🔭 astrophysics

Helium features are inconsistent with the spectral evolution of the kilonova AT2017gfo

This paper demonstrates that the 1 $\mu$ m P Cygni feature observed in the kilonova AT2017gfo is inconsistent with a helium interpretation, as self-consistent helium models fail to reproduce the feature's early emergence and subsequent spectral evolution, thereby supporting its identification as Strontium II.

Original authors: Albert Sneppen, Rasmus Damgaard, Darach Watson, Christine E. Collins, Luke Shingles, Stuart A. Sim

Published 2026-02-24

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Albert Sneppen, Rasmus Damgaard, Darach Watson, Christine E. Collins, Luke Shingles, Stuart A. Sim

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

The Cosmic Mystery: What is that "1-Micron" Line?

Imagine two neutron stars (the super-dense corpses of dead stars) crashing into each other. This collision, which happened in 2017 and was named AT2017gfo, created a spectacular explosion called a "kilonova." It was the universe's way of forging heavy elements like gold and platinum.

When astronomers looked at the light from this explosion, they saw a specific "fingerprint" in the spectrum (a rainbow of light) around a wavelength of 1 micron (a deep red/infrared color). This fingerprint is a P Cygni line—a specific shape in the light curve that tells us gas is moving away from us.

For a few years, there was a debate about what element was making this fingerprint:

The Leading Suspect: Strontium (Sr). This fits perfectly with our understanding of how heavy elements are made in these collisions.
The Alternative Suspect: Helium (He). A few scientists suggested that under very strange, non-standard conditions, Helium could mimic Strontium's fingerprint.

This paper is the detective work that proves Helium is innocent and Strontium is guilty.

The Detective Work: Why Helium Doesn't Fit

The authors, led by Albert Sneppen, decided to play detective by looking at time. They didn't just look at one picture; they watched a movie of the explosion evolving over the first few days.

Here is why the "Helium Theory" falls apart, explained with analogies:

1. The "Too Fast" Appearance (The Magic Trick)

The Observation: The fingerprint appeared suddenly. One day (0.92 days after the crash), there was nothing. The next day (1.17 days), the fingerprint was fully formed. It popped up in just a few hours.
The Helium Problem: Helium is like a slow-cooking stew. To make a Helium fingerprint visible, you need a huge amount of Helium gas to be in a specific, excited state. If you had enough Helium to make the fingerprint appear on Day 1.17, that same amount of gas would have been visible on Day 0.92.
The Analogy: Imagine you hear a loud siren. If the siren was loud enough to be heard at 1:00 PM, it should have been audible at 12:55 PM. If you didn't hear it at 12:55 PM, but it suddenly blared at 1:00 PM, the siren didn't just "turn on." It must have been something else entirely. The Helium model requires the gas to be there too early, which contradicts the data.

2. The "Fading" vs. "Growing" Paradox

The Observation: As the days went by, the fingerprint got weaker. The gas cloud expanded, became thinner, and the line faded away.
The Helium Problem: Helium behaves like a snowball rolling down a hill. As the explosion cools down (which it does rapidly), Helium actually gets better at making this fingerprint. The colder it gets, the more Helium atoms settle into the right state to show the line. So, if it were Helium, the line should have gotten stronger and brighter over time.
The Analogy: Think of a campfire. As the fire dies down (cools), the smoke usually gets thicker and more visible for a while before it clears. If you saw the smoke get thinner as the fire died, you'd know it wasn't smoke; it was something else. The Helium line should have grown stronger as the kilonova cooled, but it did the opposite.

3. The "Missing Relatives" (The Family Photo)

The Observation: If you have a family member (Helium) making a noise, their siblings should be making noise too.
The Helium Problem: In physics, if Helium is excited enough to make the 1-micron line, it must also make other lines at different colors (specifically at 587 nm and 706 nm). These are like the "siblings" of the main line.
The Analogy: If you see a person wearing a bright red hat (the 1-micron line), you should also see them wearing red shoes (the other lines). In the data, we saw the hat, but the shoes were missing. The other Helium lines were nowhere to be found, which suggests the "person" (Helium) isn't there at all.

