AstroSat observations of eclipsing high mass X-ray binary pulsar OAO 1657-415
Gaurava K. Jaisawal, Sachindra Naik, Prahlad R. Epili, Birendra Chhotaray, Arghajit Jana, P. C. Agrawal
JJ. Astrophys. Astr. (0000) :
AstroSat observations of eclipsing high mass X-ray binary pulsarOAO 1657–415
Gaurava K. Jaisawal , Sachindra Naik , Prahlad R. Epili , Birendra Chhotaray , ArghajitJana and P. C. Agrawal National Space Institute, Technical University of Denmark, Elektrovej 327-328, DK-2800 Lyngby, Denmark Astronomy and Astrophysics Division, Physical Research Laboratory, Navrangapura, Ahmedabad - 380009,Gujarat, India School of Physics and Technology, Wuhan University, Wuhan 430072, China Indian Institute of Technology, Gandhinagar - 382355, Gujarat, India Department of Astronomy and Astrophysics (Retired), Tata Institute of Fundamental Research, Colaba, Mumbai- 400005, India * Corresponding author. E-mail: [email protected] received *; accepted *
Abstract.
We present the results obtained from analysis of two
AstroSat observations of the high mass X-raybinary pulsar OAO 1657-415. The observations covered 0.681–0.818 and 0.808–0.968 phases of the ∼ Keywords. stars: neutron — pulsars: individual: OAO 1657–415 — X-rays: stars.
1. Introduction
OAO 1657–415 is an accreting high mass X-ray binarypulsar, discovered with Copernicus satellite in 1978(Polidan et al. 1978). The X-ray pulsations from theneutron star were detected at 38.2 s using HEAO A-2 observations (White & Pravdo 1979; Parmar et al.1980). Later, the system was identified as an eclips-ing binary by using 1991 and 1992 observations withthe Burst and Transient Source Experiment (BATSE)onboard Compton Gamma Ray Observatory (CGRO;Chakrabarty et al. 1993). This study also revealedthe orbital period of the system to be P orb = ff orts in the optical band, no counterpart was detected up to a limitings magnitude of V < Chandra
X-ray error box (Chakrabarty et al.2002). The spectral class of the donor star was refinedto be a Ofpe / WNL type star which is thought as a tran-sitional object between the main sequence and Wolf-Rayet stars (Mason et al. 2009; Mason et al. 2012). Ob-served X-ray eclipses and massive nature of the com-panion established the binary system as an eclipsinghigh mass X-ray binary. The source distance is mea-sured to be 4.4–12 kpc (Chakrabarty et al. 2002; Ma-son et al. 2009), consistent with the measurement of7.1 ± Gaia data suggests a relatively lower distance of2.2 + . − . kpc based on the parallax angle of the opticalcompanion (Malacaria et al. 2020).In most of the high mass X-ray binary (HMXB)systems, a neutron star resides as the compact object.The mass accretion onto the neutron star in these sys-tems takes place either via stellar wind accretion, Be- © Indian Academy of Sciences 1 a r X i v : . [ a s t r o - ph . H E ] J a n J. Astrophys. Astr. (0000) : disk accretion, or Roche-Lobe overflow from the opti-cal companion (see e.g. Reig 2011; Walter et al. 2015).Based on the mass accretion mechanisms and the na-ture of donor star, the HMXBs can be classified intoBe-X-ray binaries (BeXBs) and supergiant X-ray bina-ries (SGXBs). The compact object in BeXBs co-rotatesaround a non-supergiant companion of class III-V, andaccretes directly from a circumstellar disk of the Bestar. On the other hand, the SGXBs consist of a super-giant optical companion of luminosity class I-II. Themass accretion in SGXB systems occurs through thestellar wind accretion (wind-fed accretion) or via an ac-cretion disk formed after the capture of stellar wind orRoche-Lobe overflow (disk-fed accretion). Some of theSGXBs are also known to be eclipsing systems due toedge-on periodic obscuration of the compact object bythe super-giant companion. Typical X-ray luminosityof the neutron star in SGXBs is in the range of 10 –10 erg cm − s − (Mart´ınez-N´u˜nez et al. 2017). Thesource luminosity can also vary by a factor of 5-100within a ks time-scale due to flaring activities (see, e.g.F¨urst et al. 2010, Naik, Paul & Ali 2011, Jaisawal et al.2020).Various sub-classes of HMXBs represent a uniqueposition in spin period vs. orbital period diagram (Cor-bet diagram; Corbet 1986). The BeXB systems show astrong correlation between the orbital and spin periods,whereas the wind-fed SGXBs are distributed in a hori-zontal line. There are, however, three disk-fed SGXBsviz. Cen X-3, SMC X-1 and LMC X-4, that exhibitan anti-correlation