Saturday, February 8, 2014

HW 5, due Friday 3PM. We need more peer-to peer discussion. Please join the discussion (e.g., regarding problem 5).

This assignment looks difficult to me. I would suggest starting very soon. Also, I encourage you to use comments to clarify what is asked in these problems. Especially for something like problem 5, this is a good forum to discuss what one should actually calculate and develop a consensus on that. For problem 5 b) I would recommend factoring out an overall \(e^{-i E_1 t/\hbar}\) factor. That may make it a little easier to see what that graph looks like.


Problem order was modified slightly on Feb 11. (Problem 4)
Commentary (on 5) added Feb 12.

1. If we go back to separation of variables, then we see that states have a time dependence that multiplies their spatial wave-function. So far we have been concentrating on just the spatial wave function. Going forward, however, we will begin to include time dependence where appropriate. Figuring when and how time dependence effects the wave-function, the "probability density" and expectation values is an important part of this endeavor.
a) For an electron in the ground state of a harmonic oscillator, what is the full time-dependent wave-function?
b) Plot the wave-function as a function of x: at t=0, \(t=\pi \hbar/(4E_1)\), and at times when the time dependent part of the wave-function is -i and -1 respectively (4 plots).  What are those times?

2. Consider an electron in a 1D harmonic oscillator potential. Suppose the electron is in a state that is an equal mix of the ground state and 1st excited state. At t=0:
\( \psi  =  \frac {1} {\sqrt {2}} \psi_{1} +  \frac {1} {\sqrt {2}} \psi_{2}  \).
a) Calculate \(\langle x \rangle\).
b) Is this state time dependent in a way that might effect \(\langle x \rangle\)? Is \(\langle x \rangle\) time dependent?
c) Is  \(\langle x \rangle\) real, imaginary or both in this case?
d) Calculate and plot \(\langle x\rangle\) as a function of ___. (Fill in the missing word.)
e) (extra credit) Calculate  \(\langle p\rangle\) and graph it.

3. Consider an electron in a hydrogen atom. The electron is in a state that is an equal mix of the ground state and a 1st excited state \(\psi_{21x}\).  That is, at t=0, \( \psi  =  \frac {1} {\sqrt {2}} \psi_{100} +  \frac {1} {\sqrt {2}} \psi_{21x}  \).
a) Calculate \( \langle x \rangle\). Are there any important time dependences that go with this state. Does it effect your result for the expectation value?
b) Earlier in the quarter, we combined states to make hybrids (such as \(\frac {-1} {\sqrt {3}} \psi_{200} +  \frac {\sqrt{2}} {\sqrt {3}} \psi_{21x} \)), however, we never worried about time dependence of  any expectation values. Were we ignoring time dependence that was really there?  Is this combination state different in some fundamental way from the combination states, such as sp^2 states, that we examined earlier in the quarter?
 
4.  I moved this to lower down so people don't get bogged down in it.

5. [For this problem, part d is the key part. a, b and c lead you to d).] Consider an electron in a 1D harmonic oscillator potential. Suppose the electron is in a state that is an equal mix of the ground state and 2nd excited state. That is, at t=0, \( \psi  =  \frac {1} {\sqrt {2}} \psi_{1} +  \frac {1} {\sqrt {2}} \psi_{3}  \).
a) Calculate \(\langle x \rangle\).
b) Graph \(|\Psi (x,t)|^2\) as a function of x for t=0 and \(t = \hbar \pi /(E_3 - E_1) \). [ I would recommend factoring out an overall \(e^{-i E_1 t/\hbar}\) factor. That may make it a little easier to see what that graph looks like.]
c) Look at those graphs. How are they different?  What expectation value do you imagine might be time dependent?
d) Calculate that.

6. [For this problem also, part d) is the most important part. c) is the concept, d) is the implementation.] Consider a single electron in a double square well potential. (Aside: the double dot qubit, which is a "popular" approach to creating quantum computers of the future, is essentially a double square well based system.) Suppose both square wells are 16 eV deep and have the same width L. Let s (for separation) be the distance between these two identical wells (near edge to near edge).
a) Sketch the square well potential.
b) Sketch the ground state and 1st excited state for a single electron in this system.
Suppose that after hours of calculation you know everything about the one-electron states of this 2 well system.*
c) How would you model an electron starting out at t=0 mostly on one side of the well?  How would you calculate how long it takes an electron to mostly get to the other side
d) How long does it take for the electron to get from one side to the other?
e) Is this tunneling? What is tunneling?

* Everything includes the energies of the states and the characteristics of the spatial wave functions. Like in problem 5 from HW 2, except for the double well where the wave-function in the spatial region between the two wells is of the form:
\(cosh(x/a_n)\) for symmetric states,
\(sinh(x/a_n)\) for anti-symmetric states.
For the ground state and the 1st excited state, the form of the wave-function in the right-hand well, I think could be written as:
\(\psi_n (x) = B_n cos(k_n (x-(L+s)/2)\)
and outside the well I think we could write:
\(\psi_n (x) = C_n e^{-x/a_n}\)
[this detail may be distracting. What do you really need to know about the states?]

