MITES 2010: Physics III Survey of Modern Physics Problem Set 4 Solutions
MITES 2010: Physics III Survey of Modern Physics Problem Set 4 Solutions
MITES 2010: Physics III Survey of Modern Physics Problem Set 4 Solutions
(i.) If you measure the energy of 10,000 identical particles, all in the above ’indetermi-
nate’ state just before your measurement, how many of them do you expect to yield
energy value of E2 ?
iθ 2
e 1
Solution: The probability of obtaining E2 is √ = . So, out of 10,000 measurements, we expect
6 6
10, 000
≈ 1667 particles
6
to yield energy E2 .
(ii.) On average, what value of energy would your measurement yield?
Solution:
(iii.) For which value of θ is the probability of a particle having energy E2 zero?
Solution: There is no value of θ for which the probability of obtaining E2 is zero . First of all, this
is because the coefficient of the state |E2 i is proportional to eiθ and eiθ never equals zero. Second of
all, the magnitude of√eiθ is always 1 regardless of θ, so the probability of being in E2 is independent
of θ Prob(E2 )=|eiθ / 6|.
1
(iv.) Suppose you measure the energy of a particle. You find that it has energy E1 . Does
this mean that the particle was in state |E1 i just before your measurement? Suppose
you measure the energy of the same particle for the second time. What is the proba-
bility that you get E4 as the energy this second time?
Solution: No, the particle was not in the state |E1 i before the measurement . A postulate of quan-
tum mechanics states that a quantum state is in a superposition of many states before a measurement
and only collapses to one state after the measurement. Also, since the state is now in |E1 i then there
is zero probability that it is in any other state .
(v.) Suppose you measure the energy of the particle and find that it has energy E2 .
Write down the ket-vector representing the particle’s state immediately after this mea-
surement.
Solution: Since we measured energy and obtained the value E2 we know that the particle’s state
has now collapsed to the state corresponding to E2 . That is, the particle is in state |ψi = |E2 i .
Solution: By one of the postulates of quantum mechanics, the state has collapsed into the state
corresponding to eigenvalue a1 . So, the state is in ψ1 .
(b.) If B is now measured, what are the possible results, and what are their probabilities?
Results Probability
2
3 9
b1 =
5 25
2
4 16
b2 =
5 25
2
(c.)Right after the measurement of B, A is measured again. What is the probability of
getting a1 ? (Note that the answer would be quite different if I had told you the outcome
of the B measurement.)
Solution: There are two possible ways to get to a1 from our initial a1 measurement. We can obtain
b1 from the B measurement and then obtain a1 from the A measurement. Or, we can obtain b2 from
the B measurement and then obtain a1 from the A measurement. In order to calculate the total
probability of obtaining a1 , we must calculate the probability of these independent paths and then
add them. But first, we must write φ1 and φ2 in terms of ψ1 and ψ2 in order to obtain the probabilities
for the second measurement of A after our measurement of B. Solving Eq (1)by elimination, we find
for φ1
3φ1 + 4φ2
ψ1 =
5
5ψ1 − 4φ2
= φ1
3
3
For each path, we multiplied the probability of obtaining our first measured value of B by the
probability of obtaining our second measured value of A. The total probability of obtaining a1 is
81 256 337
the sum of these probabilities and is therefore + = We can check that this procedure
625 625 625
for calculating probability is consistent with normalization by calculating the probability other two
possible paths. Instead of obtaining a1 , we could have obtained a2 in our second measurement and
the paths and probabilities associated with this value are
Path Probability
9 16 144
3) From b1 to a2 × =
25 25 625
16 9 144
4) From b2 to a2 × =
25 25 625
We have listed all possible outcomes of our measurements. The sum of the probabilities for theses
outcomes must be one.
81 256 144 144 625
Prob(Path 1) + Prob(Path 2) + Prob(Path 3) + Prob(Path 4) = + + + = =1
625 625 625 625 625
Our outcomes are normalized, so we know that our procedure is correct.
So the possible eigenvalues of Ĥ are E1 = h̄ω, E2 = 2h̄ω, and E3 = 2h̄ω. Using these eigenvalues to
find the eigenvectors, we find
4
E1 = h̄ω
h̄ω − h̄ω 0 0
0 = (Ĥ − E1 I)|E1 i = 0 2h̄ω − h̄ω 0 |E1 i
0 0 2h̄ω − h̄ω
0 0 0
= 0 h̄ω 0 |E1 i
0 0 h̄ω
So we see that |E1 i can have no second and third component in order for it to satisfy the above
equation. Therefore, |E1 i must be of the form
a 1
|E1 i = 0 = 0
0 0
Where we chose a = 1 in order to normalize |E1 i.
