Spectral and temporal changes associated with flux enhancement in 4U 1626-67
aa r X i v : . [ a s t r o - ph . H E ] D ec Mon. Not. R. Astron. Soc. , 1–7 (2009) Printed 30 October 2018 (MN L A TEX style file v2.2)
Spectral and temporal changes associated with fluxenhancement in 4U 1626 − Chetana Jain , ⋆ , Biswajit Paul and Anjan Dutta Department of Physics and Astrophysics, University of Delhi, Delhi 110007, India Raman Research Institute, Sadashivnagar, C. V. Raman Avenue, Bangalore 560080, India
30 October 2018
ABSTRACT
4U 1626 −
67 is an accretion powered X-ray pulsar that shows remarkably stableX-ray luminosity above hours timescale and gradual intensity variation on a few yearstimescale. Unlike the other high magnetic field binary X-ray pulsars, the spin-up orspin-down rate of the neutron star is also very stable upto several years. Here we reporta significant increase in the X-ray intensity in the long term
RXTE -ASM light curveof 4U 1626 −
67. Similar enhancement in the X-ray flux has also been detected in the
Swift -BAT light curve. The increase in the X-ray flux took place over a long period ofabout 100 days and there appears to be two episodes of flux enhancement. We haveinvestigated the spectral and timing features of 4U 1626 −
67 during its current state ofenhanced flux emission with data obtained from the Proportional Counter Array andthe High-Energy X-ray Timing Explorer on board the
Rossi X-ray Timing Explorer .We report the detection of a torque reversal to spin-up in 4U 1626-67. The source hasentered a new spin-up phase with a spin-up rate of 4.02(5) × − Hz s − . The presentspin-up rate is almost half of the earlier spin-up and spin-down trends. A significantexcess in soft X-ray photon emission is observed during the enhanced flux state, whichis similar to the energy spectrum obtained during the spin-up era of the pulsar before1990. 4U 1626 −
67 is a unique accretion powered X-ray pulsar in which quasi periodicoscillations have been consistently observed over the past ∼
20 years. But during therecent observations, we did not detect a QPO at frequencies as observed in earlierobservations. Instead, we report detection of a significant broadening in the wings ofthe 130 mHz peak and a change in the shape of the continuum of the power spectrum.These results indicate that the flux enhancement is not a simple case of increased massaccretion rate, but there is also a change in the accretion geometry in the vicinity ofthe neutron star.
Key words:
X-ray: Neutron Stars - pulsars: individual (4U 1626 −
67) - X-ray: binaries
4U 1626 −
67 is an accretion powered X-ray pulsar whichwas discovered with the
Uhuru satellite in 1972 (Giacconiet al. 1972). The 7.7s X-ray pulsations were discovered byRappaport et al. (1977) during SAS3 observations and havebeen extensively monitored since then. Initially the pulsarwas found to be spinning up with a characteristic timescaleof ∼ ⋆ E-mail: [email protected]
McClintock 1983). Pulsations at 130 mHz have also been de-tected from the reprocessed optical emission (Middleditch etal. 1981). Multiple sidebands separated by about 0.4 mHz,are present in the power spectrum of the optical light curve,which was interpreted as signature of an orbital period of ∼
42 min (Middleditch et al. 1981; Chakrabarty et al. 2001).But despite extensive searches, an orbital motion of the neu-tron star (usually observed in binary pulsars in the form ofDoppler shift in the pulse period, or delay in the pulse arrivaltime) has never been detected in this source (Rappaport etal. 1977; Levine et al. 1988; Jain et al. 2007). It is believedthat the neutron star most probably, has a low mass binarycompanion and is in a nearly face-on orbit.The X-ray spectrum of 4U 1626 −
67 was first studied indetail with the
ASCA spectrometers (Angelini et al. 1995) c (cid:13) Chetana Jain, Biswajit Paul and Anjan Dutta and later with instruments on board
Beppo-SAX (Owenset al. 1997; Orlandini et al. 1998),
Chandra (Schulz et al.2001) and
XMM − Newton (Krauss et al. 2007). The X-raycontinuum spectrum is evidently correlated with the torquestate. During the 1977 − ∼
