betaFeedback

WorksPhysics and Astronomy

Indirect method for nuclear reactions and the role of the self energy

Abstract

When a nuclear species (e.g., a nucleon or a deuteron nucleus) propagating freely is made to collide with a target nucleus, its trajectory is modified by exchanging variable amounts of energy, mass, linear and angular momentum with the target, according to its interaction with the nuclear medium. By addressing this perturbation away from the free path, one hopes to learn something about the nature of the medium through which our probe propagates. This is the essence of the experimental use of nuclear reactions for the purpose of gathering information about nuclear structure. In order to deal with the structure and the reaction aspects of a specific experiment on the same footing, it is therefore desirable to identify a theoretical construct that embodies the modification of the propagation of a particle in the medium with respect to the free case, and use it both for the determination of the nuclear spectrum (structure) and for the calculation of scattering observables (reaction). A candidate for such an object is the self energy, and we will try in the present lectures to put it at the center stage in the formulation of scattering theory.

Introduction

When a nuclear species (e.g., a nucleon or a deuteron nucleus) propagating freely is made to collide with a target nucleus, its trajectory is modified by exchanging variable amounts of energy, mass, linear and angular momentum with the target, according to its interaction with the nuclear medium. By addressing this perturbation away from the free path, one hopes to learn something about the nature of the medium through which our probe propagates. This is the essence of the experimental use of nuclear reactions for the purpose of gathering information about nuclear structure. In order to deal with the structure and the reaction aspects of a specific experiment on the same footing, it is therefore desirable to identify a theoretical construct that embodies the modification of the propagation of a particle in the medium with respect to the free case, and use it both for the determination of the nuclear spectrum (structure) and for the calculation of scattering observables (reaction). A candidate for such an object is the self energy, and we will try in the present lectures to put it at the center stage in the formulation of scattering theory.

Let us be more specific, and focus on a nucleon xx colliding with some target nucleus AA. In all generality, the description of the scattering process can be written in terms of a many-body Hamiltonian, which will include the kinetic energy of the nucleon, the intrinsic Hamiltonian of the nucleus AA, and the interaction between the nucleon and all the nucleons of AA. However, as we will explicitly see below, this many-body Schrödinger equation can be reduced to a single-particle one, in which the nucleon propagates in an effective potential that includes the effects of the interaction with the medium. This effective potential is the self energy of the nucleon in the medium, and it is a complex, non-local, and energy-dependent operator. The imaginary part of this operator is responsible for the absorption of the nucleon in the target nucleus, i.e., the loss of flux with respect to the entrance (elastic) channel. The scattering wavefunction ψ0\psi_0 of the resulting (x+A)(x+A) two-body system associated with the propagation in the elastic channel can then be written in terms of the self energy, by means of the integro-differential Schrödinger equation

(E−T)ψ0(rxA)−∫Σ′(rxA,rxA′,E)ψ0(rxA′) d3rxA′=0,(1)(E-T)\psi_0(\mathbf{r}_{xA})-\int\Sigma'(\mathbf{r}_{xA},\mathbf{r}'_{xA},E)\psi_0(\mathbf{r}'_{xA})\,d^3r'_{xA}=0, \tag*{(1)}

where EE is the energy in the center of mass frame, TT is the kinetic energy operator acting on the rxA\mathbf{r}_{xA} relative coordinate1 of the nucleon with respect to the nucleus AA, and Σ′\Sigma' is the self energy (with typical units in nuclear physics applications [Σ′]=MeV⋅fm−3[\Sigma'] = \mathrm{MeV}\cdot\mathrm{fm}^{-3}, i.e., a volume energy density), which we will also refer to as the optical potential. As we noted above, it is complex, energy-dependent and non-local, as a consequence of the coupling with the intrinsic degrees of freedom of the composite (x+Ax+A) system (see Sect. 2.2). Non locality implies that Eq. (1) is an integro-differential equation, and the value of the associated eigenfunction ψ0\psi_0 at a given point rxA\mathbf{r}_{xA} depends on the values of the wavefunction at all other points rxA′\mathbf{r}'_{xA}. In order to clear up the notation, in what follows we will denote the action of the self energy (or any other non local operator) on a function f(r)f(\mathbf{r}) of the space coordinates as

∫Σ′(r,r′,E)f(r′) d3r′≡Σ(r,r′,E)f(r),(2)\int\Sigma'(\mathbf{r},\mathbf{r}',E) f(\mathbf{r}')\,d^3r' \equiv\Sigma(\mathbf{r},\mathbf{r}',E) f(\mathbf{r}), \tag*{(2)}

so that Eq. (1) can be written as

(E−T−Σ(rxA,rxA′,E))ψ0(rxA)=0,(3)\left(E-T-\Sigma(\mathbf{r}_{xA},\mathbf{r}'_{xA},E)\right)\psi_0(\mathbf{r}_{xA})=0, \tag*{(3)}

Let us make now an important point. The self energy Σ\Sigma (or Σ′\Sigma') can, in principle, be derived explicitly and with arbitrary accuracy from the many-body Hamiltonian making use of well defined quantum many-body techniques [1]. In this sense, solving Eq. (1) is equivalent to solving the many-body Schrödinger equation, and obtaining Σ\Sigma is, in general, equivalently hard (see Sect. 2.2). In the same way in which there are a number of techniques to solve the many-body Schrödinger equation, these methods can also be used to obtain the self energy. We will stress in these lectures the use of nuclear field theory (NFT, [2, 3]) in order to highlight the role that elementary modes of excitations (single-(quasi) particle states, collective pairing and surface vibrations, etc.) play both in structure and reactions.

Indirect measurements of nuclear cross sections

It is often the case that nuclear reactions that are important for societal applications or basic science are difficult to measure directly in existing facilities for accelerated beams. The difficulty might be associated with an exceedingly small cross section, as the ones obtained at beam energies well below the Coulomb barrier, or with the impossibility of devising a short-lived target, as is the case for neutron-induced reactions on unstable isotopes. The fruitful line of experimental research addressing this issue with alternative reactions (indirect measurements) has been developed in parallel to the theory needed to make the connection between the observed data and the desired cross sections of the reaction under study [4].

Within this context, the use of the self energy in Eq. (3) would provide the elastic and reaction cross sections associated with the x+Ax+A reaction, which can be computed from the asymptotic part of the wavefunction ψ0\psi_0 (see below). We will show in these lectures how the self energy of the x−Ax-A system can be used to compute the cross section associated with the indirect measurement, which is defined as the inclusive measurement of a fragment bb produced in the reaction a(=x+b)+A→b+(x+A)a(=x+b)+A\to b+(x+A). Important examples of such indirect measurements are deuteron-induced reactions like 9Li(d,p)10Li{}^{9}\mathrm{Li}(d,p){}^{10}\mathrm{Li} (see Sect. 4.1) and 40Ca(d,p)41Ca{}^{40}\mathrm{Ca}(d,p){}^{41}\mathrm{Ca} (see Sect. 4.2).

The formalism to be presented in these lectures has been essentially introduced in the 80’s [5, 6, 7] and recently revived [8, 9, 10, 11], in different contexts and variations. In order to be specific about this particular implementation, we will call it the Green’s Function Transfer (GFT) formalism. The perturbative derivation presented here differs from earlier presentations, and it might provide some different insights, including the possibility of estimating the validity of the spectator approximation (see Sect. 6), and a more transparent connection with standard 2-body scattering theory techniques such as the RR-matrix theory (see Sect. 5).

Elastic and inelastic scattering

General formalism and definitions

Let us summarize in this Section some standard scattering theory results associated with the elastic scattering between two nuclei xx and AA. For a more detailed account, we refer the reader to some textbooks [12, 13, 14, 15], where a comprehensive exposition of quantum scattering theory in general, and nuclear reactions theory in particular, can be found. Within the context of the present lectures, we will treat xx as a structureless particle, while we will take the structure of both AA and B≡A+xB\equiv A+x into full consideration. With this caveat in mind, the Hamiltonian of the system is2^{2}2

H=hB(rxA,ξ)=Tx+hA(ξ)+VxA(rxA,ξ)(4)H=h_{B}(r_{xA},\xi)=T_{x}+h_{A}(\xi)+V_{xA}(r_{xA},\xi) \tag*{(4)}

where ξ\xi stands for all the spatial and spin coordinates needed to describe the microscopic structure of AA, while rxAr_{xA} is the relative coordinate of the xx-AA system. For simplicity, we will ignore the intrinsic spins of xx and AA. The intrinsic Hamiltonians of nuclei AA and BB are hAh_{A} and hBh_{B}, respectively, while TxT_{x} is the kinetic energy associated to the xx-AA relative motion. The corresponding Schrödinger equation

(E−H)Ψ(rxA,ξ)=0(5)(E-H)\Psi(r_{xA},\xi)=0 \tag*{(5)}

can be rewritten as

(E−Tx−hA(ξ))Ψ(rxA,ξ)=VxA(rxA,ξ)Ψ(rxA,ξ)(6)(E-T_{x}-h_{A}(\xi))\Psi(r_{xA},\xi)=V_{xA}(r_{xA},\xi)\Psi(r_{xA},\xi) \tag*{(6)}

The previous expression suggests a formal solution in terms of the incident wave (defined with a suitable boundary condition specifying the beam direction, etc., see Eq. (12) below) associated with the unperturbed wavefunction,

(E−Tx−hA)Ψ0=0.(7)(E-T_x-h_A)\Psi_0=0. \tag*{(7)}

and the inverse operator of the unperturbed Hamiltonian

G0(rxA,ξ,E)=lim⁡η→0(E−Tx−hA+iη)−1,(8)G_0(r_{xA},\xi,E)=\lim_{\eta\to0}(E-T_x-h_A+i\eta)^{-1}, \tag*{(8)}

which is the unperturbed many-body Green’s function, also called the propagator. The small quantity η\eta should be made to vanish after performing the inversion operation. Without its inclusion, the Green’s function would be singular for values of the energy equal to the (real) eigenvalues of the Hamiltonian, and the inversion procedure would be ill defined. In addition, taking η\eta to be a positive real vanishingly small number ensures that the second term after the equal sign in the equation below is proportional to an outgoing spherical wave, thus enforcing the right asymptotic behaviour of the scattered wave (see, e.g., [1, 14]). It can be readily verified, by applying the operator (E−Tx−hA)(E-T_x-h_A) to both sides of the equal sign of the equation below, that the wavefunction defined by the following Lippmann-Schwinger equation

Ψ=Ψ0+G0(E)VxAΨ(9)\Psi=\Psi_0+G_0(E)V_{xA}\Psi \tag*{(9)}

indeed satisfies Eq. (6). This equation can also be expressed in an equivalent way3,

Ψ=Ψ0+G(E)VxAΨ0.(10)\Psi=\Psi_0+G(E)V_{xA}\Psi_0. \tag*{(10)}

where

G(rxA,ξ,E)=lim⁡η→0(E−Tx−hA−VxA+iη)−1(11)G(r_{xA},\xi,E)=\lim_{\eta\to0}(E-T_x-h_A-V_{xA}+i\eta)^{-1} \tag*{(11)}

is the total many-body Green’s function. The first term on the right hand side of Eqs. (9) and (10) is the incident wave, describing the incident channel imposed by our boundary condition, while the second term is the scattered wave.

The unperturbed wavefunction solution of Eq. (7) associated with standard scattering boundary conditions is

Ψ0(rxA,ξ)=Φ0(ξ)F(kx,rxA),(12)\Psi_0(r_{xA},\xi)=\Phi_0(\xi)F(\mathbf{k}_x,r_{xA}), \tag*{(12)}

where F(kx,rxA)F(\mathbf{k}_x,r_{xA}) is a free incoming incident plane wave with momentum kx\mathbf{k}_x along the zz axis4, and Φ0\Phi_0 is the ground state of the nucleus AA (see Eq. (16) below).

G=G0+G0VxAG=G0+GVxAG0.(13)G=G_0+G_0V_{xA}G=G_0+GV_{xA}G_0. \tag*{(13)}
G0=(E−Tx−hA−e2ZAZxrxA)−1.(14)G_0=\left(E-T_x-h_A-\frac{e^2Z_AZ_x}{r_{xA}}\right)^{-1}. \tag*{(14)}

Elastic scattering We can expand the wavefunction describing the xx-AA system in terms of the intrinsic states of AA as

Ψ(rxA,ξ;E)=∑iΦi(ξ)ψi(rxA),(15)\Psi(\mathbf{r}_{xA},\xi;E)=\sum_i \Phi_i(\xi)\psi_i(\mathbf{r}_{xA}), \tag*{(15)}

where {ψi(rxA)}\{\psi_i(\mathbf{r}_{xA})\} is a complete set of orthogonal (not necessarily normalized) single-particle channel wavefunctions, and {Φi(ξ)}\{\Phi_i(\xi)\} is also a complete set of eigenfunctions of hAh_A,

(ϵi−hA(ξ))Φi(ξ)=0.(16)(\epsilon_i-h_A(\xi))\Phi_i(\xi)=0. \tag*{(16)}

In order to obtain an equation for the elastic (i=0i=0) channel, we project the first (7) on Φ0\Phi_0, obtaining a Lippmann-Schwinger equation for the one-body elastic channel wavefunction,

ψ0(rxA)=F+⟨Φ0∣G(E)VxA∣Φ0⟩F.(17)\psi_0(\mathbf{r}_{xA})=F+\langle\Phi_0|G(E)V_{xA}|\Phi_0\rangle F. \tag*{(17)}

Inelastic scattering (17) can be easily generalized to the description of inelastic scattering, where the nucleus AA has been excited to a state ii of its spectrum,

ψi(rxA)=⟨Φi∣G(E)VxA∣Φ0⟩F.(18)\psi_i(\mathbf{r}_{xA})=\langle\Phi_i|G(E)V_{xA}|\Phi_0\rangle F. \tag*{(18)}

Note that in this case the free, unscattered wavefunction FF does not appear, testifying to the fact that, as part of our asymptotic boundary condition, only the ground state of AA is in the incident channel. The population of the inelastic (i≠0i\ne0) channels is entirely due to the scattering process driven by the interaction VxAV_{xA}.

