Chapter 7: Beyond the Single Particle — Residual Interactions and Nuclear Correlations

"The shell model is the beginning of nuclear structure, not the end. The real nucleus is far richer — and far more difficult — than any single-particle picture can capture."

Introduction

In Chapter 6, we achieved a remarkable success: the independent-particle shell model, with a mean-field potential plus spin-orbit coupling, explained the magic numbers and correctly predicted the ground-state spins and parities of most odd-$A$ nuclei. That was a triumph of the single-particle picture — the idea that each nucleon moves independently in a smooth potential generated by all the others.

But the triumph has limits, and an honest treatment of nuclear structure must confront them. Consider the following experimental facts that the simple shell model cannot explain:

1. Even-even nuclei always have $J^\pi = 0^+$ ground states. Every single one of the roughly 800 known even-even nuclei has zero angular momentum in its ground state. The independent-particle model offers no reason why nucleons should systematically pair off to produce $J = 0$.

2. Even-even nuclei are systematically more tightly bound than their odd-$A$ neighbors. The binding energy shows a striking odd-even staggering — about 1-2 MeV for heavy nuclei — that we parameterized in Chapter 4 with the pairing term $\delta$ in the semi-empirical mass formula. But we never explained why pairing occurs.

3. Low-lying spectra of near-closed-shell nuclei show energy patterns that cannot be reproduced by any single-particle filling. Two nucleons outside a closed shell do not simply occupy the lowest available orbits — they interact, and that interaction profoundly reshapes the spectrum.

4. Many nuclei are deformed, departing dramatically from the spherical symmetry assumed in the shell model. The rare-earth and actinide nuclei have quadrupole deformations that demand a fundamentally different single-particle basis.

5. Some excited states live for extraordinarily long times — microseconds, hours, even billions of years — creating nuclear isomers whose existence reflects the interplay of angular momentum selection rules and nuclear structure.

This chapter takes us beyond the single-particle picture. We will introduce the residual interaction — the part of the nucleon-nucleon interaction not absorbed into the mean field — and see how it gives rise to pairing correlations, configuration mixing, and the rich spectroscopy of real nuclei. We will then develop the Nilsson model, which extends the shell model to deformed potentials. The common thread is that the true nuclear state is never a simple independent-particle configuration; it is always a superposition, shaped by correlations that the mean field cannot capture.

7.1 The Pairing Interaction

7.1.1 Experimental Evidence for Pairing

The most striking evidence for pairing comes from the systematic behavior of nuclear binding energies. Recall from Chapter 4 that the semi-empirical mass formula contains a pairing term:

$$\delta = \begin{cases} +\delta_0 & \text{even-even} \\ 0 & \text{odd-}A \\ -\delta_0 & \text{odd-odd} \end{cases}$$

where $\delta_0 \approx 12/\sqrt{A}$ MeV. This term was introduced purely empirically. Now we must understand its origin.

The evidence goes deeper than binding energies:

• First excited $2^+$ states in even-even nuclei lie at high excitation energies (typically 0.5-2 MeV for medium and heavy nuclei), indicating a substantial energy gap above the correlated ground state.
• Moment of inertia anomaly: The moments of inertia of deformed nuclei are roughly half the rigid-body value, suggesting that not all nucleons participate in collective rotation — the paired nucleons form a superfluid condensate.
• Odd-even staggering of separation energies: The neutron separation energy $S_n$ shows a pronounced zigzag pattern as a function of $N$, dropping sharply when $N$ goes from even to odd.
• Two-nucleon transfer reactions such as $(t, p)$ and $(p, t)$ show strongly enhanced cross sections, indicating that nucleons move in correlated pairs.

7.1.2 Origin of the Pairing Force

The pairing interaction arises from the short-range, attractive nature of the nucleon-nucleon force, discussed in Chapter 3. When two identical nucleons (say, two neutrons) occupy time-reversed orbits — the same single-particle level with opposite magnetic quantum numbers, $(n\ell j, m)$ and $(n\ell j, -m)$ — they maximize their spatial overlap, and the short-range attraction is strongest.

More precisely, consider two identical nucleons in a single-$j$ shell interacting through a short-range (delta function) force $V(\mathbf{r}_1, \mathbf{r}_2) = -V_0 \delta(\mathbf{r}_1 - \mathbf{r}_2)$. The two-body matrix elements of this interaction in states of total angular momentum $J$ are:

$$\langle j^2; J | V | j^2; J \rangle = -V_0 \int |\psi_{n\ell j}(\mathbf{r})|^4 \, d^3r \times (2j+1) \begin{pmatrix} j & j & J \\ \frac{1}{2} & -\frac{1}{2} & 0 \end{pmatrix}^2$$

where the $3j$-symbol encodes the angular momentum coupling. For identical fermions in the same $j$-shell, the Pauli principle restricts $J$ to even values: $J = 0, 2, 4, \ldots, 2j-1$.

The key result — which the reader should verify using the properties of $3j$-symbols (reviewed in Chapter 5) — is that the $J = 0$ matrix element is by far the most attractive:

$$\langle j^2; J=0 | V | j^2; J=0 \rangle = -V_0 \frac{(2j+1)}{4\pi} \int |R_{n\ell}(r)|^4 r^2 \, dr$$

while the matrix elements for $J > 0$ are much smaller in magnitude. For a typical case with $j = 7/2$, the ratio of the $J = 0$ matrix element to the next most attractive ($J = 2$) is roughly 4:1.

This enormous preference for $J = 0$ coupling is the microscopic origin of pairing. Two nucleons in the same orbit gain the most binding energy by coupling their angular momenta to $J = 0$, and this energetic preference drives the pairing correlations that pervade nuclear structure.

The physics can be understood intuitively: a $J = 0$ pair has the two nucleons moving in exactly the same spatial orbit but with opposite angular momentum projections. This maximizes their spatial overlap, and since the nuclear force is short-range and attractive, maximum overlap means maximum binding. Any other coupling ($J > 0$) has the nucleons moving in different orientations, reducing their overlap and hence their interaction energy. The effect is analogous to the Cooper pair in superconductivity, where two electrons near the Fermi surface form a bound state through phonon-mediated attraction — except that in nuclei, the pairing is direct (no need for an intermediary like phonons) and the pairs are bound by the bare nucleon-nucleon force itself.

7.1.3 The Pairing Gap

We define the pairing gap $\Delta$ as the energy required to break a pair. Experimentally, it can be extracted from binding energies using the three-point or five-point odd-even mass difference formulas. The simplest is the three-point formula for neutrons:

$$\Delta^{(3)}(N) = \frac{(-1)^N}{2} \left[ B(Z, N-1) - 2B(Z, N) + B(Z, N+1) \right]$$

where $B(Z, N)$ is the binding energy. This quantity oscillates with $N$: it is positive for even $N$ (reflecting the cost of breaking a pair) and near zero or negative for odd $N$.

A cleaner extraction uses the five-point formula:

$$\Delta^{(5)}(N) = \frac{(-1)^N}{8} \left[ B(N-2) - 4B(N-1) + 6B(N) - 4B(N+1) + B(N+2) \right]$$

which removes the smooth $N$-dependence to isolate the staggering.

The empirical systematics yield:

$$\Delta \approx \frac{12}{\sqrt{A}} \text{ MeV}$$

For example: - $^{120}$Sn ($A = 120$): $\Delta \approx 1.1$ MeV (experimental: $\approx 1.2$ MeV) - $^{208}$Pb ($A = 208$): $\Delta \approx 0.83$ MeV (experimental: $\approx 0.9$ MeV) - $^{56}$Fe ($A = 56$): $\Delta \approx 1.6$ MeV (experimental: $\approx 1.5$ MeV)

The pairing gap decreases with mass number because the single-particle level density increases with $A$, diluting the pairing effect. The formula $12/\sqrt{A}$ is remarkably robust across the nuclear chart, though deviations occur near shell closures where level densities change abruptly.

