Ultra-Stripped Supernovae

• Ultra-stripped supernovae are core-collapse explosions of nearly bare helium or carbon–oxygen cores in close binaries, marked by very low ejecta masses and modest explosion energies.
• They exhibit rapid radiative diffusion that yields fast-evolving, low-luminosity light curves and distinctive spectral features, such as prominent Mg II and Ca II lines.
• These events are key to understanding binary stripping processes and double neutron star formation, offering practical insights into explosion dynamics in compact binaries.

Ultra-stripped supernovae are core-collapse explosions of helium-star or carbon–oxygen star remnants in close binaries whose envelopes have been almost completely removed by interaction with a compact companion. Their defining physical regime is very small ejecta mass, modest explosion energy, and rapid radiative diffusion, which together produce fast, faint stripped-envelope transients. They are of particular interest because the same evolutionary channel is a major route to the formation of double neutron star systems, and because the observed phenomenology spans fast Type Ib/c events, shock-cooling transients, and, in CSM-dominated cases, even Type Ibn supernovae (Tauris et al., 2015, Suwa et al., 2015, Moriya et al., 7 Jul 2025).

1. Binary stripping and the ultra-stripped regime

The canonical ultra-stripped channel begins after the first supernova in a massive binary has already produced a neutron star. The companion is then reduced to a naked helium star, commonly after common-envelope evolution, and later undergoes a final Roche-lobe overflow episode—usually described as Case BB, and in some models Case BC—that removes most of the remaining helium envelope. The result is a “more or less naked” metal core with only a very small residual envelope at collapse, typically in the range $0.01$–$0.2\,M_\odot$, and in extreme systems below $0.01\,M_\odot$ (Tauris, 2015, Tauris et al., 2015).

Systematic progenitor studies identify helium-star donor masses around $2.5$–$3.5\,M_\odot$ and tight pre-explosion orbital periods as the most favorable regime for extreme stripping. In this parameter space, the final outcome can be an ONeMg white dwarf, an electron-capture supernova, or an iron core-collapse supernova, depending on the final core mass. One set of models places electron-capture explosions in a narrow metal-core interval of $M_{\rm core}\approx1.37$–$1.43\,M_\odot$, with iron core collapse above $1.44\,M_\odot$ (Tauris et al., 2015). Other work on bare carbon–oxygen cores in the $1.45$–$2.0\,M_\odot$ range finds that all such models form iron cores and collapse as Fe-core supernovae rather than electron-capture events (Suwa et al., 2015).

This binary channel is not only a transient progenitor model but also a dynamical explanation for compact remnant binaries. The high-mass X-ray binary CPD $0.2\,M_\odot$0 2176/SGR 0755$0.2\,M_\odot$1 has been interpreted as a descendant of an ultra-stripped supernova because its nearly circular orbit and modest space velocity require minimal sudden mass loss and a very small natal kick (Richardson et al., 2023). A plausible implication is that ultra-stripped explosions occupy the transition between ordinary stripped-envelope core collapse and the low-kick supernovae required to preserve tight compact-object binaries.

2. Pre-collapse structure and explosion dynamics

Detailed binary-evolution calculations show how extreme the stripping can become. In a representative near-solar-metallicity model, a $0.2\,M_\odot$2 helium star is reduced by binary mass transfer to $0.2\,M_\odot$3, retaining a helium envelope of $0.2\,M_\odot$4. A violent off-centre silicon flash shortly before collapse then ejects almost all of that remaining helium, leaving only $0.2\,M_\odot$5 of He at explosion and placing $0.2\,M_\odot$6 of helium-rich CSM at radii of order $0.2\,M_\odot$7 by core collapse (Müller et al., 2018, Maunder et al., 2023).

Explosion calculations consistently find modest energies and low ejecta masses. In self-consistent neutrino-radiation hydrodynamics of ultra-stripped progenitors, the characteristic values are $0.2\,M_\odot$8, $0.2\,M_\odot$9–$0.01\,M_\odot$0, $0.01\,M_\odot$1–$0.01\,M_\odot$2, and neutron-star gravitational masses near $0.01\,M_\odot$3–$0.01\,M_\odot$4 (Suwa et al., 2015, Müller et al., 2018). For the specific $0.01\,M_\odot$5 helium-star model followed through 2D hydrodynamics and later radiative transfer, the explosion parameters are $0.01\,M_\odot$6, $0.01\,M_\odot$7, and $0.01\,M_\odot$8, with a characteristic ejecta velocity

$0.01\,M_\odot$9

consistent with line-of-sight velocity tails up to $2.5$0 in polar directions (Maunder et al., 2023).

