betaFeedback

WorksPhysics and Astronomy

Low-energy enhancement in the magnetic dipole radiation of actinide nuclei

Abstract

We present the first theoretical results of the magnetic dipole (M1) γγ-ray strength function (γγSF) for actinide nuclei within the shell-model Monte Carlo (SMMC) method. We observe a low-energy enhancement (LEE) in the M1 γγSFs of the six nuclei studied here, which serves as the first evidence, theoretical or experimental, that the LEE persists in the actinides. We also identify a scissors mode resonance in all six nuclei, which we compare with recent Oslo-method experiments.

LEE_actinides.5.pdf

Low-energy enhancement in the magnetic dipole radiation of actinide nuclei

Abstract

Introduction

The γ\gamma-ray strength function (γSF\gamma\mathrm{SF}) of a nucleus, which describes the probability of γ\gamma-ray emission and absorption, is a critical input to neutron-capture rate calculations within the Hauser-Feshbach theory [1]. These reaction rates are important for determining elemental abundances from rr-process nucleosynthesis [2]. Furthermore, structures seen in γSF\gamma\mathrm{SF}s provide insight into the collective behavior of nucleons [3]. Thus, accurate modeling of nuclear γSF\gamma\mathrm{SF}s is important for developing more precise predictions of elemental abundances in stars as well as improving our understanding of nuclear structure.

The giant dipole resonance, which is an electric dipole (E1) collective mode [4], dominates the γSF\gamma\mathrm{SF} at higher γ\gamma-ray energies. Nuclei also exhibit resonances at lower energies (below the neutron separation energy SnS_n), some of which are magnetic dipole (M1) in nature. The most prominent M1 modes are the scissors mode resonance (SR) [3, 5, 6], seen in deformed nuclei, and the spin-flip mode [3]. Accounting for these low-energy structures in the γSF\gamma\mathrm{SF} of neutron-rich nuclei significantly improves calculations of neutron radiative capture rates [7].

Most experimental measurements of the γSF\gamma\mathrm{SF} rely on either nuclear resonance fluorescence (NRF) reactions [8] or the Oslo method [9], with the latter allowing for simultaneous determination of both the γSF\gamma\mathrm{SF} and the nuclear level density. Experiments on light- and medium-mass nuclei observed a then-unexpected ‘upbend’ in the γSF\gamma\mathrm{SF} at the lowest γ\gamma-ray energies [10, 11, 12, 13, 14], now known as the low-energy enhancement (LEE). Configuration-interaction (CI) shell-model calculations suggested that the LEE is magnetic dipole (M1) in nature [15, 16, 17, 18, 19, 20].

More recently, a LEE has been observed experimentally [21, 22, 23, 24] and calculated theoretically [25, 26, 27] in the M1 γSF\gamma\mathrm{SF} in chains of neodymium and samarium isotopes. Due to the large shell model spaces associated with these heavy open-shell nuclei, CI shell-model calculations are only possible using the shell-model Monte Carlo (SMMC) method [28, 29, 30, 31, 32, 33]. The SMMC has been used extensively for calculating properties of lanthanide nuclei at finite temperature [34, 35, 36, 37]. Recently, the SMMC has been extended to calculate the level densities of actinide nuclei [38].

The possible existence of a LEE in actinide nuclei is especially important, as neutron capture rates in these actinides are relevant not only for the astrophysical rr-process, but also in processes occurring inside of nuclear reactors [39]. If the LEE persists in neutron-rich heavy nuclei, this would have profound effects on rr-process nucleosynthesis through the enhancement of neutron-capture rates near the neutron drip line [40].

While various experimental measurements were made of the γSF\gamma\mathrm{SF} in actinide nuclei, many of these experiments utilized NRF reactions [41, 42, 43, 44]. The LEE has been shown to only exist in the γSF\gamma\mathrm{SF} built on top of excited states. Thus, it cannot be measured in NRF experiments. A LEE can in principle be observed in Oslo-method experiments measuring the decay of excited nuclei. Oslo-method experiments, however, are currently limited to studying actinides down to γ\gamma-ray energies of Eγ≳1 MeVE_\gamma\gtrsim1\,\mathrm{MeV} [45, 46, 47, 48, 49, 50]. While this is sufficient for observing the scissors mode, probing the LEE may require measurements of the γSF\gamma\mathrm{SF} at even lower γ\gamma-ray energies. Thus, the existence of a LEE in the actinides has yet to be experimentally confirmed.

Here we present the first SMMC calculations of the M1 γSF\gamma\mathrm{SF} at the neutron-separation energy of six actinides that have recently been studied with the Oslo method [46, 47, 50]: 232Th{}^{232}\mathrm{Th}, 237–239U{}^{237\text{–}239}\mathrm{U}, 240Pu{}^{240}\mathrm{Pu} and 243Pu{}^{243}\mathrm{Pu}. We clearly identify a LEE and observe both the scissors and spin-flip modes in all six nuclei.

We first provide an overview of the computational methods we use to calculate the imaginary-time M1 response function and its conversion into a γSF\gamma\mathrm{SF} using the maximum entropy method [25, 26, 27]. Then we present results of our calculations in the above six actinide nuclei and compare them with experimental data [46, 47, 50].

