Sturm-Liouville Problems

Sturm-Liouville Problems
The Sturm-Liouville (S-L) theory deals with a special type of second-order linear dif-
ferential equation that arises naturally in many physical problems, such as heat conduction,
wave propagation, and quantum mechanics. It provides a framework for understanding the
eigenvalues and eigenfunctions associated with these equations, which are crucial for solving
boundary value problems.

1. Introduction to Eigenvalue Problems

In linear algebra, for a matrix A, an eigenvalue problem is to find a scalar λ (eigenvalue)
and a non-zero vector v (eigenvector) such that:

Av = λv

In the context of differential equations, we replace the matrix A with a differential operator
L. So, a differential eigenvalue problem takes the form:

L[y] = λy

where L is a differential operator, y is a function (eigenfunction), and λ is a scalar (eigen-

value). The goal is to find the values of λ for which non-trivial solutions y(x) exist, subject
to certain boundary conditions.

2. Adjoint and Self-Adjoint Operators

2.1. Adjoint Operator
Consider a linear differential operator L. Its adjoint operator, denoted by L∗ , is defined
through an integration by parts formula. Let L be a linear differential operator acting on
functions u(x) and v(x) over an interval [a, b]. The adjoint operator L∗ is defined by the
relation: Z b Z b
vL[u]dx = uL∗ [v]dx + B(u, v) (∗)
a a

where B(u, v) is a boundary term that depends on u, v and their derivatives evaluated at
the boundaries a and b.

1
 Example: Let’s find the adjoint of L[u] = u′ . Using integration by parts:
Z b Z b
′ b
vu dx = [vu]a − uv ′ dx
a a
Rb Rb
So, a
vu′ dx = a
u(−v ′ )dx + [vu]ba . Comparing with equation (∗), we see that L∗ [v] = −v ′ .

2.2. Self-Adjoint Operator

A
R b linear differential
Rb operator L is said to be self-adjoint if L = L∗ . This means that
a
vL[u]dx = a uL[v]dx for all suitable functions u and v, provided the boundary terms
B(u, v) vanish.
For a second-order linear differential operator of the form:

L[y] = a2 (x)y ′′ + a1 (x)y ′ + a0 (x)y

It can be shown that L is self-adjoint if and only if a1 (x) = a′2 (x). If this condition holds,
the operator can be rewritten in the Sturm-Liouville form:
 
d dy
L[y] = p(x) + q(x)y
dx dx

where p(x) = a2 (x) and q(x) = a0 (x) − a′1 (x).

3. Self-Adjoint Differential Equations

A differential equation is called self-adjoint if the associated differential operator is self-
adjoint. The general form of a second-order linear self-adjoint differential equation is the
Sturm-Liouville form:  
d dy
p(x) + q(x)y = λr(x)y
dx dx
Here, p(x), q(x), and r(x) are given functions, and λ is a parameter. r(x) is called the
weight function or density function, and it is usually assumed to be positive, r(x) > 0.
How to transform a general second-order linear ODE into self-adjoint form:
Consider a general second-order linear ODE:

a2 (x)y ′′ + a1 (x)y ′ + a0 (x)y = λr(x)y

Multiply by an integrating factor µ(x):

µ(x)a2 (x)y ′′ + µ(x)a1 (x)y ′ + µ(x)a0 (x)y = λµ(x)r(x)y

We want µ(x)a1 (x) = (µ(x)a2 (x))′ . This leads to:

a1 (x) − a′2 (x)

Z 
µ(x) = exp dx
a2 (x)

Then p(x) = µ(x)a2 (x), q(x) = µ(x)a0 (x), and the new weight function is r̃(x) = µ(x)r(x).

2
 Solved Example: Transform y ′′ + x1 y ′ + (1 + λ)y = 0 into Sturm-Liouville form. Here,
1
a (x)−a′ (x) −0
a2 (x) = 1, a1 (x) = x1 . 1 a2 (x)2 = x 1 = x1 µ(x) = exp
R 1

x
dx = x. (Assuming x > 0)
Multiply the equation by x:
xy ′′ + y ′ + x(1 + λ)y = 0
(xy ′ )′ + x(1 + λ)y = 0
d dy
This is in the Sturm-Liouville form dx
(p(x) dx ) + q(x)y = λr(x)y, where p(x) = x, q(x) = x,
and r(x) = −x.

