Discovery of 40.5 ks Hard X-Ray Pulse-phase Modulations from SGR 1900+14

Abstract

X-ray timing properties of the magnetar SGR 1900+14 were studied, using the data taken with Suzaku in 2009 and NuSTAR in 2016, for a time lapse of 114 and 242 ks, respectively. On both occasions, the object exhibited the characteristic two-component spectrum. The soft component, dominant in energies below ∼5 keV, showed a regular pulsation, with a period of P = 5.21006 s as determined with the Suzaku XIS, and P = 5.22669 with NuSTAR. However, in ≳ 6 keV where the hard component dominates, the pulsation became detectable with the Suzaku HXD and NuSTAR only after the data were corrected for periodic pulse-phase modulation, with a period of T = 40 − 44 ks and an amplitude of ≈1 s. Further correcting the two data sets for complex energy dependences in the phase modulation parameters, the hard X-ray pulsation became fully detectable, in 12–50 keV with the HXD and 6–60 keV with NuSTAR, using a common value of T = 40.5 ± 0.8 ks. Thus, SGR 1900+14 becomes a third example, after 4U 0142+61 and 1E 1547−5408, to show the hard X-ray pulse-phase modulation, and a second case of energy dependences in the modulation parameters. The neutron star in this system is inferred to perform free precession, as it is axially deformed by ≈ P/T = 1.3 × 10−4, presumably due to ∼ 1016 G toroidal magnetic fields. As a counterexample, the Suzaku data of the binary pulsar 4U 1626−67 were analyzed, but no similar effect was found. These results altogether argue against the accretion scenario for magnetars.

Novel Clues to the Physics of Magnetars as Probed with Detailed Pulse-Timing Studies

“…To remove a concern that the PPM might appear in any NS, we also analyzed the Suzaku HXD data of the binary pulsar 4U 1626−67; its pulse period (7.7 s) is similar to those of magnetars, and its orbital Doppler effect is negligibly small (< 13 lt-msec). As described in Makishima et al (2021b) and given in Fig. 2 (f), no evidence of PPM was detected.…”

“…1, the SXC pulses of magnetars, reflecting the spin of the neutron stars (NSs), are easily detectable, but those in the HXC often remain insignificant when we rely on the conventional epoch-folding plus Z 2 statistics searches. As we have revealed so far, this is because the HXC pulses of magnetars are very often subject to slow phase modulations, at a period T ∼ 10 4 P (Makishima et al 2014(Makishima et al , 2016(Makishima et al , 2014(Makishima et al , 2021a(Makishima et al , 2021b, where P denotes the pulse period. Since this effect is presumably caused by the magnetic deformation of NSs, its study will provide valuable clues to the magnetars' ultra-strong toroidal magnetic field B t , which cannot be easily studied with other means.…”

“…As we developed in our series of publications from Makishima et al (2014) to Makishima et al (2021b), the results described above are interpreted in the following way. According to Ioka & Sasaki (2004), strong B t will axially elongate the NS to an asphericity of…”

“…Because neither T nor α would change on short time scales, the observed changes in A may result from variations in the HXC anisotropy (Makishima et al 2021a). Makishima et al (2021b) examined some alternative interpretations of the PPM, assuming that magnetars are isolated NSs accreting from fossil disks. However, such scenarios were found difficult to explain the overall results described here.…”

A freely precessing magnetar following an X-ray outburst

“…Despite initial claims of detections of free precession in the pulsars B1642−03 and B1828−11 through the timing of their pulse arrival times [24,25], these pulsars have since then been shown to exhibit magnetospheric mode changes causing the observed profile variation [26,27], complicating the interpretation of their timing behaviour. Free precession has also been invoked to explain the pulse phase modulation with period < 1 day observed in the hard X-ray emission of the magnetars 4U 0142+61 [28], 1E 1547.0−5408 [29] and SGR 1900+14 [30].…”