betaFeedback

WorksPhysics and Astronomy

Energy-independent complex single PP-waves NNNN potential from Marchenko equation

Abstract

We extend our previous results of solving the inverse problem of quantum scattering theory (Marchenko theory, fixed-ll inversion). In particular, we apply an isosceles triangular-pulse function set for the Marchenko equation input kernel expansion in a separable form. The separable form allows a reduction of the Marchenko equation to a system of linear equations for the output kernel expansion coefficients. We show that in the general case of a single partial wave, a linear expression of the input kernel is obtained in terms of the Fourier series coefficients of q1−m(1−S(q))q^{1-m}(1-S(q)) functions in the finite range of the momentum 0≤q≤π/h0\leq q\leqπ/h [S(q)S(q) is the scattering matrix, ll is the angular orbital momentum, m=0,1,…,2lm=0,1,\dots,2l]. Thus, we show that the partial SS–matrix on the finite interval determines a potential function with hh-step accuracy. The calculated partial potentials describe a partial SS–matrix with the required accuracy. The partial SS–matrix is unitary below the threshold of inelasticity and non–unitary (absorptive) above the threshold. We developed a procedure and applied it to partial-wave analysis (PWA) data of NNNN elastic scattering up to 3 GeV. We show that energy-independent complex partial potentials describe these data for single PP-waves.

PACS numbers: 24.10.Ht, 13.75.Cs, 13.75.Gx

Introduction

Extracting the interparticle interaction potential from scattering data is a fundamental problem in nuclear physics. The various inverse problem of quantum scattering (IP) methods are used for such extraction. Fitting parameters of the phenomenological potential can solve this problem. Such an adjustment is easily implemented in the case of a small number and low accuracy of the known experimental data. An accurate description of a large number of experimental data requires exact methods of IP solving. The development of such precise and unambiguous methods remains a fundamental challenge [9, 10, 11, 12, 13]. The ill-posedness of the problem significantly complicates its numerical solution.

The fixed-ll IP considered here is usually solved within Marchenko, Krein, and Gelfand-Levitan theories [1, 2, 3, 4, 5, 6, 7, 8]. In these approaches, the IP is reduced to solving Fredholm integral equations of the second kind. H.V. von Gevamb and H. Kolholff successfully applied Marchenko and Gelfand-Levitan theories to extract NNNN partial potentials from PWA data [14, 15]. They used the PWA data up to the inelastic threshold (Elab≈280E_{\mathrm{lab}} \approx280 MeV) and approximated the corresponding partial SS–matrices (spectral densities) by rational fraction expansions (Padé approximants). In this case, the input kernels of the integral equations are represented as finite separable series of the Riccati-Hankel functions products (separable kernels). A Fredholm integral equation of the second kind with a separable kernel is solved analytically. The partial potentials, in this case, are also

expressed through Riccati-Hankel functions (Bargmann-type potentials). A similar approach in frames of the Marchenko theory was used later to extract optical model NNNN partial potentials from elastic NNNN scattering PWA data up to 3 GeV [16, 17]. This method for solving the inverse problem includes four numerical procedures. The first procedure is the PWA, the second procedure is the approximation of the SS-matrix, the third procedure is the solution of the integral equation, and the fourth procedure is the differentiation of the output kernel to calculate the potential. Errors of each step are accumulated, the estimation of the error in the calculation of the potential is not an easy task. For l≥1l \ge1, Marchenko inversion is unstable for r<1r < 1 fm, and Gelfand-Levitan is unstable for r>8r > 8 fm [14, 15]. Description of PWA data within errors for energies up to 3 GeV requires the use of high-order Padé approximants [16, 17]. We found that an increase in the accuracy of the approximation and, accordingly, an increase in the order of the Padé approximant can lead to significant changes in the potential. The Padé approximant for such a problem is not the best choice since approximants of different orders giving close SS-matrix values at the PWA points can differ significantly between these points. Thus, the convergence of methods using rational fraction expansions of SS-matrix (spectral density) is not apparent.

This paper generalizes the algebraic method [18, 19] for solving the IP. We expand the Marchenko equation input kernel into a separable series in an isosceles triangular-pulse function set. We obtained a linear expression of the expansion coefficients in terms of the Fourier series coefficients of q1−m(1−S(q))q^{1-m}(1-S(q)) (m=0,1,…,2lm = 0, 1, \ldots, 2l) functions on a finite range of the momentum 0≤q≤π/h0 \le q \le\pi/h. Thus, we solve the Marchenko equation with a separable kernel, which can be performed analytically like in Refs. [14, 15, 16, 17].

