Scribed Notes

Introduction to Quantum Information and Computation

6/02/2023

1 The Stern-Gerlach Experiment

1.1 The initial experiment:

Figure 1: Left Image shows what we would expect from a classical point of view,
Where as what we would observe is shown by the figure on the right

First, silver (Ag) atoms are heated in an oven. The oven has a small hole
through which some of the silver atoms escape. As shown in fig.1(Left), the
beam goes through a collimator and is then subjected to an inhomogeneous
magnetic field produced by a pair of pole pieces, one of which has a very sharp
edge.
From a classical Point of view, we would expect the atoms to deviate at
angles , depending upon their electorn spin direction, i.e we would expect a
neat spread of atoms over a circular [Link] what was actually observed is
fig.1(Right). The beam splits into 2 , spin up and spin down
The experiment proves that classical physics is just not sufficient to explain
such weird stuff.

1
1.2 Sequential Stern-Gerlach Experiment

Figure 2: a sequential set of 3 stern-Gerlach apparatus

The experiment goes as follows:

1. the beam of atoms is passed through S-G apparatus aligned accross Z-axis
2. the beam splits into 2 One having Sz = +1/2 and Sz = −1/2

3. We block Sz = −1/2 beam

4. only Sz = 1/2 beam is then passed through an S-G apparatus aligned on
the X-axis
5. As we would expect, the beam splits into Sx = −1/2 and Sx = 1/2

6. We again block Sx = −1/2 beam.

7. we pass the Sx = 1/2 beam again through an S-G apparatus along the
Z-axis.
8. What we would expect is that there wouldn’t be any Sz = −1/2 compo-
nent (as we thaught we filtered it out in step 3), But, Counter intuitively,...
what we would observe is that the beam again splits into Sz = +1/2 and
Sz = −1/2

1.3 Cause:
What just happened is that on measuring the Sx , we have caused a change in
its SZ , basically after measuring a property, all the memory belonging to any
other property is lost

2
1.4 similarity with polarisation:

Figure 3: Polarisation of Light

as in fig.3(a), directly passing an x-polarised light through a y-polariser leads

to no beam at the end (if the polarisers are 100% efficient).
But if we add a x’-polariser (fig.3(b))( at an angle of 45deg) with the x and
y-polarisers , we get to see light coming out from y-polariser unlike in fig.3(a)

1.5 takeaways:
1. classical physics is not enough to explail particle/atom level physics.
2. The experiment helped in extablishing the wave particle Duality

2 Hilbert Space(H )
It is a Complex d-dimensional (d≥2 and d ∈ N), along with an inner product.
If x, y ∈ H and the inner Product operator ⟨⟩, Following properties are
satisfied:
1. ⟨x, x⟩ ≥ 0 , Where ⟨x, x⟩ = 0 iff x = 0̄

2. ⟨x, αy1 + y2 ⟩ = α⟨x, y1 ⟩ + ⟨x, y2 ⟩ {Linearity in 2nd argument}

3. ⟨x, y⟩ = ⟨y, x⟩

Bra-Ket Notation
|Ψ⟩ → ket notation ∈ H
⟨Ψ| → bra notation ∈ Dual of H
T
⟨Ψ| = |Ψ⟩† = |Ψ⟩
if ⟨y|x⟩ is a scalar then y is said to belong to dual space of x

3
3 Postulates(Closed systems)
3.1 Postulate 1: The complete description of a physical
system is given by its state |Ψ⟩ ∈ H
Any valid Quantum state has norm 1, i.e ⟨Ψ|Ψ⟩ = 1 where |Ψ⟩ = ⟨Ψ|†
† is called dagger and is defined as the transpose of complex
conjugate of a matrix  
x+iy
Example: if say |Ψ⟩=
a+ib
Then:
 T
† T x − iy  
⟨Ψ| = (|Ψ⟩) = |Ψ⟩ = = x − iy a − ib
a − ib

