A Panchromatic View of Late-time Shock Power in the Type II Supernova 2023ixf

TL;DR

The paper presents a detailed, pan-chromatic study of SN 2023ixf up to +620 days, demonstrating that late-time UV-to-IR luminosity is dominated by shock-powered emission from ongoing ejecta-CSM interaction rather than radioactive decay alone. By combining multi-wavelength photometry and spectroscopy with X-ray data, the authors quantify the shock contribution as $L_{\rm sh} \sim (1-3)\times10^{40}$ erg s$^{-1}$ and constrain the CSM to a wind-like profile with $\dot M \approx 2\times10^{-4}$ $M_\odot$ yr$^{-1}$ and $v_w \approx 20-25$ km s$^{-1}$, while invoking CMFGEN radiative-transfer models to reproduce the observed SEDs with $\sim 10^{-3}$ $M_\odot$ of $1$ µm silicate dust in the CDS/inner ejecta. The results indicate nearly complete thermalization of reverse-shock energy into reprocessed emission and reveal dust formation in the late-time CDS/ejecta, enhancing our understanding of mass-loss, dust production, and shock physics in nearby Type II SNe with strong CSM interaction.

Abstract

We present multi-wavelength observations of the type II supernova (SN II) 2023ixf during its first two years of evolution. We combine ground-based optical/NIR spectroscopy with Hubble Space Telescope (HST) far- and near-ultraviolet spectroscopy and James Webb Space Telescope (JWST) near- and mid-infrared photometry and spectroscopy to create spectral energy distributions of SN 2023ixf at +374 and +620 days post-explosion, covering a wavelength range of ~0.1-30 $μ$m. The multi-band light curve of SN 2023ixf follows a standard radioactive decay decline rate after the plateau until ~500 days, at which point shock powered emission from ongoing interaction between the SN ejecta and circumstellar material (CSM) begins to dominate. This evolution is temporally consistent with 0.3-10 keV X-ray detections of SN 2023ixf and broad ''boxy'' spectral line emission from reprocessing of shock luminosity in a cold dense shell located between forward and reverse shocks. Using the expected absorbed radioactive decay power and the detected X-ray luminosity, we quantify the total shock powered emission at the +374 and +620 day epochs and find that it can be explained by nearly complete thermalization of the reverse shock luminosity as SN 2023ixf interacts with a continuous, ''wind-like'' CSM with a progenitor mass-loss rate of $\dot M \approx 10^{-4}$ M$_{\odot}$ yr$^{-1}$ ($v_w = 20 \pm 5$ km/s). Additionally, we construct multi-epoch spectral models from the non-LTE radiative transfer code CMFGEN, which contain radioactive decay and shock powers, as well as dust absorption, scattering, and emission. We find that models with shock powers of $L_{sh} = (0.5-1) \times 10^{40}$ erg s$^{-1}$ and $(0.5 - 1) \times 10^{-3}$ M$_{\odot}$ of silicate dust in the cold dense shell and/or inner SN ejecta can effectively reproduce the global properties of the late-time (>300 days) UV-to-IR spectra of SN 2023ixf.

Figures (8)

• Figure 1: Late-time spectral energy distribution of SN 2023ixf covering radio, infrared, optical, UV and X-ray wavelengths at $\delta t \approx 165 - 585$ days. Emission mechanisms and power sources are labeled at the top of the figure.
• Figure 2: Multi-band light curve of SN 2023ixf extending to $\delta t \approx 700$ days and covering UV (polygons), optical (circles), NIR (squares) and MIR (stars) wavelengths. Radioactive decay power model decline rates shown as dashed (complete $\gamma$-ray trapping) and solid (incomplete $\gamma$-ray trapping) black lines. Decline rate for free-free X-ray emission from CSM-interaction shown as dotted line. The dominance of shock powered emission from CSM-interaction over absorbed radioactive decay power is observed as excess emission in UV filters at $> 200$ days and optical/NIR filters at $> 500$ days.
• Figure 3: Time-series optical/NIR (black/red) spectra of SN 2023ixf spanning $\delta t = 272 - 708$ days. Ion identifications shown in blue are based on simulations from Dessart23b. Spectra at $\delta t > 619$ days are dominated by broad, "boxy" line profiles derived from emission in the dense shell and inner ejecta as a radiative reverse shock is powered by ongoing CSM-interaction.
• Figure 4: Left: Unabsorbed 0.3-10 keV light curve of SN 2023ixf (cyan circles from method 1 and cyan squares from method 2; §\ref{['SubSec:Xray']}) compared to SNe 1993J (red polygons) and 1998S (yellow plus signs). Black solid line is the multi-shock model for SN 1993J from Fransson96 where rising flux at $>100$ days is from the emerging RS. Black dotted and dashed lines represent the analytic predictions for free-free emission (e.g., Eqn. \ref{['eqn:Lff']}) for the forward shock luminosity assuming $\dot M = 10^{-4}$ M$_{\odot}$ yr$^{-1}$ ($v_w = 20~\textrm{km}\,\textrm{s}^{-1}$). Right: CSM density measurements from Nayana25 (blue circles) and modeling of late-time X-ray spectra (red stars). The derived CSM density profile continues to trace a steady-state mass loss rate of $\dot M = 10^{-4}$ M$_{\odot}$ yr$^{-1}$ (solid black line). Dashed grey lines represent look-back times assuming a shock velocity of $10^{4}~\textrm{km}\,\textrm{s}^{-1}$ and CSM velocity of $20~\textrm{km}\,\textrm{s}^{-1}$.
• Figure 5: Left: Pseudobolometric light curves of SN 2023ixf constructed using NUV-NIR bands ($0.16 - 2.5~\mu$m, salmon circles), NUV ($0.16 - 0.3~\mu$m, magenta circles), FUV-NUV ($0.1 - 0.3~\mu$m, magenta stars), optical ($0.3 - 1.0~\mu$m, blue circles), NIR ($1.0 - 2.5~\mu$m, orange circles), and MIR ($2.5 - 30.0~\mu$m, orange circles). Unabsorbed 0.3-10 keV X-ray light curve shown as cyan circles. Right: Fractional flux emitted in UV ($0.1 - 0.3~\mu$m, magenta), optical ($0.3 - 1.0~\mu$m, blue), IR ($1 - 30~\mu$m, orange), and X-ray ($0.3 - 10$ keV, cyan) wavelengths at $\delta t = 374$ days.