Dong S.-H. Wave Equations in Higher Dimensions

Подождите немного. Документ загружается.

74 6 Harmonic Oscillator

these, such a physical problem was discussed in arbitrary dimensions D [201–205].

Recently, Sage has studied its vibrations and rotations in order to describe the di-

atomic molecule [208], in which he brieﬂy reinvestigated some properties of this

oscillator to study the pseudogaussian oscillator. On the other hand, we notice that

the harmonic oscillator plus inverse squared potential is essentially equal to a pseu-

doharmonic oscillator except for some unimportant parameters as studied in [210].

Consequently, it is unnecessary to restudy this case for simplicity.

5.2 Exact Solutions

AsshowninRef.[3], the pseudoharmonic oscillator is taken as

V(r)=

κr



−



, (6.41)

where κ is a force constant and the r

represents the equilibrium bond length.

It should be noted that the potential taken here is slightly different from that of

Ref. [207], but they are the same essentially except for an unimportant constant.

Let

φ(r) =e

−λξ

R(ξ), ξ =r

. (6.42)

In a similar way, we may obtain the radial equation from Eqs. (3.66) as follows:

R(ξ)

dξ



2τ +

−2λξ



dR(ξ)

dξ



−1



τ(2τ +D −2)

−

l(l +D −2)

−

Mκr



+ξ



−

Mκ





Mκr

−2λ



τ +



R(ξ) =0. (6.43)

Take the parameters λ and τ to remove the terms of ξ

−1

and ξ ,

λ =

√

Mκ

,τ=−

D −2



(2l +D −2)

+Mκr



1/2

. (6.44)

Further deﬁne a new variable ρ =2λξ . We may obtain the conﬂuent hypergeo-

metric equation [192]:

R(ρ)

dρ



2τ +

−ρ



dR(ρ)

dρ





√

Mκ

−τ −



R(ρ) =0. (6.45)

Therefore, the solutions to the D-dimensional Schrödinger equation with a pseudo-

harmonic potential are written as

D−1

,...,l

(x) =N

E,l

−ρ/2

(−n

;2τ +D/2;ρ)Y

D−1

,...,l

(

x), (6.46)

5 Generalization to Pseudoharmonic Oscillator 75

where

ρ =

√

Mκ



√

Mκ

−τ −

=0, 1,...,

E,l

(Mκ)

D/8

(2τ +D/2)



(n

+2τ +D/2)

D/2−1



1/2

(6.47)

Likewise, considering the ﬁniteness of the solutions at inﬁnity leads to a quan-

tum condition that n

has to be a non-negative integer. The quantized energy E is

calculated as





−

√

Mκ

+τ +



. (6.48)

The energy spectrum is equidistant for given constants τ and r

When r

is very large, r

2/

√

Mκ,wehave





−

(l +D/2 −1)

√

Mκ



, (6.49)

and when r

2/

√

Mκ,wehave





−

√

Mκ

Mκr

16(l +D/2 −1)



. (6.50)

5.3 Ladder Operators

During the past half century, the factorization method has played an important role

in physics [3, 193, 211]. With this method, one can construct the ladder operators for

certain potentials such as the Morse and modiﬁed Pöschl-Teller potentials and then

constitute a suitable algebra su(2). This approach is different from the traditional

one, where an auxiliary non-physical variable is introduced [212, 213]. In particular,

the matrix elements of some related operators, which is of signiﬁcance in physics,

can be analytically evaluated from the ladder operators. Therefore, we are going to

establish the ladder operators for pseudoharmonic oscillator.

By using the recurrence relation between the conﬂuent hypergeometric functions

and the associated Laguerre functions as well as the orthogonal relation of the asso-

ciated Laguerre functions as Eq. (5.21), we may reexpress the radial wavefunction

of pseudoharmonic oscillator as follows:

(ρ) =N

−

2τ +D/2−1

(ρ),

=(Mκ)

D/8



D/2−1

(2τ +n +D/2)

(6.51)

where n =n

is used.