Theory of the Fourier’s series substantiates convergence of the procedure with decreasing step hh.

NNNN potentials used as an input for (semi)microscopic construction of nuclear optical potentials (OPs) are usually real and only describe the NNNN PWA data below the inelastic threshold [20, 21, 22, 23, 24]. However, in the absence of a microscopic theory, complex partial NNNN potentials are required to describe nuclear reactions with energies of the NNNN relative motion above the threshold [17]. In existing models, real partial potentials are modified by energy-dependent imaginary terms which are zero below the threshold [16, 25]. The partial wave function is real below the threshold to within an rr-independent factor in such models. For realistic optical potentials, such a restriction is unreasonable. Indeed, the optical model potential (pseudopotential) definition does not guarantee that it will be Hermitian [26, 27]. One can only assert that Hamiltonian eigenfunctions corresponding to eigenenergies below the threshold satisfy the Hermitianity condition. However, the Hermitian condition

∫ψ∗(V+−V)ψ dx=0(1)\int\psi^*(V^+ - V)\psi\,dx = 0 \tag*{(1)}

implies V+−V=0V^+ - V = 0 only if Eq. (1) holds for an arbitrary function ψ\psi, which is not true for optical potentials. Using the phase-equivalent Krein transformations, one can obtain an energy-independent complex potential giving a unitary matrix [2]. The use of optical potentials limited by the condition V+−V=0V^+ - V = 0 below the inelasticity threshold is a physically unreasonable limitation. Previously, it was shown that the Marchenko theory applies to the description of elastic nDnD scattering from zero energy up to energies well above the threshold. In this case, Marchenko theory produces energy-independent complex partial nDnD potentials [28, 29]. They used a rational parametrization, the same as in Refs. [14, 15, 16, 17] for the unitary SS–matrix. The small value of the threshold (the deuteron’s binding energy Ec.m.≈2.226E_{\mathrm{c.m.}} \approx2.226 MeV) does not allow us to judge the applicability of the Marchenko theory in the general case. We analyzed the Marchenko theory [2, 3] (l=0l = 0), [4] (l>0l > 0) and found that most of the theory applies to unitary and to non–unitary SS–matrices. Our algebraic form of the Marchenko equation allows calculating the energy-independent complex local partial potential corresponding to a partly unitary and partly non–unitary SS–matrix. We previously applied this approach to analyze 1S0{}^{1}S_{0} data of elastic NNNN scattering PWA [19]. Extending this investigation, we analyze single PP-waves NNNN PWA data (up to Elab≈3E_{\mathrm{lab}} \approx3 GeV) and show that these data are described by energy-independent complex partial potentials. The reconstructed partial SS and PP-waves potentials constitute energy-independent soft core OP that describes elastic NNNN scattering PWA data up to 3 GeV.

Marchenko equation in an algebraic form

The radial Schrödinger equation is

(d2dr2−l(l+1)r2−V(r)+q2)ψ(r,q)=0.(2)\left(\frac{d^2}{dr^2}-\frac{l(l+1)}{r^2}-V(r)+q^2\right)\psi(r,q)=0. \tag*{(2)}

The Marchenko equation [2, 3] is a Fredholm integral equation of the second kind:

F(x,y)+L(x,y)+∫x+∞L(x,t)F(t,y) dt=0(3)F(x,y)+L(x,y)+\int_x^{+\infty}L(x,t)F(t,y)\,dt=0 \tag*{(3)}

The kernel function is defined by the following expression

F(x,y)=12π∫−∞+∞hl+(qx)[1−S(q)]hl+(qy) dq+∑j=1nbhl+(q~jx)Mj2hl+(q~jy)=12π∫−∞+∞hl+(qx)Y(q)hl+(qy) dq(4)\begin{aligned} F(x,y)={}&\frac{1}{2\pi}\int_{-\infty}^{+\infty}h_l^+(qx)[1-S(q)]h_l^+(qy)\,dq\\ &+\sum_{j=1}^{n_b}h_l^+(\widetilde{q}_j x)M_j^2h_l^+(\widetilde{q}_j y)\\ ={}&\frac{1}{2\pi}\int_{-\infty}^{+\infty}h_l^+(qx)Y(q)h_l^+(qy)\,dq \tag*{(4)} \end{aligned}

where hl+(z)h_l^+(z) is the Riccati-Hankel function, and

Y(q)=[1−S(q)−i∑j=1nbMj2(q−q~j)−1],(5)Y(q)=\left[1-S(q)-i\sum_{j=1}^{n_b}M_j^2(q-\widetilde{q}_j)^{-1}\right], \tag*{(5)}

