3
$\begingroup$

Assume a spin s= 1/2 is subjected to an external magnetic field $B=Be_z$. The Hamiltonian is then given by $$ \hat{H} = -\frac{eB}{mc} \hat{S}_z = w\hat{S_z} $$ and that at t= 0, the spin of the particle is in the eigenstate of the $S_x$ operator with the eigenvalue $ \hbar/2 $, e.g. $$ \hat{S}_x|\psi (t= 0)⟩= \frac{\hbar}{2}|\psi(t= 0)⟩.$$

So first I derived the expressions for the dynamics of the spin operators and got: $$ \frac{d\hat{S}_y^H}{dt} = iw\hat{S}_x^H $$ $$ \frac{d\hat{S}_x^H}{dt} = {-iw} \hat{S}_y^H $$ $$ \frac{d\hat{S}_z^H}{dt} = 0 $$

Now I want to calculate the time-dependence of the expectation values $ ⟨ \hat{S}_x ⟩ $, $ ⟨ \hat{S}_y ⟩ $ and $ ⟨ \hat{S}_z ⟩ $. To do that I used Ehrenfest theorem (for an arbitrary $S_i$:

$$ \frac{d}{dt} ⟨ S_i ⟩_H = \frac{1}{i\hbar} ⟨ [ \hat{S}_i , \hat{H} ] ⟩ + ⟨ \frac{\partial S_i }{dt} ⟩. $$

Starting with the first term: $$ ⟨ [ \hat{S}_i , \hat{H} ] ⟩ = ⟨ {S}_i \hat{H} ⟩ - ⟨ \hat{H} \hat{S}_i ⟩ = w ( ⟨ \hat{U}^{\dagger} \hat{S}_i \hat{U} \hat{U}^{\dagger} \hat{S}_z \hat{U} ⟩ - ⟨ \hat{U}^{\dagger} \hat{S}_z \hat{U} \hat{U}^{\dagger} \hat{S}_i \hat{U} ⟩ ) =$$ $$ = w ( ⟨ \hat{U}^{\dagger} \hat{S}_i \hat{S}_z \hat{U} ⟩ - ⟨ \hat{U}^{\dagger} \hat{S}_z \hat{S}_i \hat{U} ⟩ ) $$

So my question is what is the best/easiest way to go now? I tried using the definition of expectation value and the fact that the state at t=0 is ( changing to z-basis ):

$$ |\psi (t= 0)⟩ = |+⟩_x = \frac{1}{\sqrt{2}} ( |+⟩_z + |-⟩_z ) $$

So that

$$ ⟨ [ \hat{S}_i , \hat{H} ] ⟩ = \frac{w}{2} \Bigl( ⟨ ⟨+|_z + ⟨-|_z | e^{-iw\hat{S}_zt} \hat{S}_i \hat{S}_z e^{iw\hat{S}_zt} | |+⟩_z + |-⟩_z ⟩ \Bigr) $$

But I don't really know how to continue here to find the expectation value of the exponential term with t-dependence! Any advice appreciated:)

$\endgroup$
6
  • $\begingroup$ Why not using the Schrödinger pciture, where the operators do not evolve in time, while the states do? In this case the time evolution of the state is $|\psi(t)\rangle=\exp{(-iHt)}|\psi(t=0)\rangle$. If you accept this solution, I can elaborate more on a dedicated answer :) $\endgroup$
    – Matteo
    Commented Feb 21 at 13:30
  • $\begingroup$ In the Heisenberg picture you are using, Ehrenfest's theorem is a triviality. You can already establish, by inspection, that $\hat S_x^H$ and $\hat S_y^H$ rotate into each other with frequency $w /\hbar$, in your highly nonstandard notation, so you are done. $\endgroup$ Commented Feb 21 at 14:58
  • $\begingroup$ @Matteo Of course! I'm open to try multiple approaches:) but would it be easier to use Schrödinger picture in this case? $\endgroup$
    – ilra
    Commented Feb 21 at 16:27
  • $\begingroup$ @CosmasZachos How could you see that the rotate into each other with this frequency? and how is this related to the time dependence of the expectation value of them? $\endgroup$
    – ilra
    Commented Feb 21 at 16:31
  • $\begingroup$ The frequency is actually w; I set $\hbar=1$ and garbled its reintroduction. See answer. $\endgroup$ Commented Feb 21 at 18:00