The physical interpretation of the pairing gap is straightforward: $\Delta$ is the minimum energy required to break a Cooper pair and create two independent quasiparticle excitations. In an even-even nucleus, the ground state has all nucleons paired, and any excitation must first overcome the gap $2\Delta$ (breaking one pair produces two unpaired quasiparticles, each costing energy $\Delta$). This is why the first excited states of even-even nuclei lie at relatively high energies — typically 0.5 to 2 MeV in medium and heavy nuclei — compared to the dense low-lying spectra of odd-$A$ nuclei, where one quasiparticle is already present and excitations only require rearranging the unpaired nucleon.

The separation energy staggering provides a vivid illustration. The one-neutron separation energy $S_n(Z, N)$ for the tin isotopes shows a clear zigzag: for even-$N$ isotopes, $S_n$ is larger by roughly $2\Delta \approx 2.2$ MeV compared to the adjacent odd-$N$ isotopes. This staggering is visible by eye in any tabulation of separation energies and was one of the earliest clues to the existence of pairing correlations.

7.1.4 The BCS Pairing Model

The nuclear pairing problem bears a profound analogy to superconductivity in metals. In 1958, shortly after the Bardeen-Cooper-Schrieffer (BCS) theory explained superconductivity, Bohr, Mottelson, and Pines recognized that the same formalism could be applied to nuclei.

In the BCS framework, the nuclear ground state is written as:

$$| \text{BCS} \rangle = \prod_{k > 0} \left( u_k + v_k \, a^\dagger_k a^\dagger_{\bar{k}} \right) | 0 \rangle$$

where $a^\dagger_k$ creates a nucleon in single-particle state $|k\rangle$, $a^\dagger_{\bar{k}}$ creates one in the time-reversed partner $|\bar{k}\rangle$, and the product runs over all pairs of time-reversed states. The coefficients $u_k$ and $v_k$ are real numbers satisfying $u_k^2 + v_k^2 = 1$, and $v_k^2$ gives the probability that the pair $(k, \bar{k})$ is occupied.

The variational minimization of the energy with respect to $u_k$ and $v_k$ yields the BCS equations:

$$v_k^2 = \frac{1}{2} \left( 1 - \frac{\epsilon_k - \lambda}{\sqrt{(\epsilon_k - \lambda)^2 + \Delta^2}} \right)$$

$$u_k^2 = \frac{1}{2} \left( 1 + \frac{\epsilon_k - \lambda}{\sqrt{(\epsilon_k - \lambda)^2 + \Delta^2}} \right)$$

and the gap equation:

$$\Delta = -G \sum_{k > 0} u_k v_k = \frac{G}{2} \sum_{k > 0} \frac{\Delta}{\sqrt{(\epsilon_k - \lambda)^2 + \Delta^2}}$$

Here $\epsilon_k$ are the single-particle energies, $\lambda$ is the chemical potential (Fermi energy), $G$ is the pairing strength, and $\Delta$ is the pairing gap. The gap equation is self-consistent: $\Delta$ appears on both sides.

The BCS occupation probability $v_k^2$ is a smooth function of $\epsilon_k$, in stark contrast to the sharp Fermi surface of the independent-particle model. Levels below the Fermi energy are not fully occupied, and levels above are not empty. This "smearing" of the Fermi surface by pairing correlations has direct experimental consequences:

• The ground states of even-even nuclei acquire $J^\pi = 0^+$ because all nucleons are in $J = 0$ pairs.
• The first excited state requires breaking a pair, creating a gap of $2\Delta$ in the excitation spectrum.
• The effective moment of inertia is reduced because paired nucleons do not contribute to rotation.

Quasiparticles: The BCS transformation introduces a new set of elementary excitations called quasiparticles, defined by the Bogoliubov transformation:

$$\alpha^\dagger_k = u_k a^\dagger_k - v_k a_{\bar{k}}$$ $$\alpha^\dagger_{\bar{k}} = u_k a^\dagger_{\bar{k}} + v_k a_k$$

A quasiparticle is a superposition of a particle and a hole, weighted by the BCS amplitudes. The energy of a quasiparticle excitation is:

$$E_k = \sqrt{(\epsilon_k - \lambda)^2 + \Delta^2}$$

This is always greater than or equal to $\Delta$, confirming that $\Delta$ is the minimum excitation energy. Far from the Fermi surface ($|\epsilon_k - \lambda| \gg \Delta$), the quasiparticle reduces to an ordinary particle (above the Fermi surface) or an ordinary hole (below). Near the Fermi surface, the quasiparticle is a genuine mixture of particle and hole character.

The quasiparticle picture is enormously useful. In an odd-$A$ nucleus, the ground state contains one quasiparticle, and its energy and quantum numbers determine the ground-state properties. Excited states are built from one, three, five, ... quasiparticles (odd numbers only, since pair-breaking creates two quasiparticles at a time). This replacement of real particles by quasiparticles is the key simplification that makes the BCS approach so powerful.

Limitations of BCS in nuclei: The BCS theory was developed for macroscopic superconductors containing $\sim 10^{23}$ electrons. In nuclei, the number of paired nucleons is small — typically 5 to 15 pairs in a given major shell. This creates two problems:

1. Particle-number violation: The BCS wave function does not have a definite number of particles; it is a superposition of states with different particle numbers. In condensed matter, this is irrelevant because fluctuations in particle number are negligible compared to $10^{23}$. In nuclei, the fluctuation $\delta N \sim \sqrt{G/d}$ can be 1-2 particles, which is significant. Particle-number projection — mathematical projection of the BCS state onto the correct particle number — is essential for quantitative work.

2. Pairing collapse: For very small pairing strengths ($G < G_{\text{crit}}$, the critical value), the BCS gap equation has no solution with $\Delta > 0$. The system reverts to normal (unpaired) behavior. This can be an artifact of the BCS approximation; the exact solution always shows some pairing correlations, even when BCS predicts zero gap. Higher-order methods (variation after projection, the Lipkin-Nogami method) cure this deficiency.

Despite these limitations, the BCS model remains the standard tool for treating pairing in nuclear physics. Its predictions for binding energy systematics, excitation spectra, and moments of inertia are remarkably successful.

💡 Spaced Review: In Chapter 4, you encountered the pairing term in the SEMF as an empirical correction. Now you see its microscopic origin: the short-range nucleon-nucleon force preferentially couples identical nucleons in time-reversed orbits to $J = 0$, creating a correlated condensate described by BCS theory. The $12/\sqrt{A}$ formula is not a coincidence — it reflects the competition between pairing strength and level density.

7.2 Seniority and the Pairing Model

7.2.1 The Seniority Quantum Number

Before the full BCS machinery was imported from condensed matter physics, nuclear physicists had developed a powerful classification scheme based on the concept of seniority, introduced by Giulio Racah in 1943 and further developed by Racah, Talmi, and de-Shalit in the 1950s.

The idea behind seniority is deceptively simple but remarkably powerful. For $n$ identical nucleons in a single-$j$ shell, the seniority quantum number $\nu$ counts the number of nucleons not coupled in $J = 0$ pairs. The remaining $\nu$ nucleons are the "unpaired" ones, and they carry the angular momentum of the state. The seniority scheme works because the pairing interaction — the dominant component of the residual interaction — conserves seniority exactly. Under a pure pairing force, seniority is as good a quantum number as angular momentum itself.

• $\nu = 0$: All nucleons are paired, $J = 0$. This is the ground state of even-even nuclei.
• $\nu = 1$: One unpaired nucleon. Applies to odd-$A$ nuclei; $J = j$ (the spin of the unpaired nucleon's orbit).
• $\nu = 2$: One broken pair, two unpaired nucleons. These couple to allowed values of $J$ and generate the low-lying excited states.

