Francesco Salvarani

Basic mathematics for quantum

problems

ESILV A3 - Academic year 2024/2025

2

These notes cover the first part of the course “Quantique” at the École Supérieure
d’Ingénieurs Léonard de Vinci. They have been based on these works:
- Lang, Serge. Linear algebra. Third edition. Springer-Verlag, New York, 1987, ISBN
0-387-96412-6
- Rieffel, Eleanor & Polak, Wolfgang. Quantum computing. A gentle introduction,
The MIT Press, Cambridge, Mass. 2011, ISBN 978-0-262-01506-6
- Bodin, Arnaud. Quantum. Un peu de mathématiques pour l’informatique quan-
tique, Exo7, 2021
http://exo7.emath.fr/cours/livre-quantum.pdf
 3

Contents
1. Introduction 4
2. Complex numbers 5
2.1. Algebraic form of a complex number 5
2.2. Trigonometric form of a complex number 6
2.3. Exponential form of a complex number 7
3. Qubits 8
3.1. Definition of a qubit 8
3.2. Operations 8
3.3. Manipulation of qubits 10
3.4. Bloch sphere 11
4. Vector spaces 13
4.1. Basic notions 13
4.2. Qubits as vectors 15
5. Hermitian products and orthogonality 16
5.1. Hermitian products 16
5.2. Orthogonal vectors 17
6. Tensor product of vectors 19
6.1. Product of qubits 19
6.2. Quantum entanglement 20
7. Matrices 22
7.1. Determinants 22
7.2. Adjoint matrix 23
7.3. Bra-ket notation for vectors 24
7.4. Unitary matrix 24
8. Eigenvectors and eigenfunctions 26
9. Density matrices for 1-qubits and their applications 27
9.1. Density matrix for 1-qubits 27
9.2. Pure states 27
9.3. Mixed states 27
9.4. Applications of density matrices 28
Bibliography 30
Credits 30
4

1. Introduction
Quantum mechanics represents a paradigm shift that overcame some important weak-
nesses of 19th century physics and led to the birth of modern physics. The basic ideas
of quantum mechanics have also had many positive effects in other disciplines (such as
computer science). In these notes, the basic mathematical tools needed to describe some
quantum problems will be developed. The examples will be oriented, in particular, towards
applications in the field of quantum computing. These notes describe the basic mathe-
matical methods of matrix mechanics. The mathematics of wave mechanics is beyond the
scope of these notes. Readers interested in the latter description are invited to consult more
advanced texts.

Notations

Re(z) real part of the complex number z

Im(z) imaginary part of the complex number z
|z| modulus of the complex number z
z∗ conjugate of the complex number z (other notation: z̄)
x column vector in Cn
|φi column vector in Cn
|φi qubit
kψk norm of the qubit |φi
hφ| dual of a vector in Cn
hu|vi Hermitian product of two vectors u and v in Cn
kuk norm of the vector u ∈ Cn induced by the Hermitian product
u⊗v tensor product of two vectors u and v in Cn
|φi ⊗ |ψi tensor product of two qubits |φi and |ψi
A matrix
Mn,p (C) set of n ⊗ p matrices with entries in C
Mn (C) set of square matrices n ⊗ n with entries in C
A−1 inverse of the matrix A
det(A) determinant of the matrix A
A∗ adjoint of the matrix A
σ1 , σ2 , σ3 Pauli matrices
 5

2. Complex numbers
Complex numbers are necessary to study quantum problems. In what follows, we review
some concepts which should be familiar to the reader.
The imaginary unit or unit imaginary number, denoted as i, is defined as a solution of
the algebraic equation x2 + 1 = 0. Although there is no real number with this property, the
imaginary unit allows to extend the real number system.
Note tha the defining equation has two distinct solutions (because we consider a quadratic
polynomial without multiple roots). Once a solution i of the equation has been fixed, −i is
also a solution. Of course, i 6= −i. The aforementioned equation is the only definition of i.
However, it is easy to see that, if i is a solution, −i is another solution of the same equation.
Hence, it could seem that i is not well defined. However, there is no ambiguity as long as
either solution is chosen and labelled as i, the other then being labelled as −i. Although
i 6= −i, they are algebraically equivalent and both imaginary numbers have the same right
to be the number whose square is equal to −1.

2.1. Algebraic form of a complex number. The algebraic definition of complex number
is the following.
Definition 1. A complex number is a pair (a, b) ∈ R2 , denoted a + ib.
The set of complex number is denoted C.
Complex numbers can be added together and multiplied. The results of these operations
are also complex numbers. Therefore, the set of complex numbers is closed with respect to
the operations of addition and multiplication:

(a + ib) + (a0 + ib0 ) = (a + a0 ) + i (b + b0 )

(a + ib) ⊗ (a0 + ib0 ) = (aa0 − bb0 ) + i (ab0 + ba0 ) .

For the multiplication rule, we used the property i2 = −1.

Example: Let z1 = 4 + 2i and z2 = 3 − i. Alors

z1 + z2 = (4 + 2i) + (3 − i) = 7 + i
and
z1 ⊗ z2 = (4 + 2i) ⊗ (3 − i) = (12 + 2) + i(−4 + 6)
= 14 + 2i.

Some useful definitions are the following.

Definition 2. Let z = a + ib be a complex number, its real part is the real a and is denoted
by Re(z); its imaginary part is the real b and is denoted Im(z).
Definition 3. The modulus of the complex number z = a + ib, denoted |z|, is the non-
negative real number
√
|z| = a2 + b2 .
6
q √
Example: |3 − 4i| = 32 + (−4)2 = 25 = 5.

A complex number can be represented in the complex plane, i.e. the Cartesian plane
with Re(z) on the x-axis and Im(z) on the y-axis.
We note that the set of all complex numbers of modulus 1 can be represented by the circle
of radius 1 centred at the origin in the complex plane. Of course, any non-zero complex
number can be transformed into a complex number of modulus 1 by dividing it by its
modulus.
√ √
2 2
Example: −i is a complex number of modulus 1.
2 2

Definition 4. The conjugate of the complex number z = a + ib is z ∗ = a − ib.