4
 Eigen-values of Hermitian Operator take real values and
eigenkets corresponding to different eigenvalues are or-
thogonal
let A be a Hermitian operator
let |a1 ⟩ be some eigen-vector of A with eigen-value a1

A|a1 ⟩ = a1 |a1 ⟩ (1)

similarly let |a2 ⟩ be some eigen-vector of A with eigen-value a2

A|a2 ⟩ = a2 |a2 ⟩ (2)

applying † to 2,
⟨a2 |A† = a2 ⟨a2 |
On Multiplying with |a1 ⟩ on right:

⟨a2 |A† |a1 ⟩ = a2 ⟨a2 |a1 ⟩ (3)

On multiplying Eq1 with ⟨a2 | on left:

⟨a2 |A† |a1 ⟩ = a1 ⟨a2 |a1 ⟩ (4)

subtracting Eq3 and Eq4:

(a1 − a2 )⟨a2 |a1 ⟩ = 0 (5)

If we take a1 = a2 , Since ⟨a1 |a1 ⟩:

(a1 − a¯1 ) = 0

=⇒ a1 = a¯1
=⇒ a1 is Real-Valued (6)
Otherwise if we take a1 ̸= a2 =⇒ a1 ̸= a2 :

=⇒ ⟨a2 |a1 ⟩ = 0 (7)

From Eq.6 and Eq.7, It follows that Eigen-values of Hermitian

Operator take real values and eigenkets corresponding to differ-
ent eigenvalues are orthogonal

3.2 Postulate 2: Observables(Ô) are given by Hermitian

Operators which take only real-valued eigenvalues, i.e
Observables are real valued
Defined as an Operator Ô : H → H ,
Where Ô is a Hermitian operator i.e Ô = Ô†

5
 Eigen-values of Ô take real-values
It also follows that(from properties of Hermitian Operators):

⟨x, Ôy⟩ = ⟨Ô† x, y⟩

=⇒ ⟨x, Ôy⟩ = ⟨Ôx, y⟩ Since Ô† = Ô

Ô Has the following spectral decomposition:
X
Ô = λi |ai ⟩⟨ai | ,Where |ai ⟩⟨ai | is a projection operator P(P2 = P)
i

Where |ai ⟩ are elements of orthonormal basis.

On multiplying with |ai ⟩ on right:

Ô|ai ⟩ = λi |ai ⟩⟨ai |ai ⟩ = λi |ai ⟩ , Since all the⟨aj |ai ⟩ = 0

Therefore |ai ⟩ are the eigen-vectors of Ô with corresponding eigen-values λi

P
Note: it can also be written as Ô = i ai |ai ⟩⟨ai | where ai being
the eigen-value corresponding to eigen-vector |ai ⟩

3.3 Postulate 3: Measurement M̂ corresponding to an ob-

servable Ô:
For any valid quantum state |Ψ⟩:

M̂ |Ψ⟩ → |ai ⟩ with outcome λi and probability |⟨ai |Ψ⟩|2

M̂ |ai ⟩ → |ai ⟩ with outcome λi and probability 1

Even if we have identical physical objects with same state |Ψ⟩,

On measurement we may/may not get the same outcome ( since
the outcome maybe any of the λi )
But if we measure the same quantity of same object repeateadly,
we get same output , since M̂ |ai ⟩ → |ai ⟩ with outcome λi

3.4 Postulate 4: Evolution of quantum states is given by

Unitary transformation
Let say initial state be |Ψ⟩ at time 0. After an evolution (at time t) let state
become |Ψ′ ⟩
Then there exists a Linear Transformation such that:

A|Ψ⟩ = |Ψ′ ⟩

Also, On applying †:

⟨Ψ′ | = (|Ψ′ ⟩)† = (A|Ψ⟩)† = ⟨Ψ|A†

6
 For |Ψ′ ⟩ to be Valid quantum state:

⟨Ψ′ |Ψ′ ⟩ = 1

⇒ ⟨Ψ|A† A|Ψ⟩ = 1
⇒ A† A = 1
∴ A is a Unitary Transformation

7
 Note
Say {|i⟩} is Ortho-Normal Basis
P P
1. Tr[ρ] = i ⟨i|ρ|i⟩, and 1 = i |i⟩⟨i|
2. A:HA → HB

X
A= ai,j |i⟩B ⟨j|A
i,j

Where |x⟩A ∈ HA is dual of ⟨j|A , |y⟩B ∈ HB is similar to |i⟩B

If dimension of HA is dA ,j ∈ {0, 1, ..dA − 1}; dA values
If dimension of HB is dB ,j ∈ {0, 1, ..dB − 1}; dB values

• X
1ρ = |i⟩⟨i|ρ
i
X X
⇒ T r[1ρ] = T r[ |i⟩⟨i|ρ] = ⟨i|ρ|i⟩
i i

• AB = A1B

• If ρ2 = ρ ⇒ ρ is Pure
• If ρ2 ̸= ρ ⇒ ρ is Mixed state ⇒ σ =
P
i pi |ψi ⟩⟨ψi |

e.g.

1. 5-|0⟩⟨0|,5-|1⟩⟨1|
⇒ σ = 1/2|0⟩⟨0| + 1/2|1⟩⟨1|
⇒ σ = p1 ρ1 + (1 − p1 )ρ2
Therfore,σ is a mixture of more than one state
√ √
Where |ψi ⟩ = 1/ 2 |0⟩ + 1/ 2 |1⟩,
and |ψi ⟩⟨ψi | can’t be expressed as σ
2. 5-|+⟩⟨+|,5-|−⟩⟨−|
⇒ σ = 1/2|+⟩⟨+| + 1/2|−⟩⟨−|
σ in both 1 and 2 are same in quantum and this σ = 1/2
12∗2
P
Here 1 = i |i⟩⟨i| and σ is a mixed state
⋆ For each observable ∃ a Hermitian operator

For Hermitian Operator,

1. Ô = Ô† i.e. self adjoint

2. Ô:HA → HB ,where H8A ,HB are of same dimension
3. Square matrix
4. Eigen values are real
4 Postulates (Noisy System)
4.1 Tensor Product:
When we have 2 diffrent systems,
 we  take
 thespace  as tensor product H ⊗ H
b1 a1 b1
a1 b2  a1 b2 
   
     b3  a1 b3 
   
a1 b1 
 b1  a2 b1 
 

For example, a2 ⊗ b2 = a2 b2 
     
 = a2 b2 
 
a3 b3  b3  a2 b3 
   

 b1  a3 b1 
 

a3 b2  a3 b2 
b3 a3 b3
And dim(HA ⊗ HB ) = dimA .dimB .
For any observable Â, Expectation  = ⟨ψ| Â|ψ⟩
⟨ψ|ψ⟩

4.2 Postulate 1(Density Operator):

A quantum state is represented by a density operator ρ defined on a hilbert
space H
ρ:H →H
Density Operator has following 3 Properties:
1. ρ ≥ 0{Positive Semidefinite}
2. ρ is hermitian i.e, ρ = ρ† {Linearity in 2nd argument}
3. Tr[ρ]=1 {All eigen values of ρ sums to 1}
Any operator with these 3 properties represent state of any physical system.
In closed system, density operator = |ψ⟩ ⟨ψ|
Some properties

1. Any hermitian operator has spectral decomposition.

X
ρ= λi |ϕi ⟩ ⟨ϕi |
i

Let |i⟩ ∈ H , |i⟩ → ON B and λi ≥ 0.

