Revealing physical activity of GRB central engine with macronova/kilonova data

The modeling of Li-Paczyński macronova/kilonova signals gives reasonable estimate on the neutron-rich material ejected during the neutron star mergers. Usually the accretion disk is more massive than the macronova ejecta, with which the efficiencies of converting the disk mass into prompt emission of three merger-driven GRBs can hence be directly constrained. Supposing the macronovae/kilonovae associated with GRB 050709, GRB 060614 and GRB 130603B arose from radioactive decay of the r-process material, the upper limit on energy conversion efficiencies are found to be as low as $\sim 10^{-6}-10^{-4}$. Moreover, for all three events, neutrino annihilation is likely powerful enough to account for the brief gamma-ray flashes. Neutrino annihilation can also explain the "extended" emission lasting $\sim 100$ s in GRB 050709, but does not work for the one in GRB 060614. These progresses demonstrate that the macronova can serve as a novel probe of the central engine activity.