No abstract

No abstract

We report on new and archival X-ray and near-IR observations of the anomalous X-ray pulsar 1E 1048.1À5937 performed between 2001 and 2007 with the Rossi X-Ray Telescope Explorer (RXTE ), the Chandra X-Ray Observatory, the Swift Gamma-Ray Burst Explorer, the Hubble Space Telescope (HST ), and the Very Large Telescope. Monitoring with RXTE revealed that following its $2001Y2004 active period, 1E 1048.1À5937 entered a phase of timing stability; at the same time, simultaneous observations with Chandra and HST in 2006 showed that its X-ray and near-IR radiative properties, all variable prior to 2005, stabilized. Specifically, the 2006 X-ray spectrum is consistent with a two-component blackbody plus power law, with an average kt ¼ 0:52 keV and À ¼ 2:8, at a mean flux level of $6:5 ; 10 À12 erg cm À2 s À1 (2Y10 keV ). The near-IR counterpart in 2005Y2006 was detected at H $ 22:7 mag and K s $ 21:0 mag, considerably fainter than previously measured. In 2007 March, this newfound quiescence was interrupted by sudden X-ray flux, spectral, and pulse morphology changes, simultaneous with a large glitch and near-IR enhancement. Our RXTE observations revealed a factor of $3 increase in pulsed flux (2Y10 keV ), while observations with Chandra and Swift saw the total X-ray flux increase much more than the pulsed flux, reaching a peak value of >7 times the quiescent value (2Y10 keV ). We find a strong anticorrelation between X-ray flux and pulsed fraction, and a correlation between X-ray spectral hardness and flux. Simultaneously with the radiative and timing changes, we observed the X-ray pulse profile change significantly from nearly sinusoidal to having multiple peaks. We compare these remarkable events with other magnetar outbursts and discuss implications in the context of AXP emission models.

ABSTRACT4U 0142+61 is one of the Anomalous X-ray Pulsars exhibiting hard X-ray emission above 10 keV discovered with INTEGRAL. In this paper we present detailed spectral and temporal characteristics both in the hard X-ray (>10 keV) and soft X-ray (<10 keV) domains obtained using data from INTEGRAL, XMM-Newton, ASCA and RXTE. Accumulating data collected over four years with the imager IBIS-ISGRI aboard INTEGRAL, the time-averaged total spectrum shows a power-law like shape with photon index Γ = 0.93 ± 0.06. 4U 0142+61 is detected up to 229 keV and the flux between 20 keV and 229 keV is (15.01 ± 0.82) × 10 −11 erg cm −2 s −1 , which exceeds the energy flux in the 2−10 keV band by a factor of ∼2.3. Using simultaneously collected data with the spectrometer SPI of INTEGRAL the combined total spectrum yields the first evidence for a spectral break above 100 keV. Assuming a logparabolic function for the spectral shape above 20 keV the peak energy of 4U 0142+61 is 228 +65 −41 keV. There is no evidence for significant long-term time variability of the total emission from 4U 0142+61. Both the total flux and the spectral index are stable within the 17% level (1σ). Pulsed emission is measured with ISGRI up to 160 keV. The 20−160 keV profile shows a broad double-peaked pulse with a 6.2σ detection significance. The total pulsed spectrum can be described with a very hard power-law shape with a photon index Γ = 0.40 ± 0.15 and a 20−150 keV flux of (2.68 ± 1.34) × 10 −11 erg cm −2 s −1 . To perform accurate phase-resolved spectroscopy over the total X-ray window, we produced pulse profiles in absolute phase for INTEGRAL-ISGRI, RXTE-PCA, XMM-Newton-PN and ASCA-GIS. The two known pulses in all soft X-ray profiles below 10 keV are located in the same phases. Three XMM-Newton observations in 2003−2004 show statistically identical profiles. However, we find a significant profile morphology change between an ASCA-GIS observation in 1999 following a possible glitch of 4U 0142+61. This change can be accounted for by differences in relative strengths and spectral shapes (0.8−10 keV) of the two pulses. The principle peak in the INTEGRAL pulse profile above 20 keV is located at the same phase as one of the pulses detected below 10 keV. The second pulse detected with INTEGRAL is slightly shifted with respect to the second peak observed in the soft X-ray band. We performed consistent phase-resolved spectroscopy over the total high-energy band and identify at least three genuinely different pulse components with different spectra. The high level of consistency between the detailed results from the four missions is indicative of a remarkably stable geometry underlying the emission scenario. Finally, we discuss the derived detailed characteristics of the high-energy emission of 4U 0142+61 in relation to three models for the non-thermal hard X-ray emission.

After at least 6 years of quiescence, Anomalous X-ray Pulsar (AXP) 4U 0142+61 entered an active phase in 2006 March that lasted several months and included six X-ray bursts as well as many changes in the persistent X-ray emission. The bursts, the first seen from this AXP in >11 years of Rossi X-ray Timing Explorer monitoring, all occurred in the interval between 2006 April 6 and 2007 February 7. The burst durations ranged from 0.4−1.8×10 3 s. The first five burst spectra are well modeled by blackbodies, with temperatures kT ∼ 2 − 9 keV. However, the sixth burst had a complicated spectrum that is well characterized by a blackbody plus two emission features whose amplitude varied throughout the burst. The most prominent feature was at 14.0 keV. Upon entry into the active phase the pulsar showed a significant change in pulse morphology and a likely timing glitch. The glitch had a total frequency jump of (1.9±0.4)×10 −7 Hz, which recovered with a decay time of 17±2 days by more than the initial jump, implying a net spin-down of the pulsar. Within the framework of the magnetar model, the net spin-down of the star could be explained by regions of the superfluid that rotate slower than the rest. The bursts, flux enhancements, and pulse morphology changes can be explained as arising from crustal deformations due to stresses imposed by the highly twisted internal magnetic field. However, unlike other AXP outbursts, we cannot account for a major twist being implanted in the magnetosphere.