4. Consider an electron in a hydrogen atom. The electron is in a state that is an equal mix of the ground state and a 1st excited state \(\psi_{211}\).  That is, at t=0, \( \psi  =  \frac {1} {\sqrt {2}} \psi_{100} +  \frac {1} {\sqrt {2}} \psi_{211}  \).
a) What is  \( \psi_{211}  \)?
b) Calculate \( \langle x \rangle\).
c) Calculate \( \langle y \rangle\).
d)  Draw the vector \( \langle \vec{r}  \rangle = (\langle x  \rangle, \langle y \rangle, \langle z \rangle ) \).  What plane is it in? Describe its time dependence.

7. (extra credit) For an electron in the same state as in problem 2:
a) calculate the expectation value of the momentum as a function of time.
b) Plot \(\langle x \rangle\) and \(\langle p \rangle\) vs time. What is their relationship?
(Should this be extra credit? How difficult are those derivatives? How many terms do you get?)

8. extra work/extra credit: For an electron in the state given in problem 5:
a) graph and discuss the relationship between K.E. and P.E. (as a function of time). Do whatever calculations are interesting or necessary to aid and inform this discussion.

53 comments:

  1. Just for clarification for for 1b, is it when psi(x,t) = -i and psi(x,t) = -1?

    ReplyDelete
    Replies
    1. sort of yes. it is when the t part is. does that makes sense?

      Delete
    2. So we are setting t = -i and -1 in this case?

      Delete
    3. I took it as when the time dependent portion (cos(stuff)-isin(stuff)) is equal to -1 or -i. Those two plots along with the first two Zack explicitly lays out make a total of four plots.

      Delete
  2. I like this video: http://www.youtube.com/watch?v=cV2fkDscwvY

    ReplyDelete
  3. For the expectation value of x in problem 5a, I keep getting 0 for what should be the nonzero integrals (i.e. the cross products of psi 1 and psi 3), and therefore the whole answer. I know this is what we previously had for a 1D harmonic oscillator, but I'm not sure if this answer makes any sense now since time-dependence is still fairly new to me.

    ReplyDelete
    Replies
    1. Because both psy1 and psy3 are even functions respect to x variable, so when they are added together linearly the result is still symmetric (even), and that make the expectation value of x to be zero.

      Delete
  4. Alright, after looking over the plots of the two wave functions that makes sense, especially why it's zero for this case and not the case in problem 2 when we had psy 1 and psy 2. Thanks!

    ReplyDelete
  5. So for 5c, I'm thinking we're supposed to calculate the expectation value for which x I get 0. Are we supposed to be calculating ?

    ReplyDelete
    Replies
    1. Sorry that was kind of badly worded, I meant to say:

      So for 5c, I'm thinking we're supposed to calculate the expectation value for x for which get 0. Are we supposed to be calculating this?

      Delete
    2. I think he means that you would graph the ground state and 2nd excited state for the 1D harmonic oscillator and then after comparing the two you could figure out what expectation value you would calculate that is time dependent.

      Delete
    3. It would be redundant to calculate the expectation value of x twice...

      Delete
    4. I am also curious about this, since the calculation would be wiped to 0 by the integral term before we even consider the time dependence.

      Delete
    5. As Ravipa says above, you are asked to think about what expectation value would be interesting to calculate. So you are supposed to define the problem (for part d) for yourself, rather than just follow instructions. This problem asks for thinking and decision making on your part in a way that has not been asked of you before (in this class). Does that make sense?

      "Looking at those graphs... What expectation value do you imagine might be time dependent?"

      Delete
    6. Looking at the graphs, it does look like the expectation value of x changes over time but I agree it would be redundant to calculate it twice. Is the overall energy time dependent?

      Delete
  6. For number 3b, I'm not completely sure by how the combination state is different from other combination states. Are they' different when doing separate expectation values with various combination states or do you mean something else?

    ReplyDelete
    Replies
    1. I think what the problem is asking is, what is the difference between the hybridization states and the superposition states? To answer this problem I think we should look at what we get when we solve for with time dependence. The one key term that comes out of that calculation, with respect to time, is cos[t(E2-E1)/hbar)]. If we consider the hybridized states we should notice that the energy levels of each state being combined are all the same. So its not that we ignored the time dependence, but that the time dependent term drops out because the energy levels of the state functions being hybridized are degenerate. This is also where we can get the answer to the second part of this problem. In hybridized states the energies are the same; where as in superposition states the energies are different. That is the fundamental difference that makes expectation values for superposition states so different from hybridization states.

      Delete
  7. I have begun to do problem #4 and have so far come up with 12 different (although somewhat similar) integrals to take; 6 for each and . My intuition is telling me that this is a bit high, and I just wanted to see if other people are coming up with a similar number of integrals to compute? Are there possibly a few combinations of psi functions which lead to integrals equaling 0? So far, the only integrals I have going to 0 are the psi^2 terms.

