⚛️ nuclear theory

Extracting nuclear charge radii from binding energies: a single-parameter empirical formula with structural corrections

This paper proposes a novel one-parameter empirical formula, BECR $_{1p}$ , which leverages binding energy correlations and structural corrections to accurately model nuclear charge radii, achieving a root-mean-square deviation of 0.0138 fm on experimental data and enabling predictions for over 11,000 nuclei.

Original authors: Pengfei Ma, Minghui Hu, Kai Ren, Junlong Tian, Cheng Li

Published 2026-07-17

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Pengfei Ma, Minghui Hu, Kai Ren, Junlong Tian, Cheng Li

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

The Atomic Heartbeat: Why Size and Strength Are Best Friends

Imagine the universe as a giant, bustling construction site. In the center of this site are the atoms, the tiny bricks that make up everything from your sneakers to the stars. But what holds these bricks together? Deep inside the atom's core, the nucleus, protons and neutrons are huddled in a tight, chaotic dance. They are held together by a powerful force, but they also push against each other. The measure of how tightly they are glued together is called binding energy. Think of it as the "hug strength" of the nucleus; the stronger the hug, the more stable the atom.

Now, imagine you want to know how big this nuclear dance floor is. That measurement is the charge radius. It tells us the average distance the protons (the positively charged dancers) are from the center. For a long time, scientists treated these two measurements—how strong the hug is (binding energy) and how big the dance floor is (radius)—as separate puzzles. They thought you needed complex, heavy-duty machinery to figure out the size of the nucleus, often relying on expensive experiments or super-computer simulations. But what if the size of the dance floor was actually written in the strength of the hug all along? This is the question a team of physicists from Guangxi Normal University decided to tackle. They asked a simple, playful question: If we know how tightly the nucleus is holding on, can we guess how big it is without needing a microscope?

The "One-Parameter" Magic Trick

In their new study, the researchers proposed a clever shortcut. They realized that binding energy and charge radius are like two sides of the same coin. If you know one, you have a huge clue about the other. They built a new formula, which they nicknamed BECR1p (Binding-Energy-Correlated Radius, 1-parameter).

Think of the nucleus as a balloon. Usually, to guess how big a balloon is, you might just look at how much air is inside (the number of particles). But this new formula says, "Wait, let's look at how tight the rubber is stretched!" The "tightness" is the binding energy. The team started with a basic rule: the stronger the binding energy per particle, the more the nucleus squeezes itself together. This gave them a rough estimate of the size, which they called the "macroscopic" part. On its own, this rough guess was actually pretty good, missing the true size by an average of 0.0345 fm (a femtometer is one-quadrillionth of a meter, so incredibly tiny).

But nuclei are messy. They aren't perfect spheres; they have bumps, wobbles, and special "magic" numbers of particles that make them extra stable. To fix the rough guess, the team added "structural corrections." These are like adding little adjustments for:

Shells: Like layers of an onion, where certain layers make the nucleus act differently.
Odd-Even Staggering: Whether the nucleus has an even or odd number of partners affects its shape.
Deformation: Sometimes the nucleus squishes into an egg shape instead of a ball.

When they added these corrections to their binding-energy guess, the formula became incredibly sharp. The error dropped from 0.0345 fm down to just 0.0138 fm. To put that in perspective, this new formula is more accurate than other famous, complex formulas that use many more numbers to try to solve the same puzzle.

The "Residual" Detective Work

Even with the best formula, there are still tiny wiggles left over. The authors didn't just stop there; they used a smart computer tool called anisotropic kernel ridge regression (AKRR) to act as a detective. Imagine you've solved a mystery, but there are still a few tiny footprints left on the floor. This tool looks at those remaining footprints (the "residuals") to see if there's a pattern the main formula missed.

By training this tool on the data, they could predict those tiny leftover errors. When they added this "detective work" to their formula, the accuracy got even better, with the error dropping to about 0.0081 fm. However, the authors are careful to note that this super-accurate version works best when you are looking at nuclei similar to the ones they already studied. If you try to use it on a brand-new, strange nucleus far away from the known data, the main formula (the one with just the binding energy) is the safer bet.

What They Found and What They Didn't

The team tested their formula on 893 different nuclei that had been measured in experiments. They found that their simple, one-parameter formula could reproduce the "kinks" and "jumps" in size that happen when nuclei hit special "magic" numbers or change shape. For example, they could correctly predict how the size of a nucleus changes as you add more neutrons, even in tricky regions where the nucleus suddenly stretches out.

They also tried their formula on nine chains of nuclei that they didn't use to build the formula in the first place. This is like taking a map you drew for one city and seeing if it works for a neighboring city. The formula worked well enough to show the general trends, but it struggled with some specific, systematic errors in heavy elements like Terbium or Radium. This suggests that while the link between binding energy and size is strong, there are still some complex, hidden structures in those heavy elements that a simple formula can't fully capture yet.

Finally, the team used their formula to predict the sizes of 11,205 nuclei across the entire chart of elements, many of which have never been measured. They provide these predictions as a resource for other scientists.

The Bottom Line

This paper doesn't claim to have replaced the super-computers or the deep theories of nuclear physics. Instead, it offers a surprisingly simple, transparent way to understand the size of an atom's core by listening to its "hug strength." The authors show that you don't always need a sledgehammer to crack a nut; sometimes, a single, well-chosen number derived from binding energy is enough to get you very close to the truth. While the formula isn't perfect for every single exotic nucleus in the universe, it proves that the size of a nucleus and the energy holding it together are deeply, intimately connected, offering a new, playful lens through which to view the atomic world.

