Linear Operators in Quantum Mechanics
Atanu Kumar
Assistant Professor
Department of Physics
Chandernagore College
August 29, 2020
Contents
I Space of Wave Functions
I Hermitian and Anti Hermitian Operators
I Expectation Value and Uncertainty
I Eigenvalues and Eigenvectors
I Degeneracy
Space of Wave Functions
I In Quantum Mechanics, measurable physical quantities, called
observables or dynamic variables are represented by linear
operators, which are self adjoint or Hermitian.
I That operators act on the complex vector space of square
integrable functions I(2) .
R∞
I If the integration −∞ |f (x)|2 dx has a finite value then the
function f (x) is said square integrable. Here for simplicity
we assume time independence and single degree of freedom.
I Inner Product of two functions ψ(x) and φ(x) in I(2) are
defined as
Z ∞
hψ, φi = ψ ∗ (x)φ(x)dx.
−∞
I Each of such functions represents a quantum mechanical state
of a system and known as its wave functions. A normalized
wave function means hψ, ψi = 1.
Hermitian and Anti Hermitian Operators
I For each operator  on I(2) , there exists an another operator
† on I(2) , known as Adjoint of Â, defined by
D E D E
† φ, ψ = φ, Âψ , ∀φ, ψ ∈ I(2) . (1)
I If † = Â, then  is said to be Self Adjoint or Hermitian.
So for a hermitian operator
D E D E
Âφ, ψ = φ, Âψ (2)
I If † = −Â, then  is called Anti Hermitian.
I The position operator x̂, defined by x̂ψ(x) = xφ(x), ∀φ ∈ I(2)
is a hermitian operator since
Z ∞ Z ∞
∗
xψ (x)φ(x)dx = ψ ∗ (x)xφ(x)dx,
−∞ −∞
⇒ hx̂ψ, φi = hψ, x̂φi . (3)
I The operator d/dx is anti hermitian since
Z ∞ Z ∞ Z ∞
dψ ∗ ∗ ∞ ∗ dφ dφ
φdx = [ψ φ]−∞ − ψ dx = − ψ ∗ dx,
−∞ dx −∞ dx −∞ dx
where the square integrable functions are assumed to vanish
d
at infinities. The momentum operator, defined by p̂ = −i~ dx
is hermitian since
dψ ∗
Z ∞ Z ∞
dφ
(−i)~ φdx = ψ ∗ (−i)~ dx. (4)
−∞ dx −∞ dx
Solved Problems
Show that (ÂB̂)† = B̂ † † .
Let φ, ψ ∈ I(2) . Since B̂ψ ∈ I(2) and † is adjoint to Â,
D E D E
† φ, B̂ψ = φ, ÂB̂ψ . (5)
Similarly since † φ ∈ I(2) and B̂ † is adjoint to B̂
D E D E
B̂ † † φ, ψ = † φ, B̂ψ . (6)
D E D E
Comparing (5) and (6) we obtain B̂ † † φ, ψ = φ, ÂB̂ψ . So
(ÂB̂)† = B̂ † † (7)
Show that [Â, B̂]† = [B̂ † † ].
For any ψ ∈ I(2) , [Â, B̂]ψ = ÂB̂ψ − B̂ Âψ.
D E D E D E
φ, [Â, B̂]ψ = φ, ÂB̂ψ − φ, B̂ Âψ
D E D E
= B̂ † † φ, ψ − † B̂ † φ, ψ
D E
= [B̂ † † ]φ, ψ
So [Â, B̂]† = [B̂ † † ].
Theorem
If Â, B̂ are hermitian, then ÂB̂ is not necessarily hermitian, since
(ÂB̂)† = B̂ Â. However if  and B̂ commute then ÂB̂ is hermitian.
Expectation Value and Uncertainty
I Expectation Value or mean value of a hermitian operator Â
in a normalized state ψ is given by
D E Z ∞
Āψ = ψ, Âψ = ψ ∗ Âψdx (8)
−∞
I Deviation from Mean of  in the state ψ is ∆ =  − ĀIˆ.
The Mean Square Deviation
2
(∆A)2ψ = ψ, ∆ ψ
represents the uncertainty of  in the state ψ.
I Since  is hermitian, then ∆ is also hermitian and (∆A)2 is
positive definite.
From the Schwarz Inequality
||φ||||ψ|| ≥ | hφ, ψi |, for any two vectors,
it can be shown that for any two operators  and B̂ and for any
state ψ, the following inequality holds:
iE
1 Dh
(∆A)ψ (∆B)ψ ≥ Â, B̂ .