Multidimensional effects are not peripheral in this class. Rayleigh–Taylor instabilities develop at the base of the O/Ne/Mg/C shell and at the base of the helium envelope, and 3D calculations show more efficient mixing than 2D, with iron-group plumes penetrating farther into the outer layers (Müller et al., 2018). The same simulations find modest natal kicks—$2.5$1–$2.5$2 in 2D and $2.5$3 in 3D—and small spin-up by asymmetric accretion, supporting the broader view that ultra-stripped explosions are dynamically favorable for double-neutron-star survival (Müller et al., 2018).

3. Energetics, nucleosynthesis, and scaling relations

The defining radiative consequence of a tiny ejecta mass is a short diffusion time. Across the literature, the characteristic estimate is

$2.5$4

with $2.5$5–$2.5$6 for homologous ejecta (Maunder et al., 2023, Haynie et al., 11 Mar 2025). For $2.5$7, $2.5$8, and $2.5$9, one obtains $3.5\,M_\odot$0 days, in agreement with detailed Monte Carlo radiative-transfer calculations (Maunder et al., 2023). Earlier progenitor studies gave analytic rise times from $3.5\,M_\odot$1 hr to $3.5\,M_\odot$2 days and decay times from $3.5\,M_\odot$3 to $3.5\,M_\odot$4 days for the ultra-stripped regime (Tauris et al., 2015).

Radioactive powering is usually framed by Arnett-like reasoning, $3.5\,M_\odot$5, but self-consistent explosion models show that the $3.5\,M_\odot$6Ni budget is tightly constrained by fallback and by the small mass available above the compact remnant. In a suite of ultra-stripped carbon–oxygen core explosions, the ejecta mass increases from $3.5\,M_\odot$7 to $3.5\,M_\odot$8 as $3.5\,M_\odot$9 increases, while the ejected $M_{\rm core}\approx1.37$0Ni mass drops from $M_{\rm core}\approx1.37$1 to $M_{\rm core}\approx1.37$2 because fallback becomes increasingly important (Sawada et al., 2021). This trend makes more massive ultra-stripped progenitors dimmer and slower rather than brighter.

Nucleosynthesis is not limited to $M_{\rm core}\approx1.37$3Ni. Two-dimensional neutrino-driven explosions of $M_{\rm core}\approx1.37$4 and $M_{\rm core}\approx1.37$5 carbon–oxygen cores produce $M_{\rm core}\approx1.37$6Ni masses of $M_{\rm core}\approx1.37$7 and $M_{\rm core}\approx1.37$8, while also synthesizing light trans-iron elements from Ga to Zr in neutrino-irradiated ejecta (Yoshida et al., 2017). The total yield of these light trans-iron elements is estimated to be below $M_{\rm core}\approx1.37$9, but with large uncertainty because the abundance pattern is highly sensitive to the electron-fraction distribution $1.43\,M_\odot$0 in the neutrino-heated ejecta (Yoshida et al., 2017). The same calculations find $1.43\,M_\odot$1Ca yields of order $1.43\,M_\odot$2, implying that ultra-stripped and other light-core core-collapse supernovae may contribute to the Galactic inventory of neutron-rich intermediate nuclei (Yoshida et al., 2017).

The main controversy in the energy budget concerns the brightest fast transients. One modeling study finds that SN 2019dge can be powered solely by $1.43\,M_\odot$3Ni from a progenitor with $1.43\,M_\odot$4, but argues that iPTF14gqr cannot be explained by a $1.43\,M_\odot$5Ni-powered model because $1.43\,M_\odot$6 of ejected $1.43\,M_\odot$7Ni is difficult to synthesize in ultra-stripped progenitors (Sawada et al., 2021). In that framework, fallback accretion affects only the first few days, whereas a newborn neutron-star wind with $1.43\,M_\odot$8 and $1.43\,M_\odot$9 can reproduce the light curve of iPTF14gqr (Sawada et al., 2021). This suggests that the ultra-stripped label does not uniquely specify the luminosity source.

4. Light curves, spectra, and multidimensional radiative signatures

Synthetic light-curve calculations established the basic observational expectation early: sub-luminous, rapidly evolving stripped-envelope events with peak luminosities around $1.44\,M_\odot$0, peak magnitudes near $1.44\,M_\odot$1, and rise times of $1.44\,M_\odot$2–$1.44\,M_\odot$3 days (Moriya et al., 2016). Detailed STELLA models using a $1.44\,M_\odot$4 progenitor and ejecta masses of $1.44\,M_\odot$5–$1.44\,M_\odot$6 produce $1.44\,M_\odot$7–$1.44\,M_\odot$8, bolometric peaks around $1.44\,M_\odot$9, and spectra dominated by Si II $1.45$0, O I $1.45$1, and the Ca II infrared triplet, usually with Type Ic appearance because the optical He I lines remain weak even when $1.45$2 of helium is present (Moriya et al., 2016).

