A&A 385, 884-895 (2002)
DOI: 10.1051/0004-6361:20020178

BVRIJHK photometry of post-AGB candidates[*],[*]

T. Fujii 1,2,[*] - Y. Nakada 1,3 - M. Parthasarathy 2,4


1 - Institute of Astronomy, University of Tokyo, 2-21-1 Osawa, Mitaka, Tokyo 181-0015, Japan
2 - National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan
3 - Kiso Observatory, Institute of Astronomy, University of Tokyo, Mitake, Kiso, Nagano 397-0101, Japan
4 - Indian Institute of Astrophysics, Bangalore 560034, India

Received 27 March 2000 / Accepted 20 December 2001

Abstract
BVRIJHK photometric observations are presented for 27 post-AGB candidates. Almost all objects show a double peaked SED curve in the optical to far-infrared wavelengths. Seventeen objects were classified as post-AGB stars on the basis of their spectral type, location in the IRAS color-color diagram and SED. The physical parameters of the observed post-AGB stars, the inner radius of the detached shell, the mass of the shell and the distance were derived using the simple dust shell model. We compared our observational sequence of post-AGB objects to the theoretical evolutionary sequence (Schönberner 1983; Blöcker 1995) in the stellar temperatures versus age diagram. We found that two post-AGB stars, IRAS 05040+4820 and 08187-1905, have low stellar temperature with a large dynamical age of the dust shell. They appear to provide the first observational evidence that some low-mass stars bypass the planetary nebulae stage because of their slow increase in stellar temperature.

Key words: stars: AGB and post-AGB - stars: circumstellar matter - stars: evolution - stars: mass-loss - infrared: stars - ISM: Planetary Nebulae: general


1 Introduction

Low to intermediate-mass stars cross the HR diagram horizontally from the tip of the asymptotic giant branch (AGB) to the planetary nebula (PN) region after they terminate a rapid mass-loss phase. This transition phase is called post-AGB phase of evolution. From an analysis of the IRAS point source catalog, several post-AGB stars have been identified (Parthasarathy & Pottasch 1986, 1989; Pottasch et al. 1988). The spectral energy distribution (SED) of post-AGB stars is double peaked. One peak is at far-infrared wavelengths due to the cold dust-shell (100-200 K) and the other peak is at shorter wavelengths, optical or near-infrared, from the obscured central star. The cold dust-shell was observed by IRAS. Most of the post-AGB stars, proto-planetary nebulae (PPNe) and PNe were found within the IRAS color box defined by F(12 $\mu$m)/F(25 $\mu$m) < 0.3 and F(25 $\mu$m)/F(60 $\mu$m) > 0.3 (van der Veen & Habing 1988; Pottasch et al. 1988). They are well separated from other types of objects. Based on the IRAS color-color diagrams (Pottasch et al. 1988; Preite-Martinez 1988) one can conclude that there is a good chance that an object is a PN, a PPN or a post-AGB star if it is within the box defined by the colors mentioned above. An occasional HII region, Seyfert galaxy or T-Tau star is not excluded from this range.

Like PNe, PPNe are composed of post-AGB central stars and detached circumstellar envelopes of gas and dust; however, unlike PNe, their central stars are too cool to photoionize the envelopes. The terms PPNe and post-AGB stars are often used to describe the objects evolving from the tip of the AGB to PNe. Often one uses the term PPNe to describe the circumstellar shells of post-AGB stars that are not yet photoionized.

Since the post-AGB objects span a wide range in the spectral type of their central star as well as the obscuration of their dust shell, the analysis of their SED needs a combination of the optical and near-infrared photometry in addition to the IRAS data. In order to understand the nature of the SED in the shorter wavelength, optical and near-infrared photometry has been performed for some of the post-AGB stars. Manchado et al. (1989) and García-Lario et al. (1990, 1997) carried out JHK photometric survey of several IRAS sources with colors like PNe. The SEDs of eight post-AGB candidates were studied based on BVRIJHK photometry by Hrivnak et al. (1989). They showed that multi-color photometry in combination with spectral types of the IRAS sources enables one to estimate the parameters of the stars and their circumstellar dust envelopes.

In order to increase the number of well-determined post-AGB stars and to classify them using SEDs from optical to far-infrared wave length, we have initiated a program to carry out systematic optical and near-infrared photometry of post-AGB candidates. We have selected several post-AGB candidates based on the above IRAS colors that are observable from Kiso Observatory. In this paper we report the BVRIJHK photometric observations and analysis of 27 post-AGB candidates.

2 Observations and analysis

Photometric observations in BVRIJHK were made using the CCD and infrared cameras at Kiso Observatory, University of Tokyo. The list of post-AGB candidates together with their IRAS fluxes are given in Table 1. The location of the sources in the IRAS color-color diagram is shown in Fig. 1.

  \begin{figure} \par\includegraphics[angle=-90,width=6.7cm,clip]{fig1.eps} \end{figure} Figure 1: The positions of program IRAS sources in the van der Veen & Habing (1988) IRAS color-color diagram. The IRAS colors are defined as $ [12]{-}[25]= 2.5 \log ({F}_{25\,\rm \mu m}/{F}_{12\,\rm \mu m})$, $ [25]{-}[60]= 2.5 \log ({F}_{60\,\rm \mu m}/{F}_{\rm 25\,\mu m})$. Filled circles denote objects which have IRAS quality index 3 or 2 at 12 $\rm\mu m$, 25 $\rm\mu m$ and 60 $\rm\mu m$. Open triangles denote objects which have qindex 1 in one of the 12 $\rm\mu m$, 25 $\rm\mu m$ and 60 $\rm\mu m$ bands.
Open with DEXTER

The observational results are tabulated in Table 2.
   
