Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 322, 229-233 (1997)

Previous Section Next Section Title Page Table of Contents

2. Observations and data analysis

SLX 1735-269 was observed by the ASCA satellite on 1995 March 15. The observing instruments on ASCA consist of four identical thin foil mirrors which focus X-rays onto two Solid-state Imaging Spectrometers (SIS) and two Gas Imaging Spectrometers (GIS). The GIS observed in the PH mode where the on board CPU calculates the event position and discards background events using rise time and positional information. SIS data was acquired at a high bit rate and in the Faint data mode, i.e. the pixel position of the center of the event is given along with the pulse height recorded in the nine surrounding pixels. The data were converted to bright mode using the same algorithm as used on the spacecraft. Only one of the four CCD chips available in the SIS sensor was exposed. Screening of bad events was performed using the ASCA GOF recommended values (Day et al. 1995). The minimum elevation angle was set to 10 for both the SIS and GIS instruments.

2.1. Spectral analysis

Spectra for each instrument were extracted from circular regions centered on the source. The radii of these regions were [FORMULA] for the SIS and [FORMULA] for the GIS images. Background spectra were obtained from a region of the same size using the background files made available by ASCA GOF. Because the diffuse X-ray emission from the Galactic ridge is high towards the position of SLX1735-269 (e.g. Koyama 1994), the GIS background used was checked against the background obtained from a source free region in the GIS image. In this latter case, background spectra were extracted from an annulus centered on the source within which source emission is absent. No notable effects were seen due to diffuse emission; spectra obtained by subtracting one or the other background gave statistically equal results when compared.

The sensitivity of the GIS instrument is more than 10% of its maximum value in the energy range of 0.8-12 keV. However, because no good calibration data base above 10 keV is available for ASCA, GIS spectral analysis was restricted to 0.8-10 keV. The pass band used for the SIS data was 0.6-10 keV, this avoids the O K-edge in the SIS response but the Al and Si K-edges, along with the Au M-edge, due to the instrument are present in the 2-3 keV region. However, no obvious absorption features due to the source are discernible in this region so detailed analysis was not necessary between 2-3 keV. The models employed in fitting the data were a powerlaw, a thermal bremsstrahlung, and a comptonization spectrum (Sunyaev and Titarchuk 1980).

Absorption by the interstellar medium was modeled using the photoelectric absorption cross-sections of Morrison & McCammon (1983). Best fit parameters for the powerlaw and thermal bremsstrahlung are summarized in Table 1. The [FORMULA] are relatively high but this is probably due to imprecise calibration at lower energies. The powerlaw model plus absorber provides a more satisfactory fit to the data although only at a level of a few percent. To fit the comptonization spectrum, the electron temperature was fixed at 26 keV following the results from SIGMA (Goldwurm et al. 1996) and the resulting optical depth was [FORMULA] in agreement with that obtained for the SIGMA data. This can be clearly seen in Fig. 1 where the best fit comptonization plus absorber model is plotted along with the ASCA and SIGMA data and if the SIGMA points are renormalized by a factor of [FORMULA]. The jump between the ASCA and SIGMA spectra might be due to the fact that the latter is averaged over several years. On the date of the ASCA observation the SIGMA telescope was not operating in pointing mode. On this occasion SLX 1735-269 might have been in a "low" state (see Fig. 2 of Goldwurm et al. 1996). The comptonization model fits the data to within the same confidence level as the powerlaw. In the same figure is plotted for comparison the best fit thermal bremsstrahlung model obtained from the TTM experiment.


Table 1a. Best fit parameters to spectra for powerlaw plus absorber model


Table 1b. Best fit parameters to spectra for bremsstrahlung plus absorber model

[FIGURE] Fig. 1. Spectrum of SLX 1735-269 obtained from the GIS3 detector and the corresponding best fit Sunyaev-Titarchuk plus absorber model. The SIGMA data is also plotted, circles, along with the corresponding best fit powerlaw model. The dot dashed line represents the best fit bremsstrahlung plus absorber model derived from the TTM data.

