An accreting pulsar with extreme properties drives an ultraluminous x-ray source in NGC 5907

Abstract

Ultraluminous x-ray sources (ULXs) in nearby galaxies shine brighter than any X-ray source in our Galaxy. ULXs are usually modeled as stellar-mass black holes (BHs) accreting at very high rates or intermediate-mass BHs. We present observations showing that NGC 5907 ULX is instead an x-ray accreting neutron star (NS) with a spin period evolving from 1.43 s in 2003 to 1.13 s in 2014. It has an isotropic peak luminosity of ∼1000 times the Eddington limit for a NS at 17.1 Mpc. Standard accretion models fail to explain its luminosity, even assuming beamed emission, but a strong multipolar magnetic field can describe its properties. These findings suggest that other extreme ULXs (x-ray luminosity ≥1041 erg s −1 ) might harbor NSs.Ultraluminous x-ray sources (ULXs) are observed in off-nucleus regions of nearby galaxiesand have x-ray luminosities in excess of a few 10 39 erg s −1 , which is the Eddington luminosity (L Edd ) for a black hole (BH) of 10 M (1). The L Edd sets an upper limit on the accretion luminosity (L acc ) of a compact object steadily accreting, since for L acc > L Edd accretion will be halted by radiation forces. For spherical accretion of fully ionized hydrogen, the limit can be written as, where σ T is the Thomson scattering cross section, m p is the proton mass, and M/M is the compact object mass in solar masses; for a 1.4 M neutron star (NS), the maximum accreting luminosity is ∼2×10 38 erg s −1 .The high luminosity of ULXs has thus been explained as accretion at or above the Eddington luminosity onto BHs of stellar origin (<80-100 M ), or onto intermediate-mass (10BHs (2, 3). However, if the emission of ULXs were beamed over a fraction b < 1 of the sky, their true luminosity, and thus also the compact object mass required not to exceed L Edd , would be reduced by the same factor. This possibility, together with the recent identification of two accreting NSs associated with the ∼10 40 erg s −1 M82 X-2 (4) and NGC 7793 P13 (5, 6) x-ray sources, have brought support to the view that most low-luminosity ULXs likely host a NS (7) 2 or a stellar-mass BH (8). For the most extreme ULXs with x-ray luminosity exceeding a few ×10 40 erg s −1 , BHs with masses in excess of 100 M are still commonly considered (9, 10).Despite several searches for coherent x-ray pulsations,no other ultraluminous x-ray source has been found to host a NS so far (11).Within the framework of "Exploring the X-ray Transient and variable Sky", EXTraS (12) Fig. 1 and Table 1). In all cases, a strong first period derivative term is present (see Table 1). The pulse shape is nearly sinusoidal, while the pulsed fraction (the semi-amplitude of the sinusoid divided by the average count rate)is energy dependent and increases from about 12% at low energies (<2.5 keV) to ∼20% in the hard band (>7 keV; Fig. 1).To derive constraints on the orbital period (P orb ), we applied a likelihood analysis to the two 2014 NuSTAR observations (see supplementary online text), which have the longest baseline. 3By assuming a circular orbit (as in the case of M...

Super-Eddington accretion on to a magnetized neutron star

“…There are episodes of spin-up and spin-down mentioned in the paper, so we can conclude that the object is close to the equilibrium (see also Ekşi et al 2015, Lyutikov 2014and Dall'Osso et al 2015. The most interesting ULX-pulsar in NGC 5907 has the maximal observational spin-up rate of |ṗ obs | = 9.6 × 10 −9 s s −1 (Israel et al 2017a) which is two times smaller than we obtain from equation (69) |ṗs| = 2 × 10 −8 s s −1 . If we trust the luminosity estimate, this object should be in strong spin-up.…”

“…The third ULX-pulsar in NGC 5907 (Israel et al 2017a) has a very high apparent luminosity of LX = 1.5 × 10 40 − 10 41 erg s −1 and its period was changing very rapidly from p1 = 1.43 s to p2 = 1.13 s during past 11 years. From equation (65) we can obtain the minimal magnetic field that can explain such a large luminosity, µ = 6 × 10 34 G cm 3 which is rather unphysical.…”

Spectra of black hole accretion models of ultraluminous X-ray sources

“…One possibility is that the photon emission of super-critical neutron stars and black holes is strongly collimated along the polar axis, and appears highly super-Eddington only for observers located in that direction; a beaming factor scaling as Ṁ /Ṁ Edd 2 was proposed by King (2009). Another scenario (Israel et al 2016), specific to neutron stars, is that the classical Eddington limit is not a barrier to accretion onto a highly magnetized neutron star because the electron scattering cross section (and therefore the effect of radiation pressure) is reduced for photon energies in the X-ray band, in the presence of a magnetic field B 10 12 G (Herold 1979). …”

“…While one or two ULXs may well be intermediate mass BHs (e.g., HLX-1: Farrell et al 2009;Godet et al 2009;Davis et al 2011), the more recent consensus (see Bachetti 2016;Feng & Soria 2011 for reviews, and King et al 2001;Begelman et al 2006;Poutanen et al 2007 for theoretical arguments) is that the vast majority of ULXs are stellar-mass (∼ 10M ) BHs, accreting above their Eddington limit. But not all ULXs are BHs: three objects show coherent pulsations and are thus neutron stars Israel et al 2017Israel et al , 2016Fürst et al 2017Fürst et al , 2016. It is unclear what fraction of ULXs belong to this class (King & Lasota 2016).…”

On the radiation beaming of bright X-ray pulsars and constraints on neutron star mass–radius relation

“…The recently discovered pulsating ULXs, whose accretion luminosity is known to be as high as ∼ 10 40 − 10 41 erg s −1 (Bachetti 2014;Israel 2017a), are the brightest accreting NSs known up to date. It very likely that their central engine is formed by accretion columns confined by a strong magnetic field (Mushtukov et al 2015a).…”