1

Fundamentals of Nanoelectronics
ECE495 - Session 8, Sept 11, 2009
Schrodinger Equation: Finite Difference Method
Ref: Chapter 2.2
Professor Supriyo Datta
Class notes taken by: Mehdi Salmani
Schrdinger Equation
( ) +

= r U
m
E
v h
2
2
2



1-D Schrdinger Equation
( ) +

= r U
x m
E
v h
2
2 2
2
Time independent
and 1-D Schrodinger Equation.



There is analytical solution for particle in box condition but if we make another type of potential
function how should solve Schrdinger Equation? Numerical Solution.




Finite Difference
If the potential is independent of x, then the
solution to this equation can be written as:
( )
ikx iEt
e Ae t x
h /
,

=
Generally U(x) is a
complicated function
and analytical solutions
are not achievable.
Then, we have to rely in
numerical solutions.
Fundamentals of Nanoelectronics, Session 8
Sept 11 2009

Method
The approach to solve numerically is
How can we describe a function as a vector?
value of the function at each lattice po
corresponding to each lattice point there is a value for the wave function. This can also be viewed as
sampling of a continuous function into discrete values. Remember that in order to be able to
perform a numerical method we have to a have a finite number of equations so that we can solve
them.
First create a lattice for the 1-D problem:

Then: ( ) +

= r U
x m
E
v h
2
2 2
2
For large value of N, Matlab can be used to
eigenvectors:
) ( ] , [ H eig D V = D has the eigenvlaues of matrix H as its diagonal elements.
eigenvectors of H as its columns.
Example: Particle in a box

+ =
4 4 4 3 4 4 4 2 1
h
n Hamiltonia
x U
dx
d
m
E ) (
2
2
2 2
First try writing the matrix for U(x) and then the matrix for
To justify the numerical
approach we could test it on
an example that we know the
answer. Here, this could be
particle in a box.
Fundamentals of Nanoelectronics, Session 8
2
The approach to solve numerically is transforming differential equation to matrix equation.
How can we describe a function as a vector? We set up a lattice of discrete points and record the
value of the function at each lattice point. Figure shows the discrete points. As it can be seen
corresponding to each lattice point there is a value for the wave function. This can also be viewed as
sampling of a continuous function into discrete values. Remember that in order to be able to
form a numerical method we have to a have a finite number of equations so that we can solve
D problem:

)
(
(
(
(
(

(
(
(
(

=
(
(
(
(
(

N N
N N
H
E M M
2
1
2
1


For large value of N, Matlab can be used to find eigenvalues and
has the eigenvlaues of matrix H as its diagonal elements.
eigenvectors of H as its columns.
Example: Particle in a box
First try writing the matrix for U(x) and then the matrix for
Then,
becomes a
column vector
telling the value
of
different
numerical
approach we could test it on
an example that we know the
answer. Here, this could be
x = 0
U = 0
Once one has a matrix
equation like above,
eigenvalues of the N by N
matrix
evaluated. There wil
eigenvalues and N
eigenvectors
transforming differential equation to matrix equation.
We set up a lattice of discrete points and record the
int. Figure shows the discrete points. As it can be seen
corresponding to each lattice point there is a value for the wave function. This can also be viewed as
sampling of a continuous function into discrete values. Remember that in order to be able to
form a numerical method we have to a have a finite number of equations so that we can solve

has the eigenvlaues of matrix H as its diagonal elements. V has normalized
First try writing the matrix for U(x) and then the matrix for
2
2 2
2 dx
d
m
h

(x)
ecomes a
column vector
telling the value
of at
points.
(
(
(
(
(
(
(
(
(

) (
) (
) (
) (
) (
) (
3
2
1
N
n
x
x
x
x
x
x
M
M
x = L
0
Once one has a matrix
equation like above, The
eigenvalues of the N by N
matrix (H) can be
evaluated. There will be N
eigenvalues and N
eigenvectors.
Fundamentals of Nanoelectronics, Session 8
Sept 11 2009

The total Hamiltonian should be a sum of these two. For now concentrate on the easy part which is
U(x).
Consider [ ] = ) (x U E
us the potential at each lattice point, hence
Writing above as a matrix equation:

(
(
(
(

=
=
(
(
(
(
(

N
x U H
E M
2
1
) (
Then
Now, how do we write the second derivative at a particular point?
At each particular point the Schrdinger equation (after dropping U) can be written as:
n
n
dx
d
m
E
|
|

