This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

A Class of Exactly Solvable Real and Complex PTPT Symmetric Reflectionless Potentials

Suman Banerjeea111e-mail address: [email protected] (S.B), Rajesh Kumar Yadava222e-mail address: [email protected] (R.K.Y), Avinash Khareb333e-mail address: [email protected] (A.K) and
Bhabani Prasad Mandalc444e-mail address:[email protected](BPM)
Abstract

We consider the question of the number of exactly solvable complex but PT-invariant reflectionless potentials with NN bound states. By carefully considering the XmX_{m} rationally extended reflectionless potentials, we argue that the total number of exactly solvable complex PT-invariant reflectionless potentials are 2[(2N1)m+N]2[(2N-1)m+N].

a~{}^{a}Department of Physics, Sido Kanhu Murmu University, Dumka-814110, India.
b~{}^{b}Department of Physics, Savitribai Phule Pune University, Pune-411007, India.
c~{}^{c}Department of Physics, Banaras Hindu University, Varanasi-221005, India.

PACS numbers: 11.30.Pb, 03.65.Ge, 03.65.Nk,

1 Introduction

The reflectionless potentials, also known as transparent potentials or black potentials, are of great importance in physics and engineering. In view of the numerous applications of the reflectionless potentials, it is very important to search for new reflectionless potentials. While it is well known that there are NN continuous parameter families of exactly solvable real reflectionless potentials, to the best of our knowledge, the question of the complex PT-invariant exactly solvable reflectionless potentials has not been addresses in the literature.

In the last two decades, after the discovery of the PTPT (combined parity (PP) and time reversal (TT)) symmetric non-hermitian systems [1, 2, 3], it has been shown that there are non-hermitian complex PT-invariant potentials which are also reflectionless [4, 5, 6, 7, 8]. After the discovery of the XmX_{m} exceptional orthogonal polynomials (EOPs) [9, 10, 11], a group of new (also known as rationally extended) hermitian as well as PTPT symmetric non-hermitian potentials have been constructed [12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31] with their solutions in terms of these EOPs. It is then natural to enquire how many distinct reflectionless potentials with NN bound states can be constructed using both PT symmetry and XmX_{m} EOPs. This is the task that we have undertaken in this paper.

Just to set the notation, we first consider a real exactly solvable reflectionless potential with NN bound states and using the method of supersymmetric quantum mechanics [32, 33, 34, 35] explicitly obtain one continuous (λ\lambda) parameter family of reflectionless potentials including the corresponding reflectionless Pursey and Abraham Moses (AM) potentials with N1N-1 bound states. This can be generalized and one can obtain NN continuous parameter families of real reflectionless potentials with NN bound states.

We then consider the case of the non-hermitian PTPT symmetric Scarf-II reflectionless potentials with NN bound states and discuss the role of the parametric symmetry. Finally, we consider the rationally extended complex PT-symmetric scarf-II potential whose eigenfunctions are written in the form of XmX_{m} Jacobi EOPs with m=1,2,3..m=1,2,3.., and show that these extended potentials are also reflectionless and combining all these factors we have in all 2[(2N1)m+N]2[(2N-1)m+N] number of complex PT-symmetric reflectionless potentials with NN bound states.

The organization of this manuscript is as follows: In Sec. 22, we briefly discuss the formalism of supersymmetric quantum mechanics relevant to this paper [32] and explicitly obtain one continuous parameter family of real reflectionless potentials with NN bound states. In Sec. 33, we consider the complex PT-invariant Scarf-II potential with NN bound states and obtain conditions underwhich it is reflectionless. We also discuss the role of parametric symmetry in counting the number of reflectionless PT-symmetric complex potentials. We then construct the corresponding XmX_{m} family of complex PT-invariant reflectionless potentials with NN bound states and argue that the total number of complex PT-invariant reflectionless potentials is 2[(2N1)m+N]2[(2N-1)m+N]. In Sec. 44, we consider the case of N=3N=3 explicitly. First we give explicit expression as well as suitable plots for the one parameter family of real reflectionless potentials and the corresponding eigenfunctions with three bound states. We then give explicit expression for the real and imaginary parts of the complex PT-invariant potentials and their eigenfunctions in the case of three bound states. Finally, in Sec. 55, we summarize our results.

2 Formalism

In this section, we set the basic notations of supersymmetric quantum mechanics (SQM) as relevant to the present discussion. We then discuss the case of the real reflectionless potential with NN bound states and obtain one continuous parameter family of strictly isospectral reflectionless potentials with NN bound states. This can be generalized [32] and one can obtain NN continuous parameter family of reflectioless real potentials with NN bound states.

2.1 Basic Results of Supersymmetric Quantum Mechanics

Consider a Hamiltonian

H(1)(x)=d2dx2+V(1)(x),(=2m=1)H^{(1)}(x)=-\frac{d^{2}}{dx^{2}}+V^{(1)}(x),\quad(\hbar=2m=1) (1)

with ground state energy E0(1)=0E_{0}^{(1)}=0. One can then factorize H(1)H^{(1)} in terms of the operators AA and AA^{\dagger} as

H(1)(x)=AAH^{(1)}(x)=A^{\dagger}A (2)

with

A=ddx+W(x)andA=ddx+W(x),A=\frac{d}{dx}+W(x)\quad\mbox{and}\quad A^{\dagger}=-\frac{d}{dx}+W(x), (3)

where

W(x)=ddx[lnψ0(1)(x)]W(x)=-\frac{d}{dx}[\ln\psi^{(1)}_{0}(x)] (4)

is the superpotential, which determines the two partner potentials

V(1)(x)=W2(x)W(x)andV(2)(x)=W2(x)+W(x).\displaystyle V^{(1)}(x)=W^{2}(x)-W^{\prime}(x)\quad\mbox{and}\quad V^{(2)}(x)=W^{2}(x)+W^{\prime}(x). (5)

The eigenvalues and the eigenfunctions of these two potentials (when the SUSY is unbroken) are related by

En+1(1)=En(2)E0(1)=0,E^{(1)}_{n+1}=E^{(2)}_{n}\qquad E^{(1)}_{0}=0, (6)

and

ψn(2)(x)=1[En(2)]1/2Aψn+1(1)ψn+1(1)(x)=1[En(2)]1/2Aψn(2)\psi^{(2)}_{n}(x)=\frac{1}{[E^{(2)}_{n}]^{1/2}}A\psi^{(1)}_{n+1}\qquad\psi^{(1)}_{n+1}(x)=\frac{1}{[E^{(2)}_{n}]^{1/2}}A^{\dagger}\psi^{(2)}_{n} (7)

respectively. For the one dimensional case, the transmission (T(1,2)(k))(T^{(1,2)}(k)) and reflection (R(1,2)(k))(R^{(1,2)}(k)) amplitudes for the partner potentials V(1,2)(x)V^{(1,2)}(x) are related by

