🔭 astrophysics

SN2023ixf: ultraviolet-to-infrared radiative-transfer modeling of the nebular-phase evolution until 1000 days

This paper presents non-local thermodynamic equilibrium radiative-transfer modeling of the supernova SN2023ixf from the ultraviolet to the infrared up to 1000 days, revealing that its nebular-phase evolution is driven by a partially stripped red-supergiant progenitor interacting with a massive circumstellar shell, with its late-time light curve and spectral features shaped by persistent interaction, enhanced gamma-ray escape, and significant dust formation in both the shell and inner ejecta.

Original authors: Luc Dessart, Wynn V. Jacobson-Galan, K. Azalee Bostroem, Alexei V. Filippenko, WeiKang Zheng, Thomas G. Brink, Lluis Galbany, Claudia Gutierrez, Stefano Valenti

Published 2026-05-15

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Luc Dessart, Wynn V. Jacobson-Galan, K. Azalee Bostroem, Alexei V. Filippenko, WeiKang Zheng, Thomas G. Brink, Lluis Galbany, Claudia Gutierrez, Stefano Valenti

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a massive star, about 15 times heavier than our Sun, living out its final days. Instead of a quiet fade, it explodes in a spectacular supernova known as SN 2023ixf. This paper is like a detailed detective story, tracking the "ghost" of this explosion as it evolves over nearly three years (from day 112 to day 1000 after the blast).

The scientists, led by Luc Dessart, used powerful computer models to simulate what was happening inside this cosmic fireball. Here is the story of what they found, explained simply.

1. The Setup: A Star with a "Speed Bump"

When the star exploded, it didn't just shoot debris into empty space. It had been shedding material like a heavy coat before it died. This created a dense, cold shell of gas (the CDS) surrounding the explosion, moving at about 8,000 kilometers per second.

Think of the explosion as a race car speeding down a highway. The CDS is like a thick fog bank the car is driving into. The car (the explosion debris) slams into this fog, creating a shockwave that lights up the gas, much like headlights cutting through fog.

2. The Mystery of the Fading Light

For the first few months, the light from the explosion was powered by radioactive decay (like a glowing battery inside the debris). But around day 200, something strange happened: the supernova started fading much faster than the "battery" should have allowed.

The Culprit: Dust.
The scientists realized that dust was forming inside the explosion. Imagine a thick, dark curtain being woven inside the fireball. This dust acts like a pair of sunglasses, blocking the visible light and making the supernova look dimmer.

Where did it form? It started in the outer "fog" (the CDS) and later formed in the inner debris too.
What is it made of? A mix of carbon-rich and silicon-rich grains (like soot and sand).
The Effect: By day 700, enough dust had formed to block a significant amount of light, explaining why the supernova looked so faint in visible light.

3. The "Ghost" in the Ultraviolet

While the dust was dimming the visible light, the ultraviolet (UV) light was behaving differently. It was like a spotlight shining through a hole in the curtain.

The UV light wasn't being blocked by the dust because the dust was "behind" the UV light's source. Instead, the UV light was being shaped by the shockwave hitting the outer fog.

The Lyman-alpha (Ly α) Line: This is a specific color of UV light. The models showed that the way the shockwave hit the outer fog determined how bright this line was.
The X-ray Twist: The scientists found that if they assumed the shockwave acted like X-rays hitting the gas (rather than just heating electrons), the model matched the real observations much better. It's as if the "fog" was being ionized (charged up) by invisible X-rays, changing how the UV light escaped.

4. The Great Deceleration

One of the most interesting findings was about speed.

The Start: At day 112, the outer shell (CDS) was racing at 8,000 km/s.
The End: By day 1000, it had slowed down to about 6,500 km/s.

The Analogy: Imagine a runner sprinting into a crowd. As they push through more and more people, they slow down. The fact that the shell slowed down so much suggests it was "sweeping up" more mass than originally thought—like a snowplow gathering a massive pile of snow. The shell likely grew from 0.2 solar masses to over 0.5 solar masses by the end of the observation period.

5. A Cosmic Twin: SN 2023ixf vs. SN 1993J

The paper compares SN 2023ixf to a famous older supernova, SN 1993J.

Similarities: After about a year, both looked very similar. They both had broad, boxy lines of light, suggesting they were both powered by the same "fog" interaction.
Differences: SN 2023ixf was actually fainter and cooler. Why? Because SN 2023ixf had a much thicker layer of dust and a heavier outer shell, which muffled the light and kept the temperature down. SN 1993J, having less dust, stayed brighter and hotter.

Summary of the Detective Work

The scientists built a complex computer model (a "virtual supernova") to test their theories. They found that:

Dust is the main reason the supernova got dim in visible light after day 200.
Interaction with the outer shell is what keeps the UV light bright and shapes the spectrum.
The outer shell is slowing down, proving it is gathering more mass as it expands.

In short, SN 2023ixf is a cosmic drama where a star explodes, hits a wall of its own making, creates a thick layer of cosmic dust that hides its visible face, but still shines brightly in ultraviolet light, all while slowly grinding to a halt as it sweeps up the universe around it.

Problem
SN 2023ixf is a nearby Type II supernova (SN) originating from a red supergiant (RSG) progenitor, characterized by early and persistent interaction with circumstellar material (CSM). While previous studies have constrained its photospheric evolution and progenitor properties, there remains significant disparity in inferred progenitor masses (ranging from 11 to 20 $M_\odot$ ) and explosion parameters due to fundamental degeneracies in light-curve modeling. Furthermore, the transition from a decay-powered nebular phase to an interaction-dominated phase, coupled with the onset of dust formation, presents complex challenges for radiative transfer modeling. Existing archival nebular models often relied on standard RSG progenitors and neglected the influence of interaction power or dust, failing to explain the specific spectral evolution and flux attenuation observed in SN 2023ixf out to 1000 days.