We have in particular that Re (z ∗ ) = Re(z) and Im (z ∗ ) = − Im(z). The point z ∗ is the
∗ 2 2
√ to the real axis. Since zz = (a + ib)(a − ib) = a + b
symmetric of the point z with respect
then the modulus of z is also |z| = zz ∗ .
A more classical notation for the conjugate is z̄, however in this note the notation z ∗ is
preferred.
Lemma 5. For any z 6= 0, there exists a unique z −1 ∈ C such that zz −1 = 1 (where
1 = 1 + 0 i). Moreover
1 a − ib z∗
z −1 = = 2 =
z a + b2 |z|2
The complex number z −1 is the inverse of the complex number z.
2.2. Trigonometric form of a complex number. A complex number z ∈ C can be
written as follows:
z = r cos θ + ir sin θ with r ∈ R+ and θ ∈ R.
Note that r is in fact the modulus of z : r = |z|. Moreover, θ is an argument of z, denoted
arg(z). The argument is not unique: if θ is an argument then so is θ + 2kπ(k ∈ Z). To
make the argument unique, we can impose the condition θ ∈] − π, π] (or θ ∈ [0, 2π[ ). If we
impose θ ∈] − π, π] then for a non-zero complex number z, the writing z = r cos θ + ir sin θ
is unique. In general, arg(z) is “defined modulo 2π” and the writing θ ≡ θ0 (mod2π) means
that θ = θ0 + 2kπ for a certain integer k ∈ Z.
√ π
Example: Let z = 1 + 3i. Then r = |z| = 2 and θ =
.
3
The complex number of module r = 3 and argument θ = π/4 is
√ √
π π 3 2 3 2
    
0
z = r(cos θ + i sin θ) = 3 cos + i sin = + i.
4 4 2 2
The trigonometric form facilitates the calculation of multiplications (or divisions), because
moduli multiply and arguments add.
Proposition 6. Let z and z 0 be two complex numbers. Then
|zz 0 | = |z| · |z 0 | and arg (zz 0 ) ≡ [arg(z) + arg (z 0 )](mod2π)
 7

Proof.
zz 0 = |z|(cos θ + i sin θ) |z 0 | (cos θ0 + i sin θ0 )
= |zz 0 | (cos θ cos θ0 − sin θ sin θ0 + i (cos θ sin θ0 + sin θ cos θ0 ))
= |zz 0 | (cos (θ + θ0 ) + i sin (θ + θ0 ))
hence |zz 0 | = |z| · |z 0 | et arg (zz 0 ) ≡ [arg(z) + arg (z 0 )](mod2π). 
2.3. Exponential form of a complex number. A complex number z1 ∈ C of modulus
1 can be written as follows:
z1 = eiθ = cosθ + i sin θ,
and thus, any complex number z ∈ C can be written
z = reiθ ,
where r = |z| is its modulus and θ = arg(z) is one of its arguments.
π 3π
Example: ei 2 = i, eiπ = −1, ei 2 = −i and e2iπ = e0 = 1.
In the framework of the exponential notation, calculations are performed with the usual
laws for powers. For example:  n
eiθ = einθ .
This formula, written in the trigonometric formulation, is nothing but than De Moivre’s
formula:
(cos θ + i sin θ)n = cos(nθ) + i sin(nθ).
0
For z = reiθ and z 0 = r0 eiθ , it is possible to deduce that
0 0
- zz 0 = rr0 eiθ eiθ = rr0 ei(θ+θ ) ;
 n  n
- z n = reiθ = rn eiθ = rn einθ ;
1 1 1
- = iθ = e−iθ ;
z re r
∗ −iθ
- z = re .
8

3. Qubits
A qubit is a basic unit of quantum information. It is a two-state system and shares the
main properties of two-state systems of quantum mechanics. It can exist in any quantum
superposition of two independent and distinguishable states.

3.1. Definition of a qubit. The definition of 1-qubit is the basic brick of qubit theory, and
it is summarized below. For some authors, the definition of a qubit includes its normalization
(in a sense that will be specified later). Other authors do not impose a normalization
condition to define the qubit, but treat qubits as equivalence classes and use the normalized
qubit as the representative element of the class.
In these notes, the normalization condition in the definition of qubits will be included.
For didactical purposes, these notes make the difference between (non-normalized) “states”
and (normalized) “qubits”. However, when a non-normalized quantum state will be be
manipulated, it will be meant that the considered quantity is equivalent to a qubit. The
normalization procedure allows to deduce the corresponding normalized qubit.
Definition 7. Consider two basic, independent quantum states |0i and |1i. A 1-qubit is a
quantum state obtained as linear combination of the two basic quantum states |0i and |1i,
α|0i + β|1i with α ∈ C and β ∈ C,
under the normalization condition
|α|2 + |β|2 = 1.
Remark 1. A qubit is defined by two complex numbers, α = a1 + ib1 and β = a2 + ib2 or,
equivalently, by the four real numbers a1 , b1 , a2 and b2 .
Two qubits joined together are in a quantum state |ψi, called 2-qubit when normalized,
defined as follows.
Definition 8. Consider four basic, independent quantum states |0.0i, |0.1i, |1.0i and |1.1i.
A 2-qubit is a quantum state obtained by the superposition
α|0.0i + β|0.1i + γ|1.0i + δ|1.1i with α, β, γ, δ ∈ C
under the normalization condition
|α|2 + |β|2 + |γ|2 + |δ|2 = 1.
Of course, these definitions can be generalized for defining d-qubits in higher dimension
d.
Remark 2. The notation h | (whose usefulness will become clearer when reading Section
5.1) reads “bra”, whereas the notation | i reads “ket”.
3.2. Operations. The sum of two quantum states is done coefficient by coefficient (in the
case of a sum, the normalization condition is not guaranteed). In practice, it consists in
adding pairs of complex numbers. For example if
|φi = (1 + 4i)|0i + 2i|1i et |ψi = 2|0i + (1 − i)|1i
then
|φi + |ψi = (3 + 4i)|0i + (1 + i)|1i.
 9

In the case of 2-qubits:

√ √
2 2 √
(|1.0i + |0.1i) + (|1.0i − |0.1i) ≡ 2|1.0i.
2 2
Two 1-qubits can be multiplied to obtain a 2-qubit. The calculations are done as algebraic
calculations using the basic rules
|0i ⊗ |0i = |0.0i, |0i ⊗ |1i = |0.1i,

|1i ⊗ |0i = |1.0i, |1i ⊗ |1i = |1.1i.

