Accurate transition and hyperfine data in Ag I from Multiconfiguration Dirac-Hartree-Fock and Relativistic Coupled-Cluster methods

This study employs Multiconfiguration Dirac-Hartree-Fock and relativistic coupled-cluster methods to calculate accurate excitation energies, radiative transition rates, and hyperfine structure constants for 18 states of neutral silver (Ag I), providing extensive data with quantified uncertainties to support precise r-process abundance determinations in late-type stars.

The Gist

Imagine the universe as a giant, cosmic library. In this library, every star is a book, and the "ink" used to write those books is made of chemical elements. To read the story of how the universe formed, astronomers need to understand exactly what kind of ink is in each book.

One specific "letter" in this cosmic alphabet is Silver (Ag). Silver is a special clue that helps astronomers figure out how heavy elements were forged in the violent explosions of dying stars (a process called the r-process). But to read the silver clues correctly, scientists need a perfect dictionary of how silver atoms behave.

This paper is essentially the team building that dictionary. Here is the story of what they did, explained simply:

1. The Problem: A Blurry Map

Think of a silver atom as a tiny solar system with a nucleus in the center and electrons zooming around it. When light hits these electrons, they jump up and down between orbits, creating a unique "fingerprint" of light (a spectral line) that we can see in starlight.

However, calculating exactly where these electrons jump and how bright those jumps are is incredibly hard. It's like trying to predict the exact path of a hummingbird in a hurricane while it's vibrating at a million miles per hour. If your math is slightly off, your "dictionary" is wrong, and you misread the history of the stars.

2. The Tools: Two Super-Computers

The authors used two different, high-powered mathematical "engines" to solve this puzzle:

• MCDHF (Multiconfiguration Dirac–Hartree–Fock): Think of this as a highly detailed 3D map maker. It builds a complex model of the atom by considering every possible way the electrons can wiggle and interact.
• RCC (Relativistic Coupled-Cluster): This is like a precision time-traveling calculator. It accounts for the fact that electrons move so fast (near the speed of light) that they behave differently than slow-moving objects.

The team ran both engines. It's like having two different GPS apps navigate the same route; if both apps say "turn left," you can be very confident that's the right way.

3. The Challenge: The "Ghost" Electrons

Silver is tricky because it has a "core" of electrons that are very tightly packed, and one "lonely" electron on the outside.

• The Easy Part: Calculating the jumps for the lonely outer electron is like walking on a flat sidewalk.
• The Hard Part: Some silver states involve the inner "core" electrons getting excited. This is like trying to walk on a sidewalk that is constantly shaking, shifting, and rearranging itself. The paper admits that for these specific "core-excited" states, the math is still a bit fuzzy (they call this the "E class" of uncertainty, meaning there's a lot of guesswork left).

4. The Results: A New, Better Dictionary

After crunching the numbers, the team produced a massive list of data for 18 different energy states of silver. Here is what they found:

• The "Speed Limits" (Transition Rates): They calculated exactly how fast an electron drops from a high orbit to a low one. They rated their own work on a "quality scale" (like AA, A, B, C). Most of their data is AA or A quality (less than 3% error), which is excellent.
• The "Fingerprint" (Hyperfine Structure): Silver atoms have a tiny magnetic spin that splits their light into multiple fine lines. The team calculated these splits with high precision, matching real-world experiments almost perfectly.
• The "Metastable" Mystery: They found a specific state of silver (called 4d95s2) that is "metastable." Imagine a ball balanced on a very flat hilltop; it doesn't roll down easily. This silver atom sits there for a long time (about 163 milliseconds) before finally dropping. This is crucial because it affects how much silver we think is in a star.

5. Why This Matters

Before this paper, astronomers had to use "best guesses" or older, less accurate data for silver. If your dictionary has the wrong spelling, you might think a word means "peace" when it actually means "war."

By providing this new, highly accurate data:

1. Better Star Stories: Astronomers can now read the "silver" in starlight with much higher confidence.
2. Understanding the Universe: This helps us understand how the heavy elements in our universe (like the gold in your jewelry or the silver in your jewelry) were created in the first place.
3. Non-LTE Physics: In the real world, stars aren't in a perfect, calm equilibrium. This new data allows scientists to model the chaotic, non-equilibrium environments of stars much better.

The Bottom Line

The authors didn't just guess; they built two independent, super-accurate models of the silver atom, compared them against real-world experiments, and created a "gold standard" (or perhaps a "silver standard"!) of data. This new dictionary will help astronomers rewrite the history books of the stars with much greater clarity.

Technical Summary

1. Problem Statement

Silver (Ag) is a critical tracer for understanding the "weak r-process" in late-type stars, which produces elements up to the second r-process peak ($N=82$). Accurate determination of stellar silver abundances relies heavily on precise atomic data. However, current analyses face challenges:

• Non-LTE Effects: When the assumption of Local Thermodynamic Equilibrium (LTE) is relaxed, accurate radiative transition data are required not just for diagnostic lines, but for the entire model atom to correctly solve coupled statistical-equilibrium equations.
• Data Gaps: While some resonance lines (Ag I 328 nm and 338 nm) are well-constrained, a comprehensive set of transition rates, hyperfine structure constants, and lifetimes for Ag I is lacking, particularly for core-excited states and higher-order transitions.
• Theoretical Limitations: Existing theoretical data often lack the precision required for modern 3D radiation-hydrodynamics stellar atmosphere models, and discrepancies exist between different theoretical approaches and experimental measurements.

