FQT2023 2
FQT2023 2
FQT2023 2
Sheet 2 — HT23
[Last update: 21:16 on 2023-02-04]
X
3
[Ji , Jj ] = iℏ εijk Jk .
k=1
1. Check that [J2 , Ji ] = 0 and [J3 , J± ] = ±ℏJ± . Deduce that in an irreducible representation of the angular
momentum operators, one can find a basis of joint eigenstates of J and J3 for which J takes a constant
value and if J3 |ψ⟩ = ℏm|ψ⟩ then J3 |J± ψ⟩) = ℏ(m ± 1)|J± ψ⟩.
2. Compute ⟨J± ψ|J± ψ⟩ in terms of ⟨ψ|ψ⟩. Use this to prove that if we write the J2 eigenvalue as ℏ2 j(j + 1),
then j must be a non-negative half-integer and the possible J3 eigenvalues can only be of the form ℏm
where m takes values in −j, −j + 1, . . . , j − 1, j.
3. Explain why, in an irreducible representation, each state with a given choice of quantum numbers |j, m⟩
is the unique such state (up to rescaling). Deduce the general structure of the spin-j angular momentum
representation.
Now consider the orbital angular momentum operators Li acting on wave functions in R3 . Define x± =
x1 ± ix2 = r sin θe±iφ and x3 = r cos θ. The corresponding derivatives are
1 ∂ ∂ ∂
∂± = ∓i , ∂3 = ,
2 ∂x1 ∂x2 ∂x3
so ∂± x± = 1 and ∂± x∓ = 0.
4. Show that with respect to this basis, the components of the orbital angular momentum operators L =
X ∧ P = −iℏ x ∧ ∇ are given by
6. Determine the action of the Laplacian on the functions rℓ Ym ℓ (φ, θ). (Don’t use spherical coordinates.)
Thus infer that the spherical harmonics are, up to normalisation, simply the restriction to the unit sphere
of homogeneous, harmonic polynomials of degree ℓ in three variables, with m measuring the power of
x+ minus the power of x− .
Here you should work through for yourself some of the discussion of the spin 1/2 projective representation
of the rotation group. The Pauli spin matrices are defined by
! ! !
0 1 0 −i 1 0
σ1 = , σ2 = , σ3 = .
1 0 i 0 0 −1
For a vector a, we define σ · a = σ1 a1 + σ2 a2 + σ3 a3 . This is the observable associated with measuring spin
along the a axis.
(σ · a)(σ · b) = a · b I2×2 + iσ · (a ∧ b)
2. Check by direct computation that the matrix representing a rotation by angle θ about the axis designated
by a unit vector n takes the form
iθ θ θ
exp − σ · n = cos − i sin σ·n.
2 2 2
3. Argue that a two-by-two matrix of this form is the most general two-by-two unitary matrix with deter-
minant one, and so this representation gives a two-to-one identification of elements of SU(2) with those
of SO(3).
Consider a particle moving in one dimension with potential given by a periodic function V(x) = V(x + ℓ)
for some real ℓ > 0.
1. Using the (discrete) translational symmetry of the potential, show that there will be a basis of (gener-
alised) energy eigenstates for the problem of the form
iθx
ψ θ (x) = exp ϕ(x) ,
ℓ
where ϕ(x) has the same periodicity properties as the potential. Explain why without loss of generality
you can take θ ∈ [−π, π].
2. For a(n generalised) eigenstate with energy E, show that ϕ(x) obeys the second order ODE,
ℏ2 ′′ iℏ2 θ ′ ℏ2 θ 2
− ϕ (x) − ϕ (x) + ϕ(x) = (E − V(x)) ϕ(x) .
2m m ℓ 2mℓ2
For what range of x should you solve this equation, and with what boundary conditions?
This can be thought of as a model for a one-dimensional crystal, where at the locations of the atoms in
the crystal a particle experiences an ultralocal attractive interaction.
Solve the ODE from the previous part of the question in this case for E > 0. You should leave the
expression in terms of the energy E (or better, k where k2 = 2mE/ℏ2 ), where k is implicitly determined
by θ according to (show this!)
α
cos(θ) = cos(kℓ) − sin(kℓ) ,
k
for a constant α that you determine.
[You may want to consider your freedom to choose an appropriate range of values of x for which to write
your solution. You can make a choice, for example, so the delta function appears in the middle of your
interval.]
4. Give a qualitative description of the allowed energy levels of the crystal. It will be useful to do some
investigations in a computational environment like Matlab or Mathematica. You will discover the phe-
nomenon of “electronic band structure”.
5. [Extra] For your own entertainment, think about how you would generalise this story to the case of a
three-dimensional lattice of delta functions. The generalisation of the choice of θ ∈ [0, 2π) is now the
choice of a point in what’s called the first Brillouin zone of the lattice.
2.4 Anti-unitarity
For a symmetry represented by a unitary operator U to be a dynamical symmetry, we require the condition
iHt iHt
U exp − = exp − U,
ℏ ℏ
1. If instead U is an anti-unitary operator, show that the above equation would imply that UHU∗ = −H.
Explain why this means that a system with such a dynamical, anti-unitary symmetry would have negative
energy states with energy −E for every positive energy state with energy E.
2. Consider now the anti-unitary operator T that acts on wave functions in L2 (R) by complex conjugation:
T (ψ)(x) = ψ(x) .
Explain how this evades the above issue in the case of, say, the harmonic oscillator Hamiltonian, for
which T is a true symmetry (sending energy eigenstates to energy eigenstates).
P2
H= +L·V,
2m
where L is the orbital angular momentum operator and V is a fixed (constant) vector. Is this system
T -symmetric?
Formulate a general condition for a Hamiltonian of a single-particle system (written in terms of X and P
operators) to respect T symmetry. Can you explain why this condition should hold intuitively?
1. Work out the decomposition of the composite Hilbert space into irreducible representations of the total
angular momentum operators in terms of what irreducible representations appear.
2. Consider the state 21 , 21 ⊗ 21 , 21 ⊗ 21 , − 21 . (Here the convention is, as in lectures, that (normalised)
basis states are written |j, m⟩.) You can determine how this state is expressed in terms of representations
of the total angular momentum operators in two different ways.
• First, combining the first two spins gives the state |1, 1⟩, and then combining with the third spin
gives α 23 , 21 + β 21 , 21 , for some numbers α and β.
• On the other hand, combining the second and third spin in the first instance gives γ|0, 0⟩ + δ|1, 0⟩,
whereupon taking the further tensor product with the first spin gives ε 23 , 21 + ζ 21 , 21 .
Compute α, β, γ, δ, ε, ζ (you may use values of Clebsch-Gordan coefficients given in textbooks if you
wish).
If you’ve followed the instructions, you have probably most likely β ̸= ζ. Explain what went wrong, and
identify the true answer for the resultant state.
2 2 2 (tot)
3. Show that J(1) + J(2) and J(2) + J(3) each separately commute with J(tot) and J3 , but do not
commute with one another.
Use this insight to rephrase the resolution of part two of this question in terms of two inequivalent choices
of basis for the triple tensor product of the spin- 21 representation.