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

Abstract

The spectral features observed in kilonovae (KNe) reveal the elemental composition and the velocity structures of matter ejected from neutron star mergers. In the spectra of the kilonova AT2017gfo, a P Cygni line at about 1$μ$m has been linked to Sr II, providing the first direct evidence of freshly synthesised $r$-process material. An alternative explanation to Sr II was proposed - He I $λ1083.3$nm under certain non-local-thermodynamic-equilibrium (NLTE) conditions. A key way to robustly discriminate between these identifications, and indeed other proposed identifications, is to analyse the temporal emergence and evolution of the feature. In this analysis we trace the earliest appearance of the observed feature and detail its spectro-temporal evolution, which we compare with a collisional-radiative model of helium. We show that the 1$μ$m P Cygni line is inconsistent with a He I interpretation both in emergence time and in subsequent spectral evolution. Self-consistent helium masses cannot reproduce the observed feature, due to the diminishing strength of radiative pathways leaving triplet helium.

Figures (6)

• Figure 1: Evolution of the 1 $\mu$m P Cygni feature over the first five days post merger from the unified spectral compilation Sneppen2024. The spectra were taken post merger plus 0.92 day Andreoni2017, 1.17 day Buckley2018, 1.47 day Nicholl2017Chornock2017 1.43, 2.4, 3.4, and 4.4 day Pian2017Smartt2017. The spectra prior to one day contain no strong deviation from the black-body continuum Shappee2017. However, between 0.92 and 1.17 days, strong absorption appears, while an emission feature begins to emerge with a delay of around ten hours after this (1.43-1.47 days). We note that these are the predicted emergence time framess for both absorption and emission for a feature produced by LTE Sr ii due to a recombination wave passing through the ejecta Sneppen2023_rapid. We show in this analysis that this emergence is inconsistent with a He i$\lambda 1083.0$ nm interpretation of the feature.
• Figure 2: Grotrian diagram showing the He i energy levels. The 1s2s $^3$S--1s2p $^3$P 1083 nm transition could produce a 1 $\upmu$m P Cygni feature in NLTE models Tarumi2023 because 1s2s $^3$S can be difficult to ionise (e.g. requiring UV photons $<260\,$nm) and the natural transition to the ground state is very slow (e.g. characteristic timescale of hours). However, as shown here, the other well-populated triplet state, 1s2p $^3$P, is several orders of magnitude more susceptible to ionisation and naturally decays to the ground state $10^6$ times faster, on a timescale of 0.01 second.
• Figure 3: SALT and VLT/X-shooter spectra with Swift-UVOT photometric constraints on the early UV flux. Photons with energies sufficient to ionise the population at 1s2s $^3$P ($\lambda < 340$ nm rest frame or equivalently $\lambda < 280$ nm corrected for ejecta expanding at $0.28c$) is shown by the red region. The UV flux detected by Swift-UVOT in Evans2017 decreases with time, falling below the 3$\sigma$ (2$\sigma$) detection threshold after 1.1 days (3.0 days). For comparison with the Swift observations, the maximum UV flux that allows He i to still produce the 1 $\upmu$m feature (within the Tarumi2023 NLTE helium model) is shown in black. The Swift flux is around three orders of magnitude larger than what is permitted for the helium interpretation.
• Figure 4: Network Grotrian diagram indicating the dominant pathways entering and leaving any level. The arrow-width shows the transition rate on a logarithmic scale, with the colour indicating the dominant mechanism. At early times, the dominant pathways leaving triplet He i is photoionisation. At later times (as the upper triplet levels are less populated), natural decay to the He i ground state and collisional pathways out of the triplet become increasingly important.
• Figure 5: Estimated helium mass required to produce the absorption of the 1 $\upmu$m feature in early spectra. These estimates are derived using the NLTE helium model described in Sec. \ref{['sec:he_NLTE_calculation']}. The left y-axis indicates the approximate He mass fraction (given the high velocity ejecta component is around $\sim$0.01 M$_\odot$, which is likely most representative around three to five days post merger), while the exact helium mass fraction for each epoch can be found in Table \ref{['table:table']}. The required helium mass changes drastically between epochs, and for any epoch, it is largely inconsistent with all remaining epochs (primarily due to the decreasing strength of radiative pathways leaving triplet helium). At early times, the helium mass required for a feature is comparable to the total mass of the early higher-velocity ($v\approx0.2$-$0.3c$) 'blue' ejecta. The absence of a helium feature at 0.92 days provides a $5\sigma$ upper limit on the helium mass (reported as solely the statistical uncertainty from the observed spectra). The grey dashed line indicates the mass fraction considered in Tarumi2023.