Table 1: Infrared properties of candidates from IRAS-PSC.
ID IRAS Name Association $F_{12\rm\, \mu m}$ $F_{25\rm\,\mu m}$ $F_{60\rm\,\mu m}$ $F_{100\rm\,\mu m}$ VAR a lb bb qindexc
No.   SAO or HD [Jy] [Jy] [Jy] [Jy]   [deg.] [deg.]  
1 02086+7600   0.25 2.68 7.41 10.49 9 127.9 14.2 3333
2 02143+5852   5.90 18.06 5.39 8.97 94 133.9 -1.9 3331
3 02528+4350   0.56 2.38 3.00 2.22 0 145.4 -13.3 3333
4 04269+3550   2.83 9.92 21.29 15.58 19 164.9 -8.6 3333
5 04296+3429   12.74 45.94 15.45 9.22 11 166.2 -9.1 3331
6 05040+4820 SAO 40039 0.25 7.20 20.20 11.00 - 159.8 4.8 1333
7 05089+0459   7.37 21.89 11.10 3.78 14 196.3 -19.5 3333
8 05113+1347   3.78 15.30 5.53 1.67 10 188.9 -14.3 3331
9 05170+0535 SAO 112630 0.60 4.42 14.06 9.38 31 196.8 -17.5 3333
10 05238-0626   0.59 1.74 1.16 1.87 0 208.9 -21.8 3321
11 05341+0852   4.51 9.85 3.96 8.01 0 196.2 -12.1 3331
12 05355-0117 HD 290764 0.68 3.89 10.38 9.00 0 205.5 -16.9 3333
13 05381+1012   0.85 2.93 1.39 9.44 13 195.5 -10.6 3331
14 06013-1452   1.86 4.34 3.11 1.25 0 221.2 -17.2 3331
15 06059-0632 SAO 132875 0.38 4.41 9.64 6.52 12 213.9 -12.5 3322
16 06060+2038 HD 252325 2.27 27.08 112.3 18.39 15 189.8 0.4 3321
17 06284-0937 SAO 133356 0.32 2.60 0.40 369.1 - 219.3 -8.9 1311
18 06338+5333 SAO 25845 0.46 0.38 0.40 1.15 3 162.0 19.6 3211
19 06530-0213   6.11 27.41 15.05 4.10 8 215.4 -0.1 3333
20 07077-1825   0.80 6.66 0.40 211.4 0 231.5 -4.4 3311
21 07131-0147   2.59 4.22 3.96 3.68 1 217.4 4.5 3333
22 07171+1823   0.41 1.38 0.70 1.00 - 199.5 14.4 1331
23 07253-2001   6.30 15.16 6.05 8.05 8 234.9 -1.5 3331
24 07430+1115   7.68 29.93 10.67 2.53 9 208.9 17.1 3333
25 08187-1905 HD 70379 0.71 17.62 12.31 3.68 0 240.6 9.8 3333
26 23304+6147   11.36 59.07 26.60 30.89 8 113.9 0.6 3331
27 (BD+39$^{\circ}$4926) SAO 72704 - - - - - 98.4 -16.7  
Notes: BD+39$^{\circ}$4926 has no IRAS data.
a VAR: percent likelihood of variability.
b l, b: Galactic longitude and latitude, respectively.
c qindex: Flux density quality (1 = upper limit, 2 = moderate quality, 3 = good quality). From left to right, each figure
stand for IRAS photometric band 12 $\rm\mu m$, 25 $\rm\mu m$, 60 $\mu\rm m$ and 100 $\mu\rm m$.

2.1 Optical observation and data reduction

The optical photometric observations were performed using a 105 cm Schmidt telescope with a CCD camera. The CCD camera contains a TI Japan TC215 chip with an array size of 1024 $\times$ 1024 pixels. The field of view is about $12\hbox{$.\mkern-4mu^\prime$ }5 \times 12\hbox{$.\mkern-4mu^\prime$ }5 $ and one pixel is $0\hbox{$.\!\!^{\prime\prime}$ }75$ in the sky. The CCD images were taken at B, V, $R_{\rm C}$ and $I_{\rm C}$ filters. The raw data were processed using the IRAF image data reduction software by subtracting bias and dividing by the dome flat field. We used the DoPHOT (Schechter et al. 1993) program to obtain magnitudes. The instrumental magnitudes were transformed to the Johnson-Cousins photometric system magnitudes by analyzing the frames of standard stars from Landolt (1992). The derived B, V, $R_{\rm C}$ and $I_{\rm C}$ magnitudes are shown in Table 2 Cols. 3, 4, 5 and 6. The observational date is shown in Table 2 Col. 10.
   
Table 2: Photometric observations of Post-AGB candidates.
ID IRAS Name B V $R_{\rm C}$ $I_{\rm C}$ J H K Obs. date Obs. date
No.   [mag] [mag] [mag] [mag] [mag] [mag] [mag] Optical NearIR

