Magnetic field decay of magnetars in supernova remnants

In this paper, we modify our previous research carefully, and derive a new expression of electron energy density in superhigh magnetic fields. Based on our improved model, we re-compute the electron capture rates and the magnetic fields' evolutionary timescales $t$ of magnetars. According to the calculated results, the superhigh magnetic fields may evolve on timescales $\sim (10^{6}-10^{7})$ yrs for common magnetars, and the maximum timescale of the field decay, $t\approx 2.9507 \times 10^{6}$ yrs, corresponding to an initial internal magnetic field $B_{\rm 0}= 3.0 \times 10^ {15}$ G and an initial inner temperature $T_{\rm 0}= 2.6 \times 10^ {8}$ K. Motivated by the results of the neutron star-supernova remnant(SNR) association of Zhang $\&$ Xie(2011), we calculate the maximum $B_{\rm 0}$ of magnetar progenitors, $B_{\rm max}\sim (2.0\times 10^{14}-2.93 \times 10^{15})$ G when $T_{\rm 0}= 2.6 \times 10^ {8}$ K. When $T_{\rm 0}\sim 2.75 \times 10^ {8}-~1.75 \times 10^ {8}$ K, the maximum $B_{\rm 0}$ will also be in the range of $\sim 10^{14}-10^{15}$ G, not exceeding the upper limit of magnetic field of a magnetar under our magnetar model. We also investigate the relationship between the spin-down ages of magnetars and the ages of their SNRs, and explain why all AXPs associated with SNRs look older than their real ages, whereas all SGRs associated with SNRs appear younger than they are.