼8'' × 3'' bipolar nebula with highly structured lobes (overall shape is point symmetric); jet-like features and an ansa can be seen at the lobes; a dark waist totally obscures the central star at optical wavelengths.

García Lario et al. (1991) searched for ¹²CO (J = 2–1) emission toward this source but did not detect emission at any of the positions observed with ≳0.1 K (≳3.9 Jy taking into account the beam size 30'' of the 12 m NRAO antenna used by these authors). We report detection of ¹²CO (J = 1–0) emission toward IRAS 22568+6141 for the first time: Weak emission is found over a full velocity extent of ∼55 km s⁻¹ centered at V_LSR = −85 km s⁻¹ (Figures 3.23 and 4). The line core shows a narrow absorption feature around V_LSR = −82 km s⁻¹, which probably has an interstellar origin. Our velocity-integrated ¹²CO map reveals a bipolar flow with a deconvolved half-maximum size of 67 × 12 oriented along P.A. = −65°, that is, along the optical lobes. An overall point symmetry is suggested by our CO maps. We find a clear velocity gradient along the bipolar molecular outflow, the redshifted (blueshifted) emission arising at the SE (NW) lobe.

We detect 2.6 mm continuum emission arising in an extended region with size and orientation similar within errors to that of the CO bipolar flow. Our measurement of the 2.6 mm flux shows a clear excess with respect to the expected emission from the dust component responsible for the FIR fluxes. The similar overall point-symmetric morphology and extent of the optical, CO and 2.6 mm continuum nebulae together with the relatively hot central star (B 0) suggest a significant contribution to the millimeter-continuum flux by free–free emission from the ionized gas component of the nebula.

We derive a kinematic distance of d_k = 9 kpc (l = 11012, b = +193, V^LSR_sys = −85 km s⁻¹). The bolometric flux, F_bol ∼ 250 L_☉ kpc⁻², leads to d_L = 4.9 kpc. The visual ISM extinction in the direction to this object is A_V < 4.7 mag and, therefore, the intrinsic (dereddened) bolometric flux is F_bol < 400 L_☉ kpc⁻², implying a lower limit to the distance of d_L > 3.9 kpc. García Lario et al. (1991) estimate a distance of d = 6 kpc using the Shklovskii modified method. We adopt this value since it is within the range given by our estimates of d_L and d_k. At ∼6 kpc, the visual ISM extinction is A_V = 1.8 mag, which implies a dereddened bolometric flux and total luminosity of F_bol ∼ 250 L_☉ kpc⁻² and L ∼ 9000 L_☉, respectively.

This is a well-known C-rich Mira variable enshrouded in a thick dust envelope (e.g., Volk et al. 1992). HST images show an extended envelope traced out to a distance of ∼40'' from the center and a remarkable single-armed spiral pattern (Mauron & Huggins 2006). The spiral pattern corresponds closely to the spiral shocks found in models of mass loss in binary systems (Mastrodemos & Morris 1999). The arms of this pattern are brighter to the northwest (P.A. = −30°). In the mid-IR, subarcsecond-resolution 8–12 μm images show an unresolved source (Lagadec et al. 2011). Near-IR observations using AO reveal that IRAS 23166+1655 has two point-like components at its nucleus (presumably a binary system) separated by 011 (Morris et al. 2006).

Emission from CO and other molecules have been previously reported (e.g., Knapp & Morris 1985; Bujarrabal et al. 1994; Neri et al. 1998; Woods et al. 2003; Teyssier et al. 2006; Zhang et al. 2009). Maps of the ¹²CO (J = 1–0) emission with 107 × 46 resolution by Neri et al. (1998) reveal a large outer circumstellar nebulosity, with a ∼20'' angular diameter and presumably asymmetric, surrounding a more compact (∼37 × 28) inner envelope. Our data (Figures 3.24 and 4) show strong ¹²CO (J = 1–0) emission centered at V_LSR = −30 km s⁻¹ with a pseudo-parabolic profile consistent with nebular expansion at velocity V_exp = 13 km s⁻¹. The comparison between the previous ¹²CO (J = 1–0) single-dish profile and that obtained from our interferometric data indicate flux losses of up to ∼70%–80% in the latter.