1

 02086+7600         13.54 12.74 >11.14   Oct. 19/1997
2 02143+5852 14.96 13.74 12.96 12.15 10.58 9.51 8.79 Jan. 11/1996 Oct. 19/1997
3 02528+4350 11.11 10.82 10.67 10.46 10.20 9.99 9.62 Jan. 11/1996 Oct. 19/1997
4 04269+3550         12.00 9.65 8.35   Oct. 19/1997
5 04296+3429 16.18 14.17 12.89 11.65 9.67 8.80 8.44 Dec. 02/1995 Feb. 24/1997
    16.41 14.23 12.98 11.74 9.55 8.68 8.28 Dec. 12/1995 Nov. 11/1997
6 05040+4820 10.14 9.58 9.20 8.81 8.24 8.03 7.91 Dec. 02/1995 Oct. 19/1997
    10.12 9.54 9.20 8.86       Dec. 12/1995  
7 05089+0459 16.20 14.48 13.12 11.65 10.14 9.26 8.92 Feb. 07/1996 Feb. 24/1997
            10.09 9.12 8.84   Oct. 17/1997
8 05113+1347 14.67 12.40 11.27 10.26 8.96 8.43 8.05 Dec. 02/1995 Feb. 24/1997
    14.76 12.49 11.32 10.35 9.02 8.44 8.25 Dec. 12/1995 Oct. 17/1997
9 05170+0535 9.72 9.11 8.77 8.44       Dec. 02/1995  
    9.76 9.18 8.83 8.50       Dec. 12/1995  
10 05238-0626 10.96 10.52 10.23 9.94 9.61 9.31 9.03 Dec. 02/1995 Oct. 19/1997
11 05341+0852 15.44 13.58 12.43 11.44 9.97 9.36 9.05 Dec. 02/1995 Nov. 11/1997
    15.53 13.63 12.51 11.49       Dec. 12/1995  
12 05355-0117 10.16 9.81 9.61 9.32       Mar. 06/1996  
13 05381+1012 11.50 10.59 10.04 9.52 8.80 8.44 8.18 Dec. 02/1995 Oct. 19/1997
    11.47 10.57 10.02 9.52       Dec. 13/1995  
14 06013-1452 10.34 10.22 10.17 10.02 9.87 9.72 9.34 Mar. 06/1996 Mar. 24/1997
15 06059-0632 9.33 8.84 8.59 8.30       Mar. 06/1996  
16 06060+2038 11.36 10.76 10.32 9.91       Mar. 06/1996  
17 06284-0937 9.42 8.99 8.78 8.53       Feb. 07/1996  
18 06338+5333 9.32 8.87 8.57 8.25       Dec. 12/1995  
19 06530-0213 16.26 13.99 12.63 11.41       Feb. 07/1996  
20 07077-1825 11.25 10.78 10.43 10.12       Jan. 11/1996  
21 07131-0147 16.40 14.53 13.03 11.58 9.78 9.09 8.30 Dec. 13/1995 Feb. 26/1997
22 07171+1823 12.70 12.73 12.62 12.74       Jan. 11/1996  
23 07253-2001 14.07 13.34 12.82 12.32       Dec. 13/1995  
24 07430+1115 14.22 12.38 11.38 10.50 8.95 8.29 7.87 Dec. 12/1995 Feb. 26/1997
25 08187-1905 9.70 9.03 8.56 8.20       Dec. 13/1995  
26 23304+6147 15.65 13.19 11.80 10.49 8.54 7.86 7.54 Feb. 07/1996 Oct. 19/1997
27 (BD+39$^{\circ}$4926) 9.46 9.25 9.07 8.90 8.50 8.36 8.19 Nov. 29/1995 Oct. 19/1997
Notes. The errors of magnitude are $\pm0\hbox{$.\!\!^{\rm m}$ }08$ for B, $\pm0\hbox{$.\!\!^{\rm m}$ }05$ for V, $R_{\rm C}$ and $I_{\rm C}$, $\pm0\hbox{$.\!\!^{\rm m}$ }06$ for J, $\pm0\hbox{$.\!\!^{\rm m}$ }04$ for H, $\pm0\hbox{$.\!\!^{\rm m}$ }20$ for K. On Oct. 19/1997, the errors of JHK are $\pm0\hbox{$.\!\!^{\rm m}$ }18$, $\pm0\hbox{$.\!\!^{\rm m}$ }12$ and $\pm0\hbox{$.\!\!^{\rm m}$ }38$, respectively. On Feb. 26/1997 and Mar. 24/1997, they are $\pm0\hbox{$.\!\!^{\rm m}$ }15$, $\pm0\hbox{$.\!\!^{\rm m}$ }06$ and $\pm0\hbox{$.\!\!^{\rm m}$ }20$, respectively. The errors were determined by the magnitude deviations of standard stars. The large errors of near infrared is due to bad sky conditions. The large error in K band may be due to the thermal emission from inside of Schmidt telescope.



2.2 Near-infrared observation and data reduction

The near-infrared photometric observations were carried out using J, H and $K_{\rm S}$ filters. The $K_{\rm S}$ filter has a shorter cut-off wavelength in order to reduce the contribution of the thermal background radiation. The passband of $K_{\rm S}$ is from 2.0 to 2.3 $\mu\rm m$. KONIC (Kiso Observatory near-infrared camera, Itoh et al.  1995) with a $1040 \times 1040$ elements PtSi Schottky-barrier array (Mitsubishi Electric Co.) was attached to the 105 cm Schmidt telescope. The field of view is about $18\hbox{$.\mkern-4mu^\prime$ }4 \times 18\hbox{$.\mkern-4mu^\prime$ }4$. Standard data processing (dark subtraction and flat-fielding)was performed with the IRAF software package. The frames for flat-fielding were made by combining object frames with median filters in the $JHK_{\rm S}$ band. We used the IRAF/APPHOT package to obtain magnitudes. Several standard stars in the Elias list (Elias et al. 1982) were used to correct for atmospheric absorption and to transform the instrumental magnitudes to the CTIO system. The derived J, H and K magnitudes are shown in Table 2 Cols. 7, 8 and 9. The observational date is shown in Table 2 Col. 11.
   