The next eigenvalue E2 = E3 = 2h̄ω has multiplicity two (i.e. it occurs twice) so we should get two
eigenstates.
E = 2h̄ω
h̄ω − 2h̄ω 0 0
0 = (Ĥ − EI)|Ei = 0 2h̄ω − 2h̄ω 0 |Ei
0 0 2h̄ω − 2h̄ω
−h̄ω 0 0
= 0 0 0 |Ei
0 0 0
We see that |Ei can have no first component in order for it to satisfy the above equation. There are
two ways we can accomplish this: |Ei only has a second component; |Ei only has a third component.
These two ways correspond to our two eigenstates, so that we have |E2 i and |E3 i as
0 0
|E2 i = 1 |E3 i = 0
0 1
Summarizing these results, we have for the operator Ĥ
Eigenvalues Eigenstates
1
E1 = h̄ω |E1 i = 0
0
0
E2 = 2h̄ω |E2 i = 1
0
0
E3 = 2h̄ω |E3 i = 0
1
5
Employing the same procedure for Â, we find
−a µ 0
det(Â − aI) = det µ −a 0
0 0 2µ − a
−a 0 µ 0
= −a det −µ
0 2µ − a 0 2µ − a
= −a(a − 2µ)a − µµ(2µ − a)
= a2 (2µ − a) − µ2 (2µ − a)
= (a2 − µ2 )(2µ − a)
= (a − µ)(a + µ)(2µ − a)
So, the eigenvalues of the operator  are a1 = µ, a2 = −µ, and a3 = 2µ. Computing eigenvectors,
we find
a1 = µ
−µ µ 0
0 = (Â − a1 I)|A1 i = µ −µ 0 |A1 i
0 0 2µ − µ
−µ µ 0
= µ −µ 0 |A1 i
0 0 µ
6
Lastly a3 = 2µ
−2µ µ 0
0 = (Â − a3 I)|A1 i = µ −2µ 0 |A3 i
0 0 2µ − 2µ
−2µ µ 0
= µ −2µ 0 |A3 i
0 0 0
−2g1 + g2 = 0
g1 − 2g2 = 0
these two equations are inconsistent and therefore produce the null solution g1 = g2 = 0. Therefore,
|A3 i is of the form
0 0
|A3 i = 0 = 0
g3 1
where we chose g3 = 1 for normalization. In summary, we have for the operator Â
Eigenvalues Eigenstates
1
1
a1 = µ |A1 i = √ 1
2 0
1
1
a2 = −µ |A2 i = √ −1
2 0
0
a3 = 2µ |A3 i = 0
1
Now, for B̂. Studying the matrix form of this operator and comparing it with the operator Â, we
see that B̂ and  have the same set of eigenvalues and eigenvectors. We will prove this by explicit
computation
2λ − b 0 0
det(B̂ − bI) = det 0 −b λ
0 λ −b
−b λ
= (2λ − b) det
λ −b
= (2λ − b)(b2 − λ2 )
= (2λ − b)(b + λ)(b − λ)
7
So the eigenvalues of the operator B̂ are b1 = λ, b2 = −λ, and b3 = 2λ. Computing eigenvectors, we
find
b1 = λ
2λ − λ 0 0
0 = (B̂ − b1 I)|B1 i = 0 −λ λ |B1 i
0 λ −λ
λ 0 0
= 0 −λ λ |B1 i
0 λ −λ
This result is similar to one obtained for the  case and we therefore see that
0
1
|B1 i = √ 1
2 1
b2 = −λ
2λ + λ 0 0
0 = (B̂ − b2 I)|B1 i = 0 +λ λ |B1 i
0 λ +λ
3λ 0 0
= 0 λ λ |B1 i
0 λ λ
This result is similar to one obtained for the  case and we therefore see that
0
1
|B2 i = √ 1
2 −1
b3 = 2λ
2λ2λ 0 0
0 = (B̂ − b3 I)|B1 i = 0 −2λ λ |B1 i
0 λ −2λ
λ 0 0
= 0 −2λ λ |B1 i
0 λ −2λ
Once again, we obtain inconsistent equations from two of our rows (the bottom two rows) and the
only surviving component is the top component. This result is similar to one obtained for the  case
and we therefore see that
1
|B3 i = 0
0
8
In summary, we have for the operator B̂
Eigenvalues Eigenstates
0
1
b1 = λ |B1 i = √ 1
2 1
0
1
b2 = −λ |B2 i = √ 1
2 −1
1
b3 = 2λ |B3 i = 0
0
with |c1 |2 + |c2 |2 + |c3 |2 = 1. Find the expectation values (i.e. averages) (at t = 0) of H, A
and B.
