Nonadiabatic corrections to electric quadrupole transition rates in H$_2$

This paper derives and calculates nonadiabatic corrections to electric quadrupole transition rates in the hydrogen molecule, demonstrating that these corrections, expressed via a quadrupole moment curve, can alter fundamental band transition rates by 0.4% to 12% depending on rotational quantum numbers.

The Gist

Imagine the hydrogen molecule ($H_2$) as a tiny, dancing pair of atoms. For a long time, scientists have tried to predict exactly how this pair moves and how it interacts with light. To do this, they usually use a "simplified map" called the Born-Oppenheimer approximation. Think of this map as assuming the two heavy nuclei (the dancers' feet) are frozen in place while the light electrons (the dancers' swirling skirts) move around them. It's a great first sketch, but it's not perfect.

This paper is about drawing a much more detailed, high-definition map that accounts for the fact that the feet do move, and they wiggle in sync with the skirts. This "wiggling" is called a nonadiabatic correction.

Here is the breakdown of what the authors did, using simple analogies:

1. The Problem: A Slightly Blurry Photo

Scientists want to know exactly how fast the hydrogen molecule emits light when it jumps from one energy level to another. Specifically, they are looking at a type of light emission called an electric quadrupole transition.

• The Analogy: Imagine the molecule is a spinning top. Sometimes, it doesn't just spin; it wobbles in a specific, complex way that emits a faint signal. The "standard map" (Born-Oppenheimer) predicts the speed of this wobble, but it misses a tiny detail: the fact that the heavy parts of the top aren't perfectly still. This missing detail causes the prediction to be slightly off—sometimes by a tiny bit, sometimes by a lot.

2. The Solution: A New "Correction Curve"

The authors derived a new mathematical formula to fix this.

• The Analogy: Think of the old map as a flat, 2D drawing of a mountain. It's good, but it doesn't show the bumps and valleys. The authors created a new "elevation curve" (called $D^{(1)}(R)$) that acts like a set of instructions to add those missing bumps and valleys to the drawing.
• They didn't just guess these bumps; they calculated them using a sophisticated method called Nonadiabatic Perturbation Theory (NAPT). This is like using a super-precise 3D scanner to measure the exact shape of the molecule's movement, rather than just guessing based on how heavy the atoms are.

3. The Calculation: Building a Better Model

To get these numbers, the authors used a specific type of mathematical "Lego set" (called the Kołos-Wolniewicz basis).

• The Analogy: Imagine trying to build a perfect model of a cloud. You can't just use big blocks; you need tiny, flexible pieces that can mold into every curve. The authors used millions of these tiny mathematical pieces to simulate the electron cloud. They tested two different "building styles" (James-Coolidge and Heitler-London) depending on whether the atoms were close together or far apart, ensuring the model was accurate everywhere.

4. The Results: How Much Does It Matter?

When they applied their new "correction curve" to calculate how fast the molecule emits light, they found the results changed significantly.

• The Analogy: If you were timing a race, the old map said a runner would finish in 10.00 seconds. The new map says, "Actually, because of a slight breeze we missed, it's 10.12 seconds."
• The Numbers: For some specific movements of the molecule, the speed of the light emission changed by as little as 0.4%, but for others, it changed by as much as 12%.
  • In the "S-branch" (a specific type of molecular wobble), the correction was huge (12%) because the original speed was so slow that even a tiny nudge made a big difference.
  • In the "O-branch," the change was small and steady (about 0.4%).

5. Why This Matters (According to the Paper)

The authors explain that this work is a crucial step toward primary thermometry (measuring temperature with extreme precision).

• The Analogy: Imagine trying to measure the temperature of a room by listening to how fast a specific musical note is played by a hydrogen molecule. If your map of how that note is played is slightly wrong, your temperature reading will be wrong.
• The paper suggests that by using their new, ultra-accurate map, scientists can measure temperatures as low as 10 Kelvin (very cold!) with much higher accuracy. They propose measuring the ratio of two different "notes" (transition rates) to cancel out errors, and for this to work, the theoretical map must be perfect.

Summary

In short, the authors took a standard, slightly blurry picture of how hydrogen molecules interact with light and sharpened it. They calculated the exact "wobble" of the heavy atoms that was previously ignored. This new, sharper picture changes the predicted speed of light emission by up to 12% in some cases, providing the foundation for measuring extremely low temperatures with unprecedented accuracy.

Technical Summary

Technical Summary: Nonadiabatic Corrections to Electric Quadrupole Transition Rates in H₂

Problem Statement
While the Born-Oppenheimer (BO) approximation suffices for many molecular property calculations, high-precision applications—such as primary thermometry at temperatures as low as 10 K and the analysis of accurately measured transition energies in hydrogen (H₂)—require the inclusion of adiabatic and nonadiabatic corrections. Specifically, there is a need for a rigorous derivation and numerical calculation of the leading nonadiabatic correction to the rates of electric quadrupole (E2) transitions in the hydrogen molecule. Previous work by Wolniewicz and Komasa provided the BO quadrupole moment function, $D^{(0)}(R)$, but a comprehensive treatment of the nonadiabatic correction, $D^{(1)}(R)$, and its impact on transition rates was required to reduce theoretical uncertainties.