For example, consider

1 1
|φi = √ ((1 + 3i)|0i + 2i|1i) et |ψi = √ (3|0i + (1 − i)|1i) .
14 11
By observing that
(1 + 3i)(1 − i) = 1 − i + 3i − 3i2 = 4 + 2i and 2i(1 − i) = 2i − 2i2 = 2 + 2i,
we have
1 h i
|φi ⊗ |ψi = √ ((1 + 3i)|0i + 2i|1i) ⊗ (3|0i + (1 − i)|1i)
154
1 h i
=√ 3(1 + 3i)|0i⊗|0i + (1 + 3i)(1 − i)|0i⊗|1i + 6i|1i⊗|0i + 2i(1 − i)|1i⊗|1i
154
1 1
=√ (3 + 9i)|0.0i + √ (4 + 2i)|0.1i + 6i|1.0i + (2 + 2i)|1.1i.
154 154
Remark 3. The same procedure can be applied for multiplying two non-normalized quantum
states.
The norm of a state is a real number, denoted kψk. In particular, for a state described
by the quantity
|ψi = α|0i + β|1i,
we have that it norm is q
kψk = |α|2 + |β|2 .
On the other hand, in the case of four basic state spaces, the state
|ψi = α|0.0i + β|0.1i + γ|1.0i + δ|1.1i
has norm q
kψk = |α|2 + |β|2 + |γ|2 + |δ|2 .
The normalization of a state |ψi is the qubit (of norm equal to one) given by |ψ̄i =
|ψi/kψk.

Example: Let |ψi = (1 + 2i)|0i + (1 − i)|1i. Then

   
kψk2 = |1 + 2i|2 + |1 − i|2 = 12 + 22 + 12 + (−1)2 = 7.
√
Therefore kψk = 7.
10

3.3. Manipulation of qubits. By using the exponential form, a state |ψi = α|0i + β|1i
can be written as follows:
0
|ψi = reiθ |0i + r0 eiθ |1i.
A state is a normalized qubit when r2 + r02 = 1.
π π π π
Example: If |φi = 2ei 3 |0i + 3ei 4 and |ψi = ei 2 |0i + 2ei 3 , then
π π π π
   
|φi ⊗ |ψi = 2ei 3 |0i + 3ei 4 |1i ⊗ ei 2 |0i + 2ei 3
π π π π π π π π
= 2ei 3 · ei 2 |0.0i + 2ei 3 · 2ei 3 |0.1i + 3ei 4 · ei 2 |1.0i + 3ei 4 · 2ei 3 |1.1i
π π π π π π π π
= 2ei( 3 + 2 ) |0.0i + 4ei( 3 + 3 ) |0.1i + 3ei( 4 + 2 ) |1.0i + 6ei( 4 + 3 ) |1.1i
5π 2π 3π 7π
= 2e 6 i |0.0i + 4e 4 i |0.1i + 3e 4 i |1.0i + 6e 12 i |1.1i.
Two quantum states are equivalent if it is possible to obtain the first one from the other
one by means of the following operations:
- multiplication by a real constant:
k|ψi ≡ |ψi where k ∈ R∗ ;
- multiplication by a complex number with modulus equal to one (i.e. of type eiθ ):
eiθ |ψi ≡ |ψi where θ ∈ R.
In other words, two quantum states |φi et |ψi are equivalent if there exists z ∈ C∗ such
that |φi = z|ψi. Two equivalent quantum states cannot be distinguished by measurements.

Example: Consider
2 1
2|0i + |1i ≡ √ |0i + √ |1i.
5 5
√
The second qubit is obtained by multiplying the first quantum state by k = 1/ 5.
The two equivalent quantum states i|0i + (2 − i)|1i and −|0i + 1 + 2i|1i both lead to 0
with probability 1/6 and to 1 with probability 5/6 when measured.
We underline that two equivalent states are not equal and therefore they must not be
interchanged in the intermediate calculations. However, during the final measurement, one
can replace a state by an equivalent state without changing the result, because the two
elementary operations that define the equivalence do not change the probability calculation
for the measurement.
Proposition 9. A non-zero quantum state |ψi = α|0i + β|1i is equivalent to a qubit of
norm one, given by ! !
θ θ iφ
|ψi ≡ cos |0i + sin e |1i.
2 2
Moreover, the writing is unique when 0 < θ < π and −π < φ ≤ π.
Proof. We first prove existence. We start by transforming the coefficient of |0i into a positive
real. If α = reiθ then
|ψi ≡ e−iθ |ψi
 
= e−iθ reiθ |0i + β|1i
= r|0i + βe−iθ |1i.
 11

We then normalize this quantum state by dividing by kψk :

1
|ψi ≡ |ψi
kψk
1  
≡ r|0i + βe−iθ |1i
kψk
r βe−iθ
= |0i + |1i
kψk kψk
This last quantum state is
|ψ 0 i = r0 |0i + β 0 |1i
with r0 ∈ R and β 0 ∈ C. But |ψ 0 i is a qubit of norm 1 by definition, we have moreover  
r02 + |β 0 |2 = 1 and in particular 0 ≤ r0 ≤ 1 et |β 0 | ≤ 1. We now notice that r0 = cos 2θ
   
and |β 0 | = sin 2θ . Finally, β 0 = sin 2θ eiφ for a certain argument φ ∈ R. Thus |ψi is well
equivalent to a qubit of the desired form:
! !
0 θ θ iφ
|ψi ≡ |ψ i = cos |0i + sin e |1i.
2 2
In order to prove uniqueness, suppose by contradiction that there exist (θ, φ) and (θ0 , φ0 )
such that
θ0 θ0 iφ0
! ! ! !
θ θ iφ
cos |0i + sin e |1i = cos |0i + sin e |1i.
2 2 2 2
Then, by identification, the coefficients in front of |0i are equal, as are the coefficients in
front of |1i. Thus
θ0 θ0 iφ0
! ! ! !
θ θ iφ
cos = cos and sin e = sin e .
2 2 2 2
But 0 < θ/2 < π/2 and 0 < θ0 /2 < π/2. Hence cos (θ/2) = cos (θ0 /2) implies that θ = θ0 .
0
We therefore deduce that sin (θ/2) = sin (θ0 /2), then that eiφ = eiφ . Two arguments are
equal modulo 2, but since we have imposed −π < φ0 and −π < φ0 , we have φ = φ0 . 
3.4. Bloch sphere. The Bloch sphere is a geometrical representation of the pure state
space of a qubit.
We know that a qubit |ψi = α|0i + β|1i is defined by its complex coefficients α, β. This
qubit is equivalent to a qubit of the form
! !
0 θ θ iφ
|ψ i = cos |0i + sin e |1i,
2 2
with two real parameters θ, φ such that 0 ≤ θ ≤ π and −π < φ ≤ π. This allows us to
represent a qubit on the Bloch sphere (with radius one), by a point of colatitude θ and
longitude φ.
It is easy to deduce the formulas for obtaining the coordinates (x, y, z) ∈ R3 of this point:


 x = sin θ cos φ
y = sin θ sin φ
 z = cos θ


The basic quantum state |0i corresponds to the north pole (x, y, z) = (0, 0, 1) with colatitude
θ = 0 (the longitude φ is arbitraty).
12

