Understanding Quantum Information and Computation

Basics of quantum information

Lesson 3: Quantum circuits

Contents

1. Quantum circuits
2. Inner products, orthonormality, and projections
3. Limitations of quantum information:
• Irrelevance of global phases
• No-cloning theorem
• Non-orthogonal states cannot be perfectly discriminated
1. Quantum circuits
Circuits
Circuits are models of computation:
• Wires carry information
• Gates represent operations
In this series, circuits are always acyclic — information flows from left to right.

Example: Boolean circuits

Wires store binary values, gates represent Boolean logic operations, such
as AND (∧), OR (∨), NOT (¬), and FANOUT ( ).

1
1 ∧
Y
¬ 1
1 1 1
∨
0 0
0 ¬ 0
X ∧
0
Circuits
Circuits are models of computation:
• Wires carry information
• Gates represent operations
In this series, circuits are always acyclic — information flows from left to right.

Example: arithmetic circuits

Wires store numbers and gates represent arithmetic operations, such as

addition (+) and multiplication (∗).
x
x ∗
2
x x

y + 2
x +y
y

2 2 2
y x y+x +y +y
∗
+
1 y+1
Quantum circuits
In the quantum circuit model, the wires represent qubits and the gates represent
both unitary operations and measurements.
Example

1+i
∣0⟩ H S H T 2
∣0⟩ + √1 ∣1⟩
2

⎛ √12 √1
2
⎞ 1 0 1 0
H=⎜
⎜ 1 ⎟
⎟ S=( ) T =( 1+i )
⎝√ − √1 ⎠ 0 i 0 √
2 2 2

⎛ 1+i
2
1−i
2
⎞
T HSH = ⎜
⎜ 1 ⎟
⎟
⎝ √ √i ⎠
2 2
Quantum circuits
Example Convention

In this series (and in Qiskit),

H
ordering qubits from
bottom-to-top is equivalent
+ to ordering them
left-to-right.

1 √1
⎛ √2 2
0 0 ⎞
⎜
⎜ ⎟
⎟
⎜
⎜ 0 0 √1 − √1 ⎟
⎟
⎜
⎜ 2⎟
⎟
⎜
⎜
2 ⎟
⎟
⎜
⎜ √1 √ ⎟
1 ⎟
⎜
⎜ 0 0 ⎟
2 ⎟
⎜
⎜ 2 ⎟
⎟
⎝ √1 − √1 0 0 ⎠
2 2
Quantum circuits
Example

Y H

X +
B

A
Quantum circuits
Example

Y H B

X + A
Quantum circuits
Single-qubit gates Controlled-NOT Swap gate

X
×
+ ×
Y
Toffoli gate Fredkin gate
Z

H ×

S + ×

T
Quantum circuits
It is also sometimes convenient to view arbitrary unitary operations as gates.

Unitary operation Controlled-unitary operation

U
2. Inner products, orthonormality, and projections
Inner products
When we use the Dirac notation, a ket is a column vector, and its corresponding
bra is a row vector:

⎛ α1 ⎞
⎜ ⎟
∣ψ⟩ = ⎜
⎜
⎜ ⋮ ⎟
⎟
⎟ ⟨ψ∣ = (α1 ⋯ αn )
⎜ ⎟
⎝αn ⎠

Suppose that we have two kets:

⎛ α1 ⎞ ⎛ β1 ⎞
⎜ ⎟ ⎜ ⎟
∣ψ⟩ = ⎜
⎜ ⋮ ⎟
⎜ ⎟
⎟ and ∣ϕ⟩ = ⎜
⎜ ⋮ ⎟
⎜ ⎟
⎟
⎜ ⎟ ⎜ ⎟
⎝αn ⎠ ⎝βn ⎠
Inner products
Suppose that we have two kets:

⎛ α1 ⎞ ⎛ β1 ⎞
⎜ ⎟ ⎜ ⎟
∣ψ⟩ = ⎜
⎜
⎜ ⋮ ⎟
⎟ and ∣ϕ⟩ = ⎜
⎜ ⋮ ⎟
⎟
⎜ ⎟ ⎟ ⎜
⎜ ⎟ ⎟
⎝αn ⎠ ⎝βn ⎠

We then have

⎛ β1 ⎞
⎜ ⎟
⟨ψ∣ϕ⟩ = (α1 ⋯ αn ) ⎜
⎜
⎜ ⋮ ⎟
⎟ = α1 β1 + ⋯ + αn βn
⎟
⎜ ⎟
⎝βn ⎠

This is the inner product of ∣ψ⟩ and ∣ϕ⟩.

