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: Delta potential well

  1. Oct 10, 2008 #1
    1. The problem statement, all variables and given/known data

    consider the scattering matrix for the potential

    2m/hbar2 V(x) = λ/a δ(x-b)

    show that it has the form

    (2ika/(2ika-λ) , (e-2kib) λ/(2ika-λ)
    (e2kib) λ/(2ika-λ) , 2ika/(2ika-λ)

    (I've used commas just to separate terms in the matrix)

    prove that it is unitary and that it will yield the condition for bound states when the elements of that matrix becoe infinite (this will only occur for λ < 0)

    2. Relevant equations

    suppose the matrix is expressed as

    S11 S12
    S21 S22

    where S11 = (2ika/(2ika-λ)
    S12 = (e-2kib) λ/(2ika-λ)
    S21 = (e2kib) λ/(2ika-λ)
    S22 = 2ika/(2ika-λ)

    3. The attempt at a solution

    I see that this is a delta potential well at x=b

    ok so I know that S11 = T S21= R S22 = T and S12 = R where T and R are the reflection and transmission coefficients so I figure that if I can find those then I show the s-matrix in the above form so here it goes...


    u(x) = Arekx +Bre-kx x < b

    = Ale-kx +Blekx x > b

    the boundary condition is (du/dx at x = b+) - (du/dx at x = b-) = λ/a u(b)


    k(Arekb -Bre-kb+Ale-kb -Blekb)= λ/a u(b)

    then for an incoming particle that can be either reflected or transmitted I make Ar= 1 Al = r Bl=0 and Br=t
    where r2 = R (reflection coefficient and t2= t (transmission coefficient)

    to get

    ekb-te-kb + re-kb= λ/a u(b)

    so how do I solve for r and t separately and how do I get rid of the u(b)?

    Thank you

  2. jcsd
  3. Oct 11, 2008 #2


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Aren't you missing a bunch of "i"s in your wave-function exponents?

    What is the potential for [itex]x \neq b[/itex]? What are the wave-functions in such a region of space?

    PS: These forums support [itex]\LaTeX[/itex]. See this thread: https://www.physicsforums.com/showthread.php?t=8997 or click on my tex to see the code.
  4. Oct 11, 2008 #3
    Thank you for replying to my question,

    About the i's

    I realize now that I was assuming that λ was negative which is the condition for bound states but I realize that this was not a good assumption (turns out this only a well when λ <0) and that the i's should in fact be there.

    About the potential

    I believe the potential where x does not equal b is 0 however in class my professor indicated that the particle/wave in such a potential does not actually live at b but around it and yet it is still bound (assuming a well)

    Even with the i's however, I still end up with one equation featuring r, t and u(b). I'm not sure how to get the scattering amplitudes from this or if I am even on the right track.

    Thanks for the heads up about Latex, I will familiarize myself with the system

    Thanks again,

  5. Oct 11, 2008 #4


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    So combining all of that and starting over... you should have, for an incoming particle from the left:




    [tex]\left[\frac{du_l}{dx}\right]_{x=b-}-~\left[\frac{du_r}{dx}\right]_{x=b+}=\frac{2m\lambda}{\hbar ^2} \int_{b-}^{b+} \delta (x-b)u(x)dx~~~(3)[/tex]

    with (3) evaluated in the limit at [itex]b- \rightarrow b[/itex] from the left and [itex]b+ \rightarrow b[/itex] from the right.

    Also keep in mind that while there is a discontinuity in [itex]du/dx[/itex], there is none in [itex]u(x)[/itex] itself.


    This is the second boundary condition that you need to solve for the coefficients.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook