betaFeedback

WorksPhysics and Astronomy

Reentrance of proton-neutron pairing in hot nuclear systems

Abstract

We develop a generalized finite-temperature proton-neutron BCS framework using the superoperator formalism, incorporating both isovector and isoscalar monopole pairing channels. Numerical calculations for a schematic equidistant multilevel model and realistic even-even Ge isotopes demonstrate the emergence of proton-neutron (pnpn) pairing reentrance in even-even asymmetric (N>ZN>Z) nuclei with preexisting like-nucleon pairing correlations. This nonmonotonic behavior arises from thermal excitations that partially lift Pauli blocking of single-particle orbitals near the chemical potentials, thereby enlarging the phase space for pnpn pair formation. We uncover a delicate interplay between thermal unblocking and like-nucleon pairing, which can either suppress or enhance pnpn correlations depending on temperature and shell filling. A qualitative analysis of Fermi charge-exchange strength functions in hot 72^{72}Ge, which neglects the residual interaction between thermal quasiparticles, suggests that pnpn pairing reentrance may alter the transition strength distribution around T≈1T\approx1~MeV. This indicates that finite-temperature pnpn correlations could potentially impact stellar weak-interaction rates in rprp-process and supernova environments.

INTRODUCTION

The pairing interaction lies at the heart of nuclear structure physics [1]. While like-nucleon pairing between protons (pppp) or neutrons (nnnn) dominates in neutron-rich nuclei (N≫ZN \gg Z), proton–neutron (pnpn) pairing becomes significant in nuclei with nearly equal numbers of protons and neutrons (N≈ZN \approx Z) [2], where valence protons and neutrons occupy closely spaced orbitals. Such N≈ZN \approx Z nuclei are typically located near the proton drip line and lie along the astrophysical rapid proton-capture (rprp) process path. Consequently, the study of pnpn pairing is not only essential for a complete understanding of nuclear structure but is also expected to play a crucial role in nuclear astrophysics [3].

At zero temperature, the simultaneous treatment of like-nucleon and pnpn pairing correlations has been extensively investigated, primarily within the pnpn Bardeen-Cooper-Schrieffer (pnBCS) framework. The isovector (t=1t = 1) pnBCS formalism was first derived by Goswami in 1964 [4]. This approach was subsequently extended to incorporate both isovector and isoscalar (t=0t = 0) correlations between nucleons in time-reversed states [5, 6]. A fully isospin-generalized theory, which additionally accounts for pairing correlations in identical orbitals, was later developed by Goodman [7].

In hot nuclei, the behavior of pnpn pairing has been explored in only a limited number of studies. In Ref. [3], pnpn pairing energies in odd–odd N=ZN = Z nuclei were extracted from a “thermal” odd–odd mass difference based on shell-model calculations. These results indicate an enhancement of pnpn pairing correlations at finite temperature in such nuclei. Similarly, Ref. [8] investigated both t=1t = 1 and t=0t = 0 pairing correlations in rotating deformed N=ZN = Z systems at finite temperature using an exactly solvable Hamiltonian. Their findings reveal that the pnpn pairing gap can increase when the nucleus is simultaneously hot and rotating. In both cases, the enhancement of pnpn pairing at finite temperature can be understood as a consequence of the thermal unblocking of single-particle levels near the Fermi surface, which would otherwise remain blocked at low temperatures either by an unpaired nucleon or by nucleons from pairs broken by rotation. Thermal excitations promote nucleons to higher levels and increase the number of proton-neutron configurations available for pair formation. As a result, pairing correlations that are suppressed at zero temperature may become enhanced over a certain temperature interval.

In asymmetric nuclear systems, pnpn pairing correlations are strongly suppressed in the ground state by Pauli blocking, as excess neutrons occupy single-particle orbitals that would otherwise be available for pnpn pair formation. In Ref. [9], the onset of superfluidity in asymmetric nuclear matter without like-nucleon correlations was investigated within the BCS framework. It was demonstrated that thermal excitations can actually enable the emergence of a superfluid phase at finite temperature. In contrast, path-integral studies of pnpn pairing at finite temperature in nuclei with both like-nucleon and pnpn correlations, performed for a schematic one-level model [10, 11] and for even-even N=ZN = Z nuclei [12], found no evidence of enhanced pnpn pairing at finite temperature.

While earlier studies of pnpn pairing in hot finite nuclei focused primarily on N=ZN = Z systems, the goal of the present work is to explore the emergence of pnpn pairing reentrance in asymmetric (N>ZN > Z) even-even nuclei. We investigate proton-neutron pairing correlations in hot nuclei within a finite-temperature generalization of the pnBCS framework. To this end, we employ the superoperator formalism, which naturally extends zero-temperature many-body concepts, such as vacuum states, quasiparticles, and phonons, etc, to the finite-temperature regime. This formalism has recently been successfully applied to calculate stellar weak-interaction rates and cross sections for hot nuclei (see, e.g., the review in Refs. [13] and references therein). Within this framework, we present the demonstration of temperature-induced reentrance of pnpn pairing in asymmetric even-even nuclei, explicitly accounting for the competing influence of like-nucleon correlations. To the best of our knowledge, this is the first study demonstrating pnpn pairing reentrance in N>ZN > Z nuclei with simultaneously treated like-nucleon and pnpn correlations. We show that this reentrance arises from the partial unblocking of single-particle levels that are otherwise Pauli-blocked and suppress pnpn pair formation at low temperatures. Furthermore, we provide a qualitative analysis of how the reentrance of pnpn pairing modifies the charge-exchange strength distributions in hot nuclear systems.

The paper is organized as follows. Section II presents the application of the superoperator formalism to finite-temperature pairing correlations in nuclear systems. We employ a pairing Hamiltonian that incorporates pppp, nnnn, and pnpn monopole interactions in both isovector (t=1t = 1) and isoscalar (t=0t = 0) channels, and derive the corresponding finite-temperature gap equations. For completeness, a concise overview of the superoperator formalism is provided in Appendix A. In Sec. III A we perform numerical calculations within a schematic equidistant multilevel pairing model to illustrate the non-monotonic temperature dependence of pnpn pairing and its reentrant enhancement. Section III B demonstrates that this reentrance persists in realistic calculations for even-even Ge isotopes and examines its impact on the Fermi-type charge-exchange strength function. Conclusions and future perspectives are summarized in Sec. IV.

FORMALISM

We investigate pairing correlations in hot even-even atomic nuclei, focusing specifically on a pairing Hamiltonian that incorporates constant-strength pppp, nnnn, and pnpn monopole pairing interactions between nucleons occupying time-reversed single-particle states [14]:

H(a†,a)=∑kτ(ϵkτ−λτ)(akτ†akτ+akˉτ†akˉτ)−∑τGτ(1)∑k,k′>0Skτ(1)†Sk′τ(1)−∑tGpn(t)∑k,k′>0Skpn(t)†Sk′pn(t).(1)H(a^\dagger,a)=\sum_{k\tau}(\epsilon_{k\tau}-\lambda_\tau)(a^\dagger_{k\tau}a_{k\tau}+a^\dagger_{\bar{k}\tau}a_{\bar{k}\tau})-\sum_\tau G^{(1)}_\tau\sum_{k,k'>0}S^{(1)\dagger}_{k\tau}S^{(1)}_{k'\tau}-\sum_tG^{(t)}_{pn}\sum_{k,k'>0}S^{(t)\dagger}_{kpn}S^{(t)}_{k'pn}. \tag*{(1)}

Here akτ†a^\dagger_{k\tau} and akτa_{k\tau} denote the creation and annihilation operators for a nucleon in the single-particle state kτk\tau (τ=p,n\tau= p,n) with the corresponding single-particle energy ϵkτ\epsilon_{k\tau}. The label kˉτ\bar{k}\tau denotes the time-reversed partner of the state kτk\tau, and λp\lambda_p (λn\lambda_n) represents the chemical potential introduced as a Lagrange multiplier to adjust the average number of protons (neutrons) corresponding to the nucleus under consideration. The operators Skτ(t)†S^{(t)\dagger}_{k\tau} create isovector (t=1t = 1) and isoscalar (t=0t = 0) nucleon pairs in time-reversed states:

Skτ(1)†=akτ†akˉτ†,S^{(1)\dagger}_{k\tau}=a^\dagger_{k\tau}a^\dagger_{\bar{k}\tau},
Skpn(t)†=12(akp†akˉn†−(−1)takn†akˉp†).(2)S^{(t)\dagger}_{kpn}=\frac{1}{\sqrt{2}}\left(a^\dagger_{kp}a^\dagger_{\bar{k}n}-(-1)^t a^\dagger_{kn}a^\dagger_{\bar{k}p}\right). \tag*{(2)}

The Pauli exclusion principle restricts isoscalar pairing exclusively to pnpn pairs, as identical fermions cannot form t=0t = 0 states in time-reversed states. The constant pairing strengths for like nucleons, Gn,p(1)G^{(1)}_{n,p}, and for pnpn pairs, Gpn(t)G^{(t)}_{pn}, are taken to be real. In general, we treat Gp(1)G^{(1)}_p, Gn(1)G^{(1)}_n, Gpn(1)G^{(1)}_{pn}, and Gpn(0)G^{(0)}_{pn} as independent parameters that are allowed to be different.

As briefly outlined in Appendix A (see [13] for comprehensive treatment), the superoperator formalism provides a framework for studying quantum many-body systems at finite temperature TT by recasting the calculation of equilibrium and non-equilibrium observables into an eigenvalue problem in the Liouville space. For a hot nucleus governed by the pairing Hamiltonian (1), this reduces to diagonalization of the thermal Hamiltonian H=H−H~\mathcal{H}=H-\widetilde{H}, where H~\widetilde{H} is the tilde-conjugate of HH, contracted by replacing all nucleon creation and annihilation operators with their tilded counterparts, i.e. H~=H(a~†,a~)\widetilde{H}=H(\widetilde{a}^{\dagger},\widetilde{a}). The thermal vacuum ∣0(T)⟩\lvert0(T)\rangle, the zero-eigenvalue state of the thermal Hamiltonian H\mathcal{H} satisfying the thermal state condition (TSC) (A14), fully characterizes the equilibrium properties of the hot nucleus, i.e. all thermal averages are evaluated as vacuum expectation values, ⟨X⟩=⟨0(T)∣X∣0(T)⟩\langle X\rangle=\langle0(T)|X|0(T)\rangle. Furthermore, the thermal vacuum serves as a reference vacuum state with respect to which all non-equilibrium excitations induced by external fields are constructed.

Within the independent thermal quasiparticle approximation, the thermal Hamiltonian H\mathcal{H} corresponding to the pairing Hamiltonian (1) is diagonalized in terms of thermal quasiparticle creation and annihilation operators yielding the canonical form

H≈∑k,νEkν(βkν†βkν−β~kν†β~kν).(3)\mathcal{H}\approx\sum_{k,\nu}E_{k\nu}\left(\beta^\dagger_{k\nu}\beta_{k\nu}-\widetilde{\beta}^\dagger_{k\nu}\widetilde{\beta}_{k\nu}\right). \tag*{(3)}

The thermal vacuum ∣0(T)⟩\lvert0(T)\rangle is then defined as a vacuum state of thermal quasiparticles (see Eq. (44)). Thermal quasiparticles are introduced through two successive unitary transformations. The first is the generalized Bogoliubov (u,v)(u,v)-transformation [14], which maps the original nucleon operators onto the quasiparticle operators αkν†\alpha^\dagger_{k\nu} and αkν\alpha_{k\nu} (ν=1,2\nu= 1,2):

\begin{pmatrix} a^\dagger_{kp}\\ a^\dagger_{kn}\\ a_{\bar{k}p}\\ a_{\bar{k}n} \end{pmatrix} = \begin{pmatrix} u_{k1p} & u_{k2p} & -v_{k1p} & -v_{k2p}\\ u_{k1n} & u_{k2n} & -v_{k1n} & -v_{k2n}\\ v_{k1p} & v_{k2p} & u_{k1p} & u_{k2p}\\ v_{k1n} & v_{k2n} & u_{k1n} & u_{k2n} \end{pmatrix} \begin{pmatrix} \alpha^\dagger_{k1}\\ \alpha^\dagger_{k2}\\ \alpha_{\bar{k}1}\\ \alpha_{\bar{k}2} \end{pmatrix}. \tag*{(4)}\end{pmatrix}

In general, the transformation amplitudes ukντu_{k\nu\tau} and vkντv_{k\nu\tau} are complex-valued. An analogous transformation with complex-conjugate amplitudes ukντ∗u^{*}_{k\nu\tau} and vkντ∗v^{*}_{k\nu\tau} is applied to the tilded nucleon operators a~kτ†\tilde{a}^{\dagger}_{k\tau} and a~kτ\tilde{a}_{k\tau} yielding the tilde-quasiparticle operators α~kν†\tilde{\alpha}^{\dagger}_{k\nu} and α~kν\tilde{\alpha}_{k\nu}. The second, so-called thermal (x,y)(x,y)-transformation, introduces thermal quasiparticles by mixing non-tilde and tilde quasiparticle operators

βkν†=xkναkν†−iykνα~kν,β~kν†=xkνα~kν†+iykναkν.(5)\beta^{\dagger}_{k\nu}=x_{k\nu}\alpha^{\dagger}_{k\nu}-iy_{k\nu}\tilde{\alpha}_{k\nu}, \qquad \tilde{\beta}^{\dagger}_{k\nu}=x_{k\nu}\tilde{\alpha}^{\dagger}_{k\nu}+iy_{k\nu}\alpha_{k\nu}. \tag*{(5)}

Here xkνx_{k\nu} and ykνy_{k\nu} are real and xkν2+ykν2=1x_{k\nu}^{2}+y_{k\nu}^{2}=1. It should be noted that the complex nature of the thermal transformation arises directly from the TSC (A14) (see Ref. [13] for a detailed discussion of this point). Since both transformations are unitary, the resulting thermal quasiparticle creation and annihilation superoperators preserve the canonical fermionic anticommutation relations.

To determine the coefficients of the (u,v)(u,v) and (x,y)(x,y) transformations, we first observe that if the thermal quasiparticles are eigenmodes of the thermal Hamiltonian and the vacuum ∣0(T)⟩|0(T)\rangle of thermal quasiparticles satisfies the TSC (A14), then the amplitudes ykνy_{k\nu} are fixed as

ykν=(11+eEkν/T)1/2.(6)y_{k\nu}=\left(\frac{1}{1+e^{E_{k\nu}/T}}\right)^{1/2}. \tag*{(6)}

This result follows directly by inserting, for instance, X=αkνX=\alpha_{k\nu} into the TSC (A14). On the other hand, ykν2y_{k\nu}^{2} is nothing but a fermionic occupation number of the quasiparticle state kνk\nu

ykν2=⟨0(T)∣αkν†αkν∣0(T)⟩.(7)y_{k\nu}^{2}=\langle0(T)|\alpha^{\dagger}_{k\nu}\alpha_{k\nu}|0(T)\rangle. \tag*{(7)}

To determine the amplitudes (u,v)(u,v), we impose the eigenmode condition

[H,βkν†]=Ekνβkν†.(8)[\mathcal{H},\beta^{\dagger}_{k\nu}]=E_{k\nu}\beta^{\dagger}_{k\nu}. \tag*{(8)}

This equation is most conveniently solved by applying Wick’s theorem to bring the thermal Hamiltonian into normal-ordered form with respect to the thermal vacuum. The only non-vanishing contractions are the following components of the single-particle density matrix ρkl≡⟨⟨ak†al⟩⟩=⟨⟨a~k†a~l⟩⟩∗\rho_{kl}\equiv\langle\langle a^{\dagger}_{k}a_{l}\rangle\rangle=\langle\langle\tilde{a}^{\dagger}_{k}\tilde{a}_{l}\rangle\rangle^{*} and pairing tensor κkl≡⟨⟨ak†al†⟩⟩=⟨⟨a~k†a~l†⟩⟩∗\kappa_{kl}\equiv\langle\langle a^{\dagger}_{k}a^{\dagger}_{l}\rangle\rangle=\langle\langle\tilde{a}^{\dagger}_{k}\tilde{a}^{\dagger}_{l}\rangle\rangle^{*}:

ρkτ;kτ′=∑ν(ykν2ukντukντ′∗+xkν2vkντvkντ′∗),\rho_{k\tau;k\tau'}=\sum_{\nu}\left(y_{k\nu}^{2}u_{k\nu\tau}u^{*}_{k\nu\tau'}+x_{k\nu}^{2}v_{k\nu\tau}v^{*}_{k\nu\tau'}\right),
κkτ;kτˉ′=∑ν(ykν2ukντvkντ′∗−xkν2vkντukντ′∗).(9)\kappa_{k\tau;k\bar{\tau}'}=\sum_{\nu}\left(y_{k\nu}^{2}u_{k\nu\tau}v^{*}_{k\nu\tau'}-x_{k\nu}^{2}v_{k\nu\tau}u^{*}_{k\nu\tau'}\right). \tag*{(9)}

and time-reversal invariance gives ρkτ;kτ′∗=ρkτˉ;kτˉ′\rho^{*}_{k\tau;k\tau'}=\rho_{k\bar{\tau};k\bar{\tau}'}, κkτ;kτˉ′∗=κkτ′;kτˉ\kappa^{*}_{k\tau;k\bar{\tau}'}=\kappa_{k\tau';k\bar{\tau}}. Then, performing linearization, i.e. neglecting in Eq. (8) the normal product triple terms, we arrive at the eigenvalue problem, whose solution yields the energy EkνE_{k\nu} and the (u,v)(u,v)-amplitudes of thermal quasiparticles:

\begin{pmatrix} \mathcal{E}_{kp} & \mathcal{E}_{kpn} & \Delta_{p} & \Delta_{pn} \\ \mathcal{E}_{knp} & \mathcal{E}_{kn} & \Delta_{np} & \Delta_{n} \\ \Delta_{p} & \Delta_{pn} & -\mathcal{E}_{kp} & -\mathcal{E}_{kpn} \\ \Delta_{np} & \Delta_{n} & -\mathcal{E}_{knp} & -\mathcal{E}_{kn} \end{pmatrix} \begin{pmatrix} u_{k\nu p}\\ u_{k\nu n}\\ v_{k\nu p}\\ v_{k\nu n} \end{pmatrix} = E_{k\nu} \begin{pmatrix} u_{k\nu p}\\ u_{k\nu n}\\ v_{k\nu p}\\ v_{k\nu n} \end{pmatrix}. \tag*{(10)}\end{pmatrix}

Here, the mean-field potentials and pairing gaps are given by

Ekτ=(ϵkτ−λτ)−Gτ(1)ρkτ;kτ−12(Gpn(1)+Gpn(0))ρkτ′;kτ′,\mathcal{E}_{k\tau}=(\epsilon_{k\tau}-\lambda_{\tau})-G_{\tau}^{(1)}\rho_{k\tau;k\tau}-\frac{1}{2}\left(G_{pn}^{(1)}+G_{pn}^{(0)}\right)\rho_{k\tau';k\tau'},
Ekττ′=−12(Gpn(1)−Gpn(0))ρkτ′;kτ,\mathcal{E}_{k\tau\tau'}=-\frac{1}{2}\left(G_{pn}^{(1)}-G_{pn}^{(0)}\right)\rho_{k\tau';k\tau},
Δτ=−Gτ(1)∑k>0κkτ;kτˉ,\Delta_{\tau}=-G_{\tau}^{(1)}\sum_{k>0}\kappa_{k\tau;k\bar{\tau}},
Δττ′=−12∑tGpn(t)∑k>0(κkτ;kτˉ′−(−1)tκkτ′;kτˉ).(11)\Delta_{\tau\tau'}=-\frac{1}{2}\sum_{t}G_{pn}^{(t)}\sum_{k>0}\left(\kappa_{k\tau;k\bar{\tau}'}-(-1)^{t}\kappa_{k\tau';k\bar{\tau}}\right). \tag*{(11)}

Hereafter, τ′\tau' denotes the isospin partner of τ\tau (i.e., τ′≠τ\tau'\ne\tau). From the symmetry properties of the density matrix and the pairing tensor it follows that Eτ\mathcal{E}_{\tau} and Δτ\Delta_{\tau} are real, while Ekττ′=Ekτ′τ∗\mathcal{E}_{k\tau\tau'}=\mathcal{E}^{*}_{k\tau'\tau} and Δττ′=Δτ′τ∗\Delta_{\tau\tau'}=\Delta^{*}_{\tau'\tau}. Consequently, the matrix in (10) is Hermitian, guaranteeing real eigenvalues EkνE_{k\nu}. These eigenvalues appear in pairs of opposite signs, a property well known from the zero-temperature pnBCS theory (see, e.g., Ref. [4]). In what follows, the subscript ν=1,2\nu=1,2 labels the positive-energy solutions EkνE_{k\nu}, which correspond to the non-tilde thermal quasiparticles, while their tilde-conjugate partners carry the negative energy −Ekν-E_{k\nu}.

Using explicit expressions for the pairing tensor κkτ;kτˉ′\kappa_{k\tau;k\bar{\tau}'} (see Eq. (9)), we can write the pairing gaps as

Δτ=Gτ(1)2∑ν,k>0(ukντvkντ∗+ukντ∗vkντ)(1−2ykν2),\Delta_{\tau}=\frac{G_{\tau}^{(1)}}{2}\sum_{\nu,k>0}\left(u_{k\nu\tau}v^{*}_{k\nu\tau}+u^{*}_{k\nu\tau}v_{k\nu\tau}\right)(1-2y_{k\nu}^{2}),
Δττ′=Δττ′(1)+iΔττ′(0),\Delta_{\tau\tau'}=\Delta_{\tau\tau'}^{(1)}+i\Delta_{\tau\tau'}^{(0)},
Δττ′(1)=Gpn(1)2∑ν,k>0Re⁡(ukντvkντ′∗+ukντ′vkντ∗)(1−2ykν2),\Delta_{\tau\tau'}^{(1)}=\frac{G_{pn}^{(1)}}{2}\sum_{\nu,k>0}\operatorname{Re}\left(u_{k\nu\tau}v^{*}_{k\nu\tau'}+u_{k\nu\tau'}v^{*}_{k\nu\tau}\right)(1-2y_{k\nu}^{2}),
Δττ′(0)=Gpn(0)2∑ν,k>0Im⁡(ukντvkντ′∗−ukντ′vkντ∗)(1−2ykν2).(12)\Delta_{\tau\tau'}^{(0)}=\frac{G_{pn}^{(0)}}{2}\sum_{\nu,k>0}\operatorname{Im}\left(u_{k\nu\tau}v^{*}_{k\nu\tau'}-u_{k\nu\tau'}v^{*}_{k\nu\tau}\right)(1-2y_{k\nu}^{2}). \tag*{(12)}

Thus, the real part of the pnpn pairing gap originates from isovector pairing, whereas its imaginary part stems from isoscalar pairing. This feature, well known from zero-temperature analyses [5, 6], persists at finite temperature.

Together with the particle-number constraints

Nτ=2∑k>0ρkτ;kτ,(τ=p,n).(13)N_{\tau}=2\sum_{k>0}\rho_{k\tau;k\tau},\qquad(\tau=p,n). \tag*{(13)}

Eqs. (6), (10), and (12) form a closed set of finite-temperature pnBCS (FT-pnBCS) equations. Owing to their coupled non-linear structure, these equations must be solved self-consistently. In numerical implementations, an iterative scheme is typically employed, initialized from the zero-temperature pnBCS solution, and then gradually evolved to the desired temperature. Due to their non-linear character, the FT-pnBCS equations may admit multiple stationary solutions, depending on the initial zero-temperature seed. The physically relevant solution is uniquely identified as the one that minimizes the free energy.

Analytical expressions for the quasiparticle energies EkνE_{k\nu} can be derived by employing the reduced form of the squared FT-pnBCS matrix, following the method developed for the zero-temperature case [4]. These expressions are considerably simplified under the approximation that single-particle states are not renormalized, specifically, by setting εkτ=ϵkτ−λτ\varepsilon_{k\tau}=\epsilon_{k\tau}-\lambda_\tau and Ekττ′=0\mathcal{E}_{k\tau\tau'}=0. Namely,

Ekν2=12(Ekp2+Ekn2+2∣Δpn∣2+(−1)ν+1Rk),ν=1,2,(14)E_{k\nu}^{2}=\frac{1}{2}\left(E_{kp}^{2}+E_{kn}^{2}+2|\Delta_{pn}|^{2}+(-1)^{\nu+1}R_k\right),\qquad\nu=1,2, \tag*{(14)}

where Ekτ=εkτ2+Δτ2E_{k\tau}=\sqrt{\varepsilon_{k\tau}^{2}+\Delta_\tau^{2}} (εkτ=ϵkτ−λτ\varepsilon_{k\tau}=\epsilon_{k\tau}-\lambda_\tau) are conventional BCS quasiparticle energies, and

Rk=[(Ekp2−Ekn2)2+4∣Δpn∣2((Δp+Δn)2+(εkp−εkn)2)]1/2.R_k=\left[(E_{kp}^{2}-E_{kn}^{2})^{2}+4|\Delta_{pn}|^{2}\left((\Delta_p+\Delta_n)^{2}+(\varepsilon_{kp}-\varepsilon_{kn})^{2}\right)\right]^{1/2}.

