Griffiths Quantum Mechanics 2nd Edition: Problems 4.33

We are first asked to construct the Hamiltonian matrix for the system in question. This is straightforward, as Griffiths tells us that for an electron at (transational) rest in a magentic field, the Hamiltonian is of the form:

$$H \ = \ -\gamma \textbf{B} \ \cdot \ \textbf{S}$$
This means that for $\textbf{B} \ = \ B_0 \cos(\omega t)\hat{k}$, we have:

$$H \ = \ -\gamma B_0 \cos(\omega t)\hat{k} \ \cdot \ \textbf{S} \ = \ -\gamma B_0 \cos(\omega t) \textbf{S}_z \ = \ -\frac{\gamma B_0 \cos(\omega t) \hbar}{2} \ \begin{pmatrix} 1 & 0 \\ 0 & -1 \end{pmatrix}$$
Next, we are asked to find the state of the particle, $\chi(t)$ if we are given that $\chi(0) \ = \ \chi^{(x)}_{+}$, where $\chi^{(x)}_{+}$ is the spin-up basis spinor when spin is measured with respect to the $x$-axis. In order to find $\chi(t)$, let us recall that the Schrödinger equation tells us:

$$i\hbar \frac{\partial \chi}{\partial t} \ = \ \hat{H} \chi$$
We can express $\chi$ in terms of basis spinors:

$$\chi \ = \ a\chi_{+} \ + \ b\chi_{-}$$
Thus, we have:

$$i\hbar \frac{\partial}{\partial t} \ (a\chi_{+} \ + \ b\chi_{-}) \ = \ i\hbar \Big(\frac{\partial}{\partial t} \ a\chi_{+} \ + \ \frac{\partial}{\partial t} \ b\chi_{-} \Big) \ = \ \hat{H} (a\chi_{+} \ + \ b\chi_{-}) \ = \ \hat{H} a\chi_{+} \ + \ \hat{H} b\chi_{-}$$ $$\Rightarrow \ \hat{H} a\chi_{+} \ + \ \hat{H} b\chi_{-} \ = \ -\frac{\gamma B_0 \cos(\omega t) \hbar}{2} \Big( a\chi_{+} \ - \ b\chi_{-} \Big)$$ $$\Rightarrow \ i\hbar \begin{pmatrix} \frac{\partial a}{\partial t} \\ \frac{\partial b}{\partial t} \end{pmatrix} \ = \ -\frac{\gamma B_0 \cos(\omega t) \hbar}{2} \begin{pmatrix} a \\ -b \end{pmatrix}$$
So this means that we have two differential equations (as the basis spinors are linearly independent):

$$\Rightarrow \ i\hbar \frac{\partial a}{\partial t} \ = \ -\frac{\gamma B_0 \cos(\omega t) \hbar}{2} a \ \Rightarrow \ \displaystyle\int \ \frac{1}{a} \ \text{d}a \ = \ \displaystyle\int \ \frac{i\gamma B_0}{2} \ \cos(\omega t) \text{d}t \ \Rightarrow \ \ln(a) \ + \ A \ = \ \frac{i\gamma B_0}{2\omega} \sin(\omega t) \ + \ C \ \Rightarrow \ a(t) \ = \ Ee^{\frac{i\gamma B_0}{2\omega} \sin(\omega t)}$$ $$\Rightarrow \ i\hbar \frac{\partial b}{\partial t} \ = \ \frac{\gamma B_0 \cos(\omega t) \hbar}{2} b \ \Rightarrow \ \displaystyle\int \ \frac{1}{b} \ \text{d}b \ = \ \displaystyle\int \ -\frac{i\gamma B_0}{2} \ \cos(\omega t) \text{d}t \ \Rightarrow \ \ln(b) \ + \ A \ = \ -\frac{i\gamma B_0}{2\omega} \sin(\omega t) \ + \ C \ \Rightarrow \ b(t) \ = \ Fe^{-\frac{i\gamma B_0}{2\omega} \sin(\omega t)}$$
So this means that we have:

$$\chi(t) \ = \ \begin{pmatrix} Ee^{\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \\ Fe^{-\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \end{pmatrix}$$
We determine the coefficients $E$ and $F$ by the initial condition that:

$$\chi(0) \ = \ \chi_{+}^{(x)} \ = \begin{pmatrix} \frac{1}{\sqrt{2}} \\ \frac{1}{\sqrt{2}} \end{pmatrix} \ = \ \begin{pmatrix} Ee^{\frac{i\gamma B_0}{2\omega} \sin(\omega (0))} \\ Fe^{-\frac{i\gamma B_0}{2\omega} \sin(\omega (0))} \end{pmatrix} \ = \ \begin{pmatrix} E \\ F \end{pmatrix}$$
So $E \ = \ \frac{1}{\sqrt{2}}$ and $F \ = \ \frac{1}{\sqrt{2}}$, and we have:

$$\chi(t) \ = \ \begin{pmatrix} \frac{1}{\sqrt{2}}e^{\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \\ \frac{1}{\sqrt{2}}e^{-\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \end{pmatrix}$$
To find the probability of getting $\hbar/2$ when measuring $\textbf{S}_x$, we simply have to find the number when we take the inner product of $\chi(t)$ with $\chi^{(x)}_{-}$, which has eigenvalue $\hbar/2$ (sort of like a scalar projection onto the basis vector that corresponds to the eigenvalue which we want to find the probability of measuring):

$$P \ = \ \langle \chi^{(x)}_{-} | \chi(t) \rangle \ = \ \begin{pmatrix} \frac{1}{\sqrt{2}} \\ -\frac{1}{\sqrt{2}} \end{pmatrix} \ \cdot \ \begin{pmatrix} \frac{1}{\sqrt{2}}e^{\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \\ \frac{1}{\sqrt{2}}e^{-\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \end{pmatrix} \ = \ \frac{1}{2} \ \Big( e^{\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \ + \ e^{-\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \Big)$$
Using Euler's formula, and the fact that $\cos(\theta) \ = \ -\cos(\theta)$ and $\sin(-\theta) \ = \ -\sin(\theta)$:

$$\frac{1}{2} \ \Big( e^{\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \ + \ e^{-\frac{i\gamma B_0}{2\omega} \sin(\omega t)} \Big) \ = \ i \sin \Big( \frac{\gamma B_0}{2 \omega} \ \sin(\omega t) \Big)$$
So, to find the probability, we simple find the value of $|P|^2$, which gives us:

$$|P|^2 \ = \ \sin^{2} \Big( \frac{\gamma B_0}{2 \omega} \ \sin(\omega t) \Big)$$
Finally, we are asked the minimum field $B_0$ required to completely flip our particle from its initial state of $\chi^{(x)}_{+}$ to $\chi^{(x)}_{-}$. Well, in order for this to happen with completely, the probability of the particle being in state $\chi^{(x)}_{-}$ has to be equal to $1$ at some time $t$ (the particle won't remain in this state as the magnetic field is oscillating and the probability evolves with time). This means that:

$$\sin^{2} \Big( \frac{\gamma B_0}{2 \omega} \ \sin(\omega t) \Big) \ = \ 1 \ \Rightarrow \ \sin \Big( \frac{\gamma B_0}{2 \omega} \ \sin(\omega t) \Big) \ = \ \pm 1$$
Well, the lowest value of $\frac{\gamma B_0}{2 \omega} \ \sin(\omega t)$ where this would occur would be when $\frac{\gamma B_0}{2 \omega} \ \sin(\omega t) \ = \ \frac{\pi}{2}$. Now, $\sin(\omega t)$ will constantly be oscillating back and forth between $-1$ and $1$, so it would make sense for us to choose the highest possible value of $\sin (\omega t)$, therefore getting the lowest possible value of $\frac{\gamma B_0}{2 \omega}$ and therefore $B_0$ (we're going to discard negative numbers, as when we are talking about the strength of a field, denoted as $B_0$, we think about it as some positive magnitude). This means that we set $\sin(\omega t) \ = \ 1$, meaning that:

$$\frac{\gamma B_0}{2 \omega} \ = \ \frac{\pi}{2} \ \Rightarrow \ B_0 (\text{minimal}) \ = \ \frac{\pi \omega}{\gamma}$$