between the orbital and spin peri-ods in the Corbet diagram. In exception to the knownpattern, OAO 1657-415 occupies an intermediate posi-tion in the Corbet diagram, between the wind-fed anddisk-fed SGXBs (Chakrabarty et al. 1993; Jenke et al.2012).Since discovery, the pulse period of OAO 1657–415 changes stochastically at a rate of ˙ ν ∼ × − Hz s − (Baykal 1997; Bildsten et al.1997; Baykal 2000). Steady spin-up and spin-downpatterns, as seen in Cen X-3, are also observed in thesystem (Bildsten et al. 1997). The changes in the spinepisodes, however, can not be explained by the theoryof torque reversal without considering the formationof a transient accretion disk (Baykal 1997). Basedon a positive correlation between X-ray luminosityand the spin frequency during a spin-down phase in1997, the presence of a prograde disk was suggested(Baykal 2000). On investigation of almost two decadesof long-term spin evolution with BASTE and Gamma-Ray Burst Monitor ( Fermi / GBM) observations, twoaccretion modes are inferred in this system (Jenkeet al. 2012). The first mode is due to the disk-windaccretion (formation of an accretion disk after stellar wind accretion) where a stable accretion disk producesa correlation between flux and spin-up of the neutronstar. In the latter case, a direct stellar wind accretionproduces almost no correlation between the flux andspin-down parameter of OAO 1657–415 at a lesseraccretion rate (Jenke et al. 2012). Recently, Kim& Ikhsanov (2017) proposed a magnetic levitatingdisk hypothesis to explain the spin evolution inOAO 1657–415 .Being located at a low galactic latitude, OAO 1657–415 is highly absorbed with a column density of10 cm − (Polidan et al. 1978, Kamata et al. 1990).The energy spectrum of the pulsar can be describedby a power law continuum with an exponential highenergy cuto ff along with a soft excess and prominentemission lines at 6.4, 6.7, and 7.1 keV (Kamata et al.1990; Audley et al. 2006; Barnstedt et al. 2008; Prad-han et al. 2014; Jaisawal & Naik 2014; Pradhan, Ra-man & Paul 2019). ASCA observations were performedon 1994 March 22 and 1997 September 17 betweenorbital phase ranges of -0.001–0.074 and -0.21–0.11(mid-eclipse time as phase zero), respectively (Audleyet al. 2006). Presence of a dust scattered X-ray halo inOAO 1657–415 was suggested from these data. As ex-pected, the observed source flux was very low near themid-eclipse region with a dominant 6.4 keV line com-pared to the 6.7 keV line in the data from the first
ASCA observation. The second
ASCA observation covered en-tire eclipse along with out-of-eclipse phases before andafter eclipse. The energy continuum in this phase wasfound to be absorbed along with detection of 6.4 and7.1 keV iron emission lines (Audley et al. 2006).
Suzaku observation of the source in September2011, covering 0.12–0.34 orbital phase range, revealedflaring activities in the soft and hard X-ray light curves(Jaisawal & Naik 2014; Pradhan et al. 2014). De-tailed time-resolved spectroscopy of the
Suzaku datasuggested the accretion of clumpy material as the causeof flare-like episodes during the observation. Strong6.4 and 7.1 keV lines were also detected. These linesmostly originated from the neutral and ionized ironatoms within the accretion radius of 19 lt-sec (Jaisawal& Naik 2014). Using
BeppoSAX observation, a pres-ence of a cyclotron absorption line was suggested at36 keV (Orlandini et al. 1999). However, later studieswith
INTEGRAL and
Suzaku did not confirm the feature(Barnstedt et al. 2008; Pradhan et al. 2014; Jaisawal &Naik 2014).In the present paper, we study the properties of thesource by using two
AstroSat observations, carried outat mid and late orbital phases of the orbit (with mid-eclipse time as phase zero) in 2019 . Data analysis ispresented in Section 2, followed by timing and spectralresults in Section 3 and 4. The discussion and conclu- . Astrophys. Astr. (0000)000
AstroSat observations, carried outat mid and late orbital phases of the orbit (with mid-eclipse time as phase zero) in 2019 . Data analysis ispresented in Section 2, followed by timing and spectralresults in Section 3 and 4. The discussion and conclu- . Astrophys. Astr. (0000)000 :
Table 1 . Log of observations of OAO 1657–415 with
AstroSat . ObsID Start Date Expo. φ orb (MJD) (ks) XX XX Here XX stands for A05 205T01. φ orb represents the orbital phaseof the binary system. sion are summarized in Section 5.
