Journal of Physics B: Atomic, Molecular and Optical Physics

TUTORIAL You may also like

- Quantum information processing with
Quantum computing with photons: introduction to superconducting circuits: a review
G Wendin
the circuit model, the one-way quantum computer, - Roadmap on quantum nanotechnologies
Arne Laucht, Frank Hohls, Niels
and the fundamental principles of photonic Ubbelohde et al.

experiments - Integration of highly probabilistic sources

into optical quantum architectures:
perpetual quantum computation
Simon J Devitt, Ashley M Stephens,
To cite this article: Stefanie Barz 2015 J. Phys. B: At. Mol. Opt. Phys. 48 083001 William J Munro et al.

View the article online for updates and enhancements.

This content was downloaded from IP address 161.122.203.155 on 05/09/2023 at 02:25

 Journal of Physics B: Atomic, Molecular and Optical Physics

J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 (25pp) doi:10.1088/0953-4075/48/8/083001

Tutorial

Quantum computing with photons:

introduction to the circuit model, the one-
way quantum computer, and the
fundamental principles of photonic
experiments
Stefanie Barz
Clarendon Laboratory, Department of Physics, University of Oxford, Parks Road, Oxford OX1 3PU, UK

E-mail: [email protected]

Received 24 June 2014, revised 15 December 2014

Accepted for publication 30 December 2014
Published 20 March 2015

Abstract
Quantum physics has revolutionized our understanding of information processing and enables
computational speed-ups that are unattainable using classical computers. This tutorial reviews
the fundamental tools of photonic quantum information processing. The basics of theoretical
quantum computing are presented and the quantum circuit model as well as measurement-based
models of quantum computing are introduced. Furthermore, it is shown how these concepts can
be implemented experimentally using photonic qubits, where information is encoded in the
photons’ polarization.
Keywords: quantum information, quantum computing, photonics

(Some figures may appear in colour only in the online journal)

1. Introduction Quantum computers are expected to play an important role

in future information processing since they can outperform
Over the last decades, the omnipresence of computers has classical computers at many tasks. Their importance was rea-
revolutionized our lives in the dawn of a new information age. lized as early as 1982 [1] when Feynman pointed out that they
At the same time, computers have grown smaller and faster due can simulate quantum systems, whose properties are too
to the miniaturization of transistors—the most basic computa- complex to be calculated with a classical computer. It was
tional element. A celebrated empirical trend, known as shown in the following decade that quantum computers are
Mooreʼs law, states that the number of transistors in a computer superior to classical computers in various tasks. One of the first
and thus its computing power doubles every two years. algorithms to demonstrate an improvement over the classical
Obviously, this exponential growth cannot continue forever analog was the Deutsch–Josza algorithm, which determines if a
and at some point the basic building blocks of computers will function is constant or balanced [2, 3]. While this algorithm has
reach a size where the laws of quantum physics become no direct application, it inspired the subsequent development of
important. On the other hand, it has been realized that this other algorithms like Shorʼs algorithm and Groverʼs algorithm
seemingly-fundamental limitation opens up new possibilities which both provide a practical benefit. Shorʼs factoring algo-
for information processing and paves the way for a completely rithm facilitates the factorization of large numbers into their
new kind of computing: the field of quantum computing. prime factors in polynomial time on a quantum computer [4],

0953-4075/15/083001+25$33.00 1 © 2015 IOP Publishing Ltd Printed in the UK

J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

and Groverʼs algorithm enables searching in an unordered list

with a quadratic speed-up compared to the classical case [5].
Recently, another quantum algorithm was invented which
solves certain systems of linear equations with exponential
speed-up compared to a classical computer [6, 7].
This tutorial aims for introducing the basic principles of
quantum computing and their application in experiments with
photonic systems. Photons allow the encoding of information
in various degrees of freedom; here, we will mainly focus on
polarization. Polarization-encoded systems are well-suited for
quantum computing due to their low decoherence and the
simple realization of single-qubit gates. The challenge in Figure 1. The Bloch sphere is used for the geometric visualization of
photonic quantum computing is the realization of two-qubit qubits.
gates, which are necessary for universal quantum computing.
While at first sight it seems that strong optical nonlinearities the quantum-mechanical analog of classical binary bits and
are required for realization of those gates, it was shown in can take infinitely many values. These qubits are quantum-
2001 that efficient quantum computing is possible using only mechanical states, which in experiments are represented by
linear optics, single-photon sources and detectors [8]. states of atoms, photons, nuclei, etc. A qubit can be described
as a superposition of basis states, ∣ 0〉 and ∣ 1〉:

2. Outline ψ =α 0 +β 1 , (3.1)
where α, β are complex numbers and α 2 + β 2 = 1. The
The tuturial is structured as follows. In section 3 the basic
states ∣ 0〉 and ∣ 1〉 create an orthonormal basis of a Hilbert
principles of theoretical quantum computing are presented. The
space and are often called computational-basis states [9].
quantum circuit model is introduced, where a computation is
Whereas it is possible to determine the state of a classical
performed by a quantum circuit acting on quantum states. In
bit in one single measurement, a measurement in quantum
section 4, measurement-based models of quantum computing
mechanics gives a specific result only with a certain prob-
are presented, where quantum information is processed by
ability. If a measurement on the state ∣ ψ 〉 is performed, the
sequences of adaptive measurements. The one-way quantum
outcome zero is obtained with the probability α 2 and the
computer, a special type of measurement-based quantum
result is one with the probability β 2 . After the measurement,
computer, is introduced, and it is shown that single-qubit
the qubit is in the state ∣ 0〉 or ∣ 1〉, depending on the out-
measurements on highly-entangled resource states perform
come [10].
quantum computation. Further, it is presented how this concept
can be applied to implement secure delegated quantum com-
putations, a recently discovered feature of quantum computers. 3.1.1. Representation on the Bloch sphere. The state ∣ ψ 〉 can
In section 5, the fundamental principles of photonic quantum be represented geometrically on a unit sphere in three
computing are presented and it is shown how single-qubit and dimensions (see figure 1), called the Bloch sphere [11]. For
multi-qubit gates can be implemented experimentally using this, the state ∣ ψ 〉 can be rewritten in the following form:
polarization-encoded systems. Furthermore, it is shown, how ⎛θ⎞ ⎛θ⎞
ψ = cos ⎜ ⎟ 0 + eiϕ sin ⎜ ⎟ 1 . (3.2)
single photons can be generated experimentally. The section is ⎝2⎠ ⎝2⎠
concluded with an example of a photonic quantum computing
experiment and it is shown how the introduced concepts can be In this representation, θ and ϕ are real numbers which
applied in experiments. Finally, this tutorial ends with a con- correspond to the polar angle and the azimuthal angle,
clusion and an outlook in section 6. respectively. The description of quantum states as points on
the Bloch sphere is useful for the visualization of single-
qubits and operations on single-qubits.
3. Quantum computing The most frequently used states in quantum information
lie on the axes of the Bloch sphere:
This first section briefly reviews the basic elements of quantum
computing. The fundamental units—the qubits—and the basic 1 1
+ = ( 0 + 1 ), − = (0 − 1 ) (3.3)
building blocks of a quantum computer—the quantum gates— 2 2
are introduced. Furthermore, the circuit model, the most pro-
minent circuit model of quantum computing, is introduced. on the x-axis,

1 1
3.1. Classical bit versus quantum bits +i = ( 0 + i 1 ) , −i = ( 0 − i 1 ) (3.4)
2 2
The fundamental unit of a classical computer is a bit which
can take binary values: zero or one. Quantum bits (qubits) are on the y-axis, and the basis states ∣ 0〉 and ∣ 1〉 lie on the z-axis.

2
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

If the qubits are written in a vector notation: density operator (or density matrix) is defined as:
⎡ ⎤ ⎡ ⎤
0 = ⎢ 1⎥, 1 = ⎢ 0⎥, (3.5) ρ= ∑pi ψi ψi . (3.16)
⎣ 0⎦ ⎣ 1⎦ i

it is easy to see that these states exactly correspond to the A quantum state is pure if pi = 1 for only one i and all other pj,
eigenvectors of the Pauli matrices: j ≠ i , are equal to zero. Whereas the state-vector formalism
of the previous sections describes only pure states, i.e.
( )
σx = X = 0
1
1 , σ = Y = 0 −i ,
0
y
i 0 ( ) systems that are with certainty in a state ∣ ψi 〉, density matrices
can also represent mixed states.
( )
σz = Z = 1
0
0 .
−1
(3.6) General properties of the density operator are:
• ρ is trace-preserving: Tr (ρ) = 1,
• ρ is positive semidefinite: ρ ⩾ 0 (meaning that the
3.1.2. Multi-qubit states. States of multiple qubits can be
eigenvalues are non-negative), and
described using the same formalism. For two qubits, a set of
• ρ is self-adjoint: ρ = ρ†.
four possible basis states is given by:
00 = 0 ⨂ 0 , (3.7) For completely mixed states, the density matrix becomes
ρ = 1 d I d , where I d is the d-dimensional identity matrix.
01 = 0 ⨂ 1 , (3.8) This representation is not unique, meaning that different
mixtures can lead to the same density matrix.
10 = 1 ⨂ 0 , (3.9)
Mixed states of a single qubit can also be represented on
11 = 1 ⨂ 1 . (3.10) the Bloch sphere as each density matrix can be rewritten as
follows:
They form a basis for the product Hilbert space of the two
qubits. A general two-qubit state can be written as a I + r ⃗ · σ⃗
superposition of these four basis states: ρ= . (3.17)
2
ψ = α 00 + β 01 + γ 10 + δ 11 , (3.11)
The Bloch vector r ⃗ can be calculated from r ⃗ = Tr (ρ · σ ⃗ )
where α 2 + β 2 + γ 2 + δ 2 = 1. Similar to the with σ ⃗ = (σx , σy, σz ). A state is pure and thus lies on the
single-qubit case, a measurement gives a result (00, 01, 10, surface on the sphere, if and only if r ⃗ = 1. The Bloch vector
or 11) with certain probability, α 2 , β 2 , γ 2 , or δ 2 . of a general mixed state lies inside the sphere.
However, a simple analog of the Bloch-sphere representation
for multiple qubits is not known.
3.1.4. Measures for experiments. The density matrix of a
Two-qubit states that cannot be separated or be written as
quantum state can be used to analyze various properties of a
a product of two single-qubit state are called entangled [12]:
state [12, 17]. In experiments, these properties are very useful
1 for quantitatively verifying the quality of a quantum state.
ψ− = ( 01 − 10 ) ≠ (α 0 + β 1 ) ⨂ (α′ 0 + β′ 1 ). (3.12)
2
A useful mean for the discrimination of pure and mixed
An important set of entangled two-qubit states are the state is the purity P which is defined via [18, 19]:
maximally-entangled Bell-states [13–15]:

ψ± =
1
( 01 ± 10 ) , (3.13)
P = Tr ρ2 , ( ) (3.18)
2
where Tr is the trace. P = 1 for pure states and P < 1 for mixed
1
ϕ± = ( 00 ± 11 ) . (3.14) states. For a totally mixed state of dimension d, the purity is
2 given by 1/d.
These show strict correlations or anti-correlations and also The fidelity F of a general quantum state ρ determines
form an orthonormal basis. how close that state is to a desired state. For a pure state ∣ ψ 〉 it
A general multi-qubit state, describing n qubits, can also is defined via:
be expressed in terms of state vectors:
F (ρ , ψ ) = ψ ρ ψ . (3.20)
2n
ψ = ∑α i x1x2…x n , (3.15) The fidelity of two mixed states ρ and ρ̃ is given by [20]:
i=1
2
with 2n different probability amplitudes αi, with ∑i ∣αi∣2 = 1,
and xi ∈ {0, 1}.
( (
F ( ρ , ρ˜ ) = Tr ρ˜ ρ ρ˜ )) . (3.21)

Another way to quantify the mixedness of a quantum

3.1.3. Density operators. An alternate way to describe state is the measure of entropy which determines how much
quantum states is with the density matrix formalism [16]. If information is present when compared to the possible
a quantum system is in a state ∣ ψi 〉 with probability pi, its maximum [16]. The von Neumann entropy of a quantum

3
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

state ρ is defined by: building blocks are necessary to build a universal quantum
computer, meaning that it can be programmed to perform any
S (ρ) = −Tr ρ log2 ρ = −∑λ i log2 λ i , (3.21) computational task.
i

where λi are the eigenvalues of ρ. The von Neumann entropy

is zero for a pure state and equal to log2 (d ) for a totally mixed 3.2.1. Single-qubit gates. A single qubit gate is a unitary
state of dimension d. operation U that takes a single qubit ∣ ψ 〉 = a ∣ 0〉 + b ∣ 1〉 as
For experiments, a more useful form is the linear entropy an input and transforms it into an output state
which can be calculated directly from the density matrix ∣ ψ ′〉 = a′∣ 0〉 + b′∣ 1〉 with:
without the necessity of any diagonalization. It is directly
related to the purity of a quantum state and obtained from the ψ′ = U ψ . (3.26)
von Neumann entropy by approximating the logarithm with
In the circuit formalism, this transformation is depicted as
the first-order term of its Taylor expansion. The linear entropy
[25, 26]:
is defined via
d
S (ρ) =
d−1
(
1 − Tr ρ2 ( ) ) = d −d 1 (1 − P), (3.22) (3:27)

and its values range from zero (pure state) to one (totally The gate changes the amplitude coefficients, which can
mixed state) [18, 19]. be seen when the transformation is written in form of a
The density matrix can also be used to quantify the matrix:
amount of entanglement of a state. One measure which is
a′ = u11 u12
often used in experiments is the concurrence [21, 22]. The
concurrence of a density matrix ρ of a two-qubit system is
() (
b′ u21 u22 )( )a
b
. (3.28)

defined by:
The unitarity of the transformation follows from the fact that
C = max ( λ1 − λ2 − λ3 − )
λ4 , 0 , (3.23) the norm must be preserved:

where λi are the eigenvalues of the matrix ρρ̃ in decreasing ψ ψ = ψ ′ ψ ′ = ψ U †U ψ = 1 → U †U = I . (3.29)

order with
From this it follows that all quantum gates are reversi-
( ) (
ρ˜ = σy ⨂ σy ρ* σy ⨂ σy ) (3.24)
ble [9, 24].
and ρ* being the complex conjugate of ρ. Important single-qubit gates in quantum computation are
The two-tangle τ, where τ = C2, is another value the Pauli operators σx , σy, and σz . Beyond that, there are three
commonly used to characterize density matrices obtained major gates, that are often used in quantum computing. The
experimentally [23]. Hadamard gate H turns basis states into superposition states
and vice versa:
3.2. The circuit model of quantum computation

