TMP FD67
TMP FD67
TMP FD67
Abstract
In this work the bound state and scattering problems for a spin-1/2 particle undergone to an Aharonov-Bohm potential
in a conical space in the nonrelativistic limit are considered. The presence of a δ-function singularity, which comes
from the Zeeman spin interaction with the magnetic flux tube, is addressed by the self-adjoint extension method. One
of the advantages of the present approach is the determination of the self-adjoint extension parameter in terms of
physics of the problem. Expressions for the energy bound states, phase-shift and S matrix are determined in terms of
the self-adjoint extension parameter, which is explicitly determined in terms of the parameters of the problem. The
relation between the bound state and zero modes and the failure of helicity conservation in the scattering problem and
its relation with the gyromagnetic ratio g are discussed. Also, as an application, we consider the spin-1/2 Aharonov-
Bohm problem in conical space plus a two-dimensional isotropic harmonic oscillator.
Keywords: Self-adjoint extension, Aharonov-Bohm effect, Bound State, Scattering, Helicity
1. Introduction
The Aharonov-Bohm (AB) effect [1] (first predicted by Ehrenberg and Siday [2]) is one of most weird results
of quantum phenomena. The effect reveals that the electromagnetic potentials, rather than the electric and magnetic
fields, are the fundamental quantities in quantum mechanics. The interest in this issue appears in the different contexts,
such as solid-state physics [3], cosmic strings [4–14] κ-Poincaré-Hopf algebra [15, 16], δ-like singularities [17–19],
supersymmetry [20, 21], condensed matter [22, 23], Lorentz symmetry violation [24], quantum chromodynamics [25],
general relativity [26], nanophysics [27], quantum ring [28–30], black hole [31, 32] and noncommutative theories
[33, 34].
In the AB effect of spin-1/2 particles [7], besides the interaction with the magnetic potential, an additional two
dimensional δ-function appears as the mathematical description of the Zeeman interaction between the spin and the
magnetic flux tube [18, 19]. This interaction is the basis of the spin-orbit coupling, which causes a splitting on the
energy spectrum of atoms depending on the spin state. In Ref. [17] is argued that this δ-function contribution to the
potential can not be neglected when the system has spin, having shown that changes in the amplitude and scattering
cross section are implied in this case. The presence of a δ-function potential singularity, turns the problem more
complicated to be solved. Such kind of point interaction potential can then be addressed by the self-adjoint extension
approach [35]. The self-adjoint extension of symmetric operators [36] is a very powerful mathematical method and
it can be applied to various systems in relativistic and nonrelativistic quantum mechanics, supersymmetric quantum
mechanics and vortex-like models.
This paper extends our previous report [37] on a general physical regularization method, both in details and
depth. The method has the advantage of solving problems in relativistic and nonrelativistic quantum mechanics
whose Hamiltonian is singular. The description of the formalism is based on the works of Kay-Studer (KS) [38] and
Bulla-Gesztesy (BG) [39], both using the self-adjoint extension method. The present method is based on the physics
Email addresses: fmandrade@uepg.br (F. M. Andrade), edilbertoo@gmail.com (E. O. Silva), marciano@uepg.br (M. Pereira)
where δ2 (r) is the two-dimensional δ-function in flat space [45]. This implies a two-dimensional conical singularity
symmetrical in the z-axis, which characterizes it as a linear defect.
In order to study the dynamics of the particle in a non-flat spacetime, we should include the spin connection in
the differential operator and define the respective Dirac matrices in this manifold. The modified Dirac equation in the
curved space reads [46] (~ = c = 1):
h i
iγµ (∂µ + Γµ ) − eγµ Aµ − M Ψ = 0, (3)
where e is the charge, M is mass of the particle, Ψ is a four-component spinorial wave function, and Γµ is the spin
connection given by
1
Γµ = − γ(a) γ(b) eν(a) e(b)ν;µ , (4)
4
2
and γµ = eµ(a) (x)γ(a) are the γ matrices in the curved spacetime. We take the basis tetrad [46–48],
1 0 0 0
0 cos ϕ − sin ϕ/αr 0
µ
e(a) (x) = , (5)
0 sin ϕ cos ϕ/αr 0
0 0 0 1
with gµν = diag(−, +, +, +). For this conical spacetime the spin connection can be expressed by
1−α r
γµ Γµ = − γ, (7)
2αr
and !
0 σr
γr = cos ϕγ(1) + sin ϕγ(2) = . (8)
−σr 0
Moreover the α matrices are now written as
σ(a)
!
σi
!
0 0
αi = ei(a) = , (9)
−σ(a) 0 σi 0
where σi = (σr , σϕ , σz ) are the Pauli matrices in cylindrical coordinates obtained from the basis tetrad (5).
For the specific tetrad basis used here, the spin connection is
We are interested in the nonrelativistic limit of the Dirac equation, so it is convenient to express it in terms of a
Hamiltonian formalism
Ĥψ = Ēψ, (13)
with
Ĥ = αi (−i∇i − eAi ) − iγ0 γµ Γµ + βM. (14)
Exploiting the symmetry under z translations, we can access the (2+1)-dimensional Dirac equation which follows
from the decoupling of (3+1)-dimensional Dirac equation for the specialized case where ∂z = 0 and Az = 0, into two
uncoupled two-component equations, such as implemented in Refs. [49–51]. The Dirac equation in (2+1) dimensions
reads
βγ · Π + βM ψ = Ēψ, (15)
where
1
(∇ + Γ) − eA,
Π= (16)
i
is the generalized momentum, ψ is a two-component spinor, and the (2 + 1) dimensional γ matrices are given in terms
of the Pauli matrices in cylindrical coordinates
Table 1: Summary for the physical scenarios based on the signal of the δ coupling constant for α ∈ (0, 1).
sphere in R2 . So, it is possible to express the eigenfunctions of the two dimensional Hamiltonian in terms of the
eigenfunctions of Jˆz : !
fm (r) eimϕ
ψ(r, ϕ) = , (25)
gm (r) ei(m+1)ϕ
with m + 1/2 = ±1/2, ±3/2, . . ., with m ∈ Z. Inserting this into equation (22), we can extract the radial equation for
fm (r)
H fm (r) = k2 fm (r), (26)
where
φs δ(r)
H = H0 + , (27)
α r
and
d2 1 d [m + φ + (1 − α)/2]2
H0 = − 2
− + . (28)
dr r dr α2 r2
The Hamiltonian in Eq. (27) governs the quantum dynamics of a spin-1/2 charged particle in the conical spacetime,
with a magnetic field B along the z-axis, i.e., a spin-1/2 AB problem in the conical space. We note that in the case of
flat space, α = 1 (no spin connection), we recover the radial Hamiltonian for the usual spin-1/2 AB problem in Refs.
[17, 60],
d2 1 d (m + φ)2 δ(r)
− 2− + + φs . (29)
dr r dr r2 r
For α ∈ (0, 1] we summarize the possible physical scenarios of obtaining scattering and bound states in Table 1, based
on the signal of the δ function coupling constant in (27). Since we have two possibilities for achieving bound states
and scattering, we will focus our attention first on the conditions giving bound states. Afterwards, only when we study
the scattering problem we will take into account the other two conditions.