Cross section and TT-matrix The cross section is associated with the asymptotic form of ψi\psi_i, which describes the observed system far away from the interaction region, where the detectors are located. The asymptotic form of the elastic and inelastic channel wavefunctions can be obtained from Eqs. (17) and (18), respectively, by making use of the asymptotic form of the Green’s function (see Eq. (21) below),

ψi(rxA→∞)=F(rxA→∞)δi0+⟨Φi∣G(rxA→∞)VxA∣Φ0⟩F=F(rxA→∞)δi0+O(ki,rxA)⟨ψiΦi∣VxA∣Φ0F⟩=F(rxA→∞)δi0+O(ki,rxA)Ti0,(19)\begin{aligned} \psi_i(\mathbf{r}_{xA}\to\infty)&=F(\mathbf{r}_{xA}\to\infty)\delta_{i0}+\langle\Phi_i|G(\mathbf{r}_{xA}\to\infty)V_{xA}|\Phi_0\rangle F\\ &=F(\mathbf{r}_{xA}\to\infty)\delta_{i0}+O(\mathbf{k}_i,\mathbf{r}_{xA})\langle\psi_i\Phi_i|V_{xA}|\Phi_0F\rangle\\ &=F(\mathbf{r}_{xA}\to\infty)\delta_{i0}+O(\mathbf{k}_i,\mathbf{r}_{xA})T_{i0}, \tag*{(19)} \end{aligned}

where O(ki,rxA)O(\mathbf{k}_i,\mathbf{r}_{xA}) is an outgoing wave with momentum ki=2μ(E−ϵi)/ℏ\mathbf{k}_i=\sqrt{2\mu(E-\epsilon_i)}/\hbar, the amplitude

Ti0=⟨ψiΦi∣VxA∣Φ0F⟩(20)T_{i0}=\langle\psi_i\Phi_i|V_{xA}|\Phi_0F\rangle \tag*{(20)}

In this case, F(kx,rxA)F(\mathbf{k}_x,\mathbf{r}_{xA}) would be a Coulomb function, without otherwise affecting the overall discussions in these lectures.

is an element of the TT-matrix, and we have used the asymptotic form of the Green’s function,

⟨Φi∣G(rxA→∞,rxA′,ξ,ξ′)≡∫Φi(ξ)∗G(rxA→∞,rxA′,ξ,ξ′) dξ=O(ki,rxA)Φi(ξ′)ψi(rxA′).(21)\langle\Phi_i | G(r_{xA} \to\infty, r'_{xA}, \xi, \xi') \equiv\int\Phi_i(\xi)^* G(r_{xA} \to\infty, r'_{xA}, \xi, \xi') \,\mathrm{d}\xi= O(k_i,r_{xA}) \Phi_i(\xi') \psi_i(r'_{xA}). \tag*{(21)}

The differential cross section can be expressed in terms of the TT-matrix (see, e.g., [12, 13]),

dσi0dΩ=μ2ki4π2ℏ4k0∣Ti0∣2.(22)\frac{\mathrm{d}\sigma_{i0}}{\mathrm{d}\Omega} = \frac{\mu^2 k_i}{4\pi^2 \hbar^4 k_0} \lvert T_{i0}\rvert^2. \tag*{(22)}

R-matrix parametrization of the TT-matrix. In the context of 2-body quantum scattering theory, cross sections can always be calculated in terms of the phase shift existing between the free and scattered wavefunctions in the asymptotic region, i.e., sufficiently far away from the scattering center (see, e.g. [12, 13, 14]). This phase shift uniquely determines the asymptotic behavior of the scattered wavefunction, and it can in turn be inferred from the (inverse) logarithmic derivative of the wavefunction at some arbitrary radius in the asymptotic region. This quantity (divided by the radius at which it has been calculated) becomes a dimensionless matrix Rij(E)R_{ij}(E) (the so-called RR-matrix, dependent on the center of mass energy EE) when we take into account the population of different reaction channels (labeled by the indexes ii, jj), each one of them associated with its own wavefunction.

It can be shown ([16, 17]) that the energy-dependence of the ii, jj element of the RR-matrix can be exactly parametrized in terms of an infinite set of real, energy-independent parameters γip\gamma_{ip}, γjp\gamma_{jp}, EpE_p as

Rij(E)=∑p∞γipγjpEp−E.(23)R_{ij}(E) = \sum_p^\infty\frac{\gamma_{ip}\gamma_{jp}}{E_p-E}. \tag*{(23)}

This is known as the Wigner-Eisenbud parametrization. The γip\gamma_{ip} are known as the reduced partial widths (with dimensions of [γip]=E1/2[\gamma_{ip}] = E^{1/2}), and the EpE_p are the corresponding poles of the RR-matrix.

Although these quantities can be just considered as parameters to be adjusted in order to fit the observed experimental cross section (see below), they have a specific interpretation in terms of the so-called calculable RR-matrix formalism ([16, 17]). The energies EpE_p are the eigenvalues of the eigenvalue problem associated with the Hamiltonian of the system and specific boundary conditions at the radius $a where the RR-matrix is calculated, while the γip\gamma_{ip} are related to the amplitude of the corresponding eigenfunction at this radius. A variety of boundary conditions can be chosen giving rise to different practical implementations of RR-matrix theory, but they all have in common that the resulting spectrum is discrete (see e.g. [17, 18]).

Needless to say, these quantities depend on the choice of the RR-matrix radius aa and the specific boundary conditions chosen. Because of the non-physical boundary conditions implemented, the energies EpE_p do not match the eigenvalues of the physical problem, which corresponds instead to the eigenvalue problem associated with the physical boundary conditions (exponentially decaying for bound states, oscillating for scattering states). Similarly, the energies EpE_p do not correspond to the physical resonances of the system, which are instead associated with the poles of the TT-matrix (see Eq. (26)).

Like any other asymptotic quantity, the TT-matrix can then be derived in terms of the RR-matrix parameters, resulting in {[16]}[16]

Tij=Pi(E)Pj(E)∑pq∞γip[A−1(E)]pqγjq,(24)T_{ij}=\sqrt{P_i(E)P_j(E)}\sum_{pq}^{\infty}\gamma_{ip}\left[A^{-1}(E)\right]_{pq}\gamma_{jq}, \tag*{(24)}

with

Apq(E)=(Ep−E)δpq−∑cγcpγcq(Sc(E)+iPc(E)).(25)A_{pq}(E)=(E_p-E)\delta_{pq}-\sum_c\gamma_{cp}\gamma_{cq}\left(S_c(E)+iP_c(E)\right). \tag*{(25)}

The energy dependence of the TT-matrix is thus contained in the (E−Ep)(E-E_p) term in the denominator, and in the penetrability (Pi(E)P_i(E)) and shift (Si(E)S_i(E)) factors, which are known combinations of the Bessel functions and their derivatives (or Coulomb functions, if both particles associated with the reaction channel ii are charged) {[17]}[17]. The physical resonances (i.e., sharp structures in the strength function dσ(E)/dEd\sigma(E)/dE as a function of the energy) are the complex poles EpE_p of the TT-matrix, which can be identified as the solutions of the implicit equation

App(Ep)=0,(26)A_{pp}(E_p)=0, \tag*{(26)}

for every pp. Since these are physical, experimentally observable quantities, they do not depend on the choice of the RR-matrix radius or the boundary conditions used to calculate the RR-matrix. The imaginary part of EpE_p is associated with the width of the resonance, while its real part is the resonance energy.

Any practical implementation of the so-called phenomenological R-matrix usually consists in fitting a finite number of energies EpE_p and partial widths γip\gamma_{ip} to an experimental excitation function dσ(E)/dEd\sigma(E)/dE, which is proportional to the modulus square of the TT-matrix (24) (see Eq. (22)). Since the energy dependence of the TT-matrix is then known explicitly, the cross section can be calculated at any energy, including the low energies of astrophysical interest, where direct measurements are often not feasible (see e.g. {[19, 20]}[19, 20]).

Born series and the Distorted Wave Born Approximation The expression (9) can be used to express the wavefunction as a perturbative expansion in terms of powers of the potential VxAV_{xA}, known as the Born series,

Ψ=Ψ0+G0(E)VxA(Ψ0+G0(E)VxA(Ψ0+…)).(27)\Psi=\Psi_0+G_0(E)V_{xA}\left(\Psi_0+G_0(E)V_{xA}(\Psi_0+\ldots)\right). \tag*{(27)}

The first order term of this expansion,

Ψ≈Ψ0+G0(E)VxAΨ0(28)\Psi\approx\Psi_0+G_0(E)V_{xA}\Psi_0 \tag*{(28)}

is the first order Plane Wave Born Approximation. However, this is often not the most convenient way to express the perturbation expansion for the wavefunction. Instead, one often defines an arbitrary auxiliary potential U0(rxA)U_0(\mathbf{r}_{xA}) and, instead of Eq. (6), has

(E−Tx−hA(ξ)−U0(rxA))Ψ(rxA,ξ)=(VxA(rxA,ξ)−U0(rxA))Ψ(rxA,ξ).(29)\left(E-T_x-h_A(\xi)-U_0(\mathbf{r}_{xA})\right)\Psi(\mathbf{r}_{xA},\xi)=\left(V_{xA}(\mathbf{r}_{xA},\xi)-U_0(\mathbf{r}_{xA})\right)\Psi(\mathbf{r}_{xA},\xi). \tag*{(29)}

The solution to the above equation can now be expressed in the two equivalent following ways

Ψ=Ψ~0+G~0(E)(VxA(rxA,ξ)−U0(rxA))Ψ;Ψ=Ψ~0+G(E)(VxA(rxA,ξ)−U0(rxA))Ψ~0,(30)\begin{aligned} \Psi&=\widetilde{\Psi}_0+\widetilde{G}_0(E)\left(V_{xA}(\mathbf{r}_{xA},\xi)-U_0(\mathbf{r}_{xA})\right)\Psi;\\ \Psi&=\widetilde{\Psi}_0+G(E)\left(V_{xA}(\mathbf{r}_{xA},\xi)-U_0(\mathbf{r}_{xA})\right)\widetilde{\Psi}_0, \tag*{(30)} \end{aligned}

where

G~0=lim⁡η→0(E−Tx−hA−U0+iη)−1,(31)\widetilde{G}_0=\lim_{\eta\to0}\left(E-T_x-h_A-U_0+i\eta\right)^{-1}, \tag*{(31)}

and

Ψ~0(rxA,ξ)=Φ0(ξ)F~(rxA).(32)\widetilde{\Psi}_0(\mathbf{r}_{xA},\xi)=\Phi_0(\xi)\widetilde{F}(\mathbf{r}_{xA}). \tag*{(32)}

Now, F~\widetilde{F} is the so-called distorted wave, and is the solution of

(E−Tx−hA−U0)F~=0.(33)\left(E-T_x-h_A-U_0\right)\widetilde{F}=0. \tag*{(33)}

The first order approximation to the associated power series,

Ψ≈Ψ~0+G~0(E)(VxA(rxA,ξ)−U0(rxA))Ψ~0(34)\Psi\approx\widetilde{\Psi}_0+\widetilde{G}_0(E)\left(V_{xA}(\mathbf{r}_{xA},\xi)-U_0(\mathbf{r}_{xA})\right)\widetilde{\Psi}_0 \tag*{(34)}

is know as the first order Distorted Wave Born Approximation (DWBA), and it is widely used in practice (see e.g. [12, 13]). The freedom in choosing the auxiliary potential U0U_0 can be used to pick one which makes the matrix elements of (VxA(rxA,ξ)−U0(rxA))\left(V_{xA}(\mathbf{r}_{xA},\xi)-U_0(\mathbf{r}_{xA})\right) relevant for the calculation of Ψ\Psi as small as possible, in order to speed up the convergence of the power series.

In what follows, all the derivations made using G0G_0 and FF remain valid if we make the substitutions

G0→G~0;F→F~;VxA→VxA−U0,(35)G_0\to\widetilde{G}_0;\quad F\to\widetilde{F};\quad V_{xA}\to V_{xA}-U_0, \tag*{(35)}

even when not stated explicitly. The shift of the zero-order many-body wavefunction implied in going from Ψ0\Psi_0 to Ψ~0\widetilde{\Psi}_0 is common practice in actual nuclear reaction calculations.

The optical potential

Our ability to calculate elastic and inelastic cross sections according to Eqs. (17) and (18) rely on being able to obtain matrix elements of the Green’s function GG, possibly within some reasonable approximations. We now show how the knowledge of the optical potential (which we show to be equivalent to the self energy defined in Eq. (3)) provides a practical way to address this problem. It is essentially a tautology to say that the calculation of the optical potential allows for the calculation of the elastic scattering wavefunction, but we will use this explicit connection in a less trivial context in § 3.

