General Theory of Angular Momentum & Clebsch-Gordan Coefficients

1. The Quantum Lie Algebra of Rotations

In quantum mechanics, angular momentum operators are defined as the infinitesimal Hermitian generators of spatial rotations. An infinitesimal rotation by angle $d\phi$ about unit vector $\hat{n}$ acts on state kets via the unitary operator:

$$\hat{R}_{\hat{n}}(d\phi) = \hat{I} - \frac{i}{\hbar}(\hat{\vec{J}}\cdot\hat{n})\,d\phi$$

The non-Abelian geometry of physical three-dimensional rotations requires that the components of the general angular momentum vector operator $\hat{\vec{J}} = (\hat{J}_x, \hat{J}_y, \hat{J}_z)$ satisfy the fundamental commutation relations:

$$[\hat{J}_i, \hat{J}_j] = i\hbar \sum_{k=1}^3 \epsilon_{ijk} \hat{J}_k \iff [\hat{J}_x, \hat{J}_y] = i\hbar \hat{J}_z, \quad [\hat{J}_y, \hat{J}_z] = i\hbar \hat{J}_x, \quad [\hat{J}_z, \hat{J}_x] = i\hbar \hat{J}_y$$

These commutation relations define the Lie algebra $\mathfrak{su}(2) \cong \mathfrak{so}(3)$.

2. The Casimir Invariant Operator $\hat{\vec{J}}^2$

The total angular momentum squared is defined as:

$$\hat{\vec{J}}^2 \equiv \hat{J}_x^2 + \hat{J}_y^2 + \hat{J}_z^2$$

Evaluating the commutator with any Cartesian component $\hat{J}_k$:

$$[\hat{\vec{J}}^2, \hat{J}_z] = [\hat{J}_x^2, \hat{J}_z] + [\hat{J}_y^2, \hat{J}_z] + [\hat{J}_z^2, \hat{J}_z]$$ $$= \hat{J}_x[\hat{J}_x, \hat{J}_z] + [\hat{J}_x, \hat{J}_z]\hat{J}_x + \hat{J}_y[\hat{J}_y, \hat{J}_z] + [\hat{J}_y, \hat{J}_z]\hat{J}_y + 0$$ $$= -i\hbar(\hat{J}_x\hat{J}_y + \hat{J}_y\hat{J}_x) + i\hbar(\hat{J}_y\hat{J}_x + \hat{J}_x\hat{J}_y) = 0$$

Hence, $\hat{\vec{J}}^2$ commutes with all three components: $[\hat{\vec{J}}^2, \hat{J}_i] = 0$. By Schur's lemma, $\hat{\vec{J}}^2$ is the quadratic Casimir invariant of the rotation group. A complete set of commuting observables (CSCO) can be chosen as $\{\hat{\vec{J}}^2, \hat{J}_z\}$.

3. Raising and Lowering Ladder Operators

We define the non-Hermitian ladder operators:

$$\hat{J}_+ \equiv \hat{J}_x + i\hat{J}_y, \quad \hat{J}_- \equiv \hat{J}_x - i\hat{J}_y = (\hat{J}_+)^\dagger$$

Their commutation relations with $\hat{J}_z$ and $\hat{\vec{J}}^2$ are:

$$[\hat{J}_z, \hat{J}_\pm] = [\hat{J}_z, \hat{J}_x] \pm i[\hat{J}_z, \hat{J}_y] = i\hbar\hat{J}_y \pm i(-i\hbar\hat{J}_x) = \pm\hbar(\hat{J}_x \pm i\hat{J}_y) = \pm\hbar \hat{J}_\pm$$ $$[\hat{\vec{J}}^2, \hat{J}_\pm] = 0$$