Table 3: Central star's properties and extinction properties.
ID IRAS Name Spectral Chem. a Tstar (B-V)0 $(B-V)_{\rm obs}$ E(B-V) AV1 b AV2 c ref. d Object type
No.   Type Type [K] [mag] [mag] [mag] [mag] [mag]    
1 02086+7600                   YSO?
2 02143+5852 F(5Ib e)   6900 0.33 1.22 0.89 2.75 2.0 A P-AGB
3 02528+4350                   Galaxy
4 04269+3550                   YSO?
5 04296+3429 G0Ia C 5550 0.75 2.07 1.32 4.10 2.23 B P-AGB
6 05040+4820 A4Ia   8750 0.07 0.57 0.50 1.55 1.0 C P-AGB
7 05089+0459 M(0Ib e)   3650 1.64 1.73 0.09 0.27 0.37 A P-AGB
8 05113+1347 G8Ia C 4590 1.17 2.27 1.10 3.41 1.67 B P-AGB
9 05170+0535 G0Ve               D PMS
10 05238-0626 F2II   7380 0.30 0.44 0.14 0.44 0.37 E P-AGB
11 05341+0852 F4Iab C 7060 0.29 1.88 1.59 4.93 1.21 F P-AGB
12 05355-0117 A5III               G $\delta$ Scuti?, Herbig Ae?
13 05381+1012 G(2Ib e)   4850 0.86 0.90 0.04 0.13 1.30 B P-AGB
14 06013-1452 Ae               G Herbig Ae/Be
15 06059-0632 B3               G PMS
16 06060+2038 B1V               G ii II region
17 06284-0937 B3ne               H Herbig Ae/Be
18 06338+5333 F7IVw   6250 0.50 0.45 -0.05 0.0 f 0.31 G P-AGB
19 06530-0213 F0Iab C 7700 0.17 2.28 2.11 6.53 2.0 G P-AGB
20 07077-1825 O6               G ii II region
21 07131-0147 M5III   3330 1.63 1.87 0.24 0.76 2.0 I P-AGB, Bipolar-PPN
22 07171+1823 B(5Ib e)   13600 -0.10 -0.04 0.06 0.19 0.25 J P-AGB
23 07253-2001 F5Ie   6900 0.33 0.73 0.40 1.23 1.0 E P-AGB
24 07430+1115 G5 0-Ia C 4850 1.03 1.84 0.81 2.50 0.06 K P-AGB
25 08187-1905 F6Ib/II   6630 0.42 0.67 0.25 0.78 0.33 G P-AGB
26 23304+6147 G2Ia C 5200 0.87 2.46 1.59 4.93 2.2 B P-AGB
27 BD+39$^{\circ}$4926 B8(Ib e)   11200 -0.04 0.22 0.26 0.79 0.37 G P-AGB
a  C: carbon-rich object.
b  Derived from intrinsic and observed B-V, it contains interstellar and circumstellar extinction.
c  Derived from interstellar extinction maps.
d  References for spectral types; (A) Meixner et al. (1999), (B) Hrivnak (1995), (C) Hardorp et al. (1959), (D) Zuckerman et al. (1995), (E) Reddy and
Parthasarathy (1996), (F) Parthasarathy (1993), (G) SIMBAD database, (H) van den Ancker et al. (1998), (I) Scarrott et al. (1990), (J) Vijapurkar et al.
(1998), (K) Hrivnak and Kwok (1999).
e  Assumption.
f  We assumed AV=0.0, since $(B-V)_{\rm obs}<(B-V)_{0}$.


  \begin{figure} \par\includegraphics[width=5.6cm,clip]{fig2.eps} \end{figure} Figure 2: (J-H)0 vs. (H-K)0 color-color diagram of the 12 observed IRAS sources. We used average magnitudes for the objects which were observed twice. We corrected the interstellar and circumstellar reddening for all stars. Blackbody at temperatures ranging from 2000 to 1000 K are marked on the solid line. The dotted line shows the super-giant (SG) sequence and the dashed line shows the main-sequence (MS) (Cox 2000). The location of Miras and PNe (Glass & Feast 1982; Whitelock 1985) in the color-color diagram are also indicated.
Open with DEXTER


  \begin{figure} \par\includegraphics[width=5.6cm,clip]{fig3.eps} \end{figure} Figure 3: (B-V)0 vs. (V-I)0 color-color diagram of the 17 observed IRAS sources. We used average magnitudes for the objects which were observed twice. The thick line indicates the color-color sequence of super-giants (Cox 2000).
Open with DEXTER

   
2.3 Reddening correction

In order to derive the parameters of the stars, we need to estimate the reddening in the line of sight of these sources. First, we used two methods to obtain reddening: (1) from spectral types of stars, (2) from interstellar extinction maps. While one can estimate both interstellar and circumstellar reddening by using the first method, one can obtain interstellar reddening alone by using the latter method.
Method (1): AV1; for most post-AGB candidates, their spectral types are available in the literature. From the spectral type of the star we know the intrinsic (B-V)0, and from the observed $(B-V)_{\rm obs}$ we estimated the reddening E(B-V) ( $=(B-V)_{\rm obs} -(B-V)_{0}$). To derive the AV value, we adopted RV = 3.1, i.e. $ A_{V} = 3.1 \times E(B-V)$. For the objects whose spectral subclass and/or luminosity class were not determined, we assume a moderate one. The values of intrinsic (B-V)0 and stellar effective temperature $T_{{\rm star}}$ were quoted from the lists of Schmidt-Kaler (1982). The AV1 values contain the contribution from the interstellar and circumstellar reddening.
Method (2): AV2; we estimated AV values using their galactic latitudes and longitudes and corresponding interstellar extinction maps (Neckel & Klare 1980; Burstein & Heiles 1982; Schlegel et al. 1998).

The spectral type, chemical type (carbon-rich or not), $T_{{\rm star}}$, E(B-V), two values of AV(AV1 and AV2) and the object type cited from the literature are given in Table 3. In general one can estimate the circumstellar reddening from these two values of AV. However, for some stars, the AV2 (interstellar) is larger than the AV1 (interstellar plus circumstellar). This contradiction may be due to uncertainty of interstellar extinction maps and uncertainty of circumstellar extinction law. This may also be due to uncertainty in spectral types in some cases. In particular the AV value derived from the map depends on the distance to the star and the spatial resolution of these maps is not sufficiently high. Therefore, it is dangerous to blindly accept the AV value obtained from the map in some cases. Indeed, we could not find dust shell parameters for some stars if we fixed the value of interstellar extinction and the stellar temperature simultaneously. For this reason, we did not adopt AV2 to plot SED diagrams and model calculations. We estimated interstellar reddening from model calculations, treating extinction as a parameter. We describe how to estimate interstellar AV in Sect. 3.

