Coherent Pulsations and Burst Oscillations in the Millisecond Pulsar IGR J18245-2452/PSR J1824-2452I in M28
ATel #5086; A. Riggio, L. Burderi (DSF-UNICA), T. Di Salvo, A. D'Ai' (DiFC-UNIPA), A. Papitto (IEEC-CSIC), R. Iaria, N. Robba (DiFC-UNIPA), A. Sanna (DSF-UNICA), E. Bozzo, C. Ferrigno (ISDC-Geneve), N. Rea (IEEC-CSIC), L. Pavan (ISDC-Geneve)
on 24 May 2013; 21:42 UT
Credential Certification: Alessandro Riggio (riggio@dsf.unica.it)
Subjects: X-ray, Binary, Globular Cluster, Pulsar
We analyzed Swift observations of the transient low-mass X-ray binary
IGR J18245-2452 and millisecond radio pulsar PSR J1824-2452I in M28
(Atel #4925, #4927, #4929, #4934, #4959, #4960, #4961, #4964, #4981,
#5003, #5031, #5045, #5068, #5069). Available Swift observations,
performed in timing mode, with a time resolution of about 1.78 ms, and
covering the period from 56389.92457 to 56389.93391 MJD(TDB), have
been barycentered using the coordinates from the radio counterpart
identified by the ATCA observation taken on 2013 April 5 (Atel
#4981). We corrected the Swift/XRT arrival times using the orbital
parameters derived from the XMM-Newton observation in which X-ray
coherent pulsations at about 3.93 ms were firstly detected (see Atel
#5069, Papitto et al. 2013, arXiv:1305.3884).
The considered observation contains a type-I burst (Atel #4959),
starting at 56389.9292192 MJD(TDB), with a fast rise <= 10s and an
exponential decay on a time scale of ~39 s. We folded the data
before the start of the type-I burst at the spin period of the
source reported in Papitto et al. 2013 (arXiv:1305.3884),
finding a quasi sinusoidal pulse profile with a pulse fraction
of about 7(2)%. The pulsation is detected with a significance
of 3.5 sigma. The averaged count rate was 11 c/s.
Using a moving window we determined that burst oscillations (see Atel
#5068) were present in two time intervals during the burst decay. The
first one started 44s after the beginning of the burst, and lasted 26s,
and the second one started 80s after the beginning of the burst,
and lasted 30s. The burst oscillation was detected with a significance
of 3.4 and 4.4 sigma and a pulsed fraction of 9(3)% and 20(4)%,
respectively. The averaged count rates were 128 c/s and 46 c/s,
respectively.
The phase differences between the coherent pulsations before of the
beginning of the burst and the burst oscillations are 0.06(5) and
0.09(5), respectively. We therefore conclude that the burst
oscillations, when detected, were phase-locked with the coherent
pulsations within the uncertainties.
We folded the data starting 150s after the beginning of the type-I
burst (where burst contribution was less than 50% of the intensity),
and lasting 255s, up to the end of the considered swift observation.
We did not detect the pulsation (8% 3-sigma upper limit on the pulsed
fraction). The averaged count rates was 13 c/s.
The pulse profiles derived from the Swift observations before the burst
(figure 1) and during the burst (figure 2 and 3) can be found at this
link: http://www.dsf.unica.it/~riggio/figure1.pdf http://www.dsf.unica.it/~riggio/figure2.pdf http://www.dsf.unica.it/~riggio/figure3.pdf