Methodology
The authors employ Nonadiabatic Perturbation Theory (NAPT) to derive the leading correction to the electric quadrupole transition rates.

1. Theoretical Derivation: Starting from a nonrelativistic Hamiltonian for a neutral diatomic molecule, the authors fix the reference frame at the nuclear mass center. They decompose the total wave function into an adiabatic part and a nonadiabatic correction. By applying NAPT to a Hermitian electronic operator $Q$ (the quadrupole moment), they derive a formula where the first-order nonadiabatic correction to the matrix element can be expressed as a nuclear matrix element of a single electronic function, $D^{(1)}(R)$.
2. Operator Formulation: The electric quadrupole moment operator is rewritten in terms of relative coordinates. The resulting correction function $D^{(1)}(R)$ is expressed as a sum of four terms ($Q_1$ through $Q_4$), involving expectation values of electron coordinates, derivatives with respect to the internuclear distance $R$, and angular momentum operators.
3. Numerical Implementation: To calculate these terms with high precision, the authors utilize the Kołos-Wolniewicz (KW) basis set, employing explicitly correlated exponential functions with polynomial dependence on interparticle distances.
   • For the ground electronic state ($\Sigma_g^+$), they use a variational approach with James-Coolidge (JC) basis functions for $R \le 10$ a.u. and Heitler-London (HL) basis functions for $R \ge 10$ a.u.
   • Intermediate states required for the nonadiabatic terms are constructed by solving linear equations within the same basis sets.
   • The calculation of the most difficult term, $Q_2$ (involving $R$-derivatives), is performed using a numerical differentiation technique with respect to a perturbation parameter $\xi$, validated against analytical methods.
   • Calculations are performed in quad-double arithmetic (64 decimal digits) using parallel solvers to ensure high precision.

Key Contributions

• Derivation of $D^{(1)}(R)$: The paper provides a complete formula for the leading nonadiabatic correction to the quadrupole moment in a homonuclear diatomic molecule, represented as a single curve $D^{(1)}(R)$ analogous to the BO curve $D^{(0)}(R)$.
• High-Precision Numerical Results: The authors present numerical results for both $D^{(0)}(R)$ and $D^{(1)}(R)$ over a wide range of internuclear distances ($R \le 50$ a.u.). The relative accuracy is typically better than $10^{-9}$, reaching $\sim 10^{-17}$ for $D^{(0)}(R)$ at large $R$.
• Asymptotic Analysis: The study confirms that the long-range asymptotics of the nonadiabatic correction $D^{(1)}(R)$ scales as $R^{-6}$, consistent with the BO term, despite individual components having $R^{-3}$ leading terms that cancel out in the total sum.
• Correction of Literature: The authors identify and correct a misprint in the rotational intensity factors found in Ref. [10] (specifically for the $J'' = J' + 2$ case), while confirming that the numerical rates in that reference were calculated correctly.

Results
The authors calculated the spontaneous electric quadrupole transition rates ($A_{E2}$) for the fundamental vibrational band ($v=1 \to 0$) of H₂, including the nonadiabatic correction.

• Magnitude of Corrections: The nonadiabatic corrections significantly alter the transition rates depending on the rotational quantum numbers ($J'$).
  • For the Q-branch ($J'' = J'$) and O-branch ($J'' = J' + 2$), the relative corrections range from approximately 0.4% to 1.14%.
  • For the S-branch ($J'' = J' - 2$), the correction is highly variable, starting at $\approx 0.45\%$ for $J'=2$ and reaching a maximum of 12% at $J'=15$. This large deviation is attributed to the smallness of the radial matrix element at this specific quantum number.
• Comparison with Other Effects: In the Q-branch, for $J' \ge 18$, the magnetic dipole (M1) transition rate exceeds the total nonadiabatic correction to the E2 rate. The authors note that for Q-branch transitions, the total spontaneous emission probability must include both M1 and E2 channels.
• Uncertainty: The uncertainty in the calculated E2 rates is now dominated by unknown relativistic corrections (estimated as $\alpha^2 \times A_{E2}$), rather than the nonadiabatic terms.

Significance
The paper establishes a foundational step toward accurate primary thermometry using hydrogen molecules. By providing the $D^{(1)}(R)$ curve, the authors enable the recalculation of transition rates with reduced theoretical uncertainty. This is crucial for proposals to measure temperature via the ratio of two transition rates from the same vibrational band, a method that relies on the cancellation of experimental and theoretical uncertainties. The authors state that these results allow for the most accurate determination of transition rates to date, serving as a necessary precursor to future work involving direct nonadiabatic calculations and extensions to heteronuclear molecules (like HD) and other molecules (D₂, T₂). The work also demonstrates that the nonadiabatic corrections are significantly larger (by a factor of a few to ten) than what would be expected from a simple mass scaling factor ($m_e/m_n$).