3. Self-adjoint extensions
In this section we summarize some important concepts and results from the von Neumann-Krein theory of self-
adjoint extensions. We begin by defining an essentially self-adjoint operator. An operator O, with domain D(O), is
said to be essentially self-adjoint if and only if D(O† ) = D(O) and O† = O. For smooth functions ξ ∈ C0∞ (R2 ) with
ξ(0) = 0, we should have
Hξ = H0 ξ, (30)
and hence it is reasonable to interpret [64–66] the Hamiltonian (27) as a self-adjoint extension of
Using the unitary operator V : L2 (R+ , rdr) → L2 (R+ , dr), given by (Vξ)(r) = r1/2 ξ(r), the operator H0 becomes
d2 [m + φ + (1 − α)/2]2 1 1
!
H̃0 = VH0 V −1 = − 2 + − . (32)
dr α2 4 r2
By standard results, the symmetric radial operator H̃0 is essentially self-adjoint for |m + φ + (1 − α)/2|/α ≥ 1. For those
values of m fulfilling |m + φ + (1 − α)/2|/α < 1 it is not essentially self-adjoint, admitting an one-parameter family
5
of self-adjoint extensions [36]. In order to proceed to the self-adjoint extensions of H0 , we must find its deficiency
subspaces, N± , which are defined by
n o
N± = ξ± ∈ D(H0† ), H0† ξ± = z± ξ± , ℑ z± ≷ 0 , (33)
with dimensions n± = dim N± , which are called deficiency indices of H0 [36]. A necessary and sufficient condition
for H0 being essentially self-adjoint is that n+ = n− = 0. On the other hand, if n+ = n− ≥ 1 the operator H0 has an
infinite number of self-adjoint extensions parametrized by the unitary operators U : N+ → N− . Therefore, according
to the von Neumann-Krein theory of self-adjoint extensions, the domain of H0† is given by
One observes that even if the operator is Hermitian H0† = H0 , its domains could be different. The self-adjoint extension
approach consists, essentially, in extending the domain D(H0 ) to match D(H0† ) in (34), turning H0 a self-adjoint
operator. We then have
D(Hη,0 ) = D(H0† ) = D(H0 ) ⊕ N+ ⊕ N− . (35)
where Hη,0 represents the self-adjoint extension of H0 parametrized by η ∈ [0, 2π).
In what follows, to characterize the one parameter family of self-adjoint extension of H0 , we will use the KS [38]
and the BG [39] approaches, both based on boundary conditions. In the KS approach, the boundary condition is a
match of the logarithmic derivatives of the zero-energy solutions for Eq. (26) and the solutions for the problem H0
plus self-adjoint extension. In the BG approach, the boundary condition is a mathematical limit allowing divergent
solutions for the Hamiltonian (28) at isolated points, provided they remain square integrable.
3.1. KS method
In this section, we employ the KS approach to find the bound states for the Hamiltonian in Eq. (27). Following
[38], we temporarily forget the δ-function potential and find the boundary conditions allowed for H0 . For this intent,
we substitute the problem in Eq. (26) by the eigenvalue equation for H0 ,
H0 fρ = k2 fρ , (36)
plus self-adjoint extensions. Here, fρ is labeled by the parameter ρ of the self-adjoint extension, which is related to
the behavior of the wave function at the origin. In order for the H0 to be a self-adjoint operator in Hr , its domain of
definition has to be extended by the deficiency subspace, which is spanned by the solutions of the eigenvalue equation
(cf. Eq. (33))
H0† f± = ±ik02 f± , (37)
where k02 ∈ R is introduced for dimensional reasons. Since H0† = H0 , the only square integrable functions which are
solutions of Eq. (37) are the modified Bessel functions of second kind,
√
f± = K[m+φ+(1−α)/2]/α ( ∓ik0 r), (38)
√
with ℑ ±i > 0. These functions are square integrable only in the range [m + φ + (1 − α)/2]/α ∈ (−1, 1), for which
H0 is not self-adjoint. The dimension of such deficiency subspace is (n+ , n− ) = (1, 1). So, we have two situations for
[m + φ + (1 − α)/2]/α, i.e.,
6
Thus, D(Hρ,0 ) in L2 (R+ , rdr) is given by the set of functions [36]
h √ √ i
fρ (r) = fm (r) + C K|m+φ+(1−α)/2|/α ( −ik0 r) + eiρ K|m+φ+(1−α)/2|/α ( ik0 r) , (41)
where fm (r), with fm (0) = f˙m (0) = 0 ( f˙ ≡ d f /dr), is the regular wave function and the parameter ρ ∈ [0, 2π) represents
a choice for the boundary condition. For each ρ, we have a possible domain for H0 and the physical situation is the
factor that will determine the value of ρ [7, 40, 62, 63]. Thus, to find a fitting for ρ compatible with the physical
situation, a physically motivated form for the magnetic field is preferable for the regularization of the δ-function. This
is accomplished by replacing (19) with [17, 18, 67, 68]
φ
− ϕ̂, r > r0
αr
eA = (42)
0, r < r0 .
This modification mathematically effects the replacement of idealized zero thickness filament by one of a finite very
small radius r0 smaller than the Compton wave length λC of the electron [58]. So one makes the replacement
δ(r) δ(r − r0 )
→ . (43)
r r0
Although the functional structure of δ(r)/r and δ(r − r0 )/r0 are quite different, as discussed in [17], we are free to use
any form of potential once that the specific details of the model (43) can be shown to be irrelevant provided that only
the contribution is independent of angle and has no δ-function contribution at the origin. It should be remarked that
the δ(r − r0 )/r0 is one dimensional and well defined contrary to the two dimensional δ(r)/r.
