Studies of the Relativistic Binary Pulsar PSR B1534+12. II. Origin and Evolution

Abstract

We have recently measured the angle between the spin and orbital angular momenta of PSR B1534+12 to be either 25+/-4 deg or 155+/-4 deg. This misalignment was almost certainly caused by an asymmetry in the supernova explosion that formed its companion neutron star. Here we combine the misalignment measurement with measurements of the pulsar and companion masses, the orbital elements, proper motion, and interstellar scintillation. We show that the orbit of the binary in the Galaxy is inconsistent with a velocity kick large enough to produce a nearly antialigned spin axis, so the true misalignment must be ~25 deg. Similar arguments lead to bounds on the mass of the companion star immediately before its supernova: 3+/-1 Msun. The result is a coherent scenario for the formation of the observed binary. After the first supernova explosion, the neutron star that would eventually become the observed pulsar was in a Be/X-ray type binary system with a companion of at least 10--12 Msun. During hydrogen (or possibly helium) shell burning, mass transfer occurred in a common envelope phase, leaving the neutron star in a roughly half-day orbit with a helium star with mass above ~3.3 Msun. A second phase of mass transfer was then initiated by Roche lobe overflow during shell helium burning, further reducing both the helium star mass and orbital period before the second supernova. Scenarios that avoid Roche lobe overflow by the helium star require larger helium star masses and predict space velocities inconsistent with our measurements. The companion neutron star experienced a velocity kick of 230+/-60 km/s at birth, leading to a systemic kick to the binary of 180+/-60 km/s.