Let us start by looking for a one-body Schrödinger equation for the wavefunction ψ0\psi_0 associated to elastic scattering for an energy EE. Projecting the many-body Schrödinger equation

(E−H)Ψ=0(36)(E-H)\Psi=0 \tag*{(36)}

on the states Φi\Phi_i, one obtains a set of coupled equations for the one-body states ψi\psi_i (see Eq. (15)),

(E0−Tx−V00(rxA))ψ0(rxA)=∑i≠0V0i(rxA)ψi(rxA)(Ei−Tx−Vii(rxA))ψi(rxA)=Vi0(rxA)ψ0(rxA)+∑j≠0,iVij(rxA)ψj(rxA)(i≠0),(37)\begin{aligned} (E_0-T_x-V_{00}(r_{xA}))\psi_0(r_{xA})&=\sum_{i\ne0}V_{0i}(r_{xA})\psi_i(r_{xA})\\ (E_i-T_x-V_{ii}(r_{xA}))\psi_i(r_{xA})&=V_{i0}(r_{xA})\psi_0(r_{xA})+\sum_{j\ne0,i}V_{ij}(r_{xA})\psi_j(r_{xA})\quad(i\ne0), \tag*{(37)} \end{aligned}

where the coupling potentials are defined as

Vij(rxA)=∫dξ Φi∗(ξ)VxA(rxA,ξ)Φj(ξ),(38)V_{ij}(r_{xA})=\int d\xi\,\Phi_i^*(\xi)V_{xA}(r_{xA},\xi)\Phi_j(\xi), \tag*{(38)}

and Ei=E−ϵiE_i=E-\epsilon_i. The desired solution can be obtained within the Coupled Channels approach by solving numerically the set of coupled differential equations (37) making use of some reasonable approximations [15, 21, 22]. Alternatively, one can use the propagator (Green’s function) restricted to the space of the excited states of the x+Ax+A system,

GijQ(rxA;E)=lim⁡η→0(E−(Tx+ϵj)δij−Vij+iη)−1(i,j≠0)(39)G^Q_{ij}(r_{xA};E)=\lim_{\eta\to0}\left(E-(T_x+\epsilon_j)\delta_{ij}-V_{ij}+i\eta\right)^{-1}\quad(i,j\ne0) \tag*{(39)}

to solve for the ψi\psi_i’s in terms of ψ0\psi_0,

ψi=∑j≠0GijQVj0ψ0,(40)\psi_i=\sum_{j\ne0}G^Q_{ij}V_{j0}\psi_0, \tag*{(40)}

and substitute in the first equation of (37),

(E0−Tx−V(rxA,E))ψ0(rxA)=0.(41)(E_0-T_x-\mathcal{V}(r_{xA},E))\psi_0(r_{xA})=0. \tag*{(41)}

The non-local, complex, and energy dependent operator

V(rxA,E)=V00+∑ijV0iGijQ(E)Vj0(42)\mathcal{V}(r_{xA},E)=V_{00}+\sum_{ij}V_{0i}G^Q_{ij}(E)V_{j0} \tag*{(42)}

is the optical potential. The superscript QQ in the Green’s function indicates that it is restricted to the portion of the Hilbert space spanned by the excited states, excluding the ground state ψ0\psi_0. In other words,

GQψ0=0.(43)G^Q\psi_0=0. \tag*{(43)}

Following the standard notation of [23], we will call this subspace the QQ space.

The optical potential is seen to be equivalent to the self energy by comparing Eq. (41) with Eq. (3). The discussion in the present Section also illustrates how the one-body Eq. (41) is completely equivalent to solving the coupled equations (37) for the elastic channel, and, therefore, to diagonalizing the Hamiltonian, as mentioned in the Introduction.

The direct numerical implementation of Eq. (42) in a truncated QQ model space is not the only way to address the calculation of the optical potential. For example, in the NFT approach to nuclear structure and reactions (see, e.g., [24, 25]), the self energy is calculated making use of Feynman diagrams, which provide a systematic way to take into account the coupling of single-particle motion with collective surface and pairing vibrations, which are calculated making use of the Quasi-Particle Random Phase Approximation (QRPA) wavefunctions. In this case, the self energy is obtained as a perturbative expansion in terms of the particle-vibration coupling vertices, which provide a specific model for the couplings V0iV_{0i}. As a result of this perturbative diagonalization, the self energy obtained is non-local, complex, and energy dependent, as expected from Eq. (42). In Sect. 4.1 we provide a specific application addressing the 9Li(d,p){}^{9}\mathrm{Li}(d,p) reaction making use of NFT. We refer to the contribution of F. Barranco et al. to this volume for more details on the NFT approach.

The energy dependence of the optical potential (42) is contained in the virtual propagation in the QQ space, encoded in the propagator GQ(E)G^{Q}(E). It is thus a consequence of the connection of the ground state with the excited states described by the couplings V0iV_{0i}. In order to gain some insight on the nature of its imaginary part, let us consider the matrix element of the propagator in a eigenstate ∣μ⟩|\mu\rangle of the Hamiltonian restricted to the QQ space,

(εμ−Tx−V(rxA))∣μ⟩=0,(44)\left(\varepsilon_{\mu}-T_{x}-V(r_{xA})\right)|\mu\rangle=0, \tag*{(44)}

then

GμQ(E)=⟨μ∣GQ(E)∣μ⟩=lim⁡η→01E−εμ+iη.(45)G_{\mu}^{Q}(E)=\langle\mu|G^{Q}(E)|\mu\rangle=\lim_{\eta\to0}\frac{1}{E-\varepsilon_{\mu}+i\eta}. \tag*{(45)}

Let us extract the imaginary part of this matrix element,

Im⁡GμQ(E)=Im⁡(lim⁡η→01E−εμ+iη)=Im⁡(lim⁡η→0E−εμ−iη(E−εμ+iη)(E−εμ−iη))=Im⁡(lim⁡η→0E−εμ−iη(E−εμ)2+η2)=−πδ(E−εμ),(46)\begin{aligned} \operatorname{Im}G_{\mu}^{Q}(E)=\operatorname{Im}\left(\lim_{\eta\to0}\frac{1}{E-\varepsilon_{\mu}+i\eta}\right)=\operatorname{Im}\left(\lim_{\eta\to0}\frac{E-\varepsilon_{\mu}-i\eta}{(E-\varepsilon_{\mu}+i\eta)(E-\varepsilon_{\mu}-i\eta)}\right) \\ &=\operatorname{Im}\left(\lim_{\eta\to0}\frac{E-\varepsilon_{\mu}-i\eta}{(E-\varepsilon_{\mu})^{2}+\eta^{2}}\right)=-\pi\delta(E-\varepsilon_{\mu}), \tag*{(46)} \end{aligned}

where we have used the representation of the Dirac delta function,

δ(x)=lim⁡η→01πηx2+η2.(47)\delta(x)=\lim_{\eta\to0}\frac{1}{\pi}\frac{\eta}{x^{2}+\eta^{2}}. \tag*{(47)}

In keeping with the fact that the energy dependence of the optical potential is entirely contained in the propagator GQ(E)G^{Q}(E), the above result illustrates the fact that the optical potential becomes complex whenever the energy EE matches one of the eigenvalues εμ\varepsilon_\mu of the Hamiltonian in the QQ space. As a consequence, the Hamiltonian describing the dynamics of the elastic channel is not Hermitian, and the flux in this channel is not conserved. Since it can be shown that this imaginary part is always negative (see, e.g., [26], where a more rigorous derivation of the above results can also be found), this flux non-conservation consists in a loss (rather than an increase) of flux in the elastic channel, which leaks into the QQ space. This feeding of the excited states of the x+Ax+A system is what we call a reaction event, which, of course, is an energy-conserving process. The explicit connection of the imaginary part of the optical potential with the reaction cross section is expressed by Eq. (53) below.

The elastic scattering wavefunction can then be written as the solution of the Lippmann-Schwinger equation involving exclusively single-particle operators,

ψ0(rxA)=F+G0(E0)V(E)ψ0,(48)\psi_0(\mathbf{r}_{xA}) = F + G_0(E_0)\mathcal{V}(E)\psi_0, \tag*{(48)}

where

G0(rxA,E)=(E−Tx)−1(49)G_0(\mathbf{r}_{xA},E) = (E-T_x)^{-1} \tag*{(49)}

is the free single-particle Green’s function. Eq. (48) can be written in a different exact form (see Eq. (9)),

ψ0=F+G(E0)V(E)F,(50)\psi_0 = F + G(E_0)\mathcal{V}(E)F, \tag*{(50)}

where GG is the single-particle Green’s function,

G(rxA,E)=(E−Tx−V(E))−1.(51)G(\mathbf{r}_{xA},E) = (E-T_x-\mathcal{V}(E))^{-1}. \tag*{(51)}

We can now use (17) to identify

⟨Φ0∣G(E)VxA∣Φ0⟩=G(E0)V(E).(52)\langle\Phi_0|G(E)V_{xA}|\Phi_0\rangle= G(E_0)\mathcal{V}(E). \tag*{(52)}

Another standard result of scattering theory is that the total reaction cross section, corresponding to the sum over all energetically allowed excitations of the xx-AA system, can be obtained as the expectation value of the imaginary part of the optical potential over the elastic wave function (see, e.g., [12, 13, 15]),

σR=2μℏ2kx⟨ψ0∣Im⁡V∣ψ0⟩,(53)\sigma_R = \frac{2\mu}{\hbar^2 k_x}\langle\psi_0|\operatorname{Im}\mathcal{V}|\psi_0\rangle, \tag*{(53)}

where μ\mu is the reduced mass of the x−Ax-A system, and kxk_x is the associated wave number in the center of mass frame.

Indirect measurements and GFT

Introduction: effective 3-body problem and the spectator approximation

Let us now consider a composite system a≡x+ba \equiv x+b impinging on the nucleus AA (see Figure 1). We want to deal here with measurements in which the projectile aa breaks into its constituents xx and bb, and only bb is detected with an energy EbE_b. The fragment bb will also be assumed to be structureless, and its propagation will be described within the spectator approximation. The Hamiltonian can be written in two equivalent representations,

Image not extracted

Figure 1. Schematic representation of the inclusive reaction a+A→b+(x+A)a+A \to b+(x+A), where the fragment bb is detected, and the nucleon xx is not. The interaction between bb and the rest of the system is approximated with an optical potential. The spectator approximation assumes that the interaction between xx and AA can be treated as a two-body problem, in which the fragment bb acts as a spectator.

H=HB=Tx+hA(ξ)+VxA(rxA,ξ)+Tb+UbB(rbB)+ΔVB=HA=Ta+ha(rxb)+hA(ξ)+VxA(rxA,ξ)+UbA(rbA)+ΔVA,(54)\begin{aligned} H = H_B &= T_x + h_A(\xi) + V_{xA}(\mathbf{r}_{xA},\xi) + T_b + U_{bB}(\mathbf{r}_{bB}) + \Delta V_B \\ &= H_A = T_a + h_a(\mathbf{r}_{xb}) + h_A(\xi) + V_{xA}(\mathbf{r}_{xA},\xi) + U_{bA}(\mathbf{r}_{bA}) + \Delta V_A, \tag*{(54)} \end{aligned}

where hah_a is the intrinsic Hamiltonian of the nucleus aa, and we define

ΔVB=Vxb(rxb)+VbA(rbA,ξ)−UbB(rbB);ΔVA=VbA(rbA,ξ)−UbA(rbA).(55)\begin{aligned} \Delta V_B &= V_{xb}(\mathbf{r}_{xb}) + V_{bA}(\mathbf{r}_{bA},\xi) - U_{bB}(\mathbf{r}_{bB}); \\ \Delta V_A &= V_{bA}(\mathbf{r}_{bA},\xi) - U_{bA}(\mathbf{r}_{bA}). \tag*{(55)} \end{aligned}

Methods exist that deal with the effective three-body problem (x+b+A)(x+b+A) in an essentially exact way, such as the Faddeev equations [27, 28, 29], and the Continuum Discretized Coupled Channels (CDCC) method [30, 31, 32]. However, these methods are computationally expensive, and they are not always practical for the analysis of experimental data. The spectator approximation is a relatively economical way to reduce the three-body problem to two two-body ones ((x+A)(x+A) and (b+(x+A))(b+(x+A))), in which the fragment bb acts as a spectator. In addition, the role of the interaction between xx and AA (and, therefore, of the reaction corresponding to the direct measurement we wish to study) in the experimental cross section for the observation of the fragment bb is cleanly highlighted. In the context of the spectator approximation the interaction between the fragment bb and the rest of the system is approximated with the optical potentials UbBU_{bB} and UbAU_{bA}, so we assume ΔVA≈0\Delta V_A \approx0, and ΔVB≈0\Delta V_B \approx0. Within this context, the coupling of the fragment bb to the intrinsic structure of the systems AA and BB is only phenomenologically included in terms of an imaginary part of the potentials UbBU_{bB}, UbAU_{bA}. In Sect. 6 we will give a more precise meaning to the condition that the operators ΔVA\Delta V_A and ΔVB\Delta V_B are negligible.