Now, we are in the position to determine a fitting value for ρ. To do so, we follow [38] and consider the zero-energy
solutions f0 and fρ,0 for H with the regularization in (43) and H0 , respectively, i.e.,
" 2
[m + φ + (1 − α)/2]2 φs δ(r − r0 )
#
d 1 d
− 2− + + f0 = 0, (44)
dr r dr α2 r2 α r0
" 2
[m + φ + (1 − α)/2]2
#
d 1 d
− 2− + fρ,0 = 0. (45)
dr r dr α2 r2
The value of ρ is determined by the boundary condition
f˙0 f˙ρ,0
lim+ r0 r=r = r lim r0 . (46)
r0 →0 f0 0 0 →0
+
fρ,0 r=r0
The left-hand side of this equation can be achieved integrating (44) from 0 to r0 ,
Z r0
φs r0 [m + φ + (1 − α)/2]2 r0 f0 (r)
! Z Z
1 d d f0 (r) δ(r − r0 )
r rdr = f0 (r) rdr + rdr. (47)
0 r dr dr α 0 r0 α2 0 r2
as r0 → 0+ . So, we have
f˙0 φs
lim+ r0 r=r = . (49)
r0 →0 f0 0 α
The right-hand side of Eq. (46) is calculated using the asymptotic representation for Kν (z) in the limit z → 0,
given by
z−ν zν
" #
π
Kν (z) ∼ − , (50)
2 sin(πν) 2−ν Γ(1 − ν) 2ν Γ(1 + ν)
7
in Eq. (41) 1 . Thus, we arrive at
f˙ρ,0 Ω̇ρ (r)
lim+ r0 r=r = r lim , (51)
r0 →0 fρ,0 0 0 →0 Ωρ (r) r=r0
+
where
√ √ √ √
−ik0 r −|m+φ+(1−α)/2|/α
|m+φ+(1−α)/2|/α
ik0 r −|m+φ+(1−α)/2|/α
|m+φ+(1−α)/2|/α
−ik0 r ik0 r
iρ
Ωρ (r) = −|m+φ+(1−α)/2|/α (−) − + e ,
2 Γ 2 |m+φ+(1−α)/2|/α Γ(+)
2 −|m+φ+(1−α)/2|/α Γ(−) 2 |m+φ+(1−α)/2|/α Γ (+)
(52)
where we have introduced the notation Γ(±) = Γ (1 ± |m + φ + (1 − α)/2|/α). Inserting (49) and (51) in (46) we obtain
Ω̇ρ (r) φs
lim+ r=r = , (53)
r0 →0 Ωρ (r) 0 α
which gives us the parameter ρ in terms of the physics of the problem, i.e., the correct behavior of the wave functions
when r → 0+ .
We now determine the bound states for H0 and using (53) the bound state for H will be determined. So, we write
Eq. (36) for the bound state.√In the present system the energy of a bound state has to be negative, so that k is a pure
imaginary, k = iκ, with κ = −2MEb , where Eb < 0 is the bound state energy. Then, with the substitution k → iκ we
have " 2
[m + φ + (1 − α)/2]2
!#
d 1 d 2
+ − + κ fρ (r) = 0, (54)
dr2 r dr α2 r2
The above equation is the modified Bessel equation whose general solution is given by
p
fρ (r) = K|m+φ+(1−α)/2|/α r −2MEb . (55)
Since these solutions belong to D(Hρ,0 ), it is of the form (41) for some ρ selected from the physics of the problem.
So, we substitute (55) into (41) and compute limr0 →0+ r0 f˙ρ / fρ |r=r0 using (50). After a straightforward calculation, we
have the relation
h i
|m + φ + (1 − α)/2|/α r02|m+φ+(1−α)/2|/α Γ(−) (−MEb )|m+φ+(1−α)/2|/α + 2|m+φ+(1−α)/2|/α Γ(+) φs
2|m+φ+(1−α)/2|/α (−)
= . (56)
r Γ (−MEb )|m+φ+(1−α)/2|/α − 2|m+φ+(1−α)/2|/α Γ(+) α
0
Solving the above equation for Eb , we find the sought energy spectrum
" ! #α/|m+φ+(1−α)/2|
2 φs + |m + φ + (1 − α)/2| Γ(1 + |m + φ + (1 − α)/2|/α)
Eb = − 2 . (57)
Mr0 φs − |m + φ + (1 − α)/2| Γ(1 − |m + φ + (1 − α)/2|/α)
Notice that there is no arbitrary parameter in the above equation. Moreover, to ensure that the energy is a real number,
we must have !
φs + |m + φ + (1 − α)/2| Γ(1 + |m + φ + (1 − α)/2|/α)
> 0. (58)
φs − |m + φ + (1 − α)/2| Γ(1 − |m + φ + (1 − α)/2|/α)
This inequality is satisfied if |φs| ≥ |m + φ + (1 − α)/2| and due to |m + φ + (1 − α)/2| < 1 it is sufficient to consider
|φs| ≥ 1. As shown in Table 1, a necessary condition for a δ function to generate an attractive potential, which is able
to support bound states, is that the coupling constant (φs/α) must be negative. Thus, once that α ∈ (0, 1], the existence
of bound states requires
φs ≤ −1. (59)
So, it seems that we must have φs < 0, in such way that the flux and the spin must be antiparallel, and must have a
minimum value for the |φ|.
z−ν zν
h i
1 π
In Ref. [37] the expression used for the asymptotic representation of Kν (z) it was Kν (z) ∼ 2 sin(πν) 2−ν Γ(1−ν) + 2ν Γ(1+ν) , i.e., the signal of the
second term within the brackets must be minus as in Eq. (50).
8
3.2. BG method
The KS approach used in the previews section gives us the energy spectrum in terms of the physics of the problem,
but is not appropriate for dealing with scattering problems. Furthermore, it selects the value for the parameter ρ. On
the other hand, the approach in [39] is suitable to address both bound and scattering scenarios, with the disadvantage
of allowing arbitrary self-adjoint extension parameters. By comparing the results of these two approaches for bound
states, the self-adjoint extension parameter can be determined in terms of the physics of the problem. Here, all self-
adjoint extensions H0,λ j of H0 are parametrized by the boundary condition at the origin [35, 39],
with
where λm is the self-adjoint extension parameter. In [35] it is showen that there is a relation between the self-adjoint
extension parameter λm and the parameter ρ used in the previous section. The parameter ρ is associated with the
mapping of deficiency subspaces and extend the domain of operator to make it self-adjoint, being a mathematical
parameter. The self-adjoint extension parameter λm have a physical interpretation, it represents the scattering length
[69] of H0,λm [35]. For λm = 0 we have the free Hamiltonian (without the δ-function) with regular wave functions at
the origin and for λm , 0 the boundary condition in (60) permit a r−|m+φ+(1−α)/2|/α singularity in the wave functions at
the origin.
for r , 0, taking into account both cases in (39) simultaneously, can be written in terms of the confluent hypergeo-
metric function of the first kind M(a, b, z) as
!
1 |m + φ + (1 − α)/2| |m + φ + (1 − α)/2|
fm (r) = am e−ikr (2ikr)|m+φ+(1−α)/2|/α M + ,1+2 , 2ikr
2 α α
!
−ikr −|m+φ+(1−α)/2|/α 1 |m + φ + (1 − α)/2| |m + φ + (1 − α)/2|
+ bm e (2ikr) M − ,1− 2 , 2ikr , (63)
2 α α
where am , bm are the coefficients of the regular and irregular solutions, respectively. By implementing Eq. (63) into
the boundary condition (60), we derive the following relation between the coefficients am and bm :
λm k2
λm am = (2ik)−2|m+φ+(1−α)/2|/α bm 1 + lim+ r2−2|m+φ+(1−α)/2|/α . (64)
4(1 − |m + φ + (1 − α)/2|/α) r→0
9
The bound state wave function is obtained with the substitution k → iκ. So we have
!
1 |m + φ + (1 − α)/2| |m + φ + (1 − α)/2|
fmb (r) = am eκr (−2κr)|m+φ+(1−α)/2|/α M + ,1+2 , −2κr
2 α α
!
1 |m + φ + (1 − α)/2| |m + φ + (1 − α)/2|
+ bm eκr (−2κr)−|m+φ+(1−α)/2|/α M − ,1− 2 , −2κr . (65)
2 α α
In order to be a bound state fmb (r) must vanish at large r, i.e., it must be normalizable. By using the asymptotic
representation of M(a, b, z) for z → ∞,
Γ(b) z a−b Γ(b)
M(a, b, z) ∼ ez + (−z)−a , (66)
Γ(a) Γ(b − a)
the normalizability condition yields the relation
bm Γ(+)
= −16|m+φ+(1−α)/2|/α (−) . (67)
am Γ
From Eq. (64), for |m + φ + (1 − α)/2|/α < 1 we have
bm
= λm (−2κ)2|m+φ+(1−α)/2|/α . (68)
am
Combining these two later equations, the bound state energy is determined,
" #α/|m+φ+(1−α)/2|
2 1 Γ(1 + |m + φ + (1 − α)/2|/α)
Eb = − − . (69)
M λm Γ(1 − |m + φ + (1 − α)/2|/α)
This coincides with Eq. (3.13) of Ref. [60] for α = 1, i.e., the spin-1/2 AB problem in flat space. Also, this coincides
with Eq. (26) of Ref. [59] for the bound states energy for particles with an anomalous magnetic moment (with the
replacement λ → 1/λm in that reference).
By comparing Eq. (69) with Eq. (57), we find
!
1 1 φs + |m + φ + (1 − α)/2|
= − 2|m+φ+(1−α)/2|/α . (70)
λm r φs − |m + φ + (1 − α)/2|
0
We have thus attained a relation between the self-adjoint extension parameter and the physical parameters of the
problem. It should be mentioned that some relations involving the self-adjoint extension parameter and the δ-function
coupling constant were previously obtained by using Green’s function in Ref. [61] and the renormalization technique
in Ref. [56], being both, however, deprived from a clear physical interpretation. Also, in Ref. [59] a relation between
the self-adjoint extension parameter and the anomaly magnetic moment was found and it is commented that the
dimension of the self-adjoint extension parameter is r2|m+φ+(1−α)/2|/α but does not show an explicit relation as found by
us in Eq. (70).
In Ref. [58] the authors comment, based on a result of Aharonov and Casher [72] which states that in a cylindrical
magnetic field with flux φ a charged particle with magnetic moment and gyromagnetic ratio g = 2 possesses N (N
being the number of entire flux quanta in φ) zero modes, i.e., normalizable states with zero binding energy, any
additional attractive force which occurs for g > 2 turns the zero modes into bound states. This g > 2 value is related
with the self-adjoint extension value, i.e., different values for the self-adjoint extension parameter corresponds to
different values of the g [58]. The explicit relation between the self-adjoint extension parameter and the g will be
subject of a future work.
Moreover, the bound state wave function is given by
p
fmb (r) = Nm K|m+φ+(1−α)/2|/α (− −2MEb r), (71)
Now, comparing the arguments of the cosines above, the following phase shift is achieved:
λs
δmm (k, φ) = ∆mAB (φ) + θλms , (86)
where
π
∆mAB (φ) = (|m| − |m + φ|), (87)
2
is the usual phase shift of the AB scattering and
λs
θλms = arctan µmm (k, φ) . (88)
m λs
Therefore, the scattering operator S φ,m (S matrix) for the self-adjoint extension is
s
λm
λs AB
(φ) 2iθλms
m
S φ,m = e2iδm (k,φ)
= e2i∆m e , (89)
that is, s
1 + iµλmm (k, φ)
λms 2i∆mAB (φ) .
S φ,m =e s (90)
1 − iµλmm (k, φ)
Using Eq. (80), we have
λ s k2|m+φ+(1−α)/2|/α Γ(−) ei|m+φ+(1−α)/2|/απ + 4|m+φ+(1−α)/2|/α Γ(+)
λs AB m
m
S φ,m = e2i∆m (φ)
. (91)
λms k2|m+φ+(1−α)/2|/α Γ(−) e−i|m+φ+(1−α)/2|/απ + 4|m+φ+(1−α)/2|/α Γ(+)
Hence, for any value of the self-adjoint extension parameter λms , there is an additional scattering. If λms = 0, we
achieve the corresponding result for the usual AB problem with Dirichlet boundary condition; in this case, we recover
AB
0 ∞
the expression for the scattering matrix found in Ref. [75], S φ,m = e2i∆m (φ) . If we make λms = ∞, we get S φ,m =
AB
2i∆m (φ)+2iπ|m+φ+(1−α)/2|/α
e .
In accordance with the general theory of scattering, the poles of the S matrix in the upper half of the complex
plane [76] determine the positions of the bound states in the energy scale, Eq. (69). These poles occur when the
denominator of Eq. (91) is equal to zero with the replacement k → iκ. So, we have
For the special case of α = 1 (flat space) and φ = 0 (zero magnetic flux) we have f01 (k, ϕ) = 0, as it should be.
In the above equation we can see that it differs from the usual AB scattering amplitude off a thin solenoid because
its energy dependence. As Goldhaber [77] observed, since the only length scale in the nonrelativistic problem is
set by 1/k, it follows that the scattering amplitude would be a function of the angle alone, multiplied by 1/k. This
is the manifestation of the helicity conservation. So, the inevitable failure of helicity conservation expressed in Eq.
(95) shows that the singularity must lead to inconsistencies if the Hamiltonian and the helicity operator, ĥ = Σ · Π,
are treated as well as well-defined operators whose commutation away from the singularity implies commutation
everywhere [78–81]. After separation of the variables used in (25), the helicity operator is
!
s([m + φ + (1 − α)/2 + 1]/α)
0 −i ∂ r +
r
ĥ = . (96)
!
s[m + φ + (1 − α)/2]/α
−i ∂r − 0
r
This operator suffers from the same issue as the Hamiltonian operator in the interval |m + φ + (1 − α)/2|/α < 1, i.e. it is
not self-adjoint [82, 83]. Defined on a finite interval [0, L], ĥ can be interpreted as a self-adjoint operator on functions
satisfying ξ(L) = eiθ ξ(0). However, because the helicity operator must be defined on an infinite interval [0, ∞), ĥ has
no self-adjoint extension [84], and consequently need not be conserved, and the helicity can leak at the origin [77, 79].
As already commented at the end of Section 3.2.1, this result could be compared with those obtained in Ref. [59]
where the self-adjoint extension parameter was obtained as a function of anomaly of the magnetic moment. In an
idealized version of g − 2 experiment, change in the helicity in a magnetic field becomes a measure of the departure
of the gyromagnetic ratio of the electron or muon from the Dirac value of 2e/2M [85]. For vanishing of g − 2 there
could be no change of helicity even if the magnetic field were inhomogeneous on a very short length scale. So, once
again, different values for the self-adjoint extension is related to different values of the g.
In this section, an application of our method is presented. We address the spin 1/2 AB problem in conical space
plus a two dimensional harmonic oscillator. After including the harmonic oscillator (HO) potential and by using the
angular momentum decomposition, !