A double Gaussian ellipse has been fitted to our velocity-integrated ¹²CO map. We find two components oriented nearly perpendicularly to each other: The most compact one, with a deconvolved half-maximum size of 26 × 15, is oriented along P.A. ∼ +40°, whereas the outer one, with a deconvolved half-maximum size of 56 × 46, is oriented along P.A. ∼ −45° (see Table 5). The maximum extent of the outer envelope probed by our maps is ∼14'' × 10'' measured down to a 1σ level. The CO emission distribution is also found to be asymmetric at these large scales (with the long axis oriented along P.A. ∼ −40°). The smaller size of the CO outer envelope compared with that measured by Neri et al. (1998) likely results from interferometric flux losses in our maps. We do not measure a velocity gradient in our CO maps. This prevents us from elucidating the structure of the inner and outer envelope components: We speculate that the former could represent an equatorially dense region or a bipolar flow, and the latter is probably more consistent with an elongated shell.

Our ¹³CO (J = 1–0) maps show a horn-like or double-peaked line profile (Figure 5). The ¹³CO-emitting region is partially resolved and is reasonably well fitted by one Gaussian ellipse with deconvolved half-maximum size 45 × 28 and oriented along P.A. ∼ −33°.

We detect a 2.6 mm continuum emission source with a deconvolved half-maximum size of 25 × 1'' and oriented similarly (within errors) to the outer ¹²CO envelope. The elongation of the large CO- and 2.6 mm continuum-emitting envelope is also similar to that of the nebula visible through scattered light in the optical, which points to an intrinsic large-scale asymmetry of the envelope along the NW direction. The continuum flux at 2.6 mm and in the 450–1000 μm range (measured by Groenewegen et al. 1993) are in clear excess over the value expected extrapolating the object's SED in the IRAS wavelength range. This millimeter and sub-millimeter excess points to an additional component of dust grains much cooler than those responsible for the thermal emission at shorter wavelengths (∼100–150 K).

We adopt a distance to this object of d = 1.1 kpc (e.g., Teyssier et al. 2006, and references therein), which implies a total luminosity of $L \sim \mbox{$F_{\rm bol}$}\sim 9300$ L_☉. Since IRAS 23166+1655 is a high-latitude star (b = −40°), i.e., does not belong to the disk population, and therefore does not follow a standard disk rotation curve, we have not estimated $\mbox{$d_{\rm k}$}$, which would be highly unreliable.

This object is a C-rich PPN with a G2 Ia central star (e.g., Hrivnak 1995; Bujarrabal et al. 2001; Hrivnak et al. 2007, and references therein). The optical nebula imaged with HST has a bipolar (maybe quadrupolar) shape (Sahai et al. 2007a). The lobes are surrounded by a round halo traced out to a radius of r ∼ 55 with a non-uniform radial brightness distribution suggestive of distinct mass-loss episodes. A bright, marginally unresolved (PSF = 11) 12.5 μm counterpart is reported by Meixner et al. (1999). The central star displays a regular cyclical light variation in the optical with a period of ∼80 days (Hrivnak et al. 2010).

A ¹²CO (J = 1–0) single-dish detection toward this object is reported by Likkel et al. (1991). ¹²CO J = 2–1, J = 3–2, and J = 4–3 spectra have been obtained recently (Bujarrabal et al. 2001; Hrivnak & Bieging 2005). Our data show strong ¹²CO (J = 1–0) emission centered at V^LSR_sys = −16 km s⁻¹ with an intense core plus weak and broader wings (Figures 3.25 and 4). By comparing our ¹²CO profile with the single-dish spectrum by Likkel et al. (1991), we conclude that our interferometric data are not affected by flux losses. The CO envelope has a deconvolved half-maximum size of ∼36 × 25 and its long axis is oriented perpendicularly to the main optical lobes. This suggests an equatorially enhanced density distribution in the molecular envelope. The maps of CO emission integrated over the red and blue line wings show two clumps aligned with the optical lobes and separated by ≲1''. As for the PPN IRAS 20000+3239 (Section 3.2.19), the observed gradient is probably tracing a compact bipolar flow: The approaching (receding) flow component is displaced to the SW (NE) from the center. In this interpretation, the lobe orientation inferred from the CO velocity gradient is consistent with that suggested by the brightness contrast between the optical lobes, which independently suggest that the faint, NE optical lobe is the one lying behind the dense, obscuring equatorial regions of the envelope. We detect weak ¹³CO (J = 1–0) emission from this object for the first time (Figure 5).

