0% found this document useful (0 votes)
49 views16 pages

qm2 Ts 3 Soln

This tutorial sheet contains problems on using the variational principle in quantum mechanics. The first problem obtains the ground state energy of a 1D simple harmonic oscillator using a trial wave function. After normalization and evaluating integrals, the exact ground state energy is recovered. The second problem shows that when applying the variational principle, the integral used can be written as the sum of kinetic and potential energy terms using integration by parts.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
49 views16 pages

qm2 Ts 3 Soln

This tutorial sheet contains problems on using the variational principle in quantum mechanics. The first problem obtains the ground state energy of a 1D simple harmonic oscillator using a trial wave function. After normalization and evaluating integrals, the exact ground state energy is recovered. The second problem shows that when applying the variational principle, the integral used can be written as the sum of kinetic and potential energy terms using integration by parts.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 16

PH 422: Quantum Mechanics II

Tutorial Sheet 3
This tutorial sheet contains problems related to the use of variational principle in quan-
tum mechanics.

1. Obtain the energy of the ground state of a one-dimensional (1D) simple-harmonic


−αx2
oscillator (SHO) using the trial wave function ψ(x) = ce , where c is the normal-
ization constant, and α is the variational parameter.
Soln: Let us estimate the ground state of one-dimensional simple harmonic oscillator
2
using the trial wave function of the form ψ(x) = ce−αx Because this function is of the
form exact wave function, the obtained energy should be exact ground state energy

. Let us rst normalize ψ(x)
2
Z ∞
2
hψ|ψi = c 2
e−2αx dx = 1
−∞

Substitute t = 2αx

r
c2
Z
−t2 π
⇒ hψ|ψi = √ e dt = c2 =1
2α −∞ 2α
 2α 1/4
⇒c=
π
R∞ 2 √
Above we used value of the Gaussian integral
−∞
e−t dt = π .Now
r ∞
hψ|H|ψi ~2 d2 −αx2 1
Z
2α 2

2 2 −αx2

E(α) = = e−αx − e + mω x e dx
hψ|ψi π −∞ 2m dx2 2

Now

d2  −αx2 d −αx2

e = − 2αxe
dx2 dx !
−αx2 −αx2
= − 2αe + 4α2 x2 e

r ∞
~2 d2 −αx2 1
Z
2α 2

2 2 −αx2

⇒ E(α) = e−αx − e + mω x e dx
π −∞ 2m dx2 2
Using the standard integral

Z ∞ √ a2n+1 (2n − 1)!


x2
x2n e− a2 dx = π
−∞ 2n

We have √ 
Z ∞
2 −2αx2 π 1 3
xe dx = √
−∞ 2 2α

1
So that
(r r r √ √ √ )
2απ ~2 ~2 2α 2 π 1 1 2 π 2α
E(α) = 2α − 4α + mω √
π
2α 2m 2m π 2 (2α)3/2 2 2 π (2α)3/2
( )
~2 mω 2
E(α) = α+
2m 8α

dE ~2 mω 2
=0⇒ − =0
dα 2m 8α2

⇒α=
2~

~2 mω mω 2 mω
⇒ Emin = + ×
2m 2~ 8 2~
~ω ~ω ~ω
Emin = + =
⇒ 4 4 2
= E0 (Exact GS)
and
 mω 1/4 mωx2
ψ(x) = e− 2~
π~
= ψ0 (x)
Thus, as expected, we recover the exact ground state energy and wave function for
this trial wave function

2. In the variational principle as applied to quantum mechanics, one minimizes the inte-
~2 ∗ 2
ψ ∇ ψ+V ψ ∗ ψ}d3 r, subject to the normalization condition
R
gral I
R ∗ = hψ|H|ψi = {− 2m
ψ ψd3 r = 1. Show using integration by parts, that one can also use the expression
R ~2
I = { 2m ∇ψ ∗ · ∇ψ + V ψ ∗ ψ}d3 r.
Soln:
~2 ∗ 2
Z  