R(1)(k)=(W+ikWik)R(2)(k)R^{(1)}(k)=\bigg{(}\frac{W_{-}+ik}{W_{-}-ik}\bigg{)}R^{(2)}(k) (8)

and

T(1)(k)=(W+ikWik)T(2)(k)T^{(1)}(k)=\bigg{(}\frac{W_{+}-ik^{\prime}}{W_{-}-ik}\bigg{)}T^{(2)}(k) (9)

where

k=(EW2)12andk=(EW+2)12k=(E-W^{2}_{-})^{\frac{1}{2}}\quad\mbox{and}\quad k^{\prime}=(E-W^{2}_{+})^{\frac{1}{2}} (10)

with

W±=W(x±).W_{\pm}=W(x\rightarrow\pm\infty). (11)

The one-parameter family of potentials V^(1)(λ,x)\hat{V}^{(1)}(\lambda,x) which are strictly isospectral to the given potential V(1)(x)V^{(1)}(x) are given by

V^(1)(λ,x)=V(1)(x)2d2dx2ln(I(x)+λ),\hat{V}^{(1)}(\lambda,x)=V^{(1)}(x)-2\frac{d^{2}}{dx^{2}}\ln(I(x)+\lambda)\,, (12)

where the integral I(x)I(x) in term of the normalized ground state wavefunction is given by

I(x)=x[ψ0(1)]2(x)𝑑xI(x)=\int^{x}_{-\infty}[\psi^{(1)}_{0}]^{2}(x)\,dx (13)

and λ\lambda is a constant which is either >0>0 or <1<-1. The corresponding superpotential W^(x)\hat{W}(x) with the same SUSY partner potential V(2)(x)V^{(2)}(x) is given by

W^(x)=W(x)+ddxln[I(x)+λ].\hat{W}(x)=W(x)+\frac{d}{dx}\ln[I(x)+\lambda]. (14)

The associated normalized ground state wavefunctions to the potential V^(1)(λ,x)\hat{V}^{(1)}(\lambda,x) is given by

ψ^0(1)(λ,x)=λ(1+λ)ψ0(1)(x)[I(x)+λ],\hat{\psi}^{(1)}_{0}(\lambda,x)=\frac{\sqrt{\lambda(1+\lambda)}\psi^{(1)}_{0}(x)}{[I(x)+\lambda]}, (15)

while the normalized excited-state (n=1,2,3n=1,2,3...) eigenfunctions are given by

ψ^n+1(1)(λ,x)=ψn+1(1)(x)+1En+1(1)(I(x)I(x)+λ)(ddx+W(x))ψ(1)n+1(x).\hat{\psi}^{(1)}_{n+1}(\lambda,x)=\psi^{(1)}_{n+1}(x)+\frac{1}{E^{(1)}_{n+1}}\bigg{(}\frac{I^{\prime}(x)}{I(x)+\lambda}\bigg{)}\bigg{(}\frac{d}{dx}+W_{(}x)\bigg{)}\psi^{(1)}_{n+1}(x). (16)

2.1.1 Pursey potential

The superpotential for this case is defined by putting λ=0\lambda=0 in Eq. (14)

W[P](x)=W(x)+ddxlnI(x).W^{[P]}(x)=W(x)+\frac{d}{dx}\ln I(x). (17)

and the potential (12) becomes

V[P](x)=V^(1)(λ=0,x)=V(1)(x)2d2dx2lnI(x),V^{[P]}(x)=\hat{V}^{(1)}(\lambda=0,x)=V^{(1)}(x)-2\frac{d^{2}}{dx^{2}}\ln I(x)\,, (18)

while the corresponding eigenvalues are

En[P]=En(2)=En+1(1),n=0,1,2.E^{[P]}_{n}=E^{(2)}_{n}=E^{(1)}_{n+1}\,,~{}~{}n=0,1,2...\,. (19)

The reflection and transmission amplitudes for this case are

R[P](k)=(WikW+ik)2R(1)(k)R^{[P]}(k)=\bigg{(}\frac{W_{-}-ik}{W_{-}+ik}\bigg{)}^{2}R^{(1)}(k) (20)
T[P](k)=(WikW+ik)T(1)(k).T^{[P]}(k)=-\bigg{(}\frac{W_{-}-ik}{W_{-}+ik}\bigg{)}T^{(1)}(k)\,. (21)

2.1.2 Abraham-Moses potential

In this case the superpotential and the potential (λ=1\lambda=-1) are given by

W[AM](x)=W(x)+ddxln(I(x)1),W^{[AM]}(x)=W(x)+\frac{d}{dx}\ln(I(x)-1), (22)

and

V[AM](x)=V^(1)(λ=1,x)=V(1)(x)2d2dx2ln(I(x)1).V^{[AM]}(x)=\hat{V}^{(1)}(\lambda=-1,x)=V^{(1)}(x)-2\frac{d^{2}}{dx^{2}}\ln(I(x)-1). (23)

The eigenvalues are identical to the Pursey potential and are given by Eq. (19). The reflection and transmission amplitudes for this case are

R[AM](k)=R(1)(k)R^{[AM]}(k)=R^{(1)}(k) (24)
T[AM](k)=(W++ikW+ik)T(1)(k).T^{[AM]}(k)=-\bigg{(}\frac{W_{+}+ik^{\prime}}{W_{+}-ik^{\prime}}\bigg{)}T^{(1)}(k)\,. (25)

2.2 Real potentials with NN-bound states

We consider the well known example of real reflectionless potential

V(1)(x)=N(N+1)sech2(x);x,V^{(1)}(x)=-N(N+1){\rm sech}^{2}(x);\quad-\infty\leq x\leq\infty, (26)

for any positive integer N>0N>0. The solutions of the time-independent one-dimensional schrödinger equation corresponding to this potential are well known [32] and given as

ψn(1)(x)=Cn(N)sechN(x)Pn(N12,N12)(isinh(x)),\psi^{(1)}_{n}(x)=C_{n}^{(N)}{\rm sech}^{N}(x)P^{(-N-\frac{1}{2},-N-\frac{1}{2})}_{n}(i\sinh(x)), (27)

with the energy eigenvalues

En(1)=(Nn)2,n=0,1,2nmax<NE^{(1)}_{n}=-(N-n)^{2},\quad n=0,1,2...n_{max}<N (28)

and the normalization constant

Cn(N)=2N[n!(Nn)[Γ(Nn+12)]2Γ(2Nn+1)π]1/2.C_{n}^{(N)}=2^{N}\bigg{[}\frac{n!(N-n)[\Gamma(N-n+\frac{1}{2})]^{2}}{\Gamma(2N-n+1)\pi}\bigg{]}^{1/2}. (29)

Here Pn(N12,N12)(isinh(x))P^{(-N-\frac{1}{2},-N-\frac{1}{2})}_{n}(i\sinh(x)) is the Jacobi polynomial. The corresponding reflection amplitude is zero at all positive energies while the transmission amplitude is given by

T(1)(k)=Γ(Nik)Γ(Nik+1)Γ(1ik)Γ(ik),T^{(1)}(k)=\frac{\Gamma(-N-ik)\Gamma(N-ik+1)}{\Gamma(1-ik)\Gamma(-ik)}, (30)

with k2=En(1)k^{2}=E^{(1)}_{n} while the transmission probability |T(1)(k)|2=1|T^{(1)}(k)|^{2}=1.