We adopt a distance to IRAS 23304+6147 of $d=\mbox{$d_{\rm L}$}\sim 4$ kpc considering the value of its bolometric flux ∼330 L_☉ kpc⁻². The total ISM extinction across the Galaxy in the direction to this object is A_V = 4.8 mag and, therefore, the intrinsic (dereddened) bolometric flux is F_bol < 780 L_☉ kpc⁻², implying a lower limit to the distance of d_L > 2.7 kpc. At ∼4 kpc, we find A_V ∼ 2.5 mag and, therefore, dereddened bolometric flux and luminosity of F_bol ∼ 440 L_☉ and ∼7000 L_☉, respectively. Taking into account its galactic coordinates and radial velocity (l = 11386, b = +0059, and V^LSR_sys = −16 km s⁻¹), the kinematic distance to IRAS 23304+6147 is d_k = 1.5 kpc; such a small distance is unlikely since it implies a very low luminosity for a pAGB star (<1000 L_☉).

The lack of zero spacing data in our interferometric maps may result in flux loss of large emission structures with a smooth brightness distribution. For our targets, the flux losses, whenever present, are expected to mainly affect spectral channels around the line center, which typically arise in the extended, low-velocity remnant of the AGB CSE. Considering the shortest baselines in our observations, only emission structures larger than ∼18'' are expected to be resolved out (Table 2).

For 16 of a total of 24 circumstellar CO detections in our survey, we can evaluate whether or not there are interferometric flux losses by comparing our CO profiles, integrated over the whole nebula, with results from single-dish observations, including flux upper limits from previous non-detections. As we discuss in Section 3.2, there is only one source in our sample, the C-rich AGB star IRAS 23166+1655, for which significant ¹²CO (J = 1–0) flux losses of up to 80% are confirmed. Such losses were expected given the large size of the CSE around this star, which has an angular diameter ≳20'' as derived from previous ¹²CO interferometric maps (Neri et al. 1998) and a much larger extent indicated by HST images of the scattered light in the visible (Mauron & Huggins 2006). For one source, IRAS 19566+3423, the upper limit to the CO flux obtained from the single-dish non-detection reported by Likkel et al. (1991) is larger than the flux measured by us and, therefore, interferometric flux losses cannot be evaluated from this comparison. However, the small angular size of the CO envelope of this object, which is unresolved (θ_1/2 < 2''), together with the large beam of our map (∼81) makes flux losses unlikely.

For the remaining 14 sources with single-dish data, including those with the largest envelopes (IRAS 19548+3035 and IRAS 19255+2123 with θ_1/2 > 5''), our OPACOS CO profiles are consistent with no significant flux losses within the errors/noise and flux calibration uncertainties of the data. For some of these sources, the ¹²CO (J = 1–0) single-dish flux is even smaller than the flux measured with the interferometer. We attribute such discrepancies partially to the flux error bars in both data sets but also to the poor knowledge of the target coordinates and possible pointing errors in previous single-dish studies (especially in the earliest works performed during the late 1980s and early 1990s), which may have resulted in partially missing the emission from the target. In that sense, we expect interferometric data to provide a more reliable estimate of the flux due to the more accurate location of the CO-emitting region. Moreover, as explained in Section 1, interferometric data are less severely affected by contamination of the circumstellar profile by the diffuse, extended ISM than single-dish spectra.

For one-third of our CO detections (8 out of 24), no ¹²CO (J = 1–0) single-dish data are available in the literature and, therefore, interferometric flux losses cannot be evaluated by flux comparison. (The IRAS names of these eight sources appear boldfaced and with no asterisk in Table 3.) Significant flux losses of large emission halos are unlikely to affect most of these targets given the small angular sizes of their CO envelopes compared with the beam size of their maps and the primary beam of the antennas, ∼60'' at 115 GHz (see Tables 2 and 5). In fact, these eight sources (marked with big empty circles in Figure 6) have envelopes with comparable or even smaller angular sizes, θ_1/2, than sources with no appreciable flux losses confirmed by comparison with single-dish data (see above). Moreover, they have been observed with relatively large beams (>4''), except for IRAS 22036+5306 (beam ∼ 24).