The power of the seniority scheme is that it dramatically simplifies the many-body problem. Instead of diagonalizing the full Hamiltonian matrix in the space of all possible configurations — which for $n$ nucleons in a $j = 11/2$ shell involves matrices of dimension up to several hundred — one works in a truncated space labeled by seniority. Under a pure pairing interaction (a force that only scatters $J = 0$ pairs), seniority is an exact quantum number, and the energy eigenvalues have closed-form expressions.

In real nuclei, the residual interaction is not purely pairing — the quadrupole and other multipole components break seniority. But seniority remains approximately good for nuclei near closed shells, where the pairing interaction dominates, and it provides the physical intuition for understanding the low-energy structure of even-even and odd-$A$ nuclei throughout the periodic table.

7.2.2 Energies in the Seniority Scheme

For $n$ identical nucleons in a single-$j$ shell with a pure pairing interaction of strength $G$ (often called the "pairing Hamiltonian"):

$$H_{\text{pair}} = -G \sum_{k,k' > 0} a^\dagger_k a^\dagger_{\bar{k}} a_{\bar{k}'} a_{k'}$$

the energy eigenvalues have a remarkably simple form:

$$E(n, \nu) = -\frac{G}{4}(n - \nu)(\Omega - n - \nu + 2) + E_\nu$$

where $\Omega = j + 1/2$ is the pair degeneracy (the number of available pairs of time-reversed states), and $E_\nu$ depends on $\nu$ but not on $n$. The term $E_\nu$ is the energy of the $\nu$ unpaired nucleons and is independent of how many additional $J = 0$ pairs are present.

The term proportional to $(n - \nu)$ is the pairing energy: the binding gained by forming $(n - \nu)/2$ pairs. Each additional pair contributes an energy approximately $-G\Omega/2$.

7.2.3 Predictions of the Seniority Scheme

The seniority model makes several sharp predictions that can be tested against experiment:

1. Ground-state spins of odd-$A$ nuclei: The ground state has $\nu = 1$, and the spin is determined by the single unpaired nucleon: $J = j$. This is the same prediction as the simple shell model, and it works well.

2. Constant energy of the first excited state across an isotopic chain: For seniority $\nu = 2$ states in an even-even nucleus, the excitation energy $E(2^+_1)$ should be approximately independent of $n$ (the number of nucleons in the shell). This is because both the ground state ($\nu = 0$) and the $2^+$ state ($\nu = 2$) gain the same pairing energy from additional pairs.

3. Parabolic behavior of $B(E2; 0^+ \to 2^+)$ across a shell: The seniority scheme predicts that the $E2$ transition probability from $0^+$ to $2^+$ follows a parabolic dependence on $n$:

$$B(E2; 0^+ \to 2^+) \propto n(\Omega - n + 1)$$

This parabola peaks at mid-shell and vanishes at the beginning and end of the shell. The tin isotopes ($Z = 50$) provide a celebrated test of this prediction: the $B(E2)$ values in the Sn chain from $^{102}$Sn to $^{130}$Sn trace out a nearly perfect parabola, peaking around $^{116}$Sn at mid-shell ($N = 66$, midway through the $N = 50$-$82$ shell).

4. Magnetic moments of odd-$A$ nuclei are independent of $n$: In the seniority scheme, the magnetic moment of a $\nu = 1$ state is determined solely by the unpaired nucleon and does not change as the shell fills. The experimental moments of the Sn isotopes and the Pb isotopes confirm this to good accuracy.

5. The first $2^+$ energy is approximately constant across the tin chain: The even tin isotopes from $^{104}$Sn to $^{130}$Sn have $E(2^+_1)$ values clustered in the range 1.1-1.3 MeV, with no strong dependence on neutron number. This is exactly the seniority prediction: both the $\nu = 0$ ground state and the $\nu = 2$ excited state receive the same pairing energy from additional pairs, so the excitation energy is independent of the number of valence neutrons.

Worked example — energy levels of $n$ neutrons in $g_{7/2}$: Consider the $g_{7/2}$ orbit ($j = 7/2$, $\Omega = 4$). With a pairing strength $G = 0.5$ MeV, the seniority energy formula gives:

For the $\nu = 0$ (fully paired) ground state: $$E(n, 0) = -\frac{0.5}{4} \cdot n \cdot (4 - n + 2) = -\frac{n(6-n)}{8} \text{ MeV}$$

This gives $E(2,0) = -1.0$ MeV, $E(4,0) = -1.0$ MeV, $E(6,0) = -0.0$ MeV, $E(8,0) = +2.0$ MeV (overbinding at half-filling, reduced binding as the shell fills). Wait — let us be more careful. For $n = 2$: $E = -0.5/4 \times 2 \times (4 - 0 - 2 + 2) = -0.5/4 \times 2 \times 4 = -1.0$ MeV. For $n = 4$: $E = -0.5/4 \times 4 \times (4 - 4 - 0 + 2) = -0.5/4 \times 4 \times 2 = -1.0$ MeV. For $n = 6$: $E = -0.5/4 \times 6 \times (4 - 6 + 2) = 0$ MeV. The pairing energy is quadratic in $n$, peaking at mid-shell. The first $\nu = 2$ excited state has the same pairing energy shifted by a constant $E_{\nu=2} - E_{\nu=0}$, confirming that the excitation energy is independent of $n$.

📊 Data Check: The $B(E2; 0^+ \to 2^+)$ values for the even Sn isotopes (from the Brookhaven National Nuclear Data Center, NNDC) are: $^{104}$Sn: 0.10(3) e$^2$b$^2$, $^{112}$Sn: 0.24(1), $^{116}$Sn: 0.21(1), $^{120}$Sn: 0.20(1), $^{124}$Sn: 0.16(1), $^{130}$Sn: 0.03(1). The parabolic trend is clearly visible, though the slight asymmetry and the values near shell closure show deviations from the pure seniority limit.

7.3 Two-Particle Configurations

7.3.1 Two Nucleons Outside a Closed Shell

The simplest testing ground for residual interactions is a nucleus with exactly two valence nucleons outside a doubly-magic core. The textbook example is $^{210}$Pb: two neutrons outside the $^{208}$Pb closed shell.

In the independent-particle model, the two neutrons would fill the lowest available single-particle orbits above $N = 126$. The experimental single-particle energies, extracted from $^{209}$Pb, are:

Without any residual interaction, the lowest configuration for $^{210}$Pb would be $(2g_{9/2})^2$, and the two neutrons could couple to $J = 0, 2, 4, 6, 8$ (even values only, by the Pauli principle for identical nucleons in the same orbit). All these states would be degenerate at twice the single-particle energy, $2 \times 0 = 0$ MeV.

7.3.2 Effect of the Residual Interaction

The residual interaction splits these degenerate states. Experimentally, the low-lying spectrum of $^{210}$Pb shows:

The $J = 0$ state is pushed far below the others — this is the pairing effect in action. The 0.8 MeV gap between the $0^+$ ground state and the $2^+$ first excited state is a direct measure of the pairing energy for two $g_{9/2}$ neutrons.

The pattern of energies for the $(g_{9/2})^2$ multiplet — with the $0^+$ state strongly depressed and the higher-$J$ states clustered together — is characteristic of a short-range (surface delta) interaction and is reproduced quantitatively by the matrix elements:

$$E(J) = \epsilon_0 + \langle (g_{9/2})^2; J | V_{\text{res}} | (g_{9/2})^2; J \rangle$$

where $V_{\text{res}}$ is the residual interaction. The strong depression of the $0^+$ state relative to the others directly demonstrates the pairing mechanism.

7.3.3 Two-Particle and Two-Hole Conjugates

An important symmetry relates two-particle and two-hole nuclei. Just as $^{210}$Pb has two neutrons outside $^{208}$Pb, the nucleus $^{206}$Pb has two neutron holes in the $N = 126$ shell. The residual interaction matrix elements for two holes are related to those for two particles:

$$\langle j^{-2}; J | V | j^{-2}; J \rangle = \langle j^2; J | V | j^2; J \rangle$$

so the spectrum of two-hole states mirrors the two-particle spectrum. Experimentally, $^{206}$Pb indeed shows a $0^+$ ground state and a low-lying spectrum qualitatively similar to $^{210}$Pb, with the $0^+$ state depressed by pairing.