4. The "UV Sunburn" (The Invisible Killer)

The Observation: The explosion was blasting out a lot of ultraviolet (UV) light, like a giant tanning bed.
The Helium Problem: Helium is very sensitive to UV light. If there is too much UV, it strips the electrons off the Helium atoms, "killing" the fingerprint before it can form. The authors calculated that the UV light from the explosion was so strong that it would have ionized (destroyed) the Helium needed to make the line.
The Analogy: Imagine trying to build a sandcastle (the Helium line) on a beach during a massive tsunami (the UV light). The sandcastle would be washed away instantly. The UV light was too strong to let the Helium line survive.

The Verdict: It's Strontium

The paper concludes that the only thing that fits all the clues is Strontium.

Strontium behaves exactly like the data: It appears suddenly when the gas cools just enough to recombine, it fades as the gas expands, and it doesn't require impossible amounts of mass.
Helium requires a "miracle" amount of mass (as much as the entire high-speed ejecta of the explosion) to work at the start, and then it should have gotten stronger, not weaker.

Why Does This Matter?

This isn't just about naming an element. It's about understanding how the universe creates the heavy stuff we are made of.

If it were Helium, it would mean our models of how neutron stars merge are wrong.
Since it is Strontium, it confirms our theories: Neutron star mergers are the cosmic factories that forge elements like Strontium, Gold, and Platinum, and they do it in a way that follows the standard rules of physics (Local Thermodynamic Equilibrium) rather than some exotic, chaotic exceptions.

In short: The universe showed us a fingerprint. Some thought it was a ghost (Helium), but the authors proved it was a solid, physical object (Strontium) by watching how it moved, faded, and interacted with light over time.

1. Problem Statement

The kilonova AT2017gfo, associated with the neutron star merger GW170817, exhibited a prominent P Cygni feature at approximately 1 µm (1083 nm). While the consensus interpretation identifies this feature as the Sr II (Strontium II) recombination line, an alternative hypothesis proposed by Tarumi et al. (2023) suggested it could be the He I $\lambda$ 1083.3 nm line (transition $1s2s\ ^3S - 1s2p\ ^3P$ ) under Non-Local Thermodynamic Equilibrium (NLTE) conditions.

The core problem addressed in this paper is the temporal inconsistency of the He I hypothesis. The authors aim to determine if a self-consistent NLTE helium model can reproduce the specific emergence time, spectral evolution, and mass requirements of the observed 1 µm feature in the early epochs (0.92 to 4.4 days post-merger) of AT2017gfo.

2. Methodology

The authors employed a rigorous collisional-radiative NLTE model to simulate the population of helium energy levels and the resulting spectral features.

Observational Data: They utilized a unified, high-cadence spectral compilation of AT2017gfo from 0.92 to 4.4 days post-merger (ANU, SALT, X-shooter, SOAR, Gemini). Crucially, they incorporated Swift-UVOT photometric constraints to model the UV flux, which is critical for photoionization rates.
Atomic Physics Model:
- Implemented a comprehensive network of transitions for neutral helium (He I), including singlet and triplet states up to principal quantum number $n=4$ (and tested up to $n=8$ ).
- Included radiative transitions (bound-bound absorption/emission), collisional transitions (electron impact), photoionization, and recombination.
- Key Innovation: Unlike previous models (e.g., Tarumi et al. 2023) that focused primarily on the $1s2s\ ^3S$ state, this model explicitly included the $1s2p\ ^3P$ state. The authors demonstrated that $1s2p\ ^3P$ is highly susceptible to photoionization by UV photons ( $\lambda < 340$ nm) and decays rapidly to the ground state, acting as a major depopulation channel for the triplet system.
- Non-thermal Ionization: Modeled ionization by non-thermal electrons from r-process radioactive decay (using Spencer-Fano approximations) to generate the necessary He II for recombination into triplet states.
Ejecta Structure: Assumed a spherically symmetric, homologously expanding ejecta with a density profile $\rho \propto v^{-5}$ . They tested various density slopes and electron densities to ensure robustness.
Radiation Field: Used a geometrically diluted black-body radiation field derived from the observed continuum to calculate radiative rates.