4. Eigenvalues and Eigenfunctions

For a Sturm-Liouville problem, the non-trivial solutions y(x) are called eigenfunctions,
and the corresponding values of λ are called eigenvalues.
Key Properties of Eigenvalues and Eigenfunctions:

• Real Eigenvalues: All eigenvalues are real.

• Orthogonality of Eigenfunctions: Eigenfunctions corresponding to distinct eigen-

values are orthogonal with respect to the weight function r(x). That is, if ym (x) and
yn (x) are eigenfunctions corresponding to distinct eigenvalues λm and λn (λm ̸= λn ),
then: Z b
r(x)ym (x)yn (x)dx = 0
a

• Completeness of Eigenfunctions: The set of eigenfunctions forms a complete basis.

• Simple Eigenvalues: For regular Sturm-Liouville problems, eigenvalues are usually

simple.

5. Sturm-Liouville Boundary Value Problems

A Sturm-Liouville Boundary Value Problem (SL-BVP) consists of:

1. A self-adjoint second-order linear differential equation (Sturm-Liouville equation):

 
d dy
p(x) + q(x)y = λr(x)y, on an interval [a, b]
dx dx

2. Boundary conditions at x = a and x = b. Common types include Dirichlet (y(a) =

0, y(b) = 0), Neumann (y ′ (a) = 0, y ′ (b) = 0), Robin (c1 y(a) + c2 y ′ (a) = 0, etc.), and
Periodic (y(a) = y(b), y ′ (a) = y ′ (b)).

3
6. Regular and Singular Sturm-Liouville Problems
6.1. Regular Sturm-Liouville Problems
A Sturm-Liouville problem is called regular if the following conditions are met:

1. The interval [a, b] is finite.

2. p(x), p′ (x), q(x), and r(x) are continuous on [a, b].

3. p(x) > 0 and r(x) > 0 on [a, b].

4. The boundary conditions are of the standard homogeneous type at both ends.

Properties of Regular S-L Problems:

• Infinitely many real eigenvalues.

• Eigenvalues can be ordered: λ1 < λ2 < λ3 < . . . and λn → ∞ as n → ∞.

• Each eigenvalue corresponds to a unique eigenfunction (up to a multiplicative con-

stant).

• The eigenfunctions form a complete orthogonal set.

Solved Example (Regular S-L Problem): Consider the S-L problem: y ′′ + λy = 0

on [0, π] with y(0) = 0 and y(π) = 0.
Solution: The characteristic equation is m2 + λ = 0.

• Case 1: λ < 0. Let λ = −µ2 for µ > 0. Leads to trivial solution.

• Case 2: λ = 0. Leads to trivial solution.

• Case 3: λ > 0. Let λ = ω 2 for ω > 0. y(x) = A cos(ωx) + B sin(ωx). y(0) = A = 0.

So, y(x) = B sin(ωx). y(π) = B sin(ωπ) = 0. For non-trivial solution, sin(ωπ) = 0.
This implies ωπ = nπ for n = 1, 2, 3, . . . . So, ω = n.

Eigenvalues: λn = n2 for n = 1, 2, 3, . . . Eigenfunctions: yn (x) = sin(nx) (choosing

B = 1).
Orthogonality Check: For m ̸= n,
Z π
1 π
Z
sin(mx) sin(nx)dx = [cos((m − n)x) − cos((m + n)x)]dx
0 2 0
 π
1 sin((m − n)x) sin((m + n)x)
= − =0
2 m−n m+n 0
This confirms orthogonality.

4
6.2. Singular Sturm-Liouville Problems
A Sturm-Liouville problem is called singular if one or more of the conditions for regularity
are violated. Common cases include:

• Infinite Interval: The interval [a, b] is infinite.

