Classical-limit formula for matrix elements between bound states of distinct one-dimensional potentials

AI summary

Overview

Research area: Theoretical quantum physics — semiclassical (WKB) methods for matrix elements between bound states of different one-dimensional potentials, with strong ties to atomic, molecular, and optical physics (Born-Oppenheimer potentials, optical lattice clocks).

Technical level: Advanced. The paper relies on WKB wave functions, phase functions, density of states, and Pöschl-Teller exact solutions.

Scope (one sentence): The paper derives and then validates a closed-form "classical-limit formula" (CLF) for matrix elements of a slowly varying operator between bound states belonging to two distinct one-dimensional potentials.

What This Paper Is About

When a quantum system has two different slowly varying potentials — for example potential energy curves of different electronic states of a diatomic molecule, or the motional potentials of an atom in an optical lattice — one often needs matrix elements such as Franck-Condon overlap factors between a bound state of one potential and a bound state of the other. In the classical limit each potential supports a very dense set of bound states, and computing these matrix elements state-by-state becomes impractical. The goal of this paper is to derive a simple analytical formula that gives the (phase-averaged) magnitude of such matrix elements directly from the potentials and the local classical quantities at a special point, and then to test that formula against exactly solvable cases.

Key Contributions

1. Derivation of the CLF. Starting from WKB wave functions for two potentials $U_1(x)$ and $U_2(x)$ supporting states with $n_i \gg 1$, the author derives Eq. (21), an explicit formula for $|M|^r$ in terms of the densities of states $g_1$, $g_2$, the local wavenumber $k(s)$, the slope difference $|U_1'(s) - U_2'(s)|$, and the operator $f(s)$ — all evaluated at a "common oscillation point" (COP) $s$.

2. Identification of the COP as the controlling point. The derivation shows that the matrix element is accumulated almost entirely in the neighborhood of the point $s$ where $E_1 - U_1(s) = E_2 - U_2(s)$, i.e., where the two wave functions have a common local wavenumber $k_1(s) = k_2(s)$. If no such point exists, the matrix element is assigned a value of zero.

3. A phase-averaging rule. Because a single matrix element's sign and phase are arbitrary and lose meaning as the spectrum becomes continuous, the author replaces the rapidly varying cosine factor by its average, obtaining the $\Gamma$-function factor $\Gamma!\left(\frac{1+r}{2}\right)/[\sqrt{\pi},\Gamma!\left(1+\frac{r}{2}\right)]$, which evaluates to $2/\pi$ for $r=1$ and $1/2$ for $r=2$.

4. Two independent validations. First, an exact consistency check: for $f(x)=1$ and $r=2$, summing $|M|^2$ over all states of $U_2$ and converting the sum to an integral yields exactly $I = 1$, as completeness requires. Second, a comparison against a fully analytical Pöschl-Teller model with exact matrix elements (Eq. 29), evaluated with rational arithmetic to eliminate numerical error, showing convergence toward the CLF as the classical limit is approached.

Main Findings

• The CLF takes a compact closed form: Eq. (21) gives $|M|^r$ as the $\Gamma$-function phase-average factor times $|f(s)|^r$ times $[\pi (\hbar^2/2m), g_1 g_2, k(s), |U_1'(s)-U_2'(s)|]^{-r/2}$, with all quantities evaluated at the COP $s$.
• The matrix element localizes at the COP: away from $s$ the interference term $\cos[\Phi_1(x)-\Phi_2(x)]$ oscillates rapidly and averages to zero; near $s$ the phase difference has no linear term (since $k_1(s)=k_2(s)$) and the integral reduces to the standard Fresnel-type result of Eq. (19).
• Existence conditions matter: a COP exists only if the turning points are ordered as $a_1<a_2<b_1<b_2$ (case I) or $a_2<a_1<b_2<b_1$ (case II); otherwise the matrix element is zero. Case I corresponds to $U_1'(x) > U_2'(x)$ and case II to $U_1'(x) < U_2'(x)$.
• The completeness check passes exactly: for $f(x)=1$, the CLF gives $\sum_{n_2}|M|^2 = 1$ (Eq. 22) — the derived integral $I$ reduces identically to 1 using the WKB density of states, for both case I and case II.
• The analytical Pöschl-Teller model reproduces the CLF trend: with $U_i(x) = A/\sinh^2(\alpha x) - B_i/\cosh^2(\alpha x)$, bound-state energies $E_i = -4\epsilon(\nu_i - n_i)^2$, densities of states $g_i = 1/(4\sqrt{-\epsilon E_i})$, and $f(x) = 1/\cosh^{2t}(\alpha x)$, the CLF reduces to Eq. (27), an explicit function of $E_1$ and $E_2$ and the potential parameters.
• Convergence is observed for both operator choices: Figures 3 ($t=0$, i.e., $f(x)=1$) and 4 ($t=1$) show, across a three-panel progression toward the classical limit, that the computed data for $M$, $|M|^1$, and $|M|^2$ trend clearly toward the CLF boundary. These results are described as providing strong validation of the formula.
• Model parameters used in the comparison: the three panels share the ratio $p:\nu_1:\nu_2 = 10:7:5$; a single state of $U_1$ is used, taken as the state with energy closest to $-0.8 \times 4\epsilon\nu_1^2$; all bound states of $U_2$ are included; the energy unit $4\epsilon\nu_1^2$ approximates the depth of $U_1$.
• Not reported: the paper presents no numerical error bars, no tabulated benchmark values, and no quantitative convergence metrics; validation is shown graphically.