Inner products
Alternatively, suppose that we have two column vectors expressed like this:

∣ψ⟩ = ∑ αa ∣a⟩ and ∣ϕ⟩ = ∑ βb ∣b⟩

a∈Σ b∈Σ

Then the inner product of these vectors is as follows:

⟨ψ∣ϕ⟩ = ( ∑ αa ⟨a∣)( ∑ βb ∣b⟩)

a∈Σ b∈Σ

= ∑ ∑ αa βb ⟨a∣b⟩
a∈Σ b∈Σ

= ∑ αa β a
a∈Σ
Inner products
Example

∣1⟩ √
∣ϕ⟩ = 21 ∣0⟩ + 2
3
∣1⟩

−∣0⟩ θ ∣0⟩

∣ψ⟩ = √1 ∣0⟩ − √1 ∣1⟩

2 2
−∣1⟩

The inner product of these two vectors is

√
1− 3
⟨ψ∣ϕ⟩ = √ ≈ −0.2588
2 2
Inner products
Example

∣1⟩ √
∣ϕ⟩ = 21 ∣0⟩ + 2
3
∣1⟩

◦
105
−∣0⟩ ∣0⟩

∣ψ⟩ = √1 ∣0⟩ − √1 ∣1⟩

2 2
−∣1⟩

The inner product of these two vectors is

√
1− 3 ◦
⟨ψ∣ϕ⟩ = √ = cos(105 ) ≈ −0.2588
2 2
Inner products
Example

∣1⟩ √
∣ϕ⟩ = 12 ∣0⟩ + 2
3
∣1⟩

◦
90
−∣0⟩ ∣0⟩

√
3
∣ψ⟩ = 2
∣0⟩ − 21 ∣1⟩

−∣1⟩

The inner product of these two vectors is

◦
⟨ψ∣ϕ⟩ = 0 = cos(90 )
Inner products
Relationship to the Euclidean norm

The inner product of any vector

∣ψ⟩ = ∑ αa ∣a⟩
a∈Σ

with itself is

2 2
⟨ψ∣ψ⟩ = ∑ αa αa = ∑ ∣αa ∣ = ∥∣ψ⟩∥
a∈Σ a∈Σ

That is, the Euclidean norm of a vector ∣ψ⟩ is given by

√
∥∣ψ⟩∥ = ⟨ψ∣ψ⟩
Inner products
Conjugate symmetry

For any two vectors

∣ψ⟩ = ∑ αa ∣a⟩ and ∣ϕ⟩ = ∑ βb ∣b⟩

a∈Σ b∈Σ

we have

⟨ψ∣ϕ⟩ = ∑ αa βa and ⟨ϕ∣ψ⟩ = ∑ βa αa

a∈Σ a∈Σ

and therefore

⟨ψ∣ϕ⟩ = ⟨ϕ∣ψ⟩
Inner products
Linearity in the second argument

Suppose that ∣ψ⟩, ∣ϕ1 ⟩, and ∣ϕ2 ⟩ are vectors and α1 and α2 are
complex numbers. If we define a new vector

∣ϕ⟩ = α1 ∣ϕ1 ⟩ + α2 ∣ϕ2 ⟩

then

⟨ψ∣ϕ⟩ = ⟨ψ∣(α1 ∣ϕ1 ⟩ + α2 ∣ϕ2 ⟩) = α1 ⟨ψ∣ϕ1 ⟩ + α2 ⟨ψ∣ϕ2 ⟩

Inner products
Conjugate linearity in the first argument

Suppose that ∣ψ1 ⟩, ∣ψ2 ⟩, and ∣ϕ⟩ are vectors and β1 and β2 are
complex numbers. If we define a new vector

