PHYS 357 Pset 4. Due 11:59 PM Thursday Oct.

17

In some of these problems, we’ll see that interesting behavior can emerge when we work with
states with larger angular momentum. It’s a pain to do this by hand, so feel free to carry out
calculations on a computer. If you do this, do please explain how your codes work so the TAs can
understand what you did when it comes time to mark.

1. Townsend 3.24. Please do this one by hand - Townsend helpfully supplies all the relevant
matrices for you.

2. Consider a ping pong ball (mass 2.7 grams and radius 20 mm) rotating at 10 radians per
second (in real play, the spins can be 100 times higher). What is the approximate angular
momentum (order of mangitude is fine, so you can use M R2 for the moment of inertia)? If we
say the ping pong ball is in a state |j, ji rotating about the z−axis, what is the approximate
value of j? Given that value of j, what is the total angular momentum in the x − y plane,
p 2
Jx + Jy2 ? We can express this as an uncertainty in the rotation axis of the ping pong ball,
√ 2 2
J +J
with σ(tan(θ)) ∼ Jxz y . What is your approximate value for σ(tan(θ))? For a macroscopic
object, talking about “the” axis of rotation is a very good approximation!

3. Part A) Consider a spin-1 Stern-Gerlach experiment where you send a beam of particles
from the oven down a modified SGz machine, where you block the Jz = 0 beam before
√
recombining. You might think that would leave your particles in the state Jz = (1, 0, 1)/ 2.
If that were the case, what would you see (in terms of what fraction of particles go up,
go down, or aren’t deflected at all) if you then send the beam through an SGx machine?
Through an SGy machine?

Part B) Explain why these results can’t be correct (hint - what happens if I take my SGz
into SGx apparatus, and rotate the whole setup by 90 degrees about the z−axis).

Part C) If you were to actually set this experiment up and run it, what you would see
is that 50% of the particles show up with Jx,y = 0, 25% show up with Jx,y = +~, and
25% show up with Jx,y = −~ (where Jx,y means I measure either Jx or Jy ), so indeed the
 apparatus behaves the same if I rotate my coordinate system by 90 degrees, as it must. Can
you quantitatively explain the actual observed results? You may use a computer if you like.

4. Part A) We saw that Jz commutes with Jx2 + Jy2 . Does Jx commute with Jx2 + Jy2 ? If not,
what is the commutator [Jx , Jx2 + Jy2 ]? Feel free to express as an anti-commutator where
{A, B} = AB + BA.

Part B) Let’s say I take a set of spin-2 particles, all of which have Jz = 2~, which we write
as |2, 2iz (note the z subscript, which says this state is an eigenstate of Jz ). If I were to
measure Jx2 +Jy2 , what would I observe, and what would the uncertainty in that measurement
be?

Part C) I now send my beam of |2, 2iz particles down an SGx machine. Show that the
expectation of Jx2 that you see is one half of < Jx2 + Jy2 > you calculated in part B).

Part D) What is the probability that you measure Jx = +2~? If this is non-zero, can
you explain the apparent contradiction between finding a particle witih Jx = 2~ (and hence
Jx2 = 4~2 ) and the maximum value you found for Jx2 + Jy2 from part B?

5. Part A) Sticking with spin-2 particles, express the pure state |2, 2iy in both the z-basis and
the y-basis. In the z−basis, please pick the phase that makes the amplitude of Jz = +2~
purely real and positive.

Part B) For this same |2, 2iy pure state, what are the uncertainty in Jx and Jz ? Show that
the uncertainty relation is satisfied.

6. Part A) Calculate the matrix that rotates a state about the y−axis by +90 degrees for a
spin-2 particle. Print the absolute value of the matrix.

Part B) Show that the above matrix rotates the state |2, 2iz to the pure state |2, 2ix , and
|2, 2ix to |2, −2iz .

7. Bonus: Our end goal is to derive the angular momentum commutation relation [Ja , Jb ] =
ihabc Jc but to do that we’ll first need the canonical commutation relation [x, p] = i~. In
the first bonus we’ll work that out, in the second bonus, we’ll use it to derive the angular
 momentum commutation relations. You are more than welcome to do the second bonus
only, and just use the canonical commutation relation as supplied.

In 1924, Louis De Broglie hypothesized (in his PhD thesis!) that all matter behaved like
waves, with wavelength set by their momentum. This was the key insight that opened the
door to modern quantum mechanics, and within a couple of years both wave and matrix
mechanics were basically fleshed out in their modern forms. The De Broglie relation is

p = ~k (1)

where k = 2π/λ and p is the momentum. If I have a wave function of a particle with wave
vector k, I can write that down as a function of x as follows: Ψ = exp(ikx).
∂
Part A:Show that the operator −i~ ∂x operating on Ψ returns pΨ. In other words, the plane
∂
wave is an eigenstate of momentum, and the operator p̂ = −i~ ∂x returns the wave function
times the momentum.

Part B: Now the position operator x̂ in position space (where Ψ = Ψ(x)) is, not surpris-
ingly, just x. You can now derive the canonical commutation relation [x, p] by applying the
operators to a wave function (x̂p̂ − p̂x̂)Ψ. Show that when you do this, you get i~Ψ. That is,
when you apply [x, p] to a wave function, you get i~ times the wave function, so [x, p] = i~.

8. Bonus 2. Now we’ll work out the angular momentum commutators, using the fact that
J = r × p. If you expand that out, you can see that Jx = ry pz − rz py , which will all get
applied to a wave function.

Position and momentum on the same axis do not commute, but they do commute if you
measure position along one axis, and momentum along another, perpendicular axis. More
formally, [xi , pj ] = i~δij . Use this commutation relation and the expressions for Jx, Jy, Jz
you get from J = r × p to show that [Jx, Jy] = i~Jz. The commutators for other pairs of
axes follow from cyclic permutations of the cross product, so you only need to do it once.

Note - technically we’ve only shown this for orbital angular momentum and not spin. We
did show that spin-1/2 particles obey the same commutation relation, so I hope it’s not a
surprise that the relation holds for higher spins as well.