2. Observations and Data Analysis
AstroSat is the first Indian multi-wavelength astronomi-cal satellite launched by Indian Space Research Organi-zation on 28 September 2015 (Agrawal 2006, Singh etal. 2014). It provides a broad-band coverage from op-tical to X-ray bands for exploring the nature of the cos-mic sources. There are five sets of instruments such asSoft X-ray Telescope (SXT; Singh et al. 2017), LargeArea X-ray Proportional Counters (LAXPCs; Agrawalet al. 2017, Antia et al. 2017), Cadmium Zinc TellurideImager (CZTI; Rao et al. 2017), a Scanning Sky Moni-tor (SSM; Ramadevi et al. 2018), and Ultraviolet Imag-ing Telescope (UVIT; Tandon et al. 2017), onboardthe satellite. In this paper, we study two observationsof OAO 1657–415 with SXT and LAXPC instruments.These observations were performed on 31 March and4 July 2019, covering orbital phase ranges of 0.681–0.818 and 0.808–0.968, respectively. The orbital phaserange is calculated based on the ephemeris of the binarysystem provided by Jenke et al. (2012). In our study,the CZTI data from both the epoch of observations arenot used as the source was faint for the detector. TheUVIT was not operational during these epochs. Thelog of the observations are given in Table 1.The SXT is a soft X-ray focusing telescope con-sisting of a CCD detector and sensitive in 0.3–8 keVenergy range. The e ff ective area and energy resolutionof SXT is 128 cm and 5–6% at 1.5 keV and 22 cm and 2.5% at 6 keV, respectively. The observations ofOAO 1657–415 were carried out with SXT operatingin photon counting mode, yielding a time resolution of2.4 s. We used standard pipeline for SXT data reduc-tion and merging tool sxtevtmergertool provided bythe AstroSat
Science Support Cell (ASSC ). As the pul-sar was faint for SXT during the first observation, thedata were used for spectral analysis only. The source http://astrosat-ssc.iucaa.in/ Obs-2Obs-2 B A T I n t e n s i t y ( m C r a b ) B A T I n t e n s i t y ( m C r a b ) M A X I r a t e ( m C r a b ) M A X I r a t e ( m C r a b ) Obs-1Obs-1 B A T r a t e ( m C r a b ) B A T r a t e ( m C r a b ) Obs-2Obs-2 B A T I n t e n s i t y ( m C r a b ) B A T I n t e n s i t y ( m C r a b ) M A X I r a t e ( m C r a b ) M A X I r a t e ( m C r a b ) Obs-1Obs-1 B A T r a t e ( m C r a b ) B A T r a t e ( m C r a b ) Figure 1 . Long term monitoring light curves of OAO 1657–415 with MAXI (blue) and BAT (shaded) in 2-20 keV and15-50 keV ranges, respectively. Arrows in both the panelsrepresent the dates of the
AstroSat observations. The pulsarappears to be weak during the first
AstroSat observation (toppanel). However, the second observation caught the sourcewhile changing from a relatively bright to faint phase of thebinary orbit (bottom panel). spectrum was extracted from a 5 arcmin circular regioncentered at the source coordinate on the SXT chip using
XSELECT package. On the other hand, the light curvesand spectra were extracted from the second observa-tion by considering a source circular region of 5 arcmin.The background spectrum was obtained from a sourcefree region on the SXT chip.The three LAXPCs are sensitive to X-ray photonsin the 3–80 keV range and provide a total e ff ective areaof 8000 cm at 15 keV. The time and energy resolu-tion of the LAXPC units are 10 µ s and 12% at 22 keV,respectively. During both the observations, data fromLAXPC20 were considered in our analysis. The datafrom LAXPC10 and LAXPC30 units were not useddue to the presence of high background and gain is-sues with the instrument during observations (Antia etal. 2017). Using the standard data analysis routines( LAXPCsoftware ), the event mode data are analyzedto obtain the source light curves and spectral products.The LAXPC background products are obtained fromthe observation using standard routines, recommendedby the team. A systematic uncertainty of 2% is also
J. Astrophys. Astr. (0000) : (a)(a)
Orbital Phase (Orbital Phase (φφ orborb )) R a t e R a t e (b)(b) R a t e R a t e H R H R Seg-IVSeg-IVSeg-IIISeg-IIISeg-IISeg-IISeg-ISeg-I EclipseEclipse (a)(a)
Orbital Phase (Orbital Phase (φφ orborb )) R a t e R a t e (b)(b) R a t e R a t e H R H R Figure 2 . Light curves from first and second
AstroSat / LAXPC20 observations of OAO 1657–415 on 31 March 2019 and 4July 2019 are shown in left and right panels, respectively. The top and middle panels on both sides represent light curvesin 3–10 keV and 10–80 keV ranges, respectively. The hardness-ratio, the ratio between light curves in 10–80 keV (middlepanel) and 3–10 keV (top panel) energy bands, are shown in the bottom panels. The data from second observation aredivided into four segments for further analysis and represented with di ff erent colors (top panel of the right side of the figure).The y-axis in top and middle panels represents the source rate in counts per second unit. − − − P o w e r [(r m s / m ea n ) / H z ] Frequency (Hz) Obs − Figure 3 . Power density spectrum of OAO 1657–415 ob-tained from the light curves in 3-80 keV range from theLAXPC20 data of first
AstroSat observation. Absence ofpeaks corresponding to the spin period of the pulsar can beseen. added in the LAXPC spectrum.
