The MARTINI Platform (I): Se I-X atomic calculation and expansion opacity for early-stage kilonova spectral analysis

This paper presents a comprehensive study of selenium atomic data from Se I to Se X using GRASP2018 to calculate expansion opacities and perform spectral analysis with POSSIS, revealing that selenium spectral features are only detectable in kilonova scenarios where selenium constitutes 100% of the ejecta, with all results now accessible via the open-source MARTINI platform.

The Gist

The Big Picture: Cosmic Alchemy and the "Fog" of Death

Imagine two neutron stars (the incredibly dense, dead cores of massive stars) colliding in a cosmic dance. When they smash together, they don't just make a loud noise; they create a "kilonova." This is a spectacular explosion that acts like a giant cosmic factory, forging heavy elements like gold, platinum, and selenium out of thin air.

To understand what happened in this explosion, astronomers look at the light it emits. But here's the problem: the explosion is shrouded in a thick, expanding cloud of gas. This gas acts like a fog that blocks and scatters the light. In physics, this "fog" is called opacity.

If you want to know exactly what ingredients (elements) were made in the explosion, you have to understand how thick the fog is and what it's made of. If your map of the fog is wrong, your guess about the ingredients will be wrong, too.

The Mission: Mapping the Selenium Fog

This paper is about creating a better, more detailed map of the fog, specifically for the element Selenium (Se).

Selenium is a bit of a "middle child" in the periodic table. It's not as famous as gold or uranium, but in the early moments of a kilonova (the first day or so), it is produced in huge quantities. However, scientists didn't have a very good map of how selenium interacts with light. They were using old, blurry maps.

The authors of this paper decided to build a brand new, high-definition map. They did this by:

1. Crunching the Numbers: They used a super-complex computer code (called GRASP2018) to calculate exactly how selenium atoms behave when they are stripped of their electrons (ionized) in the extreme heat of an explosion.
2. Checking the Work: They compared their new calculations against the "gold standard" database (NIST) and other recent studies to make sure they were right. They found their map was sharper and more accurate than previous attempts.
3. Simulating the Explosion: They took their new selenium data and plugged it into a 3D simulation of a kilonova (using a code called POSSIS) to see how the light would look to an observer on Earth.

The Results: The "100% vs. 10%" Test

To see if their new map mattered, they ran two different scenarios, like testing a new recipe in two different kitchens:

• Scenario A (The Pure Selenium Kitchen): They imagined a kilonova made of 100% selenium. In this case, the new map showed clear, distinct "fingerprints" (spectral features) in the light. It was easy to see that selenium was there.
• Scenario B (The Realistic Kitchen): They imagined a more realistic kilonova where selenium is only 10% of the mix, and the other 90% is a generic "gray fog" (representing other elements).

The Surprise: In the realistic 10% scenario, the selenium fingerprints disappeared. The "gray fog" of the other elements was so thick that it completely drowned out the specific signals of the selenium.

The Analogy: Think of it like trying to hear a single violin (selenium) in a concert hall.

• If the hall is empty and only the violin is playing (100% scenario), you hear it perfectly.
• If the hall is packed with a full orchestra playing loudly (the 90% other elements), the violin is still there, but you can't distinguish its sound from the rest of the music.

Why This Matters

1. Better Maps for the Future: Even though they couldn't "see" selenium in the realistic mix yet, they have now provided the scientific community with the most accurate data on selenium ever created. This is like giving astronomers a better pair of glasses.
2. The MARTINI Platform: All this new data has been uploaded to a new, free online website called MARTINI. Think of it as a digital library where anyone can download these new "fog maps" to help them decode future cosmic explosions.
3. Understanding the Universe: By refining how we calculate opacity, we get closer to understanding exactly how the heavy elements that make up our world (and our bodies) were forged in the violent collisions of the universe.

In a Nutshell

The authors built a super-accurate computer model of how selenium behaves in a star explosion. They found that while selenium leaves a clear signature if it's the only thing exploding, its signature gets hidden when it's mixed with other elements. However, by providing this high-quality data to the world, they've helped astronomers prepare for the day when our telescopes are powerful enough to spot those hidden fingerprints in the cosmic fog.

Technical Summary

1. Problem Statement

In the multi-messenger era, kilonovae (KNe) resulting from binary neutron star (BNS) mergers are critical sites for r-process nucleosynthesis. Accurate modeling of KN light curves and spectra requires precise knowledge of the expansion opacity of the ejecta.

• The Gap: While lanthanides and actinides are well-studied for their role in late-time KN opacity, light r-process elements (specifically those with atomic numbers $Z \approx 30-50$) dominate the opacity during the early stage ($\sim 0.5 - 1.5$ days post-merger).
• Specific Deficiency: Selenium (Se) is a highly abundant element in high-electron-fraction ($Y_e \approx 0.30-0.35$) ejecta (accounting for $\sim 10\%$ of the mass), yet its atomic data for highly ionized states (Se I–X) is incomplete or lacks sufficient accuracy for radiative transfer simulations. Existing datasets (e.g., Tanaka et al. 2020) show discrepancies with experimental databases (NIST ASD), leading to uncertainties in spectral predictions.