76 6 Harmonic Oscillator

Now, we study the ladder operators

with the following properties

(ρ)

dρ

(ρ),

(ρ) =m

±1

(ρ), (6.52)

where we stress that these operators only depend on the physical variable ρ.

For this purpose, we begin by applying the operator

dρ

on the wavefunction

(6.51)

dρ

(ρ) =



−



(ρ) +N

−

dρ

2τ +D/2−1

(ρ), (6.53)

which is used to construct the ladder operators

To this end, if we consider useful relations (6.35) and substitute them into (6.53),

then we are able to obtain the following relations



dρ

−

n +τ



(ρ) =−

n +2τ +D/2 −1

n−1

(ρ), (6.54)



dρ

n +τ +D/2

−



(ρ) =

n +1

n+1

(ρ), (6.55)

from which, together with N

given in Eq. (6.51), we can obtain the following ladder

operators

−

=−ρ

dρ

+τ +ˆn −

=ρ

dρ

+τ +ˆn +

−

(6.56)

where we have introduced a number operator ˆn with the property

ˆnR

(ρ) =nR

(ρ). (6.57)

The ladder operators

have the following properties

−

(ρ) =m

−

n−1

(ρ),

(ρ) =m

n+1

(ρ),

(6.58)

where

−



n(n +2τ +D/2 −1),



(n +1)(n +2τ +D/2).

(6.59)

On the other hand, we note that the radial wavefunction can be directly obtained

by applying the creation operator

on the ground state R

(ρ), i.e.,

(ρ) =N

(ρ), (6.60)

with



(2τ +D/2)

n!(n +2τ +D/2)

(ρ) =



(2τ +D/2)

−

. (6.61)

5 Generalization to Pseudoharmonic Oscillator 77

We now study a suitable Lie algebra associated with the operators

and

−

Based on the results (6.58) and (6.59), we can calculate the commutator [

−

[

−

(ρ) =2m

(ρ), (6.62)

where we have introduced the eigenvalue



n +τ +



. (6.63)

We can thus deﬁne the operator



ˆn +τ +



. (6.64)

The operators

and

thus satisfy the commutation relations

[

−

]=2

, [

]=±

, (6.65)

which correspond to an su(1, 1) algebra.

As we know, there are four series of irreducible unitary representations for the

SU(1, 1) group except for the identity representation [214]. Since this quantum sys-

tem has ground state, the representation of the dynamic algebra su(1, 1) belongs to

(j):

|j,ν=ν|j,ν,

−

|j,ν=



(ν +j )(ν −j −1)|j,ν −1,

|j,ν −1=



(ν +j )(ν −j −1)|j,ν,

ν =−j +n, n =0, 1, 2,...,j <0.

(6.66)

In comparison with Eqs. (6.58), (6.59), (6.63) and (6.64), we have j =−(τ +D/4),

ν =n +τ +D/4, and R

(ρ) =|j,ν.

Finally, from the ladder operators

we may obtain two related functions

ρ = 2

−

,ρ

dρ



−



, (6.67)

from which, together with Eqs. (6.58) and (6.59), we have



∞



(r)r

(r)dr =



2n +2τ +





−



n(n +2τ +D/2 −1)δ



,(n−1)

−



(n +1)(n +2τ +D/2)δ



,(n+1)

, (6.68)

and



∞



(r)

(r)dr =



(n +1)(n +2τ +D/2)δ



,(n+1)

−



n(n +2τ +D/2 −1)δ



,(n−1)

−



. (6.69)

78 6 Harmonic Oscillator

5.4 Recurrence Relation

To obtain the recurrence relations among the solutions R

(ρ) given in Eq. (6.51), it

is convenient to employ the associated Laguerre functions satisfying the following

recurrence relation

(n +1)L

n+1

(x) +(n +β)L

n−1

(x) +(x −2n −β −1)L

(x) =0, (6.70)

from which, together with the normalization constant N

we obtain a useful three-

term recurrence relation



(n +1)(2τ +n +D/2)R

n+1

(ρ) +



n(2τ +n +D/2 −1)R

n−1

(ρ)