As it should be Ek1(2)=Ekp(n)E_{k1(2)}=E_{kp(n)} for vanishing proton-neutron pairing.

Furthermore, within the above approximation, analytical expressions for the (u,v)(u,v)-coefficients can be derived. Substituting these expressions into Eqs. (12) and (13) then yields a coupled system of FT-pnBCS gap equations:

Nτ=∑k>0[1−∑νtanh⁡(Ekν2T)12Ekν×{εkτ+(−1)1+ν(Ekτ2−Ekτ′2)εkτ+2∣Δpn∣2(εkτ−εkτ′)Rk}],N_\tau=\sum_{k>0}\left[1-\sum_\nu\tanh\left(\frac{E_{k\nu}}{2T}\right)\frac{1}{2E_{k\nu}}\times\left\{\varepsilon_{k\tau}+(-1)^{1+\nu}\frac{(E_{k\tau}^{2}-E_{k\tau'}^{2})\varepsilon_{k\tau}+2|\Delta_{pn}|^{2}(\varepsilon_{k\tau}-\varepsilon_{k\tau'})}{R_k}\right\}\right],
4ΔτGττ(1)=∑ν,k>0tanh⁡(Ekν2T)1Ekν{Δτ+(−1)1+ν(Ekτ2−Ekτ′2)Δτ+2∣Δpn∣2(Δp+Δn)Rk},\frac{4\Delta_\tau}{G_{\tau\tau}^{(1)}}=\sum_{\nu,k>0}\tanh\left(\frac{E_{k\nu}}{2T}\right)\frac{1}{E_{k\nu}}\left\{\Delta_\tau+(-1)^{1+\nu}\frac{(E_{k\tau}^{2}-E_{k\tau'}^{2})\Delta_\tau+2|\Delta_{pn}|^{2}(\Delta_p+\Delta_n)}{R_k}\right\},
4Δpn(t)Gpn(t)=Δpn(t)∑ν,k>0tanh⁡(Ekν2T)1Ekν{1+(−1)1+νEkp2+Ekn2+2(ΔpΔn−εkpεkn)Rk},(15)\frac{4\Delta_{pn}^{(t)}}{G_{pn}^{(t)}}=\Delta_{pn}^{(t)}\sum_{\nu,k>0}\tanh\left(\frac{E_{k\nu}}{2T}\right)\frac{1}{E_{k\nu}}\left\{1+(-1)^{1+\nu}\frac{E_{kp}^{2}+E_{kn}^{2}+2(\Delta_p\Delta_n-\varepsilon_{kp}\varepsilon_{kn})}{R_k}\right\}, \tag*{(15)}

where tanh⁡(Ekν2T)=1−2ykν2\tanh\left(\frac{E_{k\nu}}{2T}\right)=1-2y_{k\nu}^{2}. The same gap equations were derived in [11] using a path-integral approach.

The key feature of the FT-pnBCS gap equations (15) is that the summation term in the last equation is identical for both the t=0t=0 and t=1t=1 isospin channels. Consequently, when Gpn(1)≠Gpn(0)G_{pn}^{(1)}\ne G_{pn}^{(0)}, the isovector and isoscalar pnpn pairing correlations cannot coexist. At zero temperature, this result was observed in [14, 15] by numerical solution of the pnBCS gap equations. For Gpn(1)>Gpn(0)G_{pn}^{(1)}>G_{pn}^{(0)}, the solution that minimizes the free energy is Δpn(1)≠0\Delta_{pn}^{(1)}\ne0 and Δpn(0)=0\Delta_{pn}^{(0)}=0. In the particular case Gpn(1)=Gpn(0)G_{pn}^{(1)}=G_{pn}^{(0)}, both t=1t=1 and t=0t=0 pairing modes coexist and ∣Δpn(t)∣=∣Δpn∣/2|\Delta_{pn}^{(t)}|=|\Delta_{pn}|/\sqrt{2}.

Furthermore, since the gap equations (15) are invariant under the transformation Δpn(t)→−Δpn(t)\Delta_{pn}^{(t)}\to-\Delta_{pn}^{(t)}, the sign of Δpn(t)\Delta_{pn}^{(t)} is arbitrary and we adopt the convention Δpn(t)≥0\Delta_{pn}^{(t)}\ge0 without loss of generality. In contrast, the relative sign of Δp\Delta_p and Δn\Delta_n has physical significance when Δpn=0\Delta_{pn}=0, as it directly influences the coupling between like-nucleon and pnpn pairing channels.

Assuming Δpn=0\Delta_{pn}=0 in (15), we obtain a critical value of the pnpn interaction strength below which pnpn-pairing correlations vanish

2Gpncr=∑τ,k>0tanh⁡(Ekτ2T)Ekτ2+ΔpΔn−εkpεknEkτ(Ekτ2−Ekτ′2).(16)\frac{2}{G_{pn}^{\mathrm{cr}}}=\sum_{\tau,k>0}\tanh\left(\frac{E_{k\tau}}{2T}\right)\frac{E_{k\tau}^{2}+\Delta_p\Delta_n-\varepsilon_{kp}\varepsilon_{kn}}{E_{k\tau}(E_{k\tau}^{2}-E_{k\tau'}^{2})}. \tag*{(16)}

In the next section, we will demonstrate that GpncrG_{pn}^{\mathrm{cr}} can exhibit a non-monotonic dependence on temperature. This, in turn, leads to an enhancement of pnpn pairing correlations in certain temperature regimes.

A further simplification of the gap equations (15) is possible for self-conjugate nuclei (Nn=NpN_n=N_p) when the Coulomb interaction is neglected. Under these conditions ϵkp=ϵkn\epsilon_{kp}=\epsilon_{kn} and Gp(1)=Gn(1)G_p^{(1)}=G_n^{(1)}. Moreover, charge symmetry requires that the pairing tensors satisfy κkp,kp=−κkn,kn\kappa_{kp,kp}=-\kappa_{kn,kn} [16]. Therefore, Δn=−Δp\Delta_n=-\Delta_p, Rk=0R_k=0 and (15) reduces to a simple form

Nτ=∑ν,k>0[1−tanh⁡(Ekν2T)εkτEkν],N_\tau=\sum_{\nu,k>0}\left[1-\tanh\left(\frac{E_{k\nu}}{2T}\right)\frac{\varepsilon_{k\tau}}{E_{k\nu}}\right],
4ΔτGττ(1)=Δτ∑ν,k>0tanh⁡(Ekν2T)1Ekν,\frac{4\Delta_\tau}{G_{\tau\tau}^{(1)}}=\Delta_\tau\sum_{\nu,k>0}\tanh\left(\frac{E_{k\nu}}{2T}\right)\frac{1}{E_{k\nu}},
4Δpn(t)Gpn(t)=Δpn(t)∑ν,k>0tanh⁡(Ekν2T)1Ekν,(17)\frac{4\Delta_{pn}^{(t)}}{G_{pn}^{(t)}}=\Delta_{pn}^{(t)}\sum_{\nu,k>0}\tanh\left(\frac{E_{k\nu}}{2T}\right)\frac{1}{E_{k\nu}}, \tag*{(17)}

where the quasiparticle energies Ekν=(εkτ2+Δτ2+∣Δpn∣2)1/2E_{k\nu}=(\varepsilon_{k\tau}^{2}+\Delta_\tau^{2}+|\Delta_{pn}|^{2})^{1/2} are twice degenerate. Thus, in self-conjugated nuclei, like-nucleon and pnpn pairing correlations do not coexist within the FT-pnBCS approximation unless the isovector strength Gn,p(1)G_{n,p}^{(1)} equals the pnpn interaction strength in either the isoscalar or isovector channel. Isospin symmetry requires the isovector pairing strengths to be equal, i.e., Gn,p(1)=Gpn(1)G_{n,p}^{(1)}=G_{pn}^{(1)}. Under this condition, isovector pairing dominates when Gn,p(1)>Gpn(0)G_{n,p}^{(1)}>G_{pn}^{(0)}, whereas isoscalar pairing becomes energetically favored for Gpn(0)>Gn,p(1)G_{pn}^{(0)}>G_{n,p}^{(1)}. Importantly, in the isospin-symmetric limit with Gpn(0)>Gn,p(1)G_{pn}^{(0)}>G_{n,p}^{(1)}, isovector like-nucleon and isoscalar pnpn pairing may coexist within the approximate Lipkin-Nogami particle-number projection scheme [15].

For completeness, we present explicit expressions for the total energy and entropy used to evaluate the free energy, F=E−TSF = E - TS, within the FT-pnBCS approximation:

E=⟨0(T)∣H∣0(T)⟩+∑τλτNτ=∑k>0(ϵpk+ϵnk)−∑ν,k>0tanh⁡(Ekν2T)12Ekν×{(ϵpkϵpk+ϵnkϵnk)+(−1)1+νR[(ϵpkϵpk−ϵnkϵnk)]×(Ekp2−Ekn2)+2∣Δpn∣2(ϵpk−ϵnk)(ϵpk−ϵnk)}−Δn2Gn(1)−Δp2Gp(1)−2(Δpn(0))2Gpn(0)−2(Δpn(1))2Gpn(1),(18)\begin{aligned} E &= \langle0(T)|H|0(T)\rangle+ \sum_{\tau} \lambda_{\tau}N_{\tau} \\ &= \sum_{k>0}(\epsilon_{pk}+\epsilon_{nk})-\sum_{\nu,k>0}\tanh\left(\frac{E_{k\nu}}{2T}\right)\frac{1}{2E_{k\nu}} \\ &\quad\times\left\{(\epsilon_{pk}\epsilon_{pk}+\epsilon_{nk}\epsilon_{nk})+\frac{(-1)^{1+\nu}}{R}\left[(\epsilon_{pk}\epsilon_{pk}-\epsilon_{nk}\epsilon_{nk})\right]\right.\\ &\quad\left.\times(E_{kp}^{2}-E_{kn}^{2})+2|\Delta_{pn}|^{2}(\epsilon_{pk}-\epsilon_{nk})(\epsilon_{pk}-\epsilon_{nk})\right\}\\ &\quad-\frac{\Delta_{n}^{2}}{G_{n}^{(1)}}-\frac{\Delta_{p}^{2}}{G_{p}^{(1)}}-2\frac{(\Delta_{pn}^{(0)})^{2}}{G_{pn}^{(0)}}-2\frac{(\Delta_{pn}^{(1)})^{2}}{G_{pn}^{(1)}}, \tag*{(18)} \end{aligned}
S=−2∑ν,k>0[ykν2ln⁡ykν2+(1−ykν2)ln⁡(1−ykν2)].(19)S=-2\sum_{\nu,k>0}\left[y_{k\nu}^{2}\ln y_{k\nu}^{2}+(1-y_{k\nu}^{2})\ln(1-y_{k\nu}^{2})\right]. \tag*{(19)}

In the absence of pnpn pairing, these expressions reduce to the corresponding FT-BCS results given in Ref. [17].

NUMERICAL CALCULATIONS AND RESULTS

In this section, we apply the formalism developed above to study pnpn pairing in hot nuclear systems, including a schematic model with equidistant doubly degenerate single-particle levels and realistic even–even germanium isotopes. Prior to presenting the results, we first outline the key technical aspects of the numerical implementation.

To solve the highly nonlinear coupled gap equations (15), we employed the NEQBF subroutine from the IMSL Fortran Numerical Library [18], which solves a system of nonlinear equations using factored secant update with a finite-difference approximation to the Jacobian. As was mentioned above, the iterative solution procedure is initialized with the zero-temperature pnBCS solution and then continued to the target temperature by incrementally increasing TT and using the solution at each step as the initial guess for the next. This continuation strategy ensures convergence across the full temperature range. Depending on the initial zero-temperature seed, the FT-pnBCS equations may admit multiple stationary solutions. A crucial numerical aspect is that a non-trivial pnpn pairing solution (Δpn≠0\Delta_{pn}\ne0) is obtained only when the iterative procedure is initialized with like-nucleon gaps of opposite sign, i.e., Δp⋅Δn<0\Delta_p\cdot\Delta_n<0. This sign asymmetry in the initial conditions reflects the properties of charge conjugation operation [16], which leads to Δp=−Δn\Delta_p=-\Delta_n in isospin symmetric nuclei. For clarity, all subsequent figures and discussions display the absolute values ∣Δp∣|\Delta_p|, ∣Δn∣|\Delta_n|, and we adopt Δpn(t)≥0\Delta_{pn}^{(t)}\ge0 by convention. To further verify that the pnpn pairing solution indeed corresponds to the global minimum of the free energy (and not to

a metastable branch selected by the continuation procedure), we performed two independent checks. First, we initialized the iterative solver at several intermediate temperatures with different starting values for the pairing gaps Δp\Delta_p, Δn\Delta_n, and Δpn\Delta_{pn}; in all cases, the iterative procedure converged to the same solution obtained by the continuation strategy. Second, we compared the free energies of the three competing stationary solutions, as illustrated in Figure 4 below.

