Danny T. Koh

Observations of accreting pulsars

We discuss recent observations of accreting binary pulsars with the all-sky BATSE instrument on the Compton Gamma Ray Observatory. BATSE has detected and studied nearly half of the known accreting pulsar systems. Continuous timing studies over a two-year period have yielded accurate orbital parameters for 9 of these systems, as well as new insights into long-term accretion torque histories.

Torque Reversal and Spin-Down of the Accretion-Powered Pulsar 4U 1626-67

Over 5 yr of hard X-ray (20-60 keV) monitoring of the 7.66 s accretion-powered pulsar 4U 1626-67 with the Compton Gamma Ray Observatory/BATSE large-area detectors has revealed that the neutron star is now steadily spinning down, in marked contrast to the steady spin-up observed during 1977-1989. This is the second accreting pulsar (the other is GX 1+4) that has shown extended, steady intervals of both spin-up and spin-down. Remarkably, the magnitudes of the spin-up and spin-down torques differ by only 15%, with the neutron star spin changing on a timescale |ν/dot ν| ≈ 5000 yr in both states. The current spin-down rate is itself decreasing on a timescale |dot ν/bar ν| ≈ 26 yr. The long-term timing history shows small-amplitude variations on a 4000 day timescale, which are probably due to variations in the mass transfer rate. The pulsed 20-60 keV emission from 4U 1626-67 is well-fitted by a power-law spectrum with photon index γ = 4.9 and a typical pulsed intensity of 1.5 × 10^(-10) ergs cm^(-2) s^(-1). The low count rates with BATSE prohibited us from constraining the reported 42 minute binary orbit, but we can rule out long-period orbits in the range 2 days lesssim Porb lesssim 900 days. We compare the long-term torque behavior of 4U 1626-67 to other disk-fed accreting pulsars and discuss the implications of our results for the various theories of magnetic accretion torques. The abrupt change in the sign of the torque is difficult to reconcile with the extremely smooth spin-down now observed. The strength of the torque noise in 4U 1626-67, ~10^(-22) Hz^2 s^(-2) Hz^(-1), is the smallest ever measured for an accreting X-ray pulsar, and it is comparable to the timing noise seen in young radio pulsars. We close by pointing out that the core temperature and external torque (the two parameters potentially relevant to internal sources of timing noise) of an accreting neutron star are also comparable to those of young radio pulsars.

Rapid Spin-Up Episodes in the Wind-Fed Accreting Pulsar GX 301-2

The accreting pulsar GX 301-2 (P = 680 s) has been observed continuously by the large-area detectors of the Burst and Transient Source Experiment (BATSE) instrument on the Compton Gamma Ray Observatory since 1991 April 5. Orbital parameters determined from these data are consistent with previous measurements, with improved accuracy in the current orbital epoch. The most striking features in the pulsar frequency history are two steady and rapid spin-up episodes, with ν˙~(3-5)×10^(-12) Hz s^(-1), each lasting for about 30 days. They probably represent the formation of transient accretion disks in this wind-fed pulsar. Except for these spin-up episodes, there are virtually no net changes in the neutron star spin frequency on long timescales. We suggest that the long-term spin-up trend observed since 1984 (ν˙~2×10^(-13) Hz s^(-1)) may be due entirely to brief (~20 days) spin-up episodes similar to those we have discovered. We assess different accretion models and their ability to explain the orbital phase dependence of the observed flux. In addition to the previously observed preperiastron peak at orbital phase 0.956 +/- 0.022, we also find a smaller peak close to apastron at orbital phase 0.498 +/- 0.057. We show that if the companion stars effective temperature is less than 22,000 K, then it must have a mass M_c < 70 M_⊙ and a radius R_c < 85 R_⊙ so as not to overfill the tidal lobe at periastron. In order not to overflow the Roche lobe at periastron, the corresponding values are M_c < 55 M_⊙ and R_c < 68 R_⊙. These constraints are nearly at odds with the reclassification by Kaper et al. of the companion as a B1 Ia + hypergiant.

On the Dramatic Spin-up/Spin-down Torque Reversals in Accreting Pulsars

Dramatic torque reversals between spin-up and spin-down have been observed in half of the persistent X-ray pulsars monitored by the BATSE all-sky monitor on the Compton Gamma Ray Observatory. Theoretical models developed to explain early pulsar timing data can explain spin-down torques via a disk-magnetosphere interaction if the star nearly corotates with the inner accretion disk. To produce the observed BATSE torque reversals, however, these equilibrium models require the disk to alternate between two mass accretion rates, with ˙M5 producing accretion torques of similar magnitude but always of opposite sign. Moreover, in at least one pulsar (GX 114) undergoing secular spin-down, the neutron star spins down faster during brief (~20 day) hard X-ray flares—this is opposite the correlation expected from standard theory, assuming that BATSE pulsed flux increases with mass accretion rate. The 10 day to 10 yr intervals between torque reversals in these systems are much longer than any characteristic magnetic or viscous timescale near the inner disk boundary and are more suggestive of a global disk phenomenon. We discuss possible explanations of the observed torque behavior. Despite the preferred sense of rotation defined by the binary orbit, the BATSE observations are urprisingly consistent with an earlier suggestion for GX 1+4: the disks in these systems somehow alternate between episodes of prograde and retrograde rotation. We are unaware of any mechanism that could produce a stable retrograde disk in a binary undergoing Roche lobe overflow, but such flip-flop behavior does occur in numerical simulations of wind-fed systems. One possibility is that the disks in some of these binaries are fed by an X-ray–excited wind.

On the Correlation of Torque and Luminosity in GX 1+4

Over 5 years of daily hard X-ray (>20 keV) monitoring of the 2 minute accretion-powered pulsar GX 1+4 with the Compton Gamma Ray Observatory/BATSE large-area detectors has found nearly continuous rapid spin-down, interrupted by a bright 200 day spin-up episode. During spin-down, the torque becomes more negative as the luminosity increases (assuming that the 20-60 keV pulsed flux traces bolometric luminosity), the opposite of what is predicted by standard accretion torque theory. No changes in the shape of the 20-100 keV pulsed energy spectrum were detected, so that a very drastic change in the spectrum below 20 keV or the pulsed fraction would be required to make the 20-60 keV pulsed flux a poor luminosity tracer. These are the first observations that flatly contradict standard magnetic disk accretion theory, and they may have important implications for understanding the spin evolution of X-ray binaries, cataclysmic variables, and protostars. We briefly discuss the possibility that GX 1+4 may be accreting from a retrograde disk during spin-down, as previously suggested.

On the Dramatic Spin-Up/Spin-Down Torque Reversals in BATSE Observations of Accretion Powered Pulsars

X-ray pulsars are the only accreting magnetic stars where rotation torques induced by accretion are large enough to be measured on short timescales ~ days. They are thus unique laboratories for studying the interaction between an accretion disk and a stellar magnetosphere. We describe 5 years of continuous pulsar timing observations by the BATSE instrument on GRO which paint a strikingly different picture of pulsar spin behavior than understood from the previous 20 years of sparse observations. In particular, we find that more than half of the persistent pulsars we observe undergo dramatic torque reversals, switching suddenly between extended periods of steady spin-up and steady spin-down. Moreover, variations in pulsed flux are anticorrelated with torque in at least one system undergoing secular spin-down, GX1+4. This behavior contradicts standard accretion torque theory (Ghosh and Lamb 1979). A simple – albeit unconventional – hypothesis which naturally explains these observations is that the disks in these systems somehow alternate between epochs of prograde and retrograde rotation.