3. Timing Studies
Swift / BAT (Burst Alert Telescope, Krimm et al. 2013)and MAXI (Monitor of All-sky X-ray Image, Mat-suoka et al. 2009) long term monitoring light curvesof OAO 1657–415 in 15–50 keV and 2–20 keV ranges,respectively, are shown in Figure 1 to examine the over-all activity of the pulsar during both the epochs of
As-troSat observations. During the first observation on 31March 2019, OAO 1657–415 was observed in a low in-tensity phase. However, during the second observationon 4 July 2019, the pulsar was observed to be brighterin the beginning and gradually entered the low fluxlevel in the later part. The orbital phases covered duringboth the epochs of observations are 0.681–0.818 and0.808–0.968 (Table 1). Background subtracted lightcurves obtained from LAXPC20 data of both the ob-servations of the pulsar are shown in Fig 2. The toppanel (a) and middle panel (b) represent the source lightcurves in 3-10 keV and 10-80 keV energy ranges. Thehardness ratio (HR), the ratio between the light curvesin 10-80 keV and 3-10 keV ranges are also shown inthe bottom panels of the figure.On comparison of light curves in 3–10 keV and . Astrophys. Astr. (0000)000
As-troSat observations. During the first observation on 31March 2019, OAO 1657–415 was observed in a low in-tensity phase. However, during the second observationon 4 July 2019, the pulsar was observed to be brighterin the beginning and gradually entered the low fluxlevel in the later part. The orbital phases covered duringboth the epochs of observations are 0.681–0.818 and0.808–0.968 (Table 1). Background subtracted lightcurves obtained from LAXPC20 data of both the ob-servations of the pulsar are shown in Fig 2. The toppanel (a) and middle panel (b) represent the source lightcurves in 3-10 keV and 10-80 keV energy ranges. Thehardness ratio (HR), the ratio between the light curvesin 10-80 keV and 3-10 keV ranges are also shown inthe bottom panels of the figure.On comparison of light curves in 3–10 keV and . Astrophys. Astr. (0000)000 : − − − P o w e r [(r m s / m ea n ) / H z ] Frequency (Hz) Seg − I − − − P o w e r [(r m s / m ea n ) / H z ] Frequency (Hz) Seg − II − − − P o w e r [(r m s / m ea n ) / H z ] Frequency (Hz) Seg − III − − − P o w e r [(r m s / m ea n ) / H z ] Frequency (Hz) Seg − IV Figure 4 . Power density spectra (PDS) of OAO 1657–415 obtained from the segmented light curves of second
AstroSat ob-servation in 3-80 keV range. Presence of peaks at frequency corresponding to the spin period of the pulsar and its harmonicscan be seen in the first, second and third segments of the observation. These peaks are, however, absent in the PDScorresponding to the fourth segment (right bottom panel).
AstroSat observa-tion, compared to that during the initial part of the sec-ond observation. During the first epoch of observation,the pulsar did not show any significant time variabil-ity in soft and hard X-ray light curves, despite beingobserved away from the eclipse regime. The hardnessratio was also found to be constant ( ∼ AstroSat observation. The hardnessratio also found to change during the second observa-tion.A search for X-ray pulsations was performed in the 3-80 keV barycentric corrected light curves, binned at0.1 s, from both the observations. For this, the power-density spectra (PDS) were generated from the lightcurves using the Fast Fourier Transformation techniquewith the powspec task of
FTOOLS . Absence of sharpnarrow peaks in the PDS from the first observation(Figure 3) suggests the non-detection of X-ray pulsa-tions in the light curve of the pulsar. Pulsations wereagain searched in light curves in di ff erent energy bandssuch as 3-10 keV, 3-25 keV, and 10-80 keV ranges.However, we failed to detect any pulsating signal fromthe neutron star during the first observation. The sec-ond observation of OAO 1657–415 was carried out bycovering out-of-eclipse and eclipse phases of the bi-nary orbit. For the pulsation search, the entire obser-vation was divided into four di ff erent segments (Seg-I,II, III, & IV) on the basis of source intensity. Thesesegments are marked in di ff erent colors in the top right J. Astrophys. Astr. (0000) : −
80 keVSeg − I −
10 keV N o r m a li ze d I n t e n s it y −
25 keV −
50 keV −
80 keV −
80 keVSeg − II −
10 keV −
25 keV −
50 keV −
80 keV −
80 keVSeg − III −
10 keV −
25 keV −
50 keV −
80 keV
Figure 5 . Energy resolved pulse profiles of OAO 1657–415 obtained by folding the light curves from LAXPC20 instrumentduring Seg-I, Seg-II and Seg-III of the second observation (as marked in top right panel of Figure 2) at the estimated spinperiod. Top panels show the pulse profiles of the pulsar in entire LAXPC energy range for di ff erent segments, whereas theother panels show the pulse profiles in narrow energy ranges (quoted in each panel). Two pulses are shown in each panel forclarity. The error-bars represent 1 σ uncertainties.