2 ψ
Using the commutator of [x̂, p̂] = i~, we obtain the Uncertainty
Principle for the position and momentum operators:
~
∆x∆p ≥ .
2
Eigenvalues and Eigenvectors
Let us consider some states ψn (x), such that ∆Âψn = 0, i.e. the
observable corresponding to  can be exactly measured in that
state. Let us denote the expectation value of  in that state be
Ā = an So
∆Âψn =  − an Iˆ ψn = 0, ⇒ Âψn = an ψn . (9)
(9) is known as the Eigenvalue Equation for the operator Â,
where an is the Eigenvalue and ψn is the Eigenfunction,
corresponding to an . In general  may have several eigenvalues an
with corresponding eingenfunctions ψn .
Theorem
Eigenvalues of Hermitian Operators are real.
D E D E
ψn , Âψn = Âψn , ψn ⇒ hψn , an ψn i = han ψn , ψn i ⇒ an∗ = an .
Theorem
Eigenfunctions belonging to different eigenvalues of hermitian
operator are orthogonal.
D E D E
ψm , Âψn = Âψm , ψn ⇒ hψm , an ψn i = ham ψm , ψn i
⇒ an hψm , ψn i = am hψm , ψn i .
Since am 6= an , this implies hψm , ψn i = 0.
Example
d 2
Consider the operator  = dx 2 on the space of functions defined
over (−L, L). Eigenvaule Equations are:
d2 nπ 2 nπx
ψ n = an ψ n ⇒ an = − , ψn = sin .
dx 2 L L
So an are real. ψn are orthogonal, since
Z L mπx nπx
hψm , ψn i = sin sin = 0.
−L L L
For a particular eigenvalue an , there exists several eigenfunctions:
nπx nπx nπx nπx nπx
sin , sin + cos , sin − cos , ...
L L L L L
These are not all orthogonal. It can be shown that the
eigenfunctions corresponding to the eigenvalue an form a vector
subspace of dimension 2 andwe can choose two orthogonal basis
vector in that space such as sin nπx nπx
L , cos L .
Degeneracy
I If an eigenvalue an has several eigenfunctions corresponding to
it, that eigenvalue is said to be Degenerate and
corresponding states are called Degenerate States.
I If the total number of linearly independent degenerate states
corresponding to an eigenvalue an is g , then the eigenvalue is
said to be g -fold degenerate.
nπ 2
I In the previous examples, an = − L are 2-fold degenerate.
Since the eigenfunctions corresponding to different eigenvalues of a
hermitian operator  are orthonormal and hence linearly
independent, and we can choose orthonormal eigenfunctions
corresponding to same eigenvalue, all linearly independent
eigenfunctions of a hermitian operator span a vector space. Any
state ψ in that space can be written as a linear combination of the
eigenfunctions {ψn,l }, where the index l stands for the degenerate
states.
XX XX
ψ= cn,l ψn,l , |cn,l |2 = 1.
n l n l
Expectation value of  in the state ψ is found to be
D E XX
Āψ = ψ, Âψ = an |cn,l |2 .
n l
|cn,l |2 represent the probability of finding the system in the state
ψn,l , provided it is in the state ψ.
Hydrogen Atom
I States of an electron in Hydrogen atom form
a vector space with a set of basis vectors
{ψn,l,m (r , θ, φ)}, that are simultaneous
eigenfunctions of the Hamiltonian operator
Ĥ, Angular Momentum Square L̂2 and the Z
component of Angular Momentum L̂z .
2 2
Z ~ 1
I Eigenvalues of hamiltonian is En = − 2ma 2 n2 , whereas
o
eigenvalues of L̂2 and L̂z are l(l + 1)~ and m~ respectively.
I Eigenvalues of L̂2 are 2l + 1 fold degenerate, and eigenvalues
Pn−1
of Ĥ are l=0 (2l + 1) fold degenerate.
I {Ĥ, L̂2 , L̂z } together form Complete Set of Commuting
Operators.
References
1. S P Kuila, “Vector Analysis, Tensor Analysis and Linear
Vector Space”
2. B.S. Vatssa, “Elements of Modern Algebra”
3. Arfken and Weber, ”Mathematical Methods for Physicists”
4. S Lipschutz, “Schaum’s Outlines: Linear Algebra”
