betaFeedback

WorksPhysics and Astronomy

229m^{229m}Th isomer from a nuclear model perspective

Abstract

The physical conditions for the emergence of the extremely low-lying nuclear isomer 229m^{229m}Th at approximately 8 eV are investigated in the framework of our recently proposed nuclear structure model. Our theoretical approach explains the 229m^{229m}Th-isomer phenomenon as the result of a very fine interplay between collective quadrupole-octupole and single-particle dynamics in the nucleus. We find that the isomeric state can only appear in a rather limited model space of quadrupole-octupole deformations in the single-particle potential, with the octupole deformation being of a crucial importance for its formation. Within this deformation space the model-described quantities exhibit a rather smooth behaviour close to the line of isomer-ground state quasi-degeneracy determined by the crossing of the corresponding single-particle orbitals. Our comprehensive analysis confirms the previous model predictions for reduced transition probabilities and the isomer magnetic moment, while showing a possibility for limited variation in the ground-state magnetic moment theoretical value. These findings prove the reliability of the model and suggest that the same dynamical mechanism could manifest in other actinide nuclei giving a general prescription for the search and exploration of similar isomer phenomena.

I Introduction

Well supporting the current strong emphasis on interdisciplinary research, a unique extremely low-lying 229mTh^{229m}\mathrm{Th} isomer at approximately 8 eV [1, 2, 3, 4] obviously disregards the recognized low-energy border of nuclear physics firmly stepping on atomic physics territory. Although another low-lying nuclear excitation in 235U^{235}\mathrm{U} also approaches this limit with an order of magnitude larger energy of 76 eV [5], currently 229mTh^{229m}\mathrm{Th} attracts much more interest since its energy lies in the range of accessibility of present vacuum ultraviolet (VUV) lasers capable to handle the wavelength of 150 nm (≈8\approx8 eV). With a relatively narrow width and excellent stability, this transition appears to be of a practical interest for a diverse community beyond nuclear physics, involving atomic, laser, plasma physics, metrology, cosmology and others, posing a number of puzzling problems and raising hopes for possible advanced applications. The main interest is related to a new frequency standard based on laser access and stabilization of this transition with sufficient accuracy through contemporary laser (frequency comb or other) techniques. This has been often referred to in the literature as a “nuclear clock” [6, 7, 8]. Such a nuclear clock is expected to have a better or at least comparable accuracy to the currently developed atomic clocks. This entails a rich variety of possible 229mTh^{229m}\mathrm{Th}-based applications such as the precise determination of temporal variations in fundamental constants [9, 10, 11, 12], the development of nuclear lasers in the VUV range [13], detection improvements in satellite and deep space navigation, gravitation waves, geodesy, precise analysis of chemical environment and others.

Towards the aforementioned applications, recent experiments have confirmed the existence of the isomer [14] and have determined the isomer mean half-life in neutral Th atoms [15]. Furthermore, the magnetic dipole moment μ\mu of the nuclear isomeric state (IS) was determined for the first time through laser spectroscopy experiments [16, 17] providing the value μIS=−0.37(6)μN\mu_{\mathrm{IS}}=-0.37(6)\mu_{\mathrm{N}}. Then, three very recent experiments proposed newly updated values for the isomer energy, EIS=8.28(17)E_{\mathrm{IS}}=8.28(17) eV [3] (from internal conversion electron spectroscopy), EIS=8.30(92)E_{\mathrm{IS}}=8.30(92) eV [4] (by determining the transition rates and energies from the above level at 29.2 keV) and EIS=8.10(17)E_{\mathrm{IS}}=8.10(17) eV [18] (from a micro-calorimetric determination of absolute γ\gamma-ray energy differences).