Methodology
The authors present non-local thermodynamic equilibrium (non-LTE) radiative-transfer modeling of SN 2023ixf from 112 to 998 days post-explosion using the code CMFGEN. The study utilizes the ejecta model x6p0, previously established by Dessart et al. (2026) to match the photospheric phase, which corresponds to a partially stripped 15 $M_\odot$ progenitor yielding 7–8 $M_\odot$ of ejecta, 1.2 $\times$ 10 $^{51}$ erg kinetic energy, and 0.05 $M_\odot$ of $^{56}$ Ni. This ejecta is augmented with a cold dense shell (CDS) of 0.2 $M_\odot$ at 8000 km s $^{-1}$ .

Key methodological components include:

Interaction Power Treatment: Two approaches were tested: injecting high-energy electrons (shock power) directly into the CDS, and injecting X-rays throughout the outer ejecta and CDS regions.
Dust Modeling: Dust formation was introduced starting at $\sim$ 200 days. The models account for dust in the inner ejecta (below 3000 km s $^{-1}$ ) and within the CDS, using a mixture of Si-rich and C-rich grains (1:4.5 ratio) with 0.1 $\mu$ m radii. An external, cool dust component was added post-processing to match mid-infrared (MIR) emission.
Observational Constraints: The models were compared against a comprehensive dataset spanning UV (HST), optical, near-infrared (NIR), and MIR (JWST) from 112 to 998 days. Spectra were rebinned to a fixed resolution of 300 km s $^{-1}$ (or degraded to $\sim$ 3000 km s $^{-1}$ for MIR comparisons).

Key Contributions and Results

Nebular Evolution and Interaction: The study confirms that interaction with CSM persists throughout the nebular phase, powering the UV flux at all times and dominating the optical flux after $\sim$ 600 days. The CDS velocity decelerates from 8000 km s $^{-1}$ at 112 days to $\sim$ 6500 km s $^{-1}$ at 998 days, suggesting a mass accumulation of several 0.1 $M_\odot$ .
Dust Formation and Attenuation: To match the observed V-band light curve and spectral energy distribution, dust formation is required starting at $\sim$ $\sim$ 200 days.
- Location: Dust forms first in the CDS and subsequently in the inner ejecta.
- Mass Evolution: The inferred dust mass in the inner ejecta rises from $1 \times 10^{-6} M_\odot$ at 208 days to $100 \times 10^{-6} M_\odot$ at 620 days. CDS dust mass follows a similar trend, reaching $100 \times 10^{-6} M_\odot$ at 620 days.
- External Component: An external, optically thin dust component (mass $\sim$ 3 $\times$ 10 $^{-4} M_\odot$ ) is required to reproduce the MIR emission, as internal dust alone is insufficient.
Spectral Characteristics:
- UV: The UV flux is largely unaffected by dust but is strongly influenced by Fe line blanketing and the structure of the CDS. The strength of Ly $\alpha$ and Mg ii $\lambda\lambda$ 2795, 2802 is highly sensitive to the treatment of shock power; injecting power as X-rays enhances Ly $\alpha$ and suppresses Mg ii relative to electron injection, better matching observations.
- Optical/IR: The models reproduce the broad, boxy H $\alpha$ and He i 1.083 $\mu$ m profiles characteristic of CDS interaction. However, the models struggle to match the exact width and strength of He i 1.083 $\mu$ m and overestimate certain IR metal lines (e.g., [Ne ii] 12.810 $\mu$ m), potentially indicating differences in clumping or metal yields not captured by the 1D shuffled-shell structure.
Comparison with SN 1993J: Despite different progenitor types (Type II vs. Type IIb), SN 2023ixf shares spectral similarities with SN 1993J at 1–3 years, particularly the dominance of broad, boxy lines. However, SN 2023ixf is fainter and cooler, lacking strong [O iii] and [N ii] lines, likely due to greater CDS and ejecta masses and significant dust extinction.

Significance
The paper demonstrates that a consistent radiative-transfer model, incorporating a partially stripped progenitor, CSM interaction, and evolving dust formation, can successfully reproduce the multi-wavelength evolution of SN 2023ixf from the photospheric to the late nebular phase. The study highlights that:

Dust is critical for explaining the rapid fading of the optical light curve after 200 days, not just $\gamma$ -ray escape.
Interaction power dominates the UV and eventually the optical output, making the SN appear as a hybrid of decay and interaction-powered emission.
Model sensitivity: The treatment of shock power (electrons vs. X-rays) significantly alters the predicted UV spectrum, particularly the ratio of Ly $\alpha$ to Mg ii, emphasizing the need for high-resolution UV observations to constrain CDS structure and power injection mechanisms.
Progenitor constraints: The results support a progenitor mass of $\sim$ 15 $M_\odot$ with a partially stripped envelope, resolving some of the degeneracies found in previous studies that ignored interaction or dust.

The authors conclude that while their 1D models provide a satisfactory match to observations, discrepancies in line widths and specific metal line strengths suggest that future work must incorporate multidimensional hydrodynamics to better resolve the 3D structure of the CDS, clumping, and the dynamical evolution of the interaction region.