A more extreme prediction emerges from multidimensional radiative transfer. Using ARTIS on a 2D neutrino-driven model with $1.45$3, the transient peaks at $1.45$4 days with $1.45$5, $1.45$6, and $1.45$7 (Maunder et al., 2023). The standout spectral features are unusually strong Mg II lines at $1.45$8Å and $1.45$9–$2.0\,M_\odot$0Å, together with the Ca II infrared triplet at late times. The Mg II $2.0\,M_\odot$1Å line is highly sensitive to geometry: its emission varies by $2.0\,M_\odot$2 at peak, $2.0\,M_\odot$3 by $2.0\,M_\odot$4 days, and $2.0\,M_\odot$5 by $2.0\,M_\odot$6 days between different lines of sight, because iron-group plumes shield parts of the Mg-rich shell and enhance UV line blanketing (Maunder et al., 2023). In a spherically averaged model with added microscopic mixing, the Mg features disappear and the spectrum becomes dominated by Ni II and Co II (Maunder et al., 2023). A central implication is that strong Mg II can be a fingerprint of limited microscopic mixing, but its absence does not rule out an ultra-stripped explosion.

Early shock cooling can add a distinct first light-curve component. iPTF 14gqr showed a double-peaked structure whose initial blue flash was modeled as shock cooling of an extended He-rich envelope with $2.0\,M_\odot$7 and $2.0\,M_\odot$8, followed by a radioactive peak with $2.0\,M_\odot$9, $0.2\,M_\odot$00, and $0.2\,M_\odot$01 (De et al., 2018). More recent one-dimensional radiation-hydrodynamic work generalizes this into a two-phase shock-cooling picture: a first blue flash from compact He-rich CSM lasting $0.2\,M_\odot$02 a few days, and a second shock-cooling bump or plateau from the extended helium-burning envelope on $0.2\,M_\odot$03–$0.2\,M_\odot$04 day timescales (Haynie et al., 11 Mar 2025). This morphology provides two independent radial diagnostics, $0.2\,M_\odot$05 and $0.2\,M_\odot$06, when both phases are observed (Haynie et al., 11 Mar 2025).

5. Circumstellar interaction, Type Ibn outcomes, and long-wavelength diagnostics

Ultra-stripped progenitors are not necessarily “clean” stripped-envelope stars. Off-centre silicon flashes can eject substantial He-rich material shortly before collapse. In the 2.8-$0.2\,M_\odot$07 helium-star sequence, $0.2\,M_\odot$08 of helium is expelled $0.2\,M_\odot$09 days before core collapse at $0.2\,M_\odot$10, reaching $0.2\,M_\odot$11 by explosion (Müller et al., 2018). That mass-loss episode implies delayed ejecta–CSM interaction, with strong interaction expected no earlier than $0.2\,M_\odot$12 days after explosion for maximum ejecta velocities of $0.2\,M_\odot$13 (Müller et al., 2018).

If the dense CSM is sufficiently massive, it can dominate the event. Radiation-hydrodynamic calculations using a He-rich shell of $0.2\,M_\odot$14, expelled $0.2\,M_\odot$15 days before explosion and extending from $0.2\,M_\odot$16 to $0.2\,M_\odot$17, show that a canonical ultra-stripped explosion with $0.2\,M_\odot$18, $0.2\,M_\odot$19, and $0.2\,M_\odot$20 is immediately decelerated because $0.2\,M_\odot$21 (Moriya et al., 7 Jul 2025). The resulting light curve rises in $0.2\,M_\odot$22 days to $0.2\,M_\odot$23, radiates $0.2\,M_\odot$24 in the first $0.2\,M_\odot$25 days, and is UV-dominated, with only $0.2\,M_\odot$26 of the bolometric luminosity in the optical at peak (Moriya et al., 7 Jul 2025). In this regime the transient is more likely to be classified as Type Ibn than as a fast Type Ib/c, because narrow or intermediate-width He-rich interaction lines dominate the spectrum (Moriya et al., 7 Jul 2025). This directly links some Type Ibn events to ultra-stripped progenitors rather than to classical Wolf–Rayet explosions.

Radio diagnostics probe a different aspect of the same binary history. If most of the Roche-lobe overflow mass is expelled into a wind-like CSM, ejecta–CSM interaction produces radio emission whose timing and luminosity reflect the density of the lost material. Models show that bright centimeter emission at $0.2\,M_\odot$27–$0.2\,M_\odot$28 days, with $0.2\,M_\odot$29, or bright millimeter emission within $0.2\,M_\odot$30 days, implies $0.2\,M_\odot$31, a final double-neutron-star separation $0.2\,M_\odot$32, and merger within the cosmic age (Matsuoka et al., 2020). Conversely, weak radio emission together with optical evidence for $0.2\,M_\odot$33 favors a wider DNS unlikely to merge within a Hubble time (Matsuoka et al., 2020).

