Intermittent Accreting Millisecond Pulsars

Introduction

An exciting new type of sources Pulsations appear and disappear either on timescal

es of hundreds of seconds or on timescales of days In the early 1980s it was predicted that the progenito

rs of millisecond radio pulsars should be accreion-powered millisecond X-ray pulsars in neutron star low-mass X-ray binaries

The first one was discovered in 1998 A total of seven AMSPs have been found Recently three intermittent AMXPs were found

HETE J1900.1-2455

A transient source Remain active for ~3 years but pulsations were dete

cted during only the first ~ 70 days On three occasions during this period, it was observ

ed an abrupt increase in the pulse amplitude, approximately coincident with the time of a thermonuclear burst, followed by a steady decrease on a timescale of 10 days

Pulsations at 377.3 Hz and in an 83.3 minute circular orbit with a companion of mass greater than 0.016 M_sun and likely less than 0.07 M_sun

No burst oscillations have been detected

ApJ638,963-967, 2006

Aql X-1

A ~550 Hz burst oscillation transient source from which pulsations were detected only for ~150 sec out of the ~1.5 Msec

The pulsations were discovered at the end of a one-orbit ~1500 seconds onservation

Given the data structrue, we cannot exclude the occurrence of X-ray burst ~1400 seconds before the pulsations or immediately after the pulsations disappeared

ApJ674, L41-L44, 2008

The Scenarios Most Often Considered The magnetic field in nonpulsating LMXBs is too we

ak to allow channeling of the matter onto the magnetic poles

The magnetic field has comparable strength inside most LMXB neutron star, but in the large majority of them, it has been “buried” by accretion

Pulsations are produced in all LMXBs, but in most of them they are attenuated by a surrounding scattering medium that washes out the coherent beamed pulsation

The pulsations are attenuated by gravitational lensing from the neutron star

SAX J 1748-2021

A transient source with a frequency of 442.36 Hz from which intermittent pulsations were detected during both 2001 and 2005 outbursts

The pulsations appeared and disappeared on timescales of hundreds of seconds

A suggestive relation between the occurrence of type-I X-ray burst and the appearance of the pulsations was found, but the relation is not strict

The reason for the intermittence of the pulsations remain unclear It is now evident that a strict division between pulsating and nonp

ulsating neutron star systems does not exist An orbit with a period of 8.7 hr The companion star might be a main-sequence or a slightly evolv

ed star with a mss of ~1 M_sun No burst oscillations have been detected

ApJ674, L45-L48, 2008

The Neutron Star Transient SAX J1748.9-2021 In NGC 6440

NGC 6440 is a GC at 8.5±0.4 kpc. Bright X-ray outbursts from a LMXB were rep

orted in 1971, 1998, 2001, and 2005 In’t Zand et al. (2001)), from X-ray and optical

observations, concluded that the 1998 and 2001 outbursts were from the same object, which they designated SAX J1748.9-2021

Colors, Light Curves, and States The PCA observations sample three different

outbursts The color-color diagrams show a pattern typic

al for atoll sources No kHz QPOs have been found No thermonuclear bursts were detected in the

first outburst, 16 during the second , and four during the third

No burst oscillations have been found

Pulsations

Best Parameters Estimates

The Pulse Frequency As A Function Of Orbital Phase

Discussion

From the Doppler shifts on the pulsations the authors determine that the system is in a near-orbit with period of 8.7 hr and projected radius of 0.39 lt-s

The stability of the pulsations strongly suggests that the pulsation frequency reflects the neutron star spin frequency and that SAX J 1748.9-2021 is an accreting millisecond X-ray pulsar

In SAX J1748.9-2021 and Aql X-1 the timescales on which the pulse amplitude can fluctuate are as short as ~100 s, too short for the properties of the neutron star core to change. So these changes must orginate in the disk or the outer layers of the neutron star envelope

Pulsations only around a mass accretion rate of ~2*10^(16) M_sun/s, indicating that instantaneous mass accretion rate rather than total accreted mass is the important quantity

Discussion

SOURCE STATES AND PULSATIONS Two sub-types of NS LMXBs: the Z sources and the

atoll sources The Z source are the most luminous while the atoll s

ources cover a much wider range in luminosities In the case of atoll sources, the main three states: E

IS, IS, BB All AMSPs with persistent pulsations have been obs

erved only during their island or extreme island state The 3 intermittent AMSPs are the only pulsating sou

rces that have been observed not only in there island states, but also in their banana states

POWER SPECTRA IN THE INTERMITTENT AMSPS The shape of the power spectra between AMSPs an

d non-AMSPs are very similar The frequency correlations for AMSPs and non-AM

SPs and non-AMSPs are found to be shifted in frequency and only one non-AMSP might show frequency shifts as those seen in AMSPs

Confirming whether the pulse mechanism in AMSPs is related or not with these frequency shifts might give important clues that help us understand the differences between AMSPs and non-AMSPs

2005, ApJ 634, 1250-1260

POWER SPECTRA IN THE INTERMITTENT AMSPS For SAX J1748-2021 most of the power spectra are

typically of the (upper) banana state, during both pulsating and non-pulsating periods; the low-frequency noise seems to be weaker when pulsations are present

The average power spectrum of the observation of Aql X-1 in which pulsations were detected is typical of the upper banana state

Most of the power spectra of HETE J 1900.1-2455 are like those typically seen during the EIS and IS

They found identical (within errors) power spectra during the pulsating and the non-pulsating periods. Their results suggest that frequency correlations within a given intermittent source will not differe between pulsating and non-pulsating periods

Questions

Can a unique scenario explain all the intermittency observed? Is there a relation between the pulse mechanism and the thermo

nuclear burning on the neutron star surface? What is the relation between the pulse mechanism and the burst

oscillation mechanism? Interestingly, bursts and burst oscillations have been detected in

the AMSPs SAX J1808.4-3658 and XTE J1814-338 and bursts without burst oscillations in the two intermittent AMSPs SAX J1748-2021 and HETE J1900.1-2455

The only connecting system might be Aql X-1, since no X-ray bursts have been detected in the other AMSPs

Does the pulse mechanism affect in any way the observed aperiodic variability?