Experimental data entering the kernel are

{S(q),(0<q<∞),q~j,Mj,j=1,…,n},(6)\{S(q),(0<q<\infty),\widetilde{q}_j,M_j,j=1,\ldots,n\}, \tag*{(6)}

where S(q)=e2iδ(q)S(q)=e^{2i\delta(q)} is a scattering matrix dependent on the momentum qq. The SS–matrix defines asymptotic behavior at r→+∞r\to+\infty of regular at r=0r=0 solutions of Eq. (2) for q≥0q\geq0; q~j2=Ej≤0\widetilde{q}_j^2=E_j\leq0, EjE_j is jj–th bound state energy (−q~j≥0-\widetilde{q}_j\geq0); MjM_j is jj-th bound state asymptotic constant.

The potential function of Eq. (2) is obtained from the solution of Eq. 3

V(r)=−2dL(r,r)dr(7)V(r)=-2\frac{dL(r,r)}{dr} \tag*{(7)}

Many methods for solving Fredholm integral equations use a series expansion of the equation kernel. [32, 33, 34, 35, 36, 37, 38]. We also use this approach.

We introduce auxiliary functions:

Fm(z)=12π∫−∞+∞eiqzY(q)qm dq,(8)F_m(z)=\frac{1}{2\pi}\int_{-\infty}^{+\infty}\frac{e^{iqz}Y(q)}{q^m}\,dq, \tag*{(8)}

then

dkFm(z)dzk=kkFm−k(z),(k=1,2,…,m).(9)\frac{d^kF_m(z)}{dz^k}=k^kF_{m-k}(z),\quad(k=1,2,\ldots,m). \tag*{(9)}

We use transformations

K^z,lf(z)=zl+1(−1zddz)l[z−1f(z)]≡(−1)l∑n=0l(2l−n)!n!(l−n)!(−2z)n−ldnf(z)dzn.(10)\hat{K}_{z,l}f(z)=z^{l+1}\left(-\frac{1}{z}\frac{d}{dz}\right)^l\left[z^{-1}f(z)\right]\equiv(-1)^l\sum_{n=0}^{l}\frac{(2l-n)!}{n!(l-n)!}(-2z)^{n-l}\frac{d^n f(z)}{dz^n}. \tag*{(10)}

From [31] (Eqs. (10.1.23)–(10.1.26)) we get

K^z,le±iqz=qlhl±(qz).(11)\hat{K}_{z,l}e^{\pm iqz}=q^l h_l^{\pm}(qz). \tag*{(11)}

and

K^y,lK^x,lF2l(x+y)=∑n1,n2=0l(2l−n1)!n1!(l−n1)!(2l−n2)!n2!(l−n2)!(−2x)n1−l(−2y)n2−ln1+n2F2l−n1−n2(x+y)=12π∫−∞+∞hl+(qx)Y(q)hl+(qy)dq≡F(x,y).(12)\begin{aligned} \hat{K}_{y,l}\hat{K}_{x,l}F_{2l}(x+y) &=\sum_{n_1,n_2=0}^{l}\frac{(2l-n_1)!}{n_1!(l-n_1)!}\frac{(2l-n_2)!}{n_2!(l-n_2)!}(-2x)^{n_1-l}(-2y)^{n_2-l}n_1+n_2 F_{2l-n_1-n_2}(x+y)\\ &=\frac{1}{2\pi}\int_{-\infty}^{+\infty}h_l^+(qx)Y(q)h_l^+(qy)dq\equiv F(x,y). \tag*{(12)} \end{aligned}

Assuming the finite range RR of the potential function V(r)V(r), we approximate Fm(x+y)F_m(x+y) as follows:

Fm(x+y)≈∑k=−2N2Nfm,kHk(x+y)(13)F_m(x+y)\approx\sum_{k=-2N}^{2N}f_{m,k}H_k(x+y) \tag*{(13)}
≈∑k,j=0NΔk(x)fm,k+jΔj(y)(14)\approx\sum_{k,j=0}^{N}\Delta_k(x)f_{m,k+j}\Delta_j(y) \tag*{(14)}

where fm,k≡Fm(kh)f_{m,k}\equiv F_m(kh), hh is some step, and R=NhR=Nh. The used basis sets are