The K-iron emission line centered at [FORMULA] keV and typically with equivalent width [FORMULA] eV and intrinsic line width [FORMULA] keV frequently observed in the spectra of binary pulsars (Nagase 1989) does not appear in the data. In fact, a fit between [FORMULA] -10 keV by a powerlaw plus gaussian line, centered at 6.4 keV with [FORMULA] and 0.5, model does not reveal this feature nor show any significant improvement to a simple powerlaw fit. Upper limits to the equivalent line width have been determined, using the SIS data, assuming a gaussian line shape, by fixing [FORMULA] and varying the intenstity until [FORMULA] became unacceptable in an F-test at a confidence level of 95%. For [FORMULA] the equivalent line width is [FORMULA] eV while for [FORMULA] it is [FORMULA] eV.

2.2. Source position

The source position was derived from the SIS images which provide a more reliable position than the GIS. Indeed, the SIS images are solely limited by the point spread function and the SIS focal plane calibration is more accurate than that of the GIS (Gotthelf 1996). ASCA situates SLX 1735-269 at R.A. = 17h 35m 8.9s, Dec. = [FORMULA] [FORMULA] (equinox 1950) with error radius of [FORMULA]. This position is compatible with the original position from Spacelab 2 (Skinner et al. 1987), the Einstein source 1ES 1735-26.9, (Elvis et al. 1992), the SIGMA position, (Goldwurm et al. 1996) as well as with the Rosat PSPC catalogue position (Zimmermann 1994). The recently available Rosat All-Sky Survey Bright Source Catalogue (1RXS) position, R.A. = 17h 35m 9.43s, Dec. = [FORMULA] [FORMULA] [FORMULA] with an error box of radius [FORMULA] (Voges W. et al. 1996) falls within the ASCA determination. All these X-ray sources are undoubtly the same object.

2.3. Timing analysis

The data were submitted to a timing analysis employing the epoch folding method described by Leahy et al. (1983). The GIS data has a higher time resolution than the SIS (62.5 ms and 4s respectively) with roughly equal integration time. Data from both instruments was submitted to the timing analysis separately. The number of bins chosen was 10. The original GIS arrival times were binned into 1 s intervals. Because a pulsation period has never been determined for SLX 1735-269, an attempt was made to limit the number of trials. The protocol followed was: first a systematic period search was effected on the arrival times from the GIS2 detector in the range 10 to 1000 s and, from the SIS0 detector over 40-1000 s. Any possible detection at a confidence level [FORMULA] 90% was noted and, then checked for, using the data from the GIS3 or SIS1 detector in order to confirm that period. Proceeding as above, no periodic behaviour was found for SLX 1735-269 in the range searched. Assuming a sinusoidal pulse profile of amplitude A, the instantaneous count rate is written as, [FORMULA]. Upper limits for the parameter A, not accounting for background counts and imposing a confidence level for detection and sensitivity greater than 90%, are listed in Table 2.


Table 2. Upper limits to pulsation A

When binned into 1s intervals, the GIS data does not easily allow period searches below 10s by epoch folding because, in this case, too few photons per phase bin would result. Likewise, the SIS data only permitted period searches above [FORMULA] s. Thus, to search for periods in the millisecond to second range a more appropriate method is provided by Fourier analysis. The method applied here is detailed in Leahy et al. (1983). A total of 225 continuous sections of length 64 s, within a total time span, including gaps, of 43036 s from the GIS3 detector were analyzed to obtain the total power spectrum. Imposing a confidence level for detection of 90%, no pulsation period was found. The upper limit to pulsations in the period range 125 ms to 64 s is listed in Table 2. Pulse periods for binary X-ray pulsars range from [FORMULA] to [FORMULA] s with roughly two thirds of known pulsars having a period exceeding 10 s (Nagase 1989, Mereghetti et al. 1996). Solitary neutron stars have periods less than one second, at most a few seconds. Thus a fairly reasonable range of possible periods was searched.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 30, 1998