Optical HST imaging supports the compactness of most of our targets with no single-dish CO spectra (see Sahai et al. 2007a, and Table 1): Three are spatially unresolved and, except for IRAS 22036+5306, the rest show scattered light envelopes with radius <5''. IRAS 22036+5306 has an optical halo that can be traced out to a distance from the star of ∼65. IRAS 22568+6141 has relatively extended (∼4''-long) lobes, however, no scattered halo has been detected in this case. The large distances to most objects (with a mean of d ∼ 5.5 kpc for the eight sources under discussion; see Table 1) also make the presence of very large CO halos unlikely: The CO photodissociation radius expected for typical values of the mass-loss rate and expansion velocity of $\dot{M}$ ≈ 10⁻⁵ M_☉ yr⁻¹ and V_exp = 14 km s⁻¹ is R_CO ∼ 2 × 10¹⁷ cm (Mamon et al. 1988; Planesas et al. 1990), which is equivalent to an angular radius of <2''–3'' at a distance of >5 kpc. Observationally, it is also well established that linear sizes of CO envelopes of most PPNs are in the range ≈10¹⁶–3 × 10¹⁷ cm (e.g., Neri et al. 1998; Teyssier et al. 2006).

In summary, given the small angular sizes of the CO envelopes expected, it is very unlikely for sources spatially unresolved or mapped with relatively large beams >4'' in our survey to be affected by significant interferometric flux losses. The only two exceptions may be IRAS 22036+5306 and IRAS 22568+6141, which have CO envelopes spatially well resolved in our OVRO maps with beams of ∼23 and ∼4'', respectively, and also show extended nebulosities in the optical images. The flux loss statistics of objects with single-dish data (1 out of 16) and their comparable (or even larger) envelope sizes to the eight objects with no single-dish data supports an equally small probability of significant flux losses in the latter.

Line profiles of circumstellar CO emission around evolved stars provide information about the nebular kinematics, although they also depend on the optical depth of the line, the angular size of the envelope relative to the telescope beam, and excitation conditions throughout the CO-emitting layers (e.g., Olofsson et al. 1982; Olofsson 1996). The CO emission profiles in pAGB objects often display an intense-narrow line core plus weak-broader wings. The line core is believed to arise in the slow remnant of the AGB envelope, whereas the wings trace fast bipolar outflows of molecular gas presumably accelerated by interaction with fast winds later developed in the early pAGB or late-AGB phase (see Section 1).

We have investigated the presence of different kinematic components, slow and fast, in our sample by comparing the values of the FWZI and FWHM of the ¹²CO (J = 1–0) transition¹⁰ (Table 3). We have defined the dimensionless parameter γ ≡ FWZI/FWHM, which happens to be a useful indicator of the type of line profile and, in particular, of the presence of wings. As can be easily demonstrated, for a parabolic profile γ ∼ $\sqrt{2}$ ∼ 1.4, whereas for a rectangular or horn-like shape γ ∼ 1. Values of γ > 1.4 result if there are wings superimposed on the narrow core or if the line has a triangular or Gaussian-like profile.¹¹