These advances, although not yet reaching the accuracy needed for a nuclear clock, pose new challenges and inspire new studies of the 229mTh^{229m}\mathrm{Th} problem from the nuclear structure side. 229Th^{229}\mathrm{Th} belongs to the light actinide nuclear mass region known for the presence of enhanced collectivity and shape dynamic properties suggesting a complicated interaction between the collective motion of the even-even core and the individual motion of the single neutron. The single-particle (s.p.) states of the latter determine the 229Th^{229}\mathrm{Th} ground state (GS) with Kπ=5/2+K^\pi=5/2^{+} and the IS with Kπ=3/2+K^\pi=3/2^{+} based on the 5/2[633]5/2[633] and 3/2[631]3/2[631] s.p. orbitals. Here, π\pi denotes the parity and KK refers to the projection of the total nuclear angular momentum on the body-fixed principal symmetry axis of the system, respectively. We use the usual Nilsson notation K[NnzΛ]K[Nn_z\Lambda] with NN, nzn_z and Λ\Lambda being the asymptotic Nilsson quantum numbers [19]. Although it is intuitively clear that the entire nuclear structure dynamics should essentially influence the appearance and the properties of the isomer, only limited work has addressed this aspect in the past. Thus, predictions for the B(M1)B(M1) and B(E2)B(E2) reduced 3/2+→5/2+3/2^{+} \to5/2^{+} transition probabilities have been made in Refs. [20, 21] using the quasiparticle-plus-phonon model [22] without particular focus on the isomer properties. Furthermore, in Refs. [23, 24] estimates were made for the isomer B(M1)B(M1) transition rate using the Alaga branching ratios [25], and for the IS magnetic moment μIS\mu_{\mathrm{IS}} based on the Nilsson model [26]. The obtained value μIS=−0.076 μN\mu_{\mathrm{IS}}=-0.076\,\mu_N essentially differs from the recently available experimental value of −0.37(6) μN-0.37(6)\,\mu_N [16].

Understanding the physical mechanism behind the 229mTh{}^{229m}\mathrm{Th} phenomenon requires a thorough investigation of the interplay of all involved collective and s.p. degrees of freedom, and identification of all structure effects which could allow the appearance of an excitation in the eV energy scale. Since the latter is beyond reach for the accuracy of nuclear models generally speaking, the implementation of such a task would require the application of a sophisticated theoretical method which can provide the necessary conclusion by juxtaposing results and information gained from different perspectives and observables such as energies, transition probabilities and magnetic moments. Motivated by the considerations above we have recently put forward a complete nuclear-structure model approach that takes into account the axial quadrupole-octupole (QO) (pear-shape) deformation modes typical for the nuclei in the actinide region both in the collective and s.p. degrees of freedom of the nucleus [27]. The formalism involves in the even-even nuclear core the so-called coherent QO model, describing collective axial quadrupole and octupole vibrations with equal oscillation frequencies non-adiabatically coupled to the rotation motion [28, 29, 30, 31], while the odd-nucleon motion is described within a deformed shell model (DSM) including reflection-asymmetric Woods-Saxon potential [32] and pairing correlations of Bardeen-Cooper-Schrieffer (BCS) type with blocking of the unpaired nucleon orbital [33]. The odd-nucleon motion is coupled to the collective motion by a Coriolis interaction taken into account through perturbation theory. The model spectrum has the form of quasi-parity-doublet bands built on the ground and excited quasi-particle (q.p.) states. In this scheme the IS appears as a q.p. band head of an excited quasi-parity-doublet. The model framework allows a rather complete and intrinsically consistent spectroscopic treatment of the nucleus including its IS.

Based on this model description we were able in Ref. [27] to predict the B(M1)B(M1) and B(E2)B(E2) reduced probabilities for the IS 3/2+→5/2+3/2^{+} \to5/2^{+} transition. For B(M1)B(M1) we have provided the limits of 0.006−0.0080.006-0.008 Weisskopf units (W.u.), well below the earlier deduced values of 0.0480.048 W.u. [23, 24] and 0.0140.014 W.u. [21], corroborating the experimental difficulties to observe radiative isomer decay [34, 35, 36]. For the electric quadrupole B(E2)B(E2) transition probability we have determined the limits of B(E2)=20−30B(E2)=20-30 W.u. At the same time the energy spectrum and several available data on other transition rates were described with reasonable accuracy.

