Astron. Astrophys. 347, 473-477 (1999)

1. Introduction

The X-ray source 4U1700+24 was firstly discovered by the Uhuru satellite (Forman et al., 1978) and reported as an X-ray emitter also in the 3A and 1H catalogs (Pye et al., 1983; Wood et al., 1984). Its optical counterpart was identified as the red giant HD 154791 by Garcia et al. (1983): these authors stressed that it was the first (and, to date, it remains the unique) case of hard X-rays detection for an M giant having a "completely normal" optical spectrum.

Until now, HD 154791/4U1700+24 has been observed in the X, UV, optical and infrared wavelength ranges. However, a definite classification of the system, based on both morphological and physical characteristics, is still controversial. This situation is caused by the very few and sparse data published to date and by the fact that the observations of this object available until now show a significant variability at all wavelengths.

Actually, Garcia et al. (1983), from data obtained by HEAO1 and Einstein Observatory, found a 2-11 keV X-ray luminosity of 4U1700+24 (for a distance of 830 pc -see later-) ranging from 10³³ to 1.410³⁴ erg s^-1; the column density value was N_H=10²² cm^{- 2} and the derived temperatures of 15 keV or more. On the other hand, the X-ray spectrum successively measured by Dal Fiume et al. (1990) using EXOSAT, in the framework of a multiwavelength campaign performed in 1985, was significantly softer and much more absorbed: the source intensity was lower by a factor 6 and the N_H value was much higher: (2.1-3.8)10²² cm^-2. Moreover, a time variability with an average fractional variation in flux of 50% was always evident at this flux level. Finally, in a systematic search for long-term variability in a large sample of X-ray sources observed with Einstein Observatory, Maccacaro et al. (1987) evidenced a variation of a factor 2 in 203 days in the flux emitted by the system.

Strong variability is also evident in the UV range from the comparison of the Garcia et al.(1983) and Dal Fiume et al. (1990) results. Strong emission lines of C IV  1550 Å, N V  1238 Å, Mg II  2800 Å, very unusual among M giants, were observed by Garcia et al. (1983) and were present also in a previous archive observation, performed in February 1981. The C IV and N V lines later disappeared and the Mg II line intesity decreased of a factor 2 (Dal Fiume et al.,1990).

In the optical range too a variation of the system behaviour points out. Actually, Garcia et al. (1983) found that the measured colors were consistent with those from a normal M giant, except that the star is 0.3 magnitudes brighter than the normal in B. These authors also found, from low dispersion measurements, a normal M type giant spectrum (M3II) with none of the emission lines (H, H, He II  4686 Å) characteristic of accretion driven X-ray sources. The Ca II K doublet was in emission and its flux level, measured from ten high dispersion echelle spectra, was found to be high, but not unusual for an M giant. From its width at the base, a luminosity class of II and a distance of 830100 pc was deduced. The search for a secondary star gaves ambiguous conclusions and it was argued that any eventual companion must have photospheric lines in the optical region much weaker than those of the primary and must induce the M giant to have a radial velocity variation less than 3 Km s^-1. On the other hand, Dal Fiume et al. (1990) found optical results consistent with a spectral type (M5-M7) later than that proposed by Garcia et al. (1983) probably indicating a real change in the temperature of the star atmosphere. The Ca II  3933 Å emission was completely desappeared two years after the Garcia et al. (1983) observations. No variability on long time scale was found in V and B bands, with only some evidence in U band. Only occasionally, small amplitude (0.15 mag) variations on short time scale (weeks) were detected.

Schaefer (1986) included this system in its IRAS observations of binaries with compact objects obtaining IR fluxes lower than most of the X-ray binaries. He thus argued that it is not clear if a compact object is present in this system. On the other hand, Brown et al. (1990) found the presence of the HeI 10830 Å spectral line, both in absorption and emission; they thus considered HD 154791 as a red giant with a probable hot companion.

Two possible scenarios can explain the just described observational data: an M giant star with an X-ray emitting corona or an M giant with a compact accreting companion.

Temperatures of 15 kev or more are typical of those found in X-ray binaries and much higher than that expected from an hot corona; also the X-ray flux is greather than the one given by coronal emission at a distance of 800 pc; the erratic time variability is a very common feature in accreting compact objects of very different classes (LMXBs, HMXBs, low luminosity HMXBs); modulations with orbit could explain the observed variations in the neutral hydrogen column density (N_H). On the other hand, a spectral evidence of any companion or a clear variation of the radial velocity are not observed until now and no periodicities have been found to date in the X-ray flux of this source.

Furthermore, Tomasella et al. (1997) have recently reported optical observations of HD 154791 during an outburst of 4U1700+24 occurred in November 1997 (Remillard, 1997, electronic announcement reported by Tomasella et al., 1997), when its 2-10 KeV flux reached the historical maximum of 35 mCrab. They found that no change in high resolution spectra were visible and no emission lines appeared in the H-H region. This fact rules out the possibility that part of the optical flux of HD 154791 comes from an accretion disk.

© European Southern Observatory (ESO) 1999

Online publication: June 30, 1999
helpdesk.link@springer.de