H0(x)={1if 0≤x≤h,0otherwise,Hn(x)=H0(x−hn).(15)\begin{aligned} H_0(x)&=\begin{cases}1 & \text{if }0\le x\le h,\\0 & \text{otherwise},\end{cases}\\ H_n(x)&=H_0(x-hn). \tag*{(15)} \end{aligned}
Δ0(x)={1−∣x−0.25∣/hif ∣x−0.25∣≤h,0otherwise,Δn(x)=Δ0(x−hn).(16)\begin{aligned} \Delta_0(x)&=\begin{cases}1-\lvert x-0.25\rvert/h & \text{if }\lvert x-0.25\rvert\le h,\\0 & \text{otherwise},\end{cases}\\ \Delta_n(x)&=\Delta_0(x-hn). \tag*{(16)} \end{aligned}

We use bases set Δi(x)Δj(y)\Delta_i(x)\Delta_j(y) shifted by the vector (0.25h,0.25h)(0.25h,0.25h) compared to the set used previously [18, 19]. The basis sets are illustrated in Figure 1. Decreasing the step hh, one can approach Fm(x+y)F_m(x+y) arbitrarily close at all points with both sets. Coefficients fm,kf_{m,k} are same for both approximations Eqs. (13), (14).

The basis set illustration

Figure 1. The basis set Hn≡Hn(x+y)H_n\equiv H_n(x+y) (Eq. (15) is shown as trapezoid (triangle for n=0n=0) regions where Hn(x+y)=1H_n(x+y)=1, and elsewhere Hn(x+y)=0H_n(x+y)=0. The regions are bounded by lines x=0x=0, y=0y=0, and x+y=h(n−1)x+y=h(n-1). The basis set Δi(x)Δj(y)\Delta_i(x)\Delta_j(y) (Eq. (16)) is shown as projections (points) of the corresponding regular square pyramids apexes on the xyxy-plane. Δi(x)Δj(y)=1\Delta_i(x)\Delta_j(y)=1 at x=(0.25+i)hx=(0.25+i)h, y=(0.25+j)hy=(0.25+j)h (apex of the ijij-pyramid). The pyramids bases are (2h×2h)(2h\times2h) squares on the xyxy-plain with sides parallel to the xx and yy axes. Δi(x)Δj(y)=0\Delta_i(x)\Delta_j(y)=0 on sides of the corresponding squares (and outside them).

The Fourier transform of the basis set Eq. (13)

H~k(q)=∫−∞∞Hk(x)e−iqx dx=e−iqh−1qeiqhk.(17)\tilde{H}_k(q)=\int_{-\infty}^{\infty}H_k(x)e^{-iqx}\,dx=\frac{e^{-iqh}-1}{qe^{iqhk}}. \tag*{(17)}

The Fourier transform of Eq. (8) yields

Y(q)qm≈∑k=−2N2Nfm,kH~k(q)=∑k=−2N2Nfm,ke−iqh−1qeiqhk.(18)\frac{Y(q)}{q^m}\approx\sum_{k=-2N}^{2N}f_{m,k}\tilde{H}_k(q)=\sum_{k=-2N}^{2N}f_{m,k}\frac{e^{-iqh}-1}{qe^{iqhk}}. \tag*{(18)}

We rearrange the last relationship

Y(q)/qm−1=ι∑k=−2N2Nfm,k(e−iqh−1)e−iqhk=ι∑k=−2N+12N(fm,k−1−fm,k)e−iqhk+ι(−fm,−2N)eiqh2N+ι(fm,2N)e−iqh(2N+1).(19)Y(q)/q^{m-1} = \iota\sum_{k=-2N}^{2N} f_{m,k}\left(e^{-iqh}-1\right)e^{-iqhk} = \iota\sum_{k=-2N+1}^{2N}\left(f_{m,k-1}-f_{m,k}\right)e^{-iqhk}+\iota\left(-f_{m,-2N}\right)e^{iqh2N}+\iota\left(f_{m,2N}\right)e^{-iqh(2N+1)}. \tag*{(19)}

Thus, the left side of the expression is represented as a Fourier series on the interval −π/h≤q≤π/h-\pi/h \le q \le\pi/h.

fm,k−1−fm,k=−ιh2π∫−π/hπ/hY(q)eiqhkqm−1 dq(20)f_{m,k-1}-f_{m,k}=-\frac{\iota h}{2\pi}\int_{-\pi/h}^{\pi/h}Y(q)\frac{e^{iqhk}}{q^{m-1}}\,dq \tag*{(20)}

Thus, the range of known scattering data defines the value of hh and, therefore, the inversion accuracy. We solve the Eq. (3) substituting