20 40 60 800.10.20.30.4 P u l s e fr ac ti on Energy (keV)
Seg − ISeg − IISeg − IIISeg − IV Figure 6 . Pulse fraction variation of the pulsar with energyobtained from pulse profiles in multiple energy bands. panel of Figure 2. It should be noted that the fourthsegment (Seg-IV) represents the duration of the eclipseof the neutron star during the second
AstroSat observa-tion. The PDS obtained from the 3-80 keV segmentedlight curves show signatures of strong pulsations alongwith its harmonics for Seg-I, II, & III (Fig. 4). The PDSobtained from Seg-IV, however, did not show any suchsignature at a frequency corresponding to the spin pe- riod of the pulsar (right bottom panel of Fig. 4). Thissuggests that pulsations are present in the light curvesduring the out-of-the eclipse phase of the pulsar and ab-sent during the eclipsing phase.We, then, applied the chi-square maximizationtechnique (Leahy 1987) using efsearch task to deter-mine the pulsation period. From the PDS analysis andthe chi-square maximization technique, the spin periodof the pulsar is estimated to be 37.0375(8) s from theout-of-eclipse phases (Seg-I, II, & III) of the second
As-troSat observation. We checked the long term spin fre-quency history of OAO 1657–415 using
Fermi / GBM data. The GBM instrument did not detect any pulsationin the source from MJD 58571.37 to MJD 58574.35during which the first AstroSat observation was carriedout. However, pulsations were detected during the sec-ond
AstroSat observation as well as with
Fermi / GBM.It is also worth to note that the count rate observed inthe first observation is relatively higher than the samein third segment of second observation. The presenceof pulsations at lower source count rate in the third seg-ment confirms that the source was not into the propellerregime during the first observation. https://gammaray.nsstc.nasa.gov/gbm/science/pulsars.html . Astrophys. Astr. (0000)000
Fermi / GBM.It is also worth to note that the count rate observed inthe first observation is relatively higher than the samein third segment of second observation. The presenceof pulsations at lower source count rate in the third seg-ment confirms that the source was not into the propellerregime during the first observation. https://gammaray.nsstc.nasa.gov/gbm/science/pulsars.html . Astrophys. Astr. (0000)000 : The background subtracted light curves in 3–80keV energy range from the LAXPC20 data for Seg-I,II, & III were folded at the estimated spin period ofthe pulsar. The pulse profiles from these segments areshown in the top panels of Figure 5 (left to right). Asingle peaked profile with a dip at the peak is observedin first three segments. Non-detection of pulsations inthe light curve from Seg-IV, as the neutron star enteredinto the eclipsing phase of the binary (Figure 2), pulseprofiles for this segment were not generated.Evolution of the pulse profile with energy was in-vestigated by generating profiles in di ff erent energyranges. The light curves were extracted in several en-ergy ranges such as 3-10 keV, 10-25 keV, 25-50 keV,and 50-80 keV, for Seg-I, II, & III. These light curveswere folded with the estimated spin period and pre-sented in Figure 5 in second, third, fourth, and fifth pan-els (from top to bottom) for respective segments. In ourstudy, pulsations are e ff ectively detected up to 80 keVin first and second segments. The dip in the peak of theprofile was also found to present up to higher energies.The third segment which is in 0.89-0.92 orbital phaserange (out-of-eclipse phase), appears like the quiescentphase of the pulsar (Figure 2, top right panel). Signif-icant decrease in pulsar intensity during this segment,a ff ected the soft X-ray pulse profile more. A singlepeaked structure with a dip appears only in the hardX-rays above 10 keV. The pulsation is detected up to ∼
50 keV in this case.To quantify the nature of these pulsating com-ponents, pulse fraction from the pulse profiles ofOAO 1657–415 is calculated and shown in Figure 6.In our study, the pulse fraction is defined as the ratiobetween the di ff erence and sum of maximum and mini-mum intensities observed in the pulse profile. The pulsefraction from Seg-IV is not estimated as the pulsationwas not detected in this segment. From the figure, a de-creasing trend in the value of pulse fraction with energy,from a maximum of ∼ Figure 7 . Comparison between source and backgroundenergy spectra obtained from the first
AstroSat observationof OAO 1657–415 . wake, or the binary companion (e.g. in Seg-III). Pulseprofiles of the pulsar were also generated from the SXTdata. However, as the source is heavily absorbed in softX-ray ranges, the profiles were not suitable to draw anymeaningful information and not shown here.