Elastic scattering and inclusive cross section

Let us write the Schrödinger equation for the total wavefunction Ψ(rxA,rbA,ξ)\Psi(r_{xA}, r_{bA}, \xi) assuming ΔVA=0\Delta V_A = 0,

(E−Ta−ha(rxb)−hA(ξ))Ψ=(VxA(rxA,ξ)+UbA(rbA))Ψ,(56)(E-T_a-h_a(r_{xb})-h_A(\xi))\Psi=(V_{xA}(r_{xA},\xi)+U_{bA}(r_{bA}))\Psi, \tag*{(56)}

which can be formally solved in terms of many-body Green’s function,

Ψ=Ψ0+GI(E)[VxA(rxA,ξ)+UbA(rbA)]Ψ0,(57)\Psi=\Psi_0+G_I(E)\left[V_{xA}(r_{xA},\xi)+U_{bA}(r_{bA})\right]\Psi_0, \tag*{(57)}

where the have defined the unperturbed wavefunction

Ψ0=ϕa(rxb)Φ0(ξ)F(ka,ra),(58)\Psi_0=\phi_a(r_{xb})\Phi_0(\xi)F(k_a,r_a), \tag*{(58)}

where rar_a is the coordinate of the center of mass of the nucleus aa, and F(ka,ra)F(k_a,r_a) is a free incoming wave (see previous section). The unperturbed wavefunction satisfies

(E−Ta−ha(rxb)−hA(ξ))Ψ0=(E−Ta−ϵ0a−ϵ0A)Ψ0=(E0−Ta)Ψ0=0,(59)(E-T_a-h_a(r_{xb})-h_A(\xi))\Psi_0=(E-T_a-\epsilon_0^a-\epsilon_0^A)\Psi_0=(E_0-T_a)\Psi_0=0, \tag*{(59)}

and

(ϵ0a−ha(rxb))ϕa=0,(ϵ0A−hA(ξ))Φ0=0,E0=E−ϵ0A−ϵ0a,(60)(\epsilon_0^a-h_a(r_{xb}))\phi_a=0,\qquad(\epsilon_0^A-h_A(\xi))\Phi_0=0,\qquad E_0=E-\epsilon_0^A-\epsilon_0^a, \tag*{(60)}

while the many-body Green’s function is obtained from HBH_B in (54) assuming ΔVB=0\Delta V_B=0,

GI(rxA,ξ,rbB,E)=(E−Tx−Tb−hA(ξ)−VxA(rxA,ξ)−UbB(rbB))−1.(61)G_I(r_{xA},\xi,r_{bB},E)=\left(E-T_x-T_b-h_A(\xi)-V_{xA}(r_{xA},\xi)-U_{bB}(r_{bB})\right)^{-1}. \tag*{(61)}

We use the subscript II to indicate that the Green’s function above corresponds to the indirect measurement, in order to distinguish it from the direct Green’s function (12), which does not depend on rbBr_{bB}. The above expression exemplifies what can be considered the essence of the spectator approximation: the factorization of the Green’s function in a product of operators in the bb and xx-AA spaces. More explicitly, (61) can be written as

GI(E)=∫d3kb∣χb(rbB,kb)⟩⟨χb(rbB,kb)∣⊗∑ij∣Φi(ξ)⟩⟨Φj(ξ)∣×⟨Φi(ξ)∣(Ei−Eb−Tx(rxA)−VxA(rxA,ξ))−1∣Φj(ξ)⟩=∫d3kb∣χb(rbB,kb)⟩⟨χb(rbB,kb)∣⊗∑ij∣Φi(ξ)⟩((Ei−Eb−Tx(rxA))δij−Vij(rxA))−1⟨Φj(ξ)∣(62)\begin{aligned} G_I(E)={}&\int d^3k_b\lvert\chi_b(r_{bB},k_b)\rangle\langle\chi_b(r_{bB},k_b)\rvert\otimes\sum_{ij}\lvert\Phi_i(\xi)\rangle\langle\Phi_j(\xi)\rvert\\ &\times\langle\Phi_i(\xi)\rvert\left(E_i-E_b-T_x(r_{xA})-V_{xA}(r_{xA},\xi)\right)^{-1}\lvert\Phi_j(\xi)\rangle\\ ={}&\int d^3k_b\lvert\chi_b(r_{bB},k_b)\rangle\langle\chi_b(r_{bB},k_b)\rvert\\ &\otimes\sum_{ij}\lvert\Phi_i(\xi)\rangle\left((E_i-E_b-T_x(r_{xA}))\delta_{ij}-V_{ij}(r_{xA})\right)^{-1}\langle\Phi_j(\xi)\rvert \tag*{(62)} \end{aligned}

where Ei=E−ϵiAE_i=E-\epsilon_i^A, and we have used the fact that the Green’s function is diagonal in the basis of bb states χb\chi_b defined by

(Eb−Tb−UbB(rbB))χb(rbB,kb)=0;(Eb=ℏ2kb22μb).(63)(E_b-T_b-U_{bB}(r_{bB}))\chi_b(r_{bB},k_b)=0;\qquad\left(E_b=\frac{\hbar^2k_b^2}{2\mu_b}\right). \tag*{(63)}

The matrix elements VijV_{ij} of the many-body potential were defined in Eq. (38). Comparing with (12), we obtain

GI(rxA,ξ,rbB,E)=∫d3kb G(rxA,ξ,E−Eb) Pb(rbB,kb),(64)G_{\mathrm{I}}(r_{xA},\xi,r_{bB},E)=\int d^{3}k_b\,G(r_{xA},\xi,E-E_b)\,\mathcal{P}_b(r_{bB},k_b), \tag*{(64)}

where we have introduced the projection operator

Pb(kb)=∣χb(rbB,kb)⟩⟨χb(rbB,kb)∣.(65)\mathcal{P}_b(k_b)=\lvert\chi_b(r_{bB},k_b)\rangle\langle\chi_b(r_{bB},k_b)\rvert. \tag*{(65)}

Eq. (57) can now be rewritten,

Ψ=Ψ0+∫d3kb G(E−Eb)Pb(kb)[VxA(rxA,ξ)+UbA(rbA)]Ψ0=Ψ0+ΨxA,(66)\Psi=\Psi_0+\int d^3k_b\,G(E-E_b)\mathcal{P}_b(k_b)\left[V_{xA}(r_{xA},\xi)+U_{bA}(r_{bA})\right]\Psi_0=\Psi_0+\Psi_{xA}, \tag*{(66)}

where the factorized Green’s function is consistent with a process in which two systems (the xx-AA system on the one hand and the bb fragment on the other hand) scatter independently. Let us now obtain the one-body elastic scattering (defined as the process in which the nucleus AA remains in its ground state) wave function corresponding to an experimental situation in which the fragment bb has been observed with momentum kbk_b. This is obtained by projecting the many-body wavefunction (66) onto the corresponding observed state ∣Φ0χb(kb)⟩\lvert\Phi_0\chi_b(k_b)\rangle,

∫Φ0(ξ)χb(rbB,kb)Ψ(rxA,rbB,ξ) d3rbB dξ≡⟨Φ0χb(kb)∣Ψ=⟨Φ0χb(kb)∣Ψ0+⟨Φ0χb(kb)∣GI(E)[VxA(rxA,ξ)+UbA(rbA)]Ψ0=⟨χb∣ϕaF+⟨Φ0∣G(E−Eb)⟨χb∣[VxA(rxA,ξ)+UbA(rbA)]Ψ0=⟨χb∣ϕaF+G(E0−Eb)⟨χb∣[V(E−Eb)+UbA(rbA)]ϕaF,(67)\begin{aligned} \int\Phi_0(\xi)\chi_b(r_{bB},k_b)\Psi(r_{xA},r_{bB},\xi)\,d^3r_{bB}\,d\xi &\equiv\langle\Phi_0\chi_b(k_b)\rvert\Psi\\ &=\langle\Phi_0\chi_b(k_b)\rvert\Psi_0+\langle\Phi_0\chi_b(k_b)\rvert G_{\mathrm{I}}(E)\left[V_{xA}(r_{xA},\xi)+U_{bA}(r_{bA})\right]\Psi_0\\ &=\langle\chi_b\vert\phi_{aF}+\langle\Phi_0\rvert G(E-E_b)\langle\chi_b\vert\left[V_{xA}(r_{xA},\xi)+U_{bA}(r_{bA})\right]\Psi_0\\ &=\langle\chi_b\vert\phi_{aF}+G(E_0-E_b)\langle\chi_b\vert\left[V(E-E_b)+U_{bA}(r_{bA})\right]\phi_{aF}, \tag*{(67)} \end{aligned}

where we have used (52). We can now write the indirect elastic channel wavefunction for the fragment xx in terms of single-particle operators,

ψ0I(rxA)=ψHM+G(E0−Eb)[V(E−Eb)ψHM+⟨χb∣UbA(rbA)ϕaF]=ψHM(rxA)+G(rxA,E0−Eb)×[V(rxA,E−Eb)ψHM(rxA)+∫χb∗(rbB,kb)UbA(rbA)ϕa(rxb)F(raA) d3rxb],(68)\begin{aligned} \psi_0^{\mathrm{I}}(r_{xA})&=\psi^{\mathrm{HM}}+G(E_0-E_b)\left[V(E-E_b)\psi^{\mathrm{HM}}+\langle\chi_b\vert U_{bA}(r_{bA})\phi_aF\right]\\ &=\psi^{\mathrm{HM}}(r_{xA})+G(r_{xA},E_0-E_b)\\ &\quad\times\left[V(r_{xA},E-E_b)\psi^{\mathrm{HM}}(r_{xA})+\int\chi_b^*(r_{bB},k_b)U_{bA}(r_{bA})\phi_a(r_{xb})F(r_{aA})\,d^3r_{xb}\right], \tag*{(68)} \end{aligned}

or, equivalently,

ψ0I(rxA)=ψHM+G(E0−Eb)⟨χb∣(V(E−Eb)+UbA(rbA))ϕaF=ψHM(rxA)+G(rxA,E0−Eb)∫χb∗(rbB,kb)(V(rxA,E−Eb)+UbA(rbA))ϕa(rxb)F(raA) d3rxb,(69)\begin{aligned} \psi_0^{\mathrm{I}}(r_{xA})&=\psi^{\mathrm{HM}}+G(E_0-E_b)\langle\chi_b\vert\left(V(E-E_b)+U_{bA}(r_{bA})\right)\phi_aF\\ &=\psi^{\mathrm{HM}}(r_{xA})+G(r_{xA},E_0-E_b)\int\chi_b^*(r_{bB},k_b)\left(V(r_{xA},E-E_b)+U_{bA}(r_{bA})\right)\phi_a(r_{xb})F(r_{aA})\,d^3r_{xb}, \tag*{(69)} \end{aligned}

where the single-particle Green’s function is the one defined in Eq. (51), and we have introduced the Hussein-McVoy wavefunction [33],

ψHM(rxA)=∫χb∗(rbB,kb)ϕa(rxb)F(raA) d3rxb.(70)\psi^{\mathrm{HM}}(r_{xA})=\int\chi_b^*(r_{bB},k_b)\phi_a(r_{xb})F(r_{aA})\,d^3r_{xb}. \tag*{(70)}

We can now use Eq. (53) to obtain the total indirect reaction cross section of the fragment xx with the nucleus AA,

dσRI(E,Eb)dEb=2μℏ2kx⟨ϕ0∣Im⁡V(E−Eb)∣ϕ0⟩ρ(Eb),(71)\frac{d\sigma_{R}^{I}(E,E_b)}{dE_b}=\frac{2\mu}{\hbar^2 k_x}\langle\phi_0|\operatorname{Im}V(E-E_b)|\phi_0\rangle\rho(E_b), \tag*{(71)}

where ρ(Eb)\rho(E_b) is the density of states of the fragment bb (in units [ρ]=MeV−1fm−3[\rho]=\mathrm{MeV}^{-1}\mathrm{fm}^{-3}), and μ\mu is the reduced mass of the xx-AA system. The above expression is the main result of this section, and it can be used to compute the cross section for the inclusive reaction a+A→b+(x+A)a+A\to b+(x+A), where only the fragment bb is detected. This result, identical to the one obtained in [8] with a quite different method, is the essence of the Green’s Function Transfer (GFT) formalism.

In practical applications, the optical potential is either obtained from a phenomenological fit ([8, 34]), or calculated within some structure formalism (Sects. 4.1, 4.2, see also [35, 36]). In addition, a distorted wave associated with an auxiliary optical potential is used instead of the free plane wave F(raA)F(r_{aA}), with the corresponding modifications indicated in (35). Then, a partial wave expansion of the wavefunction and the potential allows for the calculation of the reaction (absorption) cross section as a function of the angular momentum of the composite xx-AA system.

The expression (71) corresponds to the calculation of the reaction cross section, i.e., of the the total flux removed from the elastic channel of the xx-AA system. As we discussed in Sect. 2.2, this flux is feeding all energy-conserving excited states of the xx-AA system. Within this context, this calculation is said to be inclusive with respect to the states of the xx-AA system, as opposed to the exclusive calculations where a specific final state of the xx-AA system is selected (see Sect. 5). However, it is often the case that only one excited state exists at a particular excitation energy, and the inclusive cross section is exhausted by the population of this single state. This is always the case for bound states (see Sect. 4.2), and can also approximately happen when a resonance is rather well isolated (see Sect. 4.1).

In the spirit of the use of indirect reactions invoked in these lectures, which is to benefit from the (indirect) cross section (71) to extract information about the xx-AA system, an interesting feature of the GFT formalism is that the cross section is explicitly computed in terms of the self energy VV which determines the structure of the xx-AA system. Therefore, a theoretical calculation of the self energy of the x−Ax-A system can be used to predict the cross section (71), and be directly compared with experimental data in order to assess the validity of the structure formalism. The full consistency of the structure calculation (embodied in the self energy/optical potential VV) with the reaction calculation (the Green’s function/propagator GG) is enforced by Eq. (51). In Sects. 4.1 and 4.2 we present examples of this procedure.

Connection with the DWBA As the energy EbE_b of the fragment bb becomes larger, the argument of the optical potential in (71) will eventually be negative, and the corresponding cross section will be associated to the population of bound states of the x+Ax+A system, i.e., below the threshold for emitting the fragment xx. At negative energies, the optical potential becomes real (see [37]), which we can express by writing

V(E)=lim⁡ϵ→0Re⁡V(E)+iϵ,for E<0.(72)\mathcal{V}(E) = \lim_{\epsilon\to0} \operatorname{Re}\mathcal{V}(E) + i\epsilon,\quad\text{for } E < 0. \tag*{(72)}

For a discrete (bound) spectrum, the spectral decomposition of the Green’s function (Lehmann representation, see e.g. [1]), is then

G(E)=lim⁡ϵ→0∑i∣ψi⟩⟨ψi∣E−Ei+iϵ.(73)G(E) = \lim_{\epsilon\to0} \sum_i \frac{|\psi_i\rangle\langle\psi_i|}{E-E_i+i\epsilon}. \tag*{(73)}

where ψi,Ei\psi_i, E_i are the eigenstates and eigenenergies of the x−Ax-A system. The coordinate representation of the above Green’s function is obtained making use of the wavefunctions of the eigenstates5

G(E,rxA,rxA′)=lim⁡ϵ→0∑iψi(rxA)ψi∗(rxA′)E−Ei+iϵ.(74)G(E,r_{xA},r'_{xA}) = \lim_{\epsilon\to0} \sum_i \frac{\psi_i(r_{xA})\psi_i^*(r'_{xA})}{E-E_i+i\epsilon}. \tag*{(74)}

Let us further note that, if the optical potential V\mathcal{V} is energy-dependent (i.e., whenever the interaction between xx and AA couples the ground state of AA with its excited states, see Sect. 2.2), the eigenfunctions in (73) are not normalized to 1, but they rather verify ([1])

⟨ψi∣ψi⟩=(1−∂V∂E)−1=Si,(75)\langle\psi_i|\psi_i\rangle= \left(1-\frac{\partial\mathcal{V}}{\partial E}\right)^{-1}=S_i, \tag*{(75)}

where SiS_i is the spectroscopic amplitude of the state ii. This is an important point: the single-particle Green’s function contains the information about the proper normalization of the single-particle states (the residues of the poles of the Green’s function). Since the nucleus is a correlated many-body system, as testified by the energy dependence of the optical potential, the single-particle states are not orthonormal, and the normalization of the wavefunctions is not trivial. Because the GFT makes use of a Green’s function consistent with the optical potential (see Eq. (51)), the resulting cross sections are properly normalized, and the corresponding absolute values can be directly compared with experimental data (see Figure 2).