Free energy as a function of temperature

Figure 4. Free energy as a function of temperature for the parameter set of Figure 3(c), evaluated for three distinct solutions of the gap equations (15): (1) the trivial normal-phase solution; (2) the solution with only like-nucleon pairing; and (3) the solution with all pairing channels active. Solution (4) corresponds to Gpn=0.7G_{pn}=0.7.

As demonstrated above, within the FT-pnBCS framework, the t=1t=1 and t=0t=0 pnpn pairing channels are mutually exclusive when the coupling strength Gpn(1)G_{pn}^{(1)} and Gpn(0)G_{pn}^{(0)} are unequal. The system instead selects the channel with the stronger interaction, as it minimizes the free energy. Consequently, for the numerical calculations presented in this section we restrict ourselves to a single isospin channel, either pure isovector or pure isoscalar, and denote the corresponding pairing strength and gap simply by GpnG_{pn} and Δpn\Delta_{pn}.

Moreover, we restrict our analysis to asymmetric systems (Nn>NpN_n>N_p). In the symmetric limit (see Eqs. (17)), whenever a finite pnpn gap Δpn(t)\Delta_{pn}^{(t)} exists, its thermal evolution parallels that of the like-nucleon gaps Δp,n\Delta_{p,n}: Δpn(t)\Delta_{pn}^{(t)} decreases monotonically with temperature and vanishes at a critical temperature, exhibiting no reentrant behavior.

Schematic model

As a schematic model, we consider a system with Ω=10\Omega=10 doubly degenerate equidistant single-particle levels for both protons and neutrons. The single-particle energies ϵk\epsilon_k (k=1,…,Ωk=1,\ldots,\Omega) are identical in the proton and neutron subsystems and increase linearly with kk, with a constant spacing ϵk+1−ϵk=1\epsilon_{k+1}-\epsilon_k=1 set as the energy unit for all quantities in the calculation (pairing gaps, temperature, etc.). The double degeneracy of each level accounts for time-reversed partners kk and kˉ\bar{k}. The proton number is fixed at Np=10N_p=10, while the neutron number NnN_n, the like-nucleon pairing strengths Gn,pG_{n,p}, and the pnpn pairing strength GpnG_{pn} are treated as variable parameters.

Figure 1(a) displays the temperature dependence of the proton and neutron pairing gaps for a system with Nn=14N_n=14 and Gn,p=0.5G_{n,p}=0.5 calculated in the absence of proton–neutron pairing (Gpn=0G_{pn}=0). As expected within the finite-temperature BCS framework, both gaps decrease monotonically with increasing temperature and vanish at their respective critical temperatures. Since the proton and neutron pairing strengths are equal and the proton subsystem is exactly half-filled, the proton gap remains larger than the neutron gap at all temperatures and its critical temperature is higher.

Figure 1. (a) Temperature dependence of the proton (Δp\Delta_p) and neutron (Δn\Delta_n) pairing gaps in the absence of proton-neutron pairing (Gpn=0G_{pn}=0). (b) Ground-state (T=0T=0) pairing gaps Δp\Delta_p, Δn\Delta_n, and Δpn\Delta_{pn} as functions of the proton-neutron pairing strength GpnG_{pn}. In both panels the particle numbers are fixed at Np=10N_p=10 and Nn=14N_n=14, and the like-nucleon pairing strength is Gn,p=0.5G_{n,p}=0.5.

Keeping all other parameters fixed, Figure 1(b) illustrates the onset of pnpn pairing correlations in the ground state (T=0T=0) as the interaction strength GpnG_{pn} increases. A critical value GpncrG_{pn}^{\mathrm{cr}} exists below which Δpn=0\Delta_{pn}=0. For

Gpn>GpncrG_{pn}>G_{pn}^{\mathrm{cr}}, Δpn\Delta_{pn} becomes finite and grows monotonically with GpnG_{pn}. The like-nucleon gaps Δp\Delta_p and Δn\Delta_n remain unchanged for Gpn≤GpncrG_{pn}\leq G_{pn}^{\mathrm{cr}} and decrease only slightly once pnpn pairing sets in. This behavior is characteristic of the pnBCS approach with a constant pairing interaction: the emergence of proton-neutron correlations competes weakly with like-nucleon pairing, allowing both types of correlations to coexist above the critical strength – a feature previously reported in the literature [14, 19].

The key question is how the critical strength GpncrG_{pn}^{\mathrm{cr}} for the onset of pnpn pairing depends on temperature and pre-existing like-nucleon pairing correlations. To address this, Figure 2(a) displays the temperature dependence of GpncrG_{pn}^{\mathrm{cr}} for the Np=10N_p=10, Nn=14N_n=14 system computed using Eq. (16) for two scenarios: with and without like-nucleon pairing included. In both scenarios, GpncrG_{pn}^{\mathrm{cr}} exhibits non-monotonic behavior: it initially decreases with rising temperature, reaches a minimum at some intermediate temperature, and subsequently increases at higher temperatures. To demonstrate that the observed non-monotonic behavior is not an artifact of particular proton–neutron asymmetry, Figure 2(b) presents the temperature dependence of GpncrG_{pn}^{\mathrm{cr}} for an alternative configuration with Nn=12N_n=12 and Np=10N_p=10. The qualitative similarity with the Nn=14N_n=14 case suggests that the observed non-monotonic temperature dependence of GpncrG_{pn}^{\mathrm{cr}} is not limited to a single choice of particle numbers or asymmetry.

Figure 2. Critical pnpn pairing strength GpncrG_{pn}^{\mathrm{cr}} as a function of temperature TT for several values of the like-nucleon pairing strengths Gn,pG_{n,p}. Panel (a) corresponds to Np=10N_p=10, Nn=14N_n=14; panel (b) shows the case Np=10N_p=10, Nn=12N_n=12.

It is also evident that the increase in GpncrG_{pn}^{\mathrm{cr}} observed in Figure 2 with increasing neutron number reflects the Pauli blocking effect induced by the neutron excess above the Np=NnN_p=N_n core: excess neutrons occupy single-particle levels that would otherwise be available for proton-neutron pairing (see Figure 2 in [20] for a schematic illustration of the blocking mechanism for pnpn pairing). Furthermore, the non-monotonic dependence of GpncrG_{pn}^{\mathrm{cr}} on like-nucleon pairing strengths and temperature demonstrates that both thermal excitations and like-nucleon correlations can either suppress or enhance pnpn pairing, depending on the specific conditions. Since like-nucleon pairing is itself temperature-dependent, these two mechanisms are intrinsically coupled. Consequently, the formation of proton-neutron pairing correlations in hot nuclei is governed by a delicate interplay between thermal effects and like-nucleon pairing.

This interplay is illustrated in Figure 3, which displays the temperature evolution of the neutron, proton, and pnpn pairing gaps for several combinations of the interaction strengths GpnG_{pn} and Gn,pG_{n,p}. The figure reveals how the competition between thermal effects and the effect of like-nucleon pairing correlations governs the non-monotonic behavior of Δpn\Delta_{pn}, including its characteristic reentrance at intermediate temperatures when like-nucleon pairing is quenched. This behavior of pnpn pairing is analogous to the reentrance phenomenon of like-nucleon pairing, which appears in hot odd nuclei [21, 22] and/or hot rotating nuclei [23, 24, 25]. Notably, in contrast to the Gpn=0G_{pn}=0 case shown in Figure 1(a), the critical temperatures, at which the proton and neutron pairing gaps vanish, coincide when pnpn pairing is present. This synchronization arises because pnpn correlations couple the proton and neutron subsystems, forcing their pairing phase transitions to occur simultaneously despite the underlying particle-number asymmetry.

Figure 3. Temperature dependence of the proton, neutron, and pnpn pairing gaps for selected combinations of the interaction strengths Gn,pG_{n,p} and GpnG_{pn}. Panels (a)–(c) correspond to Gn,p=0.50G_{n,p}=0.50 with Gpn=0.62G_{pn}=0.62, 0.640.64, and 0.650.65, respectively; panel (d) shows the case Gn,p=0.53G_{n,p}=0.53, Gpn=0.66G_{pn}=0.66. The particle numbers (Np=10N_p=10, Nn=14N_n=14) and the number of doubly degenerate single-particle levels (Ω=10\Omega=10) are kept fixed. The reentrance of Δpn\Delta_{pn} at intermediate temperatures, coinciding with the quenching of like-nucleon pairing, is clearly visible in all panels.

To isolate the thermal unblocking mechanism from the influence of like-nucleon pairing correlations, Figure 3 additionally displays the temperature dependence of Δpn\Delta_{pn} in the limiting case Gn,p=0G_{n,p}=0. This reference calculation reveals how thermal effects alone govern the onset and evolution of pnpn pairing correlations. First, we observe that at moderate temperatures and in the absence of like-nucleon pairing, Δpn\Delta_{pn} evolves more rapidly with temperature and attains its maximum value at lower temperatures compared to calculations that include like-nucleon correlations. This comparison demonstrates that like-nucleon pairing suppresses pnpn correlations at moderate temperatures. In contrast, at low temperatures the influence of like-nucleon pairing on the pnpn gap is ambivalent: depending on the magnitude of Gn,pG_{n,p}, like-nucleon correlations can either enhance or suppress Δpn\Delta_{pn}.

As mentioned above, the highly nonlinear coupled gap equations (15) may admit multiple stationary solutions. To identify the physically relevant state, we select the solution that minimizes the free energy F=E−TSF=E-TS, computed within the FT-pnBCS approximation (see Eqs. (18), (19)). As an illustrative example, Figure 4 displays the free energy evaluated for the three distinct solutions of Eqs. (15), using the same parameter set as in Figure 3(c). As expected, the onset of nonzero like-nucleon pairing significantly lowers the free energy compared to the normal phase. The difference in free energy between the two superfluid branches (with and without pnpn pairing) is relatively modest, which is natural since the chosen interaction strength Gpn=0.65G_{pn}=0.65 lies only slightly above the critical value for the onset of pnpn pairing. Nevertheless, as clearly evident from the figure, the reentrant pnpn pairing solution consistently corresponds to the minimum of the free energy throughout the entire temperature range where it exists. The figure also illustrates the decrease in the free energy with increasing GpnG_{pn}.

In the limiting case Gn,p=0G_{n,p}=0, the FT-pnBCS equations admit an analytical reduction. Specifically, the quantity RkR_k defined in Eq. (14) takes the form

Rk=2∣λp−λn∣Ck,Ck=[(ϵk−12(λp+λn))2+Δpn2]1/2(20)R_k = 2\lvert\lambda_p-\lambda_n\rvert C_k,\qquad C_k = \left[\left(\epsilon_k-\frac{1}{2}(\lambda_p+\lambda_n)\right)^2+\Delta_{pn}^2\right]^{1/2} \tag*{(20)}

and the quasiparticle energies become

Ek1=Ek+,Ek2=∣Ek−∣,(21)E_{k1}=E_k^{+},\qquad E_{k2}=\lvert E_k^{-}\rvert, \tag*{(21)}

where Ek±=Ck±12∣λp−λn∣E_k^{\pm}=C_k\pm\frac{1}{2}\lvert\lambda_p-\lambda_n\rvert. Note that while Ek+E_k^{+} is always positive, Ek−E_k^{-} becomes negative for single-particle levels satisfying (ϵk−λp)(ϵk−λn)<−Δpn2(\epsilon_k-\lambda_p)(\epsilon_k-\lambda_n)<-\Delta_{pn}^{2}. All such levels lie between the proton and neutron chemical potentials, where the product (ϵk−λp)(ϵk−λn)(\epsilon_k-\lambda_p)(\epsilon_k-\lambda_n) is negative. Substituting these expressions for RkR_k and EkνE_{k\nu} into the gap equation (15) yields a compact form for the pnpn pairing gap:

4ΔpnGpn=Δpn∑k1Ck{tanh⁡(Ek12T)+Ek−Ek2tanh⁡(Ek22T)},(22)\frac{4\Delta_{pn}}{G_{pn}}=\Delta_{pn}\sum_k\frac{1}{C_k}\left\{\tanh\left(\frac{E_{k1}}{2T}\right)+\frac{E_k^{-}}{E_{k2}}\tanh\left(\frac{E_{k2}}{2T}\right)\right\}, \tag*{(22)}

or, equivalently,

2ΔpnGpn=Δpn∑k1Ck[1−f(Ek+)−f(Ek−)],(23)\frac{2\Delta_{pn}}{G_{pn}}=\Delta_{pn}\sum_k\frac{1}{C_k}\left[1-f(E_k^{+})-f(E_k^{-})\right], \tag*{(23)}

where f(E)=(1+eE/T)−1f(E)=(1+e^{E/T})^{-1} is the Fermi-Dirac distribution function. Equation (23) represents the finite-temperature generalization of the pnpn pairing gap equation originally derived by Soloviev using Green’s function techniques [26]. A closely related gap equation for asymmetric nuclear matter was subsequently obtained by Sedrakian et al. within a similar formalism [9].