Strength and response functions

For a nucleus at temperature TT, we define the strength function of the M1 spherical vector operator O^M1\hat{O}_{M1} as

SM1(T;ω)=∑αi,Ji;αf,Jfe−βEαi,JiZ∣⟨αfJf∣O^M1∣αiJi⟩∣2×δ(Eαf,Jf−Eαi,Ji−ω)(1)\begin{aligned} S_{M1}(T;\omega)=\sum_{\alpha_i,J_i;\alpha_f,J_f}\frac{e^{-\beta E_{\alpha_i,J_i}}}{Z}\left|\langle\alpha_fJ_f|\hat{O}_{M1}|\alpha_iJ_i\rangle\right|^2 \\ &\times\delta(E_{\alpha_f,J_f}-E_{\alpha_i,J_i}-\omega) \tag*{(1)} \end{aligned}

where Z=∑αi,Ji(2Ji+1)e−βEαi,JiZ=\sum_{\alpha_i,J_i}(2J_i+1)e^{-\beta E_{\alpha_i,J_i}} is the canonical partition function and O^M1\hat{O}_{M1} is the magnetic dipole transition operator defined by

O^M1=34πμNℏ(gll^+gss^),(2)\hat{O}_{M1}=\sqrt{\frac{3}{4\pi}}\frac{\mu_N}{\hbar}\left(g_l\hat{l}+g_s\hat{s}\right), \tag*{(2)}

where l^\hat{l} and s^\hat{s} are, respectively, the orbital and spin angular momentum operators. We use the free-nucleon gg-factors gl,p=1g_{l,p}=1, gl,n=0g_{l,n}=0, gs,p=5.5857g_{s,p}=5.5857, and gs,n=−3.8263g_{s,n}=-3.8263.

Oslo-method experiments do not measure directly the strength function but rather the de-excitation γ\gammaSF. For M1 transitions, we denote the corresponding de-excitation γ\gammaSF by fM1f_{M1}; it is defined by [51]

fM1(Ei;Eγ)=ρ~(Ei)Eγ−3⟨ΓM1(Ei;Eγ)⟩(3)f_{M1}(E_i;E_\gamma)=\tilde{\rho}(E_i)E_\gamma^{-3}\left\langle\Gamma_{M1}(E_i;E_\gamma)\right\rangle \tag*{(3)}

where ⟨ΓM1(Ei;Eγ)⟩\left\langle\Gamma_{M1}(E_i;E_\gamma)\right\rangle is the average partial width for a nucleus at initial energy EiE_i to emit M1 radiation with energy EγE_\gamma, and ρ~\tilde{\rho} is the average level density.

The de-excitation strength function fM1f_{M1} is related to the strength function SM1S_{M1} by [25, 27]

fM1(Ei;Eγ)≈aρ~(Ei)3ρ~(Ef)SˉM1(T;ω=−Eγ)(4)f_{M1}(E_i;E_\gamma)\approx a\frac{\tilde{\rho}(E_i)}{3\tilde{\rho}(E_f)}\bar{S}_{M1}(T;\omega=-E_\gamma) \tag*{(4)}

where SˉM1(T)\bar{S}_{M1}(T) is the smoothed analog of Eq. (1), obtained by integrating over a finite energy resolution Δω\Delta\omega, and a=16π9(ℏc)3a=\frac{16\pi}{9(\hbar c)^3}.

The γ\gammaSF cannot be directly calculated using the SMMC. Instead, we must calculate the imaginary-time response function

RM1(T;τ)=⟨O^M1(τ)⋅O^M1(0)⟩(5)R_{M1}(T;\tau)=\left\langle\hat{O}_{M1}(\tau)\cdot\hat{O}_{M1}(0)\right\rangle \tag*{(5)}

where ⟨…⟩\langle\ldots\rangle denotes a thermal average at temperature TT, O^M1(τ)⋅O^M1(0)=∑μ(−)μO^M1μ(τ)O^M1−μ(0)\hat{O}_{M1}(\tau)\cdot\hat{O}_{M1}(0)=\sum_{\mu}(-)^\mu\hat{O}_{M1}^{\mu}(\tau)\hat{O}_{M1}^{-\mu}(0) and O^M1μ(τ)=eτH^O^M1μe−τH^\hat{O}_{M1}^{\mu}(\tau)=e^{\tau\hat{H}}\hat{O}_{M1}^{\mu}e^{-\tau\hat{H}} (H^\hat{H} is the nuclear Hamiltonian).

The strength function in Eq. (1) is related to the response function with

RM1(T;τ)=∫0∞dωK(τ,ω)SM1(T;ω)(6)R_{M1}(T;\tau)=\int_{0}^{\infty}d\omega K(\tau,\omega)S_{M1}(T;\omega) \tag*{(6)}

where K(τ,ω)=e−τω+e−(β−τ)ωK(\tau,\omega)=e^{-\tau\omega}+e^{-(\beta-\tau)\omega} and we have used the symmetry relation SO^(T;−ω)=e−βωSO^(T;ω)S_{\hat{O}}(T;-\omega)=e^{-\beta\omega}S_{\hat{O}}(T;\omega). Calculating the M1 γ\gammaSF thus amounts to inverting Eq. (6).

Computational Methods.— We calculate the response function (5) in the SMMC method [33], which is based on the Hubbard-Stratonovich transformation [52, 53] to express the imaginary-time propagator e−βH^e^{-\beta\hat{H}} as a superposition of one-body propagators describing nucleons moving in external axillary fields σα\sigma_\alpha. We use the canonical ensemble of fixed numbers of protons and neutrons by applying particle-number projection [54]. The projection on an odd number of protons or neutrons results in a sign problem at low temperatures. However, near the neutron resonance energy the Monte-Carlo sign remains close to 1 [27, 55].

In the SMMC, the imaginary-time response function (5) is given by

RM1(T;τ)=∫D[σ]WσΦσ⟨O^M1(τ)⋅O^M1(0)⟩σ∫D[σ]D[σ]WσΦσ(7)R_{M1}(T;\tau)=\frac{\int D[\sigma]W_\sigma\Phi_\sigma\left\langle\hat{O}_{M1}(\tau)\cdot\hat{O}_{M1}(0)\right\rangle_\sigma}{\int D[\sigma]D[\sigma]W_\sigma\Phi_\sigma} \tag*{(7)}

