[ Previous | Next ]
ATEL # 2564; Craig B. Markwardt (UMCP/NASA GSFC), Wayne H. Baumgartner (UMBC/NASA GSFC), Gerald K. Skinner (UMCP/NASA GSFC), Robin H. D. Corbet (UMBC/NASA GSFC)
on 14 Apr 2010; 20:23 UT
Password Certification: Craig B. Markwardt (firstname.lastname@example.org)
Subjects: Far-Infra-Red, X-ray, Binaries, Neutron Stars, Pulsars, Transients
A long-term periodicity was recently reported for the X-ray source AX J1700.2-4220 (Corbet et al. ATEL #2259). Here we report that XMM data show the that the source is a 54 second X-ray pulsar and that RXTE PCA bulge scan data confirm the 44 day apparent orbital period.
We examined archival X-ray data taken by Swift XRT (5900 sec exposure on 2007-02-17), XMM Newton (6900 sec exposure on 2007-02-24), and Chandra ACIS-S (1650 sec exposure on 2007-09-30). We also used RXTE PCA bulge scan light curves covering 2004-2010.
AX J1700.2-4220 was detected in ongoing hard X-ray surveys by both INTEGRAL (Bird et al, 2010, ApJS 186 1) and Swift/BAT (Tueller et al 2010, ApJS 186 378), with a BAT flux of 2.97e-11 erg/s/cm2. In the Chandra and XMM Newton observations, a source is detected which is consistent with the ASCA, BAT and INTEGRAL positions. The refined position derived from Chandra is, R.A. = 17h00m19.30s, Decl. = -42o20'19.2" (J2000), with a position error of approximately 1 arcsec. No other X-ray sources of comparable flux lie within the BAT error circle of 5 arcmin. Therefore we believe the BAT periodicity is correctly identified with AX J1700.
A timing analysis of the XMM data reveals a period at 54.22 ± 0.03 seconds (not barycentered). This period appears independently in data from each of three detectors. The pulse semi-amplitude is about 25% r.m.s. As a cross-check for spurious read-out related periodicities, we analyzed background-only data from the same PN node; no similar modulations were detected. We interpret the 54 sec signal as the X-ray pulse period, and the 44 day signal as the orbital period of this system.
The Chandra ACIS data set for this source contains only 70 photons. With these data we can only set an upper limit on the pulsed r.m.s. of 25% (2 sigma), which is marginally consistent with the XMM result.
The Swift XRT observation did not detect the source. The Swift, Chandra and XMM observations occurred at orbital phases 0.611, 0.723, 0.778, respectively, according to the Corbet et al ephemeris. The Swift XRT observation is near the BAT hard X-ray minimum, which may explain the non-detection.
We produced an X-ray light curve of AX J1700 using the RXTE PCA bulge scan data. The 2-10 keV light curve shows a clear orbital modulation at the Corbet et al period. A periodogram shows a peak at 44.03 ± 0.14 days. Adjusting for a flux bias by assuming the minimum flux is zero, the mean flux is about 1 mCrab (2-10 keV), and the peak flux is around 3 mCrab.
This combination of spin and orbital period places AX J1700 among known Be high-mass X-ray binaries. The Chandra- and XMM-detected point source position is inconsistent with previously suggested counterparts (e.g. HD 153295; Neguerela & Schurch 2007, A&A 461, 631). A brief examination of the 2MASS full survey data does not reveal an obvious point source counterpart at the Chandra position in the J-band, but there is an indication of a faint uncataloged source in the K band (K~14.5). Further work would be needed to clarify the nature of the possible counterpart.
The soft and hard X-ray fluxes correspond to an average 2-50 keV luminosity of about ~1036 erg/s x (d/10 kpc)2, where d is distance. The peak luminosity is 2-3 times higher. The weakness of this source indicates that the mass accretion rate is low (~10-10 solar mass / yr), or that the distance is large.