The main components of a classical computer are the memory

H= ( )
1 1 1
2 1 −1
. (3.30)

and the processor. Binary logic gates are carried out on

classical bits; which and how many gates are used depends on A phase gate or S gate adds a phase of π/2 to the
the underlying program [9, 24]. In quantum physics, infor- computational basis state ∣ 1〉 :
mation is stored in the qubit and quantum logic gates acting
on qubits can process the information, similar to classical
information processing:
( )
S= 1 0 ,
0 i
(3.31)

and the T gate or π/8 gate:

⎛1 0 ⎞
T= ⎜ ⎟ (3.32)
⎝ 0 exp (iπ 4) ⎠

(3:25) adds a phase of π/4 to the computational basis state ∣ 1〉 and

A comparison of the efficiency of both concepts shows enables universal quantum computing. Two useful algebraic
that N input qubits can store 2N (classical) amplitude coeffi- identities are given by: H = (X + Z ) 2 and S = T2.
cients. Information can thus be stored and obtained much All single-qubit gates can be represented geometrically
more efficiently in a quantum circuit than in a classical circuit. on the Bloch sphere. The application of a X (Y, Z)-Pauli gate
The basic building blocks of a quantum circuit, single-qubit is equivalent to a rotation of π about the x (y, z)-axis of the
and two-qubit gates, are described in the next paragraph. As Bloch sphere. Thus, the Pauli gates can generate rotations
will be shown, only a few different types of gates or basic about the three axes of the sphere. For example, a rotation

4
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

about the x-axis can be written as: −1. The symbol

⎛ i θX ⎞
R x (θ ) = exp ⎜ − ⎟ (3:40)
⎝ 2 ⎠
⎛θ⎞ ⎛θ⎞
= cos ⎜ ⎟ I − i sin ⎜ ⎟ X , (3.33) is used in quantum circuits for the representation of CPhase
⎝2⎠ ⎝2⎠
gates.
These two gates are also called entangling gates, since
where similar equations also exist for Y and Z gates. they can perform entangling operations. For example, a
Furthermore, a general rotation Rn̂ (θ ) about an arbitrary axis quantum circuit consisting of a Hadamard gate and a CNOT
nˆ = (n x , n y , n z ) can be decomposed into Pauli gates [9]: gate:
⎛ i θnˆ · σ ⃗ ⎞
R nˆ (θ ) = exp ⎜ − ⎟
⎝ 2 ⎠
⎛θ⎞ ⎛θ⎞ (3:41)
⎝2⎠ ⎝2⎠ )
= cos ⎜ ⎟ I − i sin ⎜ ⎟ ( n x X + n y Y + n z Z . (3.34)
can transform a product state ∣ xy〉, x, y ∈ {0, 1}, into the
For example, the Hadamard gate can be created out of two
following maximally entangled Bell states:
different rotations, first a rotation of π about the z-axis,
followed by a rotation of π/2 about the y-axis. U
00 ⟶ ( 00 + 11 ) 2 (3.42)
U
3.2.2. Multi-qubit gates and controlled operations. Multi- 01 ⟶ ( 01 + 10 ) 2 (3.43)
qubit gates take multiple qubits as input and perform U
operations on them. In the circuit formalism, this is 10 ⟶ ( 00 − 11 ) 2 (3.44)
depicted as [26]: U
11 ⟶ ( 01 − 10 ) 2. (3.45)

(3:35) 3.2.3. Universal set of gates. Arbitrary multi-qubit gates can

be generated by a universal set of single- and multi-qubit
The operation can be conditioned on the state of one or gates. A widely-used set of gates consists of the CNOT gate,
more qubits. These qubits are called control qubits in contrast the Hadamard gate and the π/8 gate. Using only these three
to the qubits on which the operation is performed, the target gates, any computation can be realized. In more detail: any
qubits. For example, a two-qubit controlled unitary operation unitary operation can be approximated to arbitrary accuracy
(CU) applies a unitary operation U on the target qubit if the using only these gates [9, 24, 26–29]. Here, also other non-
control is in the state ∣ 1〉: trivial phase gates can in principle be used instead of the π/8
gate. Another universal set of gates is, for example, the set of
all single-qubit gates, together with a CNOT gate.
(3:36) This statement has particular importance to experimental
efforts. If this gate set can be physically implemented and the
Important two-qubit gates for quantum computing are the gates arbitrarily concatenated, it will thereby be possible to
controlled-NOT gate (CNOT or CX) and the controlled-phase physically realize any unitary transformation.
gate (CPhase or CZ). Acting on two input qubits ∣ i〉 and ∣ j〉,
(i, j ∈ 0, 1), the CNOT gate performs the following opera-
tion: 4. Measurement-based models
CNOT i j = i i ⊕ j , (3.37) In the circuit model, which is described in the previous
where ⊕ is the binary addition. Thus, the state of the target section, quantum information is processed by applying
qubit is changed from ∣ 0〉 to ∣ 1〉 (or vice versa) if the control quantum gates, which realize a coherent unitary evolution
qubit is in the state ∣ 1〉. In a quantum circuit, the CNOT gate is [30]. In contrast, in measurement-based models, quantum
depicted by the symbol: information is processed by sequences of adaptive measure-
ments [31, 32]. Among measurement-based models there are
two different approaches: the teleportation-based model [33],
(3:38)
which is based on Bell-pairs and two-qubit measurements,
and the one-way model which consists of highly-entangled
The CPhase gate also acts on two input qubits ∣ i〉, ∣ j〉 and multi-particle states and single-qubit measurements. Both
performs the transformation: models are equivalent [34]; it can be shown that they are
conceptually closely related and rely on the same primitives
CPhase i j = (−1)ij i j . (3.39)
[34–38]. In general, measurement-based quantum computing
If the input qubits are in the state ∣ 11〉, they acquire a phase of (MBQC) is related to different fields of physics, for example

5
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

the values of a and b:

(4:2)

The measurement symbol on the right of this

circuit denotes a measurement in the computational basis. If
the qubits are found to be in the state (∣ 00〉 + ∣ 11〉 ) 2 , then
Figure 2. The figure illustrates the principle of quantum teleporta- the output is a = b = 0 (for the state (∣ 01〉 + ∣ 01〉 ) 2 , it is
tion. Alice and Bob share an entangled state. Alice performs a so- a = 1, b = 0; for the state (∣ 00〉 − ∣ 11〉 ) 2 : a = 0, b = 1; and
called Bell-state measurement on the state ∣ α〉 and her half of the
for the state (∣ 01〉 − ∣ 01〉 ) 2 : a = b = 1). These classical
entangled state. This Bell-state measurement projects the input states
onto one of the four Bell states. Alice shares the outcome of this outputs determine whether additional Pauli gates need to be
measurement with Bob via a classical communication channel and applied to the teleported state (the unitary U in figure 2) in
Bob chooses the unitary operation U accordingly (details see text). order to obtain the state ∣ α〉.
After applying the operation U to his half of the entangled state,
Bobʼs qubit is in state ∣ α〉.
4.1.2. The Gottesman–Chuang teleportation trick. In 1999,
Gottesman and Chuang published a ‘teleportation’ trick
entanglement theory, topology, graph theory, and mathema- which enables universal quantum computation using only
tical logic [39]. single-qubit operations, Bell-basis measurements and
entangled states as resources [41]. Their scheme—also
4.1. Teleportation-based quantum computing
known as teleporting a state ‘through’ a unitary operation—
is a generalization of quantum teleportation and reduces the
The teleportation-based model uses teleportation [33, 40] as a required resources [41]. Instead of directly applying a gate to
way to realize unitary transformations. Historically, the field a state, that state is teleported using a modified resource as
started with the invention of the Gottesman–Chuang tele- compared to the original teleportation protocol [33, 40].
portation trick, described below. This trick lead to the In more detail, an operation U can either be applied to a
invention of a variety of teleportation-based concepts. Fur- state ∣ α〉, or that state ∣ α〉 can be teleported using the modified
thermore, it is the basis of a landmark paper in the field of Bell state (I ⨂ U ) ∣ ϕ+〉 as a resource, which leads to the same
photonic quantum computing which shows that limitations in output up to local Pauli corrections. The following circuit
photonic quantum computing due to missing interactions in shows a state ∣ α〉 which is first teleported (dashed box,
photonic systems can be overcome [8]. describes in the previous section) and then experiences a
unitary gate U:
4.1.1. Quantum teleportation. The aim of quantum
teleportation is to send a quantum state ∣ α〉 from A (Alice)
to B (Bob), where A and B can be far apart, and Alice is only
allowed to transmit classical information to Bob [33, 40].
Furthermore, Alice does neither know the quantum state, nor (4:3)
can she determine it since she holds only a single copy.
Teleportation enables Alice to send the state ∣ α〉 to Bob, by
utilizing an entangled photon pair and classical
communication. The basic principle of quantum
teleportation is depicted in figure 2.
This is equivalent to the following circuit, where the unitary
It is important to note that quantum teleportation does not
operation is absorbed in the entangled resource state:
allow faster-than-light communication. The teleportation
protocol requires Alice to send classical information to Bob.
This process is clearly limited by the speed of light.
In more detail, the quantum circuits that accomplishes the
(4:4)
teleportation of a state ∣ α〉 is the following:

For example, the unitary U could be a Hadamard gate

(4:1) and instead of applying this gate directly to the state ∣ α〉, one
teleports ∣ α〉 using a modified resource
(I ⨂ H ) ∣ ϕ+〉 = (∣ 0 + 〉 + ∣ 1 − 〉 ) 2 .
The advantages of this method are obvious: instead of
Here, the double lines carry classical bits and the box performing operations on unknown states, it is just necessary
‘Bell’ represents a Bell-state measurement, which determines to construct known states as offline resources. The operation

6
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

U could be a logic gate that is difficult to implement, but the teleportation, and single-qubit measurements—are easily
creation of the resource state might be much easier. realizable in optical experiments.
Furthermore, it is no longer necessary to perform probabilistic
gates. 4.1.3. The Knill–Laflamme–Milburn (KLM) scheme. In their
The advantage of this teleportation trick becomes even seminal paper in 2001, KLM showed that efficient quantum
more obvious in the case of multi-qubit gates like the CNOT computation is possible using only beam splitters, phase
gate. Applying the gate to two qubits ∣ α〉∣ β 〉 is equivalent to shifters, single-photon sources and photo-detectors [8].
absorbing it in the preparation of the resource state. If the state For many years, it was strongly believed that quantum
∣ α〉∣ β 〉 is first teleported and then the CNOT gate is applied, computing with only linear optics is not possible due to the
after implementing corrections dependent on the Bell missing interaction between photonic qubits and the resulting
measurement outcome, we obtain: lack of entangling gates. KLM revolutionized linear-optics
quantum computing (LOQC) by developing an efficient
scheme based on the Gottesman–Chuang teleportation trick.
They took advantage of the fact that there is a hidden
nonlinearity in the photon detection process and transferred
this nonlinearity to the qubits via measurements to enable
universal computing.
In their paper [8], they first show that non-deterministic
quantum computation is possible with linear optics. For this
(4:5) demonstration, they use dual-rail encoded qubits, where the
information is stored in the photon number of an optical
mode. They show that a non-deterministic sign change,
dependent on the photon number, is possible:
α 0 0 + α1 1 + α2 2 → α 0 0 + α1 1 − α2 2 . (4.7)
Their gate—the so-called NS gate—just requires photon
counters that are able to count the number of photons in one
mode. Applying the NS gate twice, they can achieve an
entangling gate—a conditional sign flip—with a success
This is again equivalent to the following circuit, where the probability of 1/16 through projective measurements. Figure 3
CNOT gate is absorbed in the resources and which leads to shows the basic principle of the NS gate and how to use it in
the same output state CNOT ∣ α〉∣ β 〉: order to achieve a conditional sign flip. A detailed description
of the NS gate and the KLM scheme in general can be found
in [42] or in [43].
By using a generalized, near-deterministic form of
teleportation and by applying the Gottesman–Chuang tele-
portation trick, they further show that this success probability
can be increased to n2/(n + 1)2 with 2n ancilla qubits. Here, it
is important to note that a complete Bell state measurement is
impossible for photonic qubits encoded in one degree of
(4:6) freedom (see [44] for Bell-state measurements using hyper-
entanglement and [45] for a review on Bell-state measure-
ments). This is the reason for the use of the 2n ancilla qubits,
which enable near-deterministic teleportation. Thus, an
arbitrarily high success probability is possible at the cost of
ancillary resources—the more ancilla qubits, the higher the
success—which makes the scheme quite resource-intensive.
Their final and main result, robust LOQC being possible with
polynomial resources, provides practical scalability of photo-
nic quantum computing experiments [46].
The resource state CNOT ∣ ϕ+〉∣ ϕ+〉 can for example be Often, the KLM model of quantum computing is referred
created out of two three-qubit Greenberger–Horne–Zeilinger to as the photonic quantum circuit model. However, a closer
(GHZ) states; where an n-qubit GHZ state is an entangled look reveals that although the KLM model superficially
quantum state of the form ∣ GHZ〉 = (∣0〉 ⨂ n + ∣1〉 ⨂ n ) 2 . resembles the circuit model, it is still a measurement-based
The Gottesman–Chuang scheme was very important for scheme [47]. The KLM scheme is based on entangled ancilla
the invention of teleportation-based concepts and for the photon pairs and thus provides entanglement from the very
development of quantum computing with linear optics since beginning. The photons do not interact as in standard circuit
their requirements—GHZ states, Bell measurements, models, but the interaction is created via the application of a