Methodology in Plain English

The author treats both potentials as slowly varying and both states as highly excited, so that each wave function can be written in the standard semiclassical (WKB) form: a slowly varying amplitude divided by the square root of the local wavenumber, times a rapidly oscillating cosine of a phase function. Multiplying two such wave functions together with the operator $f(x)$ produces two kinds of terms — one that oscillates very fast and washes out, and one that oscillates slowly only where the two local wavenumbers coincide. That special point is the COP. Expanding the phase difference to second order around the COP turns the remaining integral into a standard Fresnel-type integral with an exact closed-form answer, giving the CLF.

Because a single matrix element has an essentially arbitrary phase in the dense-spectrum limit, the author averages over states with nearby energies, which replaces the cosine factor by a $\Gamma$-function constant. To validate the result, the author chooses potentials of the Pöschl-Teller type that have exact analytical wave functions and exact matrix elements, evaluates those exactly using rational arithmetic in commercial software, and then plots the exact results against the CLF for progressively more semiclassical parameter sets. The paper also performs an independent exact check based on completeness of the states of one potential.

Why This Matters

Matrix elements between states of different potentials are the quantities behind Franck-Condon factors, transition strengths between electronic states, and — in the author's own program — transition strengths between motional states of atoms trapped in optical lattices. In the classical limit these states form an effectively continuous dense spectrum, and enumerating them individually is not feasible. A closed-form classical-limit expression gives a practical way to model such spectra.

• Optical lattice clocks: the paper explicitly anticipates using the formula to model inhomogeneous excitation spectra of atoms in one-dimensional optical lattices, with an eye toward improved clock performance. Population distributions over motional states are described as critical for state-of-the-art lattice clocks.
• Molecular spectroscopy: the potentials are envisioned as Born-Oppenheimer potential energy curves for different electronic states of a diatomic molecule, with $x$ the internuclear separation and the matrix elements being Franck-Condon factors.
• Lattice-trapped atom spectroscopy: the group has previously used the Born-Oppenheimer approximation to describe motional states of atoms in 1D optical lattices, with axial and radial degrees of freedom playing the roles of electronic and nuclear coordinates.
• Sudden changes of a confining potential: the paper notes that the two "distinct potentials" could instead be the "before" and "after" forms of a potential that changes abruptly in time.

Industry relevance: the most direct relevance is to precision timekeeping and frequency metrology (optical lattice clocks), and more broadly to modeling of spectra in atomic and molecular physics where dense bound-state manifolds make state-by-state calculation impractical.

Future Directions

• Application to lattice clocks: the author states an intent to use the formula to model inhomogeneous excitation spectra of atoms in one-dimensional optical lattices and thereby support improved optical lattice clock performance.
• Beyond the monotonic-difference assumption: the paper restricts the main derivation to potentials whose difference varies monotonically, but Appendix A indicates that essential aspects of the theory can be applied more generally.
• Testing with other potential and operator forms: the reported validation uses Pöschl-Teller potentials and $f(x) = 1/\cosh^{2t}(\alpha x)$ with $t=0$ and $t=1$; broader functional forms and operator types are natural extensions.
• Connecting to a specific physical system: the paper defers the details of the physical problem that partly motivated the choice of potentials to the Conclusion, leaving the explicit mapping onto a real experimental spectrum as an open next step.

Target Audience

Advanced graduate students and researchers in theoretical quantum physics, atomic and molecular physics, precision metrology, and optical lattice clock development. Readers should be comfortable with the WKB approximation, semiclassical wave functions, densities of states, and special-function results such as the Pöschl-Teller solutions. Experimentalists working on lattice clocks or molecular spectra will benefit from the qualitative message (the COP controls the transition strength) even if they skip the derivation.

Authors’ abstract

A classical-limit formula is derived for matrix elements between bound states of distinct one-dimensional potentials. We leave open the physical interpretation of the potentials, but generally envision them to be Born-Oppenheimer-type potentials with the position variable $x$ identifying with a slowly evolving degree of freedom of the system. For instance, $x$ could represent the internuclear separation in a diatomic molecule, with the potentials being potential energy curves for different electronic states. In this scenario, the matrix elements could be, e.g., conventional Franck-Condon factors. To test the derived formula, we assume functional forms for the potentials and operator that afford analytical solutions for the matrix elements. As the classical limit is approached, the computed matrix elements exhibit a clear tendency towards the classical-limit formula, providing strong validation for the formula. In future work, we anticipate using the formula to model inhomogeneous excitation spectra of atoms in one-dimensional optical lattices, with an eye towards improved optical lattice clock performance.