We adopted AV1 in plotting optical and near-infrared color-color diagrams to see positions of dereddened post-AGB central stars. We have corrected the photometric data for interstellar and circumstellar reddening using standard extinction laws derived by Rieke & Lebofsky (1985). For $R_{\rm C}$ and $I_{\rm C}$ bands, we interpolated values given in the mentioned above paper and adopted them. Strictly, it is not correct to apply a standard extinction law to correct for circumstellar extinction since these two (interstellar and circumstellar) extinction laws are not guaranteed to be the same. However, little is known about circumstellar extinction laws. Therefore, we presumed interstellar and circumstellar extinction laws to be similar here.

The objects for which the evolutionary status is not clear were excluded from further analysis and discussion. The IRAS name of these excluded objects are 02086+7600, 02528+4350, 04269+3550, 05170+0535, 05355-0117, 06013-1452, 06059-0632, 06060+2038, 06284-0937 and 07077-1825. The reasons for excluding these objects from the list of post-AGB candidates are given in Sect. 4.2 (notes on individual objects).

Figures 2 and 3 show the location of the program stars in the (J-H)0, (H-K)0 and (B-V)0, (V-I)0 color-color diagram with the sequence of super-giants (Cox 2000), respectively. The position of the stars in the color-color diagrams indicate the consistency of optical, near-infrared colors and spectral types.

   
3 Model fitting

In order to derive the dust shell parameters of each object, we introduce a simple model consisting of a central star and a detached shell. The entire SED is given by:

\begin{displaymath}F_{\lambda, {\rm model}} = F_{\lambda, {\rm star}} + F_{\lambda, {\rm shell}}\,, \end{displaymath} (1)

where $F_{\lambda, {\rm star}}$ is flux from the star and $F_{\lambda, {\rm shell}}$ is flux from the dust shell. We assume that the dust shell is spherically symmetric. The geometrical thickness of the dust shell was taken as $4.7\times 10^{17}~{\rm cm}$, corresponding to 10000 yr for the duration of AGB mass-loss (superwind mass-loss) with an expansion velocity of $15\,{\rm km}\,{\rm s}^{-1}$ and a constant mass loss rate.

The central star was assumed to be a blackbody source with a temperature $T_{{\rm star}}$ and the luminosity $L = 8000\,L_{\odot}$. The core-mass luminosity relation gave 0.625 $M_{\odot}$for the mass of the central star (Blöcker 1995), if we neglect the very thin envelope around the core. Flux from the star is given by:

\begin{displaymath}F_{\lambda, {\rm star}} = \pi B_{\lambda}(T_{{\rm star}}) \exp(-\tau_{{\rm shell},\lambda})\, R_{{\rm star}}^{2} / D^{2}, \end{displaymath} (2)

where $B_{\lambda}(T_{{\rm star}})$ is a blackbody function at a wavelength $\lambda$, $T_{{\rm star}}$ is the stellar temperature, $\tau_{{\rm shell},\lambda}$ is the optical depth of the dust shell at a wavelength $\lambda$, $R_{{\rm star}}$ is the radius of the star and D is the distance to the star.

A constant expansion velocity of the AGB shell ( $V_{{\rm exp}} = 15\,{\rm km}\,{\rm s}^{-1}$) and constant mass-loss rate in the AGB phase are assumed. In this case, the mass density in the shell is inversely proportional to the square of distance from the central star. The number density distribution of dust grains is given by:

\begin{displaymath}N_{{\rm dust}}(r) = N_{0} \left(\frac{R_{{\rm in}}}{r}\right)^{2}, \end{displaymath} (3)

where N0 is the number density of dust grains at the inner boundary of the shell (= $R_{{\rm in}}$) and r is the radial distance from the center of the star.
  \begin{figure} \par\resizebox{10cm}{!}{ \rotatebox{0}{\includegraphics{fig4.eps}} } \end{figure} Figure 4: Reddening-corrected (using AV values in Table 4) spectral energy distribution of program stars. The full line indicates the calculated SED curve using parameters in Table 4. Open boxes denote IRAS flux density quality = 2 and open triangles indicate IRAS flux density quality = 1. Continued in Fig. 5.
Open with DEXTER


  \begin{figure} \par\resizebox{10cm}{!}{ \rotatebox{0}{\includegraphics{fig5.eps}} } \end{figure} Figure 5: SED of program stars. The same as Fig. 4.
Open with DEXTER

We assume the shell is optically thin at all wavelengths and the dust grains are in thermal equilibrium with the radiation from the central star. Light scattering by the dust grains is neglected. Frequency dependence of the dust grain absorption cross-section is assumed as a power-law function of wavelength, which allows an analytical form of the temperature distribution of the dust grains. For the dust grain parameters, we assumed the average grain radius $a = 0.1\,\rm\mu m$ and the mass density of the grains $\rho_{{\rm dust}} = 2.5\, {\rm g}\,{\rm cm}^{-3}$ and absorption efficiency $Q(\lambda) = 0.2 / \lambda\, [\mu{\rm m}]$. The temperature distribution is given by:

\begin{displaymath}T_{{\rm dust}}(r) = \left(\frac{R_{\rm star}}{2 r}\right)^{\frac{2}{5}} T_{{\rm star}}\,. \end{displaymath} (4)

Flux from one dust particle at r is given by:

 \begin{displaymath}F_{\lambda, {\rm dust.1}} = \pi B_{\lambda}\left(T_{{\rm dust}}(r)\right)\,Q(\lambda)\,a^{2}/ D^{2}. \end{displaymath} (5)

Integrating from the inner boundary to the outer boundary of flux of all dust particles, flux from the shell is given by:

\begin{displaymath}F_{\lambda, {\rm shell}} = \int_{R_{{\rm in}}}^{R_{{\rm out}}} F_{\lambda,{\rm dust.1}} \, N_{{\rm dust}}(r) {\rm d}r \,. \end{displaymath} (6)

In the fitting process, we fixed the stellar temperature $T_{{\rm star}}$(we used the same values in Table 3). Therefore fitting parameters are $R_{{\rm in}}$, $M_{{\rm dust}}$ and D in our model. In addition to these parameters, interstellar extinction AV was also treated as a parameter, as mentioned in Sect. 2.3. Changing these four input parameters, we calculated SEDs and we determined the parameters at a minimum of difference between the model and the observation.