7
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

Figure 3. The basic principle of the NS gate and how to use it for conditional sign flips. (a) The NS gate realizes a non-deterministic phase
shift NS on one mode. (b) It is implemented by using additional modes and a network of phase shifters and beam splitters. A phase shifter
adds a phase of eiϕ to an optical mode, a beam splitter splits an incidents beam into two parts (see section 5.2 for a full mathematical
description) and adds phases to the output modes. The ratio of transmission and reflection of the beam splitter and the phases, acquired from
the phase shifter and the beam splitter, determine the phase shift NS. For certain settings [8], one can achieve that a phase shift of NS = −1 for
the case of two photons entering the input, ∣ in〉 = ∣ 2〉 and thus obtain the operation α0 ∣ 0〉 + α1 ∣ 1〉 + α2 ∣ 2〉 → α0 ∣ 0〉 + α1 ∣ 1〉 − α2 ∣ 2〉.
The phase shift has been applied successfully to the upper mode, if one and zero photons have been registered in the ancilliary mode,
respectively. (c) The NS gate can be used to implement a conditional phase shift. Two qubits are encoded into four spatial modes, 1, 2, 3 and
4, respectively. If modes 1 and 3 both contain a photon, ∣11〉13, the state after the beam splitter will be ∣02〉13 + ∣02〉13 and the NS gates will
add a phase of ‘−1’ to that state. After the second beam splitter, the state will then be −∣11〉13. If no or only one photon enter the modes 1 and
3, no phase shift will be applied.

teleportation-based scheme. Consequently, the KLM scheme eigenstate with eigenvalue +1 for all stabilizer operators:
for quantum computing is a truly teleportation-based and thus
a measurement-based scheme. Sa G = + G . (4.9)

4.2. One-way quantum computer The resource state for the one-way quantum computer is
special kind of graph state, known as cluster state, where the
The basis of the one-way quantum computer is a highly- underlying graph forms a 2D lattice. A graph or cluster state
entangled resource state. Single-qubit measurements on that can be created by preparing a qubit in the ∣ + 〉 state for each
state enable the processing of quantum information [30–32]. vertex and using CPhase gates to entangle each pair of qubits
The computational power of the one-way model is strongly which should be connected by an edge as nearest neighbors
related to the properties of the underlying resource state. It is (see figure 4).
required that every possible quantum state can be created out The choice of nearest-neighbor two-qubit interactions
of the resource state just by single-qubit measurements. Since defines the structure of the cluster state, which determines
single-qubit measurements cannot create entanglement, also the basic type of quantum circuit it can implement.
entanglement must be included in the resource state itself. Different graph states and cluster states are shown in
Possible resource states for this task are graph states in the figure 5. Certain families of cluster states (combined with
form of different two-dimensional (2D) lattices [48, 49]. single-qubit measurements and feed-forward) comprise a
set of resources sufficient for universal quantum
4.2.1. Graph states. A graph state is a multi-qubit quantum computing.
state which can be represented by a mathematical graph G(V,
E) with vertices V and edges E. The vertices of a graph state 4.2.2. One-way computation. A computation in the one-way
correspond to the physical qubits whereas the edges indicate model is described by a sequence of consecutive single-qubit
an entangling operation between the qubits. measurements and a feed-forward protocol. The basic
Mathematically, a graph state can be described in the principle of one-way computation is depicted in figure 6.
stabilizer language, which was invented by Gottesman Measuring a qubit of a one-dimensional cluster state in
[43, 49, 50]. For every vertex a of a graph, an operator Sa the basis:
can be defined:
1
Sa := σ xa ∏ σzb, (4.8)
+ϕ =
2
(0 + eiϕ 1 ) (4.10)
b ∈ Na
has the effect of applying the single-qubit rotation
where Na are all vertices in the neighborhood of vertex a. The R z (−ϕ) = exp (iϕσz 2) on an encoded qubit in the cluster
corresponding graph state ∣ G〉 is defined as the unique up to a Hadamard operation.

8
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

Figure 5. The figure shows different graphs states. (a) A three-qubit

Figure 4. General cluster state, where the blue circles denote the entangled state. (b) A four-qubit GHZ state. (c) A four-qubit cluster
physical qubits and the edges between the qubits denote state. GHZ and cluster states have different entanglement structures
entanglement. and cannot be converted into each other by local operations. For the
case of three qubits, shown in (a), the three-qubit GHZ and the three-
qubit cluster state are the same. (d) The horseshoe cluster state is a
two-dimensional cluster state and allows for the generation of two-
Measuring a qubit in the computational basis ∣ 0〉 or ∣ 1〉 qubit logic gates. (e) The brickwork state is universal resource for
measurement-based quantum computing.
disconnects the qubits from the cluster and deletes all affected
edges.
In the case where the outcome is found to be zero, the X–Y plane of the Bloch sphere:
computation is correct, however if the outcome is found to be
one, a Pauli error is introduced. This error is corrected by
applying a feed-forward mechanism that compensates for the
known errors by adapting further measurement angles [51].
In the following, I will describe in more detail how a (4:15)
measurement on an entangled state leads to a computation;
the derivation follows [52]. Starting with a simple teleporta- A measurement in the basis ∣ ±ϕ 〉 leads to a rotated
tion circuit, an input state R z (−ϕ) ∣ + 〉 is teleported to the output qubit up to a Pauli X correction depending on the
output up to a local Pauli correction [52]: measurement outcome. Thus single-qubit measurements can
implement general HR z (ϕ) rotations. These are sufficient to
implement arbitrary single-qubit rotations, since every
rotation can be decomposed into rotations about the x-axis
and the z-axis using the Euler angles, Rz(γ) Rx(β) Rz(α). By
applying the relations H 2 = I and HRz(ϕ) H = Rx(ϕ), the
(4:11) Euler angles can be rewritten in terms of single-qubit rotations
Rz:
The CNOT gate can be transformed to a CPhase gate
applying the relation R z (γ ) R x (β ) R z (α ) = H ( HR z (γ ) )( HR z (β ) ) ( HR z (α ) ). (4.16)

CNOT = (I ⨂ H ) CPhase (I ⨂ H ). (4.12) In order to obtain such a general rotation, single-qubit

teleportations can simply be concatenated:
Additionally, the rotation R z (−ϕ) can be implemented within
the circuit and all Hadamard gates can be absorbed in the
measurement basis or the target input qubit, which leads to
the following circuit:

(4:13)

with

out = HZ mR z (−ϕ) + = X mHR z (−ϕ) + . (4.14) (4:17)

The output state is equal to

The rotation R z (−ϕ) in the upper wire commutes with the
CPhase gate and thus can also be absorbed in the
out = X mHR z (−γ ) X lHR z (−β ) X kHR z (−α ) + (4.18)
measurement basis, leading to a general measurement in the

9
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

Figure 6. Principle of one-way computation. The pattern of single-qubit measurements on the cluster state determines the quantum circuits
that is implemented. The figure illustrates how two different measurement patterns (yellow and red) on the same cluster state can lead to two
different circuits. In order to implement such a pattern, one usually starts measuring the qubits from the left side of the cluster state and the
continues to the right. This figure also illustrates the differents between the horizontal and vertical lines in the cluster state. The horizontal
lines represent the logical qubits and the vertical lines are used to implement entangling gates between these qubits.

= X mZ lX kHR z (−1)l γ R x (−1)k β R z (−α ) + , (4.19) ∣ ψ 〉 = R z (−ϕ) ∣ + 〉 and ∣ ψ ′〉 = R z (−ϕ′) ∣ + 〉 converts the
( ) ( ) circuit into the following:
where the Pauli corrections were commuted through all
rotations to the left. The dependency of the measurement
bases on previous measurement outcomes also defines a
temporal direction of the computation.
Up to now, only one-dimensional cluster states were
introduced, which can still be simulated efficiently classi-
cally [53]. For universal quantum computing, 2D cluster
states and two-qubit gates such as the CPhase gate are
necessary. These CPhase gates can be implemented via
vertical lines in the cluster state. This can be illustrated with (4:21)
the following circuit where an entangled input state CPhase
∣ ψ 〉∣ ψ ′〉 is teleported up to local Pauli corrections combining This circuit prepares a cluster state (dashed box), the
two single-qubit teleportation circuits (here, the entangling horseshoe cluster state shown in figure 5(e), by applying
CPhase gate is shown in the dashed box as part of the CPhase gates to qubits in a state ∣ + 〉. Subsequent measure-
circuit): ments, specifically in the ∣ +ϕ 〉 basis (∣ +ϕ′ 〉 basis) for the
upper (lower) qubit, perform a computation on these encoded
qubits. The remaining qubits are in the output state:
out = (H ⨂ H ) ( Z m ⨂ Z n )
(
× R z (−ϕ) ⨂ R z (−ϕ′) ) CPhase + + (4.22)

= ( X m ⨂ X n ) (H ⨂ H )
(4:20)
Using the same transformations as in the above single-
(
× R z (−ϕ) ⨂ R z (−ϕ′) ) CPhase + + . (4.23)

qubit teleportation and defining the input as This confirms that the vertical lines in cluster states lead to the

10
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

generation of entangling gates on the encoded qubits. If the

underlying cluster state is large enough, any possible quantum
computation can be performed.

4.2.3. Local complementation. Different types of graphs can

be transformed into each other by applying a set of local
operations. Experiments which can only prepare a restricted Figure 7. Example of a local complementation (LC) procedure. In
set of states thus have access to a much larger variety of states the first step, LC on vertex 2 creates an edge between vertices 1 and
3. LC on vertex 3 then deletes the edge between the vertices 1 and 2
by simply applying local gates. This is a very useful and creates a new edge between the vertices 1 and 4 and between
procedure since it easily enables the preparation of a set of vertices 2 and 4. The final LC on vertex 2 then deletes the edges
quantum states and thus the realization of different quantum between the vertices 3 and 4.
algorithms.
The local complementation procedure is a simple
4.3. Comparison of the different quantum computing models
graph transformation rule which utilizes a local Clifford
operation on a quantum state to effect a transformation of its At first sight, the teleportation-based model and the one-way
associated graph [49, 54, 55]. A Clifford operation maps a quantum computer seem to be very different models of
Pauli operator to the same or another Pauli operator; or, in quantum computing. Figure 8 provides a schematic overview
other words, the Clifford group consists of all unitary of both models and their characteristics [35–37, 47, 57].
operators which map the Pauli group to the Pauli group Teleportation-based quantum computation, which arose
under conjugation. from the ideas of Gottesman and Chuang [41] and KLM [8],
A transformed graph or, more precisely, the local relies on teleportation using bipartite entangled pairs and two-
complement of a graph G at a vertex a, τa(G), can be qubit measurements. In some aspects, it is similar to the
obtained by complementing the neighborhood Na of a; the standard circuit model: interactions are required during the
neighborhood Na of a consists of all vertices which are performance of the algorithm, and no actions are performed
connected to a [49]: unless a non-identity gate is applied.
This is different from the one-way quantum computer
τa (G ) := G + Na . (4.24) [31, 32] where no quantum interactions are required after the
preparation of the multi-partite entangled resource (the cluster
In other words, all edges between the vertices of the state). The cluster state itself is independent of the computa-
neighborhood of a vertex a are erased; if there are tion, which is performed by single-qubit measurements on the
unconnected vertices, new edges are created between them cluster state. The one-way model also presents a paradigm
(see figure 7). shift in the theory of quantum computing since it is the only
The corresponding action on the graph state is described model which clearly separates the quantum and the classical
by the local complementation rule: a graph state ∣ τa (G ) 〉 parts of a computation.
equivalent to a state ∣ G〉 under local Clifford transformations However, both models are measurement-based models
Uτa(G) is obtained by local complementation of a graph G at a and it can be shown that the underlying principle is one-bit
vertex a: teleportation [37]. In this framework, deterministic quantum
computations can be performed up to local Pauli corrections
τa (G ) = Uaτ (G ) G . (4.25)
[36]. Both models have the same efficiency and are poly-
The transformation Uaτ (G ) is defined by: nomial-time equivalent to the circuit model [32], solving the
same class of problems.
Uaτ (G ) = − iσ xa − iσzNa . (4.26) There exist not only pure teleportation-based or one-way
models, but also a variety of schemes which combine prop-
Here, the operator erties of both. These hybrid models were mostly invented to
overcome the enormous resource requirements of the original
− iσ xa = exp [ −iπ 4 · σ xa ] (4.27) methods [43, 58–63].

transforms the qubit at vertex a and the operator 4.4. An application: blind quantum computing (BQC)