2.2.1 One-parameter family of reflectionless potentials

It is straight forward to obtain the one continuous parameter family of reflectionless potentials by using Eqs. (12), (13) and (26) and we obtain

V^(1)(λ,x)=N(N+1)sech2(x)2d2dx2ln(I(x)+λ),\hat{V}^{(1)}(\lambda,x)=-N(N+1){\rm sech}^{2}(x)-2\frac{d^{2}}{dx^{2}}\ln(I(x)+\lambda)\,, (31)

where

I(x)=[C0N]2xsech2N(y)𝑑y,I(x)=[C_{0}^{N}]^{2}\int_{-\infty}^{x}{\rm sech}^{2N}(y)\,dy\,, (32)

and λ>0\lambda>0 or λ<1\lambda<-1. Further, it is straight forward to obtain the corresponding reflectionless Pursey or AM reflectionless potentials with N1N-1 bound states using Eqs. (18) and (23).

This procedure can be iterated NN times to find NN continuous parameter family of strictly isospectral reflectionless potentials with NN bound states [32].

3 PTPT Symmetric Complex Reflectionless Potentials

Apart from the NN continuous parameter families of real reflectionless potentials with NN bound states, it turns out that there are a vast number of complex PTPT symmetric reflectionless potentials with NN bound states which we discuss in this section by starting from the well known complex PT-invariant Scarf-II potential.

3.1 PTPT symmetric complex Scarf-II potential

The complex PTPT symmetric Scarf-II potential giving entirely real spectrum is well-known [36, 37] and given by

V(1)(x,a,b)=[b2+a(a+1)]sech2(x)+ib(2a+1)sech(x)tanh(x);<x<.V^{(1)}(x,a,b)=-[b^{2}+a(a+1)]{\rm sech}^{2}(x)+ib(2a+1){\rm sech}(x)\tanh(x);-\infty<x<\infty. (33)

The corresponding bound state energy eigenvalues and the eigenfunctions respectively are

En(1)=(an)2,n=0,1,2.nmax<a,E^{(1)}_{n}=-(a-n)^{2},\quad n=0,1,2....n_{max}<a, (34)

and

ψn(1)(x,a,b)=Cn(a,b)(sechx)aexp(ibtan1(sinhx))Pn(α,β)(isinhx),\psi^{(1)}_{n}(x,a,b)=C^{(a,b)}_{n}({\rm sech}x)^{a}\exp(-ib\tan^{-1}(\sinh x))P^{(\alpha,\beta)}_{n}(i\sinh x), (35)

with α=ba12\alpha=b-a-\frac{1}{2} and β=ba12\beta=-b-a-\frac{1}{2}.

The transmission and the reflection amplitudes of this potential are also well known [37] and are given by

Tscarf(1)(k,a,b)=Γ(aik)Γ(1+aik)Γ(12bik)Γ(12+bik)Γ(ik)Γ(1+ik)(Γ(12ik))2,T^{(1)}_{scarf}(k,a,b)=\frac{\Gamma(-a-ik)\Gamma(1+a-ik)\Gamma(\frac{1}{2}-b-ik)\Gamma(\frac{1}{2}+b-ik)}{\Gamma(-ik)\Gamma(1+ik)(\Gamma(\frac{1}{2}-ik))^{2}}, (36)

and

Rscarf(1)(k,a,b)=Tscarf(1)(k,a,b)×i[cosπasinπbcoshπk+sinπacosπbsinhπk]R^{(1)}_{scarf}(k,a,b)=T^{(1)}_{scarf}(k,a,b)\times i\bigg{[}\frac{\cos\pi a\sin\pi b}{\cosh\pi k}+\frac{\sin\pi a\cos\pi b}{\sinh\pi k}\bigg{]} (37)

respectively, where k2=En(1)k^{2}=E^{(1)}_{n}.

3.1.1 Parametric Symmetry

This potential (33) is invariant under the parametric transformation ba+12b\leftrightarrow a+\frac{1}{2}, however the corresponding eigenvalues and eigenfunctions are different [38] i.e.,

V(1,p)(x,a,b)=V(1)(x,ab12,ba+12)=V(1)(x,a,b)V^{(1,p)}(x,a,b)=V^{(1)}(x,a\rightarrow b-\frac{1}{2},b\rightarrow a+\frac{1}{2})=V^{(1)}(x,a,b) (38)

but

En(1,p)=(bn12)2;n=0,1,2.nmax<b12,E^{(1,p)}_{n}=-(b-n-\frac{1}{2})^{2};\qquad n=0,1,2....n_{\mbox{max}}<b-\frac{1}{2}, (39)

and

ψn(1,p)(x,a,b)=ψn(1)(x,ab12,ba+12).\psi^{(1,p)}_{n}(x,a,b)=\psi^{(1)}_{n}(x,a\rightarrow b-\frac{1}{2},b\rightarrow a+\frac{1}{2}). (40)

Here pp denotes the quantities obtained after parametric transformation. However, it is easy to check that the corresponding reflection and transmission amplitudes as given by Eqs. (36) and (37) are invariant under the parametric transformation ba+12b\leftrightarrow a+\frac{1}{2}.

Since the potential (33) has two different sets of eigenvalues and eigenfunctions ψn(1)(x,a,b)\psi^{(1)}_{n}(x,a,b) and ψn(1,p)(x,a,b)\psi^{(1,p)}_{n}(x,a,b), hence there are two different superpotentials corresponding to the same potential (33) and are given by

W(x,a,b)=atanh(x)+ibsech(x),W(x,a,b)=a\tanh(x)+ib{\rm sech}(x)\,, (41)

and

W(p)(x,a,b)=(b12)tanh(x)+i(a+12)sech(x).W^{(p)}(x,a,b)=(b-\frac{1}{2})\tanh(x)+i(a+\frac{1}{2}){\rm sech}(x)\,. (42)

This in turn gives two different partner potentials

V(2)(x,a,b)\displaystyle V^{(2)}(x,a,b) =\displaystyle= (W(x,a,b))2+W(x,a,b)\displaystyle(W(x,a,b))^{2}+W(x,a,b)^{\prime} (43)
=\displaystyle= (b2+a(a1))sech2(x)+ib(2a1)sech(x)tanh(x)\displaystyle-(b^{2}+a(a-1)){\rm sech}^{2}(x)+ib(2a-1){\rm sech}(x)\tanh(x)

and

V(2,p)(x,a,b)\displaystyle V^{(2,p)}(x,a,b) =\displaystyle= (W(p)(x,a,b))2+W(p)(x,a,b)\displaystyle(W^{(p)}(x,a,b))^{2}+W^{(p)}(x,a,b)^{\prime}
=\displaystyle= ((b1)2+a(a+1))sech2(x)+i(b1)(2a+1)sech(x)tanh(x)\displaystyle-((b-1)^{2}+a(a+1)){\rm sech}^{2}(x)+i(b-1)(2a+1){\rm sech}(x)\tanh(x)

respectively. The first partner potential (43) is SI under translation of parameter aa1a\rightarrow a-1, whereas the second one (3.1.1) is SI under translation of another parameter bb1b\rightarrow b-1. The reflection and transmission amplitudes of these two partner potentials are related to those of the potential (33) by