The simplified gap equation (23) reveals that, in the absence of like-nucleon pairing, thermal effects enter exclusively through the occupation factors f(Ek±)f(E_k^{\pm}). Notably, for single-particle levels with energies lying between the proton and neutron chemical potentials, the combination 1−f(Ek+)−f(Ek−)1-f(E_k^{+})-f(E_k^{-}) acts as an effective blocking factor at low temperatures. In this energy window, where Ek−<0E_k^- < 0, the two Fermi–Dirac terms nearly cancel the unity, strongly suppressing the contribution 1/Ck1/C_k to the gap equation. As temperature increases, thermal smearing reduces this cancellation, thereby unblocking these levels and enhancing the phase space available for pnpn pair formation – a mechanism that underlies the reentrance of Δpn\Delta_{pn}.

The progressive unblocking of single-particle levels is directly reflected in the temperature evolution of the level-resolved contributions Δpn(k)\Delta_{pn}^{(k)} to the pnpn pairing gap defined by Δpn=∑k=1ΩΔpn(k)\Delta_{pn}=\sum_{k=1}^{\Omega}\Delta_{pn}^{(k)}. Tracking the redistribution of Δpn(k)\Delta_{pn}^{(k)} across the single-particle spectrum with increasing temperature reveals which levels become accessible for pnpn pair formation and identifies the spectral regions where pairing correlations are most concentrated. Figure 5 displays the contributions Δpn(k)\Delta_{pn}^{(k)} for the same parameter set used in Figure 3(b). Comparative calculations with and without like-nucleon pairing reveal how thermal unblocking and the interplay with pp/nnpp/nn correlations re-shape the spectral distribution of pnpn pairing as a function of temperature. To elucidate the unblocking mechanism, the contributions Δpn(k)\Delta_{pn}^{(k)} are displayed at four characteristic temperatures corresponding to the evolution shown in Figure 3(b): (i) T=0T=0, where pnpn pairing is absent in the Gn,p=0G_{n,p}=0 case; (ii) T=0.129T=0.129, where Δpn\Delta_{pn} takes identical values for Gn,p=0.5G_{n,p}=0.5 and Gn,p=0G_{n,p}=0; (iii) T=0.551T=0.551, where thermal unblocking maximizes Δpn\Delta_{pn} in the absence of like-nucleon pairing; and (iv) T=0.912T=0.912, where Δpn\Delta_{pn} returns to its T=0.129T=0.129 value but with like-nucleon correlations fully quenched.

Figure 5. Level-resolved contributions Δpn(k)\Delta_{pn}^{(k)} to the proton-neutron pairing gap for the parameter set of Figure 3(b). Solid bars show the results with like-nucleon pairing present (Gn,p=0.5G_{n,p}=0.5); dashed bars show the results without like-nucleon pairing (Gn,p=0G_{n,p}=0). The results are shown at four characteristic temperatures illustrating the evolution of thermal unblocking (see main text).

As shown in Figure 5, the level-resolved contributions Δpn(k)\Delta_{pn}^{(k)} differ markedly in calculations with and without like-nucleon pairing at low temperatures. When like-nucleon pairing is present, it modifies the single-particle occupation factors, effectively opening additional channels for pnpn pairing. As a result, the contribution of levels lying between the proton and neutron chemical potentials, predominantly k=6k=6 and k=7k=7 for the present particle numbers, is already significant at T=0T=0. In contrast, in the absence of like-nucleon correlations, these same levels contribute only weakly to Δpn\Delta_{pn} at low temperatures. Instead, thermal excitations must first unblock levels adjacent to the chemical potential window (k≠6,7k\ne6,7), allowing pnpn pairing to develop there before eventually migrating toward k=6,7k=6,7 as temperature increases.

The dual role of like-nucleon pairing becomes evident when comparing the distribution of Δpn(k)\Delta_{pn}^{(k)} at T=0T=0 and T=0.551T=0.551. The observed temperature-dependent crossover from constructive to destructive influence highlights the delicate balance between like-nucleon pairing and thermal effects. Furthermore, the T=0.551T=0.551 plot reveals that the finite-temperature reentrance of pnpn pairing involves contributions from all single-particle levels, although the dominant share of Δpn\Delta_{pn} originates from levels in the vicinity of the chemical potentials. Intriguingly, at T=0.912T=0.912 the distribution Δpn(k)\Delta_{pn}^{(k)} closely resembles that at T=0T=0, despite the fact that the underlying unblocking mechanisms are fundamentally different: at T=0T=0 the enhancement stems from the occupation smearing induced by like-nucleon pairing, whereas at T=0.912T=0.912 it results from the thermal smearing.

As mentioned in the Introduction, path-integral calculations [10, 11, 12] found no enhancement of pnpn pairing at finite temperature. As shown in Figure 2 of Ref. [10], ends_mid=1 when the pnpn pairing gap appears, it exhibits the conventional FT-BCS behavior, i.e., it decreases monotonically with increasing temperature without any reentrant enhancement. To explain this discrepancy with our results, we note that in Refs. [10, 11] the gap equations identical to Eqs. (15) were derived and solved for a schematic one-level model with pppp, nnnn, and pnpn pairing. We stress that, in contrast to the present study, the authors of Refs. [10, 11] considered a system with completely degenerate single-particle levels, i.e., ϵk=0\epsilon_k=0 for all kk.

In the left panel of Figure 6, we show the temperature dependence of the critical strength GpncrG_{pn}^{\mathrm{cr}} for the model parameters used in Refs. [10, 11] (Ω=11\Omega=11, Nn=6N_n=6, Np=4N_p=4, and Gp=Gn=0.242G_p=G_n=0.242). As evident from the plot, for the model with completely degenerate single-particle levels, GpncrG_{pn}^{\mathrm{cr}} does not exhibit any non-monotonic behavior. However, upon introducing a finite level spacing ϵk+1−ϵk=0.6\epsilon_{k+1}-\epsilon_k=0.6, GpncrG_{pn}^{\mathrm{cr}} displays the same non-monotonic temperature dependence as discussed above (see Figure 2). The observed decrease of GpncrG_{pn}^{\mathrm{cr}} for the non-degenerate levels in certain temperature regimes implies that, for a fixed GpnG_{pn}, the system can enhance pnpn pairing correlations. This enhancement is illustrated in the right panel of Figure 6, where we use Gpn=1.2GppG_{pn}=1.2G_{pp}. Note that the present calculations employ a definition of the pairing strength GpnG_{pn} that differs by a factor of two from that in Refs. [10, 11], therefore, for a direct comparison, our results should be matched with the upper right panel (Gpn=0.6GppG_{pn}=0.6G_{pp}) of Fig. 2 in Ref. [10]. Thus, the reason why the path-integral calculations of Refs. [10, 11] did not observe pnpn pairing reentrance is that they employed a model with completely degenerate single-particle levels, which does not capture the thermal unblocking mechanism.

Figure 6. Left panel: Critical pnpn pairing strength GpncrG_{pn}^{\mathrm{cr}} as a function of temperature TT for the completely degenerate model of Ref. [10] and for the same model with identical parameters except for the introduction of non-degenerate single-particle levels, ϵk+1−ϵk=0.6\epsilon_{k+1}-\epsilon_k=0.6. Right panel: Temperature dependence of the proton, neutron and pnpn pairing gaps in the non-degenerate model.

In Ref. [12], path-integral studies were applied to the Richardson model and to realistic nuclei. In both cases, the authors assumed Np=NnN_p=N_n. This proton–neutron symmetry leads to a typical FT-BCS behavior of the pairing gap: it exists at sufficiently large interaction strength GpnG_{pn} and then monotonically decreases with increasing temperature, exhibiting no reentrant behavior.

Realistic nuclei

To investigate finite-temperature pnpn pairing correlations in Ge isotopes, we assume that nucleons move in a deformed axially symmetric Woods–Saxon mean-field potential. Single-particle wave functions and energies are generated using the WSBETA code [27] with the universal parameter set (see Table 1 in [27]). The employed Woods–Saxon potential relaxes the charge symmetry assumption, yielding distinct single-particle energies for protons and neutrons (ϵkp≠ϵkn\epsilon_{kp}\ne\epsilon_{kn}). The particle model space for all nuclei is constructed up to N=6ℏωN=6\hbar\omega. For the description of ground-state deformation, we include only the quadrupole deformation parameter β2\beta_2 taken from Ref. [28], assuming that the deformation is temperature-independent.

β2\beta_2Δpemp\Delta_p^{\mathrm{emp}}Δnemp\Delta_n^{\mathrm{emp}}Δpnemp\Delta_{pn}^{\mathrm{emp}}GpG_pGnG_nGpnG_{pn}
66Ge{}^{66}\mathrm{Ge}0.2081.6071.7910.7780.2450.2420.248
68Ge{}^{68}\mathrm{Ge}−0.2761.5921.8680.6060.2410.2350.254
70Ge{}^{70}\mathrm{Ge}−0.2181.5451.8630.5880.2370.2220.257
72Ge{}^{72}\mathrm{Ge}−0.2171.6161.8350.5850.2390.2150.269

Table 1. Quadrupole deformation parameters, empirical pairing gaps, and pairing strengths for selected even-even Ge isotopes.

The single-particle orbitals are labeled by the Nilsson asymptotic quantum numbers k=(N,nz,Λ,Ω)k=(N,n_z,\Lambda,\Omega). Due to the time-reversal invariance, states with Ω\Omega and −Ω-\Omega are degenerate in energy. Pairing correlations are included only between nucleons in time-reversed states of the same orbital.

To determine the pairing strengths GpG_p, GnG_n, and GpnG_{pn} for a given nucleus, we follow the procedure outlined in Refs. [12, 29, 30]. Specifically, for each of the considered Ge isotopes the zero-temperature theoretical pairing gaps are fitted to the empirical pairing gaps derived from experimental odd-even mass differences [14]

Δpemp=−18[M(Z+2,N)−4M(Z+1,N)+6M(Z,N)−4M(Z−1,N)+M(Z−2,N)],\Delta_p^{\mathrm{emp}}=-\frac{1}{8}\left[M(Z+2,N)-4M(Z+1,N)+6M(Z,N)-4M(Z-1,N)+M(Z-2,N)\right],
Δnemp=−18[M(Z,N+2)−4M(Z,N+1)+6M(Z,N)−4M(Z,N−1)+M(Z,N−2)],\Delta_n^{\mathrm{emp}}=-\frac{1}{8}\left[M(Z,N+2)-4M(Z,N+1)+6M(Z,N)-4M(Z,N-1)+M(Z,N-2)\right],
Δpnemp=14{2[M(Z,N+1)+M(Z,N−1)+M(Z−1,N)+M(Z+1,N)]−[M(Z+1,N+1)+M(Z−1,N+1)+M(Z+1,N−1)+M(Z−1,N−1)]−4M(Z,N)}(24)\Delta_{pn}^{\mathrm{emp}}=\frac{1}{4}\left\{2\left[M(Z,N+1)+M(Z,N-1)+M(Z-1,N)+M(Z+1,N)\right]-\left[M(Z+1,N+1)+M(Z-1,N+1)+M(Z+1,N-1)+M(Z-1,N-1)\right]-4M(Z,N)\right\} \tag*{(24)}

where M(Z,N)M(Z,N) are the nuclear masses [31]. We note that since pnpn pairing correlations influence the like-nucleon pairing gaps and vice versa, all three pairing strengths must be determined self-consistently. The resulting values of GpG_p, GnG_n, and GpnG_{pn}, together with the empirical pairing gaps and ground-state quadrupole deformations β2\beta_2, are summarized in Table I. The table shows that the pnpn pairing gap Δpn\Delta_{pn} at T=0T=0 is considerably smaller than the like-nucleon pairing gaps Δp\Delta_p and Δn\Delta_n. Moreover, Δpn\Delta_{pn} increases with increasing neutron excess, in agreement with the trend reported in Refs. [14, 32].