− iσzNa = exp ⎡⎣ −iπ 4 · σzNa ⎤⎦ (4.28) Recently, a new application of MBQC was invented [64]. It
was shown that quantum computers [1, 3–5], besides offering
transforms all neighboring vertices. substantial computational speedups, are also expected to
If two graph states can be related by a series of local preserve the privacy of a computation [64–69] as manifested
complementations, they are equivalent under local Clifford in the BQC protocol [64].
transformations. Interestingly, it was proven that two This security is a new aspect of quantum computers
graph states that are equivalent under general local unitary which enables a client to delegate a quantum computation to a
transformations are not necessarily equivalent under local server such that the userʼs data and the whole computation
Clifford operations [56]. remain perfectly private (see figure 9). The quantum server

11
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

Figure 8. Comparison of different measurement-based models of quantum computing.

performs calculations, but has no means to find out what it is All those properties make photons ideal carriers of
doing—it knows neither the input nor the output of the information and have led to a variety of photonic experiments
computation and cannot infer what is actually being calcu- ranging from quantum computing, quantum simulation, and
lated. Remarkably, the only quantum power that is required quantum communication to the foundations of quantum
from the client is the preparation of single qubits and their mechanics [108, 109].
transmission to the server.
The BQC protocol [64] is in detail explained in figure 10 5.1. Photonic qubits and quantum gates
and uses the concept of one-way quantum computing [30–
32, 72, 73]. There are different ways to encode information in photonic
Reference [70] shows the implementation of an opti- qubits, for example path or polarization [107, 108]. Here, we
mized version of the original protocol [64] using photonic focus on the polarization degree of freedom, where a photonic
qubit can be defined as:
qubits. Photons are ideally suited for BQC as they provide the
natural choice as quantum information carrier for the client 0 = H , (5.1)
and enable quantum computing for the server. Further, it was 1 = V , (5.2)
shown, that the concept of BQC allows testing if a quantum
computation was performed correctly [74], which has also where ∣ H 〉 denotes horizontal polarization and ∣ V 〉 denotes
been demonstrated experimentally [7]. vertical polarization.
A convenient way to treat polarization states is the Jones
formalism—a matrix formalism describing polarization by a
2D polarization vector J ⃗ [110]:
5. Optical quantum information processing ⎛ a exp (iϕ
J⃗ = ⎜
H H ) ⎞⎟, (5.3)
⎜ a exp iϕ
Experimental implementations of quantum computing have
⎝ V ( ) ⎟⎠
V
been realized in many systems including photons
[46, 51, 70, 75–79], ions [80–88], atoms [89–91], nuclear where aH and aV denote the amplitude of the wave vector in
magnetic resonance [92–95], superconducting systems [96– the horizontal and vertical direction, respectively, and ϕH and
103], and solid state systems [104–106]. Each system dis- ϕV denote the corresponding phases. Operations on states can
plays very particular advantages. Photonic qubits are espe- be represented by the Jones matrices, M:
cially well-suited for quantum information processing as they
J ′⃗ = MJ ⃗, (5.4)
show low decoherence and can be easily transmitted over
large distances. Furthermore, photonic states can be where J ′⃗ is the vector obtained when M acts on the state J ⃗ .
manipulated with very high precision and photonic systems This vector definition is consistent with the definition of
are among the fastest systems available for quantum infor- qubits given in section 3 and the Jones matrices can be seen as
mation processing [43, 107, 108]. being equivalent to single-qubit gates.

12
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

If the fast axis of the wave plate is oriented along at an

arbitrary angle θ with respect to the horizontal axis, the
transformation matrix can be determined by applying rotation
matrices R(θ):
T ′ (ϕ) = R (θ ) T (ϕ) R ( − θ )
⎛ cos2 (θ ) + e −iϕ sin2 (θ )
= ⎜⎜
( 1 − e ) cos (θ ) sin (θ ) ⎞⎟ (5.7)
−iϕ

⎟
(
⎝ 1−e
−iϕ
) cos (θ ) sin (θ ) e −iϕ cos2 (θ ) + sin2 (θ ) ⎠

with R (θ ) = (cos (θ ), −sin (θ ); sin (θ ), cos (θ )). Note that

the angle θ is a physical rotation in the laboratory, not to be
Figure 9. Given the challenges inherent in physically realizing a
quantum computer, it is conceivable that, in the future, quantum confused with a logical rotation on the Bloch sphere.
computing capabilities may be limited to a few specialized facilities Half-wave plates (HWPs) and quarter-wave plates
around the world. Similar to classical cloud computing, users might (QWPs) are special cases of phase retarders which are
then interact remotely with quantum computers and delegate their particularly useful in implementing photonic single-qubit
computations.
gates. A half-wave plate has a thickness such that the phase
5.1.1. Photonic single-qubit gates on polarization-encoded retardance is ϕ = π. It can be described by the matrix (overall
qubits. Experimentally, the polarization of light can be phases, which are important only for certain interference
manipulated with phase retarders or ‘wave plates’ [111]. experiments and not for the implementation of quantum logic
These uniaxial, birefringent crystals introduce a polarization- gates, have been omitted in all calculations presented here):
dependent phase shift. Combining multiple wave plates ⎛ cos (2θ ) sin (2θ ) ⎞
facilitates the realization of every possible single-qubit gate UHWP (θ ) = ⎜ ⎟, (5.8)
⎝ sin (2θ ) −cos (2θ ) ⎠
on polarization-encoded qubits.
When light travels through a wave plate, it experiences where, again, overall phases are omitted. For θ = 0 the HWP
different refractive indices for its ordinary (o) and implements a Z-gate, for θ = π/4 it is an X-gate, whereas for
extraordinary (e) components. The extraordinary polariza- θ = π/8 it represents a Hadamard gate. The HWP can thus be
tion lies parallel to the plane which is spanned by the used to convert linearly polarized light into other linear
optical axis and the k ⃗ vector of the pump beam, whereas the polarization states. This allows the manipulation of polariza-
ordinary polarization is perpendicular to that plane. Thus, tion-encoded quantum states, e.g. a HWP can turn the states
each component travels with a different velocity (see ∣ 0〉 and ∣ 1〉 into the states ∣ + 〉 and ∣−〉 and vice versa.
figure 11). A QWP implements a phase shift of ϕ = π/2 and can
After passing through the plate, the two perpendicular mathematically be described via:
polarizations have acquired a difference in phase that is given ⎛ 1 + i cos (2θ ) i sin (2θ ) ⎞
by: UQWP (θ ) = ⎜ ⎟. (5.9)
⎝ i sin (2θ ) 1 − i cos (2θ ) ⎠
2π
ϕ= d ne − no , (5.5) For θ equal to zero, a QWP adds a phase of (−i) to the
λ vertical component, whereas for θ = π/2, the phase shift is
where λ is the wavelength, d the thickness of the crystal, and equal to (+i). QWPs can create circularly polarized light
ne and no the refractive indices for the extraordinarily and from linearly polarized light and are able to turn, for
ordinarily polarized components. example, the states ∣ + 〉 and ∣−〉 into the states ∣ +i 〉 and ∣−i 〉
Wave plates are commonly produced from quartz or and vice versa. In contrast to HWPs, the orientation (i.e.
calcite, although wave plates made from other birefringent vertical or horizontal fast axis) of a QWP has an effect on
materials including magnesium fluoride, sapphire, and some the polarization, and this must be considered in
polymers are also available. The difference of the two experiments.
refractive indices ne and no can either be positive or negative, Any general unitary transformation U can be achieved by
depending on the material. The axis along which the phase using a combination of HWPs and QWPs. In many cases, a
velocity is fastest is called the fast axis, whereas the axis HWP is sandwiched between two QWPs
along which the phase velocity is slowest is called the U = UQWP ( θ3 ) UHWP ( θ2 ) UQWP ( θ1), (5.10)
slow axis.
If the fast axis of a wave plate is oriented horizontally, which enables arbitrary transformations of the polariza-
the phase shift can be described by the following matrix tion [112].
[111]:
5.2. Beam splitters and polarizing beam splitters (PBS)
⎛1 0 ⎞
T (ϕ) = ⎜ ⎟, (5.6)
⎝ 0 exp (−iϕ) ⎠ Although photonic quantum systems do not allow for direct
interactions, the implementation of photonic two-qubit gates
where the vertical component of the beam acquires a phase is still possible using concepts like the KLM approach. Beam
shift of − ϕ . splitters, PBSs, and the Hong–Ou–Mandel (HOM) effect

13
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

Figure 10. Scheme of BQC [64, 70]. A nearly-classical client with limited computational power can delegate a computation to a quantum
server with the full power of quantum computing such that the input, the output and the whole computation remain perfectly private. To this
end, the client prepares single-qubits in a state ∣ θ j 〉 = 1 2 (∣ 0〉 + e iθ j ∣ 1〉 ), where θj is chosen uniformly at random from the set {0, π/4,
…,7π/4 }. The qubits are then sent to the server who entangles them to a blind cluster state by applying CPhase gates. Although the cluster
state changes with the underlying qubits, it can be used for any computation. The actual computation is measurement-based and performed by
applying a pattern of consecutive adaptive single-qubit measurements. The client calculates measurement instructions δj which are sent to the
server. These depend on the measurement angle ϕj that the client wants to hide, the phase of the blind qubit θj, and a random bitflip π rj.
These classical measurement angles are set in such a way to compensate for the initial random rotation θj and any other Pauli byproducts
[51, 71] produced by previous measurements. The server now holds qubits and measurement instructions, but does neither know the state of
the qubit, ∣ θ j 〉, nor the measurement angle ϕj. The server then performs measurements in the basis ∣ ±δ j 〉 = 1 2 (∣ 0〉 ± e iδ j ∣ 1〉 ) on the
blind cluster state. Without knowledge about the state ∣ θ j 〉 or the hidden measurement angle ϕj, the measurement outcomes do not reveal any
information about the computation. The results are then sent back to the client who is the only one able to interpret them.

[113] play a crucial role in experimental implementations of If a1 and b1 are input modes of a beam splitter, the state
these concepts. of the output modes a2 and b2 can be calculated using the
following relations [52, 114, 115]:
5.2.1. Beam splitters. A beam splitter is a semi-reflective
mirror which splits an incident beam into two parts: a
a 2† = cos θ · a1† + ie−iϕ sin θ · b1† , (5.11)
transmitted part and a reflected part (see figure 12(a)).

14
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

Figure 11. Working principle of a wave plate. This example shows a beam passing through a half-wave plate with an optical axis long the
vertical direction. When traveling through the wave plate, the extraordinary and the ordinary polarization components experience different
refractive indicices, ne and no, due to the birefringence of the material. Thus, the two polarization components have two different phase
velocities; in this example the optical axis defines the slow axis. After passing the wave plate, the two different polarization components have
acquired a phase shift of π, which effectively leads to a polarization rotation of 90°.

b2† = ieiϕ sin θ · a1† + cos θ · b1† , (5.12) configuration of QWPs and HWPs (see figure 12(c)) are given
in table 1.
where a† and b† are creation operators representing a photon
in mode a and b, respectively, and the angle θ specifies the 5.2.3. The HOM effect. The HOM effect is a two-photon
transmittance of the beam splitter. This can be expressed more interference effect that occurs when two indistinguishable
conveniently by the transmission coefficient T = cos2 (θ ) and photons enter a beam splitter from two input ports [113] (see
the reflection coefficient R = sin2 (θ ), which obey the relation figure 12(d)). The effect is purely quantum and does not occur
T + R = 1. The phase shift ϕ between the reflected and the in classical optics. Mathematically, two photons entering a
transmitted modes ensures the unitarity of the beam splitter symmetric beam splitter can be described by their respective
operation [114] and is defined by the physical properties of creation operators a1†, b1† acting on the vacuum state ∣ 0〉∣ 0〉
the beam splitter; i.e. the coating of the mirror. Interestingly, (for the description of the HOM effect, we use the photon
if ϕ1 (ϕ2) is the phase shift between the transmitted and the number basis, where ∣ n〉 represents a state containing n
reflected mode for a photon entering from mode a1 (b1), photons). Applying the relations that were introduced in the
unitarity requires that ϕ1 + ϕ2 be equal to π [114]. previous section, we obtain:
A symmetric beam splitter which splits the light equally
into the two output modes (θ = π/4), and which acts (
a1† b1† 0 0 ≃ a 2† − ib2† )( a †
2 + ib2† 0 0 ) (5.15)
symmetrically on the two input ports (ϕ = 0) is defined by:

a 2† =
1 †
a1 + i
1 †
b1 , (5.13)
≃ (( a 2† ) + (b ) ) 0
2 † 2
2 0 (5.16)
2 2
1 † 1 †
≃ 2( a2 0 b2 + 0 a2 2 b2 ), (5.17)
b2† =i a1 + b1 . (5.14)
2 2 where normalization factors were omitted. The two photons
exit the BS either both in the output mode a2 or both in the
The beam splitters used in experiments are manufactured to
output mode b2; they never split up and exit in different
show a behavior close to that of an ideal symmetric beam
output modes.
splitter, but given the difficulty of constructing a perfect beam
It is important to note that the HOM effect is not the
splitter it may be necessary to add additional phases to
interference of two photons (‘that meet at a beam splitter’),
equation (5.13) to achieve a full description of an actual
but the interference of the respective two-photon amplitudes
experimental situation [52, 116].
occurring in the detectors. Interestingly it was shown that the
photons do not even have to arrive at the beam splitter at the
5.2.2. Polarizing beam splitters and measurements. A PBS same time in order to interfere [117].
splits a beam depending on its polarizations, usually The HOM effect also occurs in PBSs if the information
separating an input beam into two modes with orthogonal about the polarization (and thus the which-path information)
polarization. Light that is vertically polarized is reflected, is extinguished. In experiments this can achieved with PBSs
whereas horizontally polarized light is transmitted through a and measurements in the basis {∣ + 〉, ∣−〉 } [43].
PBS (see figure 12(b)).
PBSs can be used for the analysis of a polarization state.
5.3. Multi-qubit gates
Combined with HWPs and QWPs, they facilitate measure-
ments in each possible direction on the Bloch sphere. One of the main advantages of photonic systems is their very
In experiments, measurements of σx , σy , and σz are low decoherence. Even if photonic states are transmitted over
particularly interesting as may be used to reconstruct density large distances [118–120], the quantum states remain mostly
matrices [17] which contain the full information about the unaffected. On the other hand, the low decoherence rate
underlying quantum state. The relevant settings for measure- seems, at first sight, to prevent the implementation of multi-
ments in these bases using polarization-encoded qubits and a qubit gates since the photons do not interact with each other.