∣ψ⟩ = β1 ∣ψ1 ⟩ + β2 ∣ψ2 ⟩

then

⟨ψ∣ϕ⟩ = (β1 ⟨ψ1 ∣ + β2 ⟨ψ2 ∣)∣ϕ⟩ = β1 ⟨ψ1 ∣ϕ⟩ + β2 ⟨ψ2 ∣ϕ⟩

Inner products
The Cauchy–Schwarz inequality

For every choice of vectors ∣ψ⟩ and ∣ϕ⟩ we have

∣⟨ψ∣ϕ⟩∣ ≤ ∥∣ψ⟩∥ ∥∣ϕ⟩∥

(Equality holds if and only if ∣ψ⟩ and ∣ϕ⟩ are linearly dependent.)
Orthogonality and orthonormality
Two vectors ∣ψ⟩ and ∣ϕ⟩ are orthogonal if their inner product is zero:

⟨ψ∣ϕ⟩ = 0

An orthogonal set {∣ψ1 ⟩, . . . , ∣ψm ⟩} is one where all pairs pairs are orthogonal:

⟨ψj ∣ψk ⟩ = 0 (for all j =

/ k)

An orthonormal set {∣ψ1 ⟩, . . . , ∣ψm ⟩} is an orthogonal set of unit vectors:

1 j=k
⟨ψj ∣ψk ⟩ = { (for all j =
/ k)
0 j=
/k

An orthonormal basis {∣ψ1 ⟩, . . . , ∣ψm ⟩} is an orthonormal set that forms a

basis (of a given space).
Orthogonality and orthonormality
Example

For any classical state set Σ, the set of all standard basis vectors

{∣a⟩ ∶ a ∈ Σ}

is an orthonormal basis.

Example

The set {∣+⟩, ∣−⟩} is an orthonormal basis for the 2-dimensional space
corresponding to a single qubit.

Example
+ − + −
The Bell basis {∣ϕ ⟩, ∣ϕ ⟩, ∣ψ ⟩, ∣ψ ⟩} is an orthonormal basis for the
4-dimensional space corresponding to two qubits.
Orthogonality and orthonormality
Example

The set {∣+⟩, ∣−⟩} is an orthonormal basis for the 2-dimensional space
corresponding to a single qubit.

Example
+ − + −
The Bell basis {∣ϕ ⟩, ∣ϕ ⟩, ∣ψ ⟩, ∣ψ ⟩} is an orthonormal basis for the
4-dimensional space corresponding to two qubits.

Example

The set {∣0⟩, ∣+⟩} is not an orthogonal set because

1
⟨0∣+⟩ = √ = /0
2
Orthogonality and orthonormality
Fact

Suppose that

{∣ψ1 ⟩, . . . , ∣ψm ⟩}

is an orthonormal set of vectors in an n-dimensional space.

(Orthonormal sets are always linearly independent, so these vectors span

a subspace of dimension m ≤ n.)

If m < n, then there must exist vectors

∣ψm+1 ⟩, . . . , ∣ψn ⟩

so that {∣ψ1 ⟩, . . . , ∣ψn ⟩} forms an orthonormal basis.

(The Gram–Schmidt orthogonalization process can be used to construct

these vectors.)
Orthogonality and orthonormality
Orthonormal bases are closely connected with unitary matrices.

These conditions on a square matrix U are equivalent:

† †
1. The matrix U is unitary (i.e., U U = 1 = UU ).
2. The rows of U form an orthonormal basis.
3. The columns of U form an orthonormal basis.

For example, consider a 3 × 3 matrix U:

⎛α1,1 α2,1 α3,1 ⎞ ⎛α1,1 α1,2 α1,3 ⎞

⎜
⎜α ⎟
⎟ ⎜ ⎟
U=⎜ ⎟
†
U =⎜
⎜ α2,2 α3,2 ⎟
⎟ ⎜
⎜α α2,2 α2,3 ⎟
⎟
⎜ 1,2 ⎟ ⎜ 2,1 ⎟
⎝α1,3 α2,3 α3,3 ⎠ ⎝α3,1 α3,2 α3,3 ⎠
Orthogonality and orthonormality
For example, consider a 3 × 3 matrix U:

⎛α1,1 α2,1 α3,1 ⎞ ⎛α1,1 α1,2 α1,3 ⎞

⎜ ⎟ ⎜ ⎟
U =⎜ α3,2 ⎟ U=⎜ α2,3 ⎟
†
⎜
⎜ α α2,2 ⎟
⎟ ⎜
⎜α α2,2 ⎟
⎟
⎜ 1,2 ⎟ ⎜ 2,1 ⎟
⎝α1,3 α2,3 α3,3 ⎠ ⎝α3,1 α3,2 α3,3 ⎠

