⚛️ nuclear experiments

Zemach radii and nuclear structure effects in hyperfine splitting of lithium isotopes

This paper resolves the long-standing discrepancy between effective and elastic Zemach radii in lithium isotopes by demonstrating that nuclear polarizability effects, which are negligible in $^7$ Li but dominant in $^6$ Li, constitute the primary source of isotope-dependent nuclear corrections to hyperfine splitting in light atoms.

Original authors: Yilong Yang, Evgeny Epelbaum, Chen Ji, Pengwei Zhao

Published 2026-10-20

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Yilong Yang, Evgeny Epelbaum, Chen Ji, Pengwei Zhao

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Atoms are the fundamental building blocks of the visible world, but they are not simple, solid spheres. Inside every atom lies a dense core called the nucleus, surrounded by a cloud of electrons. When scientists look closely at the energy levels of these electrons, they find that the lines are not single, sharp marks but are split into tiny, distinct components. This phenomenon, known as hyperfine splitting, acts like a high-precision fingerprint of the atom. It is driven by the magnetic interaction between the spinning electron and the spinning nucleus. Because this interaction is so sensitive to the internal structure of the nucleus, measuring these splits allows physicists to test the laws of quantum mechanics with extreme rigor and to probe the arrangement of protons and neutrons inside the nucleus. However, for decades, a stubborn puzzle has remained in the measurements of lithium atoms. Theoretical predictions based on the size of the nucleus did not match the values observed in experiments, leaving a gap in our understanding of how these tiny particles behave.

A team of researchers has now solved this long-standing discrepancy by looking at the nucleus not as a static object, but as a dynamic one that can wobble and shift. In their study, they focused on two specific versions of lithium, known as isotopes, which differ only in the number of neutrons they contain. One isotope, lithium-6, showed a large mismatch between theory and experiment, while the other, lithium-7, did not. The researchers realized that the standard way of calculating the nuclear size ignored a crucial factor: the nucleus is not a rigid ball. When an electron passes close to it, the electric and magnetic fields can momentarily distort the shape of the nucleus, a process called nuclear polarizability. This distortion creates a temporary change in the magnetic environment that the electron feels, effectively altering the measured size of the nucleus.

To investigate this, the scientists developed a new and powerful method to simulate the behavior of the nucleus from first principles. They used a technique called variational Monte Carlo, which involves creating a vast number of possible configurations for the protons and neutrons to find the one that represents the true ground state of the atom. To make this calculation accurate enough for heavy nuclei like lithium, they introduced a novel neural-network wave function. This is a type of computer algorithm that learns the complex relationships between the particles, allowing it to handle the intricate dance of six or seven nucleons without the simplifications that usually force errors. By running these simulations, they were able to calculate the magnetic and charge distributions of the lithium nuclei with unprecedented precision.

The results confirmed that the shape of the nucleus is not the source of the error. When the researchers compared their realistic, complex density distributions with the simpler, smooth shapes often assumed in previous work, they found the difference was negligible. The real culprit was the nuclear polarizability. In the case of lithium-7, this effect was so small it could be ignored. However, in lithium-6, the effect was dominant. The nucleus of lithium-6 is particularly susceptible to these distortions because it has an unpaired proton and an unpaired neutron, making it much more flexible than its neighbor. The team found that this flexibility creates a significant shift in the effective size of the nucleus, exactly enough to close the gap between the theoretical prediction and the experimental measurement.

The study also revealed a deeper pattern that connects different atomic nuclei. The researchers showed that the strength of this polarizability effect follows a universal rule based on the symmetry of nuclear forces. They found that nuclei with both an odd number of protons and an odd number of neutrons, like lithium-6, experience a much stronger effect than those with an even number of one type. This enhancement is roughly three times larger than in other configurations. This same mechanism explains similar discrepancies observed in other light atoms, such as deuterium and helium-3. By accounting for these dynamic shifts, the researchers have provided a unified explanation for why the effective size of a nucleus appears different depending on how it is measured. Their work establishes that nuclear polarizability is the primary source of these isotope-dependent corrections, turning a confusing anomaly into a clear, predictable feature of atomic structure.

Technical Summary: Zemach Radii and Nuclear Structure Effects in Hyperfine Splitting of Lithium Isotopes

Problem Statement
High-precision atomic spectroscopy provides a sensitive probe of nuclear magnetization structure through hyperfine splitting (HFS). The nuclear structure effect on HFS is driven by the two-photon exchange (TPE) between the nucleus and the lepton, characterized by the effective Zemach radius ( $r_Z^{\text{eff}}$ ). While the elastic component of this effect is described by the elastic Zemach radius ( $r_Z$ ), derived from nuclear electromagnetic densities, a significant theoretical discrepancy exists between $r_Z^{\text{eff}}$ and $r_Z$ in light nuclei.