Rscarf(2)(k,a,b)\displaystyle{R}^{(2)}_{scarf}(k,a,b) =\displaystyle= (WikW+ik)Rscarf(1)(k,a,b)\displaystyle\bigg{(}\frac{W_{-}-ik}{W_{-}+ik}\bigg{)}R^{(1)}_{scarf}(k,a,b) (45)
=\displaystyle= (a+ikaik)Rscarf(1)(k,a,b),\displaystyle\bigg{(}\frac{a+ik}{a-ik}\bigg{)}R^{(1)}_{scarf}(k,a,b)\,,
Tscarf(2)(k,a,b)=(a+ikaik)Tscarf(1)(k,a,b).{T}^{(2)}_{scarf}(k,a,b)=-\bigg{(}\frac{a+ik}{a-ik}\bigg{)}{T}^{(1)}_{scarf}(k,a,b)\,. (46)

In the parametric case, the reflection and transmission amplitudes are invariant i.e.,

Rscarf(1,p)(k,a,b)\displaystyle{R}^{(1,p)}_{scarf}(k,a,b) =\displaystyle= Rscarf(1)(k,ab12,ba+12)\displaystyle{R}^{(1)}_{scarf}(k,a\rightarrow b-\frac{1}{2},b\rightarrow a+\frac{1}{2})
Tscarf(1,p)(k,a,b)\displaystyle{T}^{(1,p)}_{scarf}(k,a,b) =\displaystyle= Tscarf(1)(k,ab12,ba+12)\displaystyle{T}^{(1)}_{scarf}(k,a\rightarrow b-\frac{1}{2},b\rightarrow a+\frac{1}{2})

and for the corresponding partner potentials these are related as

Rscarf(2,p)(k,a,b)\displaystyle{R}^{(2,p)}_{scarf}(k,a,b) =\displaystyle= (W(p)ikW(p)+ik)R(1,p)(k,a,b)\displaystyle\bigg{(}\frac{W^{(p)}_{-}-ik}{W^{(p)}_{-}+ik}\bigg{)}R^{(1,p)}(k,a,b) (48)
=\displaystyle= (b12+ikb12ik)Rscarf(1,p)(k,a,b),\displaystyle\bigg{(}\frac{b-\frac{1}{2}+ik}{b-\frac{1}{2}-ik}\bigg{)}R^{(1,p)}_{scarf}(k,a,b)\,,
Tscarf(2,p)(k,a,b)=(b12+ikb12ik)Tscarf(1,p)(k,a,b){T}^{(2,p)}_{scarf}(k,a,b)=-\bigg{(}\frac{b-\frac{1}{2}+ik}{b-\frac{1}{2}-ik}\bigg{)}{T}^{(1,p)}_{scarf}(k,a,b)\, (49)

respectively.

3.1.2 Conditions for Reflectionless potentials

From the Eq. (37), it follows that all the potential V(1)(x,a,b)V^{(1)}(x,a,b) and hence V(2)(x,a,b)V^{(2)}(x,a,b) and V(2,p)(x,a,b)V^{(2,p)}(x,a,b) are reflectionless when the potential parameters aa and bb are either both integers or both half integers. Remarkably, using parametric symmetry it turns out that there are in fact 2N2N distinct complex PTPT-invariant reflectionless potentials all of which hold NN bound states. Out of these NN complex PT-invariant reflectionless potentials have half-integral values of aa and bb while the remaining NN complex PT-invariant potentials have integral values of aa and bb which we now list one by one.

Case (A): If aa and bb both half integers

[a,b]=[(2N1)/2,1/2],[(2N3)/2,3/2],,[3/2,(2N3)/2],[1/2,(2N1)/2].[a,b]=[(2N-1)/2,1/2],[(2N-3)/2,3/2],...,[3/2,(2N-3)/2],[1/2,(2N-1)/2]\,. (50)

On using the fact that the eigenvalues of the complex Scarf-II potential are given by Eq. (34) while those of the corresponding parametric case are given by Eq. (39), one can immediately figure out about how many eigenvalues are coming from the normal scarf-II and how many from the parametric case. For example, while in the case [a=(2N1)/2,b=1/2][a=(2N-1)/2,b=1/2], all the NN eigenvalues are from normal scarf-II, in the case [a=(2N3)/2,b=3/2][a=(2N-3)/2,b=3/2], while N1N-1 eigenvalues are coming from normal Scarf-II, one eigenvalue is coming from the parametric case.

Case (B): If aa and bb both integers

[a,b]=[N1,1],[N2,2],,[1,N1],[0,N].[a,b]=[N-1,1],[N-2,2],...,[1,N-1],[0,N]\,. (51)

In the case [a=0,b=N][a=0,b=N], while all the NN eigenvalues are from the parametric case, in the case of [a=1,b=N1][a=1,b=N-1] we have 11 eigenvalue from the normal Scarf-II while N1N-1 eigenvalues are coming from the parametric case.

3.2 Rationally Extended PTPT Symmetric Complex Potential

The PTPT symmetric complex Scarf-II potential V(1)(x,a,b)V^{(1)}(x,a,b) (given by Eq. (33)) has been extended rationally [19] in terms of classical Jacobi polynomials Pm(α,β)(z)P^{(\alpha,\beta)}_{m}(z) for any positive integers of m0m\geq 0 given by

Vm,ext(1)(x,a,b)\displaystyle V^{(1)}_{m,ext}(x,a,b) =\displaystyle= V(1)(x,a,b)+2m(2bm+1)+(2bm+1)\displaystyle V^{(1)}(x,a,b)+2m(2b-m+1)+(2b-m+1)
×\displaystyle\times [(2a1)+(2b+1)isinhx](Pm1(α,β)(isinhx)Pm(α1,β1)(isinhx))\displaystyle[(-2a-1)+(2b+1)i\sinh x]\bigg{(}\frac{P^{(-\alpha,\beta)}_{m-1}(i\sinh x)}{P^{(-\alpha-1,\beta-1)}_{m}(i\sinh x)}\bigg{)}
\displaystyle- (2bm+1)2cosh2x2(Pm1(α,β)(isinhx)Pm(α1,β1)(isinhx))2.\displaystyle\frac{(2b-m+1)^{2}\cosh^{2}x}{2}\bigg{(}\frac{P^{(-\alpha,\beta)}_{m-1}(i\sinh x)}{P^{(-\alpha-1,\beta-1)}_{m}(i\sinh x)}\bigg{)}^{2}.