4. Spectral Studies
To understand the cause of non-detection of pulsa-tions in the first
AstroSat observation and the inten-sity variation during the second observation, we car-ried out spectral analysis by using SXT and LAXPC20data from both the observations. The spectral fit-ting package
XSPEC of version 12.10.0 (Arnaud 1996)was used. In the beginning, we extracted ob-served source + background and background spectrausing laxpc make spectra and laxpc find back tasks of LAXPCsoftware package, respectively, fromLAXPC20 data of Obs-1. The observed, background,and source spectra of the source from LAXPC20 ofObs-1 are shown in Figure 7. It can be seen that thesource spectrum is limited up to ∼
22 keV. As we aim tocarry out our spectral analysis using data from both theobservations, we restricted ourselves to fit the data upto 20 keV in our fitting.In our analysis, we considered 0.5-7 keV spectrumfrom SXT and 3.5-20 keV spectrum from LAXPC20.The 0.5-20 keV energy spectrum obtained from thefirst observation was fitted with several standard mod-els such as power law, high energy cuto ff power law,power law with blackbody component etc., along withthe Galactic absorption column density TBabs (Wilms,
J. Astrophys. Astr. (0000) : − − − − k e V ( P ho t on s c m − s − k e V − ) Obs −
11 100.5 2 5 20 − χ Energy (keV) − − − k e V ( P ho t on s c m − s − k e V − ) Obs − − − χ Energy (keV) − − − − k e V ( P ho t on s c m − s − k e V − ) Obs − − II+III+IV)1 100.5 2 5 20 − χ Energy (keV)
Figure 8 . Best-fitting energy spectra of OAO 1657–415 dur-ing the first (top panel) and second (middle and bottompanels)
AstroSat observations of the source. The data fromSXT and LAXPC instruments in 0.5-7 and 3.5-20 keV areused in the spectral fitting, respectively.
Allen & McCray 2000). Only an absorbed power lawwith a blackbody component ( bbodyrad in XSPEC ) wasable to fit the spectrum better, though excess residualswere observed in 6-7 keV range. This excess was re-solved into two iron emission lines at 6.4 keV and 6.7keV while fitting SXT data alone. Similar multiple ironemission lines have also been seen in other accretionpowered X-ray pulsars such as GX 1 + + χ ν = χ /ν ≈
1. The best-fit spectral parameters obtainedfrom our fitting are quoted in Table 2. The iron line pa-rameters quoted in Table 2, such as line energy, width,equivalent width and line flux are obtained by fitting theSXT data alone.The LAXPC20 light curve of the pulsar duringthe second
AstroSat observation shows di ff erent in-tensity phases of the source (Figure 2). During thenormal phase (Seg-I), the source intensity was maxi-mum which gradually decreased (Seg-II) to low inten-sity phases (Seg-III & IV). As the source flux was ex-tremely low during Seg-III & IV and the duration ofSeg-II is very short, spectral analysis was carried outby considering Seg-I and added Seg-II, III & IV, sepa-rately. While fitting the SXT and LAXPC spectra forSeg-I, the absorbed power-law model with a blackbodycomponent provided acceptable fit. However, there wasno clear signature of iron emission line in the residu-als. While fitting the spectra corresponding to addedSeg-II, III, and IV, a simple absorbed power-law modelor an absorbed power-law with blackbody model didnot fit the data well. Considering the low intensity na-ture of the neutron star at late orbital phases, a par-tial covering component was tried along with the ab-sorbed power-law model. Addition of partial coveringcomponent to the absorbed power-law model improvedthe fitting. We also detected a 6.4 keV iron fluores-cence line in the spectra for this added segment. Thebest-fitted spectral parameters obtained from the spec-tral fitting of data from both the AstroSat observationsare given in Table 2, whereas the energy spectra alongwith corresponding residuals are shown in Figure 8. Weused cflux convolution model for flux estimation inour study. . Astrophys. Astr. (0000)000