†
Forming vectors from the columns of U, we can express U U like this:

⎛α1,1 ⎞ ⎛α1,2 ⎞ ⎛α1,3 ⎞

⎜ ⎟ ⎜ ⎟ ⎜ ⎟
∣ψ1 ⟩ = ⎜
⎜ ⎟
⎜α2,1 ⎟
⎟ ∣ψ2 ⟩ = ⎜
⎜ ⎟
⎜α2,2 ⎟
⎟ ∣ψ3 ⟩ = ⎜
⎜ ⎟
⎜α2,3 ⎟
⎟
⎜ ⎟ ⎜ ⎟ ⎜ ⎟
⎝α3,1 ⎠ ⎝α3,2 ⎠ ⎝α3,3 ⎠

⎛⟨ψ1 ∣ψ1 ⟩ ⟨ψ1 ∣ψ2 ⟩ ⟨ψ1 ∣ψ3 ⟩⎞

⎜ ⎟
U U=⎜ ⟨ψ2 ∣ψ3 ⟩⎟
†
⎜
⎜⟨ψ ∣ψ ⟩ ⟨ψ2 ∣ψ2 ⟩ ⎟
⎟
⎜ 2 1 ⎟
⎝⟨ψ3 ∣ψ1 ⟩ ⟨ψ3 ∣ψ2 ⟩ ⟨ψ3 ∣ψ3 ⟩⎠
Orthogonality and orthonormality
These conditions on a square matrix U are equivalent:
† †
1. The matrix U is unitary (i.e., U U = 1 = UU ).
2. The rows of U form an orthonormal basis.
3. The columns of U form an orthonormal basis.
Fact

Given any orthonormal set of n-dimensional vectors

{∣ψ1 ⟩, . . . , ∣ψm ⟩}

there is a unitary matrix U whose first m columns are these vectors:

⎛ ⋮ ⋮ ⋮ ⋮ ⋮ ⎞
⎜ ⎟
U=⎜
⎜
⎜ ∣ψ1 ⟩ ∣ψ2 ⟩ ⋯ ∣ψm ⟩ ∣ψm+1 ⟩ ⋯ ∣ψn ⟩ ⎟
⎟
⎟
⎜ ⎟
⎝ ⋮ ⋮ ⋮ ⋮ ⋮ ⎠
Projections
A square matrix Π is called a projection if it satisfies two properties:
†
1. Π = Π
2
2. Π = Π
Example

If ∣ψ⟩ is a unit vector, then this matrix is a projection:

Π = ∣ψ⟩⟨ψ∣
† † † †
Π = (∣ψ⟩⟨ψ∣) = (⟨ψ∣) (∣ψ⟩) = ∣ψ⟩⟨ψ∣ = Π

† † †
(AB) = B A
Projections
A square matrix Π is called a projection if it satisfies two properties:
†
1. Π = Π
2
2. Π = Π
Example

If ∣ψ⟩ is a unit vector, then this matrix is a projection:

Π = ∣ψ⟩⟨ψ∣
† † † †
Π = (∣ψ⟩⟨ψ∣) = (⟨ψ∣) (∣ψ⟩) = ∣ψ⟩⟨ψ∣ = Π
2 2
Π = (∣ψ⟩⟨ψ∣) = ∣ψ⟩⟨ψ∣ψ⟩⟨ψ∣ = ∣ψ⟩⟨ψ∣ = Π
Projections
A square matrix Π is called a projection if it satisfies two properties:
†
1. Π = Π
2
2. Π = Π
Example

If {∣ψ1 ⟩, . . . , ∣ψm ⟩} is an orthonormal set, then this is a projection:

m
Π = ∑ ∣ψk ⟩⟨ψk ∣
k=1

m † m m
† †
Π = ( ∑ ∣ψk ⟩⟨ψk ∣) = ∑ (∣ψk ⟩⟨ψk ∣) = ∑ ∣ψk ⟩⟨ψk ∣ = Π
k=1 k=1 k=1
m m m
2
Π = ∑ ∑ ∣ψj ⟩⟨ψj ∣ψk ⟩⟨ψk ∣ = ∑ ∣ψk ⟩⟨ψk ∣ = Π
j=1 k=1 k=1
Projections
A square matrix Π is called a projection if it satisfies two properties:
†
1. Π = Π
2
2. Π = Π
Fact

Every projection matrix Π takes the form

m
Π = ∑ ∣ψk ⟩⟨ψk ∣
k=1

for some orthonormal set {∣ψ1 ⟩, . . . , ∣ψm ⟩}.