I= − ψ ∇ ψ + V ψ ψ d3 r
2m
The second term remains unchanged so we concentrate only on the rst term

~2
Z
I1 = − ψ ∗ ∇2 ψd3 r
2m
Z ∞Z ∞Z ∞
~2
 2
∂ ψ ∂ 2ψ ∂ 2ψ


=− ψ + 2 + 2 dxdydz (1)
2m −∞ −∞ −∞ ∂x2 ∂y ∂z
Let us consider the rst integral and apply integration by parts
Z ∞Z ∞Z ∞
~2 ∂ 2ψ
I1x =− ψ ∗ 2 dxdydz
2m −∞ −∞ −∞ ∂x
Z ∞ Z ∞ Z ∞Z ∞Z ∞
~2 ∂ψ ∗ ∂ψ


∗ ∂ψ
=− dy dzψ − dxdydz
2m −∞ −∞ ∂x −∞ −∞ −∞ −∞ ∂x ∂x

2
the rst term on the RHS vanishes because ψ ∗ (x = ±∞, y, z) = 0, because wave func-
tion (and its complex conjugate) must vanish at innity for it to be normalizable

~2 ∂ψ ∗ ∂ψ
Z Z Z
I1x = dxdydz (2)
2m ∂x ∂x
Similarly we can show
Z ∞Z ∞Z ∞
~2 ∂ 2ψ
I1y = − ψ ∗ 2 dxdydz
2m −∞ −∞ −∞ ∂y
(3)
~2 ∂ψ ∗ ∂ψ
Z Z Z
= dxdydz
2m ∂y ∂y
and Z ∞Z ∞Z ∞
~2 ∂ 2ψ
I1z = − ψ ∗ 2 dxdydz
2m −∞ −∞ −∞ ∂z
(4)
~2 Z Z Z
∂ψ ∗ ∂ψ
= dxdydz
2m ∂z ∂z
using Eq.(2),Eq.(3),Eq.(4) in Eq.(1), we have

~2 ∂ψ ∗ ∂ψ ∂ψ ∗ ∂ψ ∂ψ ∗ ∂ψ
Z  
I1 = + + d3 r
2m ∂x ∂x ∂y ∂y ∂z ∂z
but

∂ψ ∗ ∂ψ ∂ψ ∗ ∂ψ ∂ψ ∗ ∂ψ ∂ψ ∗ ∂ψ ∗ ∂ψ ∗
  
∂ψ ∂ψ ∂ψ
+ + = î + ĵ + k̂ . î + ĵ + k̂
∂x ∂x ∂y ∂y ∂z ∂z ∂x ∂y ∂z ∂x ∂y ∂z
~ ).∇ψ
= (∇ψ ∗ ~

~2
Z
⇒ I1 = ~ ∗ ) · (∇ψ)d
(∇ψ ~ 3
r
2m

3. Estimate the ground state energy of a 1D-SHO using the trial wave function of
−α|x|
form ψ(x) = Ce , treating α as a variational parameter. (Helpful integral:
Rthe
∞ −αx n n!
0
e x dx = αn+1 .)
Soln: First we normalise ψ(x)
Z ∞
c 2
e−2α|x| dx
−∞
Z 0 Z ∞
⇒ c2 e2αx + c2 e−2αx dx = 1
−∞ 0
c2 c2
⇒ + =1
2α 2α √
⇒c= α

3
√ −α|x|
ψ(x) = αe
∂2ψ
Because the slope of this wave function is discontinuous at
∂x2
x = 0, so
is not dened
there. Therefore, we use expression of Prob. 2 for computing energy expectation value