AstroSat observation shows di ff erent in-tensity phases of the source (Figure 2). During thenormal phase (Seg-I), the source intensity was maxi-mum which gradually decreased (Seg-II) to low inten-sity phases (Seg-III & IV). As the source flux was ex-tremely low during Seg-III & IV and the duration ofSeg-II is very short, spectral analysis was carried outby considering Seg-I and added Seg-II, III & IV, sepa-rately. While fitting the SXT and LAXPC spectra forSeg-I, the absorbed power-law model with a blackbodycomponent provided acceptable fit. However, there wasno clear signature of iron emission line in the residu-als. While fitting the spectra corresponding to addedSeg-II, III, and IV, a simple absorbed power-law modelor an absorbed power-law with blackbody model didnot fit the data well. Considering the low intensity na-ture of the neutron star at late orbital phases, a par-tial covering component was tried along with the ab-sorbed power-law model. Addition of partial coveringcomponent to the absorbed power-law model improvedthe fitting. We also detected a 6.4 keV iron fluores-cence line in the spectra for this added segment. Thebest-fitted spectral parameters obtained from the spec-tral fitting of data from both the AstroSat observationsare given in Table 2, whereas the energy spectra alongwith corresponding residuals are shown in Figure 8. Weused cflux convolution model for flux estimation inour study. . Astrophys. Astr. (0000)000 :
Table 2 . Best-fitting spectral parameters (with 90% errors)of OAO 1657–415 .Parameters Obs-1 Obs-2Seg-I II + III + IVN H1 a ± ± ± H2 a – – 166 ± ± Γ ± ± ± − ) 2.5 ± ± ± ± ± norm ± ± ± ± ± b ± ± ± ± + . − . – –W2 (keV) 0.4 + . − . – –Flux2 b ± ± c ± ± ± χ ν ( ν ) 0.77 (61) 1.03 (310) 1.01 (142) a : in 10 cm − unit; b : unabsorbed line flux in 10 − erg cm − s − ; c : 0.5-30 keV unabsorbed flux in 10 − erg cm − s − unit
5. Discussion and Conclusions
We studied two
AstroSat observations of OAO 1657–415 carried out in March and July 2019 covering 0.681–0.818 and 0.808–0.968 phase ranges of the 10.4 dayorbital period of the binary system, respectively. Dur-ing the first observation, no pulsation was detected inthe SXT and LAXPC data. In this observation, despitebeing significantly away from the eclipse, the sourcewas found weak in the soft and hard X-ray light curves.The 0.5-30 keV unabsorbed flux was estimated to be1.6 × − erg cm − s − , corresponding to a source lu-minosity of ≈ × and 9.4 × erg s − at a dis-tance of 2.2 and 7 kpc, respectively. The spectral anal-ysis of the data from this observation revealed a highvalue of column density N H of about 1.5 × cm − .Strong iron lines at 6.4 keV and 6.7 keV were also de-tected in the SXT spectrum obtained from the first ob-servation. The equivalent width of the lines are esti-mated to be as high as 1 keV. Detection of such strongiron lines with high equivalent width suggests the pres-ence of abundant material around the pulsar for repro-cessing. A high value of equivalent width ( ∼ ffi ciently absorb the X-ray photonsand hence reducing the source flux to lower values asobserved in this case. Similar structure has been re-ported in case of 4U 1700–37 during the out-of-eclipsepart of the binary between 0.63–0.73 phases (Borosonet al. 2003; Jaisawal & Naik 2015). An increase inthe column density is usually observed in these orbitalphases. In contrast to the present study, Jaisawal &Naik (2015) did not find any significant increase in theiron line parameters in case of 4U 1700–37. The studyof OAO 1657–415 in 0.12–0.34 orbital phase rangewith Suzaku revealed the presence of eclipse-like seg-ment in the light curve. Time-resolved spectroscopyof the
Suzaku data corresponding to 0.19–0.23 phaseshowed a high column density as well as strong 6.4 keVand 7.1 keV iron lines with equivalent width of about 1keV and 0.3 keV, respectively (Jaisawal & Naik 2014;Pradhan et al. 2014). The presence of a dense blobof material across the line of sight of neutron star wasspeculated within an accretion radius (Jaisawal & Naik2014). In the present study, the existence of a denseblob of material or the accretion wake can not be de-nied. If the material is dense, it can absorb the X-rayphotons, henceforth no pulsation is observed as seenduring the first observation. Nonetheless, the detectionof strong iron emission lines favors the hypothesis ofabsorption through a clumpy stellar wind in our study.Alternatively, the cessation of the pulsations can bepossible in X-ray pulsars when the source enters intopropeller regime (Illarionov & Sunyaev 1975) at ex-tremely low mass accretion rate (see, e.g., in case ofGX 1 + AstroSat observation. This is due tothe detection of iron emission lines during this obser-vation. Considering the presence of high column den-sity and iron emission lines at late orbital phases of thebinary orbit (first
AstroSat observation), the disappear-ance of pulsations can be interpreted as due to the pres-ence of accretion wake.From the second