Figure 1. The Bloch sphere: geometrical representation of a two-level quan-

tum system.
 13

4. Vector spaces
Let K be a field. In this chapter, we will limit ourselvers to the cases K = R or K = C.
A vector space over K is a set V together with two binary operations that satisfy a set of
eight axioms, which are listed below. The elements of V are called vectors, and the elements
of K are called scalars.
The first operation, called addition, assigns to two vectors v and w in V a vector z ∈
V , called the sum of u and v, written as v + w. The second operation, called scalar
multiplication, assigns to any scalar λ in K and to any vector v in V the vector λv ∈ V .
The axioms are the following:
(1) Associativity of vector addition: for any u, v and w in V , u + (v + w) = (u + v) + w.
(2) Commutativity of vector addition: for any u and v in V , u + v = v + u.
(3) Identity element of vector addition: there exists an element 0 ∈ V (the zero vector),
such that v + 0 = v for all v ∈ V .
(4) Inverse elements of vector addition: for any v ∈ V , there exists an element −v ∈ V
(the additive inverse of v), such that v + (−v) = 0.
(5) Compatibility of scalar multiplication with field multiplication: for any λ and µ in
K and for all v ∈ V , λ(µv) = (λµ)v.
(6) Identity element of scalar multiplication: if 1 is the multiplicative identity in K, then
1v = v for all v ∈ V .
(7) Distributivity of scalar multiplication with respect to vector addition: for any λ ∈ K
and for all v and w in V , λ(u + v) = λu + λv.
(8) Distributivity of scalar multiplication with respect to field addition: for any λ and
µ in K and for all v ∈ V , (λ + µ)v = λv + µv.

Among the possible examples of vector spaces, a particularly intersting case consists in
all ordered n-tuples (i.e. sequences of length n) (x1 , x2 , . . . , xn ) such that xi ∈ K for all
i = 1, . . . , n. These n-tuples form a vector space over the field K, usually denoted Kn , called
a coordinate space.

4.1. Basic notions. A column vector with n elements consists of a single column of n
entries, whereas a row vector with n elements consists of a single row of n entries. For
example, a column vector u can be visualized in the following way:
 
x1

 x2 

u=  ..  where x1 , x2 , . . . , xn ∈ K
.
 
 
xn

Vectors of the coordinate space Kn can be added as follows:

     
x1 y1 x1 + y1

 x2 


 y2 


 x2 + y2 

If u =  ..  and v =  .. , then u + v =  .. .
. . .
     
     
xn yn xn + yn
14

The zero vector in the coordinate space Kn is the n-tuple

 
0

 0 

 .. .
.
 
 
0
The scalar multiplication by λ ∈ K, in the coordinate space Kn , is
   
x1 λx1

 x2 


 λx2 

u=  ..  λu =  .. .
. .
   
   
xn λxn
If K = C, the dual vector of a vector u is a vector of the same size, whose coefficients are
the conjugates of those of u, and which is written as a row vector:
 
x1

x2   
u∗ = x∗1 , x∗2 , · · · , x∗n
 
u=  ..  .
.
 
 
xn

Example:
−8i
 
 1−i   
u= u∗ = 8i, 1 + i, 0, 6 − i .
 
0

 
6+i
In the study of finite-dimensional vector space, the concept of basis is crucial.
Definition 10. Let V be a vector space. The vector v ∈ V is said to be a linear combina-
tion of N vectors v1 , v2 . . . , vN ∈ V if there exist N complex numbers a1 , a2 . . . , aN ∈ C
such that
v = a1 v1 + a2 v2 + · · · + aN vN .
Definition 11. A set of vectors v1 , v2 , . . . , vN belonging to a vector space V is linearly
dependent, if there exist a set of scalars a1 , a2 , . . . , aN , not all zero, such that the linear
combination
a1 v1 + a2 v2 + · · · + aN vN = 0,
where 0 is the zero vector.
Clearly, vectors which are not linearly dependent are linearly independent.
Definition 12. A subset B of a vector space V is a basis if its elements are linearly
independent and span the vector space.
It is worth noticing that every vector space has at least one basis. If there is more than
one basis, all bases of the same vector space have the same cardinality. The cardinality of
B is called the dimension of the vector space.
 15

4.2. Qubits as vectors. A 1-qubit is a vector of size 2 whose coefficients are complex
numbers, i.e. !
x1
|ψi = avec x1 , x2 ∈ C,
x2
under the normalization condition |x1 |2 + |x2 |2 = 1.
More generally, an n-qubit is a vector of size 2n ,
 
x1
 . 
|ψi =  .. 

 where x1 , . . . , x2n ∈ C,
x2 n
which satisfies the corresponding normalization condition.
The dual of the vector |ψi is denoted hψ| :
 
hψ| = |ψi∗ = x∗1 , x∗2 , · · · , x∗2n .
! !
1 0
Denote |0i = and |1i = . These two orthogonal states are called the z-basis
0 1
states of a qubit.
In vector notation, the independent basis states of a 2-qubit are the following:
       
1 0 0 0
 0   1   0   0 
|0.0i =  ; |0.1i =  ; |1.0i =  ; |1.1i =  .
       
 0   0   1   0 
0 0 0 1
16

5. Hermitian products and orthogonality

5.1. Hermitian products. The Hermitian product generalizes the standard scalar product
in Rn . It will be denoted by using the bra-ket notation.
Definition 13. Let V be a vector space over the field K = C. A Hermitian product on V
is an operation which associates, to any pair of elements u and v of V , a complex number,
denoted by hu|vi, which satisfies the following properties:
- For all u and v in V , hu|vi = hv|ui∗ .
- Let u, v and w be three elements of V . Then, hu|v + wi = hu|vi + hu|wi.
- Let u and v in V , and λ ∈ C. Then
hλu|vi = λ∗ hu|vi and hu|λvi = λhu|vi.
Remark 4. We use in this note the convention of the physical literature, that is we ask
conjugate linearity in the first variable and linearity in the second one. In the mathematical
literature, the customary convention is the opposite one.
In the coordinate space Cn , the Hermitian product reads as follows. Let
   
x1 y1

 x2 


 y2 

u= ..  and v =  .. .
. .
   
   
xn yn
Then, the Hermitian product of two vectors u and v belonging to Cn , is
n
x∗i yi = x∗1 y1 + x∗2 y2 + · · · + x∗n yn .
X
hu|vi =
i=1
The properties listed in Definition 13 are obviously satisfied.
Remark 5. Let v and w be two vectors of Cn , with only real entries. Then their Hermitian
product (in Cn ) is equal to their scalar product in Rn .