P P P
Now
P ,Tr[ i λP i |i⟩ ⟨i|] = i λi Tr[|i⟩ ⟨i|] = i λi Tr[⟨i|i⟩] =
i λi Tr[1] = i λi = 1

2. Same Density operator → Same Quantum state.

3. if Tr[ρ2 ] = 1, Then ρ is rank-1, ρ is pure state, ρ = |ψ⟩ ⟨ψ|
4. if Tr[ρ2 ] < 1, Then ρ is mixed state

9
4.3 Postulate 2(Observables: Given by Hermitian opera-
tor)
Expectation value of an observable Ô for a quantum state ρ is Tr[Ôρ].
If ρ represents purePstate ρ = |ψ⟩ P
⟨ψ|, Tr[Ô |ψ⟩ ⟨ψ|] = ⟨ψ| ÔP|ψ⟩.
If ρ is not pure ρ = i λi |ψ⟩ ⟨ψ|, i Tr[λi Ô |ψi ⟩ ⟨ψi |] = λi i Tr[⟨ψi | Ô |ψi ⟩]
= λi ⟨ψi | Ô |ψi ⟩. P
For identity, I = i ⟨i|i⟩ where |i⟩ is ONB.

4.4 Postulate 3
Measurement operators can be represented by Postive operator-valued
measures(POVMs),Λ
• Λ(x) ≥0 ∀ x i.e. positive semidefinite(eigen values are non-negative)
where x is outcome

•
P (x)
xΛ = 1 [like sum of probabilities]
Note that x is index but not power
For a state ρ, probability of outcome x when {Λ(x) }x POVM is performed
p(x)=Tr[Λ(x) ρA ] ⇒ Borne’s rule

4.4.1 Special case: Projective measurement

X
Ô = λi |vi ⟩⟨vi | where λi → Eigen value vi → Eigen vector
i
P
1= i |vi ⟩⟨vi | [Since ONB,Spectral decomposition]

• Each of {|vi ⟩⟨vi |}i forms POVM i.e. Λ(i) ={|vi ⟩⟨vi |}
(or)
P
1= i Px
• Projective operators {Pi }i satisfy

1. P2x = Px
2. Px Py = δxyPx

e.g.
Consider i ∈ {0, 1, 2}, dimension=2

1. P1 =|0⟩⟨0| + |1⟩⟨1|
P2 =|2⟩⟨2|
2. P1 =|0⟩⟨0| + |2⟩⟨2|
P2 =|1⟩⟨1|

10
Both 1,2 are different measurements

11
 Note

• Projective measurement {Pi }i : i Pi =1

P

Projective operators Pi :

1. P2i = Pi
2. Pi Pj = δijPi

• Borne’s rule:
X
p(x) = T r[Λ(x) ρA ]; p(x) = 1
x

Consider X
T r( Λx ρ) = T r[1ρ] = 1
x

X
= T r[Λx ρ]
x

X
= p(x) = 1
x

• Consider a Hermitian operator Ô = Ô†

⇒ eO /T r[eO ] → one of the density operators.

• f:R → R, AA† = A† A
P
⇒ A = i ai |i⟩⟨i|
P
f(A) = i:ai ∈Dom(f ) f (ai )|i⟩⟨i|
Where f is function of operator.
{Pi }i
⇒ ρ −−−→ i, Pi ρP†i /T r[Pi ρP†i ] → Pi ρPi /T r[Pi ρ] → Pi ρPi /p(i)
[Since, P†i = Pi , P2i = Pi , p(i) = Tr[Pi ρ] (where p(i) is probability)]

• Consider observable Ô = i ai |i⟩⟨i|

P

{|i⟩}di=0 → {|i⟩⟨i|}i
• Set of any ONB forms projection.
P P
{|x⟩}x → also ONB i |i⟩⟨i| = 1 = x |x⟩⟨x|
• Eigen vectors depend on observables.

• ρi = Pi ρPi /T r[Pi ρ] ⇒ ρiρj = δijρi ⇒ Tr[ρiρj ] = δij

12
 e.g.
{|0⟩⟨0| + |1⟩⟨1|, |2⟩⟨2|} is a valid projective measurement.
Since, P21 = P1 , P22 = P2 , P1 P2 = 0, x Px = 1
P

1 1
{ (|0⟩⟨0| + |1⟩⟨1|), (|0⟩⟨0| + |2⟩⟨2|)}
2 2
is not a projective measurement but a POVM.
Since, Λx ≥ 0, x Λx = 1
P

4.5 Postulate 4
Transformation/Evolution of Quantum states

• In pure state, evolution is given by unitary operators. U U † = U † U = 1

Quantum Channel N : Anything that transforms quantum state is quantum

channel.