2. Methodology

The authors employed a multi-step computational approach combining atomic physics calculations with radiative transfer simulations:

• Atomic Structure Calculations (GRASP2018):

  • Code: Used the General Relativistic Atomic Structure Package (GRASP2018), which employs fully relativistic Multiconfiguration Dirac-Hartree-Fock (MCDHF) and Relativistic Configuration Interaction (RCI) methods.
  • Scope: Calculated atomic data for Selenium from neutral (Se I) to fully stripped (Se X).
  • Strategy:
    • Se I–IV: Defined Multi-Reference (MR) sets including $f$-shell orbitals (unlike some previous works) to improve accuracy. Used specific electron configurations (ECs) and optimized radial wavefunctions separately for even and odd parities.
    • Se V–X: Adapted the strategy for higher ionization. For Se VI–VII, orbital symmetry was adjusted to ensure code convergence. For Se VIII–X, the core was treated as frozen $[Ne]$, and $f$-shells were excluded as valence electrons occupy $s, p, d$ shells at high temperatures.
  • Validation: Energy levels were compared against the NIST Atomic Spectra Database (ASD). Transition probabilities were evaluated using a Quantitative and Qualitative Evaluation (QQE) method (classes AA to E).
  • Ionization Balance: Used Saha-Boltzmann equations to determine ionization fractions at various temperatures ($T = 5,000 - 100,000$ K).

• Expansion Opacity Estimation:

  • Calculated bound-bound expansion opacity using the Sobolev formalism combined with the Eastman & Pinto (1993) formula.
  • Assumed Local Thermodynamic Equilibrium (LTE).
  • Tested densities ($\rho = 10^{-13}$ to $3 \times 10^{-12}$ g cm$^{-3}$) and temperatures relevant to early KN evolution.

• Spectral Analysis (POSSIS):

  • Code: Used the 3D Monte Carlo radiative transfer code POSSIS.
  • Input: Pre-computed opacity grids based on the new Se data, covering densities ($\log \rho = -19.5$ to $-4.5$) and temperatures ($1,000 - 51,000$ K).
  • Scenarios: Simulated two ejecta compositions:
    1. 100% Se: Pure selenium ejecta to isolate spectral features.
    2. 10% Se: Realistic scenario where Se is $10\%$ of the mass, with the remaining $90\%$ modeled as a gray opacity ($\kappa = 0.5$ cm$^2$ g$^{-1}$).
  • Epochs: Analyzed spectra at $t = 0.5$ d, $1.0$ d, and $1.43$ d.

3. Key Contributions

1. New Atomic Dataset: Generated a comprehensive set of atomic data for Se I through Se X, including energy levels, transitions, and oscillator strengths.
2. MARTINI Platform: Developed and populated the MARTINI (Multi-messenger Astrophysics Research Tool for Nucleosynthesis and Opacities) open-source platform. This platform hosts the new Se data, allowing researchers to access electron configurations, energy levels, opacities, and fluxes.
3. Improved Accuracy: Demonstrated that the new calculations offer significant improvements over previous literature (specifically Tanaka et al. 2020) regarding energy level accuracy for Se I–IV.
4. Spectral Detectability Study: Provided a definitive analysis on the observability of Selenium features in KN spectra under realistic abundance conditions.

4. Key Results

• Atomic Data Accuracy:

  • Se I–IV: The new calculations show good agreement with NIST ASD. The mean relative error is reduced by $\sim 3-5\%$ compared to Tanaka et al. (2020). However, the results are slightly less accurate than specific works by Radžiūtė & Gaigalas (2022) and Kitovienė et al. (2024) for Se II and Se III, highlighting the need for further refinement of RCI layers.
  • Se V–VIII: Calculations show high precision, with deviations from NIST ASD of $< 2.5\%$ (Se V, VII) and $< 2\%$ (Se VI, VIII).
  • Se IX–X: No NIST data exists for these states; reliability is based on the high-quality transition classification (mostly AA and A classes) derived from the wavefunction accuracy.

• Expansion Opacity:

  • Opacity behavior shifts with temperature. At $T=5,000$ K, Se I and II dominate; at $T=20,000$ K, Se III–V are dominant; at $T=100,000$ K, Se VI–X drive the opacity.
  • The new data produces opacity curves that differ significantly from existing literature, particularly in the UV and optical ranges, due to the improved energy levels and transition probabilities.

• Spectral Analysis (POSSIS):

  • 100% Se Scenario: Distinct Selenium spectral features are clearly observable in the UV/optical range. The spectra peak at shorter wavelengths due to lower overall opacity compared to lanthanide-rich models.
  • 10% Se Scenario (Realistic): When Se constitutes only $10\%$ of the mass, its specific spectral features become undetectable. The gray opacity of the remaining $90\%$ of the ejecta overwhelms the Se contribution.
  • Comparison with AT2017gfo: The 100% Se model is too blue (hotter photosphere) compared to AT2017gfo observations. The 10% Se model matches the temperature better but fails to reproduce the strong flux suppression (line blanketing) seen in real data, suggesting the need for more complex multi-component ejecta models.

5. Significance

• Refining Early-Stage Models: This work provides the first robust atomic data for highly ionized Selenium, a key component of high-$Y_e$ ejecta. This is essential for interpreting early-time KN observations (e.g., from X-shooter or future JWST data).
• Benchmarking: The study establishes that while pure-element models can reveal atomic features, realistic abundance mixing ($\sim 10\%$) often masks these features, implying that detecting specific light r-process elements in mixed ejecta is challenging without extremely high signal-to-noise ratios or specific line-of-sight geometries.
• Community Resource: The integration of these results into the MARTINI platform ensures that the astrophysics community has immediate access to standardized, open-source atomic data for future nucleosynthesis and KN modeling efforts, moving beyond the limitations of static literature tables.
• Future Directions: The authors note that while the current data is a significant improvement, further refinement of RCI layers is needed to match the precision of the best existing datasets for specific ionization stages. Future work will extend this framework to other light r-process elements.