We set the dust-to-gas mass ratio to be $5.0 \times 10^{-3}$. The dust temperature $T_{{\rm dust}}$ is determined at the inner boundary of the dust shell. Derived parameters of post-AGB candidates are shown in Table 4. The interstellar extinction-corrected SEDs with a model fitting curve are shown in Figs. 4 and 5.

The IRAS data points for which flux density quality was equal to 1 were not used for fitting. BD+39$^{\circ}$4926 was not included since it is not a IRAS source. For IRAS 02143+5852, near-infrared data points were not used for calculation because of the near-infrared excess.

The fitting parameter $R_{{\rm in}}$ is transformed to $t_{{\rm dyn}}$, the dynamical time of the dust shell, by being divided by the assumed expansion velocity of the dust shell. While an average value of $V_{{\rm exp}} = 15\,{\rm km}\,{\rm s}^{-1}$ is used, the observed values in PPNe are found over some range, for example from 10 to 20 ${\rm km}\,{\rm s}^{-1}$ (Hu et al. 1994), and this is translated into an uncertainty in the dynamical time of the dust shell. $M_{{\rm dust+gas}}$ in Col. 6 of Table 4 is obtained from $M_{{\rm dust}}$ divided by the dust-to-gas mass ratio. We assume the duration time of AGB mass-loss phase is to be 10000 yr, therefore one can easily derive the mass-loss rate in AGB phase by dividing $M_{{\rm dust+gas}}$ by 104 yr.

  \begin{figure} \par\includegraphics[angle=-90,width=8.5cm,clip]{fig6.eps} \end{figure} Figure 6: V0-[25] vs. [25]-[60] color-color diagram of the 16 observed IRAS sources which are post-AGB objects. Horizontal and vertical axes mean $ V_{0}+2.5\log({F}_{25\rm \,\mu m})$ and $2.5 \log ({F}_{60\rm \,\mu m}/{F}_{25\rm \,\mu m})$, respectively. V0 means V-band magnitude minus AV1 in Table 3. Open triangles have large position uncertainty (IRAS flux density quality = 1).
Open with DEXTER

Distances of post-AGB candidates were also estimated using the equation $M_{V} = ( V - A_{V} ) + 5 - 5 \log d $and $M_{V} = M_{{\rm bol}} - {\rm B.C}$. For all the post-AGB candidates, we used $M_{{\rm bol}}= -5.12$which correspond to $8000~L_{\odot}$. The value of the bolometric correction ${\rm B.C.}$ of each object was quoted from Schmidt-Kaler (1982) according to its spectral type. The AV values used are given in Table 3 (AV1). The last column of Table 4 shows estimated distances. The values of D calculated by two different methods agree within a factor of 1.5 except for IRAS 05089+0459. The observed optical to near-infrared energy distribution of this object is very similar to that of an early M star, requiring little extinction AV1 for this star. A strong contrast between the small extinction and the large far-infrared excess suggests an oblique torus or disk model for IRAS 05089+0459, but further observations are needed to confirm this hypothesis.

Carbon-rich objects show large differences between the two extinction values AV2 and AV3 (Table 4). We found that the carbon-rich objects all have a large AV. One possible explanation is that the circumstellar extinction law of the carbon-rich dust shell is very different from the assumed one, and another possibility is, as in the case of IRAS 05089+0459, the shape of the dust shell deviates from spherical symmetry.


   
Table 4: Derived fitting parameters of each Post-AGB candidate.
ID IRAS Name Tstar a Tdust tdyn Mdust+gas b AV3 c Dfit d DMV e
No.   [K] [K] [yr] [$M_{\odot}$] [mag] [pc] [pc]
2 02143+5852 6900 205 338 0.740 1.66 16900 16400
5 04296+3429 5550 193 354 0.420 4.36 7090 10200
6 05040+4820 8750 97 2460 0.254 1.31 4010 3930
7 05089+0459 3650 189 303 1.08 1.18 13900 40500
8 05113+1347 4590 185 358 0.240 2.34 8530 5470
10 05238-0626 7380 183 464 0.040 0.56 9310 10700
11 05341+0852 7060 210 323 0.090 4.12 7420 5620
13 05381+1012 4850 183 377 0.072 0.26 9910 11900
18 06338+5333 6250 267 167 0.002 0.00 5810 5860
19 06530-0213 7700 182 480 0.129 5.40 4880 3260
21 07131-0147 3330 171 371 0.554 1.74 13900 19200
22 07171+1823 13600 203 488 0.111 0.00 22800 22000
23 07253-2001 6900 206 335 1.270 0.00 22300 27500
24 07430+1115 4850 201 300 0.242 2.51 7460 8570
25 08187-1905 6630 135 937 0.168 0.38 5070 4540
26 23304+6147 5200 173 453 0.366 4.33 4880 4310
a  Same values in Table 3.
b  Correspond to mass-loss rate $\dot{M}$ (=$\,M_{\mathrm{dust+gas}} \times 10^{-4}~M_{\odot}~\mathrm{yr}^{-1}$).
c  Interstellar extinction derived from the model fitting.
d  Derived distances using the model fitting ( $L = 8000~L_{\hbox{$\odot$ }}$) .
e  Derived distances assuming $M_{V} = M_{\mathrm{bol},8000\hbox{$\odot$ }}{-}\mathrm{B.C.}$