15
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

Figure 12. The figure shows the working principle and applications of (polarizing) beam splitters. (a) A beam splitter (BS) splits incident
photons into two output modes depending on the splitting ratio. (b) A polarization beam splitter (PBS) transmits horizontally polarized light
and reflects vertically polarized light. (c) A PBS together with half-wave and quarter-wave plates can be used for the analysis of arbitrary
polarizations. (d) Demonstration of the HOM effect: if two indistinguishable photons enter a beam splitter, they will both exit one or the other
port and will never split up into two different ports.

Table 1. This table shows the measurement settings for the

There are mainly two different approaches to overcome this
measurements of σx , σy , and σz . A half-wave and a quarter-wave plate
problem in photonic quantum information processing: con-
are combined with a PBS as shown in figure 12(c).
cepts along the line of KLM [8] using ancilla photons and
postselection to provide measurement-induced nonlinearities, Basis QWP setting HWP setting
and concepts using optical nonlinearities [121]. In this σx π/4 π/8
section 1 will briefly introduce these approaches and outline σy π/4 0
the major advantages as well as disadvantages. σz 0 0

5.3.1. Photonic CNOT and CPhase gates. To illustrate the

working principle of most linear-optics two-qubit gates, I will However, applying this type of gate to photons generated
from a down-conversion source (a probabilistic source for the
briefly review two examples.
generation of single photons, see section 5.4 for details) is still
The CPhase gate, which is shown in figure 13(a), consists
challenging since higher-order emissions can lead to incorrect
of a polarization-dependent beam splitter (PDBS) which has a
gate operations. If a double-pair emission enters the gate
different transmission coefficient for horizontally polarized
input, it can split into all four output modes even if no ancilla
light (T = 1) as for vertically polarized light (T = 1/3) [122]. If
photons are present. These events have the same generation
two vertically-polarized photons are reflected at this PDBS,
probability as the correct events, but lead to an incorrect gate
they acquire a phase shift of π. Two successive PDBSs with operation. This issue can be solved using heralded photon
the opposite splitting ratios then equalize the output pairs where a signal announces the presence of entangled
amplitudes. The gate operation has been successful if one photons in the right modes. The generation of heralded
photon exists in each of the output modes. entangled photon pairs can also be realized by using only
Whereas the experimental implementation of this gate is linear optics [124, 125].
relatively easy, its scalability is very limited. Since a Both gates presented here resemble the circuit model, but
destructive photon measurement is necessary to verify the a closer look reveals that they are actually based on
correct operation of the gate, the state of the photons is measurements—either by postselecting the output states or
destroyed which makes the realization of a subsequent gate by measurements of ancilla photons. Thus, these examples
impossible. demonstrate the necessity of measurements in photonic
In contrast, the CNOT gate shown in figure 13(b) is quantum computing and indicate that a pure circuit model
scalable, but requires an entangled ancilla photon pair as cannot be realized using only linear optics.
resource [58, 123]. If two photons are registered in the ancilla
modes, the gate has been successful without the need for a
verification of the output state. Thus, it is possible to use these 5.3.2. Measurement-based photonic quantum computing. In
gates in succession. MBQC, quantum information is processed by single-qubit

16
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

Figure 13. Sketch of two photonic entangling gates. The CPhase gate (a) uses polarization-dependent beam splitters and requires a
measurement of the output modes to verify the correct operation of the gate. The CNOT gate (b) requires an entangled ancilla photon-pair and
a measurement of two ancilla modes to herald that the gate has worked correctly. Thus, the CPhase gate is limited in scalability since a
measurement destroys the quantum state and does not allow for a subsequent gate operation. Figure adapted from [122] and [58].

measurements on cluster states [30]. For photonic systems, and thus significantly improve the scalability of photonic
the generation of cluster states requires entangling quantum computing.
operations and thus relies on postselection techniques.
This means that measurements are necessary for the 5.4. Generation of entangled photons
creation of cluster states, but the same measurements
also implement the computation in MBQC. In other words, The workhorse of almost all photonic quantum computing
in photonic systems, the measurements which are experiments is spontaneous parametric down-conversion
necessary for the processing of quantum information (SPDC)—a process where a pump photon is converted into
arise naturally from the creation of photonic cluster two daughter photons in a nonlinear crystal. The selection of
states. Thus, despite the high propagation speed of photons with a particular frequency and spatial emission can
photons and the lack of multi-photon interaction, facilitate the availability of polarization-entangled photon
photonic systems are well-suited for MBQC pairs. In the following, I will describe the process in more
[51, 70, 76, 126]. Section 5.5 shows an example of how detail and explain how multi-photon states can be created.
cluster states can be implemented experimentally.
5.4.1. SPDC. SPDC occurs when laser light interacts with a
nonlinear crystal such as β-barium borate (BBO) [130]. When
an electromagnetic field interacts with a nonlinear medium,
5.3.3. Optical nonlinearities for entangling gates. Another
the dielectric polarization P generated in the medium shows a
approach to realize photonic quantum computing is to use
nonlinear dependency on the electric field:
optical nonlinearities for the implementation of (nearly)
deterministic two-qubit gates.
Pi = χi(1) (2) (3)
,j E j + χi,j,k E j E k + χi,j,k ,l E j E k E l + ⋯ , (5.18)
For example, Kerr-nonlinearities can induce photon–
photon interactions [52] and enable a phase shift in one mode
where χm is the susceptibility of order m, Ei denotes the
depending on the number of photons in another mode.
electric field and double indices indicate a sum [131]. The
Another type of nonlinear entangling gate is based on the
first term (χ1 ≈ 1) describes linear effects such as diffraction,
Zeno effect [127, 128] where failure events in two-qubit gate
and refraction and the third term (χ3 ≈ 10−17) is very small
operations are suppressed by continuous two-photon
and describes four-wave mixing processes.
absorptions.
Of interest here is the second term (χ2 ≈ 10−10), which leads
One of the main challenges in such experiments is that
to three-wave mixing processes like SPDC. If two waves,
the available materials provide only small nonlinearities and,
E1 cos (ω1 t ) and E2 cos (ω 2 t ) interact with a nonlinear
so far, only lead to relatively small phase shifts [129].
medium, this second term can be rewritten in the following form:
Nevertheless, the application of theses schemes may sig-
nificantly reduce the number of required ancillary photons Pi = χ (2) E1 cos ( ω1t ) E2 cos ( ω2 t ) (5.19)

17
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

= χ (2) E1E2 cos ( ( ω1 + ω2 ) t ) cos ( ( ω1 − ω2 ) t ), (5.20) where under annihilation of two photons (a1 and a2), a photon
is created (a p†).
showing that sum-frequency (ω1 + ω2) and difference-frequency In the case of type-II SPDC, the output state can be
(ω1 − ω2) waves are generated. written as follows [133]:
SPDC is the reverse configuration: a pump field with a
wavelength ω = ω1 + ω2 creates two new fields with
frequencies ω1 and ω2, called signal and idler [111, 132].
( (
ψ = Z · exp −iα a H† bV† + aV† b H† )) 0 , (5.25)

Two different types of SPDC are distinguished: both created

where Z is a normalization constant and α is a parameter
photons can either have the same polarization which is
depending on χ2 and on the pump power. The creation
orthogonal to the polarization of the pump laser (type-I), or
operators a H† (bV†) describe the generation of a horizontally
both photons have orthogonal polarizations (type-II); in the
(vertically) polarized photon in mode a (b) and act on the
following, we will focus on type-II SPDC.
vacuum state ∣ 0〉. Expanding the exponential function leads to:
The down-converted photons show correlations in
frequency as well as momentum: ⎡
⎢
ωp = ωe + ωo , (5.21) ψ = Z · ⎢ 1 − iα a H† bV† + aV†b H†
( )
⎢ 
⎢⎣
k p⃗ = k e⃗ + k o⃗ . (5.22) (1)

α2 † † 2
Only certain propagation directions k ⃗ are possible for the − (

2
)
a H bV + aV† b H†
photons due to the phase matching conditions in the nonlinear
crystal. For the degenerate case, where ωe = ωo, the emission (2)

direction of the created photons is along the surface of cones ⎤

⎥
(see figure 14), where one cone has extraordinary polarization α3 † †
a H bV + aV† b H† + ⋯⎥ 0 .
3
and the other ordinary. For our experiments, this corresponds + i (
3
) ⎥
(5.26)
to vertically and to horizontally polarized light, respectively. ⎥
(3) ⎦
For certain incidence angles of the pump beam, these cones
intersect (figure 14). Photons emitted into the directions The first term (1) corresponds to the generation of a two-
of the lines of intersection can therefore not be assigned to photon Bell state with a probability ≃Z2α2. The higher orders
one of the cones. If one photon is emitted into the direction of represent multi-photon emissions, where a four-photon emis-
one of the intersections, the other one must be in the other sion (2) is generated with a probability ≃Z2α4 and a six-photon
intersection line due to momentum conservation. These photons emission (3) with probability ≃Z2α6. The probabilities to create
are entangled in polarization (figure 14) with a state given by: multi-photon states are low compared to the two-photon case
1 and depend polynomially on the pump power.
ψ =
2
(H V + eiϕ V H . ) (5.23) Continuous-wave (cw) lasers with typical powers of
about 50 mW lead primarily to the emission of two-photon
The ordinarily and the extraordinarily polarized photons states, since the power is too low to create substantial multi-
experience different refractive indices, no and ne, in the photon emissions. The high-pump powers necessary to obtain
crystal. Thus, firstly, the propagation velocities are different multi-photon events can be achieved with pulsed lasers
for both components (longitudinal walk-off effect) and having sufficiently high peak powers for the generation of
secondly, both photons experience a spatial displacement higher-order photon states (see next section).
(transversal walk-off effect). These two effects may lead to a However, increasing the pump power also affects the
distinguishability of the photons and, as a consequence, to the quality of the states, because the noise terms emerging from
annihilation of the entanglement. For the generation of the next-order emissions are also increased. The signal-to-
properly entangled pairs, it is therefore very important to noise ratio depends on the parameter α and in this way on the
take into account both effects (for details, see figure 15). pump power:

signal 1
5.4.2. Quantum-mechanical treatment. In order to fully = 2 ⟶ 0. (5.27)
characterize the down-conversion process, a quantum noise α α→∞
mechanical treatment is necessary. The following interaction
This noise is intrinsic to all SPDC sources: the higher the pump
Hamiltonian describes the process in terms of the creation and
power, the higher the effect of the noise. In experiments it is
annihilation operators, a† and a:
therefore necessary to find the right balance between a pump
H = γa1† a 2† a p + γ *a1a 2 a p†, (5.24) power that is high enough to create the desired emission and one
which is, on the other hand, low enough to minimize the noise.
where the coupling constant γ depends on the nonlinearity χ2.
The first term expresses the down-conversion process, where 5.4.3. Pulsed SPDC. Photonic quantum information
one pump photon (ap) creates two down-converted photons processing requires the coherent generation of multi-photon
(a1† and a 2†). The second term describes the opposite process states. These can be created by pumping a SPDC source with

18
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

generalization of those in which the underlying qubits exhibit

arbitrary phases θj. They are created by entangling qubits in
1
states ∣ θ j 〉 = (∣ 0〉 + e iθ j ∣ 1〉 ), where θj is chosen from {0,
2
π/4, …, 7π/4}.
For the experiment which is presented here [70], the
cluster state consists of four blind qubits, where the phases of
∣ θ1 〉 and ∣ θ 4 〉 are chosen to be zero:

1
θ1 θ2 θ3 θ4 = ( 0 + 1 ) ( 0 + eiθ 2 1 )
4
× ( 0 + eiθ3 1 ) ( 0 + 1 ). (5.28)

