Item

Abstract

Context. Tracing unstable isotopes produced in supernova nucleosynthesis provides a direct diagnostic of supernova explosion physics. Theoretical models predict an extensive variety of scenarios, which can be constrained through observations of the abundant isotopes ⁵⁶Ni and ⁴⁴Ti. Direct evidence of the latter was previously found only in two core-collapse supernova events, and appears to be absent in thermonuclear supernovae.
Aims. We aim to to constrain the supernova progenitor types of Cassiopeia A, SN 1987A, Vela Jr., G1.9+0.3, SN1572, and SN1604 through their ⁴⁴Ti ejecta masses and explosion kinematics.
Methods. We analyzed INTEGRAL/SPI observations of the candidate sources utilizing an empirically motivated high-precision background model. We analyzed the three dominant spectroscopically resolved de-excitation lines at 68, 78, and 1157 keV emitted in the decay chain of ⁴⁴Ti→⁴⁴Sc→⁴⁴Ca. The fluxes allow the determination of the production yields of ⁴⁴Ti. Remnant kinematics were obtained from the Doppler characteristics of the lines.
Results. We find a significant signal for Cassiopeia A in all three lines with a combined significance of 5.4σ. The fluxes are (3.3 ± 0.9) × 10⁻⁵ ph cm⁻² s⁻¹, and (4.2 ± 1.0) × 10⁻⁵ ph cm⁻² s⁻¹ for the ⁴⁴Ti and ⁴⁴Sc decay, respectively. This corresponds to a mass of (2.4 ± 0.7) × 10⁻⁴ M_⊙ and (3.1 ± 0.8) × 10⁻⁴ M_⊙, respectively. We obtain higher fluxes for ⁴⁴Ti with our analysis of Cassiopeia A than were obtained in previous analyses. We discuss potential differences. We interpret the line width from Doppler broadening as expansion velocity of (6400 ± 1900) km s⁻¹. We do not find any significant signal for any other candidate sources.
Conclusions. We obtain a high ⁴⁴Ti ejecta mass for Cassiopeia A that is in disagreement with ejecta yields from symmetric 2D models. Upper limits for the other core-collapse supernovae are in agreement with model predictions and previous studies. The upper limits we find for the three thermonuclear supernovae (G1.9+0.3, SN1572 and SN1604) consistently exclude the double detonation and pure helium deflagration models as progenitors.