where O^M1(τ)⋅O^M1(0)=∑μ(−)μO^M1μ(τ)O^M1−μ(0)\hat{O}_{M1}(\tau)\cdot\hat{O}_{M1}(0)=\sum_{\mu}(-)^\mu\hat{O}_{M1}^{\mu}(\tau)\hat{O}_{M1}^{-\mu}(0) and O^M1(τ)=U^σ−1(τ)O^M1U^σ(τ)\hat{O}_{M1}(\tau)=\hat{U}^{-1}_\sigma(\tau)\hat{O}_{M1}\hat{U}_\sigma(\tau) with U^σ(τ,0)\hat{U}_\sigma(\tau,0) being the propagator between times 0 and τ\tau for a given sample σ\sigma of the auxiliary fields. The expectation value ⟨…⟩σ\langle\ldots\rangle_\sigma is taken with respect to U^σ(β,0)\hat{U}_\sigma(\beta,0), Wσ=Gσ∣Tr⁡U^σ(β,0)∣W_\sigma=G_\sigma|\operatorname{Tr}\hat{U}_\sigma(\beta,0)| is a positive-definite weight (with GσG_\sigma a Gaussian factor), and Φσ=Tr⁡U^σ(β,0)/∣Tr⁡U^σ(β,0)∣\Phi_\sigma=\operatorname{Tr}\hat{U}_\sigma(\beta,0)/|\operatorname{Tr}\hat{U}_\sigma(\beta,0)| is the Monte Carlo sign. The integration in Eq. (7) is carried out using a Metropolis Monte Carlo algorithm [33].

For our calculations of the actinides, we use a valence space comprised of the 0h9/20h_{9/2}, 1f7/21f_{7/2}, 1f5/21f_{5/2}, 2p3/22p_{3/2}, 2p1/22p_{1/2}, 0i13/20i_{13/2} (82-126 shell) plus 1g9/21g_{9/2} orbitals for protons and 1g9/21g_{9/2}, 2d5/22d_{5/2}, 0i11/20i_{11/2}, 1g7/21g_{7/2}, 3s1/23s_{1/2}, 2d3/22d_{3/2}, 0j15/20j_{15/2} (126-184 shell) plus the 1h11/21h_{11/2} orbitals for neutrons. These were determined by considering orbital occupation numbers, as in previous SMMC analyses [34].

The effective residual interactions, which consist of monopole pairing and multipole terms (quadrupole, octupole and hexadecupole), are the same as those used in Ref. [38].

We compute the imaginary-time response function RM1(T;τ)R_{M1}(T;\tau) via Eq. (7). Inverting Eq. (6) to determine the strength function SM1(T;ω)S_{M1}(T;\omega) is equivalent to numerically inverting a bilateral Laplace transform, which is known to be an ill-posed problem. While a number of methods exist to perform this task [56], we use the maximum entropy method (MEM) [57, 58, 59], which has been used previously in SMMC calculations of strength functions of heavy nuclei [25, 26, 27].

The MEM is a Bayesian approach to maximizing the objective function

Q(SM1,α)=αS−12χ2(8)Q(S_{M1},\alpha)=\alpha\mathcal{S}-\frac{1}{2}\chi^2 \tag*{(8)}

with respect to SM1S_{M1}. Here, χ2\chi^2 is the likelihood function given by

χ2=(RM1SMMC−RM1)TC−1(RM1SMMC−RM1)(9)\chi^2=(R_{M1}^{\mathrm{SMMC}}-R_{M1})^{T}C^{-1}(R_{M1}^{\mathrm{SMMC}}-R_{M1}) \tag*{(9)}

where CC is the covariance matrix of the SMMC response function data RM1SMMCR_{M1}^{\mathrm{SMMC}}, and RM1R_{M1} is calculated using Eq. (6) for a given choice of SM1S_{M1}. The entropy function S\mathcal{S} in Eq. (8) is defined by

S=∫dω(SM1−SM1prior−SM1log⁡(SM1/SM1prior))(10)\mathcal{S}=\int d\omega\left(S_{M1}-S_{M1}^{\mathrm{prior}}-S_{M1}\log\left(S_{M1}/S_{M1}^{\mathrm{prior}}\right)\right) \tag*{(10)}

where SM1priorS_{M1}^{\mathrm{prior}} is a prior distribution of the strength function.

To account for the parameter α\alpha, we use Bryan’s method [60] which averages the solutions SM1S_{M1} that maximize Eq. (8) over all values of α\alpha.

The success of the MEM is depends on a suitable choice of the prior distribution. For this analysis, we will take our prior to be the static-phase approximation (SPA) [61, 62, 63, 64, 65], as previous theoretical studies of the lanthanides demonstrated that the SPA provides a good approximation for the M1 response function at temperatures near the neutron separation energy [25, 26, 27]. This choice for the prior strength function is confirmed for actinides in the Supplemental Material by comparing its imaginary-time response function with the SMMC response function [66].

Image not extracted

Figure 1. The M1 strength functions SM1S_{M1} for six actinides at temperatures near their neutron separation energy. The blue solid lines are the SMMC+MEM results with shaded blue bands representing the uncertainties, and the red dashed lines are the SPA M1 strength functions used as the prior.

Results for actinides.— We calculated the M1M1 strength functions SM1S_{M1} of six actinide nuclei: three even-even 232Th^{232}\mathrm{Th}, 238U^{238}\mathrm{U}, and 240Pu^{240}\mathrm{Pu} and three odd-even 237U^{237}\mathrm{U}, 239U^{239}\mathrm{U}, and 243Pu^{243}\mathrm{Pu} near their neutron separation energy. The results are shown in Fig. 1. For both the SMMC and SPA calculations we used 8,000 uncorrelated samples, and uncertainties were calculated using jackknife resampling.

We clearly see a large peak at ω=0\omega= 0 in all six nuclei, which we identify as the LEE (see below). We also identify two more peaks at ω≈1\omega\approx1 MeV and ω≈3.5\omega\approx3.5 MeV. We interpret these, respectively, as the scissors mode and spin-flip mode built on top of excited states. In the Supplemental Material, we also calculate the M1M1 γ\gammaSF of the six actinides at low temperatures close to the ground state [66] and observe that the LEE peak is missing. Both the scissors mode and spin-flip modes are observed in the M1M1 strength function computed near the ground state and show more structure when compared with the M1M1 strength function computed near the neutron resonance energy.

We convert SM1S_{M1} into de-excitation strength functions fM1f_{M1} using Eq. (4), and the results are shown in Fig. 2. The level densities ρ~\tilde{\rho} in Eq. (4) were calculated with the SMMC as in Ref. [38].