Applying a CPhase gate between qubits 1–2, 2–3 and 3–4

Figure 14. View of a parametric down-conversion process where the creates a linear cluster state:
down-converted photon pairs are emitted along the surfaces of
cones. For the degenerate case, where ωe = ωo, the opening angles of 1
both cones are equal and the setup can be aligned such that the cones
ˆ
Φθ =
2
( + 00 + + eiθ3 + 01 −
intersect as depicted. Photons emitted into the direction of the
intersection lines are polarization-entangled, since it cannot be + eiθ 2 − 10 +
distinguished to which cone they belong. Figure adapted from [130]. − e i( θ 2+ θ3 ) − 11 − ), (5.29)

n π n π
a pulsed laser, where the pulse length is on the order of where θˆ = (n 2 , n 3 ) and (θ2, θ3 ) = ( 42 , 43 ).
hundreds of femtoseconds. In the following, it will be shown how the state of
Pulsed lasers reduce the uncertainty of the emission time equation (5.29) can be generated in an experiment. However,
for a given down-converted pair [134] and fulfill a necessary it should be stressed that this implementation is just one
condition for coherent higher-order emissions: the variance in example how this state can be generated and other imple-
emission time, which is determined by the duration of the mentations are also possible.
pump pulse, must be smaller than the coherence time of the The experimental setup for the generation of a blind
down-converted photons [134–137]. cluster state in shown in figure 16. It consists of a SPDC
On the other hand, pulsed lasers have the disadvantage
source, which is pumped in two directions, called the forward
that the properties of down-converted photon pairs are
and the backward direction, respectively. The blind cluster
different from those generated by a cw pump. A pulsed
state is composed of four terms, which correspond to different
pump contains a broad range of frequencies; the shorter the
four-photon emissions. These are achieved by pumping the
pulse length, the broader the spectral bandwidth [138]. This
leads to down-converted photons which are no longer exactly BBO crystal with a pulsed laser system at a high laser power
anticorrelated in frequency since they do not required to fulfill (200 fs pulses at a repetition rate of 76 MHz at 394.5 nm). A
a constant frequency sum (as in equation (5.21)). Further- four-photon emission can be obtained experimentally either
more, the spectra of the ordinary and extraordinary photons by an emission of two entangled pairs, one in the forward and
are no longer identical [134]. In the temporal domain, a one in the backward mode, or by double-pair emissions into
pulsed pump leads to a reduced coherence time of the down- respective modes. As is shown below, the generation of blind
converted photons as their bandwidth increases. cluster states exploits coherent superpositions of these dif-
These effects make the down-converted photons distin- ferent four-pair contributions and utilizes the properties of the
guishable and decrease the visibility in the two-photon PBSs as well as post-selection to obtain the appropriate state.
interference [137]. The use of narrowband filters and a In the following, the equations are written in terms of
spectral postselection can recover the indistinguishability and state vectors for the sake of clarity. However, the derivation
improve the two-photon interference visibilities at the cost of of these equations should be performed in terms of creation
reduced count rates. operators to obtain the correct results. Here, I will neglect
mathematical rigor for the benefit of an intuitive under-
5.5. Example: experimental generation of blind cluster states standing and also omit the normalization factors.
In order to create a blind cluster state, the experiment is
In section 4.4, we introduced the concept of BQC. In order to
aligned such that pairs in a state
show how the experimental concepts presented in this section
∣ϕθ−3〉ab = (∣HH 〉ab − e iθ3 ∣VV 〉ab ) 2 are emitted in the for-
are used in actual experiments, we will now outline how blind
cluster states can be generated experimentally. ward direction (modes a, b), and pairs in a state
Blind quantum computing starts with the generation of ∣ϕθ+2 〉cd = (∣HH 〉cd + e iθ2 ∣VV 〉cd ) 2 are emitted in the
blind qubits that are entangled to blind cluster states [64, 70]. backward direction (modes c, d).
Standard cluster states have already been generated in a range The emission of only one entangled pair in the forward
of experiments [138, 139]. Blind cluster states are a direction (a, b) and only one pair in the backward direction (c,

19
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

Figure 15. Compensation of walk-off effects. (a) Longitudinal walk-off effect: the arrival time of the photons can reveal information about
their polarizations and destroy the entanglement. The time difference after which both photons have passed the crystal depends crucially on
the point in the crystal where the photons are created. If the photon pair is created at the beginning of the crystal, the ordinarily polarized
photon passes the crystal faster (top example of (a)) and arrives earlier. If the photons are created at the end of the crystal, they arrive
simultaneously (bottom example of (a)). The longitudinal walk off needs to be compensated if the time difference δt = (no − ne)d/c after the
two photons have passed through the crystal is larger than the coherence time tc = 2 ln 2 λ 2 (πΔλc ) of the photons. Compensation can be
accomplished by rotating the polarization of both photons by 90◦ with a HWP, which exchanges the position of the ordinary and
extraordinary photons. If both photons subsequently pass a compensation crystal with a thickness that is half that of the first crystal, the
arrival time no longer contains information about the polarization. (b) Transversal walk-off effect. The ordinary and extraordinary photons
have different propagation directions in the crystal due to polarization-dependent refractive indices of the crystal. Together with the
extraordinary polarization of the pump, this leads to a broadening of the ordinary beam which needs to be counteracted if this broadening is
larger than the beam waist. Again, by interchanging both polarizations and letting the two beams pass through a compensation crystal, the
effect can be compensated.

d) results in two different four-photon terms: All three states given in equations (5.31), (5.32), and
(5.34) build a coherent superposition with the phases defined
ϕθ−3 ϕθ+2 ≈ HHHH abcd + eiθ 2 HHVV abcd by the relative phase Δ between the forward and the backward
ab cd
emission:
− eiθ3 VVHH abcd
− e i( θ3+ θ 2 ) VVVV abcd . (5.30) ˆ
Φ θ L (Δ) = eiΔ HHHH 1234 + eiθ3 HHVV 1234

The photons then pass the PBSs and only the terms + e2iΔeiθ 2 VVHH 1234
leading to a fourfold coincidence in modes 1–4 are post-
selected: − eiΔe i( θ3+ θ 2 ) VVVV 1234 . (5.36)
PBS and postselection
ϕθ−3 ϕθ+2 ⟶ The phase Δ is set equal to multiples of 2π and the final
ab cd ˆ
output state ∣ Φ θ L 〉 obtained in the laboratory is given by:
HHHH 1234 − e i( θ3+ θ 2 ) VVVV 1234 . (5.31)
ˆ
In the same way, the emission of two photon pairs in the Φ θ L = HHHH 1234 + eiθ3 HHVV 1234
forward modes (a, b) can be calculated:
+ eiθ 2 VVHH 1234 − e i( θ3+ θ 2 ) VVVV 1234 . (5.37)
ϕθ−3 ϕθ−3 ≈ HH a HH b −e iθ 3
HV a
ab ab ˆ
The blind cluster state ∣ Φ θ L 〉 that is produced in the
i( 2θ 3 )
× HV b + e VV a VV b , (5.32) experiment is equivalent under local unitary transformations
PBS and postselection to the blind cluster state ∣ Φ θ̂ 〉. Applying Hadamard gates on
⟶ − eiθ3 HHVV 1234 (5.33) qubits 1 and 4 and using the definition ∣ H 〉 = ∣ 0〉 and
∣ V 〉 = ∣ 1〉 finally leads to the blind cluster state:
where ∣HH 〉a denotes two horizontally polarized photons in
mode a, etc. In the backward modes (c, d), the double-pair ˆ ˆ
emission leads to a state: Φ θ = (H ⨂ I ⨂ I ⨂ H ) Φ θ L . (5.38)

ϕθ+2 ϕθ+2 ≈ HH c HH d + eiθ 2 HV c These Hadamard gates can be implemented in the experiment
cd cd
by two additional HWPs; alternatively, they may be absorbed
× HV d + e i( 2θ 2 ) VV c VV d (5.34) in the measurement basis which leads to a simple reinter-
PBS and postselection pretation of the data. Note that after the PBSs two quarter-
⟶ eiθ 2 VVHH 1234 . (5.35)
wave plates were inserted in modes 3 and 4 to compensate for
In the experiment, the phase of the term birefringence effects and additional phases.
− e iθ3 ∣H 〉1∣H 〉2 ∣V 〉3 ∣V 〉4 is then shifted by π by applying an By changing the phases of the entangled pairs, the values
additional rotation using a HWP, which has the desired of θ2 and θ3 in the blind cluster state can be manipulated
effect [76]. arbitrarily, for example using a combination of additional

20
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

methods for the creation, interaction and detection of single

photons must be developed.
Firstly, the standard process of creating entangled pho-
tons, SPDC, works probabilistically and with low efficiency.
The heralded generation of entangled photon pairs can be
realized [124, 125], but the actual rates are still low.
Additionally, the quality of multi-photon states generated
in a SPDC process are intrinsically limited due to noise
caused by higher-order emissions. Therefore, the develop-
ment of a high-efficiency push-button source which produces
single-photons or multi-photons states on-demand is crucial.
Promising candidates for this task are semiconductor quantum
dots [140–146], atoms or ions [147–150], superconducting
qubits [151], and nitrogen-vacancy centers [152–155].
Secondly, the development of efficient photonic two-
qubit gates is of great importance. Although it has been
shown that quantum computing is possible with only linear
optics and photon detection, in practice these schemes
become inefficient due to an enormous amount of required
ancilla photons [8]. In order to overcome these challenges, the
advancement and utilization of optical nonlinearities [52, 129]
on the one hand, and the development of schemes which
enable photon–photon interactions on the other hand
[127, 128], are crucial tasks. Furthermore, the future of
Figure 16. The experimental setup to produce and measure blind photonic quantum information processing might lie in inte-
cluster states. The various blind cluster states are created by grated optics which enable a higher stability and better
adjusting the settings of the half-wave plates, quarter-wave plates implementation of quantum circuits while at the same time
and BBO crystals located along the path of the state emitted into the reducing photon losses [156–158]. So far, these setups use
forward (θ3) and backward (θ2) modes.
path encoded qubits with only a few modes and are thus still
limited in their complexity. In the future, also using the
QWPs and HWPs in the forward mode [112]: polarization degrees of freedom in integrated optics might
enable the implementation of much more complex circuits.
ϕθ−3 = ⎡⎣ UQWP (−π 4) UHWP ( θ 3 8) UQWP (−π 4) Efforts are currently underway to realize integrated sources
where the photons are directly generated in a waveguide
⨂ UQWP (π 4) UHWP ( −θ 3 8) UQWP (π 4) ⎤⎦ ϕ− , (5.39) [159–161]. The ultimate aim for the future is to integrate
photon generation, processing and detection on a single
with the Bell state ∣ ϕ−〉 = (∣ HH 〉 − ∣ VV 〉 ) 2 . photonic chip.
Equation (5.39) shows that by rotating both HWPs, any phase Thirdly, current experiments are limited by the low
θ3 can be obtained. For practical reasons, the phase of the detection efficiencies of avalanche photodiodes which are
backward pair is adapted by tilting one of the compensation widely-used in photonic quantum computing experiments.
crystals. The case of standard cluster states can be obtained by Much higher detection efficiencies can be obtained by
choosing θ2 = θ3 = 0. employing superconducting transition-edge detectors or
This example of a photonic experiment shows how the superconducting nano-wire detectors [162–167]. While poor
basic concepts of photonic quantum information processing time resolution of the former impedes their application in
can be used to generate cluster states. These cluster states pulsed multi-photon experiments, the latter are ideally suited
allow performing various blind delegated computations, for this task.
including one- and two-qubit gates and the Deutsch and Furthermore, it will be interesting to see in general in the
Grover quantum algorithms [70]. Further, it was shown, that future which physical system will prevail [168]. One forward-
the concept of BQC allows testing if a quantum computation looking concept might be the development of hybrid systems
was performed correctly [74], which was also demonstrated combining the advantages of photons, in particular the low
experimentally [7]. decoherence, with the advantages of other quantum systems
which enable multi-qubit interactions [169, 170]. For these
hybrid systems, the development of interfaces between the
6. Conclusion and outlook different physical systems, for example between optical and
microwave photons [171–174], or the mapping of photonic
In order to develop scalable LOQC experiments, several steps states to atomic states is a crucial task [91, 175–184].
will have to be taken in the future. Different technical chal- Besides these technical advances, the future will also
lenges need to be overcome and, in short, more efficient show which computational model is best-suited for