χm (r) eimϕ
Φ(r, ϕ) = , (97)
ζm (r) ei(m+1)ϕ
the radial equation for χm (r) becomes
Hχm (r) = k2 χm (r), (98)
where
φs δ(r)
H = H0 + M 2 ω2 r2 + (99)
α r
with ω the angular frequency, and H0 given by (28). In order to have a more detailed analysis of this problem, we
will first examine the motion of the particle considering two cases (i) excluding the r = 0 region and (ii) including the
r = 0 region afterwards. At the end, we compare with some results in the literature.
13
4.1. Solution of the problem excluding r = 0 region
In this case, the Hamiltonian (99) does not include the delta function potential. By directly solving (98) we obtain
[74]
!
1/2+[m+φ+(1−α)/2]/2α [m+φ+(1−α)/2]/α −Mωr2 /2 m + φ + (1 − α)/2 2
χm (r) = am (Mω) r e M d, 1 + , Mωr
α
!
1/2+[m+φ+(1−α)/2]/2α [m+φ+(1−α)/2]/α −Mωr2 /2 m + φ + (1 − α)/2 2
+ bm (Mω) r e U d, 1 + , Mωr , (100)
α
where !
1 m + φ + (1 − α)/2 E
d= 1+ − , (101)
2 α 2ω
U(a, b, z) is the confluent hypergeometric function of the second kind, and am , bm are constants. However, as only
M(a, b, z) is regular at the origin, it should be imposed bm = 0. Moreover, if d is 0 or a negative integer the series
terminates and the hypergeometric function becomes a polynomial of degree n [74]. This condition guarantees that
the confluent hypergeometric function is regular at the origin, which is essential for the treatment of the physical
system since the region of interest is that around the flux tube. Therefore, the series in (100) must converge if we
consider that d = −n, n ∈ Z∗ , with Z∗ denoting the set of the nonnegative integers. This condition also guarantees
the normalizability of the wave function. So, using this condition, we obtain the discrete values for the energy whose
expression is given by !
|m + φ + (1 − α)/2|
Eb = 2n + 1 + ω, n = 0, 1, 2, . . . . (102)
α
The bound state wave function is given by
!
1/2+[m+φ+(1−α)/2]/2α |m+φ+(1−α)/2|/α −Mωr2 /2 m + φ + (1 − α)/2
χbm (r) = Nm (Mω) r e M −n, 1 + 2
, Mωr , (103)
α
with Nm a normalization constant. Notice that in Eq. (102), [m + φ + (1 − α)/2]/α can assume any value. However, we
will see that this condition is no longer true when we include the δ function. Next to study the motion of the particle
in all space, including the r = 0 region, the self-adjoint extension approach is invoked.
plus self-adjoint extensions, with χϑ labeled by a parameter ϑ. The solution to this equation is given in (100). However,
the only square integrable function is U(d, 1 + [m + φ + (1 − α)/2]/α, Mωr2 ). Then, this implies that am = 0 in Eq.
(100), and we have
!
2 m + φ + (1 − α)/2
χϑ (r) = (Mω)1/2+[m+φ+(1−α)/2]/2α r[m+φ+(1−α)/2]/α e−Mωr /2 U d, 1 + , Mωr2 . (105)
α
In order to guarantee that χ(r) ∈ L2 (R, rdr), it is advisable to study their behavior as r → 0, which implies analyz-
ing the possible self-adjoint extensions. Now, to construct the self-adjoint extensions, we must find the deficiency
subspaces,
[H0 + M 2 ω2 r2 ]† χ± (r) = ±ik02 χ± (r). (106)
The solution to this equation is
!
2 m + φ + (1 − α)/2
χ± (r) = r[m+φ+(1−α)/2]/α e−Mωr /2
U d± , 1 + , Mωr2 , (107)
α
14
where !
1 m + φ + (1 − α)/2 ik0
d± = 1+ ∓ . (108)
2 α 2ω
Now considering the asymptotic behavior of U(a, b, z) as z → 0 [74], let us find under which condition the term,
Z
|χ± (r)|2 rdr, (109)
has a finite contribution near the origin region. Using Eq. (107) we found
h i
lim |χ± (r)|2 r1+2[m+φ+(1−α)/2]/α → A1 r1+2[m+φ+(1−α)/2]/α + A2 r1−2[m+φ+(1−α)/2]/α , (110)
r→0
where A1 and A2 are constants. Studying Eq. (110), we see that χ± (r) is square-integrable only for [m + φ + (1 −
α)/2]/α ∈ (−1, 1). In this case, since N+ is expanded by χ+ (r) only, we have that its dimension is n+ = 1. The same
applies to N− and χ− (r), resulting in n− = 1 . Then, the Hilbert space, for both cases of Eq. (39), contains vectors of
the form
( !
2 |m + φ + (1 − α)/2|
χϑ (r) = χm (r) + c r|m+φ+(1−α)/2|/α e−Mωr /2 U d+ , 1 + , Mωr2
α
!)
2 |m + φ + (1 − α)/2|
+ eiϑ r|m+φ+(1−α)/2|/α e−Mωr /2 U d− , 1 + , Mωr2 , (111)
α
where c is an arbitrary complex number, χm (0) = χ̇m (0) = 0 and χm (r) ∈ L2 (R+ , rdr). For a range of ϑ, the behavior
of the wave functions (111) was addressed in [86]. The boundary condition at the origin will select the value of this
parameter. The difference here is the presence of the harmonic term. However, this harmonic term does not contribute
to the BG logarithmic derivative boundary condition (cf. Eq. (46)), since the integration of the harmonic vanishes as
r0 → 0+ . After this identification, proceeding in an analogous way we did in the Section 3.1, it is found that the bound
state energy is implicitly determined by the equation
!
Γ (1/2 + |m + φ + (1 − α)/2|/2α − Eb /2ω) 1 φs + |m + φ + (1 − α)/2|
= −
Γ (1/2 − |m + φ + (1 − α)/2|/2α − Eb /2ω) (Mω)|m+φ+(1−α)/2|/α r02|m+φ+(1−α)/2|/α φs − |m + φ + (1 − α)/2|
Γ (1 + |m + φ + (1 − α)/2|/α)
× . (112)
Γ (1 − |m + φ + (1 − α)/2|/α)
The above expression is too complicated to evaluate the bound state energy, but its limiting features are interesting. If
we take limit r0 → 0 in this expression, the bound state energy are determined by the poles of the gamma functions,
i.e.,
!
m + φ + (1 − α)/2 |m + φ + (1 − α)/2|
−1 < < 0, Eb = 2n + 1 − ω, (113)
α α
!
m + φ + (1 − α)/2 |m + φ + (1 − α)/2|
0< < 1, Eb = 2n + 1 + ω, (114)
α α
or !
|m + φ + (1 − α)/2|
Eb = 2n + 1 ± ω, n = 0, 1, 2, . . . . (115)
α
The + (−) sign refers to solutions which are regular (irregular) at the origin. This result coincide with the Eq. (1) of
Ref. [87], for the special case of α = 1. Another interesting case is that of vanishing HO potential. This is achieved
using the asymptotic behavior of the ratio of gamma functions for ω → 0 [88],
15
which holds for E < 0 and this is the necessary condition for the usual AB system have a bound state. Using this limit
in the Eq. (112), one finds
" ! #α/|m+φ+(1−α)/2|
2 φs + |m + φ + (1 − α)/2| Γ(1 + |m + φ + (1 − α)/2|/α)
Eb = − , (117)
Mr02 φs − |m + φ + (1 − α)/2| Γ(1 − |m + φ + (1 − α)/2|/α)
in agreement with the result obtained in Eq. (57). Thus, in the limit of vanishing harmonic oscillator, we recover the
usual AB problem in conical space, as it should be.
