We report on robust measurements of elemental abundances of the Type IIn supernova SN 1978K, based on the high-resolution X-ray spectrum obtained with the Reflection Grating Spectrometer (RGS) onboard XMM-Newton. The RGS clearly resolves a number of emission lines, including N Ly$alpha$, O Ly$alpha$, O Ly$beta$, Fe XVII, Fe XVIII, Ne He$alpha$ and Ne Ly$alpha$ for the first time from SN 1978K. The X-ray spectrum can be represented by an absorbed, two-temperature thermal emission model, with temperatures of $kT sim 0.6$ keV and $2.7$ keV. The elemental abundances are obtained to be N $=$ $2.36_{-0.80}^{+0.88}$, O $=$ $0.20 pm{0.05}$, Ne $=$ $0.47 pm{0.12}$, Fe $=$ $0.15_{-0.02}^{+0.01}$ times the solar values. The low metal abundances except for N show that the X-ray emitting plasma originates from the circumstellar medium blown by the progenitor star. The abundances of N and O are far from CNO-equilibrium abundances expected for the surface composition of a luminous blue variable, and resemble the H-rich envelope of less-massive stars with masses of 10-25 M$_odot$. Together with other peculiar properties of SN 1978K, i.e., a low expansion velocity of 500-1000 km s$^{-1}$ and SN IIn-like optical spectra, we propose that SN 1978K is a result of either an electron-capture SN from a super asymptotic giant branch star, or a weak Fe core-collapse explosion of a relatively low-mass ($sim$10 M$_odot$) or high-mass ($sim$20-25 M$_odot$) red supergiant star. However, these scenarios can not naturally explain the high mass-loss rate of the order of $dot{M} sim 10^{-3} rm{M_{odot} yr^{-1}}$ over $gtrsim$1000 yr before the explosion, which is inferred by this work as well as many other earlier studies. Further theoretical studies are required to explain the high mass-loss rates at the final evolutionary stages of massive stars.
Download