aa r X i v : . [ a s t r o - ph ] S e p Timing the accretion flow around accretingmillisecond pulsars
Manuel Linares
Astronomical Institute “Anton Pannekoek”, University of Amsterdam, Kruislaan 403, NL-1098 SJAmsterdam, Netherlands. mail to: [email protected]
Abstract.
At present, ten years after they were first discovered, ten accreting millisecond pulsars are known.I present a study of the aperiodic X-ray variability in three of these systems, which led to thediscovery of simultaneous kHz quasi periodic oscillations in XTE J1807–294 and extremely strongbroadband noise at unusually low variability frequencies in IGR J00291+5934. Furthermore, weclassified SWIFT J1756.9–2508 as an atoll source and measured in its 2007 outburst spectral andvariability properties typical of the extreme island state. I also give detailed estimates of the totalfluence during the studied outbursts.
Keywords:
Neutron stars - X-ray binaries - Pulsars
PACS:
INTRODUCTION
During the last decade accreting millisecond pulsars (AMPs) have revealed a numberof interesting phenomena and have opened a new window to the physics of accretiononto neutron stars (NSs). The first of such systems was discovered by Wijnands andvan der Klis [1, SAX J1808.4–3658], presenting the first prove of an accreting neutronstar having both millisecond spin period and dynamically important magnetic field. TenAMPs have been discovered to date, and in three of them millisecond X-ray pulsationshave been seen to appear in and disappear from the persistent emission, producingpredominant [HETE J1900.1-2455; 2] intermittent [SAX J1748.9-2021; 3] or very rare[Aql X-1; 4] episodes of pulsations. This implies that the AMP within them is only activeor visible during a relatively small fraction of the time, which may provide a link withthe much more numerous class of non-pulsating neutron star low-mass X-ray binaries(NS-LMXBs).One way to study accretion onto compact objects is to analyze the aperiodic variabilityin the X-ray flux coming from these sources, which tells us about processes occuring inthe inner accretion flow [5, 6]. Such timing of the accretion flow, combined with a studyof the X-ray spectrum, reveals different “accretion states”. We show in this paper threedifferent AMPs in three different accretion states, we describe their aperiodic variabilityand quantify their outburst fluence.
GR J00291+5934
The sixth AMP was discovered on December 2 nd , 2004: IGR J00291+5934 [7]. Coher-ent pulsations were found at a frequency of 598.9 Hz, modulated by the ∼ RXTE (see Fig. 2). Our study of IGR J00291+5934 showed whatstill constitutes the strongest X-ray variability seen in a NS-LMXB, namely the frac-tional rms was ∼ ∼ FIGURE 1.
Power spectra (in power times frequency representation and rms normalized) of threeAMPs in outburst: IGR J00291+5934 (left) , SWIFT J1756.9–2508 (center) and XTE J1807–294 (right) .The break frequency in IGR J00291+5934 is more than two orders of magnitude lower than that ofXTE J1807–294, which shows instead simultaneous kHz QPOs.
SWIFT J1756.9–2508
On June 7 th , 2007, a new X-ray transient was discovered [12] with the burst alerttelescope (BAT) onboard Swift . Follow up
RXTE observations revealed that this wasthe eighth discovered AMP and showed a pulse frequency of ∼
182 Hz and an orbitalperiod of ∼
54 minutes [13, 14, 15]. The outburst lasted about two weeks (see Fig. 2).We analyzed the aperiodic variability of the source, comparing it with other AMPs andwith atoll sources. We thereby classified SWIFT J1756.9-2508 as an atoll source in theextreme island state. Using both PCA and HEXTE data we detected a hard tail in itsenergy spectrum extending up to 100 keV, fully consistent with such source and stateclassification [16]. It is interesting to note that so far all AMPs show spectral and timing[except for the shifts in the frequency-frequency correlations, see 17, 18] propertiesidentical to those of atoll sources, which suggests that low mass accretion rate is anecessary (even though seemingly not sufficient) ingredient to make an AMP.
TE J1807–294
A new transient X-ray source was discovered in the Galactic bulge region on February13 th , 2003. Subsequently, coherent pulsations were detected at a frequency of 190.6 Hzturning the new system, XTE J1807–294, into the fourth discovered AMP [19]. Anorbital period of ∼
40 minutes was determined [20], still the shortest among AMPs. Theoutburst was followed by
RXTE during five months (see Fig.2). We discovered sevenpairs of twin kHz quasi-periodic oscillations in XTE J1807–294 [18], with a frequencyseparation approximately equal to the spin frequency (see M. van der Klis contributionin these proceedings for further details). Lu m i no s i t y ( . + e r g / s ) Time (MJD)SWIFT J1756.9-2508 (2007 outburst) PCA+HEXTEBAT 0 0.01 0.02 0.03 0.04 0.05 0.06 53342 53344 53346 53348 53350 53352 53354 53356 53358 Lu m i no s i t y ( . + e r g / s ) Time (MJD)IGR J00291+5934 (2004 outburst) PCA+HEXTE Lu m i no s i t y ( . + e r g / s ) Time (MJD)XTE J1807-294 (2003 outburst) PCA+HEXTEPCA SCAN
FIGURE 2.
Light curves of the three AMP outbursts studied herein. Filled circles show the
RXTE -PCA&HEXTE measurements and empty squares those of
Swift -BAT or PCA scans.