Image not extracted

Figure 2. Schematic depiction of the distinctive features of the Distorted Wave Born approximation (DWBA) as compared with the Green’s Function Transfer (GFT) formalism.

Let us now write down the different terms arising in the evaluation of the cross section (71) using (69),

dσRI(E,Eb)dEb=2μℏ2kx(A1+A2+A3)ρ(Eb),(76)\frac{d\sigma_R^I(E,E_b)}{dE_b}=\frac{2\mu}{\hbar^2 k_x}(A_1+A_2+A_3)\rho(E_b), \tag*{(76)}
G(E,rxA,rxA′)f(rxA)=∫d3rxA′ G(E,rxA,rxA′)f(rxA′),G(E,r_{xA},r'_{xA})f(r_{xA})=\int d^3r'_{xA}\,G(E,r_{xA},r'_{xA})f(r'_{xA}),

at variance with the convention used in these lectures.

where

A1=⟨ψHM∣Im⁡V∣ψHM⟩=lim⁡ϵ→0ϵ⟨ψHM∣ψHM⟩=0,(77)A_1 = \langle\psi^{HM}|\operatorname{Im}V|\psi^{HM}\rangle= \lim_{\epsilon\to0} \epsilon\langle\psi^{HM}|\psi^{HM}\rangle= 0, \tag*{(77)}
A2=⟨ψxA∣Im⁡V∣ψxA⟩,(78)A_2 = \langle\psi^{xA}|\operatorname{Im}V|\psi^{xA}\rangle, \tag*{(78)}

with

ψxA=G(E0−Eb)⟨χb∣(V(E−Eb)+UbA(rbA))ϕaF,(79)\psi^{xA} = G(E_0-E_b)\langle\chi_b|(\mathcal{V}(E-E_b)+U_{bA}(r_{bA}))\phi_{aF}, \tag*{(79)}

and the cross term is

A3=2Re⁡⟨ψxA∣Im⁡V∣ψHM⟩.(80)A_3 = 2\operatorname{Re}\langle\psi^{xA}|\operatorname{Im}V|\psi^{HM}\rangle. \tag*{(80)}

We now express the term A2A_2 as

A2=lim⁡ϵ→0[G(E0−Eb)⟨χb∣(V(E−Eb)+UbA(rbA))ϕaF]†×G(E0−Eb)⟨χb∣(V(E−Eb)+UbA(rbA))ϕaF=lim⁡ϵ→0∑iϵ(E0−Eb−Ei)2+ϵ2∣Ti∣2=∑i∣Ti∣2δ(E0−Eb−Ei)(81)\begin{aligned} A_2 &= \lim_{\epsilon\to0} \left[G(E_0-E_b)\langle\chi_b|(\mathcal{V}(E-E_b)+U_{bA}(r_{bA}))\phi_{aF}\right]^\dagger\\ &\quad\times G(E_0-E_b)\langle\chi_b|(\mathcal{V}(E-E_b)+U_{bA}(r_{bA}))\phi_{aF} \\ &= \lim_{\epsilon\to0}\sum_i \frac{\epsilon}{(E_0-E_b-E_i)^2+\epsilon^2}|T_i|^2 = \sum_i |T_i|^2\delta(E_0-E_b-E_i) \tag*{(81)} \end{aligned}

where we have used (73), and we have defined the transfer TT-matrix to the bound state ψi\psi_i,

Ti=⟨ψiχb∣V(E−Eb)+UbA(rbA)∣ϕaF⟩.(82)T_i = \langle\psi_i\chi_b|\mathcal{V}(E-E_b)+U_{bA}(r_{bA})|\phi_{aF}\rangle. \tag*{(82)}

The cross term A3A_3 also vanishes, since it is proportional to the energy-dependent term

Re⁡(lim⁡ϵ→0∑iϵ(E0−Eb−Ei)+iϵ)=0.(83)\operatorname{Re}\left(\lim_{\epsilon\to0}\sum_i\frac{\epsilon}{(E_0-E_b-E_i)+i\epsilon}\right)=0. \tag*{(83)}

The cross section is then

dσRI(E,Eb)dEb=2μℏ2kakb∑i∣Ti∣2δ(E0−Eb−Ei)ρ(Eb),(84)\frac{d\sigma_R^I(E,E_b)}{dE_b}=\frac{2\mu}{\hbar^2 k_a k_b}\sum_i|T_i|^2\delta(E_0-E_b-E_i)\rho(E_b), \tag*{(84)}

and has the same structure as the transfer cross section to bound states computed in DWBA (see Sect. 2.1, see also [8]).

In practice, when making a DWBA calculation the wavefunction ψi\psi_i in Eq. (82) is usually taken to be a solution normalized to 1 of the Schrödinger equation for the bound state ii in the potential V(E−Eb)\mathcal{V}(E-E_b), which is approximated by a simple central potential, often a Woods-Saxon with a reasonable radius and a depth fitted to reproduce the binding energy of the state ii. The explicit connection of the resulting cross section with the correlated many-body system, enforced by Eq. (51), is then lost. In particular, the value of the single-particle strength has to be extracted by comparing the calculated cross section with the experimental one, (see Figure 2) a procedure which is rendered ambiguous by the arbitrariness in the choice of the wavefunction ψi\psi_i to be used in Eq. (82).

Let us stress that, although common practice, the procedure that has been just described is not an inherent part of the DWBA. It is always possible to use in the TT-matrix a wavefunction consistent (both in terms of normalization and of spatial dependence) with a microscopic self energy V\mathcal{V} calculated within some structure formalism of choice, to be also used for the calculation of the TT-matrix (82). The resulting cross section would then be equivalent to the one computed with the GFT formalism.

Let us finally comment on the population of the continuum. The spectral decomposition of the Green’s function associated with the continuum part of the spectrum of the x−Ax-A system is

G(E,rxA,rxA′)=lim⁡ϵ→0∫dEc ρ(Ec)χ(rxA,Ec)χ∗(rxA′,Ec)E−Ec+iϵ(85)G(E,\mathbf{r}_{xA},\mathbf{r}'_{xA}) = \lim_{\epsilon\to0}\int dE_c\,\rho(E_c)\frac{\chi(\mathbf{r}_{xA},E_c)\chi^{*}(\mathbf{r}'_{xA},E_c)}{E-E_c+i\epsilon} \tag*{(85)}

where χ\chi are continuum wavefunctions, and ρ(Ec)\rho(E_c) is the density of states of the continuum. This expression should substitute Eq. (74) in the expression for the Green’s function to be used in Eq. (81). The resulting expression would then not have the form of a sum over DWBA TT-matrices corresponding to the population of a discrete set of final bound states, but would rather involve an integral over the continuum states. The essentially discrete nature of the DWBA renders somehow unnatural the description of the population of continuum states, although several methods exist to discretize the continuum and describe the associated cross sections [30, 31, 32] (see Figure 2). Let us also note that the proper normalization of the continuum wavefunctions is not trivial, while the GFT formalism takes naturally care of that. In addition, since the optical potential V(E>0)\mathcal{V}(E>0) has a finite imaginary part, the terms A1,A3A_1,A_3 do not vanish, and the exact connection with the DWBA is lost. Finally, let us point out that the DWBA cannot properly account for the coherent contribution to the cross section of overlapping resonances of the same spin and parity, while this possibility is properly embodied in the Green’s function used in the GFT.

Applications

As examples of the GFT formalism, we will discuss two applications, associated with very different systems. In both cases, the implementation of the GFT implies the calculation of the self energy V\mathcal{V} within some structure formalism of choice, and the obtention of the associated Green’s function by inverting the Hamiltonian matrix according to Eq. (51). The consistency between structure (V\mathcal{V}) and reactions (GG), is thus enforced. From a technical point of view, this implies the development of methods and tools that allow us to deal with the non quite standard situation of having a non-local potential V\mathcal{V}, resulting from most self energy calculations. These methods have been developed, allowing for the computation of the solution of the Schrödinger equation for non local potentials, and the obtention of Green’s functions from matrix inversion, for positive and negative energies, as well as for charged and neutral particles (see, e.g., [38]).

The first example will describe the population of the low-lying spectrum of the unbound system 10Li{}^{10}\mathrm{Li} in the 9Li(d,p){}^{9}\mathrm{Li}(d,p) reaction. This process serves as a good illustration of the description of the population of the continuum in indirect reactions within the GFT formalism.

The second example will be the population of the 41Ca^{41}\mathrm{Ca} nucleus in the 40Ca(d,p)^{40}\mathrm{Ca}(d,p) reaction, which is a good example of the use of the GFT formalism to describe the population of bound states in a stable nucleus.

9Li(d,p)10Li^{9}\mathrm{Li}(d,p)^{10}\mathrm{Li} with NFT

The unbound nucleus 10Li^{10}\mathrm{Li} sits at the crossroads of shell evolution and halo structure. In the N=7N=7 chain, the normal ordering places the 1p1/21p_{1/2} below the 2s1/22s_{1/2} orbit, as seen in 12B^{12}\mathrm{B} and 13C^{13}\mathrm{C}; yet the paradigm of parity inversion in 11Be^{11}\mathrm{Be} — where the 1/2+1/2^{+} bound state lies below the 1/2−1/2^{-} — suggests that the isotone 10Li^{10}\mathrm{Li} should display a near-threshold virtual 1/2+1/2^{+} configuration in competition with a low-lying 1/2−1/2^{-} resonance. When coupled to the odd 3/2−3/2^{-} proton in 9Li^{9}\mathrm{Li}, these states give rise to a 1+1^{+} and 2+2^{+} doublet associated with the 1/2−1/2^{-} resonance and a 1−1^{-}, 2−2^{-} doublet for the 1/2+1/2^{+} state. Establishing this has direct consequences for the large ss–wave component of the 11Li^{11}\mathrm{Li} halo and for a unified, many–body description of this region. The reaction 9Li(d,p)10Li^{9}\mathrm{Li}(d,p)^{10}\mathrm{Li} is a specific probe of single–neutron motion built on the 9Li^{9}\mathrm{Li} core: by measuring proton spectra and angular distributions, one gains access to the energy dependence and partial–wave content of the n+9Lin+^{9}\mathrm{Li} continuum. Within this context, and in the spirit of these lectures, the 9Li(d,p)10Li^{9}\mathrm{Li}(d,p)^{10}\mathrm{Li} reaction can be viewed as an indirect measurement of the n+9Lin+^{9}\mathrm{Li} system, the direct measurement of neutron scattering on 9Li^{9}\mathrm{Li} being experimentally essentially unfeasible due to the short life of 9Li^{9}\mathrm{Li} (178 ms) and of the neutron (614 s).

In the NFT approach, the structure calculations in [36] dress the neutron–core motion with strong particle–vibration coupling to the quadrupole phonon of the 9Li^{9}\mathrm{Li} core (β2=0.72\beta_{2}=0.72, ℏω2≃3.4\hbar\omega_{2}\simeq3.4 MeV), including relevant two–phonon (anharmonic) contributions in the d5/2d_{5/2} channel. The output is an energy–dependent, nonlocal self–energy Σ(E)\Sigma(E) which shifts, fragments, and reshapes the single–particle configurations. The resulting dressed continuum states j~π\tilde{j}^{\pi} exhibit renormalized energies, widths, spectroscopic amplitudes, and — crucially for transfer — the proper radial form factors. This same Σ(E)\Sigma(E) is used to provide the single–particle Green’s function G(E)=(E−T−Σ)−1G(E)=(E-T-\Sigma)^{-1} used in the GFT approach (§ 3), ensuring that structure and reaction are consistent and treated on the same footing, allowing predictions of absolute transfer cross sections without ad hoc normalizations.

An s–wave virtual state is encoded in a large, negative scattering length aa; equivalently the s–wave phase shift behaves as

δ0(E)→E→0−ka,E=ℏ2k22μ,(86)\delta_{0}(E) \xrightarrow[E\to0]{} -ka,\qquad E=\frac{\hbar^{2}k^{2}}{2\mu}, \tag*{(86)}

implying an energy scale ϵ∼ℏ2/(2μa2)\epsilon\sim\hbar^{2}/(2\mu a^{2}) that controls the near–threshold enhancement. For a narrow resonance at ErE_{r} one has the familiar relation

Γjπ(Er)=2(dδjπdE∣Er)−1,(87)\Gamma_{j^{\pi}}(E_{r})=2\left(\left.\frac{d\delta_{j^{\pi}}}{dE}\right|_{E_{r}}\right)^{-1}, \tag*{(87)}

which we will use below to interpret the p– and d–wave structures.