E(α) = hψ(α)|H|ψ(α)i
Z ∞  2  2 
~ dψ 2
= + V ψ dx
−∞ 2m dx
Z  αx 2 2
~2 α ∞ de−αx
Z ∞
~2 α 0
Z  Z 0
de 1 1
= dx + 2
dx + mω α 2 2αx 2
x e dx + mω α x2 e−2αx dx
2m −∞ dx 2m 0 dx 2 −∞ 2 0

using
Z 0 Z ∞
2!
2 2αx
xe dx = x2 e−2αx dx =
−∞ 0 (2α)3
1
= 3

~2 α3 0 2αx ~2 α3 ∞ −2αx mω 2
Z Z
E(α) = e dx + e dx +
2m −∞ 2m 0 4α2
~2 α3 ~2 α3 mω 2
= + +
4mα 4mα 4α2

~2 α2 mω 2
E(α) = + (5)
2m 4α2
dE ~2 α mω 2
=0⇒ − =0
dα m 2α3

m2 ω 2
⇒ α4 =
2~2 r
1 mω
⇒ α = ± 1/4 (6)
2 ~
1
p mω
but only α= 21/4 ~
will lead to a normalizable wave function, using this in Eq.(5),
we have √
~2 mω mω 2 2~
Emin = × √ + ×
2m ~ 2 4 mω
1 1
Emin = √ ~ω > ~ω
2 2
2
4. Show that for a 1D-SHO, if one uses a trial wave function ψ(x) = cxe−αx ,where c
is the normalization constant and α is the variational parameter, one obtains exact

4
energy E = 23 ~ω of the rst excited state.
Soln: Let us rst normalize the trial wave function

Z ∞
2
c 2
x2 e−2αx dx
−∞

using the result


Z ∞ √ 
2 −2αx2 π 1 3
xe dx = √
−∞ 2 2α
we have above

r
c2
Z
2 π
ψ (x)dx = =1
−∞ 2 8α3
r
√ α3
⇒ c2 = 4 2
π
 3  14

⇒c=2
π

 3  14
2α 2
⇒ ψ(x) = 2 xe−αx
π
 3  14  3  41
dψ(x) 2α −αx2 2α 2
⇒ =2 e −4 αx2 e−αx
dx π π
 2  3  12  3  21  3  12
dψ 2α −2αx2 2α 2 4 −2αx2 2α 2
⇒ =4 e + 16 α xe − 16 αx2 e−2αx
dx π π π

We will use the expression of problem 2 to compute the energy expectation value

E(α) = hψ(α)|H|ψ(α)i
Z ∞  2 Z ∞
~2 dψ 1 2
= dx + mω x2 ψ 2 (x)dx
2m −∞ dx 2 −∞
2
 3 1/2 Z ∞  3 1/2 Z ∞
~ 2α −2αx2 ~2 2α 2
= ×4 e dx + 16α 2
x4 e−2αx dx
2m π −∞ 2m π −∞
2
 3
1/2 Z ∞  3 1/2 Z ∞
~ 2α 2 1 2α 2
− 16α x2 e−2αx dx + mω 2 × 4 x4 e−2αx dx
2m π −∞ 2 π −∞

using the integrals



Z r
−αx2 π
e dx =
−∞ α
and

Z r r
4 −αx2 3!! π 3 π
xe = 2 2 = 2
−∞ 2α α 4α α

5
we have

1/2 r 1/2 r 1/2 r


2~2 2α3 π 8~2 2α3 8~2 2α3
  
2 3 π 1 π
E(α) = + α − α
m π α m π 4(2α)2 2α m π 2 8α3
1/2 r
2α3

2 2 3 π
+ 2ω m α
π 4(2α)2 2α
2~2 α 8~2 × 3α × α2 8~2 α 2mω 2 × 3α
= + − +
m m × 4 × 4α2 4m 16α2
2 2
3~ α 3mω
= +
2m 8α
3~2 α 3mω 2
E(α) = +
2m 8α
2
dE 3~ 3mω 2
=0⇒ − =0
dα 2m 8α2

m2 ω 2
⇒ α2 =
4~2

⇒α=
2~

mω 3 3
Emin = E(α = ) = ~ω + ~ω
2~ 4 4
3
Emin = ~ω
2
which is exact result

5. Here we derive the linear-combination of basis functions approach, quite commonly


used in quantum mechanics, using a variational principle. Suppose that the Hamilto-
nian of a system is given by H, and we assume that the state ket |ψi corresponding
to its ground state can be approximated as