As suggested by Refs. [15, 19] we fit the LEE to an exponential form

fM1LEE(Eγ)=C0e−κEγ(11)f_{M1}^{LEE}(E_\gamma) = C_0 e^{-\kappa E_\gamma} \tag*{(11)}

The LEE fit parameters C0C_0 and κ\kappa for all six actinide nuclei care listed in Table I.

Nucleus232Th{}^{232}\mathrm{Th}237U{}^{237}\mathrm{U}238U{}^{238}\mathrm{U}239U{}^{239}\mathrm{U}240Pu{}^{240}\mathrm{Pu}243Pu{}^{243}\mathrm{Pu}
C0C_0 (10−7 MeV−310^{-7}\ \mathrm{MeV}^{-3})1.48 (1)1.38 (6)1.32 (6)1.28 (7)1.26 (1)1.19 (7)
κ\kappa (MeV−1\mathrm{MeV}^{-1})2.38 (29)1.66 (18)1.79 (14)2.10 (22)1.82 (16)2.19 (28)

Table I. Parameters for the fits of the M1M1 γ\gammaSF to the form in Eq. (11).

In Fig. 3 we show the fitted values of κ\kappa as a function of the initial energy EiE_i. We find that κ\kappa depends weakly on EiE_i. This is consistent with previous results from SMMC [25, 26, 27] and CI shell-model [19] calculations in lanthanide and mid-mass nuclei, respectively.

Subtracting the LEE fit (11) from our fM1f_{M1} results allows us to isolate the other modes in the M1 strength. In particular, we fit the SR in the residual strength to a Lorentzian form [46]

fM1SR(Eγ)=13π2(ℏc)2σEγΓ2(Eγ2−ω2)2+Eγ2Γ2(12)f_{M1}^{SR}(E_\gamma) = \frac{1}{3\pi^2(\hbar c)^2}\frac{\sigma E_\gamma\Gamma^2}{(E_\gamma^2-\omega^2)^2+E_\gamma^2\Gamma^2} \tag*{(12)}

where σ\sigma, ω\omega, and Γ\Gamma are, respectively, the strength, center, and width of the SR.

We note that the SRs measured in Ref. [46, 47] are actually comprised of two lorentzians, whereas in our calculations we see only a single peak (a possible reason for this discrepancy will be discussed below). Thus, rather than comparing the fit parameters, we compare only the total integrated strength

B(M1)=27(ℏc)316π∫0SnfM1(Eγ) dEγ.(13)B(M1) = \frac{27(\hbar c)^3}{16\pi}\int_0^{S_n} f_{M1}(E_\gamma)\,dE_\gamma. \tag*{(13)}
Image not extracted

Figure 2. The de-excitation strength functions fM1f_{M1} calculated from Eq. (4) for the six actinides in fig. 1, calculated near their neutron separation energies. The blue solid lines are the SMMC+MEM results with shaded blue bands representing the uncertainties, and the red dashed lines describe the SPA priors.

Image not extracted

Figure 3. The slope parameter κ\kappa, obtained by fitting the LEE to Eq. (11), as a function of the initial energy EiE_i for all six actinide nuclei.

Our calculated integrated strengths and the experimentally measured values from [46, 47] are tabulated in Table II. For completeness we also include the parameters of our Lorentzian fits (12). To assess the accuracy of our theoretical values for B(M1)B(M1) when compared to the experimental measurements, we note several key points. First, the experimental measurements do not distinguish the parity of the γ\gamma-rays, namely the measured strengths contain contri butions from both E1 and M1 radiation. While the experimental results account for the presence of the E1 giant dipole resonance by subtracting its extrapolated tail, theoretical calculations have suggested that thermal excitations in the continuum may contribute to an enhancement of low-energy E1 transitions [67]. Thus there might be residual E1 radiation in the experimental SR results.

Nucleusω\omega (MeV)σ\sigma (mb)Γ\Gamma (MeV)B(M1)B(M1) (μN2\mu_N^2)B(M1)B(M1) (μN2\mu_N^2)
SMMC + MEMExperiment
232Th{}^{232}\mathrm{Th}1.02 (7)0.258 (11)1.20 (26)10.7 (25)9.5 (26)
237U{}^{237}\mathrm{U}1.26 (4)0.346 (18)1.14 (17)11.1 (18)8.8 (15)
238U{}^{238}\mathrm{U}1.26 (4)0.352 (14)1.14 (18)11.3 (19)9.4 (16)
239U{}^{239}\mathrm{U}1.32 (5)0.362 (14)1.37 (21)13.2 (25)8.8 (14)
240Pu{}^{240}\mathrm{Pu}1.30 (3)0.504 (16)1.40 (14)19.2 (21)-
243Pu{}^{243}\mathrm{Pu}1.31 (5)0.557 (17)1.48 (19)22.2 (30)10.1 (15)

Table II. Fit parameters of the M1M1 SR fitted to the Lorentzian form of Eq. (12) and their integrated strengths (13) compared to experiments [46, 47].

Second, our calculations are performed at a fixed temperature (canonical ensemble) whereas the experiment is performed at a fixed initial energy EiE_i (microcanonical ensemble). Finally, the integrated strength in our calculations is sensitive to the γ\gamma-ray energy window used in our lorentzian fit, leading to additional systematic uncertainties in the fit parameters.

Accounting for these factors, we conclude that the integrated strengths of our scissors modes are in reasonable agreement with those measured by Refs. [46, 47].

The experimental measurements of the SR for the actinides considered here contain two peaks, rather than the single peak we observe in our calculations. Further, the experimental SR peaks are located at a higher energy than our theoretical calculations suggest, as was observed in previous SMMC calculations of the lanthanides [27].

It has been suggested that triaxiality can lead to a splitting of M1 modes by an amount proportional to sin⁡γ\sin\gamma [68]. Furthermore, because the SR is the counter-rotation of the deformed proton and neutron densities, the energy of the SR mode is positively correlated to the β\beta and γ\gamma deformation of the respective nucleus.