The bound state spectrum of this extended potential is the same (isospectral) as that of the conventional one but the eigenfunctions are different and written in term of exceptional XmX_{m} Jacobi polynomials P^n+m(α,β)(isinhx)\hat{P}^{(\alpha,\beta)}_{n+m}(i\sinh x) as

ψext,n,m(1)(x,a,b)sechaxexp[ibtan1(sinhx)]Pm(α1,β1)(isinhx)P^n+m(α,β)(isinhx),\psi^{(1)}_{ext,n,m}(x,a,b)\propto\frac{{\rm sech}^{a}x\exp[-ib\tan^{-1}(\sinh x)]}{P^{(-\alpha-1,\beta-1)}_{m}(i\sinh x)}\hat{P}^{(\alpha,\beta)}_{n+m}(i\sinh x), (53)

where

P^n+m(α,β)(z(x))=(1)m[(1+α+β+n)2(1+α+n)(z(x)1)Pm(α1,β1)(g)Pn1(α+2,β)(z(x))\displaystyle\hat{P}^{(\alpha,\beta)}_{n+m}(z(x))=(-1)^{m}\bigg{[}\frac{(1+\alpha+\beta+n)}{2(1+\alpha+n)}(z(x)-1)P^{(-\alpha-1,\beta-1)}_{m}(g)P^{(\alpha+2,\beta)}_{n-1}(z(x))
+(1+αm)(α+1+n)Pm(2α,β)(z(x))Pn(α+1,β1)(z(x))];n,m0.\displaystyle+\frac{(1+\alpha-m)}{(\alpha+1+n)}P^{(-2-\alpha,\beta)}_{m}(z(x))P^{(\alpha+1,\beta-1)}_{n}(z(x))\bigg{]};\quad n,m\geq 0. (54)

is the exceptional Jacobi polynomial. Similar to the Scarf-II potential, this extended potential is also SI under the translation of the parameters a(a1)a\rightarrow(a-1). The transmission and reflection amplitudes for this potential are known [39] and are given by

Text,scarf(1)(k,m,a,b)=Tscarf(1)(k,a,b)ζ(m,a,b),T^{(1)}_{ext,scarf}(k,m,a,b)=T^{(1)}_{scarf}(k,a,b)\zeta(m,a,b)\,, (55)

and

Rext,scarf(1)(k,m,a,b)=Rscarf(1)(k,a,b)ζ(m,a,b),R^{(1)}_{ext,scarf}(k,m,a,b)=R^{(1)}_{scarf}(k,a,b)\zeta(m,a,b), (56)

where ζ(m,a,b)=([b2(ik12)2]+(bik+12)(1m)[b2(ik+12)2]+(b+ik+12)(1m))\zeta(m,a,b)=\left(\frac{[b^{2}-(ik-\frac{1}{2})^{2}]+(b-ik+\frac{1}{2})(1-m)}{[b^{2}-(ik+\frac{1}{2})^{2}]+(b+ik+\frac{1}{2})(1-m)}\right).

Remarkably, it turns out that unlike the conventional Scarf-II potential (33), the corresponding rationally extended Scarf-II potential (3.2) is not invariant under the parametric transformation ba+12b\longleftrightarrow a+\frac{1}{2}, rather it generates another extended potential [40] given by

Vm,ext(1,p)(x,a,b)=Vm,ext(1)(x,ba+12).{V}^{(1,p)}_{m,ext}(x,a,b)=V^{(1)}_{m,ext}(x,b\leftrightarrow a+\frac{1}{2}). (57)

The energy eigenvalues of this potential are isospectral to that of the conventional potential obtained after the transformation ba+12b\longleftrightarrow a+\frac{1}{2} given by Eq. (39) and the eigenfunction is

ψext,n,m(1,p)(x,a,b)=ψext,n,m(1)(x,ab1/2,ba+1/2).\psi^{(1,p)}_{ext,n,m}(x,a,b)=\psi^{(1)}_{ext,n,m}(x,a\rightarrow b-1/2,b\rightarrow a+1/2). (58)

This potential (57) is also SI under the translation of parameter bb1b\longrightarrow b-1. Since the potentials obtained after parametric transformations are different, hence the scattering amplitudes corresponding to these potentials are also different which are given by

Text,scarf(1,p)(k,m,a,b)\displaystyle T^{(1,p)}_{ext,scarf}(k,m,a,b) =\displaystyle= Text,scarf(1)(k,m,ab12,ba+12)\displaystyle T^{(1)}_{ext,scarf}(k,m,a\rightarrow b-\frac{1}{2},b\rightarrow a+\frac{1}{2})
Rext,scarf(1,p)(k,m,a,b)\displaystyle R^{(1,p)}_{ext,scarf}(k,m,a,b) =\displaystyle= Rext,scarf(1)(k,m,ab12,ba+12).\displaystyle R^{(1)}_{ext,scarf}(k,m,a\rightarrow b-\frac{1}{2},b\rightarrow a+\frac{1}{2}). (59)

Thus, it turns out that the extended potentials do not respect the parametric symmetry. As a result for a given value of [a,b][a,b] (both integers or half integers) one has in fact two different sets of rationally extended potentials for a given mm which are reflectionless. The only exceptions to these are the cases when either [a=(2N1)/2,b=1/2][a=(2N-1)/2,b=1/2] or [a=0,b=N][a=0,b=N] where only one rational partner exists. Thus in all, one has 2(2N1)m+2N2(2N-1)m+2N number of complex PT-invariant reflectionless potentials, with the 2N2N potentials being the nonrational (or m=0m=0) ones.

4 Illustration For Three Bound States (N=3N=3)

In this section, as an illustration, we consider all reflectionless potentials (both real and complex PT-invariant ones) with three bound states and show the behavior of these potentials and their corresponding normalized ground state eigenfunctions through graphical representation.

