1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Introductory Quantum Mechanics exercise.

  1. Aug 19, 2016 #1
    1. The problem statement, all variables and given/known data
    We describe particle's movement with the Hamiltonian:
    [tex] H=- \frac{\Delta E}{2} |0\rangle \langle0| + \frac{\Delta E}{2} |1\rangle \langle1|, [/tex]

    where [itex] |0\rangle [/itex] and [itex] |1\rangle [/itex] are the ortonormal basis. Let:
    [tex] |a\rangle = \frac{1}{\sqrt{2}}|0\rangle +\frac{i}{\sqrt{2}} |1\rangle. [/tex]

    a) Find [itex]|b\rangle[/itex] state, so that it would form orthonormal basis of a Hilbert space with [itex]|a\rangle[/itex].
    b) Find eigenvalues and eigenstates of a projector [itex]P_b=|b \rangle \langle b|[/itex].
    c) Let a particle be in an [itex]|a \rangle[/itex] state at t=0. Find the time evolution of a wave function.
    d) At t>0 we do a measurement of an operator[itex]P_b[/itex]. What are possible results of a measurement and what are their chances?

    3. The attempt at a solution
    a) We can write:
    [tex] |b \rangle = A |0 \rangle + B |1 \rangle\\ \langle a|b \rangle=0 \\ \langle b|b \rangle=1.[/tex]
    We get:
    [tex] A=iB \\ |A|^2 + |B|^2=1,[/tex]
    and finally:
    [tex] |b \rangle = \frac{i}{\sqrt{2}} |0 \rangle +\frac{1}{\sqrt{2}} |1 \rangle. [/tex]

    b) I am not sure what to do here? Do I have to use a projector on states?
    [tex]P_b |b \rangle =|b \rangle \langle b|b \rangle [/tex]
    I get only b state (because [itex]\langle b|b \rangle=1[/itex]) as an eigenfunction (?), but I'm not sure what are the eigenvalues then?
    [tex]P_b |a \rangle =0 [/tex]

    c) [tex] | a,t \rangle = e^{-i \frac{H}{\hbar} t} | a,0 \rangle \\ H|0\rangle=-\frac{\Delta E}{2} |0\rangle \\ H|1\rangle= \frac{\Delta E}{2} |1\rangle \\ | a,t \rangle = \frac{1}{\sqrt{2}} e^{i \frac{\Delta E}{2 \hbar} t} |0 \rangle + \frac{i}{\sqrt{2}} e^{-i \frac{\Delta E}{2 \hbar} t} |1 \rangle [/tex]
    Is this OK?

    d) Here, I have got some problems. I am thinking - we can get 0, if the wave function is still in [itex] |a\rangle[/itex] state or (I don't know what) if the wave function gets in to [itex] |b\rangle[/itex] state.
    But on the other hand I think here must be something with [itex] |0\rangle[/itex] and [itex] |1\rangle[/itex] states, so that I use only square of absolute values of coefficients from c) for their chances (which are 50:50?).

    Thank you for your answers.
  2. jcsd
  3. Aug 19, 2016 #2
    Your part a) seems to be correct, although you've ignored a potential phase factor, making the general solution for [itex]|b\rangle=\frac{i}{\sqrt 2} e^{i\phi}+\frac{1}{\sqrt 2}e^{i\phi}[/itex] (not -really- an issue, just being pedantic here).

    For part b), you're on the right track: if [itex]P_b |b\rangle=|b\rangle[/itex], what does that tell you about the eigenvalues and the eigenvectors (hint: it's almost trivial)? A similar conclusion can be obtained from [itex]P_b|a\rangle=0[/itex].

    Part c) looks good.

    Regarding the last part, I think you can just rewrite state [itex]|a,t\rangle[/itex] in terms of states [itex]|a,0\rangle[/itex] and [itex]|b,0\rangle[/itex] instead, and then apply the projector to it and see which coefficients you obtain.
  4. Aug 21, 2016 #3
    a) Ok, but what does this phase factor mean? I imagine those two states [itex]|a\rangle[/itex] and [itex]|b\rangle[/itex] in a plane, and they are perpendicular (like x and y axis in coordinate plane). Does this only mean that they are randomly directed in a plane like on the next pic?

    b) I think I have got problems with this precisely because it is trivial and don't really understand what are eigenvalues and -vectors of such kind of operators. :)
    From the definition of an eigenvector and eigenvalue ([itex] A v=\lambda v[/itex], where [itex] A[/itex] is operator, [itex] v[/itex] is eigenvector and [itex] \lambda [/itex] is eigenvalue), I would say that in our case [itex]P_b |b\rangle=|b\rangle[/itex], eigenvalue is 1 and eigenvector is [itex]|b\rangle[/itex] and in [itex]P_b|a\rangle=0[/itex] eigenvalue is 0 and eigenvector still [itex]|a\rangle[/itex]? Or vector [itex]\vec{0}[/itex]?

    d) Big thanks! I rewrited:
    [tex]| a,t \rangle = A |a,0 \rangle + B |b,0 \rangle \\ =\frac{1}{\sqrt{2}}(A+iB)|0 \rangle +\frac{1}{\sqrt{2}} (iA+B)|1 \rangle, [/tex]
    from which I nicely get:
    [tex]| a,t \rangle = \cos(\frac{\Delta E}{2 \hbar}t) |a,0 \rangle + \sin(\frac{\Delta E}{2 \hbar}t) |b,0 \rangle[/tex]
    If I now apply [itex]P_b [/itex] to this, I get:
    [tex]P_b |a,t\rangle=0 |a,0\rangle + \sin(\frac{\Delta E}{2 \hbar}t) |b,0 \rangle.[/tex]
    So the results of a measurement are 0 or sine?
    Chances are just squares of absolute values of coefficients in [itex]|a,t\rangle[/itex] and the wave function collapses to the measured state?
  5. Aug 21, 2016 #4
    An absolute phase has no physical significance, it's a sort of a redundancy in the description of the state, and doesn't influence any observables, that is, expectation values of hermitian operators. However, a relative phase difference between states can be detected: see for instance the Aharonov-Bohm effect (Sakurai - Modern quantum mechanics (2010), ch. 2.7 has some info).