37 keV which they at-tributed to be the cyclotron absorption line. The spectrumis rich in bright hydrogen-like and helium-like oxygen andneon emission lines. The blue and red Doppler shifted emis-sion lines in the X-ray spectrum indicate that the lines orig-inate in the accretion disk. However, presence of the red andblue Doppler shifted emission lines from the accretion diskis difficult to explain if the binary system is face-on. Pres-ence of a metal rich accretion disk and absence of orbitalmotion led Jain et al. (2007) to suggest that 4U 1626 − − −
67 is a unique X-ray pulsar in which QPOs havebeen consistently detected in the X-ray and optical observa-tions. The QPO frequency has been evolving over the past22 years (Kaur et al. 2008). The QPO central frequency wasfound to be increasing during the spin-up phase of the pul-sar, whereas, a negative trend was observed after the torquereversal. During the spin-up phase of the source, QPOs weredetected at a frequency of 36 mHz with the EXOSAT data(Kaur et al. 2008) and at 40 mHz from the
Ginga observa-tions (Shinoda et al. 1990). During the spin-down era of thesource, QPOs have been reported to occur at a higher fre-quency of about 48 mHz from
ASCA , Beppo -SAX,
RXTE ,and
XMM-Newton (Angelini et al. 1995; Owens et al. 1997;Kommers et al. 1998; Krauss et al. 2007). Large amplitude0.3 − RXTE and
Swift satellites, show an in-crease in the X-ray intensity of 4U 1626 −
67. In the presentwork, we have investigated the spectral and temporal prop-erties of 4U 1626 −
67 during the increased flux state. In par-ticular, we have searched for pulsations and the QPO featurein the power spectrum of 4U 1626 −
67 which was persistentlypresent in the earlier observations. We have also studied theenergy spectrum of the source during the current phase andhave compared it with the earlier known spectral features.
Figure 1.
RXTE -ASM and
Swift -BAT light curves of 4U1626 −
67, binned with 30 d. The 15 −
50 keV
Swift -BAT light curveis shown with a thick line over the 1.5 −
12 keV
RXTE -ASM lightcurve (thin line). The inset figure shows the expanded view nearthe onset of the flux enhancement which occured around MJD54500 (thick line is the
Swift -BAT light curve in the same unitsas in the main panel).
4U 1626 −
67 is a medium intensity persistent X-ray sourceamong the few hundred bright X-ray sources regularly mon-itored by the All Sky Monitor (ASM) on board
Rossi X-ray Timing Explorer ( RXTE ). Figure 1 (thin line) showsthe long term 1.5 −
12 keV ASM light curve of 4U 1626 − ∼
100 days.4U 1626 −
67 was also regularly monitored by the BurstAlert Telescope (BAT; Barthelmy et al. 2005) on boardthe
Swift observatory (Gehrels et al. 2004). The long term15 −
50 keV
Swift -BAT light curve, binned with 30 d is alsoshown in Figure 1 (thick line) alongwith the ASM lightcurve. The observations covered the time range from MJD53414 to MJD 54920. A sudden increase in the X-ray fluxwas first reported by Krimm et al. (2008). The inset in Fig-ure 1 (thick line) shows the expanded view (in the sameunits) near the onset of the flux enhancement. The fluxchange on timescales of ∼
100 days is also clear in the
Swift -BAT light curve.We have also analysed data obtained from the Propor-tional Counter Array (PCA) and the High Energy X-rayTiming Explorer (HEXTE) on board the
RXTE . The PCAconsists of five xenon/methane proportional counter units(PCUs) and is sensitive in the energy range of 2 −
60 keVwith an effective area of 1300 − , depending on thenumber of PCUs ON (Jahoda et al. 1996). The HEXTE op-erates in the energy range 15 −
250 keV and consists of two c (cid:13) , 1–7 pectral and temporal changes in 4U 1626 − Table 1.
Log of
RXTE observations analyzed in the present workYear Observation ID Number of Total Exposurepointings (ks)1996 P10101 3 153P10144 1 10P20146 2 21997 P20146 12 71998 P30058 2 40P30060 7 532008 P93431 2 72009 P94423 3 9 clusters of phoswich scintillation detectors, each having acollecting area of 800 cm (Gruber et al. 1996).Table 1 gives a log of the RXTE observations used forthe present analysis. We have studied the temporal and spec-tral variations in 4U 1626 −
67, using data collected in theStandard-1 and the Standard-2 mode of the
RXTE -PCA,respectively. The archive mode data of the HEXTE was usedfor the spectral analysis.