N
X
|ψi = Cj |ji,
j=1

where |ji denote the known basis kets, while Cj are the unknown expansion coe-
cients which are also the variational parameters in this approach, and, in general, are
complex. In the r representation, the following notation is adopted ψ(r) = hr|ψi, and
φj (r) = hr|ji. Using the variational principle, show that the ground state energy E,
and the state ket |ψi can be obtained by solving the generalized eigenvalue problem

H̃ C̃ = E S̃ C̃,

6
where H̃ and S̃ N × N matrices, representing the Hamiltonian and the
denote the
overlap, with elements dened as Hij = hi|H|ji, Sij = hi|ji, respectively, while Ci
form the N elements of the column vector C̃ , denoting the ground state eigenfunction.
Note that form an orthonormal basis set, hi|ji = δij so that S̃ = I , and the previous
generalized eigenvalue problem reduces to a normal eigenvalue problem.
Soln: According to the variational principle, we should minimize

hψ|H|ψi
E(Ci , Ci∗ ) = ,
hψ|ψi

with respect to the variational coecients Ci , i = 1, 2, 3, . . . N . Using the given


expansion of |ψi, and the denitions of H̃ij = hi|H|ji and S̃ij = hi|ji,we obtain

∗ ∗
P P
∗ hψ|H|ψi j,k C k C j hk|H|ji j,k Ck Cj Hkj
E(Ci , Ci ) = = P ∗
= P ∗
. (7)
hψ|ψi j,k Ck Cj hk|ji j,k Ck Cj Skj

The variational principle in the present case implies that the conditions

∂E
=0
∂Ci
for i = 1, 2, 3, . . . N (8)
∂E
=0
∂Ci∗

must hold. Because Ci 's are complex, therefore, Ci and Ci are independent variables.
∂C ∗ ∂C ∗
∂Ci ∂Ci
Using the results
∂Cj
= ∂Ci∗ = δij and ∂Cij = ∂C ∗ = 0, we obtain on applying Eq. 8 on
j j
Eq. 7

( j,k Ck∗ δij H̃kj )( j,k Ck∗ Cj Skj ) − ( j,k Ck∗ Cj Hkj )( j,k Ck∗ δij Skj )
P P P P
∂E
= = 0.
( j,k Ck∗ Cj Skj )2
P
∂Ci

This can be written as

P ∗  P C∗C H  P
( k Ck∗ Ski )
j kj
∂E ( k Ck Hki ) − Pj,k Ck∗ Cj Skj
j,k k
= =0
( j,k Ck∗ Cj Skj )
P
∂Ci
!

P
X C
j,k k C j H kj X
=⇒ ( Ck∗ Hki ) − P ∗
( Ck∗ Ski ) = 0.
k
C C
j,k k j kj S k

Now, on using Eq. 7 in the second term above, we have

X X
( Ck∗ Hki ) − E( Ck∗ Ski ) = 0. (9)
k k

The complex conjugate of the previous equation yields

X X
∗ ∗
( Ck Hki ) − E( Ck Ski ) = 0.
k k

7
∗ ∗
Using the fact that H̃ and S̃ are Hermitian matrices, we have Hki = Hik and Ski = Sik .
On substituting these in previous equation, we obtain
X X
Hik Ck = E Ski Ck
k k

=⇒ H̃ C̃ = E S̃ C̃. (10)

This equation is called a generalized eigenvalue problem because of the presence of the
overlap matrix S̃ on the RHS, and clearly reduces to the normal eigenvalue problem
H̃ C̃ = E C̃ , for an orthonormal basis (S̃ = I ). Note that we obtained this equation by
taking the complex conjugate of the original equation 9, which actually is an eigenvalue
∗ †
problem for the complex conjugates of the coecients, Ck or C̃ (i.e., for hψ|). You can
∂E
verify that if we start with the condition
∂Ci∗
= 0, we will directly get the eigenvalue
problem of Eq. 10.