Products of the ladder operators relate directly to $\hat{\vec{J}}^2$:

$$\hat{J}_\mp \hat{J}_\pm = (\hat{J}_x \mp i\hat{J}_y)(\hat{J}_x \pm i\hat{J}_y) = \hat{J}_x^2 + \hat{J}_y^2 \pm i[\hat{J}_x, \hat{J}_y] = \hat{\vec{J}}^2 - \hat{J}_z^2 \mp \hbar\hat{J}_z$$ $$\hat{\vec{J}}^2 = \hat{J}_- \hat{J}_+ + \hat{J}_z^2 + \hbar\hat{J}_z = \hat{J}_+ \hat{J}_- + \hat{J}_z^2 - \hbar\hat{J}_z$$

1. Algebraic Derivation of the Quantum Numbers $j$ and $m$

Let $|j, m\rangle$ denote simultaneous normalized eigenstates of $\hat{\vec{J}}^2$ and $\hat{J}_z$:

$$\hat{\vec{J}}^2 |j, m\rangle = \lambda \hbar^2 |j, m\rangle, \quad \hat{J}_z |j, m\rangle = m\hbar |j, m\rangle$$

Applying the commutator $[\hat{J}_z, \hat{J}_\pm] = \pm\hbar\hat{J}_\pm$:

$$\hat{J}_z(\hat{J}_\pm |j, m\rangle) = (\hat{J}_\pm\hat{J}_z + [\hat{J}_z, \hat{J}_\pm])|j, m\rangle = (m\hbar \pm \hbar)(\hat{J}_\pm |j, m\rangle) = (m \pm 1)\hbar(\hat{J}_\pm |j, m\rangle)$$

Thus, $\hat{J}_\pm |j, m\rangle$ is an eigenstate of $\hat{J}_z$ with eigenvalue $(m \pm 1)\hbar$. Since $\langle j, m | \hat{\vec{J}}^2 - \hat{J}_z^2 | j, m \rangle = \langle j, m | \hat{J}_x^2 + \hat{J}_y^2 | j, m \rangle \ge 0$:

$$\lambda \hbar^2 - m^2\hbar^2 \ge 0 \implies m^2 \le \lambda$$

Therefore, the spectrum of $m$ must be bounded from above by a maximum value $j$, and from below by a minimum value $j'$:

$$\hat{J}_+ |j, j\rangle = 0 \implies \hat{J}_- \hat{J}_+ |j, j\rangle = (\hat{\vec{J}}^2 - \hat{J}_z^2 - \hbar\hat{J}_z)|j, j\rangle = (\lambda - j^2 - j)\hbar^2 |j, j\rangle = 0 \implies \lambda = j(j+1)$$ $$\hat{J}_- |j, j'\rangle = 0 \implies \hat{J}_+ \hat{J}_- |j, j'\rangle = (\hat{\vec{J}}^2 - \hat{J}_z^2 + \hbar\hat{J}_z)|j, j'\rangle = (\lambda - j'^2 + j')\hbar^2 |j, j'\rangle = 0 \implies \lambda = j'(j'-1)$$

Equating the two expressions for $\lambda$:

$$j(j+1) = j'(j'-1) \implies j' = -j$$

Because one transitions from $-j$ to $+j$ in discrete integer steps of $+1$:

$$j - (-j) = 2j = k \in \mathbb{N}_0 \implies j \in \left\{0, \frac{1}{2}, 1, \frac{3}{2}, 2, \dots\right\}$$

For a given $j$, the magnetic quantum number $m$ takes $2j+1$ distinct values:

$$m \in \{-j, -j+1, \dots, j-1, j\}$$

2. Matrix Elements of Angular Momentum

Evaluating the normalization of $\hat{J}_\pm |j, m\rangle$:

$$\|\hat{J}_\pm |j, m\rangle\|^2 = \langle j, m | \hat{J}_\mp \hat{J}_\pm | j, m \rangle = \langle j, m | (\hat{\vec{J}}^2 - \hat{J}_z^2 \mp \hbar\hat{J}_z) | j, m \rangle = \hbar^2 [j(j+1) - m(m \pm 1)]$$

Choosing the standard Condon-Shortley positive phase convention:

$$\hat{J}_\pm |j, m\rangle = \hbar\sqrt{j(j+1) - m(m \pm 1)} \, |j, m \pm 1\rangle$$