2 Answers 2

3
$\begingroup$

In the Schrödinger picture, you can write the Hamiltonian operator in the basis of the eigenstates of $\hat{S}_z$: $|+\rangle_z=\left(\begin{array}{c}1\\0\end{array}\right)$ and $|-\rangle_z=\left(\begin{array}{c}0\\1\end{array}\right)$ as $$ H = \frac{\hbar w}{2} \left(\begin{array}{cc}1 & 0\\0 & -1\end{array}\right) $$ and the initial state as $$ |\psi(t=0)\rangle = \frac{1}{\sqrt{2}} \left( |+\rangle_z + |-\rangle_z \right) = \frac{1}{\sqrt{2}} \left(\begin{array}{c}1\\1\end{array}\right). $$

Now from the Schrödinger equation we get the formula for the time evolution of a state: $$ i\hbar \partial_t \psi = H \psi \rightarrow |\psi(t)\rangle = \exp{\left[-iHt/\hbar\right]} |\psi(t=0)\rangle. $$ The matrix exponential of a diagonal matrix is trivial: you just exponentiate every diagonal term, so: $$ |\psi(t)\rangle = \left(\begin{array}{cc} e^{-iwt/2} & 0 \\ 0 & e^{iwt/2} \end{array}\right) \frac{1}{\sqrt{2}} \left(\begin{array}{c}1\\1\end{array}\right) = \frac{1}{\sqrt{2}} \left(\begin{array}{c}e^{-iwt/2}\\e^{iwt/2}\end{array}\right). $$ Knowing the time evolved state, you can compute expectation values of any operator, including for instance $$\hat{S}_x = \frac{\hbar}{2} \left(\begin{array}{cc} 0 & 1 \\ 1 & 0 \end{array}\right) \;\;\;\;\;\;\;\; \hat{S}_y = \frac{\hbar}{2} \left(\begin{array}{cc} 0 & -i \\ i & 0 \end{array}\right) :$$ $$ \langle \psi(t) | \hat{S}_x | \psi(t) \rangle = \frac{\hbar}{2} \cos{(wt)} \;\;\;\;\;\;\; \langle \psi(t) | \hat{S}_y | \psi(t) \rangle = \frac{\hbar}{2} \sin{(wt)} $$

Hope this helps! :)

$\endgroup$
2
  • $\begingroup$ Ahh that was very helpful! Appreciate the help:) but I have one question about the matrix exponential, you replaced the H in the exponent with w instead of wS_z, should not S_z also be in the matrix representation of the time operator? $\endgroup$
    – ilra
    Commented Feb 21 at 19:14
  • $\begingroup$ I have actually computed $\exp{(-iHt/\hbar)} = \exp{(-i \frac{wt}{2} S_z)}$, then I have exponentiated the diagonal matrix $\mathrm{diag}{(-iwt/2, +iwt/2})$, which is a diagonal matrix having the exponential of the eigenvalues as diagonal elements $\endgroup$
    – Matteo
    Commented Feb 21 at 19:18
3
$\begingroup$

Since you are working in the Heisenberg picture where states are time-independent, you may automatically take expectation values of your operator equations, resulting in the same uniform rotation structure, $$ \frac{d \langle \hat{S}_y^H\rangle }{dt} = iw \langle \hat{S}_x^H\rangle , \\ \frac{d \langle \hat{S}_x^H\rangle }{dt} = - iw \langle \hat{S}_y^H\rangle ,\\ \frac{d \langle \hat{S}_z^H \rangle }{dt} = 0 .$$

In your notation, their evident solution is $$ \langle \hat{S}_y^H (t)\rangle = \cos( wt ) \langle \hat{S}_y^H(0)\rangle + \sin(wt ) \langle \hat{S}_x^H(0)\rangle , \\ \langle \hat{S}_x^H (t)\rangle = \cos(wt ) \langle \hat{S}_x^H(0)\rangle - \sin( wt ) \langle \hat{S}_y^H(0)\rangle , \\ \langle \hat{S}_z^H (t)\rangle = \langle \hat{S}_z^H (0)\rangle .$$

$\endgroup$
2
  • $\begingroup$ So I could've got the answer by integrating my equations!:D I totally missed that, thank you so much for your help sir!:) $\endgroup$
    – ilra
    Commented Feb 22 at 10:36
  • 1
    $\begingroup$ Indeed, Ehrenfest's theorem is a triviality in the Heisenberg picture... $\endgroup$ Commented Feb 22 at 14:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.