4 Discussion

4.1 Comparison of the stellar temperature with dynamical time of the dust shell

The observed stars have double peaked SEDs, suggesting the existence of the detached cold shell around the central star. Our model calculations actually showed the  $R_{{\rm in}}$ larger than $1.5 \times 10^{16}\, {\rm cm}$, about 100 times larger than the usual radial size of the dust-forming region of the AGB stars. Similar values were estimated by Hrivnak et al. (1989). As expected from the variety of relative strength of the infrared peak to the visible one, the optical thickness of the dust shell changes from object to object. Figure 6 shows the observed stars on the V0-[25] to [25]-[60] diagram. A sequence of stars along the horizontal axis in the right part of the diagram presents the change of the optical depth of the shell.

Figure 7 compares the stellar temperature with the dynamical time of the dust shell. The theoretical evolutionary tracks of the post-AGB stars with different core masses by Schönberner (1983) and Blöcker (1995) are plotted in Fig. 7. A common feature of all those theoretical tracks is that the detached shell appears when the stellar temperature goes up to 5000 or 6000 K. This means that they assumed AGB superwind mass-loss ( $\dot{M} \sim 10^{-4}~M_{\odot}~{\rm yr}^{-1}$) terminates at these temperatures. In general, the agreement between the observational results and the theoretical evolutionary tracks is satisfactory. However, the observations appear to fit with a somewhat lower core mass indicating a somewhat lower luminosity than that used in the models. Six objects among our samples are distributed below the line of $M_{\rm core}=0.546~M_{\odot}$; two of them, IRAS 05089+0459 and 07131-0147, are peculiar M type stars while the remainders are all G type. The temperatures we used here were cited from different authors and there are uncertainties in the spectral sub-class. Overall, our observational results are in agreement with the theoretical post-AGB models (Fig. 7) (Blöcker 1995; Schönberner 1983). The same kind of figures are shown by van der Veen et al. (1989) and Schönberner & Blöcker (1993). Figure 7 is consistent with their figures.

  \begin{figure} \par\includegraphics[width=9cm,clip]{fig7.eps} \par\end{figure} Figure 7: Comparison of the stellar temperature with dynamical time of the dust shell, assuming constant expansion velocity ( $15\,{\rm km}\,{\rm s}^{-1}$). Dashed and dotted lines are theoretical evolution of hydrogen-burning post-AGB models with core mass 0.546- $0.625~M_{\odot}$. The dashed lines are cited from Blöcker (1995) and the dotted lines are cited from Schönberner (1983).
Open with DEXTER

In Fig. 7, two stars exist apart from the main group. One is IRAS 05040+4820 and the other is IRAS 08187-1905. In spite of their relatively low stellar temperatures, the dynamical ages of their dust shells are large. These dust shells most likely will disperse into the interstellar space before the stellar temperature rises to ionize the surrounding gases. We note that they are on the evolutionary track of $M_{\rm core}= 0.55~M_{\odot}$indicating low mass for their parent stars, probably one solar mass or less. Renzini (1981) predicted the fate of a low mass star to become a white dwarf bypassing the PN stage. Scarcity of PNe in globular clusters supports his hypothesis, but no direct evidence has been found as far as we know. The above two IRAS sources are the first sample of Renzini's "lazy'' AGB remnants.

   
4.2 Notes on individual objects

IRAS 02086+7600

It is a member of the dark cloud L1333, which is a molecular cloud in Cassiopeia (Obayashi et al. 1998). It is identified with CO core No. 4 and is most likely a young stellar object (YSO) embedded in a CO core. The far-infrared luminosity of this source was estimated to be about 1 $L_{\odot}$ at a distance of 180 pc. The 12 $\mu$m to 25 $\mu$m flux ratio is significantly smaller than that of a typical T Tau star.

Van de Steene & Pottasch (1995) considered it as a possible planetary nebula, however they have not found radio continuum emission from this source. Slysh et al. (1994) searched for OH maser emission and it was detected at 1667 and 1665 MHz with a velocity of $3.1\,{\rm km}\,{\rm s}^{-1}$. Preite-Martinez (1988) considered it as a possible new PN. It may be a ultra-compact ii II region, or a post-AGB star, or a YSO. We need BVRI observations and a low resolution spectrum to understand the evolutionary stage of this object.

IRAS 02143+5852

Manchado et al. (1989) and García-Lario et al. (1997) made JHK photometric observations and our JHK photometric magnitudes are in agreement with theirs. Omont et al. (1993) considered it as a carbon-rich PPN, however no CO and HCN emission is detected. It is a F type post-AGB supergiant (Meixner et al. 1999). Meixner et al. (1999) imaged it at 11.7 $\mu$m and it is not resolved.

IRAS 02528+4350

Because of its IRAS colors, it was classified as a post-AGB candidate. However, Nakanishi et al. (1997) recently found it to be a galaxy with a redshift of $33\,678\,{\rm km}\,{\rm s}^{-1}$. Crawford et al. (1996) also consider it as an ultraluminous infrared galaxy, however no radio emission is detected. The JHK photometric observations of Manchado et al. (1989), García-Lario et al. (1997) are in agreement with ours. They and Van de Steene & Pottasch (1995) considered it as a PPN/PN. However in the light of recent work of Nakanishi et al. (1997), it is a galaxy and not a post-AGB star.