    ReplyDelete
    Replies
    1. It should say for both the expectation value of x and the expectation value of y.

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

      Delete
    3. That seems like a lot. What integrals are you getting? For example, I get and integral with the integrand (x 21x 100) that is non-zero, and also (y 21y 100). Using that format, what other integrals are you getting?

      Delete
    4. I'm getting, for the x expectation value, integrands of the form (100 21X*) (i 100 21Y*) (100* 21X) (i 100* 21Y) (i 21Y* 21X) (i 21X* 21Y) with all of these being multiplied by an x.

      Delete
    5. Thanks. Some thoughts: What is the difference between (x 100 21x*) and (x 21x* 100)? Would any integral with a linear x or y be zero by symmetry? Also, in terms of organization, I would strongly suggest including the "i" in the pre-factor, rather than as part of the integral.

      Does that help reduce the number of integrals? What number do you have now?

      Delete
    6. For the x expectation value I can see how you could combine (100* 21x) and (100 21x*), if I'm not mistaken in this. but it seems to me that you are implying that terms containing (100 21y) as well as (21x 21y) go to zero, which I am not completely sure about. Is that true? Or maybe a better question is this; is (21y* 21x) = (21x* 21y)?

      Delete
    7. I think if you use a notation that includes the entire integrand it will be more clear. For example, (x 100 21y) integrates to zero. On the other hand, (y 100 21y) is not zero (and is probably the same as (x 100 21x), isn't it?).
      21x, 21y and 100 are all real, so i don't see why the * would effect any of those integrals. (Distinguishing integrals, which are real, from the prefactors, which can be complex, makes it easier.)

      Delete
    8. Okay, I think I can see now how it reduces down to two integrals. Thanks so much for the help, Professor!

      Delete
  8. I was working on number 4 and I am curious does psi star of psi(2,1,1) equal to (psi(2,1,x)-i*psi(2,1,y)? or is it just the same as psi(2,1,1)?

    ReplyDelete
    Replies
    1. What do you think? You tell me and I will comment.

      Delete
    2. I used the psi-star mentioned above, and my answer to part b ended up being the same as my answer to 3a. Is it incorrect to use that form of psi-star? I thought I would've ended up with a different answer than that of 3a.

      Delete
    3. This comment has been removed by a blog administrator.

      Delete
    4. @Monique I think it is the same as 3a. Except maybe smaller by a factor of sqrt(2). So i think you did the conjugation correctly.

      Delete
    5. @Anonymous. Sorry I deleted your comment. It was partly true, but I thought it might lead to confusion.

      Delete
  9. For problem 4c, did anyone else get 0 for the y expectation value? When I calculated this, the only non-zero terms that I got were the cross terms which ended up exactly canceling each other out. Did anyone else get this?

    ReplyDelete
    Replies
    1. Somehow I got nonzero , very similar to

      Delete
    2. Same, I did it twice and got zero both times. Not what I would've expected... but since we are working with only complex psi_21y I can believe the answer I got.

      Delete
    3. If you got the correct value, I don't see how you could get a 0 , considering how similar the integrals are. The Prof did say that the electron will be going in a circle. Mine goes clockwise. The terms can't physically cancel out, considering the time dependent parts do not equal each other. Subtract the two time dependent components and you should get i2sin(stuff).

      Either that or I'm just stupid.

      Delete
    4. Correct value for *

      Delete
    5. Thanks! I managed to get a nonzero value. I was assuming too much about things canceling and was careless with my signs.

      Delete
  10. Replies
    1. I would like to second this witchcraft theory.

      Delete
  11. In answer to your question to if 7) should be extra credit, I think it should be

    ReplyDelete
    Replies
    1. Agreed, the math gets really ugly.

      Delete
    2. It should be extra credit because you asked the same thing as extra credit for #2 already

      Delete
    3. Agreed! #7 should be extra credit.

      Delete
  12. I was wondering if the wave function we were given earlier in the quarter was incorrect???
    On psi of 3, is it composed with (1-(x^2)/(a^2)) or ((x^2)/(a^2))-1)??? (For problem number 5)

    ReplyDelete
  13. That's a good question, since one needs to check the things you are working with.
    The two terms you mention are equivalent and equally valid (since one is just -1 times the other), however, i think there is a small mistake in your terms and that the actual psi_3 is proportional to +-(1-2x^2/a^2) which could also be written as (1/2-x^2/a^2), right?
    However, I don't think that small mistake would be the cause of major problems with problem 5. The key thing is that psi3 is an even function. The key question is what expectation value are you trying to calculate for 5d?

    ReplyDelete
  14. Question 6 part d: "What is tunneling?"
    Are you asking us to describe tunneling or tell you what is 'doing' the tunneling?

    ReplyDelete
    Replies
    1. Ah, good question. I was asking you to describe tunneling. What is the essence of tunneling? What is the essential thing that makes some motion tunneling (and other motion not tunneling)?

      Delete