In previous work studying nuclear deformation in the lanthanides [69] it was found that in the restricted model space used a phenomenological factor accounting for core polarization must be included to generate sufficient deformation in the SMMC. Thus, the SMMC calculations presented here underpredict the deformation of the six actinide nuclei, which could potentially result in the SR having a lower energy than expected and the two SR peaks being too closely split to be resolved.

On the other hand, since the LEE depends only weakly on deformation [26], we expect our results for the LEE to be robust.

Summary and discussion.— We have performed SMMC calculations of the M1 γ\gammaSF for the actinide nuclei 232Th{}^{232}\mathrm{Th}, 237–239U{}^{237\text{--}239}\mathrm{U}, 240Pu{}^{240}\mathrm{Pu}, and 243Pu{}^{243}\mathrm{Pu} near their neutron separation energy. To carry out the necessary analytic continuation we used the MEM, with a prior strength function calculated in the SPA.

We observed a LEE in all six actinides. These results are the first theoretical prediction of a LEE in the M1 γ\gammaSF of actinides. We also observe a scissors mode in all six actinides.

Improvements to the results presented here could be made by studying the spin dependence of the γ\gammaSF. This would allow us to more directly compare our theoretically calculated fM1f_{M1} to experimental measurements, which populate spins differently from the intrinsic spin distribution of the nucleus.

Acknowledgments.— We thank P. Fanto for the use of the SPA and MEM codes he developed. This work was supported in part by the U.S. DOE grant No. DE-SC0019521. The calculations used resources of the National Energy Research Scientific Computing Center (NERSC), a U.S. Department of Energy Office of Science User Facility operated under Contract No. DE-AC0205CH11231.

The data that support the findings of this article are openly available [70].

References

Supplemental.3.pdf

Supplemental Material: Low-energy enhancement in the magnetic dipole radiation of actinide nuclei

M1 SF at low temperatures (near the ground state)

In this section we compare the magnetic dipole strength function at the neutron resonance energy (discussed in the main text) with the corresponding strength function at low temperatures near the ground state.

In Fig. 1 we show the M1 γ\gammaSF SM1S_{M1} for the six actinide nuclei studied in this work: 232Th^{232}\mathrm{Th}, 237U^{237}\mathrm{U}, 238U^{238}\mathrm{U}, 239U^{239}\mathrm{U}, 240Pu^{240}\mathrm{Pu}, and 243Pu^{243}\mathrm{Pu}. For the even-mass nuclei (top row), we use T=0.025T = 0.025 MeV. However, the odd-mass nuclei have a Monte Carlo sign problem at low temperatures and for these nuclei we use T=0.167T = 0.167 MeV (bottom row) for which the sign problem is still moderate [1].

As discussed in the main text, the γ\gammaSFs are calculated using the maximum entropy method (MEM). For very low temperatures we find that the quasiparticle random phase approximation (QRPA) provides a better approximation (see Sec. B below) than the static phase approximation (SPA). Thus for the even-mass nuclei we choose for the prior strength to be the QRPA M1 strength function (shown by the dashed orange lines in Fig. 1). For the odd-mass nuclei where the temperature is not so low, we use as prior the SPA strength function (dashed red lines in Fig. 1). The uncertainties were calculated using jackknife resampling.

Comparing with the M1 γ\gammaSF at the neutron resonance energy, we find that the LEE is absent for the M1 γ\gammaSF near the ground state. On the other hand, the scissors mode at ∼1.5\sim1.5 MeV and the spin-flip model at ∼3.5\sim3.5 MeV are more pronounced and show more structure.

Image not extracted

Figure 1 (PDF p. 8). The M1 γ\gammaSF for temperatures near the ground state for the even-mass (top row) and odd-mass (bottom row) actinides. The solid blue lines correspond to the MEM fit result and the blue bands describe the associated statistical uncertainties. The dashed orange lines are the QRPA prior strength functions for the even-mass actinides and the dashed red lines are the SPA prior strength functions for the odd-mass actinides.

Imaginary-time response functions

Using the SMMC, we calculate the exact (up to statistical error) imaginary-time M1 response function. We then carry out the numerical analytic continuation in the MEM using a suitable prior strength function. The success of the method depends on a good choice for this prior. In particular, we consider the SPA and QRPA M1 strength functions as possible priors.

To assess the suitability of different priors, we compare their imaginary-time response functions to the SMMC response. In Fig. 2 we compare the SMMC imaginary-time M1 response functions (green circles) for the six actinides at energies near the neutron separation energy to the SPA response (dashed red lines) and the QRPA response (dashed orange lines) functions. The latter is calculated for the even-mass nuclei only using the code HF-SHELL1 [2]. We find out that the SPA response function is much closer to the SMMC response function than the QRPA. We therefore use the SPA M1M1 strength function as the prior for temperatures that correspond to the neutron separation energies (for both the even- and odd-mass nuclei).

As is seen in Fig. 1 of the main text, the LEE is already observed in the SPA M1M1 strength function. This suggests that the large amplitude fluctuations around the mean-field configuration, which are included in the SPA, are important for generating the LEE.

In Fig. 3 we show the M1M1 imaginary-time response functions at low temperatures: T=0.025T = 0.025 MeV for the even-mass actinides and T=0.167T = 0.167 MeV for the odd-mass actinides. The SMMC results (green circles) are compared with the SPA (dashed red lines) and the QRPA (dashed orange lines) response functions. Based on the results for the even-mass actinides, we conclude that at very low temperatures (close to the ground state), the QRPA results, which account for small amplitude quantal fluctuations about the mean-field configuration are closer to the SMMC results. Therefore, in the MEM for very low temperatures, we use the QRPA strength as the prior strength function.

Image not extracted

Figure 2 (PDF p. 9). The finite temperature imaginary-time M1M1 response function for all six actinide nuclei at temperatures near the neutron separation energy. The green circles are the SMMC results, and the solid blue lines are the response functions resulting from the MEM fit. The dashed red lines are the SPA response functions and the dashed orange lines are the QRPA response functions. For all nuclei we use the SPA strength as the prior for the MEM fit.