Figures 3 (a,b) show the calculated phase shifts for the negative– and positive–parity channels, compared to the bare mean–field (“unp.”) behaviour. The dressing reverses the slope of the 2−2^{-} and 1−1^{-} phase shifts near threshold, signalling the formation of a low–lying 1/2−1/2^{-} resonance around 0.45–0.50 MeV with a width of about 0.3–0.35 MeV. In the positive–parity sector, the s1/2s_{1/2} strength is pushed into the continuum and acquires a virtual character, with a scattering length a≃−8a\simeq-8 fm (corresponding scale ϵ≈0.3\epsilon\approx0.3 MeV), while the d5/2d_{5/2} channel develops a broad 5/2+5/2^{+} resonance at a few MeV due to strong coupling to the quadrupole phonon and two–phonon states. Panels (c,d) present the corresponding spectral functions, i.e. the differences between the renormalized and bare level densities, ρˉjπ(ω)−ρˉjπunp(ω)∝dδjπ/dω\bar{\rho}_{j^{\pi}}(\omega)-\bar{\rho}^{\mathrm{unp}}_{j^{\pi}}(\omega)\propto d\delta_{j^{\pi}}/d\omega, which make the redistribution of strength transparent.

As stated above, these predictions cannot be directly addressed with a neutron scattering experiment. Instead, one can resort to the indirect (d,p)(d,p) measurement in inverse kinematics, making use of a radioactive 9Li^{9}\mathrm{Li} beam on a deuterated target. The calculated self energy Σ(E)\Sigma(E) at the basis of these results is implemented in the expressions (69) and (71) –where it takes the place of the optical potential V(E)V(E)– to compute the absolute single– and double–differential cross sections and fold it over the experimental resolutions. The results are shown in Figure 4: when the yield is integrated over the forward angular window 5.5∘≤θc.m.≤16.5∘5.5^{\circ}\leq\theta_{\mathrm{c.m.}}\leq16.5^{\circ} — the acceptance of the high–statistics measurement reanalysed in [36] — the spectrum is dominated by the p–wave resonance and exhibits an apparent absence of s–wave strength near threshold. However, if one inspects the angular distributions for θc.m.≳40∘\theta_{\mathrm{c.m.}}\gtrsim40^{\circ}, or integrates over a backward window (e.g. 50∘50^{\circ}–180∘180^{\circ}), the virtual s–wave contribution emerges clearly, with a magnitude comparable to the p–wave component close to threshold. Thus the seemingly contradictory inferences regarding parity inversion are reconciled once momentum matching and acceptance effects are accounted for within a consistent structure–reaction framework.

Image not extracted

Figure 4. Absolute (d,p)(d,p) predictions from [36]: (a) strength function folded over the forward window 5.5∘5.5^\circ–16.5∘16.5^\circ (dominated by pp–wave); (b) corresponding angular distributions in a representative energy interval; (c) strength integrated over backward angles 50∘50^\circ–180∘180^\circ, where the ss–wave becomes prominent; (d) low–energy angular distribution (E≲0.2E\lesssim0.2 MeV) emphasizing the virtual ss–wave at θc.m.≳40∘\theta_{\mathrm{c.m.}}\gtrsim40^\circ.

Image not extracted

Figure 3. Phase shifts (top) and spectral functions (bottom) for 10Li^{10}\mathrm{Li} from the renormalized NFT calculation of [36], compared with the unperturbed (“unp.”) case. (a) Negative parity: the 2−2^{-} and 1−1^{-} channels develop a low-lying 1/2−1/2^{-} resonance at ∼0.5\sim0.5 MeV. (b) Positive parity: the ss-wave becomes virtual (large negative aa) and a broad 5/2+5/2^{+} resonance appears at a few MeV. (c,d) Differences of spectral functions highlight the redistribution of strength due to particle–vibration coupling.

The dressed 1/2−1/2^{-} state in 10Li^{10}\mathrm{Li} is predominantly s1/2s_{1/2} with a small admixture of (d5/2⊗2+)(d_{5/2}\otimes2^{+}); the 1/2−1/2^{-} resonance contains (p1/2⊗2+)(p_{1/2}\otimes2^{+}) components; and the 5/2+5/2^{+} resonance around 3–4.5 MeV is strongly many–body in nature due to two–phonon couplings. This same mechanism unifies the parity inversion in 11Be^{11}\mathrm{Be}, the large ss–wave content of the 11Li^{11}\mathrm{Li} halo, and the normal ordering in 12B/13C^{12}\mathrm{B}/^{13}\mathrm{C}. From the practical viewpoint of transfer reactions, the decisive point is that the same self–energy that fixes the spectroscopy also fixes the radial form factors entering the cross sections, removing the traditional ambiguity of separate DWBA fits.

We retain Figure 5 as a compact account of the story: the near–threshold virtual 1/2+1/2^{+} and the low–lying resonant 1/2−1/2^{-} shaped by coupling to the strong 2+2^{+} core vibration, and the momentum–matching selectivity that hides (forward angles) or reveals (backward angles) the ss–wave contribution in (d,p)(d,p).

Image not extracted

Figure 5. Results obtained from the analysis of the 9Li(d,p)10Li{}^{9}\mathrm{Li}(d,p){}^{10}\mathrm{Li} reaction within the renormalized NFT+GFT framework: (i) parity inversion encoded in a virtual 1/2−1/2^{-} (large negative scattering length) alongside a low-lying 1/2−1/2^{-} resonance; (ii) dressing by the strong 2+2^{+} vibration of the 9Li{}^{9}\mathrm{Li} core and associated two–phonon effects that redistribute d5/2d_{5/2} strength; (iii) momentum matching in (d,p)(d,p) that suppresses the ss–wave at forward angles yet enhances it at θc.m.≳40∘\theta_{\mathrm{c.m.}} \gtrsim40^\circ.

40,48Ca(d,p)^{40,48}\mathrm{Ca}(d,p) with ab initio CC+GFT

A central goal of modern nuclear theory is to connect nuclear reactions directly to the underlying nuclear interactions, without introducing ad hoc parameters. For decades, transfer reactions such as (d,p)(d,p) have been analyzed with the distorted–wave Born approximation (DWBA), which relies on phenomenological spectroscopic factors and adjustable optical potentials. By contrast, the recent work of Rotureau et al. [35] demonstrates how to embed ab initio structure information from the Coupled Cluster (CC) method into the Green’s Function Transfer (GFT) framework. This unifies the description of structure and reactions: the overlaps, amplitudes and energies that govern the transfer are calculated consistently from the many–body Hamiltonian, here taken as NNLOsat_{\mathrm{sat}} chiral EFT including two– and three–nucleon forces.

On the structure side, the CC approach provides ground–state energies and particle–addition amplitudes (PA–EOM) for closed– and near–closed–shell systems. These overlap functions replace the spectroscopic factors of DWBA and directly enter the kernel of the GFT equations. On the reaction side, the GFT framework (see Sec. 3) expresses the (d,p)(d,p) cross section in terms of the single–particle Green’s function of the neutron coupled to the target plus optical potentials for deuteron and proton scattering. The crucial point is that the same Hamiltonian—and in particular the same three–body forces—determine both the overlaps and the energies, so that the calculation has predictive power once the interaction is fixed. The only phenomenological input is the choice of nucleon–target optical potentials, taken from global systematics when data are lacking.

The benchmark cases studied in [35] are the ground–state to ground–state transfers 40Ca(d,p)41Ca{}^{40}\mathrm{Ca}(d,p){}^{41}\mathrm{Ca} and 48Ca(d,p)49Ca{}^{48}\mathrm{Ca}(d,p){}^{49}\mathrm{Ca} at Ed=10E_d = 10 MeV. For these stable targets, elastic scattering data exist and were used to constrain the optical potentials. Figure 6 displays the resulting angular distributions. The curves show the CC+GFT predictions for increasing model–space truncations Nmax⁡N_{\max}. The convergence pattern is somewhat non–monotonic, but the largest space (N2=14N_2=14, N3=16N_3=16) yields angular distributions in very good agreement with the experimental data, without adjustable scaling. The calculation slightly underbinds the final ground states (∼0.5\sim0.5–0.60.6 MeV), an effect corrected by shifting the energies to the experimental values while keeping the overlaps unchanged. This procedure corresponds to adjusting the proton momentum kpk_p in the transfer kernel, not to refitting any spectroscopic parameters.

Image not extracted

Figure 6. Angular differential cross sections for 40Ca(d,p)41Cags^{40}\mathrm{Ca}(d,p)^{41}\mathrm{Ca}_{\mathrm{gs}} and 48Ca(d,p)49Cags^{48}\mathrm{Ca}(d,p)^{49}\mathrm{Ca}_{\mathrm{gs}} at Ed=10E_d=10 MeV from the ab initio CC+GFT framework of [35]. Curves correspond to different model–space truncations Nmax⁡N_{\max}; experimental data are shown as filled circles. We also indicate the neutron separation energies resulting from the corresponding calculations. The calculation associated with the larger largest space (N2=14N_2=14, N3=16N_3=16, where N2N_2, N3N_3 are the Nmax⁡N_{\max} associated with 2- and 3-body forces, respectively) reproduces the angular distributions with high accuracy. When the neutron separation energy is adjusted to the experimental values (Sn=−8.363S_n=-8.363 MeV for 41^{41}Ca, Sn=−5.146S_n=-5.146 MeV for 49^{49}Ca) the absolute value of the cross section also compares very well with the experimental measurement. This last point illustrates how the GFT provides a natural normalization of the absolute value of the cross section without the need for spectroscopic factors (see Sect. 3.2 and Figure 2).

Once benchmarked against 40,48Ca{}^{40,48}\mathrm{Ca}, the method was applied to the exotic isotopes 52Ca{}^{52}\mathrm{Ca} and 55Ca{}^{55}\mathrm{Ca}, where no (d,p)(d,p) data exist. In these cases, optical potentials were taken from global parametrizations, while the overlaps and energies were still provided by CC. The calculations predict absolute cross sections and angular distributions, offering guidance for future experiments. The emergence of shell evolution with neutron number

in the Ca chain, and the role of three–nucleon forces in shifting single–particle energies, are directly visible in the predicted transfer observables. The Ca case study illustrates how ab initio nuclear structure and reaction theories can be merged. By supplying overlaps and energies from CC into the GFT formalism, one obtains parameter–free predictions of (d,p)(d,p) angular distributions that agree with data in stable systems and extend to exotic nuclei where no measurements exist. This represents a significant step toward a unified, microscopic description of nuclear reactions based on chiral EFT interactions.

Exclusive inelastic scattering

Eq. (67) can be generalized to the calculation of the cross section for the population of a specific final state i≠0i \ne0 in the residual nucleus. This type of experiments are known as exclusive measurements, as opposed to the inclusive ones discussed in Sect. 3, where the final state is not identified and the cross section described is the total reaction one, i.e., summed for all energy-accessible final states. The corresponding cross section is obtained by projecting the wavefunction (66) onto the specific final state ii,

ψiI(rxA)=∫Φi(ξ)χb(rbB,kb)Ψ(rxA,rbB,ξ) d3rbB dξ≡⟨Φiχb(kb)∣Ψ=⟨Φi∣G(E−Eb)⟨χb∣[VxA(rxA,ξ)+UbA(rbA)]Ψ0=∫Φi(ξ)G(rxA,ξ,E−Eb)χb(rbB)[VxA(rxA,ξ)+UbA(rbA)]Ψ0(rxA,rbB,ξ) d3rbB dξ.(88)\begin{aligned} \psi_i^{I}(\mathbf{r}_{xA}) &= \int\Phi_i(\boldsymbol{\xi})\chi_b(\mathbf{r}_{bB},\mathbf{k}_b)\Psi(\mathbf{r}_{xA},\mathbf{r}_{bB},\boldsymbol{\xi})\,d^3r_{bB}\,d\boldsymbol{\xi} \equiv\langle\Phi_i\chi_b(\mathbf{k}_b)|\Psi\\ &= \langle\Phi_i|G(E-E_b)\langle\chi_b|[V_{xA}(\mathbf{r}_{xA},\boldsymbol{\xi})+U_{bA}(\mathbf{r}_{bA})]\Psi_0 \\ &= \int\Phi_i(\boldsymbol{\xi})G(\mathbf{r}_{xA},\boldsymbol{\xi},E-E_b)\chi_b(\mathbf{r}_{bB})[V_{xA}(\mathbf{r}_{xA},\boldsymbol{\xi})+U_{bA}(\mathbf{r}_{bA})]\Psi_0(\mathbf{r}_{xA},\mathbf{r}_{bB},\boldsymbol{\xi})\,d^3r_{bB}\,d\boldsymbol{\xi}. \tag*{(88)} \end{aligned}

Let us write the asymptotic wavefunction,

ψiI(rxA→∞)=O(ki,rxA)⟨ψiΦi∣VxA(rxA,ξ)∣Φ0ψHM⟩=O(ki,rxA)Ti0I,(89)\psi_i^{I}(\mathbf{r}_{xA}\to\infty)=O(\mathbf{k}_i,\mathbf{r}_{xA})\langle\psi_i\Phi_i|V_{xA}(\mathbf{r}_{xA},\boldsymbol{\xi})|\Phi_0\psi^{HM}\rangle=O(\mathbf{k}_i,\mathbf{r}_{xA})T_{i0}^{I}, \tag*{(89)}

where we have used (21) and the fact that

⟨ψiχbΦi∣UbA∣Φ0ϕaF⟩=0,(90)\langle\psi_i\chi_b\Phi_i|U_{bA}|\Phi_0\phi_a F\rangle=0, \tag*{(90)}

and we have introduced the indirect TT-matrix,

Ti0I=⟨ψiΦi∣VxA(rxA,ξ)∣Φ0ψHM⟩,(91)T_{i0}^{I}=\langle\psi_i\Phi_i|V_{xA}(\mathbf{r}_{xA},\boldsymbol{\xi})|\Phi_0\psi^{HM}\rangle, \tag*{(91)}

which differs from the direct TT-matrix (20) by the presence of ψHM(rxA)\psi^{HM}(\mathbf{r}_{xA}) instead of the free wave F(rxA;kxA)F(\mathbf{r}_{xA};\mathbf{k}_{xA}) in the ket.