We searched for pulsations using data obtained in the Stan-dard 1 mode of the
RXTE -PCA. The background countswere simulated using the ftool - runpcabackest and sub-tracted from the source light curve. The photon arrival timeswere converted to the solar system barycenter using the ftool - faxbary . We searched for the spin period of theneutron star using the efsearch tool of the HEAsoft analy-sis package, ftools ver 6.5.1. This tool folds the light curvewith a large number of trial periods around an approximateperiod and determines the best period by the χ maximiza-tion technique. Using this method, we determined a pulseperiod of 7.67941(1) s at MJD 54530.4, 7.67945(7) s at MJD54538.1 and 7.67848(2) at MJD 54984.5 (Jain & Paul 2009).This gives a spin-up rate of 4.02(5) × − Hz s − . Simi-lar behaviour based on the Fermi/GBM data has also beenreported by Camero-Arranz et al. (2009).Figure 2 shows the normalized pulse profile of 4U1626 −
67 binned into 64 phasebins. The top panel showsthe pulse profile generated from data obtained from
RXTE -PCA observations made in 1996 (ObsID 10101-01-01-00)and folded with a period of 7.66735 s (Jain et al. 2007).The bottom panel shows the normalized pulse profile of 2008data (ObsID 93431-01-01-00) folded with a period of 7.67941s. Although the two profiles resemble the characteristic bi-horned profile of the source (Levine et al. 1988), the pulsefraction is significantly different in the two cases. Moreover,the peaks in the 2008 pulse profile are significantly sharperwith a higher amplitude variation than those in the 1996pulse profile.We searched for QPOs in the 2 −
60 keV energy band,using the data collected in the Standard 1 mode of PCA,which has a time resolution of 0.125 s. We created the powerdensity spectrum (PDS) using the ftool - powspec forall the sufficiently long RXTE observations (See Table 1).The PDS were normalized such that their integral gives thesquared rms fractional variability and the white noise level
Figure 2.
The 2-60 keV normalized pulse profile of 4U 1626 − RXTE -PCA during 1996(ObsID 10101-01-01-00) and 2008 (ObsID 93431-01-01-00). Thebackground subtracted light curves were folded into 64 phasebinswith a period of 7.66735 s and 7.67941 s, respectively. was subtracted. Figure 3 shows some representative powerspectra generated from some of the observations mentionedin Table 1. The dates of observation are mentioned in eachpanel above their respective observation IDs.The narrow main peak at around 0.130 Hz correspondsto the spin frequency of the neutron star. Multiple harmon-ics are also seen in the PDS of the source. In addition tothe main peak, a QPO feature is also seen at ∼
48 mHz.However, the QPO feature is not present in the PDS gener-ated from observations made in 2008 and 2009 (Figure 3 (9),(10), (11), (12)). In all the observations prior to 2008, thepower spectrum consists of a broken power-law. As opposedto this, there is a remarkable difference in the power spectragenerated from 2008 and 2009 observations.Figure 4 shows the power spectrum generated from the2008 observations (ObsID 93431-01-01-00; Figure 3-panel 9)overlaid with a typical power spectrum from a 1996 observa-tion (ObsID 10101-01-01-00; Figure 3-panel 1). In the 2008power spectrum, absence of a QPO at ∼
48 mHz is clearlyseen. One more feature that stands out is the substantialbroadening of the wings of the pulse peak in the power spec-trum during the 2008 observation. The r.m.s variability inthe low frequency noise component of the power spectrumwas 14.6%, whereas, the r.m.s variability in the broad wingsof the fundamental pulse peak and the fundamental pulsepeak itself was 6.9% and 5.2%. This broad feature at ∼ c (cid:13)000
48 mHz is clearlyseen. One more feature that stands out is the substantialbroadening of the wings of the pulse peak in the power spec-trum during the 2008 observation. The r.m.s variability inthe low frequency noise component of the power spectrumwas 14.6%, whereas, the r.m.s variability in the broad wingsof the fundamental pulse peak and the fundamental pulsepeak itself was 6.9% and 5.2%. This broad feature at ∼ c (cid:13)000 , 1–7 Chetana Jain, Biswajit Paul and Anjan Dutta
Figure 3.