Example: Let u and v be two vectors of C2 . Then

hiu|(−1 + i)vi = −ihu|(−1 + i)vi = −i(−1 + i)hu|vi = (i + 1)hu|vi.
If we suppose that u = (i, −2) and v = (−1, 4 + i), then
hu|vi = −i · (−1) + (−2) · (4 + i) = −8 − i.

Definition 14. The norm of a vector

 
x1
 .  n
 ..  ∈ C
u= 
xn
induced by the Hermitian product, denoted by kuk, is given by
v
q u n
|x |2 .
uX
kuk = hu|ui = t
i
i=1

Remark 6. The norm of a vector u is a non-negative quantity which is zero if and only if
u = 0.
 17

5.2. Orthogonal vectors. The Hermitian scalar product can be used to define orthogo-
nality in Cn . Two vectors u and v in Cn are orthogonal if their Hermitian scalar product is
zero, i.e.
hu|vi = 0.

! !
1 0
Example: |0i = and |1i = are orthogonal qubits.
0 1

Problem: Consider the two qubits |ψi and |ψ 0 i, defined as follows:

   
cos 2θ
! !
θ θ iφ
|ψi ≡ cos |0i + sin e |1i =    
2 2 sin 2θ eiφ
  0 
θ0 θ0 iφ0 cos θ2
! !
|ψ 0 i ≡ cos |0i + sin e |1i =   0 0

2 2 sin θ2 eiφ

such that
0 < θ, θ0 < π et − π < φ, φ0 ≤ π.
Find the conditions on θ, θ0 , φ and φ0 such that these two vectors are orthogonal (with
respect to the Hermitian product).

Answer: We know that qubits are orthogonal if and only if their Hermitian product is equal
to zero. By computing the Hermitian product, we deduce
θ0 θ0 iφ0
! ! ! !
θ θ −iφ
hψ|ψ 0 i = cos · cos + sin e · sin e
2 2 2 2
θ0 θ0 i(φ0 −φ)
! ! ! !
θ θ
= cos cos + sin sin e .
2 2 2 2
0
This quantity is a real number if and only if ei(φ −φ) ∈ R, i.e. if and only if
0
ei(φ −φ) ∈ R ⇐⇒ φ0 − φ ≡ 0 ou φ0 − φ ≡ π.
In the first case, φ0 − φ ≡ 0. Hence φ = φ0 , and
θ0 θ0
! ! ! !
θ θ
hψ|ψ 0 i = 0 ⇐⇒ cos cos + sin sin =0
2 2 2 2
θ θ0
!
⇐⇒ cos − =0
2 2
0
θ θ π
⇐⇒ − =
2 2 2
⇐⇒ θ − θ0 = π.
By reading the statement of the problem, we know that 0 < θ < π and 0 < θ0 < π, hence
−π < θ − θ0 < π. Therefore the first case has no solution. On the other hand, in the second
18

case we have that φ0 − φ ≡ π. Hence, φ0 = φ + π, which implies that

θ0 θ0
! ! ! !
0 θ θ
hψ|ψ i = 0 ⇐⇒ cos cos − sin sin =0
2 2 2 2
θ θ0
!
⇐⇒ cos + =0
2 2
0
θ θ π
⇐⇒ + =
2 2 2
0
⇐⇒ θ = π − θ.
     0
We conclude that |ψi ≡ cos θ
2
|0i + sin θ
2
eiφ |1i is orthogonal to |ψ 0 i ≡ cos θ
2
|0i +
 0
θ iφ0 0 0
sin 2
e |1i if and only if (θ , φ ) = (π − θ, φ + π).
We underline that, in the representation on the Bloch sphere, the qubits |ψ 0 i and |ψi are
symmetric with respect on the origin.
We conclude this section by recalling the Cauchy-Schwarz inequality.
Theorem 15. Let u and v be two vectors of Cn . Then
|hu|vi| ≤ kukkvk.
Remark 7. The structure is the same as in the Euclidean case. Here, | · | denotes the
modulus of a complex number and k · k the norm induced by the Hermitian product.
Remark 8. The notion of orthogonality is essential for defining qubits. Indeed, in the case
of a two-state system, any pair of orthonormal different states (usually labelled |0i and |1i)
must be specified in order to represent a 1-quibit. No other requirements are needed.
 19

6. Tensor product of vectors

The tensor product of two vectors u ∈ Cn and v ∈ Cm is a vector of Cn⊗m . The precise
definition is the following.
Definition 16. Let u ∈ Cn and v ∈ Cm such that
   
x1 y1

x2 
n

y2 
∈ Cm .
   
u=  ..  ∈C and v=  .. 
. .
   
   
xn ym
The tensor product between u and v, denoted u ⊗ v, is the following vector of Cn×m :
   
y1
 .  
 ..  
 x1

x1 y1
    

x1 y2
 ym  
   
   

y1
 .. 

  .  



 . 

 x2  ..   x1 y m
  
 
..
     
y
   
u⊗v =
 m
=
  . .
.. ..

.
   
   . 
..
   
xn y1
   
 .   
 
y1

  
.. 
.
  
   
 x  ..  
   
 n .   xn y m
ym
.
Remark 9. In general, u ⊗ v 6= v ⊗ u.

Example:
 ! 
3 1×3
   
3
! ! 1
1 3 4 !  1×4 4
     

⊗ = = 
= .

2 4 3  2×3 6
   
   

2 
2×4 8
4
Proposition 17. The tensor product is bilinear:
- for any u ∈ Cn , v ∈ Cm and λ ∈ C, we have that (λu) ⊗ v = λ(u ⊗ v) = u ⊗ (λv);
- for any u1 ∈ Cn , u2 ∈ Cn and v ∈ Cm , we have that (u1 + u2 ) ⊗ v = u1 ⊗ v + u2 ⊗ v;
- for any u ∈ Cn , v1 ∈ Cm and v2 ∈ Cm , we have that u ⊗ (v1 + v2 ) = u ⊗ v1 + u ⊗ v2 .
6.1. Product of qubits. Let |φi be a n-qubit and |ψi be a m-qubit. Then, the product of
|φi by |ψi is the tensor product |φi ⊗ |ψi. We underline that |φi ⊗ |ψi is a (n + m)-qubit,
i.e. a vector with 2n × 2m = 2n+m entries.
If we remember the explicit expression of the two basic quantum states in C2 :
! !
1 0
|0i = |1i = ,
0 1
20

we deduce, consequently, that

   
! ! 1 ! ! 0
1 1  0  1 0  1 
|0.0i = |0i ⊗ |0i = ⊗ = |0.1i = |0i ⊗ |1i = ⊗ =
   
0 0 0 0 1 0
 
   
0 0
   
! ! 0 ! ! 0
0 1  0  0 0  0 
|1.0i = |1i ⊗ |0i = ⊗ = |1.1i = |1i ⊗ |1i = ⊗ = .
   