Based on the values of γ measured in our sample and taking into account their error bars, there may be as many as 12 out of 24 sources with CO emission wings (Figure 7). Among these, the wings of IRAS 19024+0044 are tentative due to strong ISM contamination of the circumstellar CO profile, which result in a somewhat uncertain value of FWHM. Particularly remarkable are the broad, Gaussian-like profiles of IRAS 22036+5306 and IRAS 19374+2359, which display the broadest wings (FWZI > 120 km s⁻¹) in our sample (Figure 4). The fast winds responsible for these wide profiles are indeed even faster than inferred from our OPACOS data as indicated by the larger full width of the ¹²CO (J = 3–2) line in IRAS 22036+5306 (FWZI ∼ 400 km s⁻¹; Sahai et al. 2006) and the single-dish ¹²CO (J = 2–1) profile observed in IRAS 19374+2359 by C.S.C. (FWZI ∼ 300 km s⁻¹; to be published elsewhere). In addition to these two sources with clear wing-dominated profiles, there are a few other objects that also show wings but significantly weaker than the line core. Some clear examples, in which the wings and the line core can be easily identified as different components of the profile, are IRAS 19475+3119, IRAS 23304+6147, and IRAS 22223+4327 (Figure 4). In the rest of cases, the presence of weak wings is more subtle and cannot be readily guessed by simple inspection of their profiles (e.g., IRAS 23166+1655 and IRAS 20000+3239). This makes γ a useful parameter for a systematic search of wings or, more generally, of deviations from the standard profile in AGB CSEs, which can be interpreted as an indication of kinematic complexity. We note that, of course, small values of γ (<1.4) do not demonstrate the absence of emission wings, which may remain undetected below the noise level.

Zoom In Zoom Out Reset image size

Obtaining an estimate of the nebular expansion velocity from the CO line profile is not straightforward (even ignoring optical depth effects or ISM contamination of the profile). In the case of multiple kinematic components or significant velocity gradients in the molecular envelope, the mere definition of a unique expansion velocity is clearly not appropriate. In these cases, ideally, one would like at least to be able to obtain separately the "characteristic" expansion velocity of the slow AGB envelope (V_AGB) and that of the fast pAGB flow projected on the line of sight (V*_pAGB).

Keeping in mind these limitations and the resulting uncertainties, we have obtained estimates of V_AGB and V*_pAGB (Table 6). The determination of V_AGB has been made by fitting a function of the type $I_v \propto 1-[(v-\mbox{$V_{\rm LSR}$})/\mbox{$V_{\rm AGB}$}]^\alpha$ to the ¹²CO (J = 1–0) line core (note that by choosing α = 2 and α ≫ 2, this function produces parabolic and horn-like profiles, respectively). For sources with ¹²CO and ¹³CO (J = 1–0) detections, the values of V_AGB measured from both profiles are consistent, with differences <2 km s⁻¹ in all cases, except for IRAS 18560+0638, IRAS 19374+2359, and IRAS 22036+5306. In these cases, the most reliable values of V_AGB derived from the ¹³CO (J = 1–0) profile are given in the table: we note the strong ISM contamination of the ¹²CO profile toward IRAS 18560+0638; in the case of IRAS 19374+2359 and IRAS 22036+5306, the line core can be best isolated from the wings in the ¹³CO profile. For objects with emission wings (γ > 1.4, Section 4.2.1), we have adopted V*_pAGB = FWZI/2. Assuming that the wings arise in a bipolar flow with a given inclination with respect to the plane of the sky, i, V*_pAGB represents a lower limit to the maximum axial velocity of the fast flow projected on the line of sight. (The lower limit arises because FWZI may be sensitivity limited.) The maximum deprojected expansion speed of the pAGB winds is given by V*_pAGB/sin (i), however, the inclination of the lobes is unknown in all cases (except for IRAS 19306+1407, for which |i| ∼ 5°; Lowe & Gledhill 2007).

The distribution of values obtained for V_AGB is given in Figure 8. Our results are in good agreement with values of the expansion velocity of AGB CSEs estimated in previous studies (e.g., Loup et al. 1993; Olofsson et al. 1993). We find that, on average, V_AGB is larger in objects with no wings detected (mean V_AGB = 17 km s⁻¹) than in objects with wings (mean V_AGB = 12 km s⁻¹). The maximum projected expansion velocities of the flows inferred from the wing widths range between V*_pAGB = 8 and 200 km s⁻¹, with a median of V*_pAGB = 25 km s⁻¹ (Figure 9). In all cases, except for IRAS 19374+2359 and IRAS 22036+5306, V*_pAGB is smaller than 35 km s⁻¹.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

We have defined and calculated two other parameters that are expected to be velocity indicators, which we designate as equivalent-width velocity, V_EW, and the mass-weighted velocity, V_MASS (Table 6). The former is given by the ratio of the ¹²CO (J = 1–0) line flux (in units of Jy km s⁻¹) to the peak intensity (Jy), that is $\mbox{$V_{\rm EW}$}= \int I_v dv/I_{\rm peak}$; whereas V_MASS is defined as the ratio of the velocity integral of the ¹²CO (J = 1–0) profile multiplied by the projected velocity relative to V^LSR_sys (in units of Jy km s⁻¹ km s⁻¹) to the CO line flux, i.e., $\mbox{$V_{\rm MASS}$}= \int I_v v dv/\int I_v dv$. Strictly, V_MASS is only a mass-weighted velocity when the emission is optically thin.