Solution: The formula for the average of an observable
hOi = hS(0)|Ô|Ŝ(0)i
9
And finally for B̂
2 0 0 c1
hS(0)|Â|Ŝ(0)i = c†1 c†2 c†3 λ 0 0 1 c2
0 1 0 c3
2µc1
† †
= c†1 c†2 †
c3 µc3 = µ(|c1 |2 + c2 c3 + c3 c2 ) = bavg
µc2
(c.) What is |S(t)i? If you measured the energy of this state (at time t), what values
might you get, and what is the probability of each? Answer the same questions for A
and for B.
Solution: From the notes we know that if |S(0)i is of the form
we see that if we measure the energy of our state, then we can get
Energy Probability
h̄ω |c1 |2
2h̄ω |c2 |2 + |c3 |2
where we added the probabilities of being in state |E2 i and |E3 i because they have the same eigen-
value.
In order to find the probabilities of obtaining certain eigenvalues of the operators  and B̂, it is
useful to write |S(0)i as a linear combination of their respective eigenvectors. For  we have
10
c1
|S(0)i = c2
c3
(c1 + c2 + c1 − c2 )/2
= (c1 + c2 − c1 + c2 )/2
c3
0 1 1
c1 + c2 c1 − c2
= c3 0 + 1 + −1
2 2
1 0 0
0 1 1
c1 + c2 1 c1 − c2 1
= c3 0 + √ √ 1 + √ √ −1
1 2 2 0 2 2 0
c1 + c2 c1 − c2
= c3 |A3 i + √ |A1 i + √ |A2 i
2 2
So, we see that if we measure observable A, then we get
Eigenvalue Probability
|c1 + c2 |2
a1 = µ
2
|c1 − c2 |2
a2 = −µ
2
2
a3 = 2µ |c3 |
11
So, we see that if we measure observable B, then we get
Eigenvalue Probability
|c2 + c3 |2
b1 = λ
2
|c2 − c3 |2
b2 = −λ
2
2
b3 = 2λ |c1 |
12
5. Problem 5. Quantum Simple Harmonic Oscillator
In this problem, we derive the properties of the simple harmonic oscillator that exists
in the realm of quantum mechanics.
The Energy (Hamiltonian operator) of a simple harmonic oscillator with mass M and
angular frequency ω in classical mechanics is
p2 mω 2 2
+ H= x
2m 2
Quantizing above classical Hamiltonian yields the quantum mechanical Hamiltonian for
a simple harmonic oscillator which is
p̂2 mω 2 2
Ĥ =+ x̂
2m 2
At a first glance, this looks just like the classical Hamiltonian. But don’t be fooled - the
important difference here is that the momentum and position are represented by the
operators p̂ and x̂ respectively, since both are observables. (Remember that in quantum
mechanics any observable (things you can measure) are represented by an operator).
(a.) Let us define the following non-Hermitian operator â and its Hermitian conjugate ↠:
ωmx̂ + ip̂ ωmx̂ − ip̂
a= √ a† = √
2ωmh̄ 2ωmh̄
Inverting these relations, derive the following position and momentum operators x̂ and p̂:
r
h̄
x̂ = (â + ↠)
2mω
r
h̄mω
p̂ = −i (â − ↠)
2
Solution: Adding â and ↠we find
ωmx̂ + ip̂ ωmx̂ − ip̂
â + ↠= √ + √
2ωmh̄ 2ωmh̄
2ωmx̂
= √
2ωmh̄
r
2mω
= x̂
h̄
r
h̄
so that x̂ = (â + ↠)
2mω
13
r
ωmh̄
so that p̂ = −i (â − ↠)
2
(b.) Using the commutation relation [x̂, p̂] that we derived in class, show the following
commutation relation:
[â, ↠] = 1
Solution: We calculate the commutator [x̂, p̂] with our position and momentum operators expressed
in terms of â and ↠.