\
|
=
2
2 2
2
h

First try to write
(

dx
d
:
2 1
2
2

a
dx
d
dx
d
dx
d n
n
+

=
(

[
1 0
2 t E
n n n
=

Fundamentals of Nanoelectronics, Session 8
3
The total Hamiltonian should be a sum of these two. For now concentrate on the easy part which is
Since U(x) is a potential function, on a discrete lattice U would tell
ach lattice point, hence it will be diagonal:
n n n
x U E = ) (
Writing above as a matrix equation:
(
(
(
(
(

N
M
2
1
and
( )
( )

=
x U
x U
x U
0 0
0
0 0
0
) (
2
1
L L L
O M M
O M M
O M
L L
L L L

Now, how do we write the second derivative at a particular point?
2
2m
E =
h
Schrdinger equation (after dropping U) can be written as:
a dx
d
n n
n

1
2 1

=
(

+
+
and
dx
d
n
n 2 1

=
(

2
1 1
2
2
2 1
2
a dx
d
n n n
n
n
+

+
=
(

(

]
2
2
0 1
2ma
t and
n
h

+

The total Hamiltonian should be a sum of these two. For now concentrate on the easy part which is
Since U(x) is a potential function, on a discrete lattice U would tell

( )
(
(
(
(
(
(
(
(

(
N
x U 0
0
0
O
M
M
M L
L

2
2
dx
d

Schrdinger equation (after dropping U) can be written as:
a
n 1

Fundamentals of Nanoelectronics, Session 8
Sept 11 2009

E

If wed want to also include the potential to the matrix we can add its corresponding values to the
diagonal elements: ( =
n n
x U E

+
=
(
(
(
(
(
(
(
(

N
n
t
U t
E
0
0
0
0
2

0
2
1
M
M
Boundary Conditions
What do we do when we get near a boundary?
E
1
= (-t
0

0
)+ (2t
0
+ U
1
)
1
- t
0

Dropping the two terms is equivalent to setting the wavefunction to 0 at the two ends:

n+1
= 0. This would be appropriate f
allowed to penetrate outside the box.
For periodic boundary conditions, such as a particle on a ring, we let
easy to solve mathematically and due to that is used
Unlike the infinite-wall scenario, periodic boundary conditions h
matrix:

[ ]

=
N
t
0
0
2
Fundamentals of Nanoelectronics, Session 8
4
Matrix diagonal Tri
t
N
n
N
n

(
(
(
(
(
(
(
(

(
(
(
(
(
(
(
(

=
(
(
(
(
(
(
(
(

2 1 0 0 0 0
1 0 0 0
0 0 0
0 0 1 0
0 0 0 1 2 1
0 0 0 0 1 2

2
1
0
2
1
M
M
O O
O O O
O O
M
M

want to also include the potential to the matrix we can add its corresponding values to the
( ) 2 )
1 1 0 +
+
n n n n
t
( )
( )
( )

(
(
(
(
(
(
(
(

(
+

N
x U t t
t
t
t x U t t
t x U
2 0 0 0 0
0 0 0
0 0 0
0 0 0
0 0 0 2
0 0 0 0
0 0
0
0
0 2 0 0
0 1
O O
O O O
O O

What do we do when we get near a boundary?

2
Dropping the two terms is equivalent to setting the wavefunction to 0 at the two ends:
This would be appropriate for the particle in a box. Problem where the wavefun
penetrate outside the box.
boundary conditions, such as a particle on a ring, we let
0
=
N
and
easy to solve mathematically and due to that is used widely. Its real application is Carbon Nanotube.
wall scenario, periodic boundary conditions have a slightly different Hamiltonian
( )
( )
+

+
+
t t t
x U t t
t x U
0 0 0
2 0 0
0 1 0
2
2
L L L
O M M
O M M
O M M
L L L
L L L
want to also include the potential to the matrix we can add its corresponding values to the
(
(
(
(
(
(
(
(

N
n

2
1
M
M

Dropping the two terms is equivalent to setting the wavefunction to 0 at the two ends:
0
= 0 and
roblem where the wavefunction is not

and
N+1
=
1
. This is
widely. Its real application is Carbon Nanotube.
ave a slightly different Hamiltonian
( )
(
(
(
(
(
(
(
(

(
+

N
x U
t
t
0
0
M
M
M