We find that the values of the FWHM and V_EW are similar in most cases and, therefore, these are similarly good or bad indicators of the mean expansion velocity. V_MASS is on average smaller than FWHM by a factor in the range 1.6–3.6 (median is ∼3). According to its definition, this is expected, since the emission from the intense but narrow line core has more weight than the weak-broad wings. There are a few outliers of the linear relationship between FWHM and V_MASS followed by most objects: IRAS 19306+1407, IRAS 19566+3423, IRAS 22568+6141, and IRAS 22036+5306. This is because all these objects have broad wings that contribute significantly to the ∫I_vvdv integral, enhancing the value of V_MASS. In this sense, V_MASS can be considered an indirect indicator of the fraction of the mass that has been accelerated (i.e., in the fast flow): The larger the intensity of the wings is relative to the core, the larger the value of V_MASS will be.

In the case of an isotropic wind expanding at constant velocity, which is usually taken to be a good approximation for the undisturbed AGB CSE, V_MASS relates to the expansion velocity by V_AGB = $2\times \mbox{$V_{\rm MASS}$}$, assuming optically thin emission (see Bujarrabal et al. 2001). We have represented $2\times \mbox{$V_{\rm MASS}$}$ versus V_AGB of the AGB remnant in Figure 10. As we can see, most data points satisfy the V_AGB = $2\times \mbox{$V_{\rm MASS}$}$ relationship with little scatter. The points below the V_AGB = $2\times \mbox{$V_{\rm MASS}$}$ line may result from moderate opacity effects on the ¹²CO (J = 1–0) profile; note also that ISM absorption or emission features in the circumstellar profile as well as any flux losses in the line core may alter the values of V_MASS. Again, there are five sources notably above the V_AGB = $2\times \mbox{$V_{\rm MASS}$}$, which is explained by the presence of relatively intense, broad wings that contribute significantly to V_MASS in these cases.

Zoom In Zoom Out Reset image size

For the 12 sources with ¹²CO (J = 1–0) emission wings denoting fast ejections (Section 4.2.1), we have obtained a rough estimate of the fraction of the mass contained in the fast wind, X_pAGB = M_pAGB/M, and in the slow remnant AGB envelope, X_AGB = M_AGB/M, which is responsible for the line core emission. For optically thin emission, X_AGB is given by the ratio of the ¹²CO (J = 1–0) emission integrated over the line core only (i.e., over the velocity range $\mbox{$V_{\rm sys}^{\rm LSR}$}\pm \mbox{$V_{\rm AGB}$}$) to the total line flux, obtained by integrating over the whole profile including the wings. The values of X_pAGB are simply taken to be 1 − X_AGB and are expressed in Table 7 as a percentage of the total mass M. We stress that these figures must be considered only as very rough guesses of the mass fraction in the fast component. The reason is twofold: First, we cannot rule out a moderate or high optical depth of the core emission in some cases, which would yield to an underestimate of X_AGB; second, the fast wind is also expected to contribute (with an unknown amount) to the line core emission due to projection effects, which would result in an overestimate of X_AGB. Both effects, with opposite actions on X_AGB, could take place simultaneously.

As shown in Figure 13, in 7 out of 12 sources the mass carried out by the fast ejections relative to the total mass in the envelope is relatively small, of the order of ∼10% or less. Nevertheless, there are five objects for which X_pAGB is significantly larger, with values ranging from 30% up to ∼60%–80% in IRAS 22036+5306 and IRAS 19374+2359, which are as expected the targets displaying the broadest, most intense emission wings.

Zoom In Zoom Out Reset image size