In the subsequent work [37] we have calculated the magnetic moment of the IS, μIS\mu_{\mathrm{IS}} and of the GS, μGS\mu_{\mathrm{GS}}, by taking into account attenuation effects in the spin and collective gyromagnetic factors, without changing the model parameters originally adjusted in Ref. [27]. The result for μIS\mu_{\mathrm{IS}} in the range from μIS=−0.25μN\mu_{\mathrm{IS}}=-0.25\mu_N to −0.35μN-0.35\mu_N is in rather good agreement with the recent experimental values (−0.3)−(−0.4)(-0.3)-(-0.4) [17] and −0.37(6)-0.37(6) [16]. On the other hand, μGS\mu_{\mathrm{GS}} was obtained in the range μGS=0.53−0.66 μN\mu_{\mathrm{GS}}=0.53-0.66\,\mu_N, overestimating the latest reported and older experimental values of 0.360(7) μN0.360(7)\,\mu_N [38] and μGS=0.45μN\mu_{\mathrm{GS}}=0.45\mu_N [39], respectively, and being in agreement with an earlier theoretical prediction μGS=0.54μN\mu_{\mathrm{GS}}=0.54\mu_N based on the modified Woods-Saxon potential [40]. Our model analysis in Ref. [37] showed that the Coriolis KK-mixing interaction lowers μGS\mu_{\mathrm{GS}} pushing it towards the experimental values, while its effect on μIS\mu_{\mathrm{IS}} is negligible due to the circumstance that the Kπ=3/2+K^{\pi}=3/2^{+} IS has no mixing partner with angular momentum Iπ=3/2+I^{\pi}=3/2^{+} in the GS band.

These results raise several important questions to our understanding of the 229mTh{}^{229m}\mathrm{Th} problem from the nuclear structure perspective, which we address in this work. (i) To which extent does the shape dynamics play a role for the emergence of such a nuclear structure phenomenon as the tiny energy difference between the IS and the GS? (ii) What is the degree of arbitrariness in the choice of parameters providing the model predictions? The basic input in DSM are the quadrupole (β2\beta_2) and octupole (β3\beta_3) deformations, which determine the s.p. orbitals on which the GS and IS are formed. It is, therefore, important to identify the regions in the (β2,β3)(\beta_2,\beta_3) deformation space of DSM which provide a relevant model treatment of the isomer and the overall spectroscopic properties of the nucleus. To clarify this question, in this work we perform DSM calculations on a grid in a wide range in the QO deformation space covering the regions of physical relevance for a nucleus in the actinide mass region around 229Th{}^{229}\mathrm{Th}. A next question that we address is (iii) whether by including the experimental GS and IS magnetic moment values in the model fits made for different pairs of DSM QO deformations, a better reproduction of μGS\mu_{\mathrm{GS}} could be achieved? How would the model predictions for the other spectroscopic quantities and in particular for B(M1)B(M1) and B(E2)B(E2) change? Finally, (iv) is 229mTh{}^{229m}\mathrm{Th} a unique phenomenon appearing by chance, or the considered dynamical mechanism could provide the presence of similar not yet observed phenomena in other nuclei? In this work we aim to provide answers to these questions, prove the degree of reliability of the model predictions, and clarify details of the mechanism which governs the appearance of the IS.

This work is structured as follows. Sec. II reviews the model formalism in a self-contained form together with details on its application to the 229mTh{}^{229m}\mathrm{Th} problem. In Sec. III results from the calculations in the QO deformation space of DSM with the corresponding behaviour of the B(M1)B(M1), B(E2)B(E2) transition rates and the IS and GS magnetic moments are presented and discussed. In Sec. IV we summarize our analysis and conclude on the reliability of the suggested model mechanism. We thereby provide our updated theoretical predictions for all discussed observables and answer the questions formulated above.