Let us introduce here a word of caution. Since an expression similar to Eq. (52) cannot be established here, we have, in general, that

⟨ψiΦi∣VxA(rxA,ξ)∣Φ0ψHM⟩≠⟨ψi∣V∣ψHM⟩.(92)\langle\psi_i\Phi_i|V_{xA}(\mathbf{r}_{xA},\boldsymbol{\xi})|\Phi_0\psi^{HM}\rangle\ne\langle\psi_i|V|\psi^{HM}\rangle. \tag*{(92)}

This is a consequence of the fact that the optical reduction process described in Sect. 2.2 consists essentially in restricting the many-body operator VxAV_{xA} to the one-dimensional sector of the Hilbert space spanned by the ground state Φ0\Phi_0 of the nucleus AA. Since the excited state Φi\Phi_i is outside this sector, the operator VxAV_{xA} cannot be reduced in Eq. (91) to a single-particle operator VV. Although a generalization of the concept of optical potential by defining it in the space of all open reaction channels can be made (see, e.g., [23, 26]), we will not touch upon this subject here. Within this context, we do not propose in this Section a calculable method to obtain (91). Instead, the main result of this Section is the explicit connection between the direct and indirect TT-matrices expressed in Eq. (95), and, more particularly, the relationship between the corresponding RR-matrix parameters shown in Eq. (99).

In order to establish a more explicit connection between the indirect and direct TT-matrices, we will expand the Hussein-McVoy function in terms of a superposition of free waves,

ψHM(rxA)=∫g(k)F(rxA,k) dk,(93)\psi^{HM}(\mathbf{r}_{xA}) = \int g(\mathbf{k}) F(\mathbf{r}_{xA}, \mathbf{k}) \,\mathrm{d}k, \tag*{(93)}

where we have introduced a “broadening factor” gg,

g(k)=∫ψHM(rxA)F∗(rxA,k) drxA.(94)g(\mathbf{k}) = \int\psi^{HM}(\mathbf{r}_{xA}) F^{*}(\mathbf{r}_{xA}, \mathbf{k}) \,\mathrm{d}\mathbf{r}_{xA}. \tag*{(94)}

If FF is a plane wave (i.e., the fragment xx is a neutron, like in [8]), the broadening factor gg is just the Fourier transform of the Hussein-McVoy function. In any case, it can be calculated if the ground state wavefunction ϕa(rxb)\phi_a(\mathbf{r}_{xb}) of the nucleus aa is known which is, at any rate, a necessary prerequisite for the applicability of the indirect method. We can now write the indirect TT-matrix in terms of the direct one,

Ti0I=∫g(k)Ti0(k) dk,(95)T^{I}_{i0} = \int g(\mathbf{k}) T_{i0}(\mathbf{k}) \,\mathrm{d}k, \tag*{(95)}

where

Ti0(k)=⟨ψiΦi∣VxA(rxA,ξ)∣Φ0F(rxA,k)⟩.(96)T_{i0}(\mathbf{k}) = \langle\psi_i \Phi_i \lvert V_{xA}(\mathbf{r}_{xA}, \xi) \rvert\Phi_0 F(\mathbf{r}_{xA}, \mathbf{k}) \rangle. \tag*{(96)}

We now take advantage of the RR-matrix parametrization (24),

Ti0(k)=Pi(Ek)P0(Ek)∑pqγip[A−1(Ek)]pqγ0q,(97)T_{i0}(\mathbf{k}) = \sqrt{P_i(E_k)P_0(E_k)} \sum_{pq} \gamma_{ip} [A^{-1}(E_k)]_{pq} \gamma_{0q}, \tag*{(97)}

with

Ek=ℏ2k22μ.(98)E_k = \frac{\hbar^2 k^2}{2\mu}. \tag*{(98)}

We can then write the explicit expression of the indirect TT-matrix in terms of the direct TT-matrix parameters,

Ti0I=∫Pi(Ek)P0(Ek)∑pqγip[A−1(Ek)]pqγ0qg(k) dk.(99)T^{I}_{i0} = \int\sqrt{P_i(E_k)P_0(E_k)} \sum_{pq} \gamma_{ip} [A^{-1}(E_k)]_{pq} \gamma_{0q} g(k) \,\mathrm{d}k. \tag*{(99)}

Discussion

The expression (99) shows how the partial widths γip\gamma_{ip}, γ0q\gamma_{0q}, as well as the energies EpE_p, can be fitted from the indirect experimental cross section and used to predict the direct one.

While the GFT expression (71) can be used to calculate the reaction cross section, (99) suggests a fitting procedure to the observed indirect cross section, proportional to ∣Ti0I∣2|T_{i0}^{I}|^2. This fitting procedure is so similar in spirit to the RR-matrix fit, that standard existing RR-matrix codes (like, e.g., AZURE) might possibly be used without essential modification. The difference between the direct and indirect RR-matrix parametrizations resides in the broadening factor gg,

g(k)=∫ψHM(rxA)F∗(rxA,k) drxA=∫χb∗(rbB,kb)ϕa(rxb)F(raA,kaA)F∗(rxA,k) drxA drxb.(100)g(\mathbf{k}) = \int\psi^{HM}(\mathbf{r}_{xA}) F^*(\mathbf{r}_{xA},\mathbf{k})\,d\mathbf{r}_{xA} = \int\chi_b^*(\mathbf{r}_{bB},\mathbf{k}_b)\phi_a(\mathbf{r}_{xb})F(\mathbf{r}_{aA},\mathbf{k}_{aA})F^*(\mathbf{r}_{xA},\mathbf{k})\,d\mathbf{r}_{xA}\,d\mathbf{r}_{xb}. \tag*{(100)}

The form of the above integral suggests that g(k)g(\mathbf{k}) is peaked around k=kaA−kb\mathbf{k} = \mathbf{k}_{aA} - \mathbf{k}_b, and has a width of the order of the momentum spread of the Fourier transform of ϕa\phi_a. Qualitatively,

g(k)∼ϕ~a(kaA−kb),(101)g(\mathbf{k}) \sim\widetilde{\phi}_a(\mathbf{k}_{aA} - \mathbf{k}_b), \tag*{(101)}

where ϕ~a\widetilde{\phi}_a is the Fourier transform of ϕa\phi_a. If the shift and penetrability factors vary slowly in such an interval, we might be able to approximate

Ti0I≈Pi(Ek0)P0(Ek0)∑pqγip[A−1(Ek0)]pqγ0q∫g(k) dk=NPi(Ek0)P0(Ek0)∑pqγip[A−1(Ek0)]pqγ0q,(102)T_{i0}^{I} \approx\sqrt{P_i(E_{k_0})P_0(E_{k_0})}\sum_{pq}\gamma_{ip}\left[A^{-1}(E_{k_0})\right]_{pq}\gamma_{0q}\int g(\mathbf{k})\,d\mathbf{k} = N\sqrt{P_i(E_{k_0})P_0(E_{k_0})}\sum_{pq}\gamma_{ip}\left[A^{-1}(E_{k_0})\right]_{pq}\gamma_{0q}, \tag*{(102)}

where Ek0E_{k_0} is the energy at the peak of the momentum distribution, and the constant NN is

N=∫g(k) dk.(103)N = \int g(\mathbf{k})\,d\mathbf{k}. \tag*{(103)}

Eq. (102) has the form suggested by Barker [39] on what seem to be purely heuristic grounds, where the role of what he calls the “feeding factor” GabqG_{abq} is played here by

Gabq=γ0qN.(104)G_{abq} = \gamma_{0q}N. \tag*{(104)}

We haven’t been able to find in [39] or elsewhere any expression or derivation of this feeding factor GabqG_{abq}. This approximation ignores completely the energy-dependence of the TT-matrix over the distribution defined by the broadening factor gg. It should also be pointed out that the modification of the entrance channel introduced by the broadening factor is also angular momentum dependent, and can be calculated from the multipolar expansion of g(k)g(\mathbf{k}). In any case, the validity of these approximations can be checked by directly computing g(k)g(\mathbf{k}), and using the exact expression (99).

A slightly better approximation could be to account for the additional energy width ΔE\Delta E introduced by the broadening phenomenon by adding a corresponding imaginary part in the denominator,

Ti0I≈NPi(E~)P0(E~)∑pqγip[A−1(E~)]pqγ0q,(105)T_{i0}^{I} \approx N\sqrt{P_i(\widetilde{E})P_0(\widetilde{E})}\sum_{pq}\gamma_{ip}\left[A^{-1}(\widetilde{E})\right]_{pq}\gamma_{0q}, \tag*{(105)}

with E~=Ek0+iΔE\widetilde{E}=E_{k0}+i\Delta E. This illustrates the fact that the factor gg broadens the resonances observed with the indirect method with respect to their “direct values”.

Beyond the spectator approximation

Within the spectator approximation, the dynamical role of the cluster bb is essentially only to “carry” the cluster xx inside the projectile aa, without altering the way in which xx and AA interact with each other. Within this context, the un-scattered state in the x−Ax-A collision doesn’t correspond, like in the direct measurement of the x−Ax-A scattering process, to a state of well defined energy described by a free wave FF, but is instead modified by the intrinsic motion of xx in the ground state of the projectile aa, as testified by the broadening factor gg (94). In other words, the role of the un-scattered wavefunction is now played by the Hussein-Mc Voy term (70), as it becomes apparent by comparing Eqs. (50) and (68). Within this approximation, the explicit coupling of bb with excited states of AA or B≡A+xB\equiv A+x is completely ignored. In order to gain some insight concerning the validity of this approximation, we need to go beyond the spectator approximation.

Let us start by writing the exact version of Eqs. (56) and (57),

(E−Ta−ha(rxb)−hA(ξ))Ψ=(VxA(rxA,ξ)+UbA(rbA)+ΔVA)Ψ,(106)(E-T_a-h_a(r_{xb})-h_A(\xi))\Psi=(V_{xA}(r_{xA},\xi)+U_{bA}(r_{bA})+\Delta V_A)\Psi, \tag*{(106)}

and

Ψ=Ψ0+G(E)[VxA(rxA,ξ)+UbA(rbA)+ΔVA]Ψ0,(107)\Psi=\Psi_0+G(E)\left[V_{xA}(r_{xA},\xi)+U_{bA}(r_{bA})+\Delta V_A\right]\Psi_0, \tag*{(107)}

Where ΔVA,ΔVB\Delta V_A,\Delta V_B are defined in Eq. (55), and the exact Green’s function G(E)\mathcal{G}(E) can be expressed in terms of the spectator approximation Green’s function,

G(E)=GI(E)+GI(E)ΔVBG(E)=GI(E)+GI(E)ΔVBGI(E)+(terms of higher order in ΔVB),(108)G(E)=G_I(E)+G_I(E)\Delta V_B G(E)=G_I(E)+G_I(E)\Delta V_B G_I(E) +(\text{terms of higher order in }\Delta V_B), \tag*{(108)}

From the two expressions above we obtain the exact wavefunction,

Ψ=Ψ0+ΨxA+GI(E)ΔVAΨ0+GI(E)ΔVBΨxA+(terms of higher order in ΔVA,ΔVB),(109)\Psi=\Psi_0+\Psi_{xA}+G_I(E)\Delta V_A\Psi_0+G_I(E)\Delta V_B\Psi_{xA}+(\text{terms of higher order in }\Delta V_A,\Delta V_B), \tag*{(109)}

where the two first terms in the right hand side correspond to the spectator approximation (66), including the scattered wavefunction ΨxA\Psi_{xA}, while the 3rd and 4th terms provide the corrections to first order in ΔVA\Delta V_A and ΔVB\Delta V_B. Let us further assume that

Ψ0≫ΨxA,(110)\Psi_0\gg\Psi_{xA}, \tag*{(110)}

i.e., that the departure away from the “elastic” channel represented by the scattered wave ΨxA\Psi_{xA} is small. This seems all the more reasonable if we remember that one can (and actually, most of the time, does) implement the distorted wave strategy (see Sect. 2.1 and, in particular, Eq. (35)), without essentially altering the present discussion. Then the wavefunction corresponding to the spectator approximation plus the first order correction is,

Ψ≈Ψ0+ΨxA+GI(E) ΔVA Ψ0=Ψ0+ΨxA+GI(E)(VbA(rbA,ξ)−UbA(rbA))Ψ0.(111)\Psi\approx\Psi_0+\Psi_{xA}+\mathbf{G}_I(E)\,\Delta V_A\,\Psi_0=\Psi_0+\Psi_{xA}+\mathbf{G}_I(E)\left(V_{bA}(r_{bA},\xi)-U_{bA}(r_{bA})\right)\Psi_0. \tag*{(111)}

In order to describe the elastic process associated with the detection of particle bb, we proceed as for the derivation of Eq. (67) projecting onto the state Φ0χb(kb)\Phi_0\chi_b(k_b),

⟨Φ0χb(kb)∣Ψ≈ψ1I=ψHM+⟨Φ0∣G(E−Eb)⟨χb∣[VxA(rxA,ξ)+UbA(rbA)]Ψ0\langle\Phi_0\chi_b(k_b)|\Psi\approx\psi_1^I=\psi^{HM}+\langle\Phi_0|\mathbf{G}(E-E_b)\langle\chi_b|\left[V_{xA}(r_{xA},\xi)+U_{bA}(r_{bA})\right]\Psi_0
+⟨Φ0∣G(E−Eb)⟨χb∣(VbA(rbA,ξ)−UbA(rbA))Ψ0=ψ0I(rxA)+⟨Φ0∣G(E−Eb)⟨χb∣(VbA(rbA,ξ)−UbA(rbA))Ψ0,(112)\begin{aligned} \quad+\langle\Phi_0|\mathbf{G}(E-E_b)\langle\chi_b|\left(V_{bA}(r_{bA},\xi)-U_{bA}(r_{bA})\right)\Psi_0 \\ &=\psi_0^I(r_{xA})+\langle\Phi_0|\mathbf{G}(E-E_b)\langle\chi_b|\left(V_{bA}(r_{bA},\xi)-U_{bA}(r_{bA})\right)\Psi_0, \tag*{(112)} \end{aligned}

where we have used Eq. (68) to identify the spectator approximation wavefunction ψ0I\psi_0^I. In order to get some insight into the meaning of the correction to the spectator approximation, let us use the explicit expressions of G\mathbf{G} and Ψ0\Psi_0 (see Eqs. (58) and (62)),