6. This problem is a simple application of the linear-combination of basis functions ap-


proach. Suppose the wave function of a given quantum mechanical system can be
expanded in terms of three basis functions {|ii, i = 1, 2, 3}, which form an orthonor-
mal set hi|ji = δij . Dening the Hamiltonian matrix elements with respect to these
basis functions as Hij = hi|H|ji, it is given that the only non-zero Hamiltonian matrix
elements are H12 = H21 = H23 = H32 = H13 = H31 = t, where t is a real positive
number. Obtain the eigenvalues and eigenvectors of this Hamiltonian. How do the
results change when we set H13 = H31 = 0?
Soln:

(a) With
 
0 t t
H = t 0 t
t t 0
the characteristic polynomial is

−λ t t
t −λ t = 0
t t −λ

⇒ −λ λ2 − t2 + t t2 + tλ + t t2 + tλ = 0
  

⇒ 2t2 (t + λ) − λ(λ − t)(λ + t) = 0


⇒ (λ + t) 2t2 − λ2 + λt = 0


⇒ λ = −t
or

⇒λ2 − λt − 2t2 = 0
⇒λ2 − 2λt + λt − 2t2 = 0
⇒λ(λ − 2t) + t(λ − 2t) = 0
⇒(λ + t)(λ − 2t) = 0
⇒λ = −t, λ = 2t

8
⇒λ = −t degenerate

λ = 2t

Let's nd the eigenvectors


(i) λ = −t
(H −
 λI)c
 = 0
t t t c1
⇒  t t t   c2  = 0
t t t c3
⇒ c1 + c2 + c3 = 0
Two possibilities

1
1
|λ = −ti1 = √  0 
2 −1
 
1
1
|λ = −ti2 = √  −2 
6 1
(ii) λ = 2t
 (H − λI)c=0 
−2t t t c1
⇒  t −2t t   c2  = 0
t t −2t c3
⇒ −2c1 + c2 + c3 = 0
c1 − 2c2 + c3 = 0
c1 + c2 − 2c3 = 0
Any two of these equations are linearly independent. A possible solution which
orthogonal to |λi1 and |λi2 is
 
1
1
|λ = 2ti3 = √  1 
3 1

(b) When H13 = H31 = 0, we have

 
0 t 0
H = t 0 t
0 t 0

|H − λI| = 0 is
−λ t 0
t −λ t = 0
0 t −λ

9
⇒ −λ λ2 − t2 + t2 λ = 0


⇒ λ(λ2 − 2t2 ) = 0

⇒ λ = 0, ±t 2

Let's nd the eigenvectors


(i) λ = 0
 (H −  λI)c
 = 0
0 t 0 c1
⇒  t 0 t   c2  = 0
0 t 0 c3
⇒ c2 = 0
c1 + c3 = 0
 
1
1
|λ = 0i = √  0 
2 −1

(ii) λ = ± 2t

√ (H − λI)c = 0
  
∓ 2t t √ t c1
⇒  t ∓ 2t t √   c2  = 0
t t √ ∓ 2t c3
⇒ ∓ √ 2c1 + c2 = 0
c1 ∓ 2c √2 + c3 = 0
c2 ∓ 2c3 = 0
Possible solutions are  
√ 1

1
|λ = 2ti =  2 
2
1
 
√ 1

1
|λ = − 2ti =  − 2 
2
1
We note that in this case, eigenvalues are symmetrically placed about λ = 0,
which is a sign of particle hole symmetry.