1 0 1 1 1 0

   
0 1

Example: Consider the two qubits

√
1 1 1 3
|φi = √ |0i + √ |1i and |ψi = |0i + |1i.
2 2 2 2
Compute |φi ⊗ |ψi.

We have that
! !! ! √ !!
1 1 1 0 1 1 3 0
|φi ⊗ |ψi = √ +√ ⊗ +
2 0 2 1 2 0 2 1
! √ ! ! !
1 1 31 1 0
= √ ⊗ + √ ⊗
2 2 0 2 2 0 0 1
! ! √ ! !
1 0 1 3 0 0
+ √ ⊗ + √ ⊗
2 2 1 0 2 2 1 1
√ √ √ √
2 6 2 6
= |0.0i + |0.1i + |1.0i + |1.1i.
4 4 4 4
6.2. Quantum entanglement. In quantum mechanics, quantum entanglement is a phe-
nomenon that occurs in a group of particles. The quantum state of a particle cannot be
described independently of the state of the other particles of the group. The same phenom-
enon may occur in quantum computing. We specialize here the definition to the case of an
entrangled 2-qubit.
Definition 18. A 2-qubit |φi is not entangled if there exist two 1-qubits |ψ1 i and |ψ2 i such
that
|φi = |ψ1 i ⊗ |ψ2 i .
If there is no |ψ1 i and |ψ2 i such that |φi = |ψ1 i ⊗ |ψ2 i, then the 2-qubit |φi is entangled.

Example: Consider the following 2-qubits:

1 1 1 1
|φi = |0.0i − |0.1i + |1.0i − |1.1i.
2 2 2 2
Let
1 1 1 1
|ψ1 i = √ |0i + √ |1i |ψ2 i = √ |0i − √ |1i.
2 2 2 2
 21

Hence,
1 1 1 1 1
|ψ1 i ⊗ |ψ2 i = (|0i + |1i) ⊗ (|0i − |1i) = |0.0i − |0.1i + |1.0i − |1.1i = |φi,
2 2 2 2 2
which means that |φi is not entrangled.
The Bell states are quantum states of 2-qubits that represent the maximal examples of
quantum entanglement. The four Bell states are listed below:
1
|Φ+ i = √ (|0.0i + |1.1i)
2
1
|Φ− i = √ (|0.0i − |1.1i)
2
1
|Ψ+ i = √ (|0.1i + |1.0i)
2
1
|Ψ− i = √ (|0.1i − |1.0i.
2
Proposition 19. The Bell states are entrangled.
Proof. We prove the result for |Φ+ i. The proof for the other Bell states is similar. Suppose,
by contradiction that |Φ+ i is not entrangled. Therefore, it would exist |ψ1 i = α|0i + β|1i
et |ψ2 i = α0 |0i + β 0 |1i such that |Φ+ i = |ψ1 i ⊗ |ψ2 i, where α, β, α0 , β 0 are complex numbers.
Hence, we have that
E
Φ+ = |ψ1 i ⊗ |ψ2 i = αα0 |0.0i + αβ 0 |0.1i + βα0 |1.0i + ββ 0 |1.1i.
E 1 1
But we have also that Φ+ = √ |0.0i + √ |1.1i. Consequently, if we identify the various
2 2
terms in the two equations written above, we have that
1 αβ 0 = 0
 
0


 αα = √ 



2 
and
0 1
 ββ = √

 

βα0 = 0.
 

2
The equations on the left imply that α, β, α0 and β 0 are all non-zero, which contradicts the
equations on the right. Thus our starting hypothesis is necessarily false, which implies that
there cannot exist such |ψ1 i and |ψ2 i, i.e. that the qubit |Φ+ i is entrangled. 
22

7. Matrices
In this notes, we assume that the reader has already some familiarity with the set of real
matrices and we will focus on matrices with complex entries.
In what follows, we denote with Mn,p (C) the set of n × p matrices with entries in C and
Mn (C) the set of square matrices with entries in C.
The sum of matrices, the products of a matrix by a scalar and the matrix product are
defined as in the case of the set of real matrices.
Definition 20. Let A = (aij ) and B = (bij ) be two n × p matrices, with 1 ≤ i ≤ n and
1 ≤ j ≤ p. The sum C = A + B is the n × p matrix such that
cij = aij + bij
for all 1 ≤ i ≤ n and 1 ≤ j ≤ p.
Definition 21. Let A ∈ n × p and λ ∈ C. Then the matrix C = λA is the n × p matrix
such that
cij = λaij
for all 1 ≤ i ≤ n and 1 ≤ j ≤ p.
Definition 22. Let A = (aij ) be a n × p matrix and B = (bij ) a p × q matrix. The product
of matrices C = AB is a n × q matix such that
p
X
cij = aik bkj
k=1

where 1 ≤ i ≤ n and 1 ≤ j ≤ q.
The identity matrix is the same as in Mn (R):
 
1 0 ... 0

 0 1 ... 0 

I= .. .. .. .. .

 . . . .


0 0 ... 1
For highlighting the dimension of I, sometimes the notation In is used.
Definition 23. Let A ∈ Mn (C). If there exists a matrix B ∈ Mn (C) such that AB = I and
BA = I, then A is invertible. The matrix B is the inverse of A and is denoted A−1 .
7.1. Determinants. The determinant is a complex number obtained as a function of the
entries of a square matrix. It is very useful for characterizing some properties of the matrix
itself.
Definition 24. Let A ∈ Mn (C), with entries aij . Then its determinant det(A) is given by
the following rule (Leibniz formula):
X n
Y X n
Y
det(A) = sgn(τ ) ai, τ (i) = sgn(σ) aσ(i), i ,
τ ∈Sn i=1 σ∈Sn i=1

where x 7→ sgn(x) is the sign function of permutations in the permutation group Sn , i.e. it
returns 1 or even permutations and −1 for odd permutations.
 23

Remark 10. In the case of a 2 × 2 matrix, the determinant is simple to compute. Let
!
a b
A= ,
c d
Then, det(A) = ad − bc ∈ C.
We have the following theorem, which gives a link between the property of invertibility
of a matrix and its determinant.
Theorem 25. A matrix A ∈ Mn (C) is invertible if and only if det(A) 6= 0.
!
a b
Remark 11. In the case of a 2 × 2 matrix, if A = and ad − bc = det(A) 6= 0,
c d
then A is invertible and !
−1 1 d −b
A = .
ad − bc −c a

