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!

Another problem - time dependent perturbation and transition probabilities

  1. Nov 28, 2007 #1
    1. The problem statement, all variables and given/known data

    (the actual question is now as an attachment)

    Assuming that the perturbation V(x,t)=[tex]\beta[/tex]x exp(-[tex]\gamma[/tex] t) is applied at t = 0 to a harmonic oscillator (HO) in the ground state, calculate to the first approximation the transition probabilty to any excited state n[tex]\geq[/tex]1 (here beta and gamma are constants). HO eigenstates are expressed in terms of the Hermite polynomials, H[tex]_{n}[/tex](y), as [tex]\left\lfloor[/tex]n[tex]\right\rangle[/tex]=[tex]\left([/tex][tex]\alpha[/tex]/(2[tex]^{2}[/tex] n! [tex]\sqrt{\pi}[/tex][tex]\right)[/tex][tex]^{1/2}[/tex] e[tex]^{-\alpha^{2} x^{2}}[/tex] H[tex]_{n}[/tex]([tex]\alpha[/tex]x) with [tex]\alpha[/tex]=[tex]\sqrt{M\varpi / \hbar}[/tex] , M is the mass of the oscillator, and [tex]\varpi[/tex] is the frequency. Use the recursion relation 2yH[tex]_{n}[/tex] = H[tex]_{n+1}[/tex] (y) = 2nH[tex]_{n-1}[/tex], and orthogonality of different eigenstates.

    2. Relevant equations
    That is the question, but what i am struggling on this question is, where to start, lol. I am completely lost as to what the recursion relation is, and how to go about finding the transition probability. Please, any help would be great, maybe a little more advise on what to read or what steps i need to take to find this probability.

    3. The attempt at a solution
    As i said, i not had much/any luck with this. i am truly struggling. I am aware the question is part of a time dependent perturbation and i am looking to find a probability of an electron making a jump from one energy level to another, but there i find myself scratching my head.

    ANy help or advise on how to approach the problem would be greatfully recieved.

    Thanks for your help...again EK

    Attached Files:

    Last edited: Nov 28, 2007
  2. jcsd
  3. Nov 28, 2007 #2


    User Avatar
    Science Advisor

    Do you know a formula for the transition probability? That would be a good place to start.
  4. Nov 28, 2007 #3
    I have the equation for the probability, "Fermi goldern rule" But the matrix element through me a little. is that the expectation value for the event occurance?

    Thanks for your help EK
  5. Nov 28, 2007 #4


    User Avatar
    Science Advisor

    This is not a "Golden Rule" problem, you need a more basic formula of time-dependent perturbation theory.

    But it will involve a matrix element of V(x,t) between the initial state (the ground state) and the final state (an energy eigenstate labeled by n). Since V(x,t) is proportional to x, you need to compute [itex]\langle n|x|0\rangle[/itex]. Any idea how to use the hints you were given to do that?
  6. Nov 29, 2007 #5
    Not really, i have not encountered the recursion relation, and am fairly confused about that. Is the eqauation for the hamiltonian just the [ hamiltonian for the stationary state + a the pertibation ] ? I am truly at a loss atm. Thanks

  7. Nov 29, 2007 #6


    User Avatar
    Science Advisor

    You may be confused by the Dirac bra-ket notation. It means
    [tex]\langle n|x|0\rangle = \int_{-\infty}^{+\infty}dx\,\psi^*_n(x)x\psi_0(x)[/tex]
    Of course, you know that, without the factor of x in the middle, the energy eigenfucntions are orthogonal:
    [tex]\langle n|n'\rangle = \int_{-\infty}^{+\infty}dx\,\psi^*_n(x)\psi_{n'}(x)=\delta_{nn'}[/tex]
    Perhaps now you can see how the recursion relations helps ...
  8. Nov 29, 2007 #7
    I understand the Kronecker and the dirac notation, but the recursion relation still has me. so far on this problem, i am thinking about the Probability being
    [tex]\frac{\left|H''\right|^2}{2 \hbar^2} g(w)t[/tex]
    where g(w) being the density of states.
    And in this case, we only deal with the ground state and the 1st state, right? As from that relation that would only give me one part of that equation. I.e. [tex]H_{0}[/tex] gives only the [tex]H_{n+1}[/tex] term?

    Thanks for your continued help.

  9. Nov 29, 2007 #8
    I'm doing the same problem - Erwin, are you at Loughborough?

    The problem is, using the orthogonality conditions and the recursion relations...

    Basically, you get an integral (I think) of this form:

    INT(-inf..inf) exp(-2a^2 * x^2) Hn Ho dx

    Even if this is the correct method, I don't know what to do next.

    PS - Apologies for the ASCII maths, I'm re-downloading LaTeX presently.
  10. Nov 29, 2007 #9

    Yeah Loughborough, i am pretty sure you take the matrix element of perturbation, at t=0, then find the constant [tex]C_{n}[/tex] with that. Then from that you can find the probability. Maybe...

    That or you go straight from [tex]V_{nk}[/tex] to probability


  11. Nov 29, 2007 #10


    User Avatar
    Science Advisor

    Can you use the recursion relation to express [itex]x\psi_0(x)[/itex] in terms of other energy eigenstates?

    And [tex]\frac{\left|H''\right|^2}{2 \hbar^2} g(w)t[/tex] is not correct. This is a "golden rule" formula that does not apply to this problem.
  12. Nov 29, 2007 #11
    I am lost as to how to use the recursion relation to be honest. I am thinking you use this to find the wave function of the problem. Is there any good internet sites or books in which the recursion relation and hermite polynomials are described fully? I am struggling to find out how to use them.

    Thank you EK
  13. Nov 30, 2007 #12
    Using Hermite, xH1 = (1/2)H2 + nHo

    But that doesn't really help.
  14. Nov 30, 2007 #13
    Thanks for your help Avodyne, i think i was being a bit retarded. now i know what i am looking for using the recursion relation, i should be fine. I think i got an answer for the solution, and its less that 1 so i am happy that its sensible. Thank you again.

  15. Nov 30, 2007 #14


    User Avatar
    Science Advisor

    Dakes: it helps a lot! See my post #6.

    EK: you're welcome!
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: Another problem - time dependent perturbation and transition probabilities
  1. Transition Probability (Replies: 6)

  2. Transition Probability (Replies: 2)