Astron. Astrophys. 330, 175-180 (1998)

4. Discussion

The region of sky containing the 1E 2259+586 pulsar and the G109.1-1.0 SNR is complex. The 1E 2259+586 spectrum is best described by the sum of a power-law and a blackbody, confirming the results of Corbet et al. (1995). These authors suggest that at least part of the power-law component could originate from a synchrotron nebula which may be visible around 1E 2259+586 in ROSAT images (Rho & Petre 1997). A similar spectral decomposition has also been reported for two other "anomalous pulsars", 4U 0142+61 and 1E 1048.1-5937 (White et al. 1996; Corbet & Mihara 1997) and has been interpreted as evidence for quasi-spherical accretion onto isolated neutron stars after common envelope evolution and spiral-in (Ghosh et al. 1997). In this case the accretion flow has two components. A low-angular momentum component giving rise to the blackbody emission from a large fraction of the neutron star surface, and a high-angular momentum one forming an accretion disk responsible for the power-law emission.

We caution that an alternative explanation for the observed two-component spectrum cannot be excluded for 1E 2259+586. If a plausible fraction of the emission arises from the part of the SNR within the pulsar's extraction region, then an acceptable fit can be obtained with a simple power-law model. This implies that the derived blackbody radius for 1E 2259+586 should be regarded as an upper limit. We note that with the power-law fit, the pulsar's N of  atoms cm^-2 is significantly greater than obtained for G109.1-1.0 of atoms cm^-2. This may indicate that the pulsar lies some distance beyond the SNR and that the two objects are unrelated, or it may indicate the presence of absorbing material local to the pulsar. The N to the SNR is consistent with the galactic column in the direction of 1E 2259+586 of atoms cm^-2 (Dickey & Lockman 1990).

The derived pulse period is consistent with the extrapolation of the long-term spin-down measured by Corbet et al. (1995) and Baykal & Swank (1996) as can be seen in Fig. 5. No large change in spin-down rate, as observed for example from 4U 1626-67 (Chakrabarty et al. 1997), is evident. This implies that the accretion torque has remained approximately constant over at least the last 19 years.

Fig. 5. Pulse period history of 1E 2259+586. See Baykal & Swank (1996) for the measured values, except for the BeppoSAX result reported here

In agreement with the results of Rho & Petre (1997), the LECS spectra of the SNR shell and of the jet-like X-ray lobe are indistinguishable, supporting a common origin for these two regions. We therefore refer to the summed spectrum as "the SNR". For an assumed distance of 4 kpc, the radius of the SNR shell corresponds to 17.5 pc, or an emitting volume of cm³, where f is the filling factor of the emitting plasma within a sphere. The mass of the SNR is then  M . The canonical filling factor for a strong shock is 0.25, although the absence of half of the remnant reduces this by a factor of 2. The clumpiness of the X-ray emission (with most of the emission coming from a few bright spots) implies an even smaller filling factor, and we adopt a value of 0.1. The total emitting remnant mass is therefore 15-20 M . The remnant age is estimated to be 13,000 yr using the hydrodynamical model of Wang et al. (1992). However this estimate uses a remnant plasma temperature of 0.4 keV, significantly lower than determined here. The higher temperature implies a faster expansion speed (900 km s^-1 rather than 590 km s^-1), and thus that the remnant is younger than previously estimated. The Wang et al. (1992) estimate is based on detailed numerical simulations, and it is not straightforward to determine the age corresponding to the updated temperature. Similarly, Hughes et al. (1981) determine an age of 17,000 yr assuming an X-ray temperature of 0.17 keV. Given their analytical approach, is possible to scale their estimated age to the current temperature determination to give a value of 3000 yr.

The low value derived for the emitting mass is consistent with a young remnant, which has not yet swept up large quantities of circumstellar material. The abundances derived from the X-ray spectrum are slightly higher than cosmic values, indicative of circumstellar material which has been mildly enriched either by the stellar wind of the progenitor in its late evolutionary stages, or by some moderate mixing-in of ejecta material. The value of the ionization parameter ( cm^-3 yr) is indicative of strong non-equilibrium conditions, with an implied ionization age of 3000 yr, in good agreement with the scaled numerical simulations of Hughes et al. (1981). The lack of strongly enriched material implies that the ejecta is not being directly observed, and is also consistent with the lack of a strong wind of enriched gas originating from the progenitor in its late stages. This is suggestive of a low-mass progenitor and a Type Ib supernova.

The LECS fit results to the whole of the SNR are similar to those of Rho & Petre (1997), who analyzed BBXRT and ROSAT spectra of a small part of the remnant, just south of the pulsar. Their fit with a NEI model in non-equipartition also shows strong non-equilibrium conditions, with a comparable (considered the differences in the instrumental responses and plasma emission models) age of 6700 yr, near-cosmic or slightly enriched abundances and an emitting mass of 85 M .

© European Southern Observatory (ESO) 1998

Online publication: January 8, 1998
helpdesk.link@springer.de