The analogous proton system provides additional testing: $^{210}$Po (two protons outside $^{208}$Pb) and $^{206}$Hg (two proton holes). The combination of these four nuclei — $^{210}$Pb, $^{206}$Pb, $^{210}$Po, $^{206}$Hg — provides a comprehensive test of the two-body residual interaction near the doubly-magic $^{208}$Pb core.

7.3.4 Mixed Configurations

When two orbits are close in energy, the two-particle states are not pure $(j_1)^2$ or $(j_2)^2$ configurations but mixed states. For example, in $^{210}$Pb the $0^+$ ground state is predominantly $(g_{9/2})^2$ but contains admixtures of $(i_{11/2})^2$, $(j_{15/2})^2$, and other configurations:

$$|0^+\rangle = \alpha_1 |(g_{9/2})^2; 0^+\rangle + \alpha_2 |(i_{11/2})^2; 0^+\rangle + \alpha_3 |(j_{15/2})^2; 0^+\rangle + \cdots$$

The mixing coefficients $\alpha_i$ are found by diagonalizing the residual interaction matrix in the space of all $0^+$ two-particle configurations. This is the simplest example of configuration mixing, which will be the central theme of the next section.

Numerical example: Consider the $0^+$ states of $^{210}$Pb formed from the $(g_{9/2})^2$ and $(i_{11/2})^2$ configurations. The unperturbed energies are $E_1 = 2 \times 0.000 = 0$ MeV and $E_2 = 2 \times 0.779 = 1.558$ MeV. If the off-diagonal matrix element of the residual interaction is $V_{12} = -0.40$ MeV, the $2 \times 2$ eigenvalue problem gives:

$$E_\pm = \frac{E_1 + E_2}{2} \pm \sqrt{\left(\frac{E_1 - E_2}{2}\right)^2 + V_{12}^2}$$

$$E_\pm = 0.779 \pm \sqrt{0.608 + 0.160} = 0.779 \pm 0.876$$

yielding $E_- = -0.097$ MeV and $E_+ = 1.655$ MeV. The lower state is pushed down by 0.097 MeV below the unperturbed $(g_{9/2})^2$ energy, and the upper state is pushed up by 0.097 MeV above the unperturbed $(i_{11/2})^2$ energy — this is level repulsion, a universal feature of configuration mixing. The mixing amplitude is $\alpha_2/\alpha_1 = V_{12}/(E_+ - E_2) = -0.40/0.097 \approx -4.1$... let us be more precise: the lower eigenstate has $\alpha_1 = \cos\theta$, $\alpha_2 = \sin\theta$ with $\tan 2\theta = 2V_{12}/(E_1 - E_2) = 2(-0.40)/(-1.558) = 0.513$, giving $\theta = 13.6°$, so $\alpha_1 = 0.972$, $\alpha_2 = 0.235$. The ground state is 94.5% $(g_{9/2})^2$ and 5.5% $(i_{11/2})^2$ — a modest but measurable admixture.

This admixture has observable consequences: it enhances the two-neutron transfer cross section for populating excited $0^+$ states and modifies the electromagnetic transition rates from the ground state. Even a 5% admixture can change a transition rate by 20-30% because transition rates depend on matrix elements, which involve the amplitude (not the probability).

7.4 Residual Interactions and Configuration Mixing

7.4.1 The Residual Interaction

The total Hamiltonian for $A$ nucleons is:

$$H = \sum_{i=1}^{A} T_i + \sum_{i

where $T_i$ is the kinetic energy and $V_{ij}$ is the two-body interaction. In the shell model, we split this as:

$$H = \sum_{i=1}^{A} \left[ T_i + U_i \right] + \left[ \sum_{i

The first term $H_0$ is the independent-particle Hamiltonian with mean-field potential $U_i$, whose eigenstates are Slater determinants of single-particle orbits. The second term $V_{\text{res}}$ is the residual interaction — the part of the nucleon-nucleon force not captured by the mean field.

The residual interaction is not small. For a schematic understanding, it is useful to decompose it by multipole:

$$V_{\text{res}} = V_{\text{monopole}} + V_{\text{pairing}} + V_{\text{quadrupole}} + V_{\text{higher multipole}}$$

Each component drives different physics:

• Monopole: Shifts single-particle energies as orbits fill, responsible for shell evolution far from stability (see Chapter 10).
• Pairing ($J = 0$ coupling): Discussed in Section 7.1. Creates the even-odd staggering and $0^+$ ground states.
• Quadrupole: Drives collective deformation and rotational motion (see Chapter 8). This is the strongest long-range component.
• Higher multipoles: Contribute to detailed spectroscopy but are less important for global properties.

7.4.2 The Configuration Interaction Framework

The true nuclear state $|\Psi\rangle$ is a superposition of many shell-model configurations:

$$|\Psi\rangle = \sum_\alpha c_\alpha |\Phi_\alpha\rangle$$

where each $|\Phi_\alpha\rangle$ is a Slater determinant (an independent-particle configuration), and the coefficients $c_\alpha$ are determined by diagonalizing the Hamiltonian matrix:

$$\sum_\beta \langle \Phi_\alpha | H | \Phi_\beta \rangle c_\beta = E \, c_\alpha$$

This is the configuration interaction (CI) or shell-model diagonalization problem. The dimension of the matrix grows combinatorially with the number of valence nucleons and the number of available orbits. For two particles in the $^{208}$Pb region, the matrix is small (tens of states) and manageable by hand. For mid-shell nuclei, the dimension can reach $10^{10}$ or more.

7.4.3 Two-Body Matrix Elements (TBME)

The building blocks of the shell-model calculation are the two-body matrix elements (TBME):

$$V_{j_1 j_2 j_3 j_4}^{J T} = \langle j_1 j_2; J T | V_{\text{res}} | j_3 j_4; J T \rangle$$

These matrix elements can be obtained in three ways:

1. From free nucleon-nucleon potentials: Start with a realistic $NN$ potential (Argonne $v_{18}$, CD-Bonn, or chiral EFT at N$^3$LO) and derive effective interactions for the shell-model space using many-body perturbation theory ($G$-matrix, $V_{\text{low-}k}$, or in-medium similarity renormalization group — IM-SRG). This is the ab initio approach. It is the most satisfying theoretically, because it connects shell-model spectroscopy to the fundamental nuclear force, but it is also the most difficult: the renormalization from the full Hilbert space to the small valence space introduces many-body effects that are hard to control. The inclusion of three-nucleon forces (3NF), which are now known to be essential for reproducing nuclear properties, has been a major advance in the past decade.

2. Empirical determination: Fit the TBME directly to experimental energy levels in the region of interest. The classic example is the set of 63 TBME for the $sd$-shell determined by Brown and Wildenthal (the USD interaction, 1988), obtained by fitting to 447 energy levels in $sd$-shell nuclei.

3. Hybrid: Start from a microscopically derived interaction and adjust selected matrix elements to improve agreement with data. The USDB interaction of Brown and Richter (2006) is an example.

The quality of the effective interaction determines the quality of shell-model predictions. The development of accurate effective interactions remains one of the central challenges in nuclear structure theory.

7.4.4 Particle-Vibration Coupling

Single-particle motion does not occur in a static potential. The core vibrates (as we will develop in Chapter 8), and the valence nucleon couples to these vibrations. This particle-vibration coupling (PVC) modifies single-particle energies and wave functions.