On longer timescales, the supernova remnant itself should be faint. Modeling of ultra-stripped supernova remnants in CSM shaped by binary mass loss predicts a hot $0.2\,M_\odot$34 bubble around the wind termination shock at $0.2\,M_\odot$35 pc, which drastically weakens the blast wave and suppresses efficient particle acceleration (Matsuoka et al., 2022). The resulting remnants have low radio luminosities and surface brightnesses compared with known Galactic SNRs, and are expected to make up only $0.2\,M_\odot$36–$0.2\,M_\odot$37 of the Galactic supernova remnant population (Matsuoka et al., 2022).

6. Observational counterparts, classification disputes, and open problems

Several observed fast transients are plausible, but not unambiguous, ultra-stripped supernovae. SN 2005ek has long been a central candidate because of its rapid decline and low ejecta mass, and early ultra-stripped Type Ic modeling explicitly proposed it as an analogue (Tauris et al., 2013). Yet the self-consistent 2D ARTIS model with $0.2\,M_\odot$38, $0.2\,M_\odot$39, and Mg-dominated spectra is even fainter and faster than SN 2005ek, and fails to reproduce the observed C II, Si II, O I, and stronger Ca II features (Maunder et al., 2023). The mismatch shows that “fast and faint” is not, by itself, a unique USSN identifier.

SN 2019wxt is a stronger phenomenological match to the helium-rich end of the class. It was a Type Ib event with $0.2\,M_\odot$40, $0.2\,M_\odot$41, $0.2\,M_\odot$42, $0.2\,M_\odot$43, and a composition dominated by He and O with trace Ca (Agudo et al., 2022). Its rapid evolution and faint peak made it a credible kilonova impostor during gravitational-wave follow-up, and its estimated rate implies that of order one such event per week can occur within a typical O4 localization volume out to $0.2\,M_\odot$44 Mpc (Agudo et al., 2022). This establishes ultra-stripped supernovae as a practical source of contamination in compact-binary counterpart searches.

SN 2023zaw has intensified the debate because different modeling assumptions lead to different ejecta masses and power sources. One study, combining shock-cooling emission and radioactive decay, infers $0.2\,M_\odot$45, $0.2\,M_\odot$46, $0.2\,M_\odot$47–$0.2\,M_\odot$48, $0.2\,M_\odot$49–$0.2\,M_\odot$50, and $0.2\,M_\odot$51–$0.2\,M_\odot$52, arguing for an ultra-stripped origin from a low-mass He star in a close binary (Das et al., 2024). A second analysis favors a much smaller ejecta mass, $0.2\,M_\odot$53, together with $0.2\,M_\odot$54 and a required additional early power source, with strong Bayesian preference for CSM interaction plus Ni over a Ni-only model and over a magnetar-plus-Ni model (Moore et al., 2024). The shared conclusion is that nickel decay alone is insufficient for the early light curve, but the inferred structure of the exploding star remains model-dependent (Moore et al., 2024). This suggests that ultra-stripped events can be both nickel-poor and power-source-diverse.

Population arguments remain similarly unsettled but broadly consistent. Event-rate estimates place iPTF14gqr-like USSNe at roughly $0.2\,M_\odot$55–$0.2\,M_\odot$56 of core-collapse supernovae, with survey-specific yields ranging from $0.2\,M_\odot$57 for iPTF to $0.2\,M_\odot$58 for ZTF and $0.2\,M_\odot$59 for LSST deep-drilling fields under specific cadence assumptions (Hijikawa et al., 2019). Broader synthesis papers emphasize that comparing future DNS merger rates from gravitational-wave detectors with USSN rates from optical surveys will directly test whether ultra-stripped explosions are the dominant contributor to the DNS population (Moriya et al., 2016, Tauris, 2015).

The principal theoretical limitations are also clear. Current predictions depend sensitively on dimensionality, microscopic versus macroscopic mixing, atomic data, $0.2\,M_\odot$60 evolution in neutrino-heated ejecta, approximate NLTE treatments, and whether CSM interaction, shock cooling, fallback accretion, or a newborn neutron-star wind is included (Maunder et al., 2023, Yoshida et al., 2017, Sawada et al., 2021). A recurring result across these studies is that multi-D radiative transfer and broader progenitor suites are essential: geometry can modulate Mg II by factors of order unity, strong Fe-group blanketing can conceal diagnostic lines, and a He-rich dense shell can transform a canonical fast Type Ib/c into a Type Ibn (Maunder et al., 2023, Moriya et al., 7 Jul 2025). Ultra-stripped supernovae are therefore best understood not as a single light-curve template, but as a compact binary explosion regime whose observables are controlled by residual envelope mass, nickel yield, mixing, and the immediate circumstellar environment.