Quadrupole-octupole core plus particle model

Hamiltonian

The model Hamiltonian of axial QO vibrations and rotations coupled to the s.p. motion with Coriolis interaction and pairing correlations can be written in the form [27]

H=Hs.p.+Hpair+Hqo+HCoriol.(1)H = H_{\mathrm{s.p.}} + H_{\mathrm{pair}} + H_{\mathrm{qo}} + H_{\mathrm{Coriol}} . \tag*{(1)}

Here Hs.p.H_{\mathrm{s.p.}} is the single-particle (s.p.) DSM Hamiltonian with the Woods-Saxon potential for fixed axial quadrupole, octupole and higher multipolarity deformations (β2\beta_2, β3\beta_3, β4\beta_4, β5\beta_5, β6\beta_6) [32] providing the s.p. energies εspK\varepsilon_{\mathrm{sp}}^K with given value of the projection KK of the total and s.p. angular momentum operators I^\hat{I} and j^\hat{j}, respectively on the intrinsic symmetry axis. HpairH_{\mathrm{pair}} is the standard BCS pairing Hamiltonian [33] which together with Hs.p.H_{\mathrm{s.p.}} determines the quasi-particle (q.p.) spectrum

εqpK=(εspK−λ)2+Δ2,\varepsilon_{\mathrm{qp}}^K = \sqrt{\left(\varepsilon_{\mathrm{sp}}^K-\lambda\right)^2+\Delta^2},

with the chemical potential λ\lambda and the pairing gap Δ\Delta determined as shown in Ref. [41]. Furthermore, HqoH_{\mathrm{qo}} describes the oscillations of the even-even core with respect to the quadrupole (β~2\widetilde{\beta}_2) and octupole (β~3\widetilde{\beta}_3) axial deformation variables mixed through a centrifugal (rotation-vibration) interaction [28, 29]. Its spectrum is obtained in an analytical form by assuming equal frequencies for the quadrupole and octupole oscillations. The latter are known as the coherent QO mode (CQOM) and will be discussed in more detail in Sec. IIB. Hereafter, we distinguish the CQOM collective (dynamical) variables β~2\widetilde{\beta}_2 and β~3\widetilde{\beta}_3 from the fixed DSM deformation parameters β2\beta_2 and β3\beta_3 considered in this work (see Sec. IIB for clarification).

Returning to the total Hamiltonian in Eq. (1), HCoriolH_{\mathrm{Coriol}} involves the Coriolis interaction between the even-even core and the unpaired nucleon [29]. It is treated as a perturbation with respect to the remaining part of Hamiltonian (1) and then incorporated into the collective QO potential of HqoH_{\mathrm{qo}} defined for a given angular momentum II, parity π\pi and s.p. bandhead projection KbK_b leading to a joint effective term [42]

HqoIKb=−ℏ22B2∂2∂β~22−ℏ22B3∂2∂β~32+12C2β~22+12C3β~32+X~(Iπ,Kb)d2β~22+d3β~32.(2)H_{\mathrm{qo}}^{I K_b} = -\frac{\hbar^2}{2B_2}\frac{\partial^2}{\partial\widetilde{\beta}_2^2} -\frac{\hbar^2}{2B_3}\frac{\partial^2}{\partial\widetilde{\beta}_3^2} +\frac{1}{2}C_2\widetilde{\beta}_2^2 +\frac{1}{2}C_3\widetilde{\beta}_3^2 +\frac{\widetilde{X}(I^\pi,K_b)} {d_2\widetilde{\beta}_2^2+d_3\widetilde{\beta}_3^2}. \tag*{(2)}

Here, B2B_2 (B3B_3), C2C_2 (C3C_3) and d2d_2 (d3d_3) are quadrupole (octupole) mass, stiffness and inertia parameters, respectively. The function X~(Iπ,Kb)\widetilde{X}(I^\pi,K_b) determines the centrifugal term in which the Coriolis mixing is taken into account and has the form:

X~(Iπ,Kb)=12[d0+I(I+1)−Kb2+(−1)I+12(I+12)a12(π,πb)δKb,12]−A∑ν≠b(Kν=12,Kb±1)[a~Kν,Kb(π,πb)(I)εqpKν−εqpKb]2,(3)\widetilde{X}(I^\pi,K_b)=\frac{1}{2}\left[d_0+I(I+1)-K_b^2 +(-1)^{I+\frac{1}{2}}\left(I+\frac{1}{2}\right) a_{\frac{1}{2}}^{(\pi,\pi_b)}\delta_{K_b,\frac{1}{2}} \right] -A\sum_{\substack{\nu\ne b\\(K_\nu=\frac{1}{2},K_b\pm1)}} \left[ \frac{\widetilde{a}_{K_\nu,K_b}^{(\pi,\pi_b)}(I)} {\varepsilon_{\mathrm{qp}}^{K_\nu}-\varepsilon_{\mathrm{qp}}^{K_b}} \right]^2 , \tag*{(3)}

where d0d_0 determines the collective QO potential origin, AA is the Coriolis mixing strength defined in Ref. [42] and the sum is performed over q.p. states with energies εqpKν\varepsilon_{\mathrm{qp}}^{K_\nu} above the Fermi level. For the sum we consider in our numerical calculations ten mixing orbitals. The quantity a12(π,πb)=ππba12,−12(π)a_{\frac{1}{2}}^{(\pi,\pi_b)}=\pi\pi_b a_{\frac{1}{2},-\frac{1}{2}}^{(\pi)} represents the decoupling factor for the case Kb=1/2K_b=1/2, while the quantities a~Kν,Kb(π,πb)(I)\widetilde{a}_{K_\nu,K_b}^{(\pi,\pi_b)}(I) represent the Coriolis mixing factors given by

a~Kν,Kb(π,πb)={(I−Kb)(I+Kb+1)aKν,Kb(πb),Kν=Kb+1,(I+Kb)(I−Kb+1)aKb,Kν(πb),Kν=Kb−1,ππb(−1)I+12(I+12)a12,−12(πb),Kν=Kb=12,(4)\widetilde{a}_{K_\nu,K_b}^{(\pi,\pi_b)}= \begin{cases} \sqrt{(I-K_b)(I+K_b+1)}a_{K_\nu,K_b}^{(\pi_b)}, & K_\nu=K_b+1,\\ \sqrt{(I+K_b)(I-K_b+1)}a_{K_b,K_\nu}^{(\pi_b)}, & K_\nu=K_b-1,\\ \pi\pi_b(-1)^{I+\frac{1}{2}}\left(I+\frac{1}{2}\right)a_{\frac{1}{2},-\frac{1}{2}}^{(\pi_b)}, & K_\nu=K_b=\frac{1}{2}, \end{cases} \tag*{(4)}

with

aKν,Kb(πb)=PKν,KbbNKν(πb)NKb(πb)⟨FKν(πb)∣j^+∣FKb(πb)⟩=PKb,KνbNKb(πb)NKν(πb)⟨FKb(πb)∣j^−∣FKν(πb)⟩.(5)a_{K_\nu,K_b}^{(\pi_b)}= \frac{P_{K_\nu,K_b}^{b}}{N_{K_\nu}^{(\pi_b)}N_{K_b}^{(\pi_b)}} \langle\mathcal{F}_{K_\nu}^{(\pi_b)}|\hat{j}_{+}|\mathcal{F}_{K_b}^{(\pi_b)}\rangle = \frac{P_{K_b,K_\nu}^{b}}{N_{K_b}^{(\pi_b)}N_{K_\nu}^{(\pi_b)}} \langle\mathcal{F}_{K_b}^{(\pi_b)}|\hat{j}_{-}|\mathcal{F}_{K_\nu}^{(\pi_b)}\rangle. \tag*{(5)}

The latter involve matrix elements of the s.p. operators j^±=j^x±ij^y\hat{j}_{\pm}=\hat{j}_x\pm i\hat{j}_y between the parity-projected components of the s.p. wave functions FKb(πb)\mathcal{F}_{K_b}^{(\pi_b)} of the bandhead state and the admixing state FKν(πb)\mathcal{F}_{K_\nu}^{(\pi_b)}. Each s.p. wave function is obtained in DSM [32] as an expansion in the axially-deformed harmonic-oscillator basis ∣NnzΛΣ⟩|Nn_z\Lambda\Sigma\rangle (with Λ+Σ=K\Lambda+\Sigma=K)

FK=∑NnzΛcNnzΛK∣NnzΛΣ⟩.(6)\mathcal{F}_K=\sum_{Nn_z\Lambda}c_{