The coupling is described by a vertex where a nucleon scatters from orbit $j_1$ to orbit $j_2$ by emitting or absorbing a collective phonon of multipolarity $\lambda$:

$$\langle j_2 | H_{\text{PVC}} | j_1 \otimes \lambda \rangle$$

The effects of PVC include:

• Energy shifts: Single-particle energies are renormalized by coupling to vibrations. An orbit near a vibrational state of appropriate multipolarity is pushed down, while one far away is pushed up.
• Spectroscopic factor fragmentation: The single-particle strength is distributed over several states. Instead of a pure single-particle state, one finds the strength spread over a principal fragment and several satellite states.
• Effective mass enhancement: The coupling increases the effective mass of the nucleon near the Fermi surface, $m^*/m \approx 1.2$–$1.4$.

Experimentally, particle-vibration coupling is most clearly seen in the fragmentation of single-particle strength in nuclei near closed shells. In $^{209}$Pb, for example, the $g_{9/2}$ orbit carries only about 60-70% of the full single-particle strength; the remaining strength is spread over higher-lying states.

The experimental tool for measuring this fragmentation is the spectroscopic factor $S$, extracted from single-nucleon transfer reactions such as $(d, p)$ or $(e, e'p)$. In a pure independent-particle model, a state that is purely single-particle would have $S = 1$. The observed values are typically $S = 0.5$ to $0.7$ for the main fragment, with the remaining strength distributed over many smaller fragments at higher excitation energies. The total strength, summed over all fragments, satisfies the sum rule $\sum S = 2j + 1$ (the number of magnetic substates).

This fragmentation is not a failure of the shell model — it is a success of the more complete picture. The coupling of the nucleon to collective vibrations of the core mixes the pure single-particle state with configurations of the form $|j' \otimes \lambda\rangle$ (nucleon in orbit $j'$ coupled to a phonon of multipolarity $\lambda$). The resulting fragmentation is a direct measure of the correlation physics beyond the independent-particle model.

Particle-vibration coupling also provides a microscopic mechanism for the effective mass enhancement. The bare nucleon mass $m$ is renormalized to an effective mass $m^*$ by the coupling to vibrations, with $m^*/m \approx 1.2$-$1.4$ near the Fermi surface. This enhancement increases the level density near the Fermi energy and affects all properties that depend on the density of states, including the pairing gap.

7.5 The Interacting Shell Model

7.5.1 The Many-Body Challenge

The shell-model diagonalization problem for realistic nuclei is a formidable computational challenge. The dimension of the Hamiltonian matrix grows combinatorially:

$$D = \binom{\Omega_\pi}{n_\pi} \times \binom{\Omega_\nu}{n_\nu}$$

where $\Omega_\pi$ ($\Omega_\nu$) is the number of proton (neutron) single-particle states in the valence space and $n_\pi$ ($n_\nu$) is the number of valence protons (neutrons).

Some representative matrix dimensions:

For the $sd$-shell ($0d_{5/2}$, $1s_{1/2}$, $0d_{3/2}$), full diagonalization is routine on modern computers. The $pf$-shell ($0f_{7/2}$, $1p_{3/2}$, $0f_{5/2}$, $1p_{1/2}$) pushes to the edge of current capability. Beyond that, approximations are essential.

7.5.2 Computational Methods

The development of large-scale shell-model codes has been one of the major achievements of computational nuclear physics. Key codes and methods include:

Lanczos algorithm: Rather than diagonalizing the full matrix (which would require storing it in memory), one uses the Lanczos iterative method to extract the lowest few eigenvalues and eigenvectors. The algorithm builds a tridiagonal matrix in a Krylov subspace, and typically converges to the lowest eigenvalues in 100-300 iterations, far fewer than the matrix dimension.

$m$-scheme codes: These work in the basis of Slater determinants labeled by the individual nucleon quantum numbers. The matrix is extremely sparse (because the two-body interaction connects at most two orbits), making storage and multiplication efficient. The ANTOINE code (Caurier and colleagues, Strasbourg) and the MSHELL/NuShellX codes (Rae, Brown) pioneered this approach. The modern KSHELL code (Shimizu, Tokyo) and BIGSTICK (Johnson and colleagues) handle dimensions exceeding $10^{10}$.

Monte Carlo shell model (MCSM): For dimensions beyond the reach of direct Lanczos diagonalization, the MCSM (Otsuka, Honma, Shimizu) uses stochastic sampling of important basis states. The method projects onto good angular momentum and parity, selects the most important configurations through auxiliary-field Monte Carlo, and has been applied successfully to nuclei as heavy as the nickel and germanium isotopes.

Importance-truncated shell model: This approach systematically truncates the basis by keeping only configurations that contribute significantly to the wave function. Natural orbital methods and tensor-network approaches are emerging as next-generation tools.

7.5.3 Successes of Large-Scale Shell Model

The interacting shell model, with carefully tuned effective interactions, provides the most detailed and accurate description of nuclear spectroscopy available:

• $sd$-shell nuclei ($A = 17$–$39$): The USD/USDB interaction reproduces hundreds of energy levels, electromagnetic moments, transition rates, and beta-decay rates to within a few percent of experiment.

• $pf$-shell nuclei ($A = 41$–$65$): The KB3G and GXPF1A interactions reproduce the onset of collectivity, the $N = 28$ shell closure, the island of inversion around $^{32}$Mg (see Chapter 10), and the structure of $^{48}$Ca and $^{56}$Ni.

• Gamow-Teller strength distributions: Critical for supernova physics and electron capture rates, these are one of the great successes of the shell model, which captures the quenching and fragmentation of GT strength observed experimentally.

• Nuclear matrix elements for double-beta decay: The shell model provides some of the most reliable calculations of the nuclear matrix elements governing neutrinoless double-beta decay ($0\nu\beta\beta$), a process whose observation would prove that neutrinos are Majorana particles. The matrix elements for candidate nuclei like $^{76}$Ge, $^{130}$Te, and $^{136}$Xe are among the highest-stakes shell-model calculations in nuclear physics (see Chapter 32).

• Isomer predictions: The shell model correctly predicts the existence, spins, and approximate excitation energies of nuclear isomers, because these depend sensitively on the detailed wave function composition and transition matrix elements.

The USD family of interactions — a case study in effective interaction development: The development of the $sd$-shell effective interaction illustrates the interplay between theory and experiment. The original USD interaction (Brown and Wildenthal, 1988) contained 63 two-body matrix elements and 3 single-particle energies, fitted to 447 experimentally known energy levels in $A = 17$-$39$ nuclei. The fit achieved a root-mean-square deviation of 185 keV — impressive agreement for a 66-parameter fit to nearly 450 data points. In 2006, Brown and Richter updated the interaction using 608 data points and a refined fitting procedure, producing the USDA and USDB interactions with RMS deviations of 170 and 130 keV, respectively. The USDB interaction is now the standard for $sd$-shell calculations and has been used in thousands of papers.

The lesson is clear: an effective interaction with a few dozen parameters, constrained by experimental data, can describe the low-energy spectroscopy of hundreds of nuclei with sub-200-keV accuracy. This empirical success is possible because the physics of the valence space is dominated by a manageable number of two-body matrix elements, and the excluded physics (core excitations, high-energy configurations) can be absorbed into effective operators.

7.5.4 Limitations

Despite its successes, the interacting shell model has fundamental limitations:

1. Truncation of the model space: Only valence nucleons are treated explicitly; the core is assumed inert. Core excitations ("core polarization") must be incorporated through effective operators or extended model spaces.

2. Effective interaction uncertainty: The effective interaction in the model space is not uniquely defined. Different starting points (free $NN$ force vs. empirical fit) can give different results, especially for observables sensitive to details of the wave function.

3. Missing physics: The standard shell model treats the nucleus as a collection of nucleons interacting through two-body forces in a fixed, bound-state model space. Three-body forces (now known to be quantitatively important), continuum effects (crucial near the drip lines where the last nucleons are barely bound), and cluster degrees of freedom (relevant for light nuclei and alpha-clustering) are all difficult to incorporate within the standard framework. Extensions to include these effects — the Gamow shell model for continuum, explicit three-body operators in the Hamiltonian, and cluster-configuration shell models — represent active areas of development.