The temperature dependence of pairing gaps in 66–72Ge{}^{66\text{–}72}\mathrm{Ge} is displayed in Figure 7. One can observe that in the presence of pnpn pairing correlations both Δp\Delta_p and Δn\Delta_n exhibit the characteristic behavior of conventional finite-temperature BCS theory: they remain nearly constant at low temperatures (T≲0.4 MeVT\lesssim0.4\,\mathrm{MeV}), then decrease rapidly and vanish at a critical temperature TcrT_{\mathrm{cr}}. On the other hand, our calculations reveal that the pnpn pairing reentrance phenomenon occurs in all Ge isotopes under consideration, analogous to the schematic case discussed above. While the like-nucleon pairing gaps decrease monotonically with temperature, Δpn\Delta_{pn} exhibits reentrant behavior: after initial plateau, it rises to a maximum at Tcr1T_{\mathrm{cr1}}. Upon further heating, the pnpn pairing is suppressed and Δpn\Delta_{pn} vanishes at the second critical temperature Tcr2>Tcr1T_{\mathrm{cr2}} > T_{\mathrm{cr1}}. Figure 7 also reveals that among the Ge isotopes considered, those with larger neutron excess, corresponding to a greater number of single-particle levels blocked for pnpn pairing at T=0T = 0, exhibit a more pronounced reentrance effect: their Δpn\Delta_{pn} persists to higher temperatures before being completely quenched.

Figure 7. Temperature dependence of the proton, neutron, and proton-neutron pairing gaps in even-even 66−72^{66-72}Ge isotopes.

We now examine the impact of temperature-induced reentrance of pnpn pairing correlations on the strength distribution of charge-exchange transitions in hot nuclei. Thermal modifications of the nuclear charge-exchange response function play a crucial role in determining the rates and cross sections of various weak-interaction processes such as electron/positron capture, β±\beta^{\pm}-decay, and (anti)neutrino absorption that occur in explosive astrophysical environments where nuclei exist at high temperatures (see Ref. [13] and references therein). In the present work, our aim is not to provide quantitatively precise strength functions but rather to perform a qualitative study of the role of pnpn pairing correlations. To this end, we neglect the residual interaction between thermal quasiparticles and restrict our analysis to Fermi-type transitions in nuclei with isovector pppp, nnnn, and pnpn pairing.

For the n→pn \to p (p→np \to n) transition, the Fermi operator reads

F∓=∑k,k′⟨kp(n)∣t∓∣k′n(p)⟩akp(n)†ak′n(p)(25)F_{\mp} = \sum_{k,k'} \langle kp(n)|t_{\mp}|k'n(p)\rangle a^\dagger_{kp(n)}a_{k'n(p)} \tag*{(25)}

where t−t_{-} (t+t_{+}) is the isospin lowering (rising) operator.

ator. In what follows, we adopt the approximation ⟨kp(n)∣t∓∣k′n(p)⟩=δkk′\langle kp(n)|t_{\mp}|k'n(p)\rangle= \delta_{kk'}. Within the independent thermal quasiparticle approximation, F∓F_{\mp} induces transitions from the thermal vacuum to two-quasiparticle states. To compute the corresponding transition probabilities (strengths), we expand the Fermi operators F∓F_{\mp} in terms of thermal quasiparticle creation operators. For the F−F_{-} operator, this yields

F−=−∑k>0,ν,ν′{β~kν†β~kν′†(ukνpvkν′n+ukν′pvkνn)xkνxkν′+β~kν†β~kν′†(vkνpukν′n+vkν′pukνn)ykνykν′+β~kν†β~kν′†(ukνpvkν′n−vkν′pukνn)xkνykν′+β~kν†β~kν′†(vkνpukν′n−ukν′pvkνn)ykνxkν′}+⋯(26)\begin{aligned} F_{-}={}&-\sum_{k>0,\nu,\nu'}\bigl\{\tilde{\beta}^{\dagger}_{k\nu}\tilde{\beta}^{\dagger}_{k\nu'}(u_{k\nu p}v_{k\nu'n}+u_{k\nu'p}v_{k\nu n})x_{k\nu}x_{k\nu'}\\ &+\tilde{\beta}^{\dagger}_{k\nu}\tilde{\beta}^{\dagger}_{k\nu'}(v_{k\nu p}u_{k\nu'n}+v_{k\nu'p}u_{k\nu n})y_{k\nu}y_{k\nu'}\\ &+\tilde{\beta}^{\dagger}_{k\nu}\tilde{\beta}^{\dagger}_{k\nu'}(u_{k\nu p}v_{k\nu'n}-v_{k\nu'p}u_{k\nu n})x_{k\nu}y_{k\nu'}\\ &+\tilde{\beta}^{\dagger}_{k\nu}\tilde{\beta}^{\dagger}_{k\nu'}(v_{k\nu p}u_{k\nu'n}-u_{k\nu'p}v_{k\nu n})y_{k\nu}x_{k\nu'}\bigr\}+\cdots \tag*{(26)} \end{aligned}

Here we omit all terms containing thermal quasiparticle annihilation operators, as they do not contribute to the strength function within the present approximation. The expression for the F+F_{+} operator is obtained by interchanging the isospin indices nn and pp in the above formula.

From the expansion (26) it follows that the transition strength Si(−)=∣⟨i∣F−∣0(T)⟩∣2S_i^{(-)} = |\langle i|F_-|0(T)\rangle|^2 from the thermal vacuum to a particular two-quasiparticle state ∣i⟩|i\rangle is given by the squared modulus of the corresponding coefficient in the expansion. For instance, the n→pn \to p transition strength to the state ∣i⟩=∣βkνβkν′⟩|i\rangle= |\beta_{k\nu}\beta_{k\nu'}\rangle is

Si(−)=(ukνpvkν′n+ukν′pvkνn)2xkν2xkν′2.(27)S_i^{(-)}=(u_{k\nu p}v_{k\nu'n}+u_{k\nu'p}v_{k\nu n})^2x_{k\nu}^2x_{k\nu'}^2. \tag*{(27)}

The transition strengths Si(±)S_i^{(\pm)} exhibit explicit temperature dependence. Notably, transitions to states containing tilde thermal quasiparticles become possible only at finite temperature, when ykν≠0y_{k\nu}\ne0. These correspond to thermally unblocked transitions from nuclear excited states that are forbidden at T=0T=0. As discussed in Appendix A, transition strengths to tilde-conjugated states satisfy the detailed balance condition (A17), ensuring thermodynamic consistency of the strength distribution at finite temperature. The total transition strength satisfies the Ikeda sum rule ∑i(Si(−)−Si(+))=Nn−Np\sum_i(S_i^{(-)}-S_i^{(+)})=N_n-N_p.

Knowing the individual transition strengths, the finite-temperature strength function for charge-exchange transitions can be expressed as a function of transition energy EE

S(∓)(E,T)=∑iSi(∓)δ(E−ωi±Λnp).(28)S^{(\mp)}(E,T)=\sum_i S_i^{(\mp)}\delta(E-\omega_i\pm\Lambda_{np}). \tag*{(28)}

Here Λnp=λn−λp+1.293 MeV\Lambda_{np}=\lambda_n-\lambda_p+1.293\ \mathrm{MeV} accounts for both the difference in neutron and proton chemical potentials and the neutron-proton mass splitting. The excitation energy ωi\omega_i of a two-quasiparticle state ∣i⟩|i\rangle is determined by the energies of the constituent thermal quasiparticles. For instance, for the state ∣i⟩=∣βkνβkν′⟩|i\rangle=|\beta_{k\nu}\beta_{k\nu'}\rangle, one has ωi=Ekν+Ekν′\omega_i=E_{k\nu}+E_{k\nu'}, while for states containing tilde quasiparticles, e.g. ∣i⟩=∣β~kνβ~kν′⟩|i\rangle=|\tilde{\beta}_{k\nu}\tilde{\beta}_{k\nu'}\rangle, the energy reads ωi=−(Ekν+Ekν′)\omega_i=-(E_{k\nu}+E_{k\nu'}). States of mixed type such as ∣βkνβekν′⟩\lvert\beta_{k\nu}\beta_{ek\nu'} \rangle have excitation energies ωi=Ekν−Ekν′\omega_i = E_{k\nu} - E_{k\nu'}. These relations follow directly from the diagonal form of the thermal Hamiltonian (3). At finite temperature, the transition energy EE can take negative values corresponding to thermally unblocked de-excitation processes that are forbidden at zero temperature.

In Figure 8 and Figure 9 we present the strength functions for Fermi p→np \to n and n→pn \to p transitions in 72Ge{}^{72}\mathrm{Ge} calculated without (Gpn=0G_{pn}=0) and with (Gpn=0.269G_{pn}=0.269 MeV) pnpn pairing. The distributions are shown at three characteristic temperatures: at T=0T=0 the pnpn pairing gap is small; at T=0.8T=0.8 MeV the pnpn gap reaches approximately half the sum of the like-nucleon gaps (Δpn≈0.5(Δp+Δn)\Delta_{pn} \approx0.5(\Delta_p+\Delta_n)), indicating that pnpn pairing becomes comparable in strength to like-nucleon pairing; and at T=1.0T=1.0 MeV the like-nucleon pairing vanishes while Δpn\Delta_{pn} approaches its maximum value.

Figure 8. Strength distribution of Fermi p→np \to n transitions in 72Ge{}^{72}\mathrm{Ge} at three temperature values. The upper row corresponds to the absence of pnpn pairing correlations, while the lower row shows the distributions with pnpn pairing included. The strength distributions are obtained by summing the transition strengths within 0.5 MeV energy bins.

Figure 9. Same as Figure 8 but for Fermi n→pn \to p transitions in 72Ge{}^{72}\mathrm{Ge}.

Note that the temperature regime around T≈1T \approx1 MeV is typically encountered in the collapsing core of a massive star during the late silicon-burning and early collapse stages, as well as in the envelope of accreting neutron stars during Type I X-ray bursts, which is the astrophysical site where the rapid proton-capture (rp) process operates.

As it clearly seen in Figure 8, at T=0T=0 the p→np \to n strength distributions obtained with and without pnpn pairing are qualitatively similar: they are concentrated in the energy interval 2.5--5 MeV, but the inclusion of pnpn pairing slightly redistributes strength toward lower energies. At T=0.8T=0.8 MeV, the distribution with pnpn pairing shows enhanced low- and negative-energy strength compared to the Gpn=0G_{pn}=0 case. The fragmentation pattern changes completely at T=1.0T=1.0 MeV, reflecting the increasing influence of pnpn correlations on the thermal quasiparticle spectra and structure. The strength function without pnpn pairing exhibits a pronounced concentration at negative energies in stark contrast to the broad distribution obtained with pnpn pairing. This demonstrates that reentrance of pnpn correlations can qualitatively reshape the p→np \to n charge-exchange response even when conventional pairing is fully quenched. Taking into account that this response governs not only supernova electron capture but also β+\beta^{+}-decay of nuclei formed in the rapid proton capture process, the reentrance effect could modify the neutronization rate of stellar matter and influence the reaction flow along the rprp-process path.

The n→pn \to p strength functions in Figure 9 display an analogous temperature dependence but with notable asymmetries relative to p→np \to n strength, reflecting the sensitivity of the response to the isospin structure of the thermal quasiparticle wave functions. First of all, the overall magnitude is larger due to the neutron excess in 72Ge{}^{72}\mathrm{Ge}. Second, the energy centroid of the n→pn \to p distribution is shifted toward higher excitation energies, a consequence of the negative value of Λnp\Lambda_{np} in 72Ge{}^{72}\mathrm{Ge}. For this reason, the bulk of the thermally unblocked n→pn \to p strength is concentrated at low positive energies, while the strength at negative energies (corresponding to de-excitation processes) remains relatively small. However, like in p→np \to n case, at T=1.0T=1.0 MeV we observe a dramatic effect of pnpn pairing reentrance: the n→pn \to p strength is concentrated in a strong peak at E≈4E \approx4 MeV when pnpn pairing is neglected, whereas the Gpn≠0G_{pn} \ne0 calculation yields low- and high-energy tails.

The present qualitative study underscores the need for further investigations into the role of pnpn pairing in determining stellar weak-interaction observables under extreme astrophysical conditions. A crucial next step in this direction is the inclusion of temperature-dependent deformation, which would require fully self-consistent calculations of the mean-field and pairing correlations at finite temperature. To address the uncertainty introduced by the assumption of a temperature-independent shape, and to demonstrate that a possible shape transition around the temperatures where pnpn-pairing reentrance occurs does not alter our main qualitative conclusions, we performed an additional test calculation. In Figure 10, we show the Fermi strength distributions in 72Ge{}^{72}\mathrm{Ge} at T=1.0T=1.0 MeV computed using a spherically symmetric mean-field (β2=0.0\beta_2=0.0), while keeping all other model parameters identical to the deformed case.