NA→B : B(HA ) → B(HB )

where B(HA ) denotes set of operators.
Properties of Quantum Channel:

• Completely positive
NA→B is positive if :

X ≥ 0 =⇒ NA→B (X) ≥ 0

(We will talk about completeness in next section(near entanglement))

• trace preserving.
State is mapped to state, as trace of state(density operator) is 1, trace is
preserved.

Everything is a quantum channel.

Note: {|i⟩A }i ⊗ {|j⟩B }j can be written as {|i⟩A ⊗ |j⟩b }i,j

5 Composite Systems:
5.1 Product State:
if
ρAB = φA ⊗ σB
Somewhat similar to 2 independent events in probability(P (A∩B) = P (A).P (B))

13
5.2 Seperable States:
if
ρAB = Σx Px φxA ⊗ σB
x

There is a classical correlation between A and B (No quantum specific rela-

tion)
this is simply a mixture of product states

5.3 Entangled State:

1 d−1
ΦAB = Σ |i⟩A ⊗ |i⟩B ⟨j|A ⊗ ⟨j|B
d i,j=0
where d is min(dim(A), dim(B))
This state cannot be decomposed into Seperable form.
Example of such ΦAB :
1
let|Ψ⟩ = √ (|0⟩A ⊗ |0⟩B + |1⟩A ⊗ |1⟩B )
2
then |Ψ⟩⟨Ψ| is a Φ of d=2
Notice that Φ is a Pure state

6 Complete Positivity:
NA→C is a Completely positive operator if

(id)B ⊗ NA→C (ΦAB ) ≥ 0∀ΦAB ≥ 0

Where (id)B is called a super operator and applied on NA→C to nullify the
affect of NA→C on state of B
NA→C Operation can be written as:

NA→C (XA ) = Σi Ki XA Ki†

Where the set of operators {Ki } (Ki :HA → HC ) are called Kraus operators
and they satisfy the property ΣKi† Ki = 1
We could call projection operators ({Pi }) as a special case of Kraus Opera-
tors ({Ki })

7 Quantum Channel:
• Trace Preserving: Tr[N (X)] = Tr[X] ∀X ∈ B(XA )
(where B(XA ) denotes set of operators with finite trace)

• Completely positive: NA→B (ΦRA ) ⩾ 0

14
 1
ΦRA = |Φ⟩⟨ Φ|RA = Σi, j | i⟩R ⊗ |i⟩A ⟨ j|R ⊗ ⟨j|A
d
d−1
1 X
|Φ⟩RA = √ |i⟩R ⊗ |i⟩A
d i=0
d−1
X √
|ψ⟩RA = pi |φ⟩R ⊗ |ϕ⟩A
i=0
where
{|φ⟩R } denotes the orthonormal basis in R
{|ϕ⟩A } denotes the orthonormal basis in A

8 Schmidt Decomposition and Purification

The vast array of techniques used for studying composite quantum systems,
which are at the heart of quantum computation and quantum information, in-
cludes density operators and partial traces. The ”Schmidt Decomposition and
Purifications” are another set of two more extremely valuable instruments.