4. Deformed nuclei: The spherical shell model requires enormous model spaces to describe well-deformed nuclei, because many major shells must be included. The Nilsson model (Section 7.7) provides a more natural starting point for deformed nuclei.

🔗 Connection: The computational challenge of the nuclear shell model is a frontier of high-performance computing. The largest calculations today use leadership-class supercomputers (Summit, Frontier, Fugaku) and new algorithmic approaches including machine learning. We will encounter this theme again in Chapter 33 (Frontiers).

7.6 Nuclear Isomers

7.6.1 Definition and Classification

A nuclear isomer is a metastable excited state of a nucleus — one that survives long enough to be observed as a distinct entity before decaying to a lower state. By convention, the term "isomer" is reserved for states with half-lives longer than about $10^{-9}$ seconds (1 nanosecond), though this boundary is somewhat arbitrary.

Nuclear isomers are classified by the mechanism that hinders their decay:

1. Spin isomers (spin traps): The most common type. The isomeric state has a spin that differs greatly from all accessible lower-lying states. Since electromagnetic transition rates decrease rapidly with increasing change in angular momentum $\Delta J$ — roughly as $(R/\lambda)^{2\Delta J}$ where $R$ is the nuclear radius and $\lambda$ is the photon wavelength — transitions with large $\Delta J$ are highly suppressed. A classic example: the $8^-$ isomer in $^{210}$Pb (at 4.895 MeV) must decay by an $M3$ transition ($\Delta J = 3$, parity change), giving it a half-life of 60 ns despite its high excitation energy.

2. Shape isomers: The isomeric state has a very different deformation from the ground state, and the transition requires a major rearrangement of the nuclear shape. The fission isomers in the actinides are the best-known examples: states in the second minimum of the potential energy surface, with deformations roughly twice the ground-state value and half-lives of nanoseconds to milliseconds.

3. K-isomers: In axially-deformed nuclei, the quantum number $K$ (the projection of angular momentum on the symmetry axis) is approximately conserved. Transitions that require a large change in $K$ are hindered by a selection rule that forbids $K$-changing transitions beyond order $\nu = |\Delta K| - \lambda$ (where $\lambda$ is the multipolarity). States with large $K$ values thus decay slowly. The $K^\pi = 16^+$ isomer in $^{178}$Hf, with a half-life of 31 years, is a dramatic example.

The boundary between "isomeric" and "normal" excited states is not sharp. The International Atomic Energy Agency (IAEA) and the Evaluated Nuclear Structure Data File (ENSDF) use thresholds ranging from 10 ns to 1 ms, depending on the context. In this textbook, we adopt the 10 ns threshold, which captures the physics: any state surviving longer than $\sim 10^{-8}$ s is living far longer than the typical electromagnetic decay time of $\sim 10^{-12}$ to $10^{-15}$ s and therefore has some structural origin for its longevity.

As of the NUBASE2020 evaluation, approximately 2,500 nuclear isomers are known, distributed across the chart of nuclides from the lightest nuclei to the superheavy elements. New isomers continue to be discovered at radioactive beam facilities, particularly in neutron-rich exotic nuclei.

7.6.2 Weisskopf Estimates and Hindrance Factors

To quantify how "isomeric" a transition is, we compare its observed rate to the Weisskopf single-particle estimate (which will be derived in detail in Chapter 9). For an electric multipole transition of order $\lambda$, the Weisskopf estimate is:

$$T_W(E\lambda) = \frac{4.4 \times 10^{21}}{\hbar} \left( \frac{3}{\lambda + 3} \right)^2 \left( \frac{E_\gamma}{197 \text{ MeV}} \right)^{2\lambda+1} \left( \frac{R}{1 \text{ fm}} \right)^{2\lambda} \text{ s}^{-1}$$

where $E_\gamma$ is the transition energy and $R = 1.2 A^{1/3}$ fm. The hindrance factor is:

$$F_W = \frac{T_W}{T_{\text{exp}}} = \frac{t_{1/2}^{\text{exp}}}{t_{1/2}^{W}}$$

For isomeric transitions, $F_W$ can be enormous — $10^2$ to $10^8$ or more — indicating that the transition is much slower than a single-particle estimate would predict.

7.6.3 Islands of Isomerism

Nuclear isomers are not uniformly distributed across the chart of nuclides. They cluster in specific regions called islands of isomerism, which correlate with shell structure:

Near magic numbers: The high-$j$ intruder orbitals (the unique-parity orbits pushed down by spin-orbit coupling, like $g_{9/2}$, $h_{11/2}$, $i_{13/2}$) are surrounded by orbits of much lower $j$. When nucleons occupy these high-$j$ orbits near a closed shell, states with large spin can form at relatively low excitation energy, and they can decay only through high-multipolarity transitions to the low-spin states below.

The principal islands of isomerism are:

7.6.4 Some Remarkable Isomers

$^{180m}$Ta ($K^\pi = 9^-$, $E^* = 77$ keV): The rarest naturally occurring nuclear isomer and one of the most extraordinary nuclei known. This isomer has never been observed to decay — its half-life exceeds $1.2 \times 10^{15}$ years, far longer than the age of the universe. It exists in nature alongside the short-lived ground state ($t_{1/2} = 8.15$ hours, $1^+$). The extreme hindrance arises because the decay from $9^-$ to $1^+$ requires $\Delta K = 8$, which is forbidden to high order. The mystery of how $^{180m}$Ta was produced in nature (likely in the neutrino process during supernovae) connects nuclear isomerism to astrophysics.

$^{99m}$Tc ($J^\pi = 1/2^-$, $E^* = 142.7$ keV, $t_{1/2} = 6.01$ hours): The single most important nuclear isomer in human history. Its decay by a 140.5 keV $M4$ gamma ray (with 89% gamma emission and 11% internal conversion) provides the basis for the majority of nuclear medicine diagnostic imaging. Over 30 million procedures per year worldwide use $^{99m}$Tc. The isomer has a half-life ideal for medical purposes: long enough to prepare and administer the radiopharmaceutical, short enough that the patient's radiation dose is manageable. Its isomeric character arises from the $\Delta J = 4$ change between the $1/2^-$ isomer and the $9/2^+$ ground state. We will examine this application in detail in Case Study 1 and in Chapter 27.

$^{178m2}$Hf ($K^\pi = 16^+$, $E^* = 2.446$ MeV, $t_{1/2} = 31$ yr): A high-$K$ isomer at remarkably high excitation energy. The configuration is approximately four aligned quasiparticles: two protons ($\pi 7/2^+[404]$ and $\pi 9/2^-[514]$) and two neutrons ($\nu 7/2^-[514]$ and $\nu 9/2^+[624]$), with $K = 7/2 + 9/2 + 7/2 + 9/2 = 16$. The 2.446 MeV stored in this isomer amounts to roughly 1.3 GJ per gram, exceeding the energy density of TNT by a factor of $\sim 3 \times 10^5$. This extraordinary energy storage led to a controversial proposal in the early 2000s (the "hafnium bomb" concept) to release the stored energy through induced deexcitation using X-ray triggers. Extensive experimental work, funded by DARPA and conducted at several laboratories, failed to demonstrate triggered energy release. Independent experiments at Argonne National Laboratory, Lawrence Livermore National Laboratory, and the Advanced Photon Source found no evidence for the claimed effect, and the physics community largely considers the idea refuted. The fundamental obstacle is that the cross section for photoexcitation to "gateway states" connecting the $K = 16$ band to lower-$K$ bands is prohibitively small.