References

Image not extracted

Figure 3 (PDF p. 10). The imaginary-time M1 response function for the six actinides at temperatures close to the ground state (T=0.025T = 0.025 MeV for the even-mass-nuclei and T=0.167T = 0.167 MeV for the odd-mass nuclei). The symbols and various lines are as in Fig. 2, except that the QRPA strength is used as a prior for the even-mass nuclei in the top row (while the SPA strength is used as a prior for the odd-mass nuclei in the bottom row).

References

  1. [1]W. Hauser and H. Feshbach, The inelastic scattering of neutrons, Phys. Rev. 87, 366 (1952).DOI
  2. [2]M. R. Mumpower, R. Surman, G. McLaughlin, and A. Aprahamian, The impact of individual nuclear properties on r-process nucleosynthesis, Prog. Part. Nucl. Phys. 86, 68 (2016).DOI
  3. [3]K. Heyde, P. Von Neumann-Cosel, and A. Richter, Magnetic dipole excitations in nuclei: Elementary modes of nucleonic motion, Rev. Mod. Phys. 82, 2365 (2010).DOI
  4. [4]M. Goldhaber and E. Teller, On nuclear dipole vibrations, Phys. Rev. 74, 1046 (1948).DOI
  5. [5]N. Lo Iudice and F. Palumbo, Positive parity isovector collective states in deformed nuclei, Nucl. Phys. A 326, 193 (1979).
  6. [6]D. Bohle, A. Richter, W. Steffen, A. E. L. Dieperink, N. L. Iudice, F. Palumbo, and O. Scholten, New magnetic dipole excitation mode studied in the heavy deformed nucleus 156Gd by inelastic electron scattering, Phys. Lett. B 137, 27 (1984).DOI
  7. [7]M. R. Mumpower, T. Kawano, J. L. Ullmann, M. Krtička, and T. M. Sprouse, Estimation of m 1 scissors mode strength for deformed nuclei in the medium- to heavy-mass region by statistical hauser-feshbach model calculations, Phys. Rev. C 96, 024612 (2017).DOI
  8. [8]U. E. P. Berg and U. Kneissl, Recent progress on nuclear magnetic dipole excitations, Annu. Rev. Nucl. Part. Sci. 37, 33 (1987).
  9. [9]A. Schiller, L. Bergholt, M. Guttormsen, E. Melby, J. Rekstad, and S. Siem, Extraction of level density and γ strength function from primary γ spectra, Nucl. Instrum. Methods Phys. Res. A 447, 498 (2000).DOI
  10. [10]A. Voinov, E. Algin, U. Agvaanluvsan, T. Belgya, R. Chankova, M. Guttormsen, G. E. Mitchell, J. Rekstad, A. Schiller, and S. Siem, Large enhancement of radiative strength for soft transitions in the quasicontinuum, Phys. Rev. Lett. 93, 124504 (2004).DOI
  11. [11]M. Guttormsen, R. Chankova, U. Agvaanluvsan, E. Algin, L. A. Bernstein, F. Ingebretsen, T. Lönnroth, S. Messelt, G. E. Mitchell, J. Rekstad, et al., Radiative strength functions in 93–98Mo, Phys. Rev. C 71, 044307 (2005).DOI
  12. [12]M. Wiedeking, L. A. Bernstein, M. Krtička, D. L. Bleuel, J. M. Allmond, M. S. Basunia, J. T. Harke, P. Fallon, R. B. Firestone, et al., Low-energy enhancement in the photon strength of 95Mo, Phys. Rev. Lett. 108, 162503 (2012).pubmed.ncbi.nlm.nih.gov/22680714
  13. [13]A. C. Larsen, N. Blasi, A. Bracco, F. Camera, T. K. Eriksen, A. Görgen, M. Guttormsen, T. W. Hagen, S. Leoni, B. Million, et al., Evidence for the dipole nature of the low-energy γ enhancement in 56Fe, Phys. Rev. Lett. 111, 242504 (2013).pubmed.ncbi.nlm.nih.gov/24483649
  14. [14]A. C. Larsen, J. E. Midtbø, M. Guttormsen, T. Renstrøm, S. N. Liddick, A. Spyrou, S. Karampagia, B. A. Brown, O. Achakovsky, S. Kamerdzhiev, et al., Enhanced low-energy γ-decay strength of 70Ni and its robustness within the shell model, Phys. Rev. C 97, 054329 (2018).DOI
  15. [15]R. Schwengner, S. Frauendorf, and A. C. Larsen, Low-energy enhancement of magnetic dipole radiation, Phys. Rev. Lett. 111, 232504 (2013).DOI
  16. [16]B. A. Brown and A. C. Larsen, Large low-energy M1 strength for 56,57Fe within the nuclear shell model, Phys. Rev. Lett. 113, 252502 (2014).DOI
  17. [17]R. Schwengner, S. Frauendorf, and B. A. Brown, Low-energy magnetic dipole radiation in open-shell nuclei, Phys. Rev. Lett. 118, 092502 (2017).DOI
  18. [18]K. Sieja, Electric and magnetic dipole strength at low energy, Phys. Rev. Lett. 119, 052502 (2017).DOI
  19. [19]S. Karampagia, B. A. Brown, and V. Zelevinsky, Low energy magnetic radiation enhancement in the f7/2 shell, Phys. Rev. C 95, 024322 (2017).DOI
  20. [20]K. Sieja, Shell-model study of the M1 dipole strength at low energy in the A > 100 nuclei, Phys. Rev. C 98, 064312 (2018).DOI
  21. [21]B. Kheswa, M. Wiedeking, F. Giacoppo, S. Goriely, M. Guttormsen, A. Larsen, F. Bello Garrote, T. Eriksen, A. Görgen, T. Hagen, et al., Galactic production of 138La: Experimental and theoretical statistical properties, Physics Letters B 744, 268 (2015).