=(2n +2τ +D/2 −ρ)R

(ρ). (6.71)

6 Position and Momentum Information Entropy

In this section we shall review the position and momentum information entropies of

the D-dimensional harmonic oscillator [58]. An information measure closely related

to the entropy is the Boltzmann-Shannon entropy deﬁned by

=−



ρ(r) ln ρ(r)d(r) (6.72)

in the position space and

=−



γ(p) ln γ(p)d(p) (6.73)

in the momentum space. The position and momentum single-particle densities are

simply given by

ρ(r) =|ψ(r)|

,γ(p) =|ψ(p)|

, (6.74)

where ψ(p) is the Fourier transform of ψ(r). The information entropies S

and S

give a measure of the spread of the single-particle density in position and momentum

space, respectively.

These two entropies have shown to play an important role in the quantum-

mechanical description of physical systems. For example, they have made

Bialynicki-Birula and Mycielski (BBM) ﬁnd a new and stronger version of the

Heisenberg uncertainty relation. The corresponding BBM inequality for single par-

ticle is given by

≥D(1 +ln π). (6.75)

Once the exact solutions of quantum systems are known, it is possible to study them

numerically.

7 Conclusions 79

7 Conclusions

In this Chapter we have made use of previous results given in Chap. 3 to present the

exact solutions of the harmonic oscillator in arbitrary dimensions. The recurrence

relations for the radial functions have been established in terms of the Hamilto-

nian. In addition, we have shown how to realize the dynamic algebra su(1, 1) by the

factorization method. As an important generalization, we have also carried out the

pseudoharmonic oscillator case and constructed the ladder operators directly from

wavefunction by factorization method. The matrix elements of some related func-

tions have been obtained analytically from the ladder operators

. We ﬁnd that

this method represents a simple and elegant approach to obtain them. On the other

hand, a useful recurrence relation among the wavefunction is derived on the basis

of the recurrence relations on the associated Laguerre polynomials. Finally, we give

a useful remark. We note that Oyewumi et al. have restudied pseudoharmonic po-

tential in N -dimensions in 2008 and carried out some expectation values for r

−2

,

r

, T , V(r), H  and p

 as well as virial theorem obtained by means of the

Hellmann-Feynman theorems [102].

Chapter 7

Coulomb Potential

1 Introduction

The exact solutions of the non-relativistic and relativistic equations with a Coulomb

ﬁeld have been the subject both in quantum mechanics and in classical mechanics.

The well-known exact solutions in almost all textbooks [1, 2] are important achieve-

ments at the beginning stage of quantum mechanics, which provided a strong evi-

dence in favor of the quantum theory being correct.

The purpose of this Chapter is three-fold. The ﬁrst is to study the analytical so-

lutions of the D-dimensional Schrödinger equation with a Coulomb potential in

arbitrary dimensions [26, 61–63, 78, 87, 99, 215], the relation between the radial

equations of the D-dimensional hydrogen atom and harmonic oscillator [34]. The

second is to realize the dynamic algebra su(1,1) for the radial Schrödinger Coulomb

potential in terms of the Sturmian bases. The third is to study a generalized case, i.e.,

the Coulomb plus an inverse squared potential and then to analyze the variation of

energy levels E(n,l,D) on the dimension D [216]. As far as the potential energy

term, we use results from scattering experiments to ﬁx its form 1/r. Indeed, since

the results of Rutherford-type scattering experiments are independent of the spatial

dimension, we can unambiguously conclude from the experimental data, that in ar-

bitrary dimension D the potential must be of the form like 1/r. This is of course

consistent with the analysis of Refs. [217, 218] that atoms with the usual kinetic

energy coupled to a modiﬁed potential of the form 1/r

D−2

are not stable, where the

exponent (D −2) is due to the requirement that Gauss’s law should be still valid in

higher dimensions.