Technical Summary: Extracting Nuclear Charge Radii from Binding Energies

Problem Statement
The nuclear charge radius ( $r_{ch}$ ) is a fundamental observable characterizing the spatial distribution of proton charge, while the binding energy per nucleon ( $\epsilon = B/A$ ) quantifies the average binding strength of a finite nucleus. Although these quantities describe the same many-body system from complementary perspectives (spatial extension vs. energetic binding), empirical radius formulas typically rely on geometric scaling ( $A^{1/3}$ ) with ad-hoc corrections, while microscopic models (e.g., HFB, RMF, DFT) depend heavily on specific effective interactions and are computationally intensive. Furthermore, experimental charge radius data is sparse compared to nuclear mass data, particularly in neutron-rich and proton-rich regions. The authors posit that since $\epsilon$ encodes information on saturation, isospin, shell structure, and deformation, it should serve as a physically motivated input to extract the charge-radius scale, offering a new empirical route to radius estimation in data-scarce regions.

Methodology
The authors propose a one-parameter empirical formula, BECR1p, which derives the leading charge-radius scale directly from the binding energy per nucleon ( $\epsilon$ ) rather than solely from mass number scaling. The methodology proceeds in three stages:

Macroscopic Term (BECR $_{mac}$ ): A phenomenological term is constructed based on a volume–surface coupling picture. The effective length scale is derived as proportional to $(N/\epsilon)^{1/5}$ , where $N$ is the neutron number. An isospin correction proportional to $- \frac{\sqrt{3}}{2} \frac{N}{Z}$ is subtracted to account for the fact that $r_{ch}$ probes protons rather than total matter. This macroscopic term alone yields a root-mean-square deviation (RMSD) of 0.0345 fm on 893 experimental radii.
Local Structural Corrections: To recover local structural details smoothed out by the macroscopic term, the authors introduce corrections for:
- Shell and Odd–Even Effects: Modeled via a Casten-like factor ( $e_{sh}$ ) based on proximity to magic numbers and a pairing factor ( $e_{pair}$ ), scaled by $A^{-5/6}$ .
- Finite-Size Effects: A $1/N$ term, crucial for light nuclei.
- Deformation: A multiplicative factor involving deformation parameters $\beta_2$ and $\beta_4$ (taken from the WS3.3 mass table).
Parameter Reduction: The full formula initially contains four fitting parameters ( $a_1, a_2, c_0, r_\beta$ ). Through global optimization, the authors identify strong empirical correlations between these parameters, allowing them to be expressed in terms of a single effective parameter, $\kappa$ . This reduces the formula to the BECR1p form.
Residual Correction (AKRR): To probe remaining systematic information in the residuals (e.g., higher-order shell effects or shape coexistence), the authors apply an Anisotropic Kernel Ridge Regression (AKRR) as a post-processing diagnostic tool. This is not a standalone predictor but a correction applied to the BECR1p residuals using $(N, Z)$ as inputs.

Key Contributions and Results

Single-Parameter Accuracy: The BECR1p formula, containing only one fitted parameter ( $\kappa = 0.7190$ ), achieves an RMSD of 0.0138 fm on a curated set of 893 experimental charge radii. This outperforms representative empirical formulas such as WS3.3-R (0.0170 fm) and Sheng (0.0247 fm).
Structural Reproduction: The formula successfully reproduces complex local features across the nuclear chart, including:
- Shell kinks at magic numbers (e.g., Ca, Pb).
- Odd–even staggering.
- Sharp radius jumps associated with shape transitions (e.g., the Sr chain near $N \approx 60$ ).
- Trends in deformed regions (Cs, Gd).
Predictive Capability: Using the formula, the authors predict charge radii for 11,205 nuclei across the nuclear chart.
Residual Analysis: Applying AKRR to the BECR1p residuals further reduces the leave-one-out cross-validation (LOOCV) RMSD to 0.0081 fm, with 98.4% of points falling within $\pm 0.02$ fm. The anisotropy in the kernel ( $\sigma_1 > \sigma_2$ ) indicates that residual correlations extend more smoothly along the neutron number ( $N$ ) than the proton number ( $Z$ ).
Limitations: When tested on nine isotopic chains excluded from the fitting set (Tb, Tm, Lu, Re, Po, Rn, Fr, Ra, Cm), the model reproduces general trends but exhibits systematic offsets (RMSD $\approx$ 0.04 fm). These offsets are attributed to experimental normalization ambiguities in these chains and complex theoretical structures (shape coexistence, octupole deformation) that global empirical formulas struggle to capture.

Significance and Claims
The paper claims that binding energy per nucleon ( $\epsilon$ ) contains intrinsic structural information sufficient to establish the leading scale of nuclear charge radii. By leveraging this correlation, the BECR1p formula offers a complementary, physically transparent approach to charge-radius modeling that does not rely on the complex effective interactions required by microscopic models.

The authors emphasize modesty regarding the method's scope:

BECR1p is presented as an empirical tool that deepens the understanding of the binding–radius correlation, not as a replacement for microscopic models.
For extrapolation to drip-line nuclei or exotic regions far from stability, the authors state that microscopic models (like HFB-29) remain the "more physically robust guide."
The AKRR correction is explicitly framed as a "residual analyser" useful for interpolation within known regions; its application to sparse-data areas is cautioned against due to reliance on nearby training coverage.

In summary, the work demonstrates that a single-parameter formula grounded in binding-energy correlations can achieve high accuracy in reproducing charge radii, providing a viable baseline for estimates in regions where experimental data is lacking, while acknowledging the limits of global systematics in the face of complex nuclear structures.