2. Methodology

The authors employed two advanced, independent relativistic many-body methods to compute atomic data for 18 states of neutral silver (Ag I), covering configurations up to $4d^{10}8s$:

• Multiconfiguration Dirac–Hartree–Fock (MCDHF) and Relativistic Configuration Interaction (RCI):

  • Performed using the GRASPG package.
  • Utilized a systematic expansion of Configuration State Functions (CSFs) via single and double (SD) excitations from reference configurations ($4d^{10}\{5s, \dots, 4f\}$) into increasingly larger orbital sets (up to layer 6).
  • Included Breit interactions and Quantum Electrodynamics (QED) effects in the RCI step.
  • Applied a Quantitative and Qualitative Evaluation (QQE) approach to estimate uncertainties based on NIST ASD terminology, utilizing length and velocity gauge comparisons.
  • Fine-tuned (FT) Hamiltonian matrix elements for specific core-excited states ($4d^9 5s^2$) to correct energy overestimations.

• Relativistic Coupled-Cluster (FSMRCC):

  • Utilized Fock-space Multi-Reference Coupled Cluster theory.
  • Included triple excitations and self-consistent Breit interactions.
  • Employed Gaussian Type Orbitals (GTOs) and considered Bohr-Weisskopf (BW) corrections for magnetic dipole hyperfine constants.
  • Results were generated at the RCCSD (singles and doubles) and RCCSDT (singles, doubles, and triples) levels.

3. Key Contributions

• Comprehensive Dataset: Generated a complete set of excitation energies, electric dipole (E1) transition rates, weighted oscillator strengths ($gf$), hyperfine structure constants ($A_{hfs}$), and lifetimes for 18 states.
• Uncertainty Quantification: Provided rigorous uncertainty estimates for 57 E1 transitions, classifying them according to NIST ASD standards (ranging from AA $\le$ 1% to E > 50%).
• Metastable State Analysis: Calculated the lifetime of the $4d^9 5s^2 \ ^2D_{5/2}$ metastable state, which decays via an E2 transition, a critical parameter for ionization balance in stellar models.
• Methodological Benchmarking: Conducted a direct, high-level comparison between MCDHF/RCI and FSMRCC methods, identifying specific regimes where each method excels.

4. Key Results

Excitation Energies

• MCDHF/RCI: Generally overestimated core-excited $4d^9 5s^2$ states by ~9000 cm$^{-1}$ in the core-valence (CV) model, requiring manual fine-tuning.
• FSMRCC: Showed better agreement with experiment for most states and correctly reproduced the ordering of the closely spaced $4d^{10}6d$ and $4d^{10}4f$ configurations, which MCDHF/RCI struggled with.
• Agreement: FSMRCC values were generally closer to experimental NIST data than MCDHF/RCI, except for the $4d^9 5s^2$ states where FSMRCC still overestimated by ~4800 cm$^{-1}$.

Hyperfine Interaction Constants ($A_{hfs}$)

• Ground/First Excited States: FSMRCC results agreed better with high-precision experimental data (e.g., for $5s \ ^2S_{1/2}$ and $5p$ states) than MCDHF/RCI.
• Core-Excited States: For the $4d^9 5s^2 \ ^2D_{5/2}$ state, MCDHF/RCI provided a better estimate than FSMRCC, likely due to the difficulty of handling triple excitations in the current FSMRCC implementation for these specific configurations.
• Recommendation: The authors recommend FSMRCC values for ground/low-lying states and MCDHF/RCI values for the $4d^9 5s^2$ metastable states.

Transition Rates and Oscillator Strengths

• General Agreement: For transitions between ordinary Rydberg states (s, p, d, f sequences), MCDHF/RCI and FSMRCC agreed well (median relative difference ~1.7–3.5%).
• Uncertain Transitions: The $6p \to 5s$ and $7p \to 5s$ transitions showed significant discrepancies. MCDHF/RCI length-gauge results were sensitive to basis completeness, while velocity-gauge results aligned better with FSMRCC. These were flagged as having higher uncertainty.
• Two-Electron One-Photon (TEOP) Transitions: Transitions involving the $4d^9 5s^2$ core-excited states are weak and computationally challenging. FSMRCC could not compute these; MCDHF/RCI provided them but assigned them to Class E (uncertainty > 50%) due to their weak nature and lack of experimental validation.
• Comparison with Experiment: The computed weighted oscillator strengths for resonance lines ($5p \to 5s$) perfectly matched experimental values derived from picosecond delayed-coincidence measurements ($gf = 0.464$ and $0.952$).

Lifetimes

• Agreement: Computed lifetimes from both methods generally fell within the error bars of experimental Laser-Induced Fluorescence (LIF) measurements.
• Metastable State: The $4d^9 5s^2 \ ^2D_{5/2}$ state has a calculated lifetime of 163 ms (MCDHF/RCI), which is in reasonable agreement with a recent CI-MBPT calculation (192 ms). No experimental technique currently exists to measure this specific lifetime.

5. Significance

• Stellar Abundance Analysis: The provided dataset, particularly the recommended mean values of transition parameters and their associated uncertainties, enables more accurate non-LTE abundance determinations of silver in late-type stars. This is crucial for constraining the weak r-process nucleosynthesis models.
• Atomic Physics Benchmark: The study serves as a rigorous benchmark for relativistic many-body theories. It highlights the strengths of FSMRCC for ground states and MCDHF/RCI for specific core-excited configurations, guiding future theoretical developments.
• Data Quality: By assigning NIST ASD uncertainty classes, the authors provide a transparent framework for astronomers to weigh the reliability of specific spectral lines in their models, moving beyond the "black box" of theoretical data.

Final Recommendation: The authors recommend using the mean of FSMRCC and MCDHF/RCI values for transition parameters where both methods are applicable, with uncertainties assigned conservatively based on the worst-case scenario between the two methods or the QQE gauge dependence analysis. For hyperfine constants, FSMRCC is preferred for ground states, while MCDHF/RCI is preferred for the $4d^9 5s^2$ metastable states.