7. Estimate the ground state energy of a particle of mass m in a box with V = 0, for
0 ≤ x ≤ a, and V = ∞, otherwise, using variational principle. For the purpose, take
a wave function consisting of two linear components ψ1 (x) and ψ2 (x) dened by: (i)
ψ1 (0) = 0, ψ1 (x = α) = C for 0 ≤ x ≤ α, and (ii) ψ2 (x = α) = C, ψ2 (x = a) = 0, for
α ≤ x ≤ a, where C is the normalization constant, and α is the variational parameter.
Soln: We estimate the ground state energy of a particle of mass m, in a one dimen-
sional box of length a. We consider the trial function to be a linear function which is
zero at x=0 and x=a, and is peaked at x=α, 0 ≤ α ≤ a, where α is the variational
parameter.

10
Clearly

Nx
ψ(x) = ψ1 (x) = f or 0 ≤ x ≤ α
α
N (a − x)
ψ(x) = ψ2 (x) = f or α ≤ x ≤ a
(a − α)
ψ(x) = 0 elsewhere
To obtain normalization constant
a
N2 α 2
Z a
N2
Z Z
2
ψ (x)dx = 2 x dx + (a − x)2 dx
0 α 0 (a − α)2 α
N2 N2
= α+ (a − α) = 1
3 3

N2
a=1
3 r
3
N=
a
r
3x
⇒ ψ1 (x) =

r
3 (a − x)
ψ2 (x) =
a (a − α)
Now the standard form

~2 2
Z
ψ∗ −

E = hψ|H|ψi = ∇ + V ψdτ
2m
is not valid here because ψ 0 (x) is discontinuous at x = α. For such cases one uses the
alternative expression

 ~2
Z
E= (∇ψ ~
~ ∗ ).(∇ψ) + V ψ ∗ ψ dτ
2m

11
which can be obtained by integrating by parts the rst term and using the fact that

the wave function vanishes at innity.For the present case ψ = ψ and V = 0, so that
Z a  2
~2 dψ
E(α) = dx
2m 0 dx
Z α Z a
~2 dψ1 2 ~2 dψ2 2
= dx + dx
2m 0 dx 2m α dx
or
Z α Z a
~2  3  1 ~2  3  1
E(α) = dx + dx
2m a α2 0 2m a (a − α)2 α
~2  3 n 1 1 o
= +
2m a α a − α

dE ~2  3 n 1 1 o
⇒ = − 2+ =0
dα 2m a α (a − α)2
⇒ a − α = ±α
The only meaningful solution is

a
2α = a ⇒ α =
2
and
~2  3  4 6~2
Emin = × =
2m a a ma2
~2 π 2 5~2
E0 (exact) = 2 ≈ ⇒ Emin > E0
2a m ma2
q
2
If we plot the true ground state wave function ψ0 =
a
sin( πx
a
) along with the ap-

proximate wave function ψ(x) is obtained above, we have

Figure 1: Comparison between the exact and the approximate wave functions

12
a
We note that α=
2
obtained through variational principle ensures that the variational
a
wave function peaks at the same x = , as an exact wave function.
2

8. Consider the Hamiltonian of a particle moving in a 1D Gaussian potential well H =


p2 2
2m
− V0 e−ax , with V0 and a > 0. Estimate its ground-state energy employing varia-
−αx2
tional principle, with a trial wave function of the form ψ(x) = Ce , with α as the
variational parameter.
Soln: Let us compute

∞  2
~2 dψ
Z  
2
E(α) = + V ψ dx
−∞ 2m dx
2 2α 1/4
ψ(x) = Ce−αx

Here where C= π
was computed in problem 1. Now

 1/4 n
dψ(x) 2α 2
o
= − 2αxe−αx
dx π
so
r ∞ ∞
~2
 Z Z 
2α 2 2 −2αx2 −(2α+a)x2
E(α) = 4α x e dx − V0 e dx
π 2m −∞ −∞
r
~2 2 1
 r r 
2α π π
E(α) = 4α − V0
π
2m 2 (2α)3 2α + a
r
~2 2α
= α − V0
2m 2α + a
With this
√ √
dE ~2 2 2 2α
= − V0 p + V0 =0
dα 2m 2 α(2α + a) 2(2α + a)3/2
( √ )
V0 1 2α ~2
=⇒ p √ − =
(2α + a) 2α (2α + a) 2m
aV0 ~2
=⇒ p =
2α(2α + a)3 2m
4m2 a2 V02
=⇒ 2α(2α + a)3 =
~4
After solving this equation for α, we can obtain the value of ground state energy E(α).