Some representative X-ray power spectra of 4U 1626 −
67 generated from the
RXTE -PCA observations. The date of observationis mentioned in each panel above the respective observation ID.
The pulse-phase averaged spectra of 4U 1626 −
67 were gen-erated from the data collected in the Standard 2 mode of the
RXTE -PCA and the archive mode of the
RXTE -HEXTE.The background counts for the PCA data were simulated us-ing the ftool - runpcabackest and subtracted from thesource spectrum. In case of the HEXTE data, the source andthe background spectra were generated using the ftool - hxtback . These were then corrected for the detector deadtime using the ftool - hxtdead . In order to bring out thesignificant evolution in the X-ray spectrum, we have gener-ated the spectrum from one of the 1996 observation (Ob-sID 10101 − − −
00) and the 2008 observation (ObsID93431 − − − −
20 keV and 15 −
50 keV energyspectrum of the source from the 1996
RXTE -PCA and
RXTE -HEXTE observations. A simultaneous fit of the twospectrum consists of an intrinsic absorption, a powerlaw, agaussian, a high energy cutoff and cyclotron resonance fea-ture (CRF). The bottom panel of Figure 5 shows the residualof the fit. The best fit spectral parameters (with 1% system-atic errors) are given in Table 2. The spectrum is similarto the broadband continuum spectrum obtained from the
Beppo -SAX observations made after the torque reversal in1990 (Owens et al. 1997; Orlandini et al. 1998). The
Beppo -SAX energy spectrum was well fit by a model consisting of alow energy absorption, a blackbody, a powerlaw and a highenergy cutoff. However, the best fit blackbody temperaturewas ∼ c (cid:13) , 1–7 pectral and temporal changes in 4U 1626 − −3 P o w e r Frequency (Hz)
Figure 4.
X-ray power spectra of 4U 1626 −
67 from the
RXTE observations of 1996 and 2008. The upper, dotted curve is the X-ray spectrum of 1996 observation (ObsID 10101-01-01-00; Figure3-panel 1) and the lower, solid curve is from 2008 observation(ObsID 93431-01-01-00; Figure 3-panel 9). N o r m a li ze d C oun t s s − k e V −
105 20−202 χ Energy (keV)RXTE−PCA RXTE−HEXTE1996
Figure 5.
The X-ray spectrum of the 1996
RXTE -PCA and
RXTE -HEXTE observations of 4U 1626 −
67. The model compo-nents are intrinsic absorption, powerlaw, a gaussian and a highenergy cutoff. The bottom panel shows the residual of the fit. the
RXTE -PCA detector. We have obtained a high energycutoff of 6.45 keV which is also quite small as compared to ∼
20 keV obtained from the
Beppo -SAX observations.The 3 −
20 keV
RXTE -PCA and the 15 −
50 keV
RXTE -HEXTE energy spectrum obtained from the 2008 observa-tions are shown in Figure 6. The spectra were simultaneouslyfitted with a model consisting of a powerlaw, a blackbody, agaussian, high energy cutoff and cyclotron resonance feature(CRF). In Table 2, we have given the best fit spectral param-eters (with 1 % systematic errors). The observations madewith
HEAO-1 (Pravdo et al. 1979) and X-ray astronomysatellite
Tenma (Kii et al. 1986) during the earlier spin-upphase of 4U 1626 −
67 also showed similar energy spectrum.Figure 7 shows the evolution of the 3 −
50 keV X-ray N o r m a li ze d C oun t s s − k e V −
105 20−202 χ Energy (keV)RXTE−PCA RXTE−HEXTE2008
Figure 6.
Energy spectrum of 4U 1626 −
67 obtained from the2008
RXTE -PCA observations. The model consists of a black-body, a powerlaw and a gaussian component.
Table 2.