⚠️ Physical Insight: Nuclear isomers illustrate a key theme of this chapter: the detailed structure of nuclear states — their spins, parities, and wave function composition — controls transition rates by many orders of magnitude. A few units of angular momentum difference can mean the difference between a transition that occurs in femtoseconds and one that takes billions of years.

7.7 The Nilsson Model for Deformed Nuclei

7.7.1 Evidence for Nuclear Deformation

The shell model of Chapter 6 assumed a spherically symmetric potential. This is an excellent approximation near closed shells, where nuclei are indeed nearly spherical. But many nuclei — particularly those in the rare-earth ($150 \lesssim A \lesssim 190$) and actinide ($A \gtrsim 220$) regions — have large permanent quadrupole deformations.

The experimental evidence for deformation includes:

1. Large electric quadrupole moments $Q$: Deformed nuclei have quadrupole moments 5-10 times larger than single-particle estimates. For example, $^{176}$Lu has $Q = +8.0$ b (barns), compared to a single-particle estimate of $\sim 1$ b.

2. Rotational band structure: Deformed nuclei show sequences of states with energies $E(I) \propto I(I+1)$, characteristic of a quantum rigid rotor. These rotational bands, discussed in Chapter 8, are the clearest fingerprint of deformation.

3. Enhanced $E2$ transitions: The $B(E2; 0^+ \to 2^+)$ values in deformed nuclei are 10-100 times larger than single-particle estimates, reflecting the coherent contribution of many nucleons to the quadrupole transition.

4. Isotope shifts: Laser spectroscopy reveals sudden changes in the mean-square charge radius at the onset of deformation, notably near $N = 90$ (the onset of deformation in the rare-earth region).

7.7.2 The Nilsson Hamiltonian

In 1955, Sven Gösta Nilsson proposed extending the shell model to deformed potentials. The starting point is a deformed harmonic oscillator, modified with spin-orbit and centrifugal-flattening corrections:

$$H_{\text{Nilsson}} = -\frac{\hbar^2}{2m}\nabla^2 + \frac{1}{2}m\left[\omega_\perp^2(x^2 + y^2) + \omega_z^2 z^2\right] - \kappa\hbar\omega_0 \left[ 2\boldsymbol{\ell}\cdot\mathbf{s} + \mu(\boldsymbol{\ell}^2 - \langle \boldsymbol{\ell}^2 \rangle_N) \right]$$

The frequencies $\omega_\perp$ and $\omega_z$ parameterize the deformation. For a prolate (cigar-shaped) nucleus, $\omega_z < \omega_\perp$; for an oblate (disk-shaped) nucleus, $\omega_z > \omega_\perp$. The deformation parameter $\epsilon$ (or equivalently $\delta$ or $\beta_2$) relates the frequencies:

$$\omega_z = \omega_0 \left(1 - \frac{2}{3}\epsilon\right), \quad \omega_\perp = \omega_0 \left(1 + \frac{1}{3}\epsilon\right)$$

with volume conservation: $\omega_\perp^2 \omega_z = \omega_0^3$.

The last two terms are the spin-orbit coupling (with strength $\kappa$) and the $\ell^2$ term (with strength $\mu$), which corrects the level ordering from harmonic oscillator to something closer to a realistic Woods-Saxon potential. The parameters $\kappa$ and $\mu$ are fit to reproduce the spherical magic numbers when $\epsilon = 0$.

7.7.3 Symmetries and Quantum Numbers

The Nilsson Hamiltonian has axial symmetry about the $z$-axis (the symmetry axis of the deformed potential) but not full spherical symmetry. Consequently:

• Good quantum number: $\Omega = m_j = m_\ell + m_s$, the projection of the total angular momentum on the symmetry axis.
• Not good quantum numbers: $\ell$, $j$ (broken by the deformation).
• Approximately good: $N$ (the major oscillator quantum number) remains approximately good because the deformation mixes orbits primarily within the same $N$-shell.

Each Nilsson level is labeled by the asymptotic quantum numbers $[N n_z \Lambda] \Omega$:

• $N$: principal oscillator quantum number
• $n_z$: number of oscillator quanta along the symmetry axis ($z$)
• $\Lambda$: projection of orbital angular momentum on the symmetry axis
• $\Omega$: projection of total angular momentum on the symmetry axis

These quantum numbers are exact only in the limit of pure harmonic oscillator with no spin-orbit coupling. In the Nilsson model, they are approximate but remain useful labels. The notation $\Omega^\pi [N n_z \Lambda]$ is commonly used, where $\pi = (-1)^{\ell}$ is the parity.

7.7.4 Nilsson Diagrams

The Nilsson diagram plots single-particle energies as a function of the deformation parameter $\epsilon$ (or $\beta_2$). At $\epsilon = 0$, the levels are the familiar spherical shell-model levels, labeled by $n\ell j$. As $\epsilon$ increases (or decreases), the levels split according to their $\Omega$ values.

The splitting follows a simple rule for a pure harmonic oscillator: an orbit with quantum numbers $n_z$ and $n_\perp = N - n_z$ has energy:

$$E = \hbar\omega_z \left(n_z + \frac{1}{2}\right) + \hbar\omega_\perp \left(n_\perp + 1\right)$$

For prolate deformation ($\omega_z < \omega_\perp$): - Orbits with large $n_z$ (elongated along the symmetry axis, small $\Omega$) decrease in energy. - Orbits with small $n_z$ (concentrated in the equatorial plane, large $\Omega$) increase in energy.

The reverse is true for oblate deformation.

When the spin-orbit and $\ell^2$ terms are included, the level ordering becomes more complex, but the general trend persists. Level crossings occur at specific deformations, and these can create new shell gaps — new "deformed magic numbers" — that stabilize particular deformation values.

The most important deformed shell gaps occur at particle numbers:

These deformed shell gaps explain the onset of deformation in specific regions of the nuclear chart and the existence of superdeformed rotational bands.

The Nilsson diagram is one of the most important tools in nuclear structure physics. A well-drawn Nilsson diagram — and every nuclear experimentalist has one pinned to their office wall — allows one to identify the expected ground-state configuration, the available excited states, and the expected transition rates for any deformed nucleus.

Several important features of the Nilsson diagram deserve emphasis:

The no-crossing rule for levels of the same quantum numbers: Levels with the same $\Omega$ and parity cannot cross; they repel each other (von Neumann-Wigner theorem). This creates characteristic avoided crossings that are visible as the levels "bounce" off each other in the diagram. The gaps between levels at a given deformation are the deformed shell gaps.

Intruder orbitals: The unique-parity orbital from each major shell (e.g., $h_{11/2}$ in the $N = 4$ shell, $i_{13/2}$ in the $N = 5$ shell) plays a special role. Because this orbital has opposite parity from the rest of its shell, it cannot mix with them, and it traces a distinctive, often steeply sloping trajectory across the Nilsson diagram. These intruder orbitals are often responsible for the low-$\Omega$ states that drive deformation, and they produce the high-$K$ configurations that give rise to $K$-isomers.

Deformation driving: An orbit with large $n_z$ (elongated along the symmetry axis) not only decreases in energy with prolate deformation but actively drives the deformation. When such an orbit lies near the Fermi surface, filling it lowers the total energy at finite deformation, making the deformed configuration energetically favorable. This is the microscopic mechanism behind the onset of deformation. The competition between the spherical shell gap (which favors zero deformation) and the gain from deformation-driving orbitals determines the equilibrium shape of each nucleus.

7.7.5 Ground-State Predictions from the Nilsson Model

For an odd-$A$ deformed nucleus, the ground-state spin and parity are determined by the Nilsson orbital occupied by the unpaired nucleon. The Nilsson model thus provides a direct prediction: find the Nilsson diagram, fill orbitals up to the Fermi surface at the appropriate deformation, and read off $\Omega^\pi$ for the last filled orbit.