Specifically, in $^6$ Li, $r_Z^{\text{eff}}$ extracted from HFS measurements yields a much smaller value than $r_Z$ derived from experimental root-mean-square (rms) charge and magnetic radii. Additionally, the $r_Z^{\text{eff}}$ value for $^6$ Li is about 40% smaller than that of the stable isotope $^7$ Li, reversing the trend observed in their charge radii. Previous explanations suggested this might arise from non-Gaussian nuclear density shapes or inelastic TPE effects (nuclear polarizability). However, nuclear polarizability had only been calculated for nuclei with mass number $A \le 3$ , leaving lithium isotopes unaddressed due to the computational challenge of modeling six- and seven-body wave functions ab initio.

Methodology
The authors employ a novel ab initio framework combining deep-learning variational Monte Carlo (VMC) with chiral effective field theory (EFT) interactions to resolve these discrepancies.

Neural-Network Wave Functions: To overcome the limitations of conventional Jastrow-type wave functions (which struggle with larger $A$ ), the authors utilize a deep-learning VMC (VMC-NN) approach. They introduce a novel neural-network wave function ansatz comprising a sequence of correlators incorporating two- and three-body correlations, alongside an antisymmetric part controlling quantum numbers and long-range behavior. This ansatz is optimized using stochastic reconfiguration to minimize the energy expectation value.
Nuclear Interactions: Calculations are performed using local chiral N2LO(1.0) and N2LO(1.2) interactions, allowing for the estimation of chiral-truncation uncertainties. Phenomenological forces (AV8'+UIX') are also used for benchmarking.
Elastic Zemach Radii ( $r_Z$ ): Using the optimized ground-state wave functions, the authors compute charge ( $\rho_E$ ) and magnetic ( $\rho_M$ ) density distributions. The elastic Zemach radius is calculated via the integral $r_Z = \int d^3r d^3r' \rho_E(r)\rho_M(r')|r' - r|$ .
Nuclear Polarizability ( $\delta r_{\text{pol}}$ ): The authors derive a formalism for nuclear polarizability contributions to the effective Zemach radius without taking the $\omega_N \to 0$ limit (which overestimates effects in light nuclei). They employ the closure approximation, replacing the excitation energy $\omega_N$ with a state-independent average $\bar{\omega}$ , reducing the calculation to a ground-state observable. The polarizability contribution is expressed as a sum over single-nucleon magnetic moments and charge densities.
Symmetry Analysis: The relative magnitudes of polarizability contributions are analyzed using the approximate SU(4) symmetry of nuclear forces to establish systematic trends across different nuclei.

Key Results

Benchmarking: The VMC-NN method yields ground-state energies for $^4$ He and $^6$ Li that are significantly lower and more accurate than conventional Jastrow-based VMC, reaching the accuracy of Green's function Monte Carlo (GFMC) and resolving the unphysical breakup of $^6$ Li into $^4$ He + $^2$ H subclusters.
Density Shapes: The study confirms that the deviation between $r_Z^{\text{eff}}$ and $r_Z$ in $^6$ Li is not caused by non-Gaussian nuclear density shapes. The elastic Zemach radii calculated from realistic VMC-NN densities match those estimated from Gaussian distributions fixed by rms radii.
Resolution of the $^6$ Li Discrepancy: The calculations reveal that nuclear polarizability effects are negligible in $^7$ Li but dominate in $^6$ Li. When the polarizability contribution ( $\delta r_{\text{pol}}$ ) is included, the theoretical $r_Z^{\text{eff}}$ aligns with experimental HFS data.
Universal Scaling: The authors demonstrate that the ratio of nuclear polarizability contributions between different nuclei is universal in the limit of closure and SU(4) symmetry. Specifically, the contribution in odd-odd nuclei (like $^6$ Li and $^2$ H) is enhanced by a factor of $\mu_p/(\mu_p + \mu_n) \approx 3$ compared to odd- $A$ isotopes (like $^7$ Li and $^3$ He).
Quantitative Agreement: By adjusting the closure energy $\bar{\omega}$ to experimental data ( $\bar{\omega} \approx 23$ MeV for $^6$ Li and $\approx 29$ MeV for $^7$ Li), the unified framework successfully reproduces the observed $r_Z^{\text{eff}} - r_Z$ differences for $^2$ H, $^3$ He, $^6$ Li, and $^7$ Li.

Significance and Claims
The paper establishes nuclear polarizability as the dominant source of isotope-dependent nuclear corrections to hyperfine splitting in light atoms. It resolves the long-standing discrepancy between effective and elastic Zemach radii in $^6$ Li, which was previously unexplained.

The authors claim that their findings provide a unified explanation for similar anomalies observed in $^2$ H and $^3$ He. By leveraging the approximate SU(4) symmetry, they offer a predictive framework for nuclear polarizability effects, suggesting that odd-odd nuclei will consistently exhibit enhanced corrections compared to their odd- $A$ counterparts. This work advances the understanding of nuclear structure effects in light atoms and provides testable benchmarks for future studies of other light nuclei, such as $^7$ Be and $^9$ Be. The study also highlights the capability of neural-network wave functions to solve high-precision few-body nuclear problems with chiral interactions.

Technical Summary: Zemach Radii and Nuclear Structure Effects in Hyperfine Splitting of Lithium Isotopes

More like this