Other basic properties of the determinant are the following:

- det(AB) = det(A) det(B);
- det(I) = 1;
- if A is invertible, then det (A−1 ) = 1/ det(A).
7.2. Adjoint matrix. We suppose that the reader is familiar with the notion of transpose
of a matrix: it is an operator which switches the row and column indices of a matrix A and
produces another matrix, denoted by AT :
   
a11 a12 . . . a1p a11 a21 . . . an1

 a21 a22 . . . a2p 
 T

 a12 a22 . . . an2 

A=  .. .. ..  A =  .. .. .
.. 
. . . . . . 
  
  
an1 an2 . . . anp a1p a2p . . . anp
Of course, if the matrix A has n rows and p columns, the tanspose AT has p rows and n
columns.
Definition 26. Let A be a n × p matrix. The adjoint matrix, denoted by A∗ , is the p × n
matrix such that  ∗
a11 a∗21 . . . a∗n1

 a∗ a∗ . . . a∗ 
∗
 12 22 n2 
A =  .. .. .
.. 
 . . . 
a∗1p a∗2p . . . a∗np

Example:  
1−i i !
∗ 1 + i −1 1 − i
A =  −1 2  A = .
 
−i 2 1−i
1+i 1+i
Some properties on adjoint matrices are the following.
Proposition 27. Let A be a n × p matrix and B be a p × n matrix. Let c ∈ C. Then
(1) (A∗ )∗ = A;
24

(2) (A + B)∗ = A∗ + B ∗ ;
(3) (AB)∗ = B ∗ A∗ ;
(4) (cA)∗ = c∗ A∗ ;
(5) If A ∈ Mn (C), then det(A∗ ) = det(A)∗ ;
(6) The matrix A is invertible if and only if A∗ is invertible and (A∗ )−1 = (A−1 )∗ .
Definition 28. A matrix A = (aij ) in Mn (C) is Hermitian if and only if aij = a∗ji for all
indices i and j, i.e. the element in the i-th row and j-th column is equal to the complex
conjugate of the element in the j-th row and i-th column, for all i and j.
7.3. Bra-ket notation for vectors. A column vector x ∈ Cn , in the bra-ket notation, is
denoted |xi and its dual (i.e. the adjoint of |xi) is denoted hx|. Note the parallelism with
the bra-ket notation used for qubits in Section 3. This is due to the fact that qubits are
described by vectors with complex entries (see also Section 4.2).
This notation is very practical. If |xi and |yi are two vectors of Cn , then their (non-
commutative) matrix product gives a scalar, which is nothing but the hermitian product x
and y, i.e. hx|yi, whose notation coincides with that introduced in Subsection 5.1 on the
Hermitian product.
We end this subsection with another property which links the adjoint of a matrix and the
Hermitian product.
Proposition 29. If x and y are two vectors of Cn and A ∈ Mn (C), then hAx|yi = hx|A∗ yi.
7.4. Unitary matrix. An unitary matrix is the compex analogue of an orthogonal matrix
in Rn .
Definition 30. A matrix A ∈ Mn (C) is unitary if and only if A∗ A = I.
Hence, a complex square matrix A is unitary if and only if its ajoint matrix A∗ is also its
inverse. The set of unitary n × n matrices is denoted Un .

Example: The Pauli matrices are a set of three 2 × 2 unitary complex matrices, whose
explicit form is the following:
! ! !
0 1 0 −i 1 0
σ1 = σ2 = σ3 = .
1 0 i 0 0 −1
Remark 12. Sometimes, the Pauli matrices are denoted with the indices x, y and z instead
of 1, 2 and 3, i.e.
σ1 = σx σ2 = σy σ3 = σz .
The following properties of unitary matrices hold.
Proposition 31. If A is an unitary matrix, then hAu|Avi = hu|vi.
Proof. We simply notice that
hAu|Avi = hu|A∗ Avi = hu|vi

We will often deal with 2 × 2 matrices. Consider the matrix
!
a b
A= .
c d
 25

By highlighting the columns of the matrix, A can be written in the following form:
! !
  a b
A= u v avec u= , v= .
c d
The following result gives a necessary and sufficient condition for characterizing an unitary
matrix.
Proposition 32. The matrix A is unitary if and only if the vectors (u, v) form an orthonor-
mal basis, i.e. satisfy the conditions
kuk = 1, kvk = 1 and hu|vi = 0
Proof. We limit ourselvers to consider the 2 × 2 case. We have that
! !
a b ∗ a∗ c ∗
A= A =
c d b∗ d ∗
!
∗ aa∗ + cc∗ ba∗ + dc∗
A A= .
ab∗ + cd∗ bb∗ + dd∗
We suppose that A is unitary. Hence
!
∗ 1 0
A A=I=
0 1
and, consequently,
aa∗ + cc∗ = 1
 






kuk2 = |a|2 + |c|2 = 1

 

bb∗ + dd∗ = 1 hence kvk2 = |b|2 + |d|2 = 1

 

 
a∗ b + c ∗ d = 0 hu|vi = a∗ b + c∗ d = 0.

 


On the other hand, if we suppose that kuk = 1, kvk = 1 and hu|vi = 0, then we recognize
that A∗ A = I. 

Example: It is immediate to see that the column vectors of this matrix are orthogonal and
that their norm is equal to one:
     
θ θ
cos 2
− sin 2
eiλ
U (θ, φ, λ) =  .

    
θ iφ θ i(φ+λ)
sin 2
e cos 2
e
Unitary matrices preserve lengths. This feature characterises unitary matrices:
Proposition 33. Soit A ∈ Mn (C). A is a unitary matrix if and only if kAuk = |uk for any
vector u ∈ Cn .
Another interesting property is the following.
Proposition 34. Let x ∈ Cn and y ∈ Cn be two complex vectors and A ∈ Mn (C) be a
unitary matrix. Then, multiplication by A preserves their inner product:
hAx|Ayi = hx|yi.
This result can be useful.
Lemma 35. The Pauli matrices, together with the identity matrix, form a basis for the real
vector subspace of Hermitian matrices in Mn (C).
26