4.1 Real reflectionless potential

In this case, we fix the value of parameter N=3N=3 which gives the potential (26)

V(1)(x)=12sech2(x),V^{(1)}(x)=-12{\rm sech}^{2}(x), (60)

with three bound states with the corresponding binding energies being E0(1)=9,E1(1)=4E^{(1)}_{0}=-9,E^{(1)}_{1}=-4 and E2(1)=1E^{(1)}_{2}=-1. The normalized ground state is

ψ0(1)(x)=154sech3x.\psi^{(1)}_{0}(x)=\frac{\sqrt{15}}{4}{\rm sech}^{3}x\,. (61)

Thus the partner potential with its normalized ground state eigenfunction are

V(2)(x)=6sech2(x),ψ0(2)(x)=32sech2(x).V^{(2)}(x)=-6{\rm sech}^{2}(x)\,,~{}~{}\psi_{0}^{(2)}(x)=\frac{\sqrt{3}}{2}{\rm sech}^{2}(x)\,. (62)

It is straight forward to calculate the integral (13) using the potential (60) and one obtains

I(x)=116(8+[8+4sech2(x)+3sech4(x)]tanh(x)),I(x)=\frac{1}{16}\bigg{(}8+[8+4{\rm sech}^{2}(x)+3{\rm sech}^{4}(x)]\tanh(x)\bigg{)}\,, (63)

which gives one continuous parameter family of real reflectionless potentials

V^(1)(x,λ)\displaystyle\hat{V}^{(1)}(x,\lambda) =\displaystyle= 6sech2(x)[2+1(8+16λ+(8+4sech2(x)+3sech4(x))tanh(x))2\displaystyle 6{\rm sech}^{2}(x)\bigg{[}-2+\frac{1}{(8+16\lambda+(8+4{\rm sech}^{2}(x)+3{\rm sech}^{4}(x))\tanh(x))^{2}} (64)
×\displaystyle\times [15sech4(x){(1+3cosh(2x)+cosh(4x))sech6(x)\displaystyle[15{\rm sech}^{4}(x)\{(1+3\cosh(2x)+\cosh(4x)){\rm sech}^{6}(x)
+\displaystyle+ 16tanh(x)(1+2λ+tanh(x))}]].\displaystyle 16\tanh(x)(1+2\lambda+\tanh(x))\}]\bigg{]}.

The normalized ground state eigenfunction for this potential is obtained as

ψ^0(1)(x,λ)=4[15λ(λ+1)]12sech3(x)(8+16λ+(8+4sech2(x)+3sech4(x))tanh(x)).\displaystyle\hat{\psi}^{(1)}_{0}(x,\lambda)=\frac{4[15\lambda(\lambda+1)]^{\frac{1}{2}}{\rm sech}^{3}(x)}{(8+16\lambda+(8+4{\rm sech}^{2}(x)+3{\rm sech}^{4}(x))\tanh(x))}. (65)

In the limit of λ0\lambda\rightarrow 0 and 1-1, we get the reflectionless Pursey and the AM potentials respectively with two bound states. The expressions for these two potentials are given by

V[P/AM](x)=24sech2(x)[25cosh(2x)+13cosh(4x)3(5+5sinh(2x)+4sinh(4x))](5+11cosh(2x)9sinh2(2x)),V^{[P/AM]}(x)=-\frac{24{\rm sech}^{2}(x)[25\cosh(2x)+13\cosh(4x)\mp 3(\mp 5+5\sinh(2x)+4\sinh(4x))]}{(5+11\cosh(2x)\mp 9\sinh^{2}(2x))}, (66)

where upper sign corresponds to the Pursey and the lower one for the AM potential. The plots of V^(1)(x,λ)\hat{V}^{(1)}(x,\lambda) for positive and negative λ\lambda are shown in Fig. 1(a) and 1(b) respectively. The AM (V[AM](x)V^{[AM]}(x)), the Pursey (V[P](x)V^{[P]}(x)) and the partner potential (V(2)(x)V^{(2)}(x)) are shown in Fig. 1(c). The normalized ground state wavefunctions for some positive values of λ\lambda are also shown in Fig. 1(d).
[Uncaptioned image]
Fig.1(a) One-parameter family of potential V^(1)(x,λ)\hat{V}^{(1)}(x,\lambda) for positive λ=0.1,0.01,0.0001,0\lambda=0.1,0.01,0.0001,0 and \infty. The Pursey potential is shown for λ=0\lambda=0.
[Uncaptioned image]
Fig.1(b) One-parameter family of potential V^(1)(x,λ)\hat{V}^{(1)}(x,\lambda) for negative λ=1.1,1.01,1.0001,1\lambda=-1.1,-1.01,-1.0001,-1 and -\infty. The AM potential is shown for λ=1\lambda=-1.
 [Uncaptioned image]
Fig.1(c) The Pursey potential V[P](x){V}^{[P]}(x), the AM potential V[AM](x){V}^{[AM]}(x) and the partner potential V(2)(x){V}^{(2)}(x).
[Uncaptioned image]
Fig.1(d) Normalized ground-state wavefunctions ψ^0(1)(x,λ)\hat{\psi}^{(1)}_{0}(x,\lambda) for some potentials (with positive λ=0.1,0.01,0.001\lambda=0.1,0.01,0.001 and \infty.)

4.2 PT symmetric complex Scarf-II potentials

4.2.1 The conventional potential

In this case, for three bound states (N=3N=3), we have six different possible combinations of [a,b][a,b] given by

[a,b]=[5/2,1/2],[3/2,3/2],[1/2,5/2],[2,1],[1,2],[0,3],[a,b]=[5/2,1/2],[3/2,3/2],[1/2,5/2],[2,1],[1,2],[0,3]\,, (67)

and hence we have six reflectionless potentials. If we plot the potentials for these combinations of aa and bb, while the potentials [5/2,1/2][5/2,1/2], [3/2,3/2][3/2,3/2] and [1/2,5/2][1/2,5/2] are same as that of [0,3][0,3], [1,2][1,2] and [2,1][2,1] respectively as can be seen from the plots, the corresponding eigenfunctions are different for half-integer and integer combinations of both aa and bb. As mentioned in Eqs. (50) and (51), the integer combination [0,3][0,3] is not acceptable for the potential V(1)(x,a,b)V^{(1)}(x,a,b), however this is acceptable for the parametric case V(1,p)(x,a,b)V^{(1,p)}(x,a,b). Similarly, the first combination [5/2,1/2][5/2,1/2] is well acceptable for the first potential, but not for the parametric case. The plots of these potentials (real and imaginary parts) are shown in Fig. 22. The corresponding eigenfunctions with their parametric forms are also shown in Fig. 33 and 44 receptively. We also compare the eigenfunctions of conventional PTPT symmetric potentials with their parametric counterparts graphically (shown in Fig. 55).

Refer to caption
(a) a=52,b=12a=\frac{5}{2},b=\frac{1}{2}
Refer to caption
(b) a=32,b=32a=\frac{3}{2},b=\frac{3}{2}
Refer to caption
(c) a=12,b=52a=\frac{1}{2},b=\frac{5}{2}

Fig.2: (a)-(c) Real and imaginary parts of the conventional PTPT symmetric Scarf-II potential (V(1)(x,a,b)V^{(1)}(x,a,b)) vs xx for different aa and bb.

[Uncaptioned image]

(a) a=52,b=12a=\frac{5}{2},b=\frac{1}{2}

[Uncaptioned image]

(b) a=32,b=32a=\frac{3}{2},b=\frac{3}{2}

[Uncaptioned image]

(c) a=12,b=52a=\frac{1}{2},b=\frac{5}{2}

Fig.3 (i): (a)-(c) Real and imaginary parts of the normalized ground state eigenfunctions ψ0(1)(x,a,b)\psi^{(1)}_{0}(x,a,b) vs xx for half-integer values of aa and bb.