21
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

experimental quantum computing. Also on the theoretical [18] Wei T, Nemoto K, Goldbart P M, Kwiat P G, Munro W J and
side approaches combining advantages of different models Verstraete F 2003 Maximal entanglement versus entropy for
might be beneficial [61, 62]. Breaking through the boundaries mixed quantum states Phys. Rev. A 67 022110
[19] Williams C P 2010 Explorations in Quantum Computing
of different fields, e.g. physics and computer science, might (Berlin: Springer)
inspire future research. [20] Uhlmann A 1976 The ‘transition probability’ in the state
space of a*-algebra Rep. Math. Phys. 9 273–9
[21] Hill S and Wootters W K 1997 Entanglement of a pair of
quantum bits Phys. Rev. Lett. 78 5022–5
Acknowledgments [22] Wootters W K 1998 Entanglement of formation of an
arbitrary state of two qubits Phys. Rev. Lett. 80 2245–8
The author is grateful to Philip Walther and Anton Zeilinger [23] Coffman V, Kundu J and Wootters W K 2000 Distributed
entanglement Phys. Rev. A 61 052306
for discussions; and to Wolfgang Duer, Marissa Giustina, Tim [24] Deutsch D 1989 Quantum computational networks Proc. R.
Langen, and Michael Zwerger for critical reading of the Soc. A 425 73–90
manuscript. This work was supported by the Marie Curie [25] Feynman R P 1986 Quantum mechanical computers Found.
Actions within the Seventh Framework Programme for Phys. 16 507–31
Research of the European Commission, under the Initial [26] Barenco A, Bennett C H, Cleve R, DiVincenzo D P,
Margolus N, Shor P, Sleator T, Smolin J A and Weinfurter H
Training Network PICQUE, Grant No. 608062. 1995 Elementary gates for quantum computation Phys. Rev.
A 52 3457–67
[27] Kitaev A Y 1997 Quantum computations: algorithms and
error correction Russ. Math. Surv. 52 1191–249
References [28] Shi Y 2002 Both toffoli and controlled-not need little help to
do universal quantum computation arXiv:quant-ph/0205115
[1] Feynman R 1982 Simulating physics with computers Int. J. [29] Aharonov D 2003 A simple proof that toffoli and hadamard
Theor. Phys. 21 467–88 are quantum universal arXiv:quant-ph/0301040
[2] Deutsch D 1985 Quantum theory, the church-turing principle [30] Briegel H-J, Browne D E, Dür W, Raussendorf R and
and the universal quantum computer Proc. R. Soc. A 400 Van den Nest M 2009 Measurement-based quantum
97–117 computation Nat. Phys. 5 19–26
[3] Deutsch D and Jozsa R 1992 Rapid solution of problems by [31] Raussendorf R and Briegel H J 2001 A one-way quantum
quantum computation Proc. R. Soc. A 439 553–8 computer Phys. Rev. Lett. 86 5188–91
[4] Shor P W 1997 Polynomial-time algorithms for prime [32] Raussendorf R, Browne D E and Briegel H J 2003
factorization and discrete logarithms on a quantum computer Measurement-based quantum computation with cluster
SIAM J. Comput. 26 1484–509 states Phys. Rev. A 68 022312
[5] Grover L K 1996 A fast quantum mechanical algorithm for [33] Bennett C H, Brassard G, Crépeau C, Jozsa R, Peres A and
database search Proc. 28th Annual ACM Symp. on the Wootters W K 1993 Teleporting an unknown quantum state
Theory of Computing pp 212–9 via dual classical and Einstein–Podolsky–Rosen channels
[6] Harrow A W, Hassidim A and Lloyd S 2009 Quantum Phys. Rev. Lett. 70 1895–9
algorithm for linear systems of equations Phys. Rev. Lett. [34] Verstraete F and Cirac J I 2004 Valence-bond states for
103 150502 quantum computation Phys. Rev. A 70 060302
[7] Barz S, Kassal I, Ringbauer M, Lipp Y O, Dakić B, [35] Leung D W 2001 Two-qubit projective measurements are
Aspuru-Guzik A and Walther P 2014 A two-qubit photonic universal for quantum computation arXiv:quant-ph/0111122
quantum processor and its application to solving systems of [36] Aliferis P and Leung D W 2004 Computation by
linear equations Sci. Rep. 4 6115 measurements: a unifying picture Phys. Rev. A 70 062314
[8] Knill E, Laflamme R and Milburn G J 2001 A scheme for [37] Childs A M, Leung D W and Nielsen M A 2005 Unified
efficient quantum computation with linear optics Nature 409 derivations of measurement-based schemes for quantum
46–52 computation Phys. Rev. A 71 032318
[9] Nielsen M A and Chuang I L 2000 Quantum Computation [38] Jorrand P and Perdrix S 2005 Unifying quantum computation
and Quantum Information (Cambridge: Cambridge with projective measurements only and one-way quantum
University Press) computation Proc. SPIE 5833 44–51
[10] Sakurai J J 2003 Modern Quantum Mechanics (Reading, MA: [39] van den Nest M and Briegel H J 2008 Measurement-based
Addison-Wesley) quantum computation and undecidable logic Found. Phys.
[11] Fox M 2006 Quantum Optics. An Introduction (Cambridge: 38 448–57
Cambridge University Press) [40] Bouwmeester D, Pan J-W, Mattle K, Eibl M,
[12] Horodecki R, Horodecki P, Horodecki M and Horodecki K Weinfurter H and Zeilinger A 1997 Experimental quantum
2009 Quantum entanglement Rev. Mod. Phys. 81 865–942 teleportation Nature 390 575–9
[13] Einstein A, Podolski B and Rosen N 1935 Can quantum- [41] Gottesman D and Chuang I L 1999 Demonstrating the
mechanical description of physical reality be considered viability of universal quantum computation using
complete? Phys. Rev. 47 777–80 teleportation and single-qubit operations Nature 402
[14] Bell J 1964 On the Einstein–Podolsky–Rosen paradox 390–3
Physics 1 195–200 [42] Myers C R and Laflamme R 2005 Linear optics quantum
[15] Clauser J F, Horne M A, Shimony A and Holt R A 1969 computation: an overview arXiv:quant-ph/0512104
Proposed experiment to test local hidden-variable theories [43] Kok P, Munro W J, Nemoto K, Ralph T C, Dowling J P and
Phys. Rev. Lett. 23 880–4 Milburn G J 2007 Linear optical quantum computing with
[16] von Neumann J 1932 Mathematische Grundlagen der photonic qubits Rev. Mod. Phys. 79 135–74
Quantenmechanik vol 42 (Berlin: Springer) [44] Kim Y-H, Kulik S P and Shih Y 2001 Quantum teleportation
[17] James D F V, Kwiat P G, Munro W J and White A G 2001 of a polarization state with a complete bell state
Measurement of qubits Phys. Rev. A 64 52312 measurement Phys. Rev. Lett. 86 1370–3

22
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

[45] Lee S-W and Jeong H 2013 Bell-state measurement and [71] Danos V and Kashefi E 2006 Determinism in the one-way
quantum teleportation using linear optics: two-photon pairs, model Phys. Rev. A 74 052310
entangled coherent states, and hybrid entanglement [72] Danos V, Kashefi E and Panangaden P 2007 The
arXiv:1304.1214 measurement calculus J. ACM 54 8
[46] O’Brien J L, Pryde G J, White A G, Ralph T C and [73] Gross D, Eisert J, Schuch N and Perez-Garcia D 2007
Branning D 2003 Demonstration of an all-optical quantum Measurement-based quantum computation beyond the one-
controlled-not gate Nature 426 264–7 way model Phys. Rev. A 76 052315
[47] Popescu S 2007 Knill–Laflamme–Milburn linear optics [74] Fitzsimons J F and Kashefi E 2012 Unconditionally verifiable
quantum computation as a measurement-based computation blind computation arXiv:1203.5217
Phys. Rev. Lett. 99 250501 [75] Walther P, Resch K J, Rudolph T, Schenck E, Weinfurter H,
[48] Hein M, Eisert J and Briegel H J 2004 Multiparty Vedral V, Aspelmeyer M and Zeilinger A 2005
entanglement in graph states Phys. Rev. A 69 062311 Experimental one-way quantum computing Nature 434
[49] Hein M, Dür W, Eisert J, Raussendorf R, Nest M and 169–76
Briegel H J 2006 Entanglement in graph states and its [76] Kiesel N, Schmid C, Weber U, Tóth G, Gühne O, Ursin R and
applications arXiv:quant-ph/0602096 Weinfurter H 2005 Experimental analysis of a four-qubit
[50] Gottesman D 1997 Stabilizer codes and quantum error photon cluster state Phys. Rev. Lett. 95 210502
correction PhD Thesis Caltech [77] Lu C-Y, Browne D E, Yang T and Pan J-W 2007
[51] Prevedel R, Walther P, Tiefenbacher F, Böhi P, Kaltenbaek R, Demonstration of a compiled version of Shorʼs quantum
Jennewein T and Zeilinger A 2007 High-speed linear optics factoring algorithm using photonic qubits Phys. Rev. Lett. 99
quantum computing using active feed-forward Nature 445 250504
65–69 [78] Lanyon B P, Weinhold T J, Langford N K, Barbieri M,
[52] Kok P 2007 Lecture notes on optical quantum computing James D F V, Gilchrist A and White A G 2007 Experimental
arXiv:0705.4193 demonstration of a compiled version of shorʼs algorithm
[53] Nielsen M A 2006 Cluster-state quantum computation Rep. with quantum entanglement Phys. Rev. Lett. 99 250505
Math. Phys. 57 147–61 [79] Gao W B, Lu C Y, Yao X C, Xu P, Gühne O, Goebel A,
[54] van den Nest M, Dehaene J and de Moor B 2004 Graphical Chen Y A, Peng C Z, Chen Z B and Pan J W 2010
description of the action of local Clifford transformations on Experimental demonstration of a hyper-entangled ten-qubit
graph states Phys. Rev. A 69 022316 schrödinger cat state Nat. Phys. 6 331–5
[55] van den Nest M, Dehaene J and de Moor B 2004 Efficient [80] Cirac J I and Zoller P 1995 Quantum computations with cold
algorithm to recognize the local clifford equivalence of trapped ions Phys. Rev. Lett. 74 4091–4
graph states Phys. Rev. A 70 034302 [81] Monroe C, Meekhof D M, King B E, Itano W M and
[56] Ji Z, Chen J, Wei Z and Ying M 2010 The LU–LC conjecture Wineland D J 1995 Demonstration of a fundamental
is false Quantum Inf. Comput. 1 97–108 quantum logic gate Phys. Rev. Lett. 75 4714–7
[57] Leung D W 2004 Quantum computation by measurements [82] Kielpinski D, Monroe C and Wineland D J 2002 Architecture
Int. J. Quantum Inf. 2 33–43 for a large-scale ion-trap quantum computer Nature 417
[58] Pittman T B, Jacobs B C and Franson J D 2001 Probabilistic 709–11
quantum logic operations using polarizing beam splitters [83] Blinov BB, Moehring DL, Duan L M and Monroe C 2004
Phys. Rev. A 64 062311 Observation of entanglement between a single trapped atom
[59] Pittman T B, Jacobs B C and Franson J D 2002 and a single photon Nature 428 153–7
Demonstration of non-deterministic quantum logic [84] Gerritsma R, Kirchmair G, Zähringer F, Solano E, Blatt R and
operations using linear optical elements Phys. Rev. Lett. 88 Roos CF 2010 Quantum simulation of the Dirac equation
257902 Nature 463 68–71
[60] Yoran N and Reznik B 2003 Deterministic linear optics [85] Kim K, Chang M-S, Korenblit S, Islam R, Edwards E E,
quantum computation with single photon qubits Phys. Rev. Freericks J K, Lin G-D, Duan L-M and Monroe C 2010
Lett. 91 037903 Quantum simulation of frustrated Ising spins with trapped
[61] Nielsen M A 2004 Optical quantum computation using cluster ions Nature 465 590–3
states Phys. Rev. Lett. 96 040503 [86] Schindler P, Barreiro J T, Monz T, Nebendahl V, Nigg D,
[62] Browne D E and Rudolph T 2005 Resource-efficient linear Chwalla M, Hennrich M and Blatt R 2011 Experimental
optical quantum computation Phys. Rev. Lett. 95 10501 repetitive quantum error correction Science 332 1059–61
[63] Kieling K, Gross D and Eisert J 2007 Minimal resources for [87] Lanyon BP et al 2011 Universal digital quantum simulation
linear optical one-way computing J. Opt. Soc. Am. B 24 184–8 with trapped ions Science 334 57–61
[64] Broadbent A, Fitzsimons J and Kashefi E 2009 Universal [88] Blatt R and Roos C 2012 Quantum simulations with trapped
blind quantum computation Proc. 50th Annual Symp. on ions Nat. Phys. 8 277–84
Foundations of Computer Science pp 517–26 [89] Wieman C E, Pritchard D E and Wineland D J 1999 Atom
[65] Childs A 2005 Secure assisted quantum computation cooling, trapping, and quantum manipulation Rev. Mod.
Quantum Inf. Comput. 5 456–66 Phys. 71 253–62
[66] Arrighi P and Salvail L 2006 Blind quantum computation Int. [90] Duan L M, Lukin M D, Cirac J I and Zoller P 2001 Long-
J. Quantum Inf. 4 883–98 distance quantum communication with atomic ensembles
[67] Giovannetti V, Lloyd S and Maccone L 2008 Quantum and linear optics Nature 414 413–8
private queries Phys. Rev. Lett. 100 230502 [91] Wilk T, Webster S C, Kuhn A and Rempe G 2007 Single-
[68] de Martini F, Giovannetti V, Lloyd S, Maccone L, atom single-photon quantum interface Science 317
Nagali E, Sansoni L and Sciarrino F 2009 Experimental quantum 488–90
private queries with linear optics Phys. Rev. A 80 10302 [92] Cory D G et al 2000 Nmr based quantum information
[69] Aharonov D, Ben-Or M and Eban E 2010 Interactive proofs processing: achievements and prospects Fortschr. Phys. 48
for quantum computations Proc. Innovations in Computer 875–907
Science pp 453–69 [93] Vandersypen L M K, Steffen M, Breyta G, Yannoni C S,
[70] Barz S, Kashefi E, Broadbent A, Fitzsimons J F, Sherwood M H and Chuang I L 2001 Experimental
Zeilinger A and Walther P 2012 Demonstration of blind realization of Shorʼs quantum factoring algorithm using
quantum computing Science 335 303–8 nuclear magnetic resonance Nature 414 883–7