  22. [22]F. Naqvi, A. Simon, M. Guttormsen, R. Schwengner, S. Frauendorf, C. S. Reingold, J. T. Burke, N. Cooper, R. O. Hughes, S. Ota, and A. Saastamoinen, Nuclear level densities and γ-ray strength functions in samarium isotopes, Phys. Rev. C 99, 054331 (2019).DOI
  23. [23]A. Simon, M. Guttormsen, A. C. Larsen, C. W. Beausang, P. Humby, J. T. Burke, R. J. Casperson, R. O. Hughes, T. J. Ross, J. M. Allmond, et al., First observation of low-energy γ-ray enhancement in the rare-earth region, Phys. Rev. C 93, 034303 (2016).DOI
  24. [24]M. Guttormsen, K. O. Ay, M. Ozgur, E. Algin, A. C. Larsen, F. L. Bello Garrote, H. C. Berg, L. Cre-spo Campo, T. Dahl-Jacobsen, F. W. Furmyr, et al., Evolution of the γ-ray strength function in neodymium isotopes, Phys. Rev. C 106, 034314 (2022).DOI
  25. [25]P. Fanto and Y. Alhassid, Low-energy enhancement in the magnetic dipole γ-ray strength functions of heavy nuclei, Phys. Rev. C 109, L031302 (2024).DOI
  26. [26]A. Mercenne, P. Fanto, W. Ryssens, and Y. Alhassid, Magnetic dipole γ-ray strength functions in the crossover from spherical to deformed neodymium isotopes, Phys. Rev. C 110, 054313 (2024).arxiv.org/abs/2407.06161
  27. [27]D. DeMartini and Y. Alhassid, Low-energy enhancement of the magnetic dipole radiation in odd-mass lanthanides, Phys. Rev. C 111, 034315 (2025).arxiv.org/abs/2410.00109
  28. [28]C. W. Johnson, S. E. Koonin, G. H. Lang, and W. E. Ormand, Monte carlo methods for the nuclear shell model, Phys. Rev. Lett. 69, 3157 (1992).DOI
  29. [29]G. H. Lang, C. W. Johnson, S. E. Koonin, and W. E. Ormand, Monte carlo evaluation of path integrals for theDOI
  30. [30]Y. Alhassid, D. J. Dean, S. E. Koonin, G. Lang, and W. E. Ormand, Practical solution to the monte carlo sign problem: Realistic calculations of 54Fe, Phys. Rev. Lett. 72, 613 (1994).DOI
  31. [31]S. E. Koonin, D. J. Dean, and K. Langanke, Shell model monte carlo methods, Phys. Rep. 278, 1 (1997).DOI
  32. [32]Y. Alhassid, Quantum monte carlo methods for nuclei at finite temperature, Int. J. Mod. Phys. B 15, 1447 (2001).DOI
  33. [33]Y. Alhassid, Auxiliary-field quantum monte carlo methods in nuclei, in Emergent Phenomena in Atomic Nuclei from Large-Scale Modeling (World Scientific, 2017) p. 267–298.arxiv.org/abs/1607.01870
  34. [34]Y. Alhassid, L. Fang, and H. Nakada, Heavy deformed nuclei in the shell model monte carlo method, Phys. Rev. Lett. 101, 082501 (2008).DOI
  35. [35]C. Öżen, Y. Alhassid, and H. Nakada, Crossover from vibrational to rotational collectivity in heavy nuclei in the shell-model monte carlo approach, Phys. Rev. Lett. 110, 042502 (2013).DOI
  36. [36]Y. Alhassid, C. Gilbreth, and G. Bertsch, Nuclear deformation at finite temperature, Phys. Rev. Lett. 113, 262503 (2014).DOI
  37. [37]M. Guttormsen, Y. Alhassid, W. Ryssens, K. Ay, M. Ozgur, E. Algin, A. Larsen, F. Bello Garrote, L. Crespo Campo, T. Dahl-Jacobsen, et al., Strong enhancement of level densities in the crossover from spherical to deformed neodymium isotopes, Phys. Lett. B 816, 136206 (2021).DOI
  38. [38]D. DeMartini and Y. Alhassid, Nuclear state and level densities of actinides with the shell-model Monte Carlo, (2025), arXiv:2509.26571 [nucl-th].arxiv.org/abs/2509.26571
  39. [39]M. B. Chadwick, M. W. Herman, P. Oblozinsky, M. E. Dunn, Y. Danon, A. Kahler, D. L. Smith, B. P. Ritchychenko, G. Arbanas, R. Arcilla, et al., ENDF/B-VII.1 nuclear data for science and technology: Cross sections, covari-ances, fission product yields and decay data, Nucl. Data Sheets 112, 2887 (2011).DOI
  40. [40]A. C. Larsen and S. Goriely, Impact of a low-energy enhancement in the γ-ray strength function on the neutron-capture cross section, Phys. Rev. C 82, 014318 (2010).DOI
  41. [41]A. S. Adekola, C. T. Angell, S. L. Hammond, A. Hill, C. R. Holder, H. J. Karwowski, J. H. Kelley, and E. Kwan, Discovery of low-lying E1 and M1 strengths in 232Th, Phys. Rev. C 83, 034615 (2011).DOI
  42. [42]R. Heil, H. Pitz, U. Berg, U. Kneissl, K. Hummel, G. Kilgus, D. Bohle, A. Richter, C. Wesselborg, and P. Von Brentano, Observation of orbital magnetic dipole strength in the actinide nuclei 232Th and 238U, Nucl. Phys. A 476, 39 (1988).DOI
  43. [43]J. Margraf, A. Degener, H. Friedrichs, R. D. Heil, A. Jung, U. Kneissl, S. Lindenstruth, H. H. Pitz, H. Schacht, U. Seemann, et al., Photoexcitation of low-lying dipole transitions in 236U, Phys. Rev. C 42, 771 (1990).pubmed.ncbi.nlm.nih.gov/9966770
  44. [44]O. Yevetska, J. Enders, M. Fritzsche, P. von Neumann-Cosel, S. Oberstedt, A. Richter, C. Romig, D. Savran, and K. Sonnabend, Dipole strength in the 235U(γ,γ′) reaction up to 2.8 MeV, Phys. Rev. C 81, 044309 (2010).