IRAS 04269+3550

Van de Steene & Pottasch (1995) considered it as a PN candidate, however they did not detect radio continuum emission. Wouterloot et al. (1993) searched for H2O, OH, CH3OH and CO and did not detect any of these emissions. Preite-Martinez (1988) considered it as a possible new PN. From our JHK photometry and IRAS data, the SED of this object seems to be a single peaked curve rather than a double peaked curve. It maybe a post-AGB star obscured by the thick dust-shell. To confirm the object type, we need other observations, such as spectroscopy.

IRAS 04296+3429

The unidentified emission feature at 21 $\mu\rm m$ was first discovered in the IRAS LRS spectra of four carbon-rich post-AGB stars (Kwok et al. 1989), including IRAS 04296+3429. It also possesses a strong 30 $\mu$m emission (Szczerba et al. 1999). In the optical spectrum emission bands (0, 0) and (0, 1) of the Swan system of the C2 molecule were detected (Klochkova et al. 1999). The effective temperature of the star from high resolution spectrum was estimated to be 6300 K. The star is metal-poor ([Fe/H]=-0.9) and overabundant in carbon and s-process elements (Decin et al. 1998; Klochkova et al. 1999), van Winckel & Reyniers (2000) similar to that of other 21 $\mu$m post-AGB stars such as IRAS 05341+0852 (Reddy et al. 1997).

IRAS 05113+1347

It is a 21 $\mu$m carbon-rich post-AGB star (Kwok et al. 1995). It shows C2, C3 and 11.3  $\mu\rm m$ emission features.

IRAS 05170+0535

Van den Ancker et al. (1998) classified it as a low mass young stellar object. Coulson et al. (1998) considered it as a Vega-excess star. The G0Ve spectrum and Hipparcos parallax suggests that it is not a post-AGB star. It is most likely a young G dwarf at the end of the T Tauri phase. The Hipparcos parallax yields a distance of 180 pc and a luminosity of about 1.23 $L_{\odot}$.

IRAS 05238-0626

Its spectral type is F2II (Reddy & Parthasarathy 1996). The photometric observations made by García-Lario et al. (1997) and Torres et al. (1995) are in agreement with our observations.

IRAS 05341+0852

It is a post-AGB star with 21 $\mu$m emission. Reddy et al. (1997), van Winckel & Reyniers (2000) found it to be metal-poor and overabundant in carbon and s-process elements. It appears to have evolved from the AGB carbon star stage to post-AGB stage only recently. Our BVRI data of this star is in good agreement with the photometric data of Hrivnak & Kwok (1999). However, our JHK data differ by about 0.2 mag from the JHK data reported by Hrivnak & Kwok (1999). They classified the low resolution spectrum of this star and assigned a spectral type G2 0-Ia.

IRAS 05355-0117

Its optical counterpart is HD 290764. Its spectral type is A5III (Schild & Cowley 1971). It is considered as a $\delta$ Scuti star with a full amplitude of 0.016 mag. However, the presence of cold detached dust shell is not consistent with the $\delta$ Scuti type variability. It is likely a pre-main-sequence star in the Orion stellar ring. High resolution spectroscopic analysis may help us to understand the evolutionary stage of this star.

IRAS 06013-1452

It is a high galactic latitude Ae star. Thé et al. (1994) listed it in the catalog of Herbig Ae/Be stars. García-Lario et al. (1997) also classified it as a Herbig Ae/Be star.

IRAS 06059-0632

It is a B3 star in the direction of Orion OB1 association (de Geus et al. 1990). It is listed in the Hipparcos cataloge ( $\pi = 1.68\,{\rm mas}$). This source is probably a nearby pre-main-sequence star and not a post-AGB supergiant.

IRAS 06060+2038

It is a low galactic latitude IRAS source with B1V star in the direction of Gemini OB1 molecular cloud complex. It is in the Sharpless 252 which is an extended ii II region (Haikala 1994).

IRAS 06284-0937

Van den Ancker et al. (1998) considered it as a Herbig Ae/Be star. It is in the Hipparcos cataloge ( $\pi = 4.6\,{\rm mas}$). It may be a variable star (NSV 2998).

IRAS 07077-1825

It is a O6 star listed in the LS catalog as LS 207 (Reed & Beatty 1995) and is in the direction of the Sharpless 301 which is an extended ii II region (Moffat et al. 1979).

IRAS 07131-0147

It is a bipolar object with a M5III central star (Scarrott et al. 1990). The evolutionary status of this star is not clear. The M5III spectral type suggests that it may be a first ascent red giant and may not be a post-AGB star.

IRAS 07171+1823

It is found to be very low excitation planetary nebula with a hot (B-type) post-AGB central star (Vijapurkar et al. 1998). It shows nebular emission lines of [ii II] and [ii II]. The Balmer lines are also in emission. It is a high galactic latitude hot-post-AGB star.

IRAS 07430+1115

It is a high galactic latitude carbon-rich post-AGB star (Hrivnak & Kwok 1999). Our BVRI photometry of this star is in agreement within 0.25 mag with the photometric data reported by Hrivnak & Kwok (1999). The difference may be due to small amplitude light variations of this star.

5 Conclusions

From the BVRIJHK photometry of 27 IRAS sources with far-infrared colors similar to planetary nebulae, 17 objects are found to be most likely post-AGB stars. From the analysis of SED we have obtained some of the parameters of their circumstellar dust shells. We plotted our objects on the diagram of the stellar temperature against the dynamical time of the dust shell. Comparing with the theoretical evolutionary tracks of the post-AGB stars, two objects are classified as slowly evolving post-AGB stars which may evolve into white dwarfs without experiencing the PNe phase.

Acknowledgements
We would like to thank Dr. B. J. Hrivnak for careful reading of this manuscript and valuable comments. We used the SIMBAD database operated at CDS, Strasbourg, France. M.P. thanks Prof. Keiichi Kodaira, Prof. Shuji Deguchi, Prof. Norio Kaifu and Prof. Hiroshi Karoji for their kind encouragement, support and hospitality.

References

 

Copyright ESO 2002