23
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

[94] Vandersypen L M K and Chuang I L 2005 Nmr techniques for [119] Yin J et al 2012 Quantum teleportation and entanglement
quantum control and computation Rev. Mod. Phys. 76 1037–69 distribution over 100 kilometre free-space channels Nature
[95] Bloch I, Dalibard J and Nascimbène S 2012 Quantum 488 185–8
simulations with ultracold quantum gases Nat. Phys. 8 [120] Ma X-S et al 2012 Quantum teleportation over 143 kilometres
267–76 using active feed-forward Nature 489 269–73
[96] Yamamoto T, Pashkin Y A, Astafiev O, Nakamura Y and [121] Franson J D, Donegan M M and Jacobs B C 2004 Generation
Tsai J S 2003 Demonstration of conditional gate operation of entangled ancilla states for use in linear optics quantum
using superconducting charge qubits Nature 425 941–4 computing Phys. Rev. A 69 052328
[97] Makhlin Y, Schön G and Shnirman A 2001 Quantum-state [122] Kiesel N, Schmid C, Weber U, Ursin R and Weinfurter H
engineering with Josephson-junction devices Rev. Mod. 2005 Linear optics controlled-phase gate made simple Phys.
Phys. 73 357–400 Rev. Lett. 95 210505
[98] Neeley M et al 2009 Emulation of a quantum spin with a [123] Gasparoni S, Pan J-W, Walther P, Rudolph T and Zeilinger A
superconducting phase qudit Science 325 722–5 2004 Realization of a photonic CNOT gate sufficient for
[99] DiCarlo L et al 2009 Demonstration of two-qubit algorithms quantum computation Phys. Rev. Lett. 93 020504
with a superconducting quantum processor Nature 460 [124] Wagenknecht C, Li C M, Reingruber A, Bao X H, Goebel A,
240–4 Chen Y A, Zhang Q, Chen K and Pan J W 2010
[100] DiCarlo L, Reed MD, Sun L, Johnson BR, Chow JM, Experimental demonstration of a heralded entanglement
Gambetta JM, Frunzio L, Girvin SM, Devoret MH and source Nat. Photonics 4 549–52
Schoelkopf RJ 2010 Preparation and measurement of three- [125] Barz S, Cronenberg G, Zeilinger A and Walther P 2010
qubit entanglement in a superconducting circuit Nature 467 Heralded generation of entangled photon pairs Nat.
574–8 Photonics 4 553–6
[101] Bialczak R C et al 2010 Quantum process tomography of a [126] Walther P, Pan J-W, Aspelmeyer M, Ursin R,
universal entangling gate implemented with josephson phase Gasparoni S and Zeilinger A 2004 De Broglie wavelength of
qubits Nat. Phys. 6 409–13 a non-local four-photon state Nature 429 158–61
[102] Mariantoni M et al 2011 Implementing the quantum von [127] Franson J D, Jacobs B C and Pittman T B 2004 Quantum
Neumann architecture with superconducting circuits Science computing using single photons and the zeno effect Phys.
334 61–65 Rev. A 70 062302
[103] Reed MD, DiCarlo L, Nigg SE, Sun L, Frunzio L, [128] Franson J D, Pittman T B and Jacobs B C 2007 Zeno logic
Girvin SM and Schoelkopf RJ 2012 Realization of three- gates using microcavities J. Opt. Soc. Am. B 24 209–13
qubit quantum error correction with superconducting circuits [129] Langford N K, Ramelow S, Prevedel R, Munro W J,
Nature 482 382–5 Milburn G J and Zeilinger A 2011 Efficient quantum
[104] Berezovsky J, Mikkelsen M H, Stoltz N G, Coldren L A and computing using coherent photon conversion Nature 478
Awschalom D D 2008 Picosecond coherent optical 360–3
manipulation of a single electron spin in a quantum dot [130] Kwiat P G, Mattle K, Weinfurter H, Zeilinger A,
Science 320 349–52 Sergienko A V and Shih Y 1995 New high-intensity source
[106] Fushman I, Englund D, Faraon A, Stoltz N, Petroff P and of polarization-entangled photon pairs Phys. Rev. Lett. 75
Vučković J 2008 Controlled phase shifts with a single 4337–41
quantum dot Science 320 769–72 [131] Gerry C and Knight P 2004 Introductory Quantum Optics
[106] Hanson R and Awschalom D D 2008 Coherent manipulation (Cambridge: Cambridge University Press)
of single spins in semiconductors Nature 453 1043–9 [132] Boyd R W 1992 Nonlinear Optics (San Diego, CA:
[107] O’Brien J L 2007 Optical quantum computing Science 318 Academic)
1567–70 [133] Weinfurter H and Żukowski M 2001 Four-photon
[108] O’Brien J L, Furusawa J A and Vučković J 2009 Photonic entanglement from down-conversion Phys. Rev. A 64
quantum technologies Nat. Photonics 3 687–95 010102
[109] Aspuru-Guzik A and Walther P 2012 Photonic quantum [134] Grice WP and Walmsley IA 1997 Spectral information and
simulators Nat. Phys. 8 285–91 distinguishability in type-II down-conversion with a
[110] Jones R C 1941 A new calculus for the treatment of optical broadband pump Phys. Rev. A 56 1627–34
systems J. Opt. Soc. Am. 31 500–3 [135] Żukowski M, Zeilinger A, Horne M A and Ekert A K 1993
[111] Saleh B E A and Teich M C 1991 Fundamentals of Photonics ‘Event-ready-detectors’ Bell experiment via entanglement
(New York: Wiley) swapping Phys. Rev. Lett. 71 4287–90
[112] Langford N K 2007 Encoding, manipulating and measuring [136] Zukowski M, Zeilinger A and Weinfurter H 1995 Entangling
quantum information in optics PhD Thesis University of photons radiated by independent pulsed sources Annals of
Queensland the New York Academy of Sciences (New York: New York
[113] Hong C K, Ou Z Y and Mandel L 1987 Measurement of Academy of Sciences) pp 91–102
subpicosecond time intervals between two photons by [137] Di Giuseppe G, Haiberger L, De Martini F and Sergienko AV
interference Phys. Rev. Lett. 59 2044–6 1997 Quantum interference and indistinguishability with
[114] Zeilinger A 1981 General properties of lossless beam splitters femtosecond pulses Phys. Rev. A 56 21–24
in interferometry Am. J. Phys. 49 882–3 [138] Prevedel R 2009 Experimental all-optical one-way quantum
[115] Holbrow CH, Galvez E and Parks ME 2002 Photon quantum computing PhD Thesis University of Vienna
mechanics and beam splitters Am. J. Phys. 70 260–5 [139] Walther P 2005 Entanglement and nonlocality in multi-
[116] Zetie KP, Adams SF and Tocknell RM 2000 How does a photon systems PhD Thesis University of Vienna
Mach–Zehnder interferometer work? Phys. Educ. 35 [140] Michler P, Kiraz A, Becher C, Schoenfeld W V, Petroff P M,
46–48 Zhang L, Hu E and Imamoglu A 2000 A quantum dot
[117] Pittman TB, Strekalov DV, Migdall A, Rubin MH, single-photon turnstile device Science 290 2282–5
Sergienko AV and Shih YH 1996 Can two-photon [141] Santori C, Fattal D, Vuckovic J, Solomon G S and
interference be considered the interference of two photons? Yamamoto Y 2002 Indistinguishable photons from a single-
Phys. Rev. Lett. 77 1917–20 photon device Nature 419 594–7
[118] Ursin R et al 2007 Entanglement-based quantum [142] Yuan Z, Kardynal B E, Stevenson R M, Shields A J,
communication over 144 km Nat. Phys. 3 481–6 Lobo C J, Cooper K, Beattie N S, Ritchie D A and Pepper M

24
J. Phys. B: At. Mol. Opt. Phys. 48 (2015) 083001 Tutorial

2002 Electrically driven single-photon source Science 295 [163] Verevkin A, Zhang J, Sobolewski R, Lipatov A, Okunev O,
102–5 Chulkova G, Korneev A, Smirnov K, Gol’tsman G N and
[143] Lodahl P, Van Driel A F, Nikolaev I S, Irman A, Overgaag K, Semenov A 2002 Detection efficiency of large-active-area
Vanmaekelbergh D and Vos W L 2004 Controlling the nbn single-photon superconducting detectors in the
dynamics of spontaneous emission from quantum dots by ultraviolet to near-infrared range Appl. Phys. Lett. 80 4687
photonic crystals Nature 430 654–7 [164] Rosenberg D, Lita A E, Miller A J and Nam S W 2005 Noise-
[144] Stevenson RM, Young RJ, Atkinson P, Cooper K, free high-efficiency photon-number-resolving detectors
Ritchie DA and Shields AJ 2006 A semiconductor source of Phys. Rev. A 71 061803
triggered entangled photon pairs Nature 439 179–82 [165] Gol’tsman G et al 2007 Middle-infrared to visible-light
[145] Akopian N, Lindner NH, Poem E, Berlatzky Y, Avron J, ultrafast superconducting single-photon detectors IEEE
Gershoni D, Gerardot BD and Petroff PM 2006 Entangled Trans. Appl. Supercond. 17 246–51
photon pairs from semiconductor quantum dots Phys. Rev. [166] Lita A E, Miller A J and Nam S W 2008 Counting near-
Lett. 96 130501 infrared single-photons with 95% efficiency Opt. Express 16
[146] Shields A J 2007 Semiconductor quantum light sources Nat. 3032–40
Photonics 1 215–23 [167] Divochiy A et al 2008 Superconducting nanowire photon-
[147] Kuhn A, Hennrich M and Rempe G 2002 Deterministic number-resolving detector at telecommunication
single-photon source for distributed quantum networking wavelengths Nat. Photonics 2 302–6
Phys. Rev. Lett. 89 067901 [168] Ladd T D, Jelezko F, Laflamme R, Nakamura Y,
[148] McKeever J, Boca A, Boozer AD, Miller R, Buck JR, Monroe C and O’Brien J L 2010 Quantum computers
Kuzmich A and Kimble HJ 2004 Deterministic generation of Nature 464 45–53
single photons from one atom trapped in a cavity Science [169] Mandel O, Greiner M, Widera A, Rom T, Hansch T W and
303 1992–4 Bloch I 2003 Controlled collisions for multi-particle
[149] Keller M, Lange B, Hayasaka K, Lange W and Walther H entanglement of optically trapped atoms Nature 425 937–40
2004 Continuous generation of single photons with [170] Weitenberg C, Endres M, Sherson J F, Cheneau M, Schauß P,
controlled waveform in an ion-trap cavity system Nature Fukuhara T, Bloch I and Kuhr S 2011 Single-spin
431 1075–8 addressing in an atomic mott insulator Nature 471 319–24
[150] Eisaman MD, André A, Massou F, Fleischhauer M, [171] André A, DeMille D, Doyle J M, Lukin M D, Maxwell S E,
Zibrov AS and Lukin MD 2005 Electromagnetically Rabl P, Schoelkopf R J and Zoller P 2006 A coherent all-
induced transparency with tunable single-photon pulses electrical interface between polar molecules and mesoscopic
Nature 438 837–41 superconducting resonators Nat. Phys. 2 636–42
[151] Houck A A et al 2007 Generating single microwave photons [172] Schoelkopf RJ and Girvin SM 2008 Wiring up quantum
in a circuit Nature 449 328–31 systems Nature 451 664–9
[152] Kurtsiefer C, Mayer S, Zarda P and Weinfurter H 2000 Stable [173] Verdú J, Zoubi H, Koller C, Majer J, Ritsch H and
solid-state source of single photons Phys. Rev. Lett. 85 Schmiedmayer J 2009 Strong magnetic coupling of an
290–3 ultracold gas to a superconducting waveguide cavity Phys.
[153] Rugar D, Budakian R, Mamin HJ and Chui BW 2004 Single Rev. Lett. 103 43603
spin detection by magnetic resonance force microscopy [174] Togan E, Chu Y, Trifonov A S, Jiang L, Maze J, Childress L,
Nature 430 329–32 Dutt MVG, Soerensen AS and Hemmer PR 2010 Quantum
[154] Neumann P, Mizuochi N, Rempp F, Hemmer P, Watanabe H, entanglement between an optical photon and a solid-state
Yamasaki S, Jacques V, Gaebel T, Jelezko F and spin qubit Nature 466 730–4
Wrachtrup J 2008 Multipartite entanglement among single [175] Lukin MD 2003 Colloquium: trapping and manipulating
spins in diamond Science 320 1326–9 photon states in atomic ensembles Rev. Mod. Phys. 75
[155] Wrachtrup J and Jelezko F 2006 Processing quantum 457–72
information in diamond J. Phys.: Condens. Matter 18 [176] Sherson J F, Krauter H, Olsson R K, Julsgaard B,
807–24 Hammerer K, Cirac I and Polzik E S 2006 Quantum
[156] Politi A, Cryan M J, Rarity J G, Yu S and O’Brien J L 2008 teleportation between light and matter Nature 443 557–60 10
Silica-on-silicon waveguide quantum circuits Science 320 [177] Moehring DL, Maunz P, Olmschenk S, Younge KC,
646–9 Matsukevich DN, Duan LM and Monroe C 2007
[157] Matthews J, Politi A, Stefanov A and O’Brien J 2009 Entanglement of single-atom quantum bits at a distance
Manipulation of multiphoton entanglement in waveguide Nature 449 68–71
quantum circuits Nat. Photonics 3 346–50 [178] Kimble H J 2008 The quantum internet Nature 453 1023–30
[158] Politi A, Matthews J C F and O’Brien J L 2009 Shorʼs [179] Olmschenk S, Matsukevich D N, Maunz P, Hayes D,
quantum factoring algorithm on a photonic chip Science 325 Duan L-M and Monroe C 2009 Quantum teleportation
1221–1221 between distant matter qubits Science 323 486–9
[159] Fulconis J, Alibart O, Wadsworth W, Russell P and Rarity J [180] Duan L M and Monroe C 2010 Colloquium: quantum
2005 High brightness single mode source of correlated networks with trapped ions Rev. Mod. Phys. 82 1209
photon pairs using a photonic crystal fiber Opt. Express 13 [181] Hafezi M, Kim Z, Rolston SL, Orozco LA, Lev BL and
7572–82 Taylor JM 2012 Atomic interface between microwave and
[160] Rarity J, Fulconis J, Duligall J, Wadsworth W and Russell P optical photons Phys. Rev. A 85 020302
2005 Photonic crystal fiber source of correlated photon pairs [182] Vitali D, Barzanjeh S, Abdi M, Tombesi P and Milburn G J
Opt. Express 13 534–44 2012 A reversible optical to microwave quantum interface
[161] Cohen O, Lundeen J S, Smith B J, Puentes G, Mosley P J and Quantum Information and Measurement (Washington, DC:
Walmsley I A 2009 Tailored photon-pair generation in Optical Society of America)
optical fibers Phys. Rev. Lett. 102 123603 [183] Ritter S et al 2012 An elementary quantum network of single
[162] Gol’tsman G N, Okunev O, Chulkova G, Lipatov A, atoms in optical cavities Nature 484 195–200
Semenov A, Smirnov K, Voronov B, Dzardanov A, [184] Hofmann J, Krug M, Ortegel N, Gérard L, Weber M,
Williams C and Sobolewski R 2001 Picosecond Rosenfeld W and Weinfurter H 2012 Heralded
superconducting single-photon optical detector Appl. Phys. entanglement between widely separated atoms Science 337
Lett. 79 705–7 72–75

25