9. Using the trial wave function ψ(r) = Ce−αr , where C is the normalization constant,
and α is the variational parameter, estimate the ground state energy of the hydrogen
atom.
Soln:
ψα (r) = Ce−αr

13
Z Z ∞
ψα2 (r)d3 r =C 2
4πr2 e−2αr dr = 1
0
Z ∞
⇒ 4πc2 r2 e−2αr dr = 1
0
2!
⇒ 4πC 2 =1
(2α)3
πC 2
=1
α3 r
α3
C=
π
r
α3 −αr
ψ(r) = e
π
The Hamiltonian for the hydrogen atom is

~2 2 e2
H=− ∇ − ,
2m r
so that

E(α) = hψ(α)|H|ψ(α)i
~2 2 e2 2
Z    
= ψα − ∇ ψα − ψα (r) dr
2m r

but

~2 2 ~2 ∂ ~2 ~2 ∂2
   
2 ∂ ∂ ∂
− ∇ =− r − sin θ −
2m 2mr2 ∂r ∂r 2mr2 sin θ ∂θ ∂θ 2mr2 sin2 θ ∂ϕ2

As there is no angular dependence of ψα (r) since the ground state functionis spherically

14
symmetric, so the last two terms give us zero, so we are left with

~2 ∂ e2 2
Z     
2 ∂ψα
E(α) = ψα − r − ψα (r) dr
2mr2 ∂r ∂r r
−αr
α3 ~2 ∂ e2 −2αr
Z     
−αr 2 ∂e
= e − r − e dr
π 2mr2 ∂r ∂r r
−αr Z −2αr
~2 α3 e2 α3
Z   
−αr 1 ∂ 2 ∂e e
=− e 2
r dr − dr
2m π r ∂r ∂r π r
Z −2αr
~2 α3 e2 α3
Z  
−αr 1 ∂ 2 −αr
 e
=− e 2
r [−αe ] dr − dr
2m π r ∂r π r
Z −2αr
~2 α4 e2 α3
Z  
−αr 1 ∂ 2 −αr  e
= e r e dr − dr
2m π r2 ∂r π r
Z −2αr
~2 α4 e2 α3
Z  
−αr 1  −αr 2 −αr
 e
= e 2
2re − αr e dr − dr
2m π r π r
Z −2αr
~2 α4 e2 α3
Z  
−αr 1  −αr 2 −αr 2 e
4πr2 dr

= e 2
2re − αr e 4πr dr −
2m π r π r
2~2 α4
Z Z
−αr −αr 2 −αr 3 2
re−2αr rdr

= e 2re − αr e dr − 4α e
m
2~2 α4
 Z Z  Z
−2αr 2 −2αr
= 2 re dr − α r e dr − 4α e 3 2
re−2αr rdr
m

Using the denition of Gamma function

Z ∞
Γ(n + 1)
xn e−ax dx = where a > 0
0 an+1
we obtain

2~2 α4
      
1! 2! 3 2 1!
E(α) = 2 −α − 4α e (11)
m (2α)2 (2α)3 (2α)2
4~2 α4
 
1 1
= 2
− 2 − αe2
m 4α 8α
2 2

= − αe2
2m

dE ~2 α
=0⇒ − e2 = 0
dα m
me2 1
⇒α= 2 = (12)
~ a0

15
~2
where a0 = me2
is the Bohr radius. Substituting Eq.(12) in Eq.(11), we get

~2 m2 e4 me4
 
Emin = −
2m ~4 ~2
me4 me4
= − 2
2~2 ~
4
me
=− 2
2~

me4
Emin = −
2~2
This is the exact value of the ground state energy of the hydrogen atom, obtained after
solving the Schrödinger equation.

16

You might also like