[Uncaptioned image]

(a) a=1,b=2a=1,b=2

[Uncaptioned image]

(b) a=2,b=1a=2,b=1

Fig.3 (ii): (a)-(b) Real and imaginary parts of the normalized ground state eigenfunctions for ψ0(1)(x,a,b)\psi^{(1)}_{0}(x,a,b) vs xx for integer values of aa and bb.

[Uncaptioned image]

(a) a=32,b=32a=\frac{3}{2},b=\frac{3}{2}

[Uncaptioned image]

(b) a=12,b=52a=\frac{1}{2},b=\frac{5}{2}

Fig.4 (i): (a)-(b) Real and imaginary parts of the normalized ground state eigenfunctions (for half-integer values of aa and bb) for the parametric case ψ0(1,p)(x,a,b)\psi^{(1,p)}_{0}(x,a,b).

[Uncaptioned image]

(a) a=1,b=2a=1,b=2

[Uncaptioned image]

(b) a=2,b=1a=2,b=1

[Uncaptioned image]

(b) a=0,b=3a=0,b=3

Fig.4 (ii): (a)-(b) Real and imaginary parts of the normalized ground state eigenfunctions (for integer values of aa and bb) for the parametric case ψ0(1,p)(x,a,b)\psi^{(1,p)}_{0}(x,a,b).

[Uncaptioned image]

(a)

[Uncaptioned image]

(b)

Fig.5 (i): (a)-(b) Comparison of real and imaginary parts of the normalized ground state eigenfunctions for the conventional ψ0(1)(x,a,b)\psi^{(1)}_{0}(x,a,b) and parametric cases ψ0(1,p)(x,a,b)\psi^{(1,p)}_{0}(x,a,b) for a=12,b=52a=\frac{1}{2},b=\frac{5}{2}.

[Uncaptioned image]

(a)

[Uncaptioned image]

(b)

Fig.5 (ii): (a)-(b) Comparison of real and imaginary parts of the normalized ground state eigenfunctions for the conventional ψ0(1)(x,a,b)\psi^{(1)}_{0}(x,a,b) and parametric cases ψ0(1,p)(x,a,b)\psi^{(1,p)}_{0}(x,a,b) for a=2,b=1a=2,b=1.

4.2.2 Rationally extended PTPT symmetric Scarf-II potential

In this case, unlike the conventional PTPT symmetric Scarf-II potential, the extended potentials as well as the corresponding eigenfunctions are completely different under parametric transformations. For m=1m=1, the expression of potential and the normalizable eigenfunction are given as

V1,ext(1)(x,a,b)=V(1)(x,a,b)+(2(2a+1))(2ibsinh(x)+2a+1)+(2((2a+1)24b2))(2ibsinh(x)+2a+1)2V^{(1)}_{1,ext}(x,a,b)=V^{(1)}(x,a,b)+\frac{(-2(2a+1))}{(-2ib\sinh(x)+2a+1)}+\frac{(2((2a+1)^{2}-4b^{2}))}{(-2ib\sinh(x)+2a+1)^{2}} (68)

and

ψext,n,m(1)(x,a,b)sechaxexp[ibtan1(sinhx)](2ibsinh(x)+2a+1)P^n+1(α,β)(isinhx),\psi^{(1)}_{ext,n,m}(x,a,b)\propto\frac{{\rm sech}^{a}x\exp[-ib\tan^{-1}(\sinh x)]}{(-2ib\sinh(x)+2a+1)}\hat{P}^{(\alpha,\beta)}_{n+1}(i\sinh x), (69)

where P^n+1(α,β)(isinhx)\hat{P}^{(\alpha,\beta)}_{n+1}(i\sinh x) is the X1X_{1} exceptional Jacobi polynomial.

We consider the same sets of parameters [a,b][a,b] (half-integers as well as integers) as discussed in the above conventional case and show the behaviors of potentials V1,ext(1)(x,a,b)V^{(1)}_{1,ext}(x,a,b), V1,ext(1,p)(x,a,b)V^{(1,p)}_{1,ext}(x,a,b) and the corresponding eigenfunctions ψ1,ext(1)(x,a,b)\psi^{(1)}_{1,ext}(x,a,b), ψ1,ext(1,p)(x,a,b)\psi^{(1,p)}_{1,ext}(x,a,b) respectively in Figs. 66, 77 and 88.

[Uncaptioned image]
[Uncaptioned image]
[Uncaptioned image]
[Uncaptioned image]

(a) a=12,b=52a=\frac{1}{2},b=\frac{5}{2}

[Uncaptioned image]

(b) a=32,b=32a=\frac{3}{2},b=\frac{3}{2}

[Uncaptioned image]

(c) a=52,b=12a=\frac{5}{2},b=\frac{1}{2}

Fig.6 (i): (a)-(c) Real and imaginary parts of the RE PTPT symmetric complex Scarf-II potentials and corresponding eigenfunctions for half-integer values of aa and bb.

[Uncaptioned image]
[Uncaptioned image]
[Uncaptioned image]

(a) a=2,b=1a=2,b=1

[Uncaptioned image]

(b) a=1,b=2a=1,b=2

Fig.6 (ii): (a)-(c) Real and imaginary parts of the RE PTPT symmetric complex Scarf-II potentials and their corresponding eigenfunctions for integer values of aa and bb.

[Uncaptioned image]
[Uncaptioned image]
[Uncaptioned image]

(a) a=32,b=32a=\frac{3}{2},b=\frac{3}{2}

[Uncaptioned image]

(b) a=12,b=52a=\frac{1}{2},b=\frac{5}{2}

Fig.7 (i): (a)-(b) Real and imaginary parts of the RE PTPT symmetric complex Scarf-II potentials and their corresponding eigenfunctions obtained after parametric transformation for half-integer values of aa and bb.

[Uncaptioned image]
[Uncaptioned image]
[Uncaptioned image]
[Uncaptioned image]

(a) a=2,b=1a=2,b=1

[Uncaptioned image]

(b) a=1,b=2a=1,b=2

[Uncaptioned image]

(c) a=0,b=3a=0,b=3

Fig.7 (ii): (a)-(b) Real and imaginary parts of the RE PTPT symmetric complex Scarf-II potentials and their corresponding eigenfunctions obtained after parametric transformation for integer values of aa and bb.

[Uncaptioned image]
[Uncaptioned image]
[Uncaptioned image]

(a) a=12,b=52a=\frac{1}{2},b=\frac{5}{2}

[Uncaptioned image]

(b) a=12,b=52a=\frac{1}{2},b=\frac{5}{2}

Fig.8 (i): (a)-(b) Comparison between real and imaginary parts of RE PTPT symmetric complex Scarf-II potential and their corresponding eigenfunctions obtained after parametric transformation for half integer combination of aa and bb.