L(x,y)≈∑j=0NPj(x)Δj(y)(21)L(x,y)\approx\sum_{j=0}^{N}P_j(x)\Delta_j(y) \tag*{(21)}
∑m=0N(δjm+∑n=0N[∫xmax⁡((m+0.25)h,(n+0.25)h)Δm(t)Δn(t) dt]Fn,j)Pm(x)=−∑k=0NΔk(x)Fk,j(22)\sum_{m=0}^{N}\left(\delta_{jm}+\sum_{n=0}^{N}\left[\int_x^{\max((m+0.25)h,(n+0.25)h)}\Delta_m(t)\Delta_n(t)\,dt\right]F_{n,j}\right)P_m(x)=-\sum_{k=0}^{N}\Delta_k(x)F_{k,j} \tag*{(22)}

We define

ζnmp=∫(p+0.25)hmax⁡((m+0.25)h,(n+0.25)h)Δm(t)Δn(t) dt=h6(2δnm(δnp+2ηn≥p+1)+δn(m−1)ηn≥p+δn(m+1)ηm≥p)(23)\zeta_{nmp}=\int_{(p+0.25)h}^{\max((m+0.25)h,(n+0.25)h)}\Delta_m(t)\Delta_n(t)\,dt=\frac{h}{6}\left(2\delta_{nm}\left(\delta_{np}+2\eta_{n\ge p+1}\right)+\delta_{n(m-1)}\eta_{n\ge p}+\delta_{n(m+1)}\eta_{m\ge p}\right) \tag*{(23)}

where δkp\delta_{kp} are the Kronecker symbols δkp\delta_{kp}, and

ηa={1if a is true,0otherwise,(24)\eta_a= \begin{cases} 1 & \text{if }a\text{ is true},\\ 0 & \text{otherwise}, \end{cases} \tag*{(24)}

Since Δk(hp)≡δkp\Delta_k(hp)\equiv\delta_{kp}, we finally get a system of equations

∑m=0N(δjm+∑n=pNζnmpFn,j)Pp,m=−Fp,j(25)\sum_{m=0}^{N}\left(\delta_{jm}+\sum_{n=p}^{N}\zeta_{nmp}F_{n,j}\right)P_{p,m}=-F_{p,j} \tag*{(25)}

for Pk(h(p+0.25))≡Pp,kP_k(h(p+0.25))\equiv P_{p,k}, (p,k=0,…,N)(p,k=0,\ldots,N) (j,p=0,…,N)(j,p=0,\ldots,N).

for k=−2N,…,2Nk=-2N,\ldots,2N. Recursive solving of the Eq. (20) from k=2N+1k=2N+1 (fm,2N+1=0f_{m,2N+1}=0) gives

fm,k=hπ∫−π/hπ/h(∑ν=k+12N+1eiqhν)Y(q)qm−1 dq=−ιh2π∫−π/hπ/heiqh(k+1)(1−eiqh(2N−k+1))1−eiqhY(q)qm−1 dq.(26)\begin{aligned} f_{m,k}=\frac{h}{\pi}\int_{-\pi/h}^{\pi/h}\left(\sum_{\nu=k+1}^{2N+1}e^{iqh\nu}\right)\frac{Y(q)}{q^{m-1}}\,dq \\ &=-\frac{\iota h}{2\pi}\int_{-\pi/h}^{\pi/h}\frac{e^{iqh(k+1)}\left(1-e^{iqh(2N-k+1)}\right)}{1-e^{iqh}}\frac{Y(q)}{q^{m-1}}\,dq. \tag*{(26)} \end{aligned}

The F(x,y)F(x,y) is defined by fm,kf_{m,k} (m=0,1,…,2lm=0,1,\ldots,2l), k=0,1,…,2Nk=0,1,\ldots,2N from Eqs. (12), and (14) as