Now we have to remark that this result contains a subtlety that must be interpreted as follows: the presence of the
singularity in the problem establishes the range |m + φ + (1 − α)/2|/α < 1. If we ignore the singularity and impose
that the wave function is regular at the origin (χm (r) ≡ χ̇m (r) ≡ 0), we achieve the same spectrum of Eq. (115), but
with [m + φ + (1 − α)/2]/α assuming any value [89–91]. In this sense the self-adjoint extension prevents us from
obtaining a spectrum incompatible with the singular nature of the Hamiltonian when we take into account the singular
δ function [92, 93]. We have to take into account that the true boundary condition is that the wave function must be
square-integrable through all space, regardless it is irregular or regular at the origin [38, 93].
From Eq. (119), for |m + φ + (1 − α)/2|/α < 1 we have bm = λm (Mω)|m+φ+(1−α)/2|/α am and by using Eq. (120), the
bound state energy is implicitly determined by the equation
Γ (1/2 + |m + φ + (1 − α)/2|/2α − Eb /2ω) 1 Γ(1 + |m + φ + (1 − α)/2|/α)
=− |m+φ+(1−α)/2|/α
(121)
Γ (1/2 − |m + φ + (1 − α)/2|/2α − Eb /2ω) λm (Mω) Γ(1 − |m + φ + (1 − α)/2|/α)
This results coincides with Eq. (53) of Ref. [61] for α = 1, and using the result in Eq. (116), it is easy to see that in
the limit of vanishing oscillator potential, the spectrum of the usual AB is recovered (cf. Eq. (69)). By comparing Eq.
(121) with Eq. (112), we find
!
1 2 φs + |m + φ + (1 − α)/2|
= 2|m+φ+(1−α)/2|/α . (122)
λm r φs − |m + φ + (1 − α)/2|
0
16
Then, the relation between the self-adjoint extension parameter and the physics of the problem for the usual AB has
the same mathematical structure as for the AB plus HO. However, we must observe that the self-adjoint extension
parameter is negative for the usual AB, confirming the restriction of negative values of the self-adjoint extension
parameter made in [61], in such way we have an attractive δ-function. It is a necessary condition to have a bound state
in the usual AB system.
5. Conclusions
We have presented a general regularization procedure to address systems endowed with a singular Hamiltonian
(due to localized fields sources or quantum confinement). Using the KS approach, the bound states were determined
in terms of the physics of the problem, in a very consistent way and without any arbitrary parameter. In the sequel,
we employed the BG approach. By comparing the results of these approaches, we have determined an expression for
the self-adjoint extension parameter for the bound state problem, which coincides with the one for scattering problem.
We have thus obtained the S matrix in terms of the physics of the problem as well. In this point, we remark that the
important results of Refs. [7, 52, 60] are given in terms of an arbitrary self-adjoint extension parameter. In our work
this parameter was determined in terms of the physics of the problem. The outcomes obtained by Park are a particular
case of our results for a fixed value of the self-adjoint extension parameter. To our knowledge, it was not known in
the literature an expression for the bound state energies for the AB with a defined self-adjoint extension parameter. In
Ref. [37] this expression was presented by the first time, whose details are derived here.
To illustrate the applicability of our approach to other physical systems, we deal with the spin-1/2 AB problem
in conical space plus a two dimensional HO. Two cases were considered: (i) without and (ii) with the inclusion of
the δ function potential in the nonrelativistic Hamiltonian. Even though we have obtained an equivalent mathematical
expression for both cases, it has been shown that, in (i) [m + φ + (1 − α)/2]/α can assume any value while in (ii) it is
in the range |m + φ + (1 − α)/2|/α| < 1. In the first case, it is reasonable to impose that the wave function vanish at the
origin. However, this condition does not give a correct description of the problem in the r = 0 region. Therefore, the
energy spectrum obtained in the second case is physically acceptable. The presence of the singularity establishes that
the effective angular momentum must obey the condition |m + φ + (1 − α)/2|/α < 1 and implies that irregular solutions
must be taken into account in this range. The only situation in which we can neglect the δ function potential is that
one in which one looks only for topological phases. A natural extension of this work is the inclusion of the Coulomb
potential, which naturally appears in two-dimensional systems, such as graphene [94] and anyons systems [95, 96].
This will be reported elsewhere.
Acknowledgments
The authors would like to thank M. M. Ferreira Jr., E. R. Bezerra de Mello and F. Moraes for the critical reading
of the manuscript and for helpful discussions. F. M Andrade acknowledges research grants by Fundação Araucária
project no. 205/2013, E. O. Silva acknowledges research grants by CNPq-(Universal) project no. 484959/2011-5 and
M. Pereira acknowledges research grants by Fundação Araucária.
References
[1] Y. Aharonov, D. Bohm, Phys. Rev. 115 (1959) 485. doi:10.1103/PhysRev.115.485.
[2] W. Ehrenberg, R. E. Siday, Proc. Phys. Soc. B 62 (1949) 8–. doi:10.1088/0370-1301/62/1/303.
[3] R. L. Kronig, W. G. Penney, Proc. R. Soc. Lond. A 130 (1931) 499. doi:10.1098/RSPA.1931.0019.
[4] M. Nouri-Zonoz, A. Parvizi, Phys. Rev. D 88 (2013) 023004–. doi:10.1103/PhysRevD.88.023004.
[5] A. Vilenkin, E. P. S. Shellard, Cosmic Strings and Other Topological Defects, Cambridge University Pres, Canbridge, 2000.
[6] Y.-Z. Chu, H. Mathur, T. Vachaspati, Phys. Rev. D 82 (2010) 063515–. doi:10.1103/PhysRevD.82.063515.
[7] P. de Sousa Gerbert, Phys. Rev. D 40 (1989) 1346. doi:10.1103/PhysRevD.40.1346.
[8] V. B. Bezerra, Braz. J. Phys. 36 (2006) 141 – 156. doi:10.1590/S0103-97332006000200006.
[9] M. A. Hohensee, B. Estey, P. Hamilton, A. Zeilinger, H. Müller, Phys. Rev. Lett. 108 (2012) 230404–.
doi:10.1103/PhysRevLett.108.230404.
[10] V. Bezerra, Ann. Phys. (NY) 203 (1990) 392 – 409. doi:10.1016/0003-4916(90)90175-N.
[11] V. B. Bezerra, J. Math. Phys. 38 (1997) 2553–2564. doi:10.1063/1.531995.