Image not extracted

Figure 10. Same as Figure 8 and Figure 9, but for Fermi transitions in spherical 72Ge{}^{72}\mathrm{Ge} at T=1.0T=1.0 MeV.

As follows from our calculations, even in the spherical mean-field at T=1.0T=1.0 MeV, the like-nucleon pppp and nnnn pairing correlations are completely quenched, whereas the pnpn pairing gap survives and reaches a value comparable to the deformed case (Δpn≈1.3\Delta_{pn}\approx1.3 MeV). More importantly, the inclusion of pnpn pairing produces the same profound qualitative effect on the Fermi strength functions as observed in the deformed nucleus. Specifically, for the n→pn \to p transitions (right panels of Figure 10), the sharp, highly concentrated peak located at E≈5E\approx5 MeV in the absence of pnpn pairing is completely fragmented into a broad distribution with distinct peaks at both lower and higher energies. Similarly, the p→np \to n strength distribution (left panels) is noticeably reshaped and redistributed across the energy spectrum.

These results suggest that the significant impact of pnpn pairing reentrance on the charge-exchange response is likely a general feature, rather than an artifact of a specific deformation assumption. Consequently, while quantitative astrophysical rates will eventually require the inclusion of temperature-dependent deformation (and residual interactions), the qualitative modifications to the strength functions discussed above are expected to persist regardless of shape evolution.

CONCLUSIONS

In the present work, we have developed a generalized finite-temperature proton-neutron BCS (FT-pnBCS) framework for both isovector and isoscalar pairing channels using the superoperator formalism. Numerical calculations for a schematic doubly degenerate equidistant multilevel pairing model as well as for realistic even-even Ge isotopes demonstrate the emergence of pnpn pairing reentrance in N>ZN>Z even-even nuclei with preexisting like-nucleon pairing. This phenomenon is driven by thermal excitations, which reduce Pauli blocking through the partial unblocking of single-particle levels that otherwise suppress pnpn correlations at low temperatures.

The reentrance phenomenon we observe occurs in the temperature regime T≈1T\approx1 MeV, where the mean-field FT-BCS approximation predicts a sharp phase transition and both thermal fluctuations and particle-number projection effects may become non-negligible. However, as discussed in the Introduction, the temperature-induced pnpn pairing reentrance due to unblocking of single-particle levels near the Fermi surface has been demonstrated in odd–odd [3] and rotating [8] N=ZN = Z nuclear systems using particle number conserving approaches that avoid sharp phase transitions entirely. This provides strong evidence that thermally-induced pnpn pairing reentrance is a robust qualitative feature that persists beyond the mean-field approximation, regardless of whether the low-temperature blocking arises from an odd/unpaired nucleon or from the neutron excess. We acknowledge, however, that quantitative aspects, such as the sharpness of the pnpn pairing enhancement and the precise boundaries of the reentrant temperature window, may be modified by a particle-number conserving shell-model Monte-Carlo treatment [33] or by projection techniques (e.g., Lipkin-Nogami [34, 35]). This constitutes an important direction for future investigations.

Another limitation of the FT-pnBCS framework is that it does not allow the simultaneous presence of isoscalar (t=0t = 0) and isovector (t=1t = 1) pairing fields when their coupling strengths are different. The possibility of coexistence of isovector and isoscalar pairing correlations in realistic nuclei remains a subject of ongoing theoretical and experimental debate [2]. Consequently, the specific reentrance behavior reported here should be regarded as a model-dependent prediction. In this respect, an extension of the present FT-pnBCS approach incorporating particle-number projection, similar to that developed at zero temperature [15], would provide a natural framework for investigating the coexistence and competition of isovector and isoscalar pairing correlations at finite temperature. Nevertheless, the reentrance mechanism discussed here is driven primarily by the thermal unblocking of single-particle levels and this mechanism is not specific to the isospin character of the pairing channel. This suggests that the thermal-unblocking mechanism should remain operative even within more general theoretical frameworks, although the quantitative characteristics of the reentrance phenomenon may depend on the treatment of competing pairing channels.

The qualitative analysis of Fermi charge-exchange strength distributions in hot 72^{72}Ge performed within independent thermal quasiparticle approximation motivates further systematic studies to clarify how pnpn pairing correlations modify stellar weak-interaction observables in high-temperature astrophysical environments. Besides the inclusion of temperature-dependent deformation, this will also require taking into account residual interactions, as was done for spherical nuclei at finite temperature within the thermal quasiparticle RPA [13].

The impact of pnpn pairing reentrance on key nuclear observables such as the level density, γ\gamma-ray strength function, and β\beta-decay rates of thermally excited nuclei will be addressed in future works. Furthermore, the present formalism can naturally be extended to hot, rapidly rotating nuclei, thereby incorporating additional collective degrees of freedom.

Acknowledgments

This work was funded by the National Foundation for Science and Technology Development (NAFOSTED) of Vietnam (Grant No. 103.04-2025.22).

Overview of the superoperator formalism

One of the principal technical complications in statistical quantum mechanics arises from the fact that systems are described by the density operator ρ\rho rather than by a state vector. The superoperator formalism offers a mathematical framework that restores a state-vector description for statistical systems. In this Appendix, we briefly discuss details of the formalism that are relevant to our paper, while the comprehensive description is given in [13].

Consider a quantum system described by a Hamiltonian HH acting on the Hilbert space H\mathcal{H}. The mixed state of the system is characterized by the density operator ρ\rho, which can be expanded in an orthonormal basis {∣n⟩}\{|n\rangle\} of H\mathcal{H} as

ρ=∑nmρnm∣n⟩⟨m∣,(29)\rho= \sum_{nm} \rho_{nm}|n\rangle\langle m|, \tag*{(29)}

where ρnm=⟨n∣ρ∣m⟩\rho_{nm} = \langle n|\rho|m\rangle are the matrix elements. The time evolution of the density operator is governed by the Liouville–von Neumann equation:

dρdt=−i[H,ρ].(30)\frac{d\rho}{dt} = -i[H,\rho]. \tag*{(30)}

At thermal equilibrium and finite temperature TT, the mixed state is characterized by the Gibbs density operator

ρ(T)=e−H/TZ(T),(31)\rho(T) = \frac{e^{-H/T}}{Z(T)}, \tag*{(31)}

where the partition function Z(T)=Tr⁡(e−H/T)Z(T) = \operatorname{Tr}(e^{-H/T}) ensures the normalization of the density operator Tr⁡ρ=1\operatorname{Tr}\rho= 1. The time-independent equilibrium density operator ((A3)) is a stationary solution of the Liouville–von Neumann equation and it satisfies the Kubo-Martin-Schwinger (KMS) boundary condition (see, e.g., Sect. 2.2.e in Ref. [36]). Specifically, for any two Heisenberg-picture operators A(t)A(t) and B(t′)B(t'), the equilibrium correlation functions at inverse temperature β=1/T\beta= 1/T obey the relation

⟨⟨A(t)B(t′)⟩⟩=⟨⟨B(t′)A(t+iβ)⟩⟩,(32)\langle\langle A(t) B(t')\rangle\rangle= \langle\langle B(t') A(t+i\beta)\rangle\rangle, \tag*{(32)}

where the double-bracket notation denotes the statistical average with respect to ρ(T)\rho(T):

⟨⟨X⟩⟩≡Tr⁡(ρ(T)X).(33)\langle\langle X\rangle\rangle\equiv\operatorname{Tr}(\rho(T)X). \tag*{(33)}

The cornerstone of the superoperator formalism is the isomorphism L(H)≅H⊗H∗\mathcal{L}(\mathcal{H}) \cong\mathcal{H} \otimes\mathcal{H}^{*} between the Liouville space L\mathcal{L} of linear operators on the Hilbert space H\mathcal{H} and vectors in the extended Hilbert space H⊗H∗\mathcal{H}\otimes\mathcal{H}^{*}.1 Here H∗\mathcal{H}^{*} denotes the dual space to H\mathcal{H}, i.e. it is the space of bras instead of the space of kets. Thus, given an orthonormal basis {∣n⟩}\{|n\rangle\} in the Hilbert space H\mathcal{H}, the operators ∣nm⟩⟩≡∣n⟩⟨m∣|nm\rangle\rangle\equiv|n\rangle\langle m| form a complete orthonormal basis in L(H)\mathcal{L}(\mathcal{H}). This correspondence allows us to represent linear operators on H\mathcal{H} as vectors (pure states) in the associated Liouville space. In particular, the density operator (A1) is mapped to the vector

∣ρ⟩⟩=∑m,nρnm∣nm⟩⟩.(34)|\rho\rangle\rangle= \sum_{m,n}\rho_{nm}|nm\rangle\rangle. \tag*{(34)}

Superoperators are linear transformations acting on the Liouville space. The central object in this formalism is the Liouvillian superoperator L\mathcal{L}, defined by its action on the vectorized density operator: L∣ρ⟩⟩=∣[H,ρ]⟩⟩\mathcal{L}|\rho\rangle\rangle= |[H,\rho]\rangle\rangle. Within the superoperator formalism, the Liouville-von Neumann equation can be cast into a Schrödinger-like form:

ddt∣ρ(t)⟩⟩=−iL∣ρ(t)⟩⟩.(35)\frac{d}{dt}|\rho(t)\rangle\rangle= -i\mathcal{L}|\rho(t)\rangle\rangle. \tag*{(35)}

Consequently, the thermal equilibrium state satisfies the stationary condition

L∣ρ(T)⟩⟩=0.(36)\mathcal{L}|\rho(T)\rangle\rangle= 0. \tag*{(36)}

To construct an explicit representation of the Liouvillian superoperator L\mathcal{L}, one must first specify a basis in the Hilbert space H\mathcal{H}. For the nuclear many-body problem, a natural choice is the set of particle-number eigenstates {∣n⟩}\{|n\rangle\} defined by ak†ak∣n⟩=nk∣n⟩a_k^\dagger a_k|n\rangle= n_k|n\rangle, where ak†a_k^\dagger and aka_k denote, respectively, the particle creation and annihilation operators, and nk∈{0,1}n_k \in\{0,1\} denotes the occupation number of the single-particle state kk. It was shown in Ref. [13] that once the nuclear Hamiltonian H(a⃗†,a⃗)H(\vec{a}^{\dagger},\vec{a}) is given in second-quantized form the Liouvillian superoperator can be expressed as

L=H(a⃗†,a⃗)−H(aˉ†,aˉ),(37)\mathcal{L}=H(\vec{a}^{\dagger},\vec{a})-H(\bar{a}^{\dagger},\bar{a}), \tag*{(37)}

where a⃗†,a⃗\vec{a}^{\dagger},\vec{a} and aˉ†,aˉ\bar{a}^{\dagger},\bar{a} denote the left- and right-action creation and annihilation superoperators, respectively. Their action on the vectorized basis states ∣nm⟩⟩|nm\rangle\rangle is defined by

a⃗k†∣nm⟩⟩≡ak†∣n⟩⟨m∣,aˉk†∣nm⟩⟩≡αnm∣n⟩⟨m∣ak,a⃗k∣nm⟩⟩≡ak∣n⟩⟨m∣,aˉk∣nm⟩⟩≡αnm∣n⟩⟨m∣ak†,(38)\begin{aligned} \vec{a}_{k}^{\dagger}|nm\rangle\rangle&\equiv a_{k}^{\dagger}|n\rangle\langle m|,\qquad&\bar{a}_{k}^{\dagger}|nm\rangle\rangle&\equiv\alpha_{nm}|n\rangle\langle m|a_k,\\ \vec{a}_{k}|nm\rangle\rangle&\equiv a_k|n\rangle\langle m|,\qquad&\bar{a}_{k}|nm\rangle\rangle&\equiv\alpha_{nm}|n\rangle\langle m|a_k^{\dagger}, \tag*{(38)} \end{aligned}

where the phase factor αnm=i(−1)Nn+Nm\alpha_{nm}=i(-1)^{N_n+N_m}, with Nn=∑knkN_n=\sum_k n_k being the total particle number in the state ∣n⟩|n\rangle, ensures the correct fermionic structure of the algebra

generated by the creation and annihilation superoperators in the Liouville space. Thus, within the superoperator framework, the original set of fermionic creation and annihilation operators is effectively doubled, while their canonical anticommutation relations are preserved. This extension of the second quantization to the Liouville space is sometimes referred to in the literature as the third quantization [38].

