1

Homework 9
PHYS 370, Fall 2023
Due for 100%: 10am, October 25
Due for 80%: 10am, November 01

1. “Bosonic error correction” and “cat codes” [10 pts]. One of the most mature approaches to quantum
error correction is based on harmonic systems such as microwave photons in a resonator or vibrations of harmonically-
trapped ions. This paradigm is generally referred to as “bosonic error correction” because, from a statistical mechanics
perspective, we can think of phonons and photons as bosonic particles. Perhaps the most prominent bosonic error
correction scheme is the so-called “cat code” in which we encode a qubit into a superposition of macroscopically
distinct states reminiscent of the Schrödinger cat thought experiment.
The logical qubit states (“code words”) of the cat code are based on coherent states |αi:
•Even photon number:

|0+
L i ≡ |αi + | − αi (1)

|1+
L i ≡ |iαi + | − iαi (2)

•Odd photon number:

|0−
L i ≡ |αi − | − αi (3)

|1−
L i ≡ |iαi − | − iαi. (4)

The reason to have even and odd logical qubits is based on the dominant error mechanism: single-photon loss. Since
it is possible to measure whether the photon number is even or odd (i.e. the parity), the cat code allows us to detect
single-photon loss errors [see Ofek et al. Nature 536, 441-445 (2016)].
−
(a) Show the effect of single-photon loss by proving that â|0+ L i = α|0L i. Calculate â|αi and â| − αi [3 pts].
+ −
(b) Prove that â|1L i = iα|1L i by calculating â|iαi and â| − iαi [4 pts].
(c) Show that cat codes cannot detect two-photon loss errors by proving that â2 |0+ 2 +
L i = α |0L i [3 pts].
 2

2. Doppler cooling of atoms [10 pts]. How cold can we get our atoms by laser cooling? If the atomic transition
was infinitely narrow, then the answer in principle would be absolute zero. It turns out that spontaneous emission
from the upper state with a rate γ gives the line a finite “width”, meaning that any detuning δ . γ is essentially
“on resonance”. The finite width of a realistic atomic transition imposes a temperature limitation called the Doppler
temperature:
hγ
TD = . (5)
2kB

(a) For alkali species [Group I of the periodic table; including cesium (see below)], γ = 6 MHz. Calculate their Doppler
temperature [2 pts].
(b) Alkaline earth species (Group II of the periodic table) have a “broad” and a “narrow” transition. The broad
transition has γ = 30 MHz. Calculate its Doppler temperature [2 pts].
(c) The width of the narrow transition depends on the species. Calculate the Doppler temperature of the narrow
transition in strontium (γ = 7 kHz) [2 pts].
(d) Calculate the Doppler temperature of the narrow transition in calcium (γ = 400 Hz) [2 pts].
(e) In certain cases, it is possible to laser cool below the Doppler temperature. The absolute limit is set by the recoil
energy of a scattered photon:

~2 k 2
Trecoil = , (6)
mkB

where kB = 1.38 × 10−23 J/K, m is the atomic mass in kg, and k = 2π/λ is the magnitude of the photon’s wavevector.
Calculate Trecoil for cesium-133 with an optical transition λ = 852 nm and mass of 133 atomic mass units (1 amu
= 1.66 × 10−27 kg) [2 pts].
 3

3. Optical tweezers for neutral atoms [10 pts]. As we learned in Lecture 4, the Rabi frequency is proportional
to the electric field amplitude: Ω ∼ E. Also as we learned in Lecture 4, the light shift is given by

Ω2
∆= , (7)
4δ
where δ = ω − ω01 is the detuning of the EM field’s frequency ω from the qubit (atomic) resonance ω01 . Since the
potential felt by the atom is simply the light shift and since E 2 ∼ I for intensity I, we can rewrite this as

aI(r)
U (r) = , (8)
4δ
where we have allowed the intensity to have a spatial profile, and where the constant a which has units of Hz2 /(W/cm2 )
includes the dipole matrix element squared.

(a) A typical optical tweezer requires P0 = 1 mW and is focused to a beam size (“waist”) of w0 = 1 µm. Knowing
that the intensity of a Gaussian beam is given by
2P0
I0 = , (9)
πω02

calculate I0 in units of W/cm2 [3 pts].

2r2
 
I(r) = I0 exp − 2 . (10)
w0

The “waist” of the beam corresponds to the position r = w0 . Calculate I(r = w0 ) in units of I0 [3 pts].
(c) A typical optical tweezer is detuned from resonance by δ = −50 THz. For the power and waist listed above, what
must be the coefficient a in order for the “depth” of the tweezer to be the typical |U (r = 0)| = 10 MHz [2 pts]? (Note:
a/4δ is often called the “polarizability”).
(d) It is common to refer to trap depths in temperature units by equating kB T = hf , where kB = 1.38 × 10−23 J/K.
Calculate the temperature corresponding to the trap depth of U (0) = 10 MHz [2 pts].
 4

4. IBM Quantum Experience [10 pts]. In Lecture 14, we discussed quantum gates and circuits. In this problem,
we will learn how to use the IBM circuit composer to create a simple quantum circuit.

(a) First, we must obtain an IBM quantum account. Go to the lBM quantum website https://quantum-computing.
ibm.com, and create an IBMid (Please see the pdf version of Homework 9 on the course website to access the links).
(b) You can watch this video or use the IBM composer user guide to become familiar with creating small quantum
circuits.
(c) Use the Hadamard and the CNOT gate to create the Bell state |Φ+ i = √12 (|00i + |11i) [5 pts].
(d) Add a σ̂x or a NOT gate to the circuit to create the Bell state |Ψ+ i = √12 (|01i + |10i) [5 pts].

For parts (c) and (d), download the final statevector and the circuit as pdf or png files (or take a screenshot) and sub-
mit them. You can print out the results and attach them to your homework or email the files (tahereh2@illinois.edu).