1. 1 ACCRETING X-RAY MILLISECOND PULSARS M A U R I Z I O F A L A N G A & E R I N W. B O N N I N G NS day, ParisJune 27, 2007 Service d‘Astrophysique, CEA –Saclay LUTH

2. 2 MSPs hosted in LMXBs  SXT: L ~ 10 31 -10 32 erg/s in quiescent L ~ 10 36 -10 38 erg/s in outburst, recurence time 2-5 yr. Close X-ray binaries: Companion: M << M sun, Accretion disk, Compact object NS: B~10 8 G  Rich time variability, such as twin QPOs at kHz frequencies (400 - 1300 Hz, increasing with Mdot); kHz QPOs are thought to reflect Kepler at the inner accretion disk. (Van der Klis, 2000, astro-ph/00001167) (The Power spectra obtained for SAX J1808.4-3658 during 2002 outburst.)  7 SXT which show X-ray millisecond coherent modulation.  Spin frequencies lye between 180 and 600 Hz. (see review by Wijnands 2004, astro-ph/0403409)  Type-I X-ray bursts, with nearly coherent oscillations in the range 300-600 Hz.  Burst oscillations reflect the NS spin frequency (D. Chakrabarty, Nature, 2003) (Burst oscilation from SAX J1808.4-3658 during 2002 outburst.) NS day, ParisJune 27, 2007

3. 3 …now we know 7 LMXBs (transients) which show X- ray millisecond coherent modulation: SAX J1808.4-3658: P s = 2.5ms, P orb = 2hr (Wijnands & van der Klis 1998) XTE J1751-306: P s = 2.3ms, P orb = 42min (Markwardt et al. 2002) XTE J0929-314: P s = 5.4ms, P orb = 43.6min (Galloway et al. 2002) XTE J1807-294: P s = 5.3ms, P orb = 40min (Markwardt et al. 2003) XTE J1814-388: P s = 3.2ms, P orb = 4.3hr (Markwardt et al. 2003) IGR J00291+5934: Ps = 1.67ms, Porb = 2.46hr (Eckert et al. 2004) HETE J1900.1-2455: Ps = 2.65ms, Porb = 1.4hr (Markwardt et al. 2005) The growing family of the X-ray millisecond pulsars NS day, ParisJune 27, 2007

4. 4 Companion mass Mc/Msun Companion radius Rc/Rsun Brown dwarfs 0.1 Gyr 5 Gyr 1 Gyr White dwarfs XTE J0929-314 XTE J1751-305 XTE J1807-294 IGR J00291+5934 SAX J1808..4-3658 XTE J1814-338 Assuming that the companion star should fill its Roche lobe to allow sufficient accretion on the compact star Companion Star  Brown dwarf models at different ages (Chabrier et al. 2000)  Cold low-mass white dwarfs with pure-helium composition  IGR J00291+5934  SAX J1808.4-3658 H-rich donor, brown dwarf  XTE J1814-338  XTE J0292-314  XTE J1751-305 H-poor, highly evolved dwarf  XTE J1807-294 NS day, ParisJune 27, 2007

5. 5 Recycling model for MSPs  LMXB phase preceding the MSP stage;  mass transfer stops;  the radio MSP switches on  Most binary MSPs have short orbital periods and mass function identifying the companions as low mass evolved dwarfs X-ray transients can be the missing link between LMXBs and MSPs! Old Neutron stars spin up by accretion from a companion Radio Pulsar Millisecond Radio Pulsar Spin up by mass accretion Accreting NS in LMXBs are conventionally thought to be the progenitors of millisecond or „recycled“ radio pulsars (Alpar et al. 1982) NS day, ParisJune 27, 2007

6. 6 INTEGRAL IBIS/ISGRI Observation IGR J00291+5934 (20 - 40 keV) V709 Cas Cas Gamma 2S0114+650 Cas A IGR J00291+5934 (80 - 200 keV) (40-80 keV) significance level ~51σ (80-200 keV) significance level ~17σ Rev 261/262/263/264 ; Exposure 343 ks December 2004 Outburst (20-40 keV) significance level ~88σ derived angular distance: 18´ NS day, ParisJune 27, 2007

7. 7 OUTBURST PROFILE Discovery (Eckert et al. 2004) From RXTE (Galloway et al. 2005) (Falanga, Kuiper, Poutanen et al. 2005) ISGRI 20-100 keV PULSE PROFILE IGR J00291+5934 Rev 261/262/263, ~205 Porbit = 2.457 hr Ps = 1.67 ms Pdot = +8.4 x 10 -13 Hz/s NS day, ParisJune 27, 2007

8. 8 Geometry of the emission region XTE J1807-294 Thermal disk emission The plasma is heated by the accretion shock as the material collimated by the hotspot on to the surface. The seed photons for Comptonization are provided by the hotspot. Seed photons from the hotspot Thermal Comptonization in plasma of Temperature ~ 40 keV B ~ 10 8 G RmRm (Falanga, Bonnet-Bidaud, poutanen et al. 2005) θ NS day, ParisJune 27, 2007

9. 9 Pulsed fraction and Time lag : IGR J00291+5934 (Falanga et al. 2005) If the spectrum has a sharp cutoff, the rms amplitude of the pulse at energies above the cutoff increases dramatically. F(E) ≈E -(Γ 0-1 ) exp(-[E/E c ] β ), Componization photon index Γ(E) = Γ 0 + β(E/E c ) β (Falanga, Kuiper, Poutanen et al. 2005) NS day, ParisJune 27, 2007

