Quantum Computing Notes: M-2

Pauli Operators
Pauli's operators are the three fundamental 2x2 Hermitian and unitary matrices - X, Y, Z that
act as a single qubit quantum gate in quantum computing.
• Pauli X (σx = X ): (0 1)
1 0
​ ​ ​

• Pauli Y (σy = Y ): (0i −i

)
0
​ ​ ​

• Pauli Z (σz = Z ): (10 0

)
−1
​ ​ ​

Properties
All Pauli operators square to identity (X 2 = Y 2 = Z 2 = I ), have eigenvalues ±1, and
anticommute pairs. XY = iZ (cyclic permutation).
Example
For ∣ψ⟩ = (αβ): ​

• X∣ψ⟩ = (β )
α
​

• Y ∣ψ⟩ = (−iβ )
iα
​

• Z∣ψ⟩ = ( α )
−β
​

Outer Product
The outer product ∣ψ⟩⟨ϕ∣ of states ∣ψ⟩ and ⟨ϕ∣ yields a matrix that maps any ∣χ⟩ to the
scaled projection ⟨ϕ∣χ⟩∣ψ⟩.
Properties
• Applies as (∣ψ⟩⟨ϕ∣)∣χ⟩ = ∣ψ⟩⟨ϕ∣χ⟩
• Hermitian if ∣ψ⟩ = ∣ϕ⟩: ∣ψ⟩⟨ψ∣ is a projector (P 2 = P , eigenvalue 0 or 1).
 • Trace: T r(∣ψ⟩⟨ϕ∣) = ⟨ϕ∣ψ⟩
Example
For ∣+⟩ = 1
2
​

​
(∣0⟩ + ∣1⟩) , the projector onto ∣+⟩ is:
1 1
P+ = ∣+⟩⟨+∣ = 1
( )
2 1 1
​ ​ ​ ​

Closure Relation
The closure relation is also known as the completeness relation or resolution of the
identity. It states that an orthonormal basis spans the entire Hilbert space, summing to the
identity operator.
For a discrete orthonormal basis {∣n⟩} with ⟨n∣m⟩ = δnm : ​

∑n ∣n⟩⟨n∣ = I
​

where I is the identity operator.

Derivation of Closure Relation from Completeness
Take any ∣ψ⟩, insert completeness: ∣ψ⟩ = ∑n ∣n⟩⟨n∣ψ⟩. ​

Apply ⟨ϕ∣ from left: ⟨ϕ∣ψ⟩ = ∑n⟨ϕ∣n⟩⟨n∣ψ⟩. ​

Since this holds for all ∣ψ⟩, ∣ϕ⟩, the operator I = ∑n ∣n⟩⟨n∣ is defined. ​

Matrix Elements
Pauli Operator Example
Pauli-X in {∣0⟩, ∣1⟩} basis:
⟨0∣X∣0⟩ ⟨0∣X∣1⟩ 0 1
Xij = ⟨i∣X∣j⟩ ⇒ X = ( )=( )
⟨1∣X∣0⟩ ⟨1∣X∣1⟩ 1 0
​ ​ ​ ​ ​

General Construction
Any operator A^ = ∑ij Aij ∣i⟩⟨j∣. ​ ​

Hermitian operators have Aij = A∗ji , yielding real eigenvalues.

​ ​

Diagonal Form
In eigenbasis {∣uk ⟩}, A^ = ∑k λk ∣uk ⟩⟨uk ∣. The matrix is diagonal with eigenvalues λk .
​ ​ ​ ​ ​ ​

Outer Product Matrix

For ket ∣ψ⟩ (column) and bra ⟨ϕ∣ (row), ∣ψ⟩⟨ϕ∣ is the matrix whose (i, j)-th entry is ψi ϕ∗j ,
equivalent to column times row vector multiplication.
​ ​

Operator Decomposition
In orthonormal basis {∣i⟩}, operator A^ = ∑ij Aij ∣i⟩⟨j∣, where matrix element
Aij = ⟨i∣A^∣j⟩.
​ ​

Qubit Examples
• Pauli X: X = ∣0⟩⟨1∣ + ∣1⟩⟨0∣ = (01 1
)
0
​ ​

• Projector ∣0⟩⟨0∣ : (10 0

)
0
​ ​

• Hadamard H: H = 1
(∣0⟩⟨0∣ + ∣0⟩⟨1∣ + ∣1⟩⟨0∣ − ∣1⟩⟨1∣) = 1
(
1 1
)
2 2 1 −1
​ ​ ​ ​

​ ​

Relation between Outer Product and Tensor Product

Outer product and tensor product are closely related, but they live in slightly different
"levels" of structure. The outer product is a concrete matrix construction, while the tensor
product is a more abstract vector space operation. In quantum mechanics, the outer
product of state vectors can be viewed as the "matrix realization" of a particular tensor.