[Uncaptioned image]
[Uncaptioned image]
[Uncaptioned image]

(a) a=2,b=1a=2,b=1

[Uncaptioned image]

(b) a=2,b=1a=2,b=1

Fig.8 (ii): (a)-(b) Comparison between real and imaginary parts of RE PTPT symmetric complex Scarf-II potential and their corresponding eigenfunctions obtained after parametric transformation for integer combination of aa and bb.

5 Conclusions

In this work, we have made an attempt to obtain all possible exactly solvable complex PT-invariant reflectionless potentials. As a simple exercise, we first started with a well-known reflectionless real potential with NN bound states and generated one continuous parameter (λ\lambda) family (which can be easily generalized to NN-parameter family) of strictly isospectral reflectionless potentials. As a special case we have also obtained expressions for the corresponding reflectionless Pursey and the AM potentials corresponding to λ=0\lambda=0 and 1-1 respectively and with N1N-1 bound states.

In the PTPT symmetric case, we started with the well known complex PT-invariant Scarf-II potential and showed that it has novel parametric symmetry. We then showed that there are NN number of reflectionless potentials when both aa and bb are either integers or half-integers, thereby obtaining 2N2N number of complex PT-invariant reflectionless potentials in total. Further, we considered the rationally extended PTPT symmetric reflectionless scarf-II potential, whose solutions are in terms of XmX_{m}-Jacobi EOPs and shown that unlike the usual one, this extended potential is not invariant under the parametric symmetry but instead generates another set of reflectionless potentials whose solutions are also in terms of XmX_{m}-EOPs. By combining all these factors we then showed that there are 2[(2N1)m+N]2[(2N-1)m+N] number of complex PT-invariant reflectionless exactly solvable potentials.

This paper raises few questions. Some of these are, have we really exhausted the number of complex PT-invariant reflectionless exactly solvable potentials or are there are still more? While we believe that the answer to the question is no, one can never be sure. Secondly, since reflectionless potentials have found wide applications, it would be interesting if one of these complex reflectionless potential finds some application.

Acknowledgments
AK is grateful to Indian National Science Academy (INSA) for awarding INSA Honorary Scientist position at Savitribai Phule Pune University. BPM acknowledges the research grant for faculty under IoE scheme (Number 6031) of Banaras Hindu University Varanasi.

References

  • [1] C. M. Bender and S. Boettcher, Phys. Rev. Lett. 80 (1998) 5243.
  • [2] C. M. Bender, Rep. Progr. Phys. 70 (2007) 947.
  • [3] A. Mostafazadeh, Int. J. Geom. Method. Mod. Phys. 7 (2010) 1191.
  • [4] Z. Ahmed, C. M. Bender and M. V. Berry, J. Phys. A: Math. Gen. 38 (2005) L627.
  • [5] Ming-huang Sang et al., Int. J. Theor. Phys. 50 (2011) 3459.
  • [6] A. Ghatak, M. Hasan, B. P. Mandal, Phys. Lett. A 379(20-21) (2015) 1326.
  • [7] S. Longhi, J. Phys. A: Math. Theor. 44 (2011) 485302.
  • [8] A. Mostafazadeh, Phys. Rev. A 87 (2013) 012103.
  • [9] D. Gomez-Ullate, N. Kamran and R. Milson, J. Math. Anal.Appl. 359 (2009) 352.
  • [10] D. Gomez-Ullate, N. Kamran and R. Milson, J. Phys. A 43 (2010) 434016.
  • [11] D. Gomez-Ullate, N. Kamran and R. Milson, Contemp. Math. 563 (2012) 51.
  • [12] C. Quesne, J.Phys.A 41 (2008) 392001.
  • [13] B. Bagchi, C. Quesne and R. Roychoudhary, Pramana J. Phys. 73(2009) 337, C. Quesne, SIGMA 5 (2009) 84.
  • [14] S. Odake and R. Sasaki, Phys. Lett. B, 684 (2010) 173; ibid 679 (2009) 414. J. Math. Phys, 51, 053513 (2010).
  • [15] S. Odake and R. Sasaki, Phys. Lett. B 702 (2011) 164.
  • [16] Y. Grandati, J. Math. Phys. 52 103505 (2011).
  • [17] Y. Grandati, Ann. Phys. 326 2074 (2011); 327 2411 (2012); 327 185 (2012).
  • [18] C. Quesne, SIGMA 8 080 (2012).
  • [19] B Midya and B Roy, J. Phys. A: Math. Theor. 46 (17) (2013) 175201 .
  • [20] Y. Grandati and C. Quesne, SIGMA 11 (2015) 061.
  • [21] R. K. Yadav, B. P. Mandal and A. Khare, Acta Polytechnica 57(6) (2017) 477.
  • [22] R. K. Yadav, A. Khare and B. P. Mandal, Ann. Phys. 331 (2013) 313.
  • [23] R. K. Yadav, A. Khare and B. P. Mandal, Phys. Lett. B 723 433 (2013); Phys. Lett. A 379 (2015) 67.
  • [24] C. L. Ho, J. C. Lee and R. Sasaki, Annals of Physics 343 (2014) 115.
  • [25] R. K. Yadav, N. Kumari, A. Khare and B. P. Mandal, Ann. Phys. 359 (2015) 46.
  • [26] A. Ramos et al., Ann. Phys.382 (2017) 143.
  • [27] N. Kumari, R. K. Yadav, A. Khare and B. P. Mandal, Ann. Phys.385 (2017) 57.
  • [28] N. Kumari, R. K. Yadav, A. Khare and B. P. Mandal, J. Math. Phys.59 (2018) 062103-1.
  • [29] B. Basu-Mallick, B. P. Mandal and P. Roy, Ann. Phys.380 (2017) 206.
  • [30] R. K. Yadav, A. Khare, N. Kumari and B. P. Mandal, Ann. Phys.400 (2019) 189.
  • [31] S. Yadav, A. Khare and B. P. Mandal, Ann. Phys.444 (2022) 169064.
  • [32] F. Cooper, A. Khare, U. Sukhatme Phys. Rep. 251 (1995) 267; ”SUSY in Quantum Mechanics” World Scientific (2001).
  • [33] A. Khare and U. P. Sukhatme J. Phys. A: Math. Gen.22(1989) 2847
  • [34] D. L. Pursey, Phys. Rev. D 33 1048 (1986) 2267.
  • [35] P. B. Abraham and H.E. Moses, Phys. Rev. A 22 (1980) 1333.
  • [36] B. Bagchi and C. Quesne, Phys. Lett. A 300 (2002) 18.
  • [37] Z. Ahmed, Phys. Lett. A 282 (2001) 343. 287 (2001) 295.
  • [38] B. Bagchi and C. Quesne, Phys. Lett. A 273 (2000) 285.
  • [39] N. Kumari, R. K. Yadav, A. Khare, B. Bagchi and B. P. Mandal, Ann. Phys.373 (2016) 163.
  • [40] R. K. Yadav, A. Khare, N. Kumari, B. Bagchi and B. P. Mandal, J. Math. Phys.57 (2016) 062106.