[12] C. Furtado, V. B. Bazerra, F. Moraes, Mod. Phys. Lett. A 15 (2000) 253. doi:10.1142/S0217732300000244.
17
[13] G. de A. Marques, V. B. Bezerra, Mod. Phys. Lett. A 19 (2004) 49. doi:10.1142/S021773230401237X.
[14] A. Aliev, D. Gal’tsov, Ann. Phys. 193 (1989) 142–165. doi:10.1016/0003-4916(89)90355-2.
[15] P. Roy, R. Roychoudhury, Phys. Lett. B 359 (1995) 339 – 342. doi:10.1016/0370-2693(95)01079-6.
[16] F. M. Andrade, E. O. Silva, Phys. Lett. B 719 (2013) 467–471. doi:10.1016/j.physletb.2013.01.062.
[17] C. R. Hagen, Phys. Rev. Lett. 64 (1990) 503. doi:10.1103/PhysRevLett.64.503.
[18] C. R. Hagen, Phys. Rev. D 48 (1993) 5935. doi:10.1103/PhysRevD.48.5935.
[19] C. R. Hagen, D. K. Park, Ann. Phys. (NY) 251 (1996) 45. doi:10.1006/aphy.1996.0106.
[20] F. Correa, H. Falomir, V. Jakubský, M. S. Plyushchay, Ann. Phys. (NY) 325 (2010) 2653–2667. doi:10.1016/j.aop.2010.06.005.
[21] V. Jakubský, L.-M. Nieto, M. S. Plyushchay, Phys. Lett. B 692 (2010) 51–56. doi:10.1016/j.physletb.2010.07.014.
[22] A. O. Slobodeniuk, S. G. Sharapov, V. M. Loktev, Phys. Rev. B 84 (2011) 125306. doi:10.1103/PhysRevB.84.125306.
[23] A. O. Slobodeniuk, S. G. Sharapov, V. M. Loktev, Phys. Rev. B 82 (2010) 075316–. doi:10.1103/PhysRevB.82.075316.
[24] H. Belich, E. O. Silva, M. M. Ferreira Jr., M. T. D. Orlando, Phys. Rev. D 83 (2011) 125025–. doi:10.1103/PhysRevD.83.125025.
[25] K. Hashimoto, N. Iizuka, Phys. Rev. D 82 (2010) 105023–. doi:10.1103/PhysRevD.82.105023.
[26] S. R. Dolan, E. S. Oliveira, L. C. Crispino, Physics Letters B 701 (2011) 485–489. doi:http://dx.doi.org/10.1016/j.physletb.2011.06.013.
[27] A. P. Dmitriev, I. V. Gornyi, V. Y. Kachorovskii, D. G. Polyakov, Phys. Rev. Lett. 105 (2010) 036402–.
doi:10.1103/PhysRevLett.105.036402.
[28] J. Schelter, B. Trauzettel, P. Recher, Physical Review Letters 108 (2012) 106603–. doi:10.1103/PhysRevLett.108.106603.
[29] A. Tanaka, T. Cheon, Phys. Rev. A 82 (2010) 022104–. doi:10.1103/PhysRevA.82.022104.
[30] D. Baltateanu, Phys. Lett. A 375 (2011) 2952–2957. doi:10.1016/j.physleta.2011.06.033.
[31] M. A. Anacleto, F. A. Brito, E. Passos, Phys. Rev. D 87 (2013) 125015–. doi:10.1103/PhysRevD.87.125015.
[32] M. A. Anacleto, F. A. Brito, E. Passos, Phys. Rev. D 86 (2012) 125015. doi:10.1103/PhysRevD.86.125015.
[33] A. Das, H. Falomir, M. Nieto, J. Gamboa, F. Méndez, Phys. Rev. D 84 (2011) 045002. doi:10.1103/PhysRevD.84.045002.
[34] H. Falomir, J. Gamboa, M. Loewe, F. Méndez, J. C. Rojas, Phys. Rev. D 66 (2002) 045018–. doi:10.1103/PhysRevD.66.045018.
[35] S. Albeverio, F. Gesztesy, R. Hoegh-Krohn, H. Holden, Solvable Models in Quantum Mechanics, second ed., AMS Chelsea Publishing,
Providence, RI, 2004.
[36] M. Reed, B. Simon, Methods of Modern Mathematical Physics. II. Fourier Analysis, Self-Adjointness., Academic Press, New York - London,
1975.
[37] F. M. Andrade, E. O. Silva, M. Pereira, Phys. Rev. D 85 (2012) 041701(R). doi:10.1103/PhysRevD.85.041701.
[38] B. S. Kay, U. M. Studer, Commun. Math. Phys. 139 (1991) 103. doi:10.1007/BF02102731.
[39] W. Bulla, F. Gesztesy, J. Math. Phys. 26 (1985) 2520. doi:10.1063/1.526768.
[40] C. Filgueiras, E. O. Silva, F. M. Andrade, J. Math. Phys. 53 (2012) 122106. doi:10.1063/1.4770048. arXiv:1205.1155.
[41] E. O. Silva, F. M. Andrade, Europhys. Lett. 101 (2013) 51005–. doi:10.1209/0295-5075/101/51005.
[42] E. O. Silva, F. M. Andrade, C. Filgueiras, H. Belich, Eur. Phys. J. C 73 (2013) 2402. doi:10.1140/epjc/s10052-013-2402-1.
[43] F. M. Andrade, E. O. Silva, T. Prudêncio, C. Filgueiras, J. Phys. G: Nucl. Part. Phys. 40 (2013) 075007–.
doi:10.1088/0954-3899/40/7/075007.
[44] C. Filgueiras, F. Moraes, Phys. Lett. A 361 (2007) 13. doi:10.1016/j.physleta.2006.09.030.
[45] D. D. Sokolov, A. A. Starobinski, Sov. Phys. Dokl. 22 (1977) 312.
[46] E. R. Bezerra de Mello, J. High Energy Phys. 2004 (2004) 016–. doi:10.1088/1126-6708/2004/06/016.
[47] K. Bakke, J. R. Nascimento, C. Furtado, Phys. Rev. D 78 (2008) 064012–. doi:10.1103/PhysRevD.78.064012.
[48] K. Bakke, C. Furtado, Ann. Phys. (Berlin) 522 (2010) 447–455. doi:10.1002/andp.201000043.
[49] H. J. de Vega, Phys. Rev. D 18 (1978) 2932–2944. doi:10.1103/PhysRevD.18.2932.
[50] R. H. Brandenberger, A.-C. Davis, A. M. Matheson, Nucl. Phys. B 307 (1988) 909–923. doi:10.1016/0550-3213(88)90112-5.
[51] M. G. Alford, F. Wilczek, Phys. Rev. Lett. 62 (1989) 1071–1074. doi:10.1103/PhysRevLett.62.1071.
[52] K. Kowalski, K. Podlaski, J. Rembieliński, Phys. Rev. A 66 (2002) 032118. doi:10.1103/PhysRevA.66.032118.