Example: $^{177}$Hf ($Z = 72$, $N = 105$). The ground state has $J^\pi = 7/2^-$. In the Nilsson model, with a prolate deformation of $\epsilon \approx 0.28$, the 105th neutron occupies the $\Omega^\pi[Nn_z\Lambda] = 7/2^-[514]$ orbital. This corresponds to a state derived from the $h_{9/2}$ spherical orbit ($\ell = 5$, negative parity), with the asymptotic quantum numbers $N = 5$, $n_z = 1$, $\Lambda = 4$, giving $\Omega = \Lambda + 1/2 = 7/2$. The model correctly predicts the observed $7/2^-$ ground state.

Example: $^{175}$Lu ($Z = 71$, $N = 104$). The odd proton occupies the $7/2^+[404]$ Nilsson orbital, derived from the $g_{7/2}$ spherical orbit, giving $J^\pi = 7/2^+$, in agreement with experiment.

The Nilsson model's predictive power for ground-state spins and parities of deformed nuclei is comparable to the spherical shell model's success near closed shells. The combination of the two — spherical shell model near magic numbers, Nilsson model in deformed regions — covers most of the nuclear chart.

A systematic comparison: The following table lists several deformed odd-$A$ nuclei, their experimental ground-state spins and parities, and the Nilsson model predictions:

The agreement is essentially perfect. In each case, the spherical shell model would predict a different (and wrong) spin-parity. The Nilsson model correctly identifies the orbital occupied by the unpaired nucleon at the equilibrium deformation and reads off $\Omega^\pi$ as the ground-state spin-parity.

Note that the asymptotic quantum numbers carry physical information. For $^{235}$U with the orbital $7/2^-[743]$: $N = 7$ places the neutron in the highest oscillator shell, $n_z = 4$ means four of the seven quanta are along the symmetry axis (a highly elongated orbit), $\Lambda = 3$ gives three units of orbital angular momentum about the axis, and $\Omega = \Lambda + 1/2 = 7/2$ with spin parallel to $\Lambda$. The large $n_z$ means the unpaired neutron's wave function extends strongly along the nuclear symmetry axis, which has direct consequences for the angular distribution of fission fragments and the neutron capture cross section.

7.7.6 Connections and Limitations

The Nilsson model is an independent-particle model in a deformed potential. All the complications discussed earlier in this chapter — pairing, residual interactions, configuration mixing — apply equally to the Nilsson model, and the full treatment of deformed nuclei requires:

1. Deformed BCS or Hartree-Fock-Bogoliubov (HFB): Pairing correlations in the deformed basis.
2. Particle-rotor model: Coupling of single-particle Nilsson states to the collective rotational motion, accounting for the Coriolis force.
3. Projected shell model: Angular momentum projection from the deformed intrinsic state, restoring the broken rotational symmetry.

These developments connect this chapter to Chapter 8 (collective motion) and illustrate the central theme: nuclear structure requires a synthesis of single-particle, pairing, and collective degrees of freedom. No single model captures the full picture; the art of nuclear structure is knowing which approximation to use where.

💡 Spaced Review: In Chapter 6, the shell model predicted ground-state spins by filling spherical orbits. Now you have a second tool: the Nilsson model, which does the same for deformed nuclei by filling orbitals in a deformed potential. The key new quantum number $\Omega$ replaces $j$ as the good quantum number when spherical symmetry is broken. The two approaches are complementary — one works near closed shells, the other in deformed regions — and together they cover most of the chart of nuclides.

7.8 Summary: The Hierarchy of Nuclear Models

This chapter has revealed the nuclear many-body problem in its true complexity. Let us organize what we have learned as a hierarchy of approximations, each building on the last:

Level 0 — The liquid drop (SEMF, Chapter 4): Nuclei as structureless drops of nuclear matter. Captures bulk binding energy, the valley of stability, and the energetics of fission. No shell structure, no spectroscopy.

Level 1 — The independent-particle shell model (Chapter 6): Nucleons in a mean-field potential with spin-orbit coupling. Explains magic numbers, ground-state spins and parities, and magnetic moments. Misses pairing, collectivity, and deformation.

Level 2 — Shell model with pairing (this chapter, Sections 7.1–7.2): Adding the pairing interaction via seniority or BCS. Explains even-odd staggering, $0^+$ ground states, the pairing gap, reduced moments of inertia. Still works in a spherical basis.

Level 3 — Interacting shell model (this chapter, Sections 7.3–7.5): Full configuration mixing in the valence space. The state-of-the-art for detailed spectroscopy of spherical and weakly-deformed nuclei. Limited by computational cost to moderate model spaces.

Level 4 — Nilsson model and collective extensions (this chapter, Section 7.7, and Chapter 8): Deformed mean field + pairing + collective rotation and vibration. Essential for the rare-earth and actinide regions. The particle-rotor model, cranked shell model, and projected shell model are the workhorses.

Level 5 — Ab initio nuclear structure (Chapter 33): Start from the nuclear force and solve the many-body problem without empirical adjustments. Currently feasible up to $A \approx 100$ with coupled-cluster, in-medium SRG, and valence-space approaches. The ultimate goal of the field.

The progression from Level 0 to Level 5 is the story of nuclear structure physics over the past 90 years. Each level builds on the one below, and each has its domain of validity:

• The liquid drop (SEMF) is best for global mass systematics and fission energetics.
• The independent-particle shell model is best near closed shells for ground-state properties.
• The pairing model adds the even-odd distinction and the superfluid properties.
• The interacting shell model provides the gold standard for detailed spectroscopy of spherical and weakly-deformed nuclei.
• The Nilsson model and its extensions are essential for the large regions of the nuclear chart where nuclei are deformed.
• Ab initio methods are the frontier, providing parameter-free predictions from the underlying nuclear force.

The nucleus is not a simple system — it is a strongly-interacting quantum many-body system with no small parameter, where single-particle, pairing, and collective modes compete and coexist. The beauty of the field lies in the interplay of these modes, and in the fact that relatively simple models — when used in the right regime — capture the essential physics remarkably well. A working nuclear physicist must be fluent in all of these models, knowing when each is appropriate and where each breaks down. This is the art of nuclear structure, and it is a theme we will return to throughout the remainder of this textbook.

Chapter Summary

• The pairing interaction, arising from the short-range nucleon-nucleon force, preferentially couples identical nucleons in time-reversed orbits to $J = 0$. It produces the even-odd binding energy staggering, the universal $0^+$ ground states of even-even nuclei, and a pairing gap $\Delta \approx 12/\sqrt{A}$ MeV.

• The BCS model, adapted from superconductivity, provides a quantitative description of nuclear pairing, with occupation probabilities $v_k^2$ that smoothly smear the Fermi surface.

• The seniority scheme classifies states by the number of unpaired nucleons, dramatically simplifying the many-body problem. It predicts constant excitation energies and parabolic $B(E2)$ trends across isotopic chains, well confirmed by the tin isotopes.

• Two-particle configurations near doubly-magic nuclei (e.g., $^{210}$Pb) provide the cleanest test of residual interaction matrix elements, showing the characteristic depression of $J = 0$ states.

• Configuration mixing means the true nuclear state is a superposition of many shell-model configurations. The interacting shell model diagonalizes the residual interaction in the full valence space, achieving remarkable spectroscopic accuracy but facing exponentially growing computational costs.

• Nuclear isomers are metastable excited states, classified as spin traps, shape isomers, or $K$-isomers. They cluster in "islands of isomerism" near magic numbers and have profound applications ($^{99m}$Tc in medicine) and fundamental interest ($^{180m}$Ta, $^{178m2}$Hf).

• The Nilsson model extends the shell model to deformed potentials, introducing the quantum number $\Omega$ and the asymptotic labeling $[N n_z \Lambda]\Omega$. Nilsson diagrams show how single-particle levels evolve with deformation, creating deformed shell gaps that stabilize specific shapes.

• Nuclear structure is a hierarchy of models, each capturing different aspects of the many-body problem. The art lies in knowing which model to apply where.