G(E)=∑ij∣Φi(ξ)⟩((Ei−Eb−Tx(rxA))δij−Vij(rxA))−1⟨Φj(ξ)∣;\mathbf{G}(E)=\sum_{ij}|\Phi_i(\xi)\rangle\left((E_i-E_b-T_x(r_{xA}))\delta_{ij}-V_{ij}(r_{xA})\right)^{-1}\langle\Phi_j(\xi)|;
∣Ψ0⟩=∣ϕa(rxb)Φ0(ξ)F(ka,ra)⟩.(113)|\Psi_0\rangle=|\phi_a(r_{xb})\Phi_0(\xi)F(k_a,r_a)\rangle. \tag*{(113)}

Substituting in (112), the corrected wavefunction can be written as

ψ1I=ψ0I(rxA)+∑i((Ei−Eb−Tx(rxA))δ0i−V0i(rxA))−1\psi_1^I=\psi_0^I(r_{xA})+\sum_i\left((E_i-E_b-T_x(r_{xA}))\delta_{0i}-V_{0i}(r_{xA})\right)^{-1}
×⟨Φiχb∣(VbA(rbA,ξ)−UbA(rbA))∣ϕaΦ0F⟩=ψ0I(rxA)+∑i((Ei−Eb−Tx(rxA))δ0i−V0i(rxA))−1\begin{aligned} \quad\times\langle\Phi_i\chi_b|\left(V_{bA}(r_{bA},\xi)-U_{bA}(r_{bA})\right)|\phi_a\Phi_0F\rangle\\ &=\psi_0^I(r_{xA})+\sum_i\left((E_i-E_b-T_x(r_{xA}))\delta_{0i}-V_{0i}(r_{xA})\right)^{-1} \end{aligned}
×[⟨Φiχb∣VbA(rbA,ξ)∣ϕaΦ0F⟩−⟨χb∣UbA(rbA)∣ϕaF⟩δ0i].(114)\quad\times\left[\langle\Phi_i\chi_b|V_{bA}(r_{bA},\xi)|\phi_a\Phi_0F\rangle-\langle\chi_b|U_{bA}(r_{bA})|\phi_aF\rangle\delta_{0i}\right]. \tag*{(114)}

We will not try here to give a quantitative estimate of the importance of the correction to the spectator approximation contained in the 2nd term in the right hand side of the above expression. However, it is still possible to extract some physical insight from it, which will help assessing the regime of validity of the spectator approximation.

The terms with i≠0i\ne0 represent the inelastic excitations of AA due to the interaction of bb with AA, and they contribute to the scattered wave with an amplitude proportional to the matrix elements

⟨Φi≠0χb∣VbA(rbA,ξ)∣ϕaΦ0F⟩.(115)\langle\Phi_{i\ne0}\chi_b|V_{bA}(r_{bA},\xi)|\phi_a\Phi_0F\rangle. \tag*{(115)}

Part of this contribution is compensated by the removal of flux associated with the imaginary part of the optical potential UbAU_{bA}. However, standard optical potentials with a smooth energy dependence are associated with an energy-averaged description of the interaction of bb with AA. Therefore, they do not account for the contribution of narrow, tightly spaced resonances of the b+Ab+A system present at low energies, which give rise to a rapidly varying energy dependence of the scattering amplitude. The difference between the energy-averaged cross section accounted for by standard optical potentials and the rapidly-varying one is called the fluctuation cross section, and is characteristic of compound nuclear reactions [40].

The term corresponding to i=0i=0 also contributes to the correction term, with an amplitude proportional to

⟨Φ0χb∣VbA(rbA,ξ)∣ϕaΦ0F⟩−⟨χb∣UbA(rbA)∣ϕaF⟩.(116)\langle\Phi_0\chi_b \mid V_{bA}(r_{bA},\xi) \mid\phi_a\Phi_0 F\rangle-\langle\chi_b \mid U_{bA}(r_{bA}) \mid\phi_a F\rangle. \tag*{(116)}

This term represents the contribution associated with elastic processes not accounted for by the optical potential UbAU_{bA}. As for the inelastic processes discussed above, the origin of this difference lies in the contribution of narrow resonances of the b+Ab+A system which, in this case, decay by emitting bb back into the elastic channel. In this context, the fluctuation cross section is called the compound elastic contribution [40].

This discussion suggests that the spectator approximation might break whenever the incident effective energy of bb, determined by the peak of the broadening factor g(kb)g(k_b) (see Eq. (94)), is in the region of low-energy resonances of the b+Ab+A system. Therefore, in order to take advantage of the spectator approximation one should devise experiments with a high enough incident energy of the projectile aa. The same considerations apply to the exit channel, suggesting that the selected events should correspond to a high enough energy of the detected particle bb.

A possible physical picture that emerges from the structure of (109) and this discussion is the following. The spectator approximation accounts exactly for the scattering of xx with AA, while the interaction of bb with AA is treated in an effective way through the optical potential UbAU_{bA}. The resulting scattered wave is ΨxA\Psi_{xA}. The first correction (third term in (109)) consists in accounting for the scattering of bb with AA to first order in the difference between the real interaction VbAV_{bA} and the optical potential UbAU_{bA}. This correction accounts for the excitation of AA by bb and for elastic processes not accounted for by the optical potential UbAU_{bA}. Up to now, we have considered single-scattering events, while the next correction (fourth term in 109) accounts for a double-scattering process: first, xx scatters with AA, and then the interaction ΔVB=Vxb(rxb)+VbA(rbA,ξ)−UbB(rbB)\Delta V_B=V_{xb}(r_{xb})+V_{bA}(r_{bA},\xi)-U_{bB}(r_{bB}) drives a simultaneous scattering event of bb with AA and xx. This is the lowest order at which the genuinely three-body character of the system appears, arising from the three-body nature of the interaction ΔVB\Delta V_B.

Let us finally emphasize again that the soundness of this perturbative scheme rely on ΔVA\Delta V_A and ΔVB\Delta V_B being small. More specifically, the matrix elements of the kind of those appearing in (116) should be small. We have tried to show here that this rely on two basic assumptions: (i) the optical potential UbAU_{bA} should provide a good energy-averaged description of the interaction of bb with AA, and (ii) the incident energy of aa, and the final energy of bb, should be high enough to avoid the region of low-energy resonances of the AA and x+Ax+A systems, where standard optical potentials fail to describe the fluctuation cross section.

Acknowledgments

I wish to thank E. Vigezzi and F. Barranco for a critical reading of the manuscript and for many insightful suggestions.

References

  1. [1]W.H. Dickhoff, D. Van Neck, Many–Body Theory Exposed: Propagator Description of Quantum Mechanics in Many–Body Systems (World Scientific, 2005)
  2. [2]P.F. Bortignon, R.A. Broglia, D.R. Bes, R. Liotta, Phys. Rep. 30, 305 (1977)
  3. [3]D.R. Bes, Prog. Theor. Phys. Suppl. 74–75, 1 (1983). DOI 10.1143/PTPS.74.1. URL https://doi.org/10.1143/PTPS.74.1
  4. [4]J. Escher, J. Burke, F. Dietrich, N. Scielzo, I. Thompson, W. Younes, Rev. Mod. Phys. 84, 353
  5. [5]T. Udagawa, T. Tamura, Physical Review C 33, 494 (1986). DOI 10.1103/PhysRevC.33.494DOI
  6. [6]M. Ichimura, N. Austern, C.M. Vincent, Phys. Rev. C 32, 431 (1985). DOI 10.1103/PhysRevC.32.431DOI
  7. [7]M. Ichimura, Phys. Rev. C 41, 834 (1990). DOI 10.1103/PhysRevC.41.834DOI
  8. [8]G. Potel, F.M. Nunes, I.J. Thompson, Phys. Rev. C 92, 034611 (2015)
  9. [9]J. Lei, A.M. Moro, Phys. Rev. C 92(4), 044616 (2015)
  10. [10]B.V. Carlson, R. Capote, M. Sin, URL https://doi.org/10.48550/arXiv.1508.01466
  11. [11]G. Potel, G. Perdikakis, B. Carlson, M.C. Atkinson, W. Dickhoff, J. Escher, M. Hussein, J. Lei, W. Li, A.O. Macchiavelli, A.M. Moro, F.M. Nunes, S.D. Pain, J. Rotureau, Eur. Phys. J. A 53, 178 (2017)
  12. [12]I.J. Thompson, F.M. Nunes, Nuclear Reactions for Astrophysics: Principles, Calculation and Applications of Low-Energy Reactions, 1st edn. (Cambridge University Press, 2009). DOI 10.1017/CBO9781139152150. URL https://www.cambridge.org/core/product/identifier/9781139152150/type/bookDOI
  13. [13]D.F. Jackson, Nuclear Reactions (Methuen, 1970)DOI
  14. [14]A. Messiah, Mécanique Quantique (Dunod, 1995)
  15. [15]G. Satchler, Direct Nuclear Reactions (Clarendon Press, 1983)
  16. [16]A.M. Lane, R.G. Thomas, Rev. Mod. Phys. 30(2), 257 (1958). DOI 10.1103/RevModPhys.30.257. URL https://link.aps.org/doi/10.1103/RevModPhys.30.257DOI
  17. [17]P. Descouvemont, D. Baye, Rep. Prog. Phys. 73(3), 036301 (2010)
  18. [18]C. Bloch, Nucl. Phys. 4, 503 (1957). DOI 10.1016/0029-5582(87)90058-7. URL https://www.sciencedirect.com/science/article/pii/0029558287900587DOI
  19. [19]R.J. deBoer, J. Görres, M. Wiescher, et al., Rev. Mod. Phys. 89, 035007 (2017). DOI 10.1103/RevModPhys.89.035007DOI
  20. [20]A. Formicola, G. Imbriani, H. Costantini, C. Angulo, et al., Phys. Lett. B 591(1–2), 61 (2004). See also arXiv:nucl-ex/0312015DOI
  21. [21]T. Tamura, Rev. Mod. Phys. 37, 679 (1965)
  22. [22]I.J. Thompson, Comput. Phys. Rep. 7, 167 (1988)
  23. [23]H. Feshbach, Ann. of Phys. 19, 287 (1962)
  24. [24]R.A. Broglia, P.F. Bortignon, F. Barranco, E. Vigezzi, A. Idini, G. Potel, Phys. Scr. 91, 063012 (2016)
  25. [25]G. Potel, R.A. Broglia, The Nuclear Cooper Pair (Cambridge University Press, 2021)DOI
  26. [26]H. Feshbach, Ann. of Phys. 5, 357 (1958)
  27. [27]E.O. Alt, P. Grassberger, W. Sandhas, Nucl. Phys. B 2, 167 (1967). DOI 10.1016/0550-3213(67)90016-8DOI
  28. [28]W. Glöckle, The Quantum Mechanical Few-Body Problem. Texts and Monographs in Physics (Springer-Verlag, Berlin, Heidelberg, 1983). DOI 10.1007/978-3-642-82081-6DOI
  29. [29]L.D. Faddeev, S.P. Merkuriev, Quantum Scattering Theory for Several Particle Systems, Mathematical Physics and Applied Mathematics, vol. 11 (Springer, Dordrecht, 1993). DOI 10.1007/978-94-017-2832-4DOI
  30. [30]N. Austern, Y. Iseri, M. Kamimura, M. Kawai, G. Rawitscher, M. Yahiro, Phys. Rep. 154(3), 125 (1987). DOI 10.1016/0370-1573(87)90094-9DOI
  31. [31]M. Kamimura, M. Yahiro, Y. Iseri, Y. Sakuragi, H. Kameyama, M. Kawai, Prog. Theor. Phys. Suppl. 89, 1 (1986). DOI 10.1143/PTPS.89.1
  32. [32]M. Yahiro, K. Ogata, T. Matsumoto, K. Minomo, Prog. Theor. Exp. Phys. 2012(1), 01A206 (2012). DOI 10.1093/ptep/pts004
  33. [33]M. Hussein, K. McVoy, Nucl. Phys. A 445, 124 (1985)
  34. [34]A. Ratkiewicz, J.A. Cizewski, J.E. Escher, G. Potel, J.T. Burke, R.J. Casperson, M. McCleskey, R.A.E. Austin, S. Burcher, R.O. Hughes, B. Manning, S.D. Pain, W.A. Peters, S. Rice, T.J. Ross, N.D. Scielzo, C. Shand, K. Smith, Phys. Rev. Lett. 122, 052502 (2019)
  35. [35]F. Rotureau, G. Potel, W. Li, F.M. Nunes, J. Phys. G: Nucl. Part. Phys. 47, 065103 (2020)
  36. [36]F. Barranco, G. Potel, E. Vigezzi, R.A. Broglia, Phys. Rev. C 101(3), 031305 (2020)
  37. [37]W.H. Dickhoff, R.J. Charity, Prog. Part. Nucl. Phys. 105, 252 (2019)
  38. [38]G.H. Sargsyan, G. Potel, K. Kravvaris, J.E. Escher. DOI 10.48550/arXiv.2410.21714. URL http://arxiv.org/abs/2410.21714
  39. [39]F.C. Barker, Aust. J. Phys. 20(3), 341 (1967). DOI 10.1071/ph670341. URL https://www.publish.csiro.au/ph/ph670341DOI
  40. [40]F.L. Friedman, V.F. Weisskopf, in Niels Bohr and the Development of Physics: Essays Dedicated to Niels Bohr on the Occasion of His Seventieth Birthday, ed. by W. Pauli, L. Rosenfeld, V.F. Weisskopf (McGraw-Hill, New York, 1955), pp. 134–162

Paper details

vlnc:W_kpbbz6pc
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