Title: Braking indices of young radio pulsars: theoretical perspective

Abstract: Recently, Parthsarathy et al. analysed long-term timing observations of 85 young radio pulsars. They found that 11 objects have braking indices ranging $\sim 10-100$, far from the classical value $n=3$. They also noted a mild correlation between measured value of $n$ and characteristic age of a radio pulsar. In this article we systematically analyse possible physical origin of large braking indices. We find that a small fraction of these measurements could be caused by gravitational acceleration from an unseen ultra-wide companion of a pulsar or by precession. Remaining braking indices cannot be explained neither by pulsar obliquity angle evolution, nor by complex high-order multipole structure of the poloidal magnetic field. The most plausible explanation is a decay of the poloidal dipole magnetic field which operates on a time scale $\sim 10^4-10^5$ years in some young objects, but has significantly longer time scale in other radio pulsars. This decay can explain both amplitude of measured $n$ and some correlation between $n$ and characteristic age. The decay can be caused by either enhanced crystal impurities in the crust of some isolated radio pulsars, or more likely, by enhanced resistivity related to electron scattering off phonons due to slow cooling of low-mass neutron stars. If this effect is indeed the main cause of the rapid magnetic field decay manifesting as large braking indices, we predict that pulsars with large braking indices are hotter in comparison to those with $n\approx 3$.