F(x,y)≈∑k,j=0NΔk(x)Fk,jΔj(y),(27)F(x,y)\approx\sum_{k,j=0}^{N}\Delta_k(x)F_{k,j}\Delta_j(y), \tag*{(27)}

where

Fk,j=∑n1,n2=0l(2l−n1)!n1!(l−n1)!(2l−n2)!n2!(l−n2)!(−2(k+0.25)h)n1−l(−2(j+0.25)h)n2−lf2l−n1−n2,k+j(28)F_{k,j}=\sum_{n_1,n_2=0}^{l}\frac{(2l-n_1)!}{n_1!(l-n_1)!}\frac{(2l-n_2)!}{n_2!(l-n_2)!}(-2(k+0.25)h)^{n_1-l}(-2(j+0.25)h)^{n_2-l}f_{2l-n_1-n_2,k+j} \tag*{(28)}
=−ιh2π∫−π/hπ/hh+−1(q(k+0.25)h)eiqh(1−eiqh(2N−k−j+1))1−eiqhY(q)h+†(q(j+0.25)h)q dq.(29)=-\frac{\iota h}{2\pi}\int_{-\pi/h}^{\pi/h}h_+^{-1}(q(k+0.25)h)\frac{e^{iqh}\left(1-e^{iqh(2N-k-j+1)}\right)}{1-e^{iqh}}Y(q)h_+^{\dagger}(q(j+0.25)h)q\,dq. \tag*{(29)}

Solution of the Eq. (25) gives Pp,kP_{p,k}. We calculate potential values at points r=hpr=hp (p=0,…,Np=0,\ldots,N) from Eq. (7) by some finite difference formula.

We tested the developed approach by restoring the potential function V(r)=−3exp⁡(−3r/2)V(r)=-3\exp(-3r/2) from the corresponding scattering data. Results are presented in Figure 2, Figure 3, where h=0.04h=0.04, R=4R=4. The input SS–matrix was calculated at points shown in Figure 2 up to q=8q=8. The SS-matrix was interpolated by a quadratic spline in the range 0<q<80<q<8. The SS–matrix was approximated as asymptotic S(q)≈exp⁡(−2iα/q)S(q)\approx\exp(-2i\alpha/q) for q>8q>8, where α\alpha was calculated at q=8q=8.

Data points used to reconstruct the potential

Figure 2. Data used to reconstruct V(r)=V0exp⁡(−ar)V(r)=V_0\exp(-ar), where V0=−3 fm−2=−124.5 MeVV_0=-3\ fm^{-2}=-124.5\ MeV, a=1.5 fm−1a=1.5\ fm^{-1}. Angular orbital momentum l=1l=1. Units correspond to the NNNN system.

Initial and reconstructed potentials

Figure 3. Initial and reconstructed potentials: V(r)=V0exp⁡(−ar)V(r)=V_0\exp(-ar), where V0=−3 fm−2=−124.5 MeVV_0=-3\ fm^{-2}=-124.5\ MeV, a=1.5 fm−1a=1.5\ fm^{-1}. Units correspond to the NNNN system.

Energy-independent complex partial potentials

Two-particle relativistic potential models can be represented in a non–relativistic form [39]. Thus, the applicability of methods for solving the inverse problem is not limited only to nonrelativistic quantum mechanics. As we showed earlier [19], modern data of the NNNN partial-wave analysis up to Elab=3 GeVE_{\mathrm{lab}}=3\ \mathrm{GeV} can be described by the energy-independent complex partial potential for 1S0{}^{1}S_{0} single wave at least.

After analyzing the Marchenko theory [2, 3], including for the case of l>0l>0 [4], we found that Eqs. (2), (4) apply to non–unitary SS–matrices describing absorption. In this case, the absorbing partial SS–matrix should be defined as

S(q)={Su(q)+Sn(q)for q>0,Su+(−q)−Sn+(−q)for q<0,(30)S(q)= \begin{cases} S_u(q)+S_n(q) & \text{for } q>0,\\ S_u^{+}(-q)-S_n^{+}(-q) & \text{for } q<0, \end{cases} \tag*{(30)}

where superscript ++ means hermitian conjugation. For q>0q>0, we define

Su(q)=e2iδ(q),Sn(q)=−sin⁡2(ρ(q))e2iδ(q),(31)S_u(q)=e^{2i\delta(q)},\qquad S_n(q)=-\sin^2(\rho(q))e^{2i\delta(q)}, \tag*{(31)}

where δ(q)\delta(q) and ρ(q)\rho(q) are phase shift and inelasticity parameter correspondingly. With the SS–matrix defined in this way, all Eqs. (29)–(25) remain valid and allow calculating local and energy-independent OP from an absorptive SS–matrix and corresponding bound states’ characteristics.

Results and conclusions

Extending our previous results [19], we analyzed modern single PP-wave NNNN phase shift data up to 3 GeV [40, 41] (SW16, single-energy solutions, Figure 4).

P-wave phase shifts and inelasticity parameters

Figure 4. Data used to reconstruct PP-wave partial potentials are phase shifts δ\delta (solid circles) and inelasticity parameters ρ\rho (solid squares) (from Ref [40, 41]). Circles and squares stand for δ\delta and ρ\rho correspondingly calculated from the reconstructed partial potentials.