• Singularity in p(x) or r(x): p(x) or r(x) vanishes or becomes infinite at one or both
endpoints.

• Singularity in q(x): q(x) becomes infinite at an endpoint.

For singular problems, boundary conditions are often replaced by a requirement that solu-
tions remain bounded at singular points.
Examples of Singular S-L Problems:

• Legendre’s Equation: dx d dy


(1 − x2 ) dx + ν(ν + 1)y = 0 on [−1, 1]. Here p(x) = 1 − x2
is zero at x = ±1.

 2

• Bessel’s Equation: dxd dy
x dx + x − νx y = 0 on [0, R]. Here p(x) = x is zero at
x = 0.

• Hermite’s Equation:
2 2
d
dx
(e−x y ′ ) + 2λe−x y = 0 on (−∞, ∞). The interval is infinite.

Practice Questions
1. Transforming to Sturm-Liouville Form: Transform the following differential equa-
tions into Sturm-Liouville form, identifying p(x), q(x), and r(x): a) xy ′′ + y ′ + x1 y = λy on
[1, 2] b) x2 y ′′ − 2xy ′ + 2y = λy on [1, 3] c) (1 + x2 )y ′′ + 2xy ′ + y = λy on [0, 1]
2. Identifying Regular/Singular S-L Problems: For each of the following problems,
determine if it is a regular or singular Sturm-Liouville problem. Justify your answer. a)
y ′′ + λy = 0 on [0, 1] with y(0) = 0, y ′ (1) = 0. b) dxd
(ex y ′ ) + λex y = 0 on [0, ∞) with y(0) = 0
′′ ′
and y(x) bounded as x → ∞. c) xy + (1 − x)y + λy = 0 on [0, 1] with y(1) = 0 and y(0)
bounded. (Hint: Transform to S-L form first.)
3. Solving a Regular S-L Problem: Find the eigenvalues and eigenfunctions for the
following Sturm-Liouville problem: y ′′ + λy = 0 on [0, L] with y ′ (0) = 0 and y ′ (L) = 0.
Check the orthogonality of the eigenfunctions.
4. Conceptual Questions: a) Explain why the self-adjoint property is important in
the context of Sturm-Liouville problems. b) What is the significance of the weight function
r(x) in the orthogonality relation of eigenfunctions?
—

5
Solutions to Practice Questions
1. Transforming to Sturm-Liouville Form:
a) xy ′′ + y ′ + x1 y = λy This is already in Sturm-Liouville form: (xy ′ )′ + x1 y = λy. So,
p(x) = x, q(x) = x1 , r(x) = 1. R −2x−2x

b)R x2 y ′′ −
2xy ′ + 2y = λy Here, a2 (x) = x2 , a1 (x) = −2x. µ(x) = exp x2

dx =
−4 1 ′′ 2 ′ 1 ′
exp − x dx = x = x4 . Multiply by x4 : x2 y − x3 y + x4 y = λ x4 y This is dx x2 y + x24 y =
4 1 1 2 1 d

λ x14 y. So, p(x) = x12 , q(x) = x24 , r(x) = x14 .

c) (1+x2 )y ′′ +2xy ′ +y = λy This is already in Sturm-Liouville form: ((1+x2 )y ′ )′ +y = λy.
So, p(x) = 1 + x2 , q(x) = 1, r(x) = 1.
2. Identifying Regular/Singular S-L Problems:
a) y ′′ +λy = 0 on [0, 1] with y(0) = 0, y ′ (1) = 0. This is in S-L form with p(x) = 1, q(x) =
0, r(x) = 1.

• Interval is finite [0, 1].

• p, p′ , q, r are continuous.

• p(x) = 1 > 0, r(x) = 1 > 0.

• Boundary conditions are standard homogeneous type.

