Chapter 14 Spin

14.1 Introduction

14.2 What is the Nature of Spin?

14.3 Kinematics of Spin

Exercise 14.3.1 Let us verify the above corollary explicitly. Take some spinor with components $\alpha=\rho_1 \mathrm{e}^{\mathrm{i} \phi_1}$ and $\beta=\rho_2 \mathrm{e}^{\mathrm{i} \phi_2}$. From $\langle\chi \mid \chi\rangle=1$, deduce that we can write $\rho_1=\cos (\theta / 2)$ and $\rho_2=\sin (\theta / 2)$ for some $\theta$. Next pull out a common phase factor so that the spinor takes the form in Eq. (14.3.28a). This verifies the corollary and also fixes $\hat{n}$.

Exercise 14.3.2 (1) Show that the eigenvectors of $\boldsymbol{\sigma}\cdot\hat{n}$ are given by Eq. (14.3.28).

(2) Verify Eq. (14.3.29).

Exercise 14.3.3 Using Eqs. (14.3.32) and (14.3.33) show that the Pauli matrices are traceless.

Exercise 14.3.4 Derive Eq. (14.3.39) in two different ways.

(1) Write $\sigma_{1}\sigma_{2}$ in terms of $[\sigma_{i},\sigma_{j}]_{+}$ and $[\sigma_{i},\sigma_{j}]$.

(2) Use Eqs. (14.3.42) and (14.3.43).

Exercise 14.3.5 Express the following matrix $M$ in terms of the Pauli matrices:

Exercise 14.3.6 (1) Argue that $|\hat{n},+\rangle=U[R(\phi \mathbf{k})] U[R(\theta \mathbf{j})]\left|s_z=\hbar / 2\right\rangle$.

(2) Verify by explicit calculation.

Exercise 14.3.7 Express the following as linear combinations of the Pauli matrices and $I$:

(1) $\left(I+\mathrm{i} \sigma_x\right)^{1 / 2}$. (Relate it to half a certain rotation.)

(2) $\left(2 I+\sigma_x\right)^{-1}$.

(3) $\sigma_x^{-1}$.

Exercise 14.3.8 (1) Show that any matrix that commutes with $\boldsymbol{\sigma}$ is a multiple of the unit matrix.

(2) Show that we cannot find a matrix that anticommutes with all three Pauli matrices. (If such a matrix exists, it must equal zero.)

14.4 Spin Dynamics

Exercise 14.4.1 Show that if $H=-\gamma \mathbf{L} \cdot \mathbf{B}$, and $\mathbf{B}$ is position independent,

Exercise 14.4.2 Derive (14.4.31) by studying Fig. 14.3.

Exercise 14.4.3 We would like to study here the evolution of a state that starts out as $\binom{1}{0}$ and is subject to the $\mathbf{B}$ field given in Eq. (14.4.27). This state obeys

where $H=-\gamma \mathbf{S} \cdot \mathbf{B}$, and $\mathbf{B}$ is time dependent. Since classical reasoning suggests that in a frame rotating at frequency ($-\omega \mathbf{k}$) the Hamiltonian should be time independent and governed by $\mathbf{B}_r$ [Eq. (14.4.29)], consider the ket in the rotating frame, $\left|\psi_r(t)\right\rangle$, related to $|\psi(t)\rangle$ by a rotation angle $\omega t$:

Combine Eqs. (14.4.34) and (14.4.35) to derive Schrödinger's equation for $\left|\psi_r(t)\right\rangle$ in the $S_z$ basis and verify that the classical expectation is borne out. Solve for $\left|\psi_r(t)\right\rangle=U_r(t)\left|\psi_r(0)\right\rangle$ by computing $U_r(t)$, the propagator in the rotating frame. Rotate back to the lab and show that

Compare this to the state $|\hat{\mathbf{n}},+\rangle$ and see what is happening to the spin for the case $\omega_0=\omega$.

Calculate $\left\langle\mu_z(t)\right\rangle$ and verify that it agrees with Eq. (14.4.31).

Exercise 14.4.4 At $t=0$, an electron is in the state with $s_z=\hbar/2$. A steady field $\mathbf{B}=B \mathbf{i}$, $B=100 \mathrm{G}$, is turned on. How many seconds will it take for the spin to flip?

Exercise 14.4.5 We would like to establish the validity of Eq. (14.4.26) when $\boldsymbol{\omega}$ and $\mathbf{B}_0$ are not parallel.

(1) Consider a vector $\mathbf{V}$ in the inertial (nonrotating) frame which changes by $\Delta \mathbf{V}$ in a time $\Delta t$. Argue, using the results from Exercise 12.4.3, that the change as seen in a frame rotating at an angular velocity $\boldsymbol{\omega}$, is $\Delta \mathbf{V}-\boldsymbol{\omega} \times \mathbf{V} \Delta t$. Obtain a relation between the time derivatives of $\mathbf{V}$ in the two frames.

(2) Apply this result to the case of $\mathbf{l}$ [Eq. (14.4.8)], and deduce the formula for the effective field in the rotating frame.

Exercise 14.4.6 (A Density Matrix Problem) (1) Show that the density matrix for an ensemble of spin-1/2 particles may be written as

where $\mathbf{a}$ is a $c$-number vector.

(2) Show that $\mathbf{a}$ is the mean polarization, $\langle\overline{\boldsymbol{\sigma}}\rangle$.

(3) An ensemble of electrons in a magnetic field $\mathbf{B}=B \mathbf{k}$, is in thermal equilibrium at temperature $T$. Construct the density matrix for this ensemble. Calculate $\langle\overline{\boldsymbol{\mu}}\rangle$.

14.5 Return of Orbital Degrees of Freedom

Exercise 14.5.1 (1) Why is the coupling of the proton's intrinsic moment to B an order $m / M$ correction to Eq. (14.5.4)?

(2) Why is the coupling of its orbital motion an order $(m / M)^2$ correction? (You may reason classically in both parts.)

Exercise 14.5.2 (1) Estimate the relative size of the level splitting in the $n=1$ state to the unperturbed energy of the $n=1$ state, when a field $\mathbf{B}=1000 \mathrm{kG}$ is applied.

(2) Recall that we have been neglecting the order $B^2$ term in $H$. Estimate its contribution in the $n=1$ state relative to the linear ($-\boldsymbol{\mu} \cdot \mathbf{B}$) term we have kept, by assuming the electron moves on a classical orbit of radius $a_0$. Above what $|\mathbf{B}|$ does it begin to be a poor approximation?

Exercise 14.5.3 A beam of spin-1/2 particles moving along the $y$ axis goes through two collinear SG apparatuses, both with lower beams blocked. The first has its B field along the $z$ axis and the second has its $\mathbf{B}$ field along the $x$ axis (i.e., is obtained by rotating the first by an angle $\pi / 2$ about the $y$ axis). What fraction of particles leaving the first will exit the second? If a third filter that transmits only spin up along the $z$ axis is introduced, what fraction of particles leaving the first will exit the third? If the middle filter transmits both spins up and down (no blocking) the $x$ axis, but the last one transmits only spin down the $z$ axis, what fraction of particles leaving the first will leave the last?

Exercise 14.5.4 A beam of spin-1 particles, moving along the $y$ axis, is incident on two collinear SG apparatuses, the first with $\mathbf{B}$ along the $z$ axis and the second with $\mathbf{B}$ along the $z^{\prime}$ axis, which lies in the $x-z$ plane at an angle $\theta$ relative to the $z$ axis. Both apparatuses transmit only the uppermost beams. What fraction leaving the first will pass the second?