In what follows, for convenience, we adopt the notation and terminology of Thermo Field Dynamics (TFD) [39].2 In the TFD notation, the left-action superoperators correspond to the physical operators acting on the original (physical) Hilbert space, whereas the right-action superoperators are associated with their counterparts acting on the fictitious (tilde) copy of the Hilbert space. In other words, within the TFD framework, every physical operator X=X(a†,a)X=X(a^{\dagger},a) possesses a tilde counterpart X~=X∗(a~†,a~)\widetilde{X}=X^{*}(\widetilde{a}^{\dagger},\widetilde{a}) obtained from XX by the tilde conjugation operation: each nucleon creation (annihilation) operator is replaced by its tilde partner (a→a~a\to\widetilde{a}, a†→a~†a^{\dagger}\to\widetilde{a}^{\dagger}), and all complex coefficients are complex-conjugated (c→c∗c\to c^{*}).

In TFD, the equilibrium density matrix ρ(T)\rho(T) is encoded as a temperature-dependent vector, referred to as the thermal vacuum ∣0(T)⟩|0(T)\rangle and the statistical average of any observable XX takes the form of a vacuum expectation value:

⟨X⟩=⟨0(T)∣X∣0(T)⟩.(39)\langle X\rangle=\langle0(T)|X|0(T)\rangle. \tag*{(39)}

The Liouvillian superoperator L\mathcal{L} is identified with the thermal Hamiltonian H\mathcal{H}:

H=H−H~,(40)\mathcal{H}=H-\widetilde{H}, \tag*{(40)}

and the stationarity condition (A8) for thermal equilibrium is equivalently expressed as

H∣0(T)⟩=0.(41)\mathcal{H}|0(T)\rangle=0. \tag*{(41)}

The KMS condition (A4) finds its counterpart in TFD notation as the thermal-state condition (TSC):

X∣0(T)⟩=σXeH/(2T)X~†∣0(T)⟩,(42)X|0(T)\rangle=\sigma_X e^{\mathcal{H}/(2T)}\widetilde{X}^{\dagger}|0(T)\rangle, \tag*{(42)}

where the phase factor σX\sigma_X equals +1+1 for bosonic-like operators and −i-i for fermionic-like operators.3 Thus, within the superoperator formalism, the construction of the thermal vacuum and the evaluation of statistical averages is reduced to finding the zero-eigenvalue eigenstate of the thermal Hamiltonian H\mathcal{H} that satisfies the thermal-state condition (A14). Eigenstates of H\mathcal{H} with nonzero eigenvalues describe non-equilibrium states of the system at finite temperature induced by external perturbations. This interpretation follows from the fact that H\mathcal{H} generates time evolution in the Liouville space. The presence of two-body interactions in H\mathcal{H} generally precludes exact diagonalization; however, an approximate diagonal form can be achieved within a suitable many-body scheme:

H≈∑iωi(Oi†Oi−O~i†O~i).(43)\mathcal{H} \approx\sum_i \omega_i \left(\mathcal{O}_i^\dagger\mathcal{O}_i-\widetilde{\mathcal{O}}_i^\dagger\widetilde{\mathcal{O}}_i\right). \tag*{(43)}

Note that at finite temperature both excitation and de-excitation processes become physically accessible, corresponding respectively to the positive- and negative-energy eigenmodes of H\mathcal{H}.

Once the thermal Hamiltonian is diagonalized within a given approximation, it is natural, by a direct analogy with the zero-temperature case, to define the corresponding (approximate) thermal vacuum ∣0(T)⟩\lvert0(T)\rangle as a vacuum state for the annihilation superoperators:

Oi∣0(T)⟩=0,O~i∣0(T)⟩=0.(44)\mathcal{O}_i\lvert0(T)\rangle=0,\qquad\widetilde{\mathcal{O}}_i\lvert0(T)\rangle=0. \tag*{(44)}

The thermal-state condition (A14) then constrains not only the equilibrium state but also the structure of non-equilibrium excitations. In particular, it enforces a detailed-balance relation between excitation and de-excitation amplitudes: for any transition operator T\mathcal{T},

∣⟨O~i∣T∣0(T)⟩∣2=e−ωi/T∣⟨Oi∣T†∣0(T)⟩∣2.(45)\left\lvert\langle\widetilde{\mathcal{O}}_i\lvert\mathcal{T}\rvert0(T)\rangle\right\rvert^2=e^{-\omega_i/T}\left\lvert\langle\mathcal{O}_i\lvert\mathcal{T}^{\dagger}\rvert0(T)\rangle\right\rvert^2. \tag*{(45)}

In this way, the superoperator formalism offers a thermodynamically consistent framework for describing both equilibrium and non-equilibrium properties of atomic nuclei at finite temperature within a suitable many-body approximation. Following the conceptual framework of the quasiparticle-phonon nuclear model [41], Ref. [13] provides a detailed description of three successive many-body schemes for diagonalizing the thermal Hamiltonian: (i) thermal quasiparticles, (ii) thermal phonons, and (iii) the coupling between thermal quasiparticles and thermal phonons.

References

References

  1. [1]R. A. Broglia and V. Zelevinsky, eds., Fifty Years of Nuclear BCS. (World Scientific, 2013).DOI
  2. [2]S. Frauendorff and A. Macchiavelli, Overview of neutron-proton pairing, Progress in Particle and Nuclear Physics 78, 24 (2014).
  3. [3]K. Kanelko and M. Hasegawa, Proton-neutron pairing energies in N = Z nuclei at finite temperature, Physical Review C 72, 031302 (2005).
  4. [4]A. Goswami, Treatment of neutron-proton correlations, Nuclear Physics 60, 228 (1964).DOI
  5. [5]H. Chen and A. Goswami, Generalized treatment of neutron-proton pairing in N = Z nuclei, Physics Letters B 24, 257 (1967).DOI
  6. [6]A. Goodman, G. Struble, and A. Goswami, Restoration of axial symmetry of the equilibrium shape of 24Mg by pairing correlations, Physics Letters B 26, 10 (1968).
  7. [7]A. Goodman, Generalized gap equations and the coherence of the α-α pair field, Nuclear Physics A 186, 475 (1972).
  8. [8]J. A. Sheikh, R. Palit, and S. Frauendorff, Reappearance of the pairing correlations at finite temperature, Phys. Rev. C 72, 041301(R) (2005).DOI
  9. [9]A. Sedrakian, T. Alm, and U. Lombardo, Superfluidity in asymmetric nuclear matter, Phys. Rev. C 55, R582 (1997).DOI
  10. [10]M. Fellah, N. H. Allal, M. Belabbas, M. R. Oudih, and N. Benhamouda, Temperature-dependent isovector pairing gap equations using a path integral approach, Phys. Rev. C 76, 047306 (2007).
  11. [11]D. Mokhtari, N. H. Allal, and M. Fellah, Isovector plus isoscalar neutron-proton pairing effect on nuclear statistical quantities using a path integral approach, International Journal of Modern Physics E 27, 1850054 (2018).
  12. [12]M. Fellah, N. H. Allal, and M. R. Oudih, Neutron-proton isovector pairing correlations treatment in heated nuclei within the path integral formalism, Chinese Physics C 49, 084101 (2025).
  13. [13]A. A. Dzhioev and A. I. Vdovin, Superoperator approach to the theory of hot nuclei and astrophysical applications: I–Spectral properties of hot nuclei, Physics of Particles and Nuclei 53, 885 (2022); Superoperator approach to the theory of hot nuclei and astrophysical applications: II–Electron capture in stars, Physics of Particles and Nuclei 53, 939 (2022); Superoperator approach to the theory of hot nuclei and astrophysical applications: III–Neutrino-nucleus reactions in stars, Physics of Particles and Nuclei 53, 1051 (2022).
  14. [14]F. Šimkovic, C. C. Moustakidis, L. Pacaerescu, and A. Faessler, Proton-neutron pairing in the deformed BCS approach, Physical Review C 68, 054319 (2003).DOI
  15. [15]W. Satula and R. Wyss, Competition between T = 0 and T = 1 pairing in proton-rich nuclei, Physics Letters B 393, 1 (1997).
  16. [16]H. Wolter, A. Faessler, and P. Sauer, HFB calculations with T = 1 and T = 0 pairing correlations, Nuclear Physics A 167, 108 (1971).
  17. [17]L. Moretto, Statistical description of deformation in excited nuclei and disappearance of shell effects with excitation energy, Nuclear Physics A 182, 641 (1972).
  18. [18]Rogue Wave Software, IMSL Numerical Library.
  19. [19]O. Civitarese, M. Reboiro, and P. Vogel, Neutron-proton pairing in the BCS approach, Physical Review C 56, 1840 (1997).DOI
  20. [20]W. Satula and R. Wyss, A number projected model with generalized pairing interaction, Nuclear Physics A 676, 120 (2000).DOI
  21. [21]N. Q. Hung, N. D. Dang, and L. T. Q. Huong, Improved treatment of blocking effect at finite temperature, Phys. Rev. C 94, 024341 (2016).
  22. [22]T. V. Dong, L. T. Phuc, N. Q. Hung, N. D. Dang, T. D. Huyen, and N. A. Tuan, Pairing reentrance in odd nuclei at finite temperature, Physical Review C 107, 064319 (2023).
  23. [23]N. Q. Hung and N. D. Dang, Pairing in hot rotating nuclei, Phys. Rev. C 78, 064315 (2008).DOI
  24. [24]N. Q. Hung and N. D. Dang, Pairing reentrance in hot rotating nuclei, Phys. Rev. C 84, 054324 (2011).DOI
  25. [25]N. Q. Hung, N. D. Dang, and L. G. Moretto, Pairing in excited nuclei: a review, Reports on Progress in Physics 82, 056301 (2019).DOI
  26. [26]V. G. Soloviev, Effect of superconducting pairing correlations on nuclear properties, in Selected topics in nuclear theory, edited by F. Janouch (International Atomic Energy Agency, Vienna, 1963) pp. 233–289.
  27. [27]C. Cwiok, J. Dudek, W. Nazarewicz, J. Skalski, and T. Werner, Single-particle energies, wave functions, quadrupole moments and g-factors in an axially deformed Woods-Saxon potential with applications to the two-centre-type nuclear problems, Computer Physics Communications 46, 379 (1987).DOI
  28. [28]P. Möller, A. Sierk, T. Ichikawa, and H. Sagawa, Nuclear ground-state masses and deformations: FRDM(2012), Atomic Data and Nuclear Data Tables 109-110, 1 (2016).
  29. [29]A. Baran and K. Sieja, Neutron–proton pairing in 64Ge, International Journal of Modern Physics E 14, 445 (2005).
  30. [30]X. Lian, C. L. Bai, H. Sagawa, and H. Q. Zhang, Proton-neutron pairing correlations in N = Z nuclei with Deformed Skyrme+pnBCS model (2025), arXiv:2507.18124 [nucl-th].arxiv.org/abs/2507.18124
  31. [31]W. Huang, M. Wang, F. G. Kondev, G. Audi, and S. Naimi, The AME 2020 atomic mass evaluation (i). Evaluation of input data, and adjustment procedures, Chinese Physics C 45, 030002 (2021).DOI
  32. [32]O. Civitarese and M. Reboiro, Proton-neutron pairing effects in medium and heavy mass nuclei, Physical Review C 56, 1179 (1997).DOI
  33. [33]S. Koonin, D. Dean, and K. Langanke, Shell model Monte Carlo methods, Physics Reports 278, 1 (1997).
  34. [34]D. Kosov and A. I. Vdovin, Approximate number projection at finite temperature, Zeitschrift für Physik A Hadrons and Nuclei 355, 17 (1996).DOI
  35. [35]N. D. Dang, P. Ring, and R. Rossignoli, Lipkin-Nogami method at finite temperature in the static-path approximation, Phys. Rev. C 47, 606 (1993).pubmed.ncbi.nlm.nih.gov/9968477
  36. [36]G. Emch, Algebraic methods in statistical mechanics and quantum field theory (Wiley, New York, 1974).
  37. [37]M. Frembs and E. G. Cavalcanti, Variations on the Choi-Jamiolkowski isomorphism, Journal of Physics A: Mathematical and Theoretical 57, 265301 (2024).
  38. [38]T. Prosen, Third quantization: a general method to solve master equations for quadratic open fermi systems, New Journal of Physics 10, 043026 (2008).DOI
  39. [39]Y. Takahashi and H. Umezawa, Thermo field dynamics, International journal of modern Physics B 10, 1755 (1996).DOI
  40. [40]M. Schmutz, Real-time Green’s functions in many body problems, Zeitschrift für Physik B Condensed Matter 30, 97 (1978).DOI
  41. [41]V. G. Soloviev, Theory of atomic nuclei: quasiparticles and phonons (Institute of Physics Publishing, Bristol and Philadelphia, 1992).

Paper details

vlnc:W_dbsux796
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