This is a regular Sturm-Liouville problem.

d
b) dx (ex y ′ ) + λex y = 0 on [0, ∞) with y(0) = 0 and y(x) bounded as x → ∞. Here,
p(x) = ex , q(x) = 0, r(x) = ex . The interval is infinite [0, ∞). This is a singular Sturm-
Liouville problem.
c) xy ′′ + (1 − x)y ′ + λy = 0 on [0, 1] with y(1) = 0 and y(0) bounded. First, transform to
S-L form. µ(x) = e−x . dx d
(xe−x y ′ ) + λe−x y = 0. So, p(x) = xe−x , q(x) = 0, r(x) = e−x . At
0
x = 0, p(0) = 0 · e = 0. This violates p(x) > 0. This is a singular Sturm-Liouville problem
at x = 0.
3. Solving a Regular S-L Problem: y ′′ + λy = 0 on [0, L] with y ′ (0) = 0 and
y ′ (L) = 0.

• Case 1: λ < 0. Let λ = −µ2 . Leads to trivial solution.

• Case 2: λ = 0. y ′′ = 0 =⇒ y(x) = Ax + B. y ′ (x) = A. y ′ (0) = A = 0. So y(x) = B.

y ′ (L) = 0 is satisfied. Eigenvalue: λ0 = 0. Eigenfunction: y0 (x) = 1.

• Case 3: λ > 0. Let λ = ω 2 . y(x) = A cos(ωx) + B sin(ωx). y ′ (x) = −Aω sin(ωx) +

Bω cos(ωx). y ′ (0) = Bω = 0 =⇒ B = 0. So, y(x) = A cos(ωx). y ′ (L) =
−Aω sin(ωL) = 0. For non-trivial solution, sin(ωL) = 0. ωL = nπ for n = 1, 2, 3, . . . .
So, ω = nπ
L
.

2
Eigenvalues: λn = nπ nπx


L
for n = 1, 2, 3, . . . . Eigenfunctions: yn (x) = cos L
.
Orthogonality Check:
RL
L
• For m = 0 and n ̸= 0: 0 1 · cos nπx
L
sin nπx


L
dx = nπ L 0
= 0.

6
 RL R L h  (m−n)πx   i
• For m ̸= n (both non-zero): mπx nπx 1 (m+n)πx



0
cos L
cos L
dx = 2 0
cos L
+ cos L
dx =
0.

Orthogonality confirmed.
4. Conceptual Questions:
a) Explain why the self-adjoint property is important in the context of Sturm-
Liouville problems. The self-adjoint property of the differential operator in a Sturm-
Liouville problem is crucial because it guarantees several fundamental properties of the
eigenvalues and eigenfunctions:

• Real Eigenvalues: All eigenvalues are real, which is essential for physical quantities.

• Orthogonality of Eigenfunctions: Self-adjointness leads directly to the orthogo-

nality of eigenfunctions corresponding to distinct eigenvalues. This property is vital
for expanding arbitrary functions in terms of these eigenfunctions.

• Completeness of Eigenfunctions: Self-adjoint operators typically yield a complete

set of eigenfunctions, forming a basis for a function space.

b) What is the significance of the weight function r(x) in the orthogonal-

ity relation of eigenfunctions? The weight function r(x) in the orthogonality relation
Rb
a
r(x)ym (x)yn (x)dx = 0 defines the inner product for the function space.

• Physical Meaning: In many physical applications, r(x) represents a physical quantity

like mass density. The orthogonality then accounts for this varying density.

• Mathematical Consistency: It ensures the integral remains consistent with the

self-adjoint form of the differential equation.

• Generalized Fourier Series: The orthogonality with respect to r(x) allows for the
expansion of arbitrary functions into a generalized Fourier series, which is funda-
mental for solving initial/boundary value problems.

1. Introduction to Eigenvalue Problems

An eigenvalue problem involves finding scalar values λ (eigenvalues) and nontrivial functions
y(x) (eigenfunctions) such that:

L[y] = λy
In Sturm-Liouville form, the equation becomes:
 
d dy
p(x) + [λw(x) − q(x)]y = 0
dx dx

7
Solved Example
Solve the eigenvalue problem:

y ′′ + λy = 0, y(0) = 0, y(π) = 0
General solution:
√ √
y(x) = A sin( λx) + B cos( λx)
Applying boundary conditions:

y(0) = 0 ⇒ B = 0
√ √
y(π) = 0 ⇒ A sin( λπ) = 0 ⇒ λ = n, λn = n2
Eigenfunctions: yn (x) = sin(nx), for n = 1, 2, 3, . . .