We smoothed the phase shift and inelasticity parameter data for q>3 fm−1q>3\ \mathrm{fm}^{-1} by the following functions, correspondingly:

δ(q)∼∑k=38Ak/qk,ρ(q)∼∑k=18Ak/qk,(32)\delta(q)\sim\sum_{k=3}^{8} A_k/q^k,\qquad\rho(q)\sim\sum_{k=1}^{8} A_k/q^k, \tag*{(32)}

where the coefficients were fitted by the least-squares method. The SW16 data and asymptotics (32) above q>3 fm−1q>3\ \mathrm{fm}^{-1} were used to calculate coefficients of Eqs. (29) with h=0.02 fmh=0.02\ \mathrm{fm} corresponding to qmax⁡≈157.08 fm−1q_{\max}\approx157.08\ \mathrm{fm}^{-1}.

The used data are described by energy-independent complex partial potentials (Figure 5, Figure 6). Thus, we presented an IP solution algorithm for finite range potentials (fixed-ll, single partial wave inversion, in frames of the Marchenko theory).

Real and imaginary parts of the reconstructed partial potentials

Figure 5. Real and imaginary parts of the reconstructed partial potentials.

Real and imaginary parts of the reconstructed partial potentials

Figure 6. Real and imaginary parts of the reconstructed partial potentials.

Our method gives a geometric glimpse into the NNNN interaction. As noted earlier by P. Fernández-Soler and E. Ruiz Arriola [30], ”short-wavelength fluctuations/oscillations are inherent to the maximum energy or CM momentum being fixed for the phase shift”. In this conclusion, they refer to the results of our calculations [16]. Then, we did not pay attention to this feature of the IP solution, and we considered the oscillations an artifact of the used IP solution in which the potential is explicitly, algebraically expressible in terms of Hankel functions. Reconstructed 1S0NN{}^{1}S_{0}NN partial potential [19] shows that short-wavelength oscillations may indeed be a manifestation of a physical phenomenon. There are also similar oscillations in the reconstructed PP-wave partial potentials (Figure 6). Microscopic calculations also predict such oscillations [42].

We show that IP solutions give optical model energy independent NNNN partial potentials not only in the 1S0NN{}^{1}S_{0}NN wave, but also in other single NNNN partial waves. The optical model partial NNNN potentials with a repulsive core does not necessarily exhibit a strong energy dependence up to 3 GeV, as previously stated in [30]. Instead of the soft repulsive core that was reconstructed in our model for the real part of the 1S0NN{}^{1}S_{0}NN partial potential, the barriers at about 0.2-1.3 fm were reconstructed for the real parts of the PP-waves partial potentials.

The reconstructed PP-wave NNNN complex potentials may be requested from the author in the Fortran code.

