A nuclear model is a simplified picture of how protons and neutrons are arranged and behave inside the atomic nucleus. No single model captures everything, because the nucleus is a system of dozens or hundreds of particles interacting through forces that are fiendishly complicated to calculate exactly. Instead, physicists use several complementary models, each highlighting a different aspect of nuclear behavior. The liquid drop model treats the nucleus like a tiny charged droplet of fluid. The shell model treats it more like an atom, with protons and neutrons filling discrete energy levels. These and other models overlap, sometimes conflict, and together form a patchwork that has proven remarkably successful at explaining nuclear masses, shapes, decay rates, and reactions.
The Liquid Drop Model
The oldest and most intuitive nuclear model imagines the nucleus as a small drop of incompressible liquid. Just as water molecules in a droplet are held together by short-range forces and the droplet has a surface tension, nucleons (the collective name for protons and neutrons) are bound by the strong nuclear force and the nucleus has a surface energy cost. This analogy was developed in the 1930s and quickly proved useful for explaining why nuclear binding energy follows certain smooth trends as you move across the periodic table.
The model’s mathematical backbone is the semi-empirical mass formula, which adds up several energy contributions: a volume term (more nucleons means more binding), a surface term (nucleons on the surface are less tightly bound), a Coulomb term (protons repel each other electrically), and a symmetry term (nuclei with very unequal numbers of protons and neutrons are less stable). The liquid drop picture was also the first framework that could explain nuclear fission, because a charged droplet can be stretched until the electrical repulsion overcomes the surface tension and the drop splits in two.1International Journal of Mass Spectrometry. 80 Years of the liquid drop—50 years of the macroscopic–microscopic model
The formula works surprisingly well for predicting bulk nuclear properties. Recent work fitting an expanded version of the formula to all known nuclear masses found that the interplay between its different terms is significant: for instance, the fitted surface energy and symmetry energy coefficients shift substantially depending on which terms are included, with a ten-term version yielding surface and symmetry coefficients of about 26 and 33 MeV respectively, compared to roughly 17 and 23 MeV for a classic five-term formula.2Physica Scripta. Theoretical predictions of nuclear binding energy for the observed nuclei: the influence of coefficients and terms in a semi-empirical mass formula That sensitivity tells you the liquid drop picture is not a finished product; the terms compensate for each other, and how you carve up the physics matters.
The Shell Model
The liquid drop model’s smooth predictions miss something important: certain nuclei are far more tightly bound and stable than their neighbors. Nuclei with 2, 8, 20, 28, 50, 82, or 126 protons or neutrons (the so-called magic numbers) stand out as especially robust, much the way atoms with filled electron shells (the noble gases) are chemically inert. This pattern demanded a different model, one that treats nucleons as individual quantum particles moving in energy levels inside the nucleus.
In the shell model, each proton or neutron occupies a specific quantum state, and the magic numbers correspond to large gaps between groups of states. When a gap is filled, the nucleus gains extra stability. The model was developed in the late 1940s and won its creators the Nobel Prize. It explains not only the magic numbers but also nuclear spins, magnetic moments, and patterns in radioactive decay.
One of the most exciting developments in recent decades is the discovery that magic numbers can shift in exotic, neutron-rich nuclei far from the stable isotopes we encounter on Earth. A study of calcium-54 provided direct evidence for a sizable shell closure at neutron number 34, a “magic” gap that does not appear in stable nuclei.3PubMed. Evidence for a new nuclear ‘magic number’ from the level structure of 54Ca The implication is that the energy levels inside the nucleus rearrange themselves depending on the ratio of protons to neutrons, a phenomenon sometimes called shell evolution. The shell model’s magic numbers are not universal constants carved in stone; they are emergent features that depend on context.
Bridging the Two Pictures
The liquid drop model and the shell model seem almost contradictory. One treats the nucleus as a featureless blob; the other treats each nucleon as an independent particle in a quantum orbit. In practice, both contain truths. The macroscopic-microscopic approach, developed in the 1960s, literally adds the two together: you start with a liquid drop energy for the bulk shape and then add shell corrections that account for the quantum graininess. This hybrid has been one of the most successful tools in nuclear physics, accurately reproducing masses, fission barriers, and nuclear shapes across the periodic table.
A more formally unified treatment is the collective model, which describes how the whole nucleus can vibrate and rotate while still being built from individual nucleon orbits. The collective model interprets low-energy rotational bands (sequences of nuclear excited states with regularly spaced energies) and vibrational modes as coherent motions of many nucleons acting in concert. A microscopic derivation shows how these collective degrees of freedom emerge naturally from the shell model itself: the full shell model space factors into collective and intrinsic parts, so that rotational bands, vibrational excitations, and giant resonances all have a clear shell model interpretation.4Reports on Progress in Physics. Microscopic theory of the nuclear collective model In other words, the collective model is not a separate theory layered on top; it is the shell model viewed through a wide-angle lens.
