IGR J18245-2452

• Swift refinement of the X-ray position
• Inside the core of the globular cluster M28
• inconsistent with a previously known quiescent X-ray binary
• Substantial X-ray variability > factor 2 in less than 100s
• Spectral analysis => N_H=4.4 x 10²¹ cm-2 Γ=1.41

• Swift observation, after trigger by BAT
• Γ=1.35, N_H=4.4 x10²¹ cm-2, in window timing mode
• Count rate around 10 cts/s showing two large flares reaching 20 and 40 cts/s

• Search for timing signatures in XRT window timing mode data
• Possible modulation at 6.6 mHz (150s) in first observation
• => possible quasi periodic signal or orbital Doppler modulation

• Swift detection of a type I X-ray burst from a position consistent with that of the IGR
• Other source not excluded but no other active source in M28 at the moment
• Burst spectrum => 2.5 keV bbody

• Time resolved spectroscopy of the burst
• =>thermal spectrum kT decreasing from 2.7 to 1.1 keV
• Cooling tail => thermonuclear burst
• => LMXB +NS; 1st of such object in M28
• Γ of the XRT spectrum consistent with low luminosity (Atoll) LMXBs

• 2 X-ray bursts detected by Maxi

• Optical 'amateur' observations => brightened star at Dec = 18h 24m 32.93s, -24deg 51' 59.6" (with an astrometric uncertainty of less than <0.03" in RA and <0.5" in Dec)
• Inferred absolute visual magnitude ~ -2 => in the range of that found for active low-mass X-ray binaries
• Source is however 11.6" away from Swift position
• no optical sources found at Swift position

• ATCA observations at 5.5 and 9 GHz
• Detection of a source at both frequency at 4.8" from the best XRT position (that has 3.5" error)
• Mean flux densities 0.62 & 0.75 mJy at 9 & 5.5 GHz respectively.
• Mean spectral index ~ 0.4
• ATCA source = only radio source detected within the inner core of the cluster
• Strongly variability over the 90 min observation
• None of the 12 reported pulsars in this area has position compatible with the radio source
• Association with the IGR suggested even if radio detection puzzling given the onset of thermonuclear bursts subsequent to the radio observation.

• HST observations of the field at 3 different epochs
• Possible optical counterpart => Main sequence star with unexpectedly large variations
• Source at 2.5σ from Swift, consistent with ATCA and a Chandra source (source 23 in Becker et al. 2003, ApJ, 594, 798)
• Variability during each observations

• Chandra ACIS-S observation
• 1 single source within 0.5" ATCA error circle=> source 23 from Becker et al. 2003, ApJ, 594, 798
• Fit with absorbed power law => N_H~ 1.8e21cm-2 Γ~1.52, with an obtained flux significantly higher than that of source 23 in Becker et al. 2003

• Detection of burst oscillations at a frequency of 254.4 Hz
• Burst oscillations observed by Swift/XRT during a thermonuclear (type I) burst

• Radio monitoring with GBT at 2 GHz and Parkes and WSRT at 1.4 GHz
• Radio pulsar detected sporadically by each telescope
• Non detection of the source at orb phase 0.25 probably due to eclipsing
• => These systems switch back and forth between accreting binary and radio pulsar, on short time scales

• Identification of otpical counterpart
• Multi epoch HST observations => detection during quiescent and outburst state
• Quiescence => unperturbed main sequence star
• Active=> Hα excess suggest active accretion

• IGR source and PSR 1824-2425I (in cluster M28) have identical ephemerides => sources associated
• Pulsations in IGR not powered by rotation of B field (as in the radio pulsar)
• IGR compatible with X-ray accreting ms pulsar
• XMM observations => rapid switch between low and high intensity sates (factor of 100 variations on ~s times scales)
• => transition between accretion and rotation powered regimes.