8.1 Schmidt Decomposition

Schmidt decomposition is a mathematical tool in quantum information theory
that represents a pure quantum state as a linear combination of orthogonal
states. It is used to quantify the entanglement between two subsystems of a
larger quantum system. Any pure state of a bipartite quantum system can be
expressed in terms of a set of orthogonal states that are shared by the two
subsystems. These orthogonal states are known as the Schmidt states, and the
coefficients in the linear combination that represents the state are known as
the Schmidt coefficients. The Schmidt decomposition provides a measure of the
entanglement between the two subsystems, as the number of non-zero Schmidt
coefficients is equal to the dimension of the smallest [Link] Schmidt
decomposition is important in the field of quantum information processing be-
cause it provides a way to quantify and manipulate the entanglement between
quantum systems. It has applications in quantum communication and quantum
computation, as well as in the study of quantum phase transitions and quantum
field theory.
Given a pure state |ψ⟩ of a composite system AB, it can be expressed in
terms of orthogonal states |iA ⟩ for system A and |iB ⟩ for system B:
X
|ψ⟩ = λi |iA ⟩|iB ⟩ (8)
i

where λi are the non-negative real numbers called Schmidt coefficients and
X
(λi )2 = 1 (9)
i

15
 Given a pure state |ψ⟩ of a composite system AB, then by the Schmidt
Decomposition is:
X X
ρA = (λi )2 |iA ⟩⟨iA |, ρB = (λi )2 |iB ⟩⟨iB | (10)
i i

so the eigenvalues of (ρA ) and (ρB ) are identical, namely (λi )2 for both density
operators. Many important properties of quantum systems are completely de-
termined by the eigenvalues of the reduced density operator of the system, so
for a pure state of a composite system such properties will be the same for both
systems.

8.2 Purification
Purification on the other hand is a mathematical procedure which allows us to
associate pure states with mixed states. The close relationship of the Schmidt
decomposition to purification is the procedure used to purify a mixed state of
system A is to define a pure state whose Schmidt basis for system A is just the
basis in which the mixed state is diagonal, with the Schmidt coefficients being
the square root of the eigenvalues of the density operator being purified.

9 Partial Tracing
Partial tracing in quantum mechanics refers to the process of tracing out or
eliminating the degrees of freedom of a subsystem in a quantum system. The
reduced density matrix of the remaining subsystem is obtained after partial
tracing. This operation allows for a simplified description of the state of a
subsystem of interest, while ignoring the other degrees of freedom.
Partial tracing is a key concept in quantum information and quantum com-
putation, where it is used to obtain reduced density matrices of composite quan-
tum systems, study entanglement, and analyze the performance of quantum
algorithms and protocols. It is also used in quantum field theory, quantum
thermodynamics, and other areas of quantum physics.
Here are some examples to help illustrate partial tracing in quantum com-
puting: Bell state: Consider the Bell state, which is a maximally entangled
two-qubit state given by |ψ⟩ = (|00⟩ + |11⟩), If we trace out one of the qubits,
we obtain a reduced density matrix for the other qubit. For example, if we trace
out qubit 1, we have:
ρ1 = T r(|ψ⟩ ⟨ψ|) = T r(|00⟩ ⟨00| + |11⟩ ⟨11| + |00⟩ ⟨11| + |11⟩ ⟨00|) = |0⟩ ⟨0|
Which is a pure state. This means that the remaining qubit is in a definite
state, even though the two-qubit system was entangled.
Quantum state of a qubit system: Consider a quantum system composed of
two qubits with state,
ψ = α |00⟩ + β |01⟩ + γ |10⟩ + δ |11⟩
If we trace out qubit 2, we obtain a reduced density matrix for qubit 1:

16
 ρ1 = T r(|ψ⟩ ⟨ψ|) = T r(α |00⟩ ⟨00| + β |11⟩ ⟨11| + γ |00⟩ ⟨11| + δ |11⟩ ⟨00|) =
α |0⟩ ⟨0| + γ |1⟩ ⟨1|
This shows that partial tracing can simplify the description of a complex
quantum system by eliminating the influence of one of the subsystems.

Team Members
• Shravya Kukkapalli - 2021101051
• Siya Puttagunta - 2021101062

• Bocha venkata krishna sai anirudh- 2021101003

• Abhinav Reddy Boddu - 2021101034
• Santhoshini thota - 2021101097
• Chitturi Meghana Sai Rama Lakshmi - 2021101006

• Vishna Panyala - 2021101044

• Aishani Pandey- 2022121009

17