Best fit spectral parameters of 4U 1626 −
67, obtainedfrom the
RXTE -PCA and
RXTE -HEXTE observations.Parameter Value1996 2008N H (10 cm − ) 0.65 0.82Photon index (Γ) 0.59(4) 0.74(7)norm (photons keV − cm − s − ) 0.0079(6) 0.012(2)Gaussian line (keV) 6.56(40) 6.33(39)Gaussian width (keV) 1.15(21) 1.43(38)Gaussian norm (photons cm − s − ) 0.00054(32) 0.0019(18)Blackbody, kT (keV) - 0.66(12)Blackbody, norm (km kpc − ) - 44(7)cutoff E (keV) 6.45775(95) 17.7(8)foldE (keV) 32(5) 26(12)Energy ( CRF ) (keV) 39.5(1.8) 40.3(1.2)Width ( CRF ) (keV) 13.8(3.0) 4.8(4.5)Depth ( CRF ) χ (d.o.f.) 0.59 (57) 0.61 (43) spectrum from 1996 observation to the 2008 observation.In the figure, the best fit spectrum of the 1996 observation(dotted curve) is shown with the spectrum generated fromthe 2008 observations. The bottom panel in the figure givesthe ratio between the two spectra. It is clear that there is anexcess soft-photon emission in the spectrum generated fromthe 2008 observations. The powerlaw photon index has alsoincreased during the recent observation.
4U 1626 −
67 is a unique system in which the X-ray flux hasbeen gradually decreasing over the past ∼
30 years. Sincethe discovery of X-ray pulsations in 1977, the source hasshown a smooth spin evolution history. Initially, it was foundto be spinning up with a characteristic timescale of 5000 c (cid:13) , 1–7 Chetana Jain, Biswajit Paul and Anjan Dutta N o r m a li ze d C oun t s s − k e V −
105 201.31.41.51.6 R a ti o Energy (keV) 20081996
Figure 7.
The best fit spectrum of 4U 1626 −
67, generated fromthe 1996 (dotted line) and 2008 (solid line)
RXTE observations.The bottom panel shows the ratio of the two spectra. yr, but after an abrupt torque reversal, the neutron starstarted spinning down at about the same rate (Chakrabartyet al. 1997; Krauss et al. 2007). The recent flux enhancementhas opened up many questions regarding the stability of theaccretion disk. We have found a sudden increase in the X-rayflux of 4U 1626 −
67, which could have also caused significantchanges in the geometry of the accretion column.The power spectra of X-ray pulsars consist of narrowpeaks originating from the periodic signal and its harmoniclines which are often accompanied with aperiodic featureslike broad bumps, wiggles and steep low frequency excess(Wijnands & van der Klis 1999). The periodic variabilitesarise due to the rotation induced motion of the accretioncolumn, through which the infalling matter is funneled to-wards the magnetic poles of the neutron star. Whereas, anyinstability in the emissivity of the accretion column, can giverise to an aperiodic variability (Burderi et al. 1997). If theinhomogeneities occur far away from the surface of the neu-tron star, then the periodic and aperiodic variations occurindependent of each other (Angelini et al. 1989). But if theaperiodic variability occurs near the accretion column, thenthey are modulated at the pulsar spin frequency (Makishimaet al. 1988). Lazzati & Stella (1997) found a significant cou-pling between the aperiodic and periodic variabilities in VelaX-1 and 4U 1145 −
62. A significant coupling was also re-ported in the accretion powered millisecond binary pulsarSAX J1808.4 − −
67, Yi &Vishniac (1999) had found that the torque reversal in 1990 was accompanied by spectral transition and a change in theluminosity. During the spin-up phase of 4U 1626 −
67, thephase averaged spectrum was well fit by a blackbody tem-perature k T of ∼ ∼ −
10 keV energy spectrum. The powerlaw photon in-dex changed from 1.6 (Pravdo et al. 1979) to ∼ ∼ ∼ ACKNOWLEDGMENTS
We thank the anonymous referee for some useful commentson improving this paper. This research has made use of thedata obtained through the High Energy Astrophysics Sci-ence Archive Research Center Online Service, provided bythe NASA/Goddard Space Flight Center. In particular, wethank the
RXTE -ASM teams at MIT for provision of theASM data.
