1 Introduction

The bursting pulsar GRO J1744-28 is a unique source that presents both the characteristics of a type II X-ray burster and a X-ray pulsar. It was discovered near the Galactic center by the Burst and Transient Source Experiment (BATSE) on board the Compton Gamma-Ray Observatory (CGRO) on 1995 December 2 (Fishman et al. 1995; Kouveliotou et al. 1996b), when it was experiencing a very active bursting phase with a rate of about 20 hard X-ray bursts per hour. After one day, the source entered a regime of hourly bursting which lasted until 1996 April 26. Many of these bursts were also detected by the Proportional Counter Array (PCA) instrument on board the Rossi X-ray Timing Explorer (RXTE) (Swank 1996; Giles et al. 1996) and the Oriented Scintillation Spectrometer Experiment (OSSE) instrument on board CGRO (Strickman et al. 1996). Then the burst rate decreased rapidly and the bursting activity ceased for the next 7 months. On 1996 December 1, a second bursting phase started with a similar burst rate to that seen in the first outburst (Woods et al. 1999). It lasted until 1997 April 7. The source was no longer detected by BATSE or RXTE after this date. More than 5800 bursts were observed by BATSE in total.

During the first outburst, persistent hard X-ray emission from GRO J1744-28 was detected by BATSE and OSSE (Paciesas et al. 1996). Finger et al. (1996) measured that GRO J1744-28 is a 467 ms X-ray pulsar in a binary system of orbital period 11.8 days. The pulsar period determination was confirmed with ASCA (Dotani et al. 1996a,b). It has also been found that the X-ray bursts were showing pulsations with the same 467 ms period (Kouveliotou et al. 1996c) and had similar hard X-ray / soft gamma-ray spectra than the persistent emission (Briggs et al. 1996; Briggs et al. 1996), so that the association is firmly established.

The initial source localization was a $6^\circ$ radius error circle (Fishman et al. 1995), which was rapidly refined to a 24' wide annulus by the Interplanetary Network using time-delay measurements between CGRO and other satellites, mainly Ulysses (Hurley et al. 1995). Observations with ASCA provided a $\sim$1' error circle. ROSAT (Kouveliotou et al. 1996a; Augusteijn et al. 1997) improved the source localization to RA $17^{\rm h}44^{\rm m}33\hbox{$.\!\!^{\rm s}$ }1$ and Dec $-28^{\circ}44'29''$ (J2000) with a $\sim$10'' error circle. An optical/near-infrared variable object was observed in 1996 at the periphery of this ROSAT error circle (Augusteijn et al. 1997; Cole et al. 1997) and proposed as a possible counterpart. More recently, Hurley et al. (2000) derived a more precise IPN localization of GRO J1744-28 with a $3\sigma$ error ellipse of $532\ {\rm arcsec^{2}}$ fully contained in the ASCA error circle and partially overlapping with the ROSAT error circle. The optical/near-infrared variable source lies outside the new IPN $3\sigma$ ellipse, which makes it unlikely to be associated with GRO J1744-28.

Finger et al. (1996) determined the mass function of the binary system to be $1.36\times 10^{-4}~{M_{\odot}}$. This supports that GRO J1744-28 is very likely a Low-Mass X-ray Binary (LMXB), where the central neutron star accretes matter through Roche lobe overflow from an evolved low-mass stellar companion (Daumerie et al. 1996; Sturner & Dermer 1996). This is however the only known example of such a system showing both the characteristics of a X-ray pulsar (the persistent pulsed emission) and of a type II X-ray burster. The relative fluences in the bursts and in the persistent emission probably rule out a thermonuclear origin of the bursts, which are more likely of type II. Such bursts are accretion powered and are possibly triggered by an instability in the accretion flow, the nature of which is not clearly identified (Kouveliotou et al. 1996c; Lewin et al. 1996; Cannizzo 1996). This is supported by several similarities between the bursts of GRO J1744-28 and those of the Rapid Burster MXB 1730-355 (Lewin et al. 1996), i.e. the first and unique object where type II bursts were discovered.

In this paper, we report new observations of GRO J1744-28 in the X-ray band, made with XMM-Newton in April 2001 as it was in its quiescent state.

 

 

Table 1: Position of the reference star TYC 6840-38-1 as detected by each of the EPIC instruments. The first three columns (right ascension, declination and statistical error at $1 \sigma$ level) are the result of the standard SAS source detection procedure. The last two columns show the measured offsets with respect to the Tycho-2 catalogue position RA $17^{\rm h}43^{\rm m}51\hbox{$.\!\!^{\rm s}$ }3$and Dec $-28^{\circ }46'38''$.

Inst.	RA	Dec	stat.	error	error
			error	(RA)	(Dec)
PN	$17^{\rm h}43^{\rm m}51\hbox{$.\!\!^{\rm s}$ }3$	$-28^{\circ}46'40''$	$\sim$0.6''	-0.3''	+1.8''
MOS1	$17^{\rm h}43^{\rm m}51\hbox{$.\!\!^{\rm s}$ }4$	$-28^{\circ }46'38''$	$\sim$0.7''	-1.2''	+0.1''
MOS2	$17^{\rm h}41^{\rm m}51\hbox{$.\!\!^{\rm s}$ }4$	$-28^{\circ}46'37''$	$\sim$0.7''	-1.7''	-0.8''