Exotic Nuclei and Where Models Struggle
Standard nuclear models were built on data from stable or near-stable isotopes. Push far enough from stability, adding many extra neutrons for instance, and the models start to creak. Two phenomena illustrate this vividly: halo nuclei and islands of inversion.
A halo nucleus has one or two neutrons orbiting at an enormous distance from the rest of the nucleus, held on by the barest thread of binding energy. The neutron cloud extends so far outward that the nucleus can be physically larger than a much heavier stable isotope. Calculations using the Gamow shell model, which handles the coupling of loosely bound states to the continuum of unbound states, found that neon-29 is a strong candidate for a one-neutron halo, with its outermost neutron in a spatially extended wave orbit around a compact neon-28 core.5Physics Letters B. One-neutron halo structure of 29Ne Two-neutron halos exist as well: carbon-22, for example, can be described as a carbon-20 core surrounded by two weakly bound neutrons forming a diffuse cloud, with the neutron-neutron interaction alone strong enough to locate carbon-22 right on the neutron drip line, the edge beyond which neutrons can no longer be bound.6Physics Letters B. Two-neutron “halo” from the low-energy limit of neutron–neutron interaction: Applications to drip-line nuclei 22C and 24O
Islands of inversion are regions of the nuclear chart where the expected shell model ordering of energy levels is upended. Nucleons that “should” be in higher-energy orbits according to the standard magic numbers actually drop below the normal ordering, changing the nucleus’s shape and properties. Recent calculations suggest that two such islands, located around neutron numbers 40 and 50, actually merge in the chromium isotopic chain, forming a broad zone where standard shell closures lose their meaning.7Physics Letters B. Merging of the island of inversion at N = 40 and N = 50 These findings have forced theorists to develop more flexible shell model calculations that allow orbits to be mixed and rearranged rather than treated as fixed.
Clustering Inside Nuclei
Another phenomenon that no simple mean-field picture captures well is nuclear clustering: the tendency of nucleons inside certain nuclei to temporarily group themselves into alpha particles (clusters of two protons and two neutrons). The most famous example involves a specific excited state of carbon-12 known as the Hoyle state. This state sits just above the energy threshold where carbon-12 would break apart into three alpha particles, and it plays a critical role in how stars synthesize carbon from helium. Without it, the universe would contain almost no carbon.
The internal structure of the Hoyle state has been debated for decades. Electron scattering data compared with several theoretical models, including a microscopic alpha-cluster model, indicate that the Hoyle state has a remarkably dilute density and a large spatial extent, consistent with it being an almost gaseous arrangement of three alpha clusters.8PubMed. Structure of the Hoyle state in 12C Some researchers have even described it as resembling a Bose-Einstein condensate of alpha particles. Whether or not that analogy holds exactly, the Hoyle state is a reminder that nucleons inside a nucleus can organize themselves in ways that neither a simple liquid drop nor a straightforward shell model would predict.
Superheavy Elements and the Island of Stability
At the heavy end of the periodic table, shell effects become a matter of existence rather than just extra stability. Elements beyond about atomic number 104 owe their very survival to shell corrections. Without the quantum structure predicted by the shell model, the Coulomb repulsion among their many protons would tear these nuclei apart almost instantly through fission. The idea that there should be an “island of stability,” a cluster of superheavy nuclei with lifetimes dramatically longer than their neighbors, has driven experimental programs for decades.
The predicted center of this island lies near proton number 114 and neutron number 184, corresponding to the next set of expected magic shell closures beyond lead-208 (which has 82 protons and 126 neutrons). Experimental progress in the 1990s extended the known elements and collected data on fission transitions in the region around the doubly magic proton and neutron numbers, confirming that shell stabilization is real in this regime.9Journal of Physics G. Stability of heavy and superheavy elements Several superheavy elements, including flerovium (element 114) and oganesson (element 118), have since been synthesized. Their half-lives, while still very short by everyday standards, are orders of magnitude longer than naive liquid drop calculations would predict, confirming the shell model’s stabilizing influence.
The Neutron Skin and Its Cosmic Implications
One of the most striking connections between nuclear models and the wider universe runs through a measurement you can do on a single nucleus in the lab: the neutron skin thickness. In a heavy nucleus like lead-208, the neutrons extend slightly further from the center than the protons, forming a thin neutron-rich “skin.” The thickness of that skin depends on the symmetry energy, a term in the nuclear energy formula that quantifies how the energy cost changes as you shift the balance between protons and neutrons.
The PREX-II experiment at Jefferson Lab measured this skin thickness using parity-violating electron scattering, a technique in which the weak nuclear force (which interacts differently with neutrons than with protons) is used to map the neutron distribution. The result was a neutron skin thickness of about 0.28 femtometers, with an uncertainty of roughly 0.07 femtometers.10PubMed. Accurate Determination of the Neutron Skin Thickness of 208Pb through Parity-Violation in Electron Scattering That number matters far beyond the lab, because the same symmetry energy that sets the neutron skin thickness in lead also governs the properties of neutron stars, which are essentially giant nuclei held together by gravity. A thicker neutron skin in lead implies a stiffer symmetry energy, which in turn predicts larger neutron star radii.11Nuclear Physics A. The role of nuclear symmetry energy and neutron skin thickness of 208Pb in controlling the underlying physics of neutron star This is one of the clearest cases where a nuclear model parameter measured on a tabletop-scale target constrains the structure of objects ten kilometers across.