This Chapter is organized as follows. Section 2 is devoted to the study of the

D-dimensional Schrödinger equation with a Coulomb potential. We establish the

shift operators in Sect. 3. In Sect. 4 we illustrate the mapping between the Coulomb

potential and harmonic oscillator radial functions. In Sect. 5 we show how to realize

the dynamic algebra su(1, 1) relying only on the radial Schrödinger equation. The

generalized case, i.e., the Coulomb potential plus an inverse squared potential shall

be investigated in Sect. 5. Some concluding remarks are given in Sect. 6.

S.-H. Dong, Wave Equations in Higher Dimensions,

82 7 Coulomb Potential

2 Exact Solutions

The D-dimensional Schrödinger equation ( =M =1) with a spherically symmet-

ric potential V(r)can be written as

−



(x) =[E −V(r)](x). (7.1)

For central potentials, we may separate the angular variables from the radial

one by taking the wavefunction (x) = R

(r)Y

D−2

...l

(

x). Substitution of this into

Eq. (7.1) allows us to obtain the D-dimensional radial Schrödinger equation with a

spherically symmetric potential V(r)

D−1



D−1



(r) +



2E −2V(r)−

l(l +D −2)



(r) =0, (7.2)

where the potential V(r)is taken as a Coulomb potential

V(r)=−

Zα

=−

,ξ=Zα. (7.3)

Upon taking a new variable

ρ = r

√

−8E (7.4)

for the bound states, Eq. (7.2) is rearranged as

(ρ)

dρ

D −1

(ρ)

dρ



−

l(l +D −2)



(ρ) =0, (7.5)

where

τ ≡ξ



−2E

. (7.6)

From the behaviors of the radial function at the origin and at inﬁnity, we deﬁne

R(ρ) =ρ

−ρ/2

F(ρ). (7.7)

Substituting of Eq. (7.7) into Eq. (7.5), we ﬁnd that F(ρ) satisﬁes

F(ρ)

dρ

+(2l +D −1 −ρ)

dF(ρ)

dρ



τ −l −

D −1



F(ρ)=0, (7.8)

whose solutions are nothing but the conﬂuent hypergeometric functions

(a;b;ρ)

with

a =l −τ +

D −1

,b=2l +D −1. (7.9)

Thus, the eigenfunctions can be expressed as

R(ρ) =Nρ

−ρ/2



l −τ +

D −1

;2l +D −1;ρ



, (7.10)

where N is the normalization factor to be determined.

2 Exact Solutions 83

We now discuss the eigenvalues. From the consideration of the ﬁniteness of the

solutions at inﬁnity, the general quantum condition is given by

τ −l −

D −1

=0, 1, 2,.... (7.11)

Introduce a principal quantum number

n =n

+l +1. (7.12)

As a result, we have



−2E

=τ =n +

D −3

. (7.13)

From this we obtain the energy levels

E(n,D) =−

2[n +(D −3)/2]

. (7.14)

Notice that the E(n,D) is independent of the quantum number l. On the other hand,

this result means that a stable hydrogen atom exists when the Coulomb-like potential

is taken as the form Eq. (7.3) in higher dimensions.

For a large D,wehave

E(n,D) −2ξ

−2

−2(2n −3)D

−3

+3(2n −3)

−4

−···}. (7.15)

We now calculate the normalization factor N . Note that n

=τ −l −(D −1)/2

is a non-negative integer. We can express the conﬂuent hypergeometric functions

(−n

;β +1;ρ) by the associated Laguerre polynomials L

(ρ). Based on for-

mulas (5.21), we express the radial function as

R(ρ) =Nρ

−ρ/2

2l+D−2

n−l−1

(ρ). (7.16)

The N can be obtained from the normalization condition



∞

R(ρ)

D−1

dr =1. (7.17)

Before proceeding to do so, we recall a generalized Coulomb-like integral J

(γ )

n,α

given in Ref. [219]

(γ )

n,α



∞

−x

α+γ

(x)]

(α +n +1)



k=0

(−1)

(n −k −γ)(α+k +γ +1)

(−k −γ)(α+k +1)k!(n −k)!

, (7.18)

with Re(α +γ +1)>0. With the help of this integral formula, we are able to obtain

(1)

n,α

(2n +α +1)(α +n +1)

, (7.19)