[53] C. R. de Oliveira, Intermediate Spectral Theory and Quantum Dynamics, Birkhäuser Basel, Switzerland, 2008.
doi:10.1007/978-3-7643-8795-2.
[54] N. I. Akhiezer, I. M. Glazman, Theory of Linear Operators in Hilbert Space, Dover, New York, 1993.
[55] P. de Sousa Gerbert, R. Jackiw, Commun. Math. Phys. 124 (1989) 229. doi:10.1007/BF01219196.
[56] R. Jackiw, Diverse topics in theoretical and mathematical physics, Advanced Series in Mathematical Physics, World Scientific, Singapore,
1995.
[57] S. Voropaev, D. Galtsov, D. Spasov, Phys. Lett. B 267 (1991) 91–94. doi:10.1016/0370-2693(91)90529-Y.
[58] M. Bordag, S. Voropaev, Phys. Lett. B 333 (1994) 238–244. doi:10.1016/0370-2693(94)91037-5.
[59] M. Bordag, S. Voropaev, J. Phys. A 26 (1993) 7637–. doi:10.1088/0305-4470/26/24/032.
[60] D. K. Park, J. G. Oh, Phys. Rev. D 50 (1994) 7715. doi:10.1103/PhysRevD.50.7715.
[61] D. K. Park, J. Math. Phys. 36 (1995) 5453. doi:10.1063/1.531271.
[62] C. Filgueiras, F. Moraes, Ann. Phys. (NY) 323 (2008) 3150–3157. doi:10.1016/j.aop.2008.08.002.
[63] C. Filgueiras, E. O. Silva, W. Oliveira, F. Moraes, Ann. Phys. (NY) 325 (2010) 2529. doi:10.1016/j.aop.2010.05.012.
[64] F. Gesztesy, S. Albeverio, R. Hoegh-Krohn, H. Holden, J. Reine Angew. Math. 380 (1987) 87. doi:10.1515/crll.1987.380.87.
[65] L. Dabrowski, P. Stovicek, J. Math. Phys. 39 (1998) 47–62. doi:10.1063/1.532307.
[66] R. Adami, A. Teta, Lett. Math. Phys. 43 (1998) 43–54. doi:10.1023/A:1007330512611.
[67] C. R. Hagen, Phys. Rev. Lett. 64 (1990) 2347–2349. doi:10.1103/PhysRevLett.64.2347.
[68] C. R. Hagen, Int. J. Mod. Phys. A 6 (1991) 3119. doi:10.1142/S0217751X91001520.
[69] J. J. Sakurai, J. Napolitano, Modern Quantum Mechanics, 2nd ed. ed., Addison-Wesley, 2011.
[70] J. Audretsch, U. Jasper, V. D. Skarzhinsky, J. Phys. A 28 (1995) 2359. doi:10.1088/0305-4470/28/8/026.
[71] F. A. B. Coutinho, Y. Nogami, J. Fernando Perez, Phys. Rev. A 46 (1992) 6052. doi:10.1103/PhysRevA.46.6052.
18
[72] Y. Aharonov, A. Casher, Phys. Rev. A 19 (1979) 2461–2462. doi:10.1103/PhysRevA.19.2461.
[73] C. R. de Oliveira, M. Pereira, J. Phys. A 43 (2010) 354011. doi:10.1088/1751-8113/43/35/354011.
[74] M. Abramowitz, I. A. Stegun (Eds.), Handbook of Mathematical Functions, New York: Dover Publications, 1972.
[75] S. N. M. Ruijsenaars, Ann. Phys. (NY) 146 (1983) 1. doi:10.1016/0003-4916(83)90051-9.
[76] K. Bennaceur, J. Dobaczewski, M. Ploszajczak, Phys. Rev. C 60 (1999) 034308–. doi:10.1103/PhysRevC.60.034308.
[77] A. S. Goldhaber, Phys. Rev. D 16 (1977) 1815–1827. doi:10.1103/PhysRevD.16.1815.
[78] Y. Kazama, C. N. Yang, A. S. Goldhaber, Phys. Rev. D 15 (1977) 2287–2299. doi:10.1103/PhysRevD.15.2287.
[79] B. Grossman, Phys. Rev. Lett. 50 (1983) 464–467. doi:10.1103/PhysRevLett.50.464.
[80] N. Ganoulis, Phys. Lett. B 298 (1993) 63–68. doi:10.1016/0370-2693(93)91708-U.
[81] A. Davis, A. Martin, N. Ganoulis, Nucl. Phys. B 419 (1994) 323–340. doi:10.1016/0550-3213(94)90045-0.
[82] F. A. B. Coutinho, J. F. Perez, Phys. Rev. D 49 (1994) 2092–2097. doi:10.1103/PhysRevD.49.2092.
[83] V. S. Araujo, F. A. B. Coutinho, J. F. Perez, J. Phys. A 34 (2001) 8859–. doi:10.1088/0305-4470/34/42/310.
[84] G. Bonneau, J. Faraut, G. Valent, Am. J. Phys. 69 (2001) 322. doi:10.1119/1.1328351.
[85] J. J. Sakurai, Advanced Quantum Mechanics, Addison Wesley, 1967.
[86] H. D. Doebner, H. J. Elmers, W. F. Heidenreich, J. Math. Phys. 30 (1989) 1053–1059. doi:10.1063/1.528375.
[87] T. Blum, C. R. Hagen, S. Ramaswamy, Phys. Rev. Lett. 64 (1990) 709–712. doi:10.1103/PhysRevLett.64.709.
[88] F. W. J. Olver, D. W. Lozier, R. F. Boisvert, C. W. Clark (Eds.), NIST Handbook of Mathematical Functions, Cambridge University Press,
2010.
[89] C. Furtado, B. G. da Cunha, F. Moraes, E. de Mello, V. Bezzerra, Phys. Lett. A 195 (1994) 90–94. doi:10.1016/0375-9601(94)90432-4.
[90] C. Furtado, F. Moraes, J. Phys. A 33 (2000) 5513–. doi:10.1088/0305-4470/33/31/306.
[91] C. Furtado, F. Moraes, Europhys. Lett. 45 (1999) 279–. doi:10.1209/epl/i1999-00159-8.
[92] C. R. Hagen, Phys. Rev. A 77 (2008) 036101–. doi:10.1103/PhysRevA.77.036101.
[93] B. Allen, B. S. Kay, A. C. Ottewill, Phys. Rev. D 53 (1996) 6829. doi:10.1103/PhysRevD.53.6829.
[94] A. H. Castro Neto, F. Guinea, N. M. R. Peres, K. S. Novoselov, A. K. Geim, Rev. Mod. Phys. 81 (2009) 109–162.
doi:10.1103/RevModPhys.81.109.
[95] F. Wilczek, Phys. Rev. Lett. 49 (1982) 957–959. doi:10.1103/PhysRevLett.49.957.
[96] Y.-H. Chen, F. Wilczek, E. Witten, B. I. Halperin, Int. J. Mod. Phys. B3 (1989) 1001. doi:10.1142/S0217979289000725.
19