Nuclear Density Functional Theory
For heavier nuclei, solving the full shell model exactly is computationally impossible: the number of possible configurations of nucleons grows astronomically. Nuclear density functional theory (DFT) sidesteps this by working with the average density of nuclear matter rather than tracking every single nucleon. It is the workhorse of modern large-scale nuclear structure calculations, capable of predicting masses, radii, and deformation shapes across the entire nuclear chart.
One active frontier is connecting DFT to more fundamental calculations rooted in the actual forces between nucleons. Recent work has fitted relativistic mean-field DFT models to predictions from chiral effective field theory, a modern framework for deriving nuclear forces from the underlying theory of quarks and gluons. The results showed that model extensions beyond the simplest version are important for capturing the physics, particularly for charge radii and neutron skin thicknesses of closed-shell nuclei.12Physical Review C. Connecting relativistic density functional theory to microscopic calculations This kind of benchmarking is essential for building confidence that DFT predictions for unmeasured nuclei, especially the very neutron-rich ones produced in stellar explosions, are trustworthy.
Symmetry as a Guiding Principle
Nuclear models lean heavily on symmetries. One of the most useful is isospin symmetry, the idea that the strong force treats protons and neutrons almost identically. If you swap every proton for a neutron and vice versa in a nucleus (creating its “mirror” partner), the resulting energy levels should look nearly the same. This approximate symmetry lets physicists predict properties of hard-to-study nuclei from measurements of their mirror partners.
The symmetry is not perfect, though, because protons carry electric charge and neutrons do not. In most nuclei the breaking is small and predictable. But occasionally it is dramatic enough to be genuinely surprising. A recent study of the mirror pair bromine-71 and krypton-71 found evidence that their ground states have different quantum spins, a situation that violates the most basic expectation of mirror symmetry.13Physical Review Letters. Isospin Symmetry Breaking in the ^{71}Kr and ^{71}Br Mirror System Cases like this are valuable precisely because they are rare: they highlight the limits of the symmetry and force models to account for electromagnetic and other effects more carefully.
Another test of nuclear models comes from electromagnetic transitions, the gamma rays nuclei emit when they drop from one excited state to another. The rates of these transitions are sensitive to details of the nuclear wave function, making them a sharp diagnostic. Systematic comparisons of measured transition rates in deformed nuclei with theoretical estimates have revealed consistent deviations from simple single-particle predictions, signaling that collective effects and configuration mixing play important roles even in states that look like they should be dominated by a single nucleon.14Nuclear Physics. Systematics of absolute gamma-ray transition probabilities in deformed odd-mass nuclei
Phase Transitions Under Extreme Conditions
Nuclear models also extend into territory where nuclei themselves dissolve. In heavy-ion collisions at facilities like CERN and Brookhaven, nuclei are smashed together at close to the speed of light, producing temperatures and densities so extreme that protons and neutrons lose their individual identities. The quarks and gluons that make up nucleons are liberated into a new state of matter called a quark-gluon plasma.
Modeling this transition requires connecting a description of ordinary nuclear matter (often based on relativistic mean-field theory) to a description of the quark-gluon phase, with a first-order phase transition between them.15Nuclear Physics A. Relativistic hydrodynamics for heavy-ion collisions. II. Compression of nuclear matter and the phase transition to the quark-gluon plasma The details of this transition, its temperature, density, and order, depend on the nuclear model used for the hadronic side. Getting it right matters for understanding what happened microseconds after the Big Bang, when the entire universe was a quark-gluon plasma cooling into the protons and neutrons that eventually formed atoms.
How Nuclear Properties Are Actually Measured
None of these models would be useful without experimental data to test them, and measuring the properties of short-lived nuclei is an extraordinary technical challenge. One of the most precise tools available is Penning trap mass spectrometry, which confines a single ion in a combination of electric and magnetic fields and measures its mass by monitoring how fast it orbits. The mass of a nucleus encodes information about all the internal interactions between its constituents, so a precise mass measurement is a direct window into binding energy without needing to know anything about the nucleus’s internal level structure.16Nuclear Physics A. Direct mass measurements of the heaviest elements with Penning traps This technique has been extended to superheavy elements and very short-lived isotopes, providing the raw data that every nuclear model must ultimately confront.
Electron scattering, meanwhile, maps out how charge and weak charge are distributed inside nuclei. Because electrons interact through the well-understood electromagnetic and weak forces rather than the complicated strong force, they act as clean probes. Combining charge-radius measurements from electron scattering with mass measurements from Penning traps gives theorists two independent constraints that a successful model must satisfy simultaneously. When a model passes both tests across hundreds of nuclei, physicists gain some confidence in its extrapolations to regions of the nuclear chart that cannot yet be reached experimentally.