  45. [45]M. Guttormsen, L. A. Bernstein, A. Bürger, A. Görgen, F. Gunsing, T. W. Hagen, A. C. Larsen, T. Renstrøm, S. Siem, M. Wiedeking, and J. N. Wilson, Observation of large scissors resonance strength in actinides, Phys. Rev. Lett. 109, 162503 (2012).
  46. [46]M. Guttormsen, L. A. Bernstein, A. Görgen, B. Jurado, S. Siem, M. Aiche, Q. Ducasse, F. Giacoppo, F. Gunsing, T. W. Hagen, et al., Scissors resonance in the quasicon-tinuum of Th, Pa, and U isotopes, Phys. Rev. C 89, 014302 (2014).DOI
  47. [47]T. A. Laplace, F. Zeiser, M. Guttormsen, A. C. Larsen, D. L. Bleuel, L. A. Bernstein, B. L. Goldblum, S. Siem, F. L. B. Garrote, J. A. Brown, et al., Statistical properties of ²⁴³Pu , and ²⁴²Pu(n,γ) cross section calculation, Phys. Rev. C 93, 014323 (2016).
  48. [48]T. G. Tornyi, M. Guttormsen, T. K. Eriksen, A. Görgen, F. Giacoppo, T. W. Hagen, A. Krasznahorkay, A. C. Larsen, T. Renstrøm, S. J. Rose, S. Siem, and G. M. Tveten, Level density and γ-ray strength function in the odd-odd ²³⁸Np nucleus, Phys. Rev. C 89, 044323 (2014).DOI
  49. [49]F. Bello Garrote, A. Lopez-Martens, A. Larsen, I. Deloncle, S. Perú, F. Zeiser, P. Greenlees, B. Kheswa, K. Auranen, D. Bleuel, et al., Experimental observation of the M1 scissors mode in ²⁵⁴No, Phys. Lett. B 834, 137479 (2022).DOI
  50. [50]F. Zeiser, G. M. Tveten, G. Potel, A. C. Larsen, M. Guttormsen, T. A. Laplace, S. Siem, D. L. Bleuel, B. L. Goldblum, L. A. Bernstein, et al., Restricted spin-range correction in the oslo method: The example of nuclear level density and γ-ray strength function from ²³⁹Pu(d,pγ)²⁴⁰Pu, Phys. Rev. C 100, 024305 (2019).
  51. [51]G. A. Bartholomew, E. D. Earle, A. J. Ferguson, J. W. Knowles, and M. A. Lone, Gamma-ray strength functions, in Advances in Nuclear Physics, edited by M. Baranger and E. Vogt (Springer US, Boston, MA, 1973) p. 229–324.
  52. [52]R. L. Stratonovich, On a Method of Calculating Quantum Distribution Functions, Sov. Phys. Dokl. 2, 416 (1957).
  53. [53]J. Hubbard, Calculation of partition functions, Phys. Rev. Lett. 3, 77 (1959).DOI
  54. [54]W. E. Ormand, D. J. Dean, C. W. Johnson, G. H. Lang, and S. E. Koonin, Demonstration of the auxiliary-field monte carlo approach for sd-shell nuclei, Phys. Rev. C 49, 1422 (1994).DOI
  55. [55]Y. Alhassid, P. Fanto, and C. Özen, Circumventing the odd-particle-number sign problem in the shell model Monte Carlo method, Phys. Rev. C 110, L061303 (2024).arxiv.org/abs/2308.14698
  56. [56]R.-A. Tripolt, P. Gubler, M. Ulybyshev, and L. Von Smekal, Numerical analytic continuation of euclidean data, Comput. Phys. Commun. 237, 129 (2019).DOI
  57. [57]J. E. Gubernatis, M. Jarrell, R. N. Silver, and D. S. Sivia, Quantum monte carlo simulations and maximum entropy: Dynamics from imaginary-time data, Phys. Rev. B 44, 6011 (1991).pubmed.ncbi.nlm.nih.gov/9998463
  58. [58]M. Jarrell and J. Gubernatis, Bayesian inference and the analytic continuation of imaginary-time quantum monte carlo data, Phys. Rep. 269, 133 (1996).
  59. [59]J. E. Gubernatis and P. Werner, Quantum Monte Carlo methods : algorithms for lattice models (Cambridge University Press, Cambridge :, 2016).
  60. [60]R. K. Bryan, Maximum entropy analysis of oversampled data problems, Eur. Biophys. J. 18, 165 (1990).DOI
  61. [61]G. Puddu, P. Bortignon, and R. Broglia, The rpa-spa approximation to level densities, Ann. Phys. 206, 409 (1991).
  62. [62]B. Lauritzen, P. Arve, and G. F. Bertsch, Nuclear level densities in the static-path approximation, Phys. Rev. Lett. 61, 2835 (1988).pubmed.ncbi.nlm.nih.gov/10039240
  63. [63]Y. Alhassid and B. Bush, Nuclear level densities in the static-path approximation: (i). a solvable model, Nucl. Phys. A 549, 43 (1992).
  64. [64]R. Rossignoli and P. Ring, The RPA strength function in the presence of thermal fluctuations, Nucl. Phys. A 633, 613 (1998).DOI
  65. [65]R. Rossignoli, N. Canosa, and P. Ring, Microscopic description of hot nuclei: the SPA+RPA approach, Nucl. Phys. A 654, 719c (1999).DOI
  66. [66]See the Supplemental Material accompanying this article.
  67. [67]E. Litvinova and N. Belov, Low-energy limit of the radiative dipole strength in nuclei, Phys. Rev. C 88, 031302 (2013).DOI
  68. [68]E. Lipparini and S. Stringari, Sum rules and giant resonances in nuclei, Phys. Rep. 175, 103 (1989).DOI
  69. [69]M. T. Mustonen, C. N. Gilbreth, Y. Alhassid, and G. F. Bertsch, Statistical theory of deformation distributions in nuclear spectra, Phys. Rev. C 98, 034317 (2018).DOI
  70. [70]https://doi.org/10.5281/zenodo.20148823.

Paper details

vlnc:W_63zrseyp
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 5

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