r r
h̄ ωmh̄
[x̂, p̂] = −i (â + ↠)(â − ↠)
2mω 2
r r
h̄ ωmh̄
i (â − ↠)(â + ↠)
2mω 2
h̄ †
= −i (â2 + ↠â − â↠+ aˆ2 )
2
h̄ 2 †
+i (â − ↠â + â↠+ aˆ2 )
2
= ih̄(â↠− ↠â)
= ih̄[â, ↠]
14
The derivation for ↠is exactly the same as the one for â except we have a minus, instead of a plus,
sign infront of the momentum operator. Therefore for ↠we have
† 1 x̂ d
â = √ − x0
2 x0 dx
(d.) Now, putting everything together, show that the (quantum) Hamiltonian for a sim-
ple harmonic oscillator is
† 1
Ĥ = h̄ω â â +
2
So we have now reduced the problem to that of finding the eigenvalues of the occupation
number operator
n̂ = ↠â
p̂2 mω 2 2
Ĥ = + x̂
2m 2
h̄mω (â − ↠)2 mω 2 h̄
= − + (â + ↠)2
2 2m 2 2mω
h̄ω †
= − (â2 − â↠− ↠â + aˆ2 )
4
h̄ω 2 †
+ (â + â↠+ ↠â + aˆ2 )
4
h̄ω
= (â↠+ ↠â)
2
h̄ω †
= (â â + 1 + ↠â)
2
h̄ω † † 1
= (2â â + 1) = h̄ω â â +
2 2
In the second to last line, we used the commutation relation [â, ↠] = 1 to write â↠= ↠â + 1.
(e.) Using the result from problem 4a, show that
[n̂, ↠] = â†
[n̂, â] = −â
15
Similarly, for the other commutation relation
[â, ↠] = 1
[â, â] = [↠, ↠] = 0
The first result was derived in (b). The second result is the obvious statement that ââ = ââ and
↠↠= ↠↠. (f.) Now, show that if |ni is an eigenstate of n̂ with an eigenvalue n, then
â|ni is an eigenstate of n̂ with an eigenvalue n + 1.
So
n̂↠|ni = (n + 1)↠|ni
From Eq,(2), we may be tempted to write ↠|ni = |n + 1i so that our final result can be written as
n̂↠|n + 1i = (n + 1)↠|n + 1i
However, this definition of |n + 1i does not fit our requirement for normalization. We can see this by
assuming the n + 1 and n state are both normalized, that is hn + 1|n + 1i = 1 and hn|ni = 1. If we
have ↠|ni = |n + 1i then
So if ↠|ni = |n + 1i then hn + 1|n + 1i = n + 1 6= 1 and our n + 1 state is not properly normalized.
We must include some numerical factor in order to ensure the normalization of the state. So, we let
↠|ni = c1 |n + 1i where c1 is our arbitrary numerical factor. If we go through the same calculation
for hn + 1|n + 1i we find that
hn + 1|n + 1i|c1 |2 = (n + 1)
16
√
So, in order to ensure the normalization of |n + 1i, c1 must equal n + 1 and we have therefore have
the formula √
↠|ni = n + 1|n + 1i (3)
(g.) It turns out that |0i is an eigenstate of n̂ with the lowest possible eigenvalue n = 0
(we’re going to take this as given, and skip the proof of this fact). This means that
the lowest possible energy of the simple harmonic oscillator is E0 = 12 h̄ω. Using (f.),
construct all other eigenstates of the Hamiltonian and their corresponding eiegnenergies.
In particular, show that the allowed energies and the energy eigenstates of a quantum
simple harmonic oscillator are
1
En = n+ h̄ω (where n = 0, 1, 2, ..)
2
1
|ni = √ (↠)n |0i
n!
Solution: If |0i is an eigenstate of the hamiltonian with eigenvalue E0 = 12 h̄ω. Then by definition
we have
1 1
Ĥ|0i = h̄ω ↠â + |0i = h̄ω|0i
2 2
Using Eq (3)we can construct states excited states higher than |0i. For example,
1
|n + 1i = √ ↠|ni
n+1
1
|1i = √ ↠|0i
1
1 1
|2i = √ ↠|1i = √ ↠|0i
2 2·1
1 † 1 1
|3i = √ â |2i = √ ↠|1i = √ ↠|0i
3 3·2 3·2·1
1 † 1 1 1
|4i = √ â |3i = √ ↠|2i|2i = √ ↠|1i = √ ↠|0i
4 4·3 4·3·2 4·3·2·1
17
To compute the eigenvalue of the state |ni we apply the hamiltonian operator to our arbitrary state
† 1
Ĥ|ni = h̄ω â â + |ni
2
1
= h̄ω n̂ + |ni
2
1
= h̄ω n̂|ni + |ni
2
1
= h̄ω n|ni + |ni
2
1
= h̄ω n + |ni
2
In the foruth line, we replaced the operator n̂ with the number n because of the relation n̂|ni = n|ni stated
in (f). So if we have our eigenstate-eigenenergy definition as Ĥ|ni = En |ni then we see that
1
En = h̄ω n +
2
18