The Cartesian matrix elements are:

$$\langle j, m' | \hat{J}_x | j, m \rangle = \frac{\hbar}{2}\left[\sqrt{j(j+1)-m(m+1)}\delta_{m', m+1} + \sqrt{j(j+1)-m(m-1)}\delta_{m', m-1}\right]$$ $$\langle j, m' | \hat{J}_y | j, m \rangle = \frac{\hbar}{2i}\left[\sqrt{j(j+1)-m(m+1)}\delta_{m', m+1} - \sqrt{j(j+1)-m(m-1)}\delta_{m', m-1}\right]$$ $$\langle j, m' | \hat{J}_z | j, m \rangle = m\hbar \delta_{m', m}$$

1. Differential Operators in Spherical Polar Coordinates

Orbital angular momentum is defined classically as $\vec{L} = \vec{r} \times \vec{p}$. In coordinate space, substituting $\hat{\vec{p}} = -i\hbar\nabla$:

$$\hat{\vec{L}} = -i\hbar(\vec{r} \times \nabla)$$

In spherical polar coordinates $(r, \theta, \phi)$, the Cartesian components become:

$$\hat{L}_x = i\hbar\left(\sin\phi\frac{\partial}{\partial\theta} + \cot\theta\cos\phi\frac{\partial}{\partial\phi}\right)$$ $$\hat{L}_y = i\hbar\left(-\cos\phi\frac{\partial}{\partial\theta} + \cot\theta\sin\phi\frac{\partial}{\partial\phi}\right)$$ $$\hat{L}_z = -i\hbar\frac{\partial}{\partial\phi}$$

The Casimir invariant $\hat{\vec{L}}^2$ takes the differential form:

$$\hat{\vec{L}}^2 = -\hbar^2\left[\frac{1}{\sin\theta}\frac{\partial}{\partial\theta}\left(\sin\theta\frac{\partial}{\partial\theta}\right) + \frac{1}{\sin^2\theta}\frac{\partial^2}{\partial\phi^2}\right]$$

2. The Spherical Harmonics $Y_{lm}(\theta, \phi)$

The simultaneous eigenfunctions of $\hat{\vec{L}}^2$ and $\hat{L}_z$ are the spherical harmonics:

$$\hat{\vec{L}}^2 Y_{lm}(\theta, \phi) = l(l+1)\hbar^2 Y_{lm}(\theta, \phi), \quad \hat{L}_z Y_{lm}(\theta, \phi) = m\hbar Y_{lm}(\theta, \phi)$$

Separating variables, $Y_{lm}(\theta, \phi) = \Theta_{lm}(\theta)\Phi_m(\phi)$:

$$\hat{L}_z \Phi_m(\phi) = -i\hbar\frac{d\Phi_m}{d\phi} = m\hbar\Phi_m(\phi) \implies \Phi_m(\phi) = \frac{1}{\sqrt{2\pi}}e^{im\phi}$$

The single-valuedness requirement $\Phi_m(\phi + 2\pi) = \Phi_m(\phi)$ restricts $m$ strictly to integers ($m \in \mathbb{Z}$). Consequently, orbital angular momentum quantum numbers must be strictly integer-valued:

$$l \in \{0, 1, 2, 3, \dots\}, \quad m \in \{-l, -l+1, \dots, +l\}$$

The $\theta$-dependent factor solves the associated Legendre differential equation:

$$Y_{lm}(\theta, \phi) = (-1)^m \sqrt{\frac{2l+1}{4\pi}\frac{(l-m)!}{(l+m)!}} P_l^m(\cos\theta) e^{im\phi}$$

Orthonormality over the unit sphere is:

$$\int_0^{2\pi} d\phi \int_0^\pi \sin\theta\,d\theta \, Y_{l'm'}^*(\theta, \phi) Y_{lm}(\theta, \phi) = \delta_{l'l}\delta_{m'm}$$

1. The Spin-1/2 Degree of Freedom

Unlike orbital angular momentum, which describes spatial motion and is restricted to integer eigenvalues, elementary particles possess an intrinsic angular momentum termed spin $\hat{\vec{S}}$ with half-integer quantum numbers. For electrons, protons, and neutrons, $s = 1/2$.