8. Eigenvectors and eigenfunctions

An eigenvector is a non-zero vector that changes by at most one scalar factor after being
multiplied by a matrix. The corresponding eigenvalue is the scaling factor of the eigenvector.
Definition 36. Let A ∈ Mn (C). The scalar λ ∈ C is an eigenvalue associated to the
eigenvector x if
(1) x is different from the zero vector;
(2) Ax = λx.
The following results hold.
Proposition 37. Let A ∈ Mn (C) and n ≥ 2. If λ is an eignenvalue associated to the
eigenvector x then An x = λn x.
Proof. We prove this result by induction. We first consider the base case, and hence prove
the statement for n = 2. If n = 2, we indeed have that
A2 x = A(Ax) = A(λx) = λ(Ax) = λ(λx) = λ2 x.
We then handle the induction step. We suppose that there exists n ∈ N∗ such that An x =
λn x. Hence,
An+1 x = A(An x) = A(λn x) = λn (Ax) = λn (λx) = λn+1 x.
This proves that the property is true for n + 1. Hence, the proposition is proved. 
Proposition 38. Let A ∈ Mn (C) a unitary matrix. If λ is an eignenvalue of A, then
|λ| = 1.
Proof. Let λ be an eigenvalue associated to the eigenvector x (which is non-zero and con-
sequently kxk 6= 0). Hence, by using all the property previously described, we deduce
that
hAx | Axi = hx | xi =⇒ hλx | λxi = hx | xi
=⇒ λ∗ hx | λxi = hx | xi
=⇒ λ∗ λhx | xi = hx | xi
=⇒ |λ|2 · kxk2 = kxk2
=⇒ |λ|2 = 1
=⇒ |λ| = 1.

Remark 13. By the previous result, is always possible to suppose that λ = e2iπθ . Hence, an
eigenvalue of a given unitary matrix can be individuated by its phase θ.
Each Pauli matrix has two eigenvalues: λ1 = +1 and λ2 = −1. Pairs of corresponding
normalized eigenvectors are the following:
! !
1 1 1 1
ψx+ =√ , ψx− = √ ,
2 1 2 −1
! !
1 1 1 1
ψy+ =√ , ψy− = √ ,
2 i 2 −i
! !
1 0
ψz+ = , ψz− = .
0 1
 27

9. Density matrices for 1-qubits and their applications

9.1. Density matrix for 1-qubits. The density matrix (or density operator) is a funda-
mental tool in quantum mechanics, especially useful for representing quantum systems in
mixed states or statistical ensembles. It generalizes the concept of a quantum state vector
and provides a more flexible framework, capable of describing both pure and mixed quantum
states.

9.2. Pure states. A pure state is a quantum state that can be fully described by a single
state vector |ψi. For a 1-qubit system, the general pure state is:

|ψi = α|0i + β|1i,

where α and β are complex numbers, and |α|2 + |β|2 = 1 to ensure normalization.
The density matrix for a pure state |ψi is given by:

ρ = |ψihψ|,

which, in matrix form, is:

!
|α|2 αβ ∗
ρ= .
α∗ β |β|2
For pure states, the trace of the square of the density matrix satisfies:

tr(ρ2 ) = 1.

Example: Consider a qubit in the state |ψi = √1 (|0i + |1i). The density matrix for this
2
state is:
1 1
 

2 2
ρ = |ψihψ| = 
 .

1 1
2 2

9.3. Mixed states. A mixed state describes a quantum system that is in a probabilistic
mixture of different pure states, rather than being in a single pure state. In such a case, the
density matrix is a weighted sum of the pure state density matrices:
X
ρ= pi |ψi ihψi |,
i

where pi is the probability of the system being in the pure state |ψi i, and
P
i pi = 1. For
mixed states, the trace of the square of the density matrix satisfies:

tr(ρ2 ) < 1.
1
Example: If a qubit is in the state |0i with probability 3
and in the state |1i with probability
2
3
, the density matrix is:
!
1 2 1
0
ρ = |0ih0| + |1ih1| = 3 2 .
3 3 0 3
28

9.4. Applications of density matrices. The density matrix formalism plays a crucial
role in various applications in quantum mechanics, quantum computing, and quantum in-
formation theory. Below are some key applications:
- Describing mixed states. In many real-world quantum systems, particularly in cases
where the system interacts with its environment, the quantum system may not be
in a pure state. The density matrix provides a complete description of such mixed
states. This is particularly useful when the system is in a thermal state or subject
to noise.
- Quantum decoherence. Quantum decoherence describes the process by which a quan-
tum system loses its quantum coherence due to interaction with the environment.
As a result, the system transitions from a pure state to a mixed state. The density
matrix formalism is essential for modeling this phenomenon.
For instance, consider a qubit initially in the superposition state |ψi = √12 (|0i +
|1i). Due to decoherence, the off-diagonal elements of the density matrix decay over
time:
1 −γt 1
 
e

2 2
ρ(t) = 
 ,

−γt 1 1 

e
2 2
where γ is the decoherence rate. As t → ∞, the off-diagonal terms vanish, leaving a
fully mixed state.
- Quantum measurement. In quantum mechanics, measurements are probabilistic,
and the density matrix formalism is used to calculate the probability of different
outcomes. For a system described by a density matrix ρ, the probability of measuring
a state corresponding to a projector P is:
P (outcome) = tr(P ρ).
After the measurement, the density matrix collapses to the state corresponding to
the measurement result.
- Quantum entanglement. In multi-qubit systems, the density matrix plays a vital
role in describing quantum entanglement. For example, for an entangled state such
as the Bell state:
1
|ψi = √ (|00i + |11i),
2
the density matrix for the entire system is:
 
1 0 0 1
10 0 0 0
ρ = |ψihψ| =  .

2 0 0 0 0
1 0 0 1

- Quantum computing algorithms. In quantum computing, especially in noisy inter-

mediate-scale quantum (NISQ) devices, the density matrix is used to describe qubit
states in the presence of noise. During quantum computation, qubits are subject to
noise and decoherence, and the density matrix helps model the imperfect evolution
of quantum states.
 29

For example, after applying a quantum gate, the qubit may not remain in a pure
state due to noise. The density matrix captures the probabilistic nature of the state
and allows tracking of errors throughout the computation.
30

Bibliography
- Lang, Serge. Linear algebra. Third edition. Springer-Verlag, New York, 1987, ISBN
0-387-96412-6
- Rieffel, Eleanor & Polak, Wolfgang. Quantum computing. A gentle introduction,
The MIT Press, Cambridge, Mass. 2011, ISBN 978-0-262-01506-6
- Bodin, Arnaud. Quantum. Un peu de mathématiques pour l’informatique quan-
tique, Exo7, 2021
http://exo7.emath.fr/cours/livre-quantum.pdf

Credits
- Figure 1. Author: Smite-Meister, Own work, CC BY-SA 3.0, Created: 30 January
2009