References

  1. [1]B. M. Levitan, Generalized Translation Operators and Some of Their Applications (Israel Program for Scientific Translations, Jerusalem, 1964).
  2. [2]Z. S. Agranovich and V. A. Marchenko, The Inverse Problem of Scattering Theory (Gordon and Breach, New York, 1963).
  3. [3]V. A. Marchenko, Sturm-Liouville Operators and their Applications (Naukova Dumka, Kiev,1977)
  4. [4]M. Blazek, Matem.-Fys. Casopis 13, 147 (1963); Commun. Math. Phys. 3, 282 (1966).
  5. [5]M. G. Krein, DAN SSSR 105, 433 (1955).
  6. [6]B. M. Levitan, Sturm-Liouville Inverse Problems (Nauka, Moscow, 1984)
  7. [7]R. G. Newton, Scattering Theory of Waves and Particles (Springer, New York, 1982).DOI
  8. [8]K. Chadan and P. C. Sabatier, Inverse Problems in Quantum Scattering Theory (Springer, New York, 1989).
  9. [9]J. M. Sparenberg and D. Baye, Phys. Rev. C 55, 2175 (1997).
  10. [10]D. Baye and J. M. Sparenberg, J. Phys. A: Math. Gen 37, 10223 (2004).
  11. [11]A. Pupasov, B. F. Samsonov, J. M. Sparenberg, and D. Baye, Phys. Rev. Lett. 106, 152301, (2011).
  12. [12]V. I. Kukulin, and R. S. Mackintosh, J. Phys. G30, R1 (2004).
  13. [13]R.S. Mackintosh, Scholarpedia, 7(11):12032 (2012); http://www.scholarpedia.org/article/Inverse_scatter
  14. [14]H. V. von Geramb and H. Kolhhoff, in Lecture Notes in Physics 427, edited by H. V. von Geramb (Springer Verlag, Berlin, 1994), p. 285.
  15. [15]H. KolhhoJJ and H. V. von Geramb, in Lecture Notes in Physics 427, edited by H. V. von Geramb (Springer Verlag, Berlin, 1994), p. 314.
  16. [16]N. A. Khokhlov and V. A. Knyr, Phys. Rev. C 73, 024004 (2006).
  17. [17]N. A. Khokhlov, V. A. Knyr, V. G. Neudatchin, Phys. Rev. C 75, 064001 (2007).
  18. [18]N.A. Khokhlov, arXiv:2102.01464 [quant-ph].
  19. [19]N.A. Khokhlov, L. I. Studenikina, Phys. Rev. C 104, 014001 (2021). arXiv:2104.03939 [quant-ph].
  20. [20]M. Burrows, R. B. Baker, Ch. Elster, S. P. Weppner, K. D. Launey, P. Maris, and G. Popa, Phys. Rev. C 102, 034606 (2020).
  21. [21]Matteo Vorabbi, Paolo Finelli, and Carlotta Giusti, Phys. Rev. C 98, 064602 (2018).
  22. [22]Matteo Vorabbi, Paolo Finelli, and Carlotta Giusti, Phys. Rev. C 96, 044001 (2017).
  23. [23]Hairui Guo, Haiying Liang, Yongli Xu, Yinlu Han, Qingbiao Shen, Chonghai Cai, and Tao Ye, Phys. Rev. C 95, 034614 (2017).
  24. [24]Matteo Vorabbi, Paolo Finelli, and Carlotta Giusti, Phys. Rev. C 93, 034619 (2016).
  25. [25]A. Funk, H. V. von Geramb, and K. A. Amos, Phys. Rev. C 64, 054003 (2001).
  26. [26]J. R. Taylor, Scattering Theory: The Quantum Theory on Nonrelativistic Collisions (John Wiley & Sons, Inc., New York, 1972).DOI
  27. [27]M. L. Goldberger, Collision Theory (John Wiley & Sons, Inc., New York, 1967).
  28. [28]A. Papastylianos, S. A. Sofianos, H. Fiedeldey, and E. O. Alt, Phys. Rev. C 42, 142 (1990).
  29. [29]E. O. Alt, L. L. Howell, M. Rauh, and S. A. Sofianos, Phys. Rev. C 49, 176 (1994).
  30. [30]P. Fernandez-Soler and E. Ruiz Arriola, Phys. Rev. C 96, 014004 (2017).
  31. [31]H. A. Antosiewicz, Bessel Functions of Fractional Order; in Handbook of Mathematical Functions With Formulas, Graphs, and Mathematical Tables, edited by Milton Abramowitz and Irene A. Stegun (National Bureau of Standards Applied Mathematics 55, 1972), p. 435.
  32. [32]K. Maleknejad and Kajani M. Tavassoli, Applied Mathematics and Computation 145, 623 (2003).
  33. [33]K. Maleknejad and Y. Mahmoudi, Applied Mathematics and Computation 149, 799 (2004).
  34. [34]B. Tavassoli Asady, M. Hadi Kajani, A. Vencheh, Applied Mathematics and Computation 163, 517 (2005).
  35. [35]K. Maleknejad and F. Mirzaee, Applied Mathematics and Computation 160, 579 (2005).
  36. [36]S. Yousefi and M. Razaghi, Mathematics and Computers in Simulation 70, 1 (2005).
  37. [37]E. Baboolian and A. Shahsavaran, Journal of Computational and Applied Mathematics 225, 87 (2009).
  38. [38]C. Cattania and A. Kudreyko, Applied Mathematics and Computation 215, 4164 (2010).
  39. [39]B. D. Keister and W. N. Polyzou, Adv. Nucl. Phys. 20, 225 (1991).
  40. [40]R. A. Arndt, I. I. Strakovsky and R.L. Workman, Phys. Rev. C 62, 034005 (2000).
  41. [41]R. A. Arndt, W. J. Briscoe, R. L. Workman, I. I. Strakovsky, http://gwdac.phys.gwu.edu/.
  42. [42]K. A. Wendt, R. J. Furnstahl, and S. Ramanan, Phys. Rev. C 86, 014003 (2012).

Paper details

vlnc:W_d89brex3
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents