Saturday, January 26, 2013

Homework 4: kinetic energy.

(due Friday by 3 PM) 
Wednesday, 2 pm:  (i just added an extra-credit problem (10) which will be relevant to understanding molecules and crystals.)

Up to this point, we have focused a lot of our attention on wave-functions and degeneracy, and applied that to understanding the some of the periodic table and the geometry of bonding for a few elements (using the 1st-excited states of a -1/r potential).

At this point we can deepen our understanding of quantum physics by turning our attention to energy. Kinetic energy, particularly when associated with confinement, plays a huge role in the quantum world. The interplay and competition between kinetic and potential energy is very important to the phenomena of quantum physics. It can help us understand the size of quantum systems, including the hydrogen atom, the length of chemical bonds and why some materials are insulators instead of metals, as well as some other things. This assignment helps us begin that endeavor.


1. Referring to the enclosed (scanned) notes for clarification as needed,
a) graph exp[-rb/a] as a function of x along the x axis, where rb=sqrt[(x-b)^2+y^2+z^2]
b) do a contour plot to illustrate the nature of exp[-rb/a]

2.
a) graph exp[-r/a] + exp[-rb/a] as a function of x along the x axis.
b) do a contour plot to illustrate the nature of (exp[-r/a] + exp[-rb/a]).

3. 
a) Calculate the expectation value of the kinetic energy (T) for the ground state of a 1D infinite square well.
b) Calculate the expectation value of the potential energy (U) for the ground state of a 1D infinite square well.

4. Note that one can calculate T without reference to any particular potential function. In the world of 1D quantum, consider an electron (of mass m) in a state in the form of a 1D normalized gaussian, i.e., $\psi (x) = (a \sqrt \pi)^{-1/2} e^{-x^2/2a^2}$:
a) calculate T (the expectation value of the kinetic energy) and,
b) graph the expectation value of T as a function of a.

5. On the other hand, potential energy (U) expectation values require both a potential, U(x) and a wave-function, $\psi(x)$.
a) Considering a as a parameter able to take on any positive, real value, calculate the expectation value of U for an electron in a state of the form $\psi (x) = (a \sqrt \pi)^{-1/2} e^{-x^2/2a^2}$,  in a potential $U(x) = (1/2)kx^2$.
b) graph the expectation value of U as a function of a.
c) find the minimum of T + U (expectation values).  At what value of a does that occur? What are the values of T and U at that minimum?

6. For a state that looks like the hydrogen atom electron ground state, but with a allowed to take any value (as in 4 and 5), one can show that T=$\hbar^2/(2ma^2)$ and that
U=$-e^2/a$ (see problem 8 for how one gets this). You can include a $1/(4 \pi \epsilon_o)$ if you prefer that unit system).
a) Show that $\hbar^2/(2ma^2)$ has units of energy,
b) Find the value of a that minimizes E=T + U and the value of T, U and E at that value of a.
c) Are E, T and U positive or negative? Why? (for each one)

7. Consider a 1D finite square well for which the potential is zero everywhere outside the well and -V inside the well. (V, the depth of our attractive well, is a positive number with units of energy.) Assume a width of L, and put the center of the well at x=0.
a) Assuming the ground state is of the form $A cos(kx)$ in the well (A and k undetermined) and $B e^{-\gamma (x-L/2)}$ outside the well, show that you can get a relation between k and $\gamma$ of the form -k*tan(kL/2) =$-\gamma$.
b) Additionally, gamma and k are related since $\hbar^2 k^2/2m = E + V$ and
$\hbar^2 \gamma^2 / (2m) = -E$ (where V is positive and E, the global separation parameter called the energy, is less than zero).
Write $\gamma$ as a function of V and k.
Why is E negative?
c) Anyway you can assume gamma is pretty much constant for small k (why?), and then graph the tan function and 1/k on the same graph to get the actual value of k for the ground state (graphically and approximately). Here is the question:
Is this value of k larger or smaller than the one you would get for an infinite square well of the same width?
d) and even more important: How do you explain that???

8. Extra credit. Calculate the expectation value of T for a hydrogen atom ground state wave function. (take 2 derivatives, integrate...). Do the same for U (i.e., integrate -1/r ...). (This is not actually that difficult as i thought at first since it is only radial derivatives and integrals, i think.)

9. Extra credit/Advanced problem.  Consider an electron in a 1DHO in the mixed state sqrt(1/2)(Psi0(x,t) + Psi1(x,t)).  Calculate and graph U and T (expectation values) as a function of t.

10. extra credit.  For an electron in the ground state of a 1DHO, the calculation of T (the expectation value of the kinetic energy, involves an integral of the product of the ground state wave-function $\psi_o (x)$ and the second derivative of the ground state wave-function (all of which is multiplied by $-\hbar^2/(2m)$.
Carefully sketch the integrand as a function of x. Which regions of x are positive and which are negative? (This will be important.)

39 comments:

  1. Question about #1. Is 'a' defined as [(hbar^2)/(mk)]^(1/4) or as the Bohr radius? I'm assuming the former, but I'm not completely sure...

    ReplyDelete
    Replies
    1. Well,if we had to choose one, the Bohr radius might make more sense, but really, for these, problems, it is just a generic "length scale".
      (and in 4, 5 and 6 we will be letting "a" vary. Trying out different values, seeing what fits best.

      Delete
  2. This comment has been removed by the author.

    ReplyDelete
  3. By the way, for 3, for the infinite square well expectation values, just integrate from zero to L. (Technically you might think you should integrate to infinity, and you would be correct, but because the wave function is zero outside the well, in practice you just want to integrate from 0 to L.

    ReplyDelete
  4. I think that we can now post equations by putting them inside dollar signs and using latex format. For example

    $\Psi_1r = e^{-r/a}$

    ReplyDelete
    Replies
    1. Now adding a second dollar sign at the end we get:

      $psi{1r} = e^{-r/a}$

      Delete
    2. You can practice stuff at the blog
      Latex embed
      http://lateximbed.blogspot.com/

      (I probably should have done that.)

      By the way, if you put \hbar inside dollar signs, then it will write $\hbar$

      Same with \psi or \Psi (which become $\psi$ and $\Psi$

      Delete
  5. For problem 3, are we looking at a 1D or 3D infinite well?

    ReplyDelete
  6. Problem 5 says "calculate the expectation value of U for an electron in this state". Which state is "this state" referring to?

    ReplyDelete
  7. For 6 did you mean to write that the average U is $\frac{e^2}{4\pi\epsilon a^2}$?

    ReplyDelete
    Replies
    1. That is $U=-\frac{e^2}{4\pi\epsilon a^2}$
      I think it needs to be negative

      Delete
    2. Thanks for that feedback. It is negative, as you say, so I fixed that. It is, however, a 1/a dependence for the potential energy (not 1/a^2). Does that make sense?

      Delete
    3. This comment has been removed by the author.

      Delete
    4. Yes, I didn't mean to square the a.

      Delete
  8. Would you mind clarifying what you mean in 7(c)? Are we supposed to solve for $k$ with the two formulas that we found with $\gamma$ and $k$?

    ReplyDelete
    Replies
    1. well you can't exactly solve for k, but one can graph those two things and see where their curves intersect (a graphical method of solving), and then see what you can learn from that. (a lot I think)

      Delete
    2. This comment has been removed by the author.

      Delete
  9. When i look up these integrals for expectaion values, I am getting erf(x/a). Can someone confirm that it just goes to zero for the bounds of integration. Or does it go to 1?

    ReplyDelete
    Replies
    1. try google searching "wolfram alpha" and then then try this (for example):
      integral of x^2 e^(-x^2/a^2) from -infinity to infinity

      Delete
  10. I think I might be over thinking question #1, as nobody is asking about it. I currently have e^-[(x-b)^(1/2)/a] for part a, and don't really know how to graph this. Could somebody nudge me in the right direction? Do I need to relate a to b or something?

    ReplyDelete
  11. I would suggest that you start by graphing r as a function of x. Then try graphing rb as a function of x.

    "I currently have e^-[(x-b)^(1/2)/a] for part a."
    I don't think that is correct.

    It won't help to try and relate a to b. a and b are independent length scales. That is, input parameters with units of length that influence different aspects of the function you are graphing.

    ReplyDelete
    Replies
    1. Thank you for answering this in class, I appreciate it.

      Delete
  12. I am wondering about number 7 c) and d). I know for an infinite square well $k=\frac{2\pi}{\lambda}=\frac{n\pi}{L}$. It also seems that for this problem $k ~ \frac{4}{3L}$. I think that this makes some sense as a smaller $k$ would suggest a greater $\lambda$ which in turn suggests less confinement. And that's what we're looking for, right?

    ReplyDelete
  13. im having trouble with graphing these, mostly confused when you say graph as a function of x or a, can you explain that a little? It seems any graphing program including wolfram alpha doesnt want to graph any of these.

    ReplyDelete
  14. above i was mostly referring to #1 and #2 graphs. completely lost on those.

    ReplyDelete
    Replies
    1. I posted a graph at the end of the "Guide to HW 4" post. Does that help clarify?

      Delete
  15. for 5b im getting an expectation value of (1/4k)a^2...which is confusing to me because it shows the expectation value of the potential as a parabola...is that right?

    ReplyDelete
    Replies
    1. Well, does that make sense in terms of how the expectation value of the potential energy depends on a? Would you expect the expectation value of the potential energy to increase or decrease with increasing a?

      Delete
    2. This comment has been removed by the author.

      Delete
  16. n a way $a$ is kind of like a one dimensional radius I believe (since were dealing with oscillations, which are somewhat circular). If $a$ is set the expected potential energy is set. In 101A I remember learning that $\kappa$ was actually related to $a$. I think it would have to have units of $\frac{J}{m^2}$ or maybe $\frac{\hbar^2}{2ma^4}$. I get from $U = \frac{\kappa}{4}a^2$ that as we increase $a$, which is weird to do because it is not normally the variable, we must increase our potential perhaps to keep the particle there. Any thoughts?

    ReplyDelete
  17. Hmmm. It is important to deeply ponder the distinction between the potential energy (PE) as a function of x, U(x), and the potential energy expectation value. It is natural that these things might get confused when you are first learning about them.
    The PE function, U(x), is a fixed characteristic that we start with. It is the environment that the electron exists in. It is fixed (by fiat) when we start the problem.
    On the other hand, the potential energy expectation value is a property of the electron. It depends both on the electron's environment, U(x), and on the electron's state (wave-function). A highly localized electron wave-function has a lower PE expectation value because the wave-function is confined primarily to regions of low U(x).
    We are not considering changing U(x). We are exploring the potential energy expectation value for different possible electron states, as defined by the wave-function and, in particular, by the parameter, "a" (which controls the "size" of the electron wave-function.

    ReplyDelete
  18. Just verifying (or not) that in problem #6b, when we look for the minimum total energy E= T + U, we are adding the expectation values of the kinetic and potential energies.

    ReplyDelete
  19. Is this going to be extended to Monday as well? Similar to the last assignment? :)

    ReplyDelete