In the two-dimensional Hilbert space spanned by the basis states $\{|+\rangle, |-\rangle\} \equiv \{|\uparrow\rangle, |\downarrow\rangle\}$, the spin operator is represented in terms of the dimensionless Pauli spin matrices $\vec{\sigma} = (\sigma_x, \sigma_y, \sigma_z)$:

$$\hat{\vec{S}} = \frac{\hbar}{2}\vec{\sigma}$$ $$\sigma_x = \begin{pmatrix} 0 & 1 \\ 1 & 0 \end{pmatrix}, \quad \sigma_y = \begin{pmatrix} 0 & -i \\ i & 0 \end{pmatrix}, \quad \sigma_z = \begin{pmatrix} 1 & 0 \\ 0 & -1 \end{pmatrix}$$

2. Algebraic Properties of Pauli Matrices

The Pauli matrices satisfy the fundamental product identity:

$$\sigma_i \sigma_j = \delta_{ij}\hat{I} + i\sum_k \epsilon_{ijk} \sigma_k$$

From this identity directly follow their anticommutation and commutation relations:

$$\{\sigma_i, \sigma_j\} = \sigma_i \sigma_j + \sigma_j \sigma_i = 2\delta_{ij}\hat{I}$$ $$[\sigma_i, \sigma_j] = \sigma_i \sigma_j - \sigma_j \sigma_i = 2i\sum_k \epsilon_{ijk} \sigma_k$$

Additionally: $\sigma_i^2 = \hat{I}$, $\text{Tr}(\sigma_i) = 0$, and $\det(\sigma_i) = -1$.

3. Finite Rotations and the 4π Spinor Symmetry

A finite rotation of a spin-1/2 state by angle $\theta$ about unit axis $\hat{n}$ is represented by the $2\times 2$ unitary matrix $\hat{U}(\hat{n}, \theta) = \exp\left(-\frac{i}{\hbar}\hat{\vec{S}}\cdot\hat{n}\theta\right) = \exp\left(-\frac{i\theta}{2}(\vec{\sigma}\cdot\hat{n})\right)$. Expanding the matrix exponential:

$$\exp\left(-\frac{i\theta}{2}(\vec{\sigma}\cdot\hat{n})\right) = \sum_{k=0}^\infty \frac{1}{k!}\left(-\frac{i\theta}{2}\right)^k (\vec{\sigma}\cdot\hat{n})^k$$

Since $(\vec{\sigma}\cdot\hat{n})^2 = \hat{I}$, even powers yield $\cos(\theta/2)$ and odd powers yield $-i(\vec{\sigma}\cdot\hat{n})\sin(\theta/2)$:

$$\hat{U}(\hat{n}, \theta) = \cos\left(\frac{\theta}{2}\right)\hat{I} - i(\vec{\sigma}\cdot\hat{n})\sin\left(\frac{\theta}{2}\right)$$

For a complete $2\pi$ spatial rotation ($\theta = 2\pi$):

$$\hat{U}(\hat{n}, 2\pi) = \cos(\pi)\hat{I} - i(\vec{\sigma}\cdot\hat{n})\sin(\pi) = -\hat{I}$$

A rotation by $360^\circ$ inverts the sign of a fermion spinor ket: $|\psi\rangle \to -|\psi\rangle$. A full $4\pi$ ($720^\circ$) rotation is required to return the state vector to its original mathematical identity, demonstrating the double-covering topology $\text{SU}(2) \to \text{SO}(3)$ verified by neutron interferometry.

1. The Composite Two-Particle Hilbert Space

Consider a composite quantum system consisting of two independent angular momenta $\hat{\vec{J}}_1$ and $\hat{\vec{J}}_2$ acting in Hilbert spaces $\mathcal{H}_1$ and $\mathcal{H}_2$ of dimensions $(2j_1+1)$ and $(2j_2+1)$. Because they act on distinct degrees of freedom, their operators commute:

$$[\hat{J}_{1i}, \hat{J}_{2j}] = 0 \quad \forall i, j$$

The total angular momentum operator is the vector sum:

$$\hat{\vec{J}} = \hat{\vec{J}}_1 + \hat{\vec{J}}_2 \equiv \hat{\vec{J}}_1 \otimes \hat{I}_2 + \hat{I}_1 \otimes \hat{\vec{J}}_2$$

The total operator $\hat{\vec{J}}$ satisfies the standard rotation Lie algebra $[\hat{J}_i, \hat{J}_j] = i\hbar \epsilon_{ijk} \hat{J}_k$.

2. The Uncoupled Basis

The uncoupled basis diagonalizes the four mutually commuting operators $\{\hat{\vec{J}}_1^2, \hat{J}_{1z}, \hat{\vec{J}}_2^2, \hat{J}_{2z}\}$:

$$|j_1, m_1; j_2, m_2\rangle \equiv |j_1, m_1\rangle \otimes |j_2, m_2\rangle$$

The dimension of the product space is:

$$N = (2j_1 + 1)(2j_2 + 1)$$

3. The Coupled Basis and Triangle Inequality

Alternatively, the coupled basis diagonalizes the complete set of commuting observables $\{\hat{\vec{J}}_1^2, \hat{\vec{J}}_2^2, \hat{\vec{J}}^2, \hat{J}_z\}$, denoted by $|J, M; j_1, j_2\rangle$ (or simply $|J, M\rangle$ when $j_1, j_2$ are fixed):

$$\hat{\vec{J}}^2 |J, M\rangle = J(J+1)\hbar^2 |J, M\rangle, \quad \hat{J}_z |J, M\rangle = M\hbar |J, M\rangle$$

Since $\hat{J}_z = \hat{J}_{1z} + \hat{J}_{2z}$, the magnetic quantum number is strictly additive:

$$M = m_1 + m_2$$

The allowed values of total angular momentum $J$ satisfy the triangle rule:

$$|j_1 - j_2| \le J \le j_1 + j_2$$

with $J$ stepping by integers from $|j_1 - j_2|$ to $j_1 + j_2$. We verify conservation of total dimensionality:

$$\sum_{J=|j_1-j_2|}^{j_1+j_2} (2J + 1) = (2j_1 + 1)(2j_2 + 1)$$

1. Definition of Clebsch-Gordan Coefficients

The unitary transformation connecting the uncoupled and coupled bases is mediated by the Clebsch-Gordan (CG) coefficients:

$$|J, M\rangle = \sum_{m_1=-j_1}^{j_1} \sum_{m_2=-j_2}^{j_2} \langle j_1, m_1; j_2, m_2 | J, M \rangle |j_1, m_1; j_2, m_2\rangle$$

The CG coefficients $\langle j_1, m_1; j_2, m_2 | J, M \rangle$ are the projection overlaps between the two bases. They satisfy two fundamental selection rules:

1. $M = m_1 + m_2$ (otherwise the coefficient is identically zero).
2. $|j_1 - j_2| \le J \le j_1 + j_2$ (triangle inequality).

2. Orthonormality and Phase Conventions

Because the change of basis is a unitary transformation, the CG coefficients satisfy dual completeness relations:

$$\sum_{m_1, m_2} \langle J, M | j_1, m_1; j_2, m_2 \rangle \langle j_1, m_1; j_2, m_2 | J', M' \rangle = \delta_{JJ'}\delta_{MM'}$$ $$\sum_{J, M} \langle j_1, m_1; j_2, m_2 | J, M \rangle \langle J, M | j_1, m_1'; j_2, m_2' \rangle = \delta_{m_1 m_1'}\delta_{m_2 m_2'}$$

Under the standard Condon-Shortley phase convention:

1. All Clebsch-Gordan coefficients are chosen to be strictly real.
2. The highest state $\langle j_1, j_1; j_2, J - j_1 | J, J \rangle > 0$ is chosen to be strictly positive.

3. Construction via Lowering Ladder Operators

The construction begins by identifying the unique state of maximum total angular momentum:

$$|J_{\max}, M_{\max}\rangle = |j_1 + j_2, j_1 + j_2\rangle = |j_1, j_1; j_2, j_2\rangle$$

Applying the lowering operator $\hat{J}_- = \hat{J}_{1-} + \hat{J}_{2-}$ to both sides generates $|j_1+j_2, j_1+j_2-1\rangle$:

$$\hbar\sqrt{(j_1+j_2)(j_1+j_2+1) - (j_1+j_2)(j_1+j_2-1)}|j_1+j_2, j_1+j_2-1\rangle = (\hat{J}_{1-} + \hat{J}_{2-})|j_1, j_1; j_2, j_2\rangle$$ $$\sqrt{2(j_1+j_2)}|j_1+j_2, j_1+j_2-1\rangle = \sqrt{2j_1}|j_1-1, j_1; j_2, j_2\rangle + \sqrt{2j_2}|j_1, j_1; j_2-1, j_2\rangle$$

The orthogonal state with the same $M = j_1 + j_2 - 1$ corresponds to the state $|j_1+j_2-1, j_1+j_2-1\rangle$. Lowering operators are then iteratively applied to generate all lower $M$ states.

1. Spherical Tensor Operators

An irreducible spherical tensor operator of rank $k$, denoted by $\hat{T}_q^{(k)}$ (with $q = -k, -k+1, \dots, +k$), is defined by its commutation relations with the angular momentum operators:

$$[\hat{J}_z, \hat{T}_q^{(k)}] = q\hbar \hat{T}_q^{(k)}$$ $$[\hat{J}_\pm, \hat{T}_q^{(k)}] = \hbar\sqrt{k(k+1) - q(q \pm 1)} \, \hat{T}_{q \pm 1}^{(k)}$$

Under spatial rotations, the $2k+1$ components of $\hat{T}_q^{(k)}$ transform among themselves identically to the angular momentum eigenstates $|k, q\rangle$.

Examples:

• Rank 0 (Scalar Operator $\hat{S}$): $[\hat{\vec{J}}, \hat{S}] = 0$ (e.g., $\hat{\vec{r}}^2, \hat{\vec{p}}^2, \hat{H}$).
• Rank 1 (Vector Operator $\hat{\vec{V}}$): Related to Cartesian components by $\hat{T}_0^{(1)} = \hat{V}_z, \hat{T}_{\pm 1}^{(1)} = \mp \frac{\hat{V}_x \pm i\hat{V}_y}{\sqrt{2}}$.
• Rank 2 (Quadrupole Tensor): Electric quadrupole moments $Q_{ij} = 3x_i x_j - r^2 \delta_{ij}$.

2. The Wigner-Eckart Theorem

The Wigner-Eckart Theorem states that the matrix element of an irreducible spherical tensor operator between angular momentum eigenstates factors into two independent terms:

$$\langle j', m' | \hat{T}_q^{(k)} | j, m \rangle = \frac{\langle j, m; k, q | j', m' \rangle}{\sqrt{2j' + 1}} \langle j' \| \hat{T}^{(k)} \| j \rangle$$

where:

1. $\langle j, m; k, q | j', m' \rangle$ is a standard Clebsch-Gordan coefficient carrying all dependence on the magnetic projection quantum numbers $m, m', q$ (geometric orientation).
2. $\langle j' \| \hat{T}^{(k)} \| j \rangle$ is the reduced matrix element, which depends solely on the physical dynamics and radial integrals, completely independent of $m, m', q$.

3. Selection Rules from the Wigner-Eckart Theorem

Because the CG coefficient vanishes unless the coupling conditions are met:

1. $m' = m + q \iff \Delta m = q$.
2. $|j - k| \le j' \le j + k$ (triangle rule between initial angular momentum $j$, tensor rank $k$, and final angular momentum $j'$).

For a scalar operator ($k = 0, q = 0$): $\langle j', m' | \hat{S} | j, m \rangle \propto \delta_{j'j}\delta_{m'm}$.