Matrix Representation of Operators in 2D Space

Let A^ be a linear operator on a 2D space and let {∣u1 ⟩, ∣u2 ⟩} be an orthonormal basis. The
matrix representation of A^ in this basis is:
​ ​

⟨u ∣A^∣u ⟩ ⟨u1 ∣A^∣u2 ⟩

A=( 1 ^ 1 )
​ ​ ​

⟨u2 ∣A∣u1 ⟩ ⟨u2 ∣A^∣u2 ⟩

​

​ ​ ​

The matrix element Aij is the coefficient you get when you apply A^ to basis vector ∣j⟩ and
project the result onto ∣i⟩.
​
 • Identity I^: I^∣0⟩ = ∣0⟩, I^∣1⟩ = ∣1⟩ ⇒ I = (10 0
)
1
​ ​

• Pauli X: X∣0⟩ = ∣1⟩, X∣1⟩ = ∣0⟩ ⇒ X = (01 1

)
0
​ ​

• Pauli Z: Z∣0⟩ = ∣0⟩, Z∣1⟩ = −∣1⟩ ⇒ Z = (10 0

)
−1
​ ​

• Hadamard H: H∣0⟩ = ∣0⟩+∣1⟩ , H∣1⟩ = ∣0⟩−∣1⟩

⇒H = 1
(
1 1
)
2 2 2 1 −1
​ ​ ​ ​ ​

​ ​ ​

Example: Operator Action

An operator A^ acts on a 2D Hilbert space with orthonormal basis {∣0⟩, ∣1⟩} as follows:
A^∣0⟩ = 2∣0⟩ + ∣1⟩, A^∣1⟩ = −∣0⟩ + 3∣1⟩
1. Find the 2x2 matrix representation of A in the basis.
2. Compute A^∣ψ⟩ given ∣ψ⟩ = 0.6∣0⟩ + 0.8∣1⟩.
Solution:
1. Matrix A is defined by elements Aij = ⟨i∣A^∣j⟩. ​

• First column: A^∣0⟩ = (2)

1
​

• Second column: A^∣1⟩ = (−1 )

3
​

2 −1
A=( )
1 3
​ ​

2. ∣ψ⟩ = (0.6)
0.8
​

2 −1 0.6 2(0.6) − 1(0.8) 0.4

A^∣ψ⟩ = ( )( ) = ( )=( )
1 3 0.8 1(0.6) + 3(0.8) 3.0
​ ​ ​ ​ ​

So, A^∣ψ ⟩ = 0.4∣0⟩ + 3.0∣1⟩ .

Pauli's Matrix
The Pauli matrices are three 2x2 complex matrices denoted by σx , σy , σz . ​ ​ ​

0 1 0 −i 1 0
σx = ( ) , σy = ( ) , σz = ( )
1 0 i 0 0 −1
​ ​ ​ ​ ​ ​ ​ ​ ​
Key Properties
• Hermitian: σj† = σj (they are observables) ​ ​

• Unitary: σj σj† = I ​ ​

• Involutory: σj2 = I ​

• Traceless: T r(σj ) = 0 ​

Example: Compute Eigenvalues & Vectors for σx ​

Solution:
Eigenvalue: det(σx − λI ) = 0 ⇒ det (−λ 1
1 −λ
) = λ2 − 1 = 0 ⇒ λ = ±1 ​ ​ ​

Eigenvector for λ = 1:
σx ( ) = ( ) = ( ) ⇒ a = b. Normalize:
a b a
b a b
​ ​ ​ ​

a + b = 1 ⇒ 2a = 1 ⇒ a = b = 1/ 2.
2 2 2
​

1
∣+⟩ = 1
( )
2 1
​ ​

Eigenvector for λ = −1:

b = −a, 2a2 = 1 ⇒ a = 1/ 2, b = −1/ 2 ​ .
​

1
∣−⟩ = 12 ( )
−1
​ ​

Given Qubit State ∣ψ⟩ = 0.8∣0⟩ + 0.6∣1⟩

Compute ⟨ψ∣σx ∣ψ⟩ and ⟨ψ∣σz ∣ψ⟩. ​ ​

0 1 0.8 0.6
σx ∣ψ⟩ = ( )( ) = ( )
1 0 0.6 0.8
​ ​ ​ ​ ​

0.6
⟨ψ∣σx ∣ψ⟩ = (0.8, 0.6) ( ) = 0.48 + 0.48 = 0.96
0.8
​ ​

1 0 0.8 0.8
σz ∣ψ⟩ = ( )( ) = ( )
0 −1 0.6 −0.6
​ ​ ​ ​ ​

0.8
⟨ψ∣σz ∣ψ⟩ = (0.8, 0.6) ( ) = 0.64 − 0.36 = 0.28
−0.6
​ ​

Hermitian Operator
An operator A^ is Hermitian if it equals its adjoint: A^† = A^.
Equivalently, ⟨ϕ∣A^∣ψ⟩ = ⟨ψ∣A^∣ϕ⟩∗ .
Unitary Operator
An operator U^ is unitary if its inverse is equal to its adjoint: U^ †U^ = U^ U^ † = I .
Key Properties of Hermitian
• Real Eigenvalues: All eigenvalues of a Hermitian operator are real.
• Physical Observable: In quantum mechanics, an observable must be Hermitian;
measurement outcomes are real numbers.
• Orthonormal Eigenvalues: There exists an orthonormal basis of eigenvectors; A^ is
diagonalizable. A^ = UDU †.
• Expectation values are real: For any state ∣ψ⟩, ⟨ψ ∣A^∣ψ ⟩ ∈ R.

Key Properties of Unitary

• Norm-preserving: For any state ∣ψ⟩, ∣∣U^ ∣ψ⟩∣∣ = ∣∣∣ψ⟩∣∣.
• Inner product preserving: ⟨U^ ϕ∣U^ ψ ⟩ = ⟨ϕ∣ψ⟩.
• Eigenvalues on the unit circle: All eigenvalues λ satisfy ∣λ∣ = 1.
• Column form an orthonormal set: In any basis, the matrix columns of a unitary
operator are orthonormal.

Normal Operator
A normal operator is a linear operator on a complex inner-product space that commutes
with its own adjoint.
An operator N^ on a Hilbert space is called normal if it satisfies N^ †N^ = N^ N^ †.
Key Properties
1. Generalize both Hermitian and Unitary operators.
• If N^ is Hermitian (N^ † = N^ ), then clearly N^ † N^ = N^ N^ † .
• If N^ is Unitary (N^ † N^ = N^ N^ † = I ), then it is also normal.
 2. Spectral Theorem applies: A normal operator has an orthonormal basis of
eigenvectors, so it is diagonalizable: N^ = UDU †, where U is unitary and D is diagonal.
3. Eigenvectors from different eigenvalues are orthogonal: If N^ ∣ψ⟩ = λ∣ψ⟩ and
N^ ∣ϕ⟩ = μ∣ϕ⟩ with λ =  μ, then ⟨ϕ∣ψ⟩ = 0.
4. Normal comparison for any vector ∣ψ⟩.

Eigenvectors
In linear algebra, an eigenvector of a linear operator A^ is a non-zero vector ∣ψ⟩ that, when
A^ acts on it, yields a scalar multiple of itself: A^∣ψ ⟩ = λ∣ψ⟩.
Where the scalar λ is the eigenvalue. This equation is called the eigenvalue equation. It
characterizes the operator's intrinsic behavior, such as scaling and direction invariance. For
a 2D operator, eigenvalue solves the characteristic equation det(A^ − λI ) = 0, and
eigenvectors are found by plugging λ back into the equation and solving for the
components.
Significance in QC
In QC, Hermitian operators represent physical observables; their eigenvalues are the
possible measurement outcomes (always real), while eigenvectors represent stationary
states with definite observable values. Measurement collapses the system into an
eigenvector.

Spectral Decomposition
The spectral decomposition theorem is a central result in linear algebra and quantum
mechanics. It says that certain operators can be "diagonalized" in a very clean, basis-
independent way using their eigenvalues and eigenvectors.
Spectral Decomposition Theorem says:
1. There exists an orthonormal basis of eigenvectors of A^.
2. In that basis, A^ becomes diagonal with eigenvalues on the diagonal.
3. In Dirac language, A^ can be written as: A^ = ∑k λk ∣uk ⟩⟨uk ∣
​ ​ ​ ​

Where λk is the eigenvalue and ∣uk ⟩ are the corresponding orthonormal eigenvectors.
​ ​

In finite dimensional terms:

Let A be an n × n Hermitian matrix, then:
1. There exists a unitary matrix U whose columns are the orthonormal eigenvectors of A.
2. There exists a diagonal matrix Λ whose entries are the eigenvalues λk . ​

3. The spectral decomposition: A = UΛU †.

Spectral Theorem for Hermitian Matrices

Step 1: Eigenvalues are real
Let A = A† be Hermitian, and let λ be an eigenvalue with eigenvector ∣ψ⟩ = 0.
A∣ψ⟩ = λ∣ψ⟩
Multiply both sides from the left by ⟨ψ∣:
⟨ψ∣A∣ψ⟩ = λ⟨ψ∣ψ⟩
Take the conjugate transpose of both sides:
ˉ ⟨ψ∣ψ⟩
⟨ψ∣A† ∣ψ⟩ = λ
But A† = A, so ⟨ψ∣A∣ψ⟩ = λˉ⟨ψ∣ψ⟩.
Since ⟨ψ∣ψ⟩ > 0, λ = λˉ ⇒ λ ∈ R.
Step 2: Eigenvectors from different eigenvalues are orthogonal
Let ∣ψ⟩, ∣ϕ⟩ be eigenvectors with values λ, μ ∈ R and λ = μ.
A∣ψ⟩ = λ∣ψ⟩, A∣ϕ⟩ = μ∣ϕ⟩
Compute ⟨ϕ∣A∣ψ⟩ in two ways:
Using the left equation: ⟨ϕ∣A∣ψ⟩ = λ⟨ϕ∣ψ⟩
Using the right equation: ⟨ϕ∣A∣ψ⟩ = ⟨Aϕ∣ψ⟩ = μ⟨ϕ∣ψ⟩ (Since A† = A and μ ∈ R)
So, λ⟨ϕ∣ψ⟩ = μ⟨ϕ∣ψ⟩ ⇒ (λ − μ)⟨ϕ∣ψ⟩ = 0.
Because λ = μ, we must have ⟨ϕ∣ψ⟩ = 0.

Trace of an Operator
For a linear operator A^ on a finite Hilbert space, the trace of A^, denoted by T r(A^), is the
sum of its diagonal matrix elements in any orthonormal basis.
T r(A^) = ∑i ⟨i∣A^∣i⟩
​

where {∣i⟩} is an orthonormal basis.

Equivalently, if A is the matrix representation of A^, then T r(A) = ∑i Aii . ​

In the spectral theorem language, for a diagonalizable operator with eigenvalues λi ,

T r(A^) = ∑i λi .
​

​ ​

Important Properties
1. Linearity: T r(αA^ + β B^ ) = αT r(A^) + βT r(B^ )
2. Cyclic: T r(A^B^ ) = T r(B^ A^)
3. Basis independent: T r(S −1 A^S) = T r(A^)
4. Trace of adjoint and transpose: T r(A^†) = T r(A^), T r(AT ) = T r(A)
​

5. Trace of identity & projection:

• In dimension d: T r(I ) = d
• For rank-1 projector ∣ψ⟩⟨ψ∣ (normalized): T r(∣ψ⟩⟨ψ∣) = 1
6. Connection to Eigenvalue

Expectation Value of an Operator

In quantum mechanics, the expectation value of an operator A^ is the average result of
many measurements of the corresponding observable on a quantum system in a given
state.
Definition
For a pure state ∣ψ⟩, the expectation value of an operator A^ is:
⟨A^⟩ = ⟨ψ∣A^∣ψ⟩
In bra-ket notation, this is the inner product of ∣ψ⟩ with A^∣ψ⟩, yielding a scalar.
For a mixed state described by a density operator ρ, the expectation value is:
⟨A^⟩ = T r(ρA^)

Projection Operator
In linear algebra and quantum computing, a projection operator is a linear operator P that
"projects" vectors into a subspace in the same way a shadow is cast.
Key Properties
 1. P maps the whole space into a subspace and leaves that subspace invariant (P 2 = P ).
2. Eigenvalues of an orthogonal projector are only 0 or 1.
3. If P^ is a rank-1 orthogonal projector onto a normalized vector ∣ψ⟩, then P^ = ∣ψ⟩⟨ψ∣.

Positive Operator
A positive operator is an operator that acts like a non-negative number in the sense that all
its expectation values are non-negative.
An operator A^ is positive if for any state ∣ψ⟩, ⟨ψ∣A^∣ψ⟩ ≥ 0.
Positive operators are necessarily Hermitian.