AstroSat data, pulsations wereclearly detected in the out-of-eclipse data. The pulseprofiles of OAO 1657–415 were found to be singlypeaked with a dip like structure at the peak of the pro-
J. Astrophys. Astr. (0000) : file. Usually this source is subjected to strong and densestellar wind of the supergiant companion as it is in aclose binary system. Inhomogenous distribution of stel-lar wind as well as clumpy wind accretion in SGXBscan a ff ect the pulsed emission. The profile can be af-fected mostly in the soft X-rays due to absorption. Inthe present and previous studies of OAO 1657–415, thepresence of a dip in the profile is seen up to higherenergies (Pradhan et al. 2014). Such type of profileswith absorption-like feature(s) at certain pulse phasesare common in transient BeXRB pulsars (Maitra, Paul& Naik 2012; Naik et al. 2013; Jaisawal, Naik & Epili2016). The complex shape of the pulse profiles in theseBe / X-ray pulsars is interpreted as an e ff ect of couplingbetween accreted material and magnetic field lines.We also carried out spectral investigation of thesource during the second AstroSat observation by di-viding the data into two parts. The first part of thedata (Seg-I, Figure 2) was relatively brighter by a fac-tor of 34 than the first
AstroSat observation. We didnot detect any iron emission line in the data from thissegment. The SXT and LAXPC spectra from the sec-ond part of the second observation (added Seg-II, III &IV, Figure 2), covering 0.87–0.968 orbital phase range,were highly absorbed. Only a weak iron emission lineat 6.4 keV was detected in this segment. As the spectralfitting is limited to 0.5-20 keV range, detection of thecyclotron absorption features (Jaisawal & Naik 2017;Staubert et al. 2019) is not expected in the data.In summary, we have studied two
AstroSat obser-vations in March and July 2019 at late orbital phases ofthe binary system. We did not detect any strong pulsa-tion at the spin period of the pulsar in the first data set,observed in 0.681–0.818 phase range. The presence ofrelatively high column density and strong iron emissionlines at 6.4 keV and 6.7 keV with an equivalent width of ≈ Acknowledgements
We thank the anonymous referee for suggestions onour paper. This publication uses the data from
As-troSat mission of the ISRO, archived at the IndianSpace Science Data Centre. We thank members ofSXT, LAXPC, and CZTI instrument teams for theircontribution to the development of the instruments andanalysis software. LAXPC data were processed by the Payload Operation Centre at TIFR, Mumbai. This workwas performed utilizing the calibration data-bases andauxiliary analysis tools developed, maintained and dis-tributed by the AstroSat-SXT team with members fromvarious institutions in India and abroad, and the SXTPayload Operation Center (POC) at the TIFR, Mumbai(https: // / astrosat sxt / index.html). SXTdata were processed and verified by the SXT POC. SN,BC and AG acknowledges the support from PhysicalResearch Laboratory which is funded by the Depart-ment of Space, India. References
Agrawal P. C. 2006, Adv. Space Res., 38, 2989Agrawal P. C., et al. 2017, J. Astrophys. Astron., 38, 30Antia H. M., et al. 2017, ApJS, 231, 10Arnaud K. A. 1996, in Jacoby G. H., Barnes J., eds, ASPConf. Ser. Vol. 101, Astronomical Data Analysis Softwareand Systems V. Astron. Soc. Pac., San Fransisco, p. 17Audley M. D., Nagase F., Mitsuda K., Angelini L., Kelley R.L. 2006, MNRAS, 367, 1147Barnstedt J., Staubert R., Santangelo A., et al. 2008, A&A,486, 293Baykal A. 1997, A&A, 319, 515Baykal A. 2000, MNRAS, 313, 637Bildsten L., Chakrabarty D., Chiu J., et al. 1997, ApJS, 113,367Blondin J. M., Kallman T. R., Fryxell B. A., Taam R. E.1990, ApJ, 356, 591Boroson B., Vrtilek S. D., Kallman T., Corcoran M. 2003,ApJ, 592, 516Chakrabarty D., Grunsfeld J. M., Prince T. A., et al. 1993,ApJ, 403, L33Chakrabarty D., Wang Z., Juett A. M., Lee J. C., Roche P.2002, ApJ, 573, 789Corbet R. H. D. 1986, MNRAS, 220, 1047Cui W. 1997, ApJ, 482 163F¨urst F., et al. 2010, A&A, 519, 37Illarionov A. F., Sunyaev R. A. 1975, A&A, 39, 185Inoue H. 1985, Space Sci. Rev., 40, 317Jaisawal G. K., Naik S. 2014, BASI, 42, 147Jaisawal G. K., Naik S. 2015, MNRAS, 448, 620Jaisawal G. K., Naik S., Epili P. 2016, MNRAS, 457, 2749Jaisawal G. K., Naik S. 2017, in Serino M., Shidatsu M.,Iwakiri W., Mihara T., eds, 7 years of MAXI: monitor-ing X-ray Transients, held 5-7 December 2016 at RIKEN,RIKEN, Saitama, Japan. p. 153 . Astrophys. Astr. (0000)000