REFERENCES
Angelini L., Stella L., Parmar A. N., 1989, ApJ, 346, 906Angelini L., White N. E., Nagase F., Kallman T. R.,Yoshida A., Takeshima T., Becker C., Paerels F., 1995ApJL, 449, 41Barthelmy S. D., Barbier L. M., Cummings J. R. et al.,2005, Space Science Reviews, 120, 143Bradt H. V. D., McClintock J. E., 1983, ARA&A, 21, 13Burderi L., Robba N. R., Cusumano G., 1993, Advances inSpace Research, 13, 291Burderi L., Robba N. R., La Barbera N., Guainazzi M.,1997, ApJ, 481, 943Camero Arranz A., Finger M. H., Ikhsanov N. R., Wilson-Hodge C. A., Beklen E., 2009, arXiv0906.4224Chakrabarty D., Bildsten L., Grunsfeld J. M. et al., 1997,ApJ, 474, 414Chakrabarty D., Homer L., Charles P. A., O’Donoghue D.,2001, ApJ, 562, 985Finger M. H., Wilson R. B., Harmon B. A., 1996, ApJ, 459,288Gehrels N., Chincarini G., Giommi P. et al., 2004, ApJ,611, 1005Giacconi R., Murray S., Gursky H., Kellogg E., SchreierE., Tananbaum H., 1972, ApJ, 178, 281Gruber D. E., Blanco P. R., Heindl W. A., Pelling M. R.,Rothschild R. E., Hink P. L., 1996, A&AS, 120, 641Jahoda K., Swank J. H., Giles A. B., Stark M. J.,Strohmayer T., Zhang W., Morgan E. H., 1996, SPIE, 2808,59Jain C., Paul B., Joshi K., Dutta A., Raichur H., 2007,JA&A, 28, 175Jain C., Paul B., 2009, ATel 2095Kaur R., Paul B., Raichur H., Sagar R., 2007, ApJ, 660,1409 c (cid:13) , 1–7 pectral and temporal changes in 4U 1626 − Kaur R., Paul B., Kumar B., Sagar, R., 2008, ApJ, 676,1184Kii T., Hayakawa S., Nagase F., Ikegami T., Kawai N.,1986, PASJ, 38, 751Kommers J. M., Chakrabarty D., Lewin W. H. G., 1998,ApJL, 497, 33Krauss M. I., Schulz N. S., Chakrabarty D., Juett A. M.,Cottam J. 2007, ApJ, 660, 605Krimm H. A., Barthelmy S. D., Baumgartner W. et al.,2008, ATel, 1426, 1Lazzati D., Stella L. 1997, ApJ, 476, 267Levine A., Ma C. P., McClintock J., Rappaport S., van derKlis M., Verbunt F., 1988, ApJ, 327, 732Makishima K., 1988, in Physics of Neutron Stars and BlackHoles, ed. Y. Tanaka (Tokyo: Universal Acad.), 175McClintock J. E., van Paradijs J., Hidajat B., HendricksH., 1977, IAUC, 3084, 3Menna M. T., Burderi L., Stella L., Robba N., van der KlisM., 2003, ApJ, 589, 503Middleditch J., Mason K. O., Nelson J. E., White N. E.,1981, ApJ, 244, 1001Moon Dae-Sik, Eikenberry S. S., 2001, ApJL, 552, 135Mukherjee U., Bapna S., Raichur H., Paul B., Jaaffrey S.N. A., 2006, JA&A, 27, 25Orlandini M., Fiume D. D., Frontera F. et al., 1998, ApJL,500, 163Owens A., Oosterbroek T., Parmar A. N., 1997, A&A, 324,L9Paul B., Rao A. R., 1998, A&A, 337, 815Pravdo S. H., White N. E., Boldt E. A. et al., 1979, ApJ,231, 912Raichur H., Paul B., 2008, ApJ, 685, 1109Rappaport S., Markert T., Li F. K., Clark G. W., JerniganJ. G., McClintock J. E., 1977, ApJL, 217, 29Schulz N. S., Chakrabarty D., Marshall H. L., CanizaresC. R., Lee J. C., Houck J., 2001, ApJ, 563, 941Shinoda K., Kii T., Mitsuda K., Nagase F., Tanaka Y.,Makishima K., Shibazaki N., 1990, PASJ, 42, L27Takeshima T., Dotani T., Mitsuda K., Nagase F., 1991,PASJ, 43, L43Vaughan B. A., Kitamoto S., 1997, astro-ph/9707105Wijnands R., van der Klis M., 1999, ApJ, 514, 939Yi I., Vishniac E. T., 1999, ApJL, 516, 87 c (cid:13)000