OUTBURST FLUENCES
An important question that still remains open is why AMPs show pulsations whereasmost of NS-LMXBs do not. The solution proposed by Cumming et al. [21] invokesscreening of the magnetic field in “classical” NS-LMXBs by a time-averaged massaccretion rate higher than that of AMPs. In order to test this and other theories (NS crustcooling, binary evolution) a careful estimate of how much mass falls onto the neutronstar surface is of capital importance [22]. For this purpose we measure the unabsorbed2-200 keV flux during the three AMP outbursts mentioned above, using data from bothPCA and HEXTE onboard
RXTE . We fit the broadband, background and deadtimecorrected, energy spectra with an absorbed disk blackbody plus power law model (fixingthe column density to the Galactic value in the source direction). For XTE J1807–294we use data from the PCA scans of the Galactic bulge [23] to cover the rise and the finaldecay, as there are no pointed
RXTE observations during those parts of the outburst. Inthe case of SWIFT J1756.9–2508 we use data from the
Swift -BAT transient monitor , asmost of the outburst had no RXTE pointings. We calibrate the conversion between PCA-scan/BAT and PCA&HEXTE fluxes in those parts of the outbursts where both fluxes areavailable, thus implicitly assuming that the spectral shape does not vary drastically. Theresulting lightcurves, translated to luminosities in units of 2.5 × erg/s, are shownin Figure 2. Table 1 shows our measurements of the fluence, radiated energy and peak http://swift.gsfc.nasa.gov/docs/swift/results/transients/ uminosity for the three outbursts studied (see note on distances used therein). We alsoshow in Table 1 the range spanned by the break frequency in the power spectrum, whichindicates the relative change in X-ray variability frequencies, as well as the orbital periodof each system. TABLE 1.
Parameters of the AMP outbursts studied in this work.Outburst Outburst Peak Break Orbitalfluence energy ∗ luminosity ∗ frequency period(10 − erg/cm ) (10 erg) (% Eddington) (Hz) (min)XTE J1807–294 7.2 5.5 [8] 9.5 [8] 5.3-10.2 40IGR J00291+5934 1.4 0.6 [6] 5 [6] 0.03-0.05 150SWIFT J1756.9–2508 1.5 1.2 [8] 7 [8] 0.09-0.12 54 ∗ The fiducial distance used is indicated, in kiloparsecs and between brackets. These three AMPs have notshown X-ray bursts and their distances are therefore uncertain [see, however, 24, 25].
ACKNOWLEDGMENTS
I thank my colleagues and members of the organizing committee, as well as all theparticipants of the Amsterdam workshop, for making this event a fruitful and enjoyableone.
REFERENCES
1. R. Wijnands, and M. van der Klis,
Nature , 344–346 (1998).2. D. K. Galloway et al.,
ApJL , L73–L76 (2007).3. D. Altamirano, P. Casella, A. Patruno, R. Wijnands, and M. van der Klis,
ApJL , L45–L48 (2008).4. P. Casella, D. Altamirano, A. Patruno, R. Wijnands, and M. van der Klis,
ApJL , L41–L44 (2008).5. M. van der Klis, in "X-ray binaries" ,Cambridge University Press, eds. Lewin, W.H.G. et al., 1995.6. M. van der Klis, in "Compact Stellar X-ray Sources", ed. W. H. G. Lewin & M. van der Klis(Cambridge Univ. Press) pp. 39–112 (2006).7. D. Eckert et al.,
The Astronomer’s Telegram , 1–+ (2004).8. C. B. Markwardt, J. H. Swank, and T. E. Strohmayer,
The Astronomer’s Telegram , 1–+ (2004).9. C. B. Markwardt et al.,
The Astronomer’s Telegram , 1–+ (2004).10. M. Linares, M. van der Klis, and R. Wijnands,
ApJ , 595–604 (2007).11. G. Hasinger, and M. van der Klis,
A&A , 79–96 (1989).12. H. A. Krimm et al.,
The Astronomer’s Telegram , 1–+ (2007).13. C. B. Markwardt, H. A. Krimm, and J. H. Swank,
The Astronomer’s Telegram , 1–+ (2007).14. C. B. Markwardt, and H. A. Krimm,
The Astronomer’s Telegram , 1–+ (2007).15. H. A. Krimm et al.,
ApJL , L147–L150 (2007), .16. M. Linares et al.,
ApJ , 515–519 (2008).17. S. van Straaten, M. van der Klis, and R. Wijnands,
ApJ , 455–482 (2005).18. M. Linares, M. van der Klis, D. Altamirano, and C. B. Markwardt,
ApJ , 1250–1260 (2005).19. C. B. Markwardt, E. Smith, and J. H. Swank,
IAUC , 2–+ (2003).20. C. B. Markwardt, M. Juda, and J. H. Swank,
IAUC , 2–+ (2003).21. A. Cumming, E. Zweibel, and L. Bildsten,
ApJ , 958–966 (2001).22. D. K. Galloway, edited by F. D’Amico et al., 2006, vol. 840 of
AIP Conference Series , pp. 50–54.23. C. B. Markwardt, and F. J. Swank,
APS APRX , 11007 (2002).24. M. Falanga, J. M. Bonnet-Bidaud, J. Poutanen, R. Farinelli, A. Martocchia, and P. Goldoni (2005).25. D. K. Galloway et al.,
ApJL622