3. Key Contributions

Identification of the $1s2p\ ^3P$ Bottleneck: The paper establishes that the $1s2p\ ^3P$ state is a critical, previously underappreciated depopulation channel. Due to its lower ionization threshold ( $\lambda < 340$ nm) compared to the ground triplet state ( $\lambda < 260$ nm) and its rapid natural decay ( $\sim 0.01$ s vs. hours for $1s2s\ ^3S$ ), it effectively prevents the accumulation of triplet helium populations required to produce a strong 1 µm feature in the presence of the observed UV flux.
Quantitative Mass Constraints: The authors calculated the specific helium mass ( $M_{He}$ ) required to produce the observed optical depth ( $\tau$ ) at each epoch. They found that the required mass varies by orders of magnitude between epochs, rendering a single physical model impossible.
UV Flux Constraints: By comparing Swift-UVOT data with model limits, they showed that the actual UV flux in the early days is $\sim 3$ orders of magnitude higher than what is required to sustain a He I triplet population, effectively photoionizing the triplet states.

4. Results

The analysis yields four primary lines of evidence that rule out the He I interpretation:

Emergence Time Inconsistency:
- Observation: The 1 µm feature was absent at 0.92 days but appeared strongly by 1.17 days.
- Model Prediction: To fit the 1.17-day spectrum with He I, a helium mass of $M_{He} \approx 0.017 M_\odot$ is required. However, this mass would have produced a detectable feature at 0.92 days (which was not observed). The non-detection at 0.92 days rules out $M_{He} > 0.004 M_\odot$ at $5\sigma$ .
- Conclusion: The feature appears too suddenly for a helium model, which would require a steady-state population build-up inconsistent with the data.
Evolutionary Trend Mismatch:
- Observation: The feature's optical depth decreases over time (fading absorption).
- Model Prediction: As the kilonova cools, photoionization rates drop, and the $1s2s\ ^3S$ level becomes more populated relative to higher levels. Consequently, a He I feature should grow stronger (increase in optical depth) over time.
- Conclusion: The observed fading is the exact opposite of the predicted He I evolution.
Missing Corroborating Lines:
- If the triplet population were sufficient to create the 1 µm line, the He I $\lambda$ 587.7 nm and $\lambda$ 706.7 nm lines should also be visible (with $\lambda$ 587.7 nm being $\sim 10\%$ the strength of the 1 µm line). These features are not observed in the early spectra.
Unphysical Mass Requirements:
- At early times (1.17 days), the required helium mass ( $\sim 0.01-0.02 M_\odot$ ) is comparable to or exceeds the total mass of the high-velocity "blue" ejecta component. This implies a helium mass fraction ( $X_{He}$ ) near 100%, which is inconsistent with nucleosynthesis simulations (which predict $X_{He} \ll 1\%$ for dynamical ejecta or at most a few percent for high- $Y_e$ ejecta).

5. Significance

Confirmation of Strontium: The results strongly reinforce the identification of the 1 µm feature as Sr II formed under LTE conditions. The observed temporal evolution (sudden appearance, fading absorption) aligns perfectly with the passage of a recombination wave through Sr II in the ejecta.
Constraints on NLTE in Kilonovae: The study demonstrates that while NLTE effects are crucial for certain species at late times, the early photospheric phase of AT2017gfo is well-described by LTE. The failure of the He I model highlights the sensitivity of NLTE triplet populations to UV photoionization and the difficulty of decoupling specific energy levels without extreme, unphysical assumptions.
Implications for Nucleosynthesis: By ruling out a helium origin for the 1 µm feature, the paper places tighter constraints on the helium abundance in the ejecta. It suggests that the "blue" component of the ejecta is dominated by low- $A$ r-process elements (like Sr, Y, Zr) rather than helium, and that non-thermal ionization models predicting the depletion of low-ionization species (like Sr II) may need revision.

In summary, the paper provides a definitive physical argument that the 1 µm P Cygni feature in AT2017gfo cannot be helium, thereby solidifying the evidence for the first direct detection of r-process nucleosynthesis products (Strontium) in a neutron star merger.