10. 10 Time/Phase Lag Model Accretion column Disk Disk soft photons Soft photons Neutron Star Hard photons 1-C ill Hard photons C ill θ hot θ ref Compton cloud (Falanga & Titarchuk 2007) ∆t(C ill,  ref,  hot,n e ref,n e hot ) = upscattering lag + downscattering lag NS day, ParisJune 27, 2007

11. 11 Spin-up IGR J00291+5934 We measured for the first time a spin-up for an accreting X-ray millisecond Pulsar υ = + 8.4 × 10 -13 Hz s -1 υ = + 3.7 × 10 -13 (L 37 /η -1 I 45 ) (R m /R co ) 1/2 (M/1.4M sun ) (υ spin /600) -1/3 Hz s -1 NS day, ParisJune 27, 2007

12. 12 Is the Spin-up real? An error in the source coordinates can give rise to timing error which may introduce a spurious spin-up or spin-down 1 Year Our observation υ = + 5.8 × 10 -14 Hz s -1 0.2 arcsec 0.092 arcsec 0.2 arcsec 0.092 arcsec 0.092 arcsec source position error would introduce a non-existant spin-up rate of Such an apparent spin-up would require a fairly large ~ 0.7 arcsec source position error during our observation YES NS day, ParisJune 27, 2007

13. 13 Pulsar spin-up Animation NASA, D. Barry NS day, ParisJune 27, 2007

14. 14 Important Questions  Missing link between LXMB and ms radio pulsar ?  Analysis suggests that the spin frequency is limited to 760 Hz (95% confidence; Chakrabarty et al. 2003 )  Several have suggested that gravitational radiation from a non- spherical neutron star might limit the maximum fraquency (Bildsten et al. 1998)  Detection by LISA?  Detecting more of these source with more instrument than before NS day, ParisJune 27, 2007

15. 15 Thank You… NS day, ParisJune 27, 2007

16. 16 Spectral analysis IGR J00291+5934 JEM-X/ISGRI Model Compps N H (cm -2 ) 0.28 x 10 22 (f) kT e (keV) 49 ± 4 kTseed (keV) 1.49 ± 0.24 Optical depth 1.12 ± 0.05 A seed (km 2 ) 20.7 ± 8.5 Cos θ 0.6 ± 0.07 L (0.1-300keV) (10 36 erg s -1 ) 3.7 NS day, ParisJune 27, 2007

17. 17 HETE J1900.1-2455 NS day, ParisJune 27, 2007

18. 18 SAX J1808.4-3658 First millisecond accretion powered X-ray pulsar Rotation period: 2.5 ms Orbital period: 2 hrs Low mass companion Wijnands & van del Klis 1998 Chakrabarty & Morgan 1998 Light curve folded at the spin period For a millisecond pulsar, e.g. SAX-J1808.4- 3658,  can take ~ 20% of 2.5 ms pulse period the time of flight delay cannot be neglected Geometry illustration

19. 19 Constraints on the neutron star mass-radius relation obtained by fitting the pulse profile of SAX J1808.4-3658 (filled circles with error bars) toghether with a set of equations of state ans strange star (Poutantn & Gierlinski 2003) NS day, ParisJune 27, 2007

20. 20 OUTBURST PROFILE XTE J1807-294 Discovery (Eckert et al. 2004) From RXTE (Galloway et al. 2005) (Falanga et al. 2005) (Wijnands 2005, astro-ph/0403409) XTE J1807-294 ISGRI 20-100 keV Outburst are extended as a consequence of X-ray irradiation of the disk ? (King & Ritter 1998) Distinct knee Pulsar Workshop, IAP, ParisMay 3-4, 2007

21. 21 Pulse profile IGR J00291+5934 ISGRI 9.1σ 20-35 keV ISGRI 7.3 σ 35-60 keV ISGRI 5.0 σ 60-100 keV Rev 261/262/263, ~205 Porbit = 2.457 hr Ps = 1.67 ms Pdot = +8.4 x 10 -13 Hz/s ISGRI 2.0 σ 100-150 keV HEXTE 1.1 σ 100.9-151.1 keV HEXTE 3.3 σ 60.1-100.9 keV HEXTE 8.3 σ 35.2-60.1 keV HEXTE 11.4 σ 20.3-35.2 keV JEM-X 4.0 σ 5-10 keV (Falanga et al. 2005) Pulsar Workshop, IAP, ParisMay 3-4, 2007

22. 22 Pulsar spin-up R(magnetosphere) The accreting matter transfers its specific angular momentum (the Keplerian AM at the magnetospheric radius) to the neutron star: L=(GMR m ) 1/2 M The process goes on until the pulsar reaches the keplerian velocity at R m (equilibrium period); P min when R m = R ns The conservation of AM tells us how much mass is necesssary to reach P min starting from a non-rotating NS Accretion regime R m < R cor (Illarionov & Sunyaev 1975) R(corotation) Propeller regime R m > R cor Pulsar Workshop, IAP, ParisMay 3-4, 2007

23. 23 (Falanga et al. 2005) Hard X-ray Soft X-ray Hot coronaAccretion disk Time lag IGR J00291+5934 Compton scattering model Time lag are normally hard The energy spectra often observed in LMXRBs suggests that the dominant radiative mechanism in the system is Compton scattering of soft photons in a hot plasma. (For a review of models for spectral variability and time lags see Poutanen 2001) Pulsar Workshop, IAP, ParisMay 3-4, 2007

24. 24 INTEGRAL CODED MASK (IBIS, SPI & JEM -X) Observation Coded MaskShadow Deconvolved Image Corrected Image End Image

25. 25 « Une étoile cannibale » « Star eats companion »