(This includes the case Π = 0.)

Projective measurements
A collection of projections {Π1 , . . . , Πm } that satisfies

Π1 + ⋯ + Πm = 1

describes a projective measurement.

When such a measurement is performed on a system in the state ∣ψ⟩, two

things happen:
1. The outcome k ∈ {1, . . . , m} of the measurement is chosen randomly:

2
Pr(outcome is k) = ∥Πk ∣ψ⟩∥ = ⟨ψ∣Πk ∣ψ⟩

2. The state of the system becomes

Πk ∣ψ⟩
∥Πk ∣ψ⟩∥
Projective measurements
We can also choose different names for the measurement outcomes. Any
collection of projections {Πa ∶ a ∈ Γ } that satisfies the condition

∑ Πa = 1
a∈Γ

describes a projective measurement having outcomes in the set Γ . The rules are
the same as before:
1. The outcome a ∈ Γ of the measurement is chosen randomly:

2
Pr(outcome is a) = ∥Πa ∣ψ⟩∥

2. The state of the system becomes

Πa ∣ψ⟩
∥Πa ∣ψ⟩∥
Projective measurements
Example

Standard basis measurements are projective measurements:

• The outcomes are the classical states of the system being
measured.
• The measurement is described by the set {∣a⟩⟨a∣ ∶ a ∈ Σ}.

Suppose that we measure the state

∣ψ⟩ = ∑ αa ∣a⟩
a∈Σ

2 2
Each outcome a appears with probability ∥∣a⟩⟨a∣ψ⟩∥ = ∣αa ∣ .

Conditioned on the outcome a, the state becomes

∣a⟩⟨a∣ψ⟩ αa
= ∣a⟩
∥∣a⟩⟨a∣ψ⟩∥ ∣αa ∣
Projective measurements
Example

Standard basis measurements are projective measurements:

• The outcomes are the classical states of the system being
measured.
• The measurement is described by the set {∣a⟩⟨a∣ ∶ a ∈ Σ}.

Example

Performing a standard basis measurement on a system X and doing

nothing to a system Y is equivalent to performing the projective
measurement

{∣a⟩⟨a∣ ⊗ 1Y ∶ a ∈ Σ}

on the system (X, Y).

Projective measurements
Example

Performing a standard basis measurement on a system X and doing

nothing to a system Y is equivalent to performing the projective
measurement

{∣a⟩⟨a∣ ⊗ 1Y ∶ a ∈ Σ}

on the system (X, Y).

Each measurement outcome a appears with probability

2
∥(∣a⟩⟨a∣ ⊗ 1)∣ψ⟩∥

The state of the system (X, Y) then becomes

(∣a⟩⟨a∣ ⊗ 1)∣ψ⟩
∥(∣a⟩⟨a∣ ⊗ 1)∣ψ⟩∥
Projective measurements
Example

Define two projections as follows:

+ − − + +
Π0 = ∣ϕ⟩⟨ϕ ∣ + ∣ϕ ⟩⟨ϕ ∣ + ∣ψ ⟩⟨ψ ∣
− −
Π1 = ∣ψ ⟩⟨ψ ∣

The projective measurement {Π0 , Π1 } is an interesting one…

Every projective measurements can be implemented using unitary operations

and standard basis measurements.

×
×
∣0⟩ H H
3. Limitations of quantum information
Irrelevance of global phases
Definition

Suppose that ∣ψ⟩ and ∣ϕ⟩ are quantum state vectors satisfying

∣ϕ⟩ = α∣ψ⟩

The states ∣ψ⟩ and ∣ϕ⟩ are then said to differ by a global phase.
iθ
(This requires ∣α∣ = 1. Equivalently, α = e for some real number θ.)

Imagine that two states that differ by a global phase are measured. If we start
with the state ∣ϕ⟩, the probability to obtain any chosen outcome a is

2 2 2 2 2
∣⟨a∣ϕ⟩∣ = ∣α⟨a∣ψ⟩∣ = ∣α∣ ∣⟨a∣ψ⟩∣ = ∣⟨a∣ψ⟩∣

That’s the same probability as if we started with the state ∣ψ⟩.