    You are right for part b), the eigenvalues of the operator are just 0 and 1, and the eigenvectors are [itex]|a\rangle[/itex] and [itex]|b\rangle[/itex] respectively (for the first one just note that [itex]P_b|a\rangle=0\cdot |a\rangle[/itex]). Note that the null vector cannot be an eigenvector by definition, since any linear map acting on it would result in 0 trivially.

    The last part is right, but it seems a bit strange that the probabilities do not add up to 1 at all times. However, the operator in question is a projector, so given a basis of a Hilbert space denoted as [itex]|c_i\rangle[/itex], the completeness relation is [itex]\sum\limits_i|c_i\rangle\langle c_i|=1[/itex], in your case being [itex]|a\rangle\langle a|+|b\rangle\langle b|=1[/itex], so nothing is violated, see Sakurai ch. 1.3 for more details.
  6. Aug 22, 2016 #5
    Why? Isn't the sum of squares of sine and cosine in [itex]| a,t \rangle[/itex] always 1?
    Thank you!

    Another question.
    We've got a free particle with spin 1/2. We mesure z-component of its spin and get [itex]\frac{\hbar}{2}[/itex].
    a) After this, we measure the x-component, too. What are possible results and what are their probabilities?
    b) What if we measure the component of its spin at angle [itex]\theta[/itex] about z axis?

    a) I am not sure here. Firstly, I would say that its spin wave function collapses after first measurement in z component so then we get 0 with 100%, but in other hand if we use [itex]S_x[/itex] operator, we get something:
    [tex]S_x |\uparrow \rangle = \frac{S_++S_-}{2} |\uparrow \rangle = \frac{\hbar}{2} |\downarrow \rangle[/tex]

    b) At lectures we said:
    [tex]| \psi \rangle = \cos(\frac{\theta}{2}) |\uparrow \rangle + \sin(\frac{\theta}{2}) e^{i \phi} |\downarrow \rangle[/tex]
  7. Aug 22, 2016 #6


    User Avatar
    Science Advisor
    Homework Helper

    Are you familiar with working with Pauli matrices? If yes, you can first try to find the eigenvectors of ##\sigma_x## in the basis of eigenvectors of ##\sigma_z##.
  8. Aug 22, 2016 #7
    \sigma_{x} = \left(
    0 & 1\\
    1 & 0
    \sigma_{3} = \left(
    1 & 0\\
    0 & -1
    So, the eigenvectors of [itex]\sigma_x[/itex] are [itex] x_1=
    [/itex] and [itex] x_2=
    [/itex]. The eigenvectors of[itex]\sigma_z[/itex] are [itex] x_3=
    [/itex] and [itex] x_4=
    So if I write eigenvectors of [itex]\sigma_x[/itex] in the basis of eigenvectors of [itex]\sigma_z[/itex], I get:
    What does these eigenvectors tell me?
  9. Aug 22, 2016 #8


    User Avatar
    Science Advisor
    Homework Helper

    Almost correct, you just forgot that a state vector must be normalized. After normalizing them, express ##x_3## (not ##x_4## because initially the state is known to be corresponding to the eigenvalue ##\hbar/2##) in terms of ##x_1## and ##x_2##. This will allow you to calculate the probability of obtaining the probabilities of the measurement as mentioned in your problem statement.
  10. Aug 22, 2016 #9
    [itex]x_1={1 \over \sqrt{2}}
    \right) [/itex] and
    [itex]x_2={1 \over \sqrt{2}}
    \right) [/itex]. From this, I get:
    [tex]x_3={1 \over 2} x_1 - {1 \over 2} x_2.[/tex]

    So, if I understand this correctly, possible results of a x-component measurement are again [itex]\pm \frac{\hbar}{2}[/itex] with the probabilities of square of the coefficients, so both with [itex]\frac{1}{4}[/itex] chance.
    Aha, so I would use [itex]x_4[/itex] if the result of a measurement would be [itex] -\hbar \over 2 [/itex] which is the other eigenvalue?
    But why do I have to express [itex]x_3[/itex] with [itex]x_1[/itex] and [itex]x_2[/itex]? -->
    Do I just understand this wrong? Because I first expressed just the opposite. :)
  11. Aug 22, 2016 #10


    User Avatar
    Science Advisor
    Homework Helper

    The coefficients are still off although their ratio is correct, remember any state vector must be normalized - the sum of the modulus square of each coefficient must add up to unity.
    Because you want to measure ##S_x## and the only possible outcomes are given by its eigenvalues. Therefore, you need to know how ##x_3## (the initial state) looks like when expressed in the basis of eigenvectors of ##S_x## (or ##\sigma_x##).
    No you didn't do something wrong. It's just because the most Pauli matrices are written in the basis of eigenvectors of ##\sigma_z##. Therefore, when you calculated the eigenvectors of ##\sigma_x## in this form you will automatically get these eigenvectors in basis of eigenvectors of ##\sigma_z##.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted