We observed eight pulsars with high spin down flux and searched for pulsation in the X-ray band (0.6-10 keV) with ASCA. We detected X-ray emission from all eight pulsars, including new detection in the 2-10 keV band for four pulsars. We also found pulsed emission from three pulsars including first detection of millisecond pulsar PSR B1821-24. The pulse of PSR B1821-24 has sharp double peaks similar to that of the Crab pulsar and power-law spectrum which strongly suggest the magnetospheric origin.
We found an empirical relation between the pulsed luminosity LX(pulse) and the spin down luminosity Lspin in the X-ray band among all X-ray pulsars observed with ASCA. The pulsed luminosity is consistent with the relation LX(pulse) ∝ L 3/2 spin, while those for the pulsed emission in the γ-ray band Lγ(pulse) is Lγ(pulse) ∝ L½spin. It should be noted that the empirical relation in the X-ray band holds also for the millisecond pulsar PSR B1821-24, which suggests same pulse emission mechanism works for millisecond pulsars.
On these results, we speculated the pulse emission mechanism. From the fact that the millisecond pulsar showed similar pulse profile to that of the high field cousins, the X-ray pulse emission region is likely to be located apart from the neutron star surface. On the hypothesis that the emission mechanism is not related to the central object, but only related to the structure of the magnetosphere generated by the synchrotron emission and 2) it is from the region which relative position and relative size to the light cylinder is same for all pulsars and 3) the emitting electron has same energy spectrum for all pulsars. In these assumptions, the relation LX(pulse) ∝ L3/2spin is naturally derived.
We also found the empirical relation that the phase averaged luminosity LX(average) ∝ L3/2spin as mentioned in Kawai et al. (1997) for the extended nebulae, while we analyzed only for compact regions around pulsars.