Irrelevance of global phases
Definition

Suppose that ∣ψ⟩ and ∣ϕ⟩ are quantum state vectors satisfying

∣ϕ⟩ = α∣ψ⟩

The states ∣ψ⟩ and ∣ϕ⟩ are then said to differ by a global phase.
iθ
(This requires ∣α∣ = 1. Equivalently, α = e for some real number θ.)

Imagine that two states that differ by a global phase are measured. If we start
with the state ∣ϕ⟩, the probability to obtain any chosen outcome a is

2 2 2 2 2
∥Πa ∣ϕ⟩∥ = ∥αΠa ∣ψ⟩∥ = ∣α∣ ∥Πa ∣ψ⟩∥ = ∥Πa ∣ψ⟩∥

That’s the same probability as if we started with the state ∣ψ⟩.

Irrelevance of global phases
Definition

Suppose that ∣ψ⟩ and ∣ϕ⟩ are quantum state vectors satisfying

∣ϕ⟩ = α∣ψ⟩

The states ∣ψ⟩ and ∣ϕ⟩ are then said to differ by a global phase.
iθ
(This requires ∣α∣ = 1. Equivalently, α = e for some real number θ.)

Suppose we apply a unitary operation to two states that differ by a global phase:

U∣ϕ⟩ = αU∣ψ⟩ = α(U∣ψ⟩)

They still differ by a global phase…

Consequently, two quantum state vectors ∣ψ⟩ and ∣ϕ⟩ that differ by a global
phase are completely indistinguishable and are considered to be equivalent.
Irrelevance of global phases
Example

The quantum states

1 1 1 1
∣−⟩ = √ ∣0⟩ − √ ∣1⟩ and − ∣−⟩ = − √ ∣0⟩ + √ ∣1⟩
2 2 2 2

differ by a global phase.

Irrelevance of global phases
Example

The quantum states

1 1 1 1
∣+⟩ = √ ∣0⟩ + √ ∣1⟩ and ∣−⟩ = √ ∣0⟩ − √ ∣1⟩
2 2 2 2

do not differ by a global phase. (This is a relative phase difference.)

This is consistent with the observation that these states can be

discriminated perfectly:

2 2
∣⟨0∣H∣+⟩∣ = 1 ∣⟨0∣H∣−⟩∣ = 0
2 2
∣⟨1∣H∣+⟩∣ = 0 ∣⟨1∣H∣−⟩∣ = 1
No-cloning theorem
Theorem (No-cloning theorem)

Let X and Y both have the classical state set {0, . . . , d−1}, where d ≥ 2.
There does not exist a unitary operation U on the pair (X, Y) such that

∀∣ψ⟩ ∶ U(∣ψ⟩ ⊗ ∣0⟩) = ∣ψ⟩ ⊗ ∣ψ⟩

∣00000⟩ ∣ψ⟩

∣ψ⟩ ∣ψ⟩
No-cloning theorem
Theorem (No-cloning theorem)

Let X and Y both have the classical state set {0, . . . , d−1}, where d ≥ 2.
There does not exist a unitary operation U on the pair (X, Y) such that

∀∣ψ⟩ ∶ U(∣ψ⟩ ⊗ ∣0⟩) = ∣ψ⟩ ⊗ ∣ψ⟩

The operation U must clone the standard basis states ∣0⟩ and ∣1⟩:

U(∣0⟩ ⊗ ∣0⟩) = ∣0⟩ ⊗ ∣0⟩

U(∣1⟩ ⊗ ∣0⟩) = ∣1⟩ ⊗ ∣1⟩

Therefore, by linearity,

1 1 1 1
U(( √ ∣0⟩ + √ ∣1⟩) ⊗ ∣0⟩) = √ ∣0⟩ ⊗ ∣0⟩ + √ ∣1⟩ ⊗ ∣1⟩
2 2 2 2
No-cloning theorem
Theorem (No-cloning theorem)

Let X and Y both have the classical state set {0, . . . , d−1}, where d ≥ 2.
There does not exist a unitary operation U on the pair (X, Y) such that