Practice Question
Solve the eigenvalue problem:

y ′′ + λy = 0, y(0) = 0, y(L) = 0

2. Adjoint and Self-Adjoint Operators

An operator L is adjoint to another operator L∗ if:
Z b Z b
uL[v] dx = vL∗ [u] dx + boundary terms
a a
A self-adjoint operator satisfies L = L∗ .

Self-Adjoint Form
 
d dy
L[y] = − p(x) + q(x)y
dx dx
is self-adjoint if:

• p(x), q(x), w(x) are real-valued and continuous

• Boundary conditions ensure integration by parts eliminates boundary terms

8
Solved Example
Check if L[y] = y ′′ is self-adjoint on [0, π] with y(0) = y(π) = 0.
Integration by parts:
Z π Z π Z π
′′ ′ ′
uy dx = − u y dx = yu′′ dx
0 0 0
∗ ′′
Thus, L [u] = u = L[u], hence L is self-adjoint.

Practice Question
d dy
Is the operator L[y] = dx
(x dx ) self-adjoint?

3. Sturm-Liouville Differential Equations

General form:
 
d dy
p(x) + [λw(x) − q(x)]y = 0
dx dx
Where:

• p(x), q(x), w(x): real-valued, continuous

• w(x) > 0: weight function

Solved Example
Given:
d dy
[(1 − x2 ) ] + λy = 0, −1 < x < 1
dx dx
This is a Legendre equation, with:

p(x) = 1 − x2 , q(x) = 0, w(x) = 1

Solutions: Legendre polynomials Pn (x), λn = n(n + 1)

Practice Question
Identify the weight function and p(x) in:
d x dy
(e ) + λex y = 0
dx dx

9
4. Eigenvalues and Eigenfunctions
Eigenvalues: values of λ for which non-trivial solutions exist. Eigenfunctions: corre-
sponding non-zero solutions y(x).
Orthogonality: Z b
ym (x)yn (x)w(x)dx = 0, m ̸= n
a

Solved Example
Solve:

y ′′ + λy = 0, y(0) = y(π) = 0
Eigenvalues: λn = n2 , Eigenfunctions: yn (x) = sin(nx)
Check orthogonality:
Z π
sin(mx) sin(nx)dx = 0 if m ̸= n
0

Practice Question
Show that cos(nπx/L) are orthogonal on [0, L]

5. Sturm-Liouville Boundary Value Problems

General form:

L[y] = λw(x)y
with boundary conditions such as y(a) = y(b) = 0, or Neumann/Robin conditions.

Solved Example
Solve:

y ′′ + λy = 0, y ′ (0) = 0, y ′ (π) = 0
Solution:
√ √
y(x) = A sin( λx) + B cos( λx)
√ √ √ √
y ′ (x) = A λ cos( λx) − B λ sin( λx)
Boundary conditions: √
y ′ (0) = A λ = 0 ⇒ A = 0
√ √ √
y ′ (π) = −B λ sin( λπ) = 0 ⇒ λ = n ⇒ λn = n2
Eigenfunctions: yn = cos(nx)

10
Practice Question
Solve:
d dy
(x ) + λxy = 0, y(1) = 0, y(e) = 0
dx dx

6. Regular and Singular Sturm-Liouville Problems

Regular:

• Coefficients are continuous

• p(x) > 0, w(x) > 0

• Finite interval, appropriate boundary conditions

Singular:

• Interval is infinite or includes a singularity (e.g., p(x) = 0)

Examples
Regular:
d dy
[(1 − x2 ) ] + λy = 0, x ∈ (−1, 1)
dx dx
Singular:
d dy
(x ) + λxy = 0, x ∈ (0, 1)
dx dx
(Singular at x = 0, since p(x) = x → 0)

Practice Question
Classify the problem:
d 2 dy
(x ) + λx2 y = 0, x ∈ (0, 1)
dx dx

11