∀∣ψ⟩ ∶ U(∣ψ⟩ ⊗ ∣0⟩) = ∣ψ⟩ ⊗ ∣ψ⟩

Therefore, by linearity,

1 1 1 1
U(( √ ∣0⟩ + √ ∣1⟩) ⊗ ∣0⟩) = √ ∣0⟩ ⊗ ∣0⟩ + √ ∣1⟩ ⊗ ∣1⟩
2 2 2 2

But this is not the correct behavior — we must have

1 1
U(( √ ∣0⟩ + √ ∣1⟩) ⊗ ∣0⟩)
2 2
1 1 1 1
= ( √ ∣0⟩ + √ ∣1⟩) ⊗ ( √ ∣0⟩ + √ ∣1⟩)
2 2 2 2
No-cloning theorem
Theorem (No-cloning theorem)

Let X and Y both have the classical state set {0, . . . , d−1}, where d ≥ 2.
There does not exist a unitary operation U on the pair (X, Y) such that

∀∣ψ⟩ ∶ U(∣ψ⟩ ⊗ ∣0⟩) = ∣ψ⟩ ⊗ ∣ψ⟩

Remarks:
• Approximate forms of the cloning theorem are known.
• Copying a standard basis state is possible — the no-cloning theorem does
not contradict this.

∣0⟩ + ∣a⟩

∣a⟩ ∣a⟩

• Cloning a probabilistic state (classically) is also impossible.

Discriminating non-orthogonal states
It is not possible to perfectly discriminate two non-orthogonal quantum states.
Equivalently, if we can discriminate two quantum states perfectly, then they must be
orthogonal.

Two states ∣ψ⟩ and ∣ϕ⟩ can be discriminated perfectly if there is a unitary operation
U that works like this:

0 1
∣ψ⟩ ∣ϕ⟩

U U

∣0⋯0⟩ ∣0⋯0⟩
Discriminating non-orthogonal states
0 1
∣ψ⟩ ∣ϕ⟩

U U

∣0⋯0⟩ ∣0⋯0⟩

U(∣0⋯0⟩∣ψ⟩) = ∣π0 ⟩∣0⟩ U(∣0⋯0⟩∣ϕ⟩) = ∣π1 ⟩∣1⟩

† †
∣0⋯0⟩∣ψ⟩ = U (∣π0 ⟩∣0⟩) ∣0⋯0⟩∣ϕ⟩ = U (∣π1 ⟩∣1⟩)

⟨ψ∣ϕ⟩ = ⟨0⋯0∣0⋯0⟩⟨ψ∣ϕ⟩
†
= (⟨π0 ∣⟨0∣)UU (∣π1 ⟩∣1⟩) = ⟨π0 ∣π1 ⟩⟨0∣1⟩ = 0
Discriminating non-orthogonal states
Conversely, orthogonal quantum states can be perfectly discriminated.

In particular, if ∣ψ⟩ and ∣ϕ⟩ are orthogonal, then any unitary matrix whose first
two columns are ∣ψ⟩ and ∣ϕ⟩ will work.

⎛ ⎞
⎜
⎜ ⋮ ⋮ ⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜ ∣ψ⟩ ∣ϕ⟩
U=⎜ ⎟
⎟
⎜ ? ⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜ ⎟
⎟
⎜
⎜ ⋮ ⋮ ⎟
⎟
⎝ ⎠

∣ψ⟩ U
† ∣0⋯00⟩ ∣ϕ⟩ †
U ∣0⋯01⟩
Discriminating non-orthogonal states
Alternatively, we can define a projective measurement {Π0 , Π1 } like this:

Π0 = ∣ψ⟩⟨ψ∣ Π1 = 1 − ∣ψ⟩⟨ψ∣

If we measure the state ∣ψ⟩…

2 2
Pr[outcome is 0] = ∥Π0 ∣ψ⟩∥ = ∥∣ψ⟩∥ = 1
2 2
Pr[outcome is 1] = ∥Π1 ∣ψ⟩∥ = ∥0∥ = 0

If we measure any state ∣ϕ⟩ orthogonal to ∣ψ⟩…

2 2
Pr[outcome is 0] = ∥Π0 ∣ϕ⟩∥ = ∥0∥ = 0
2 2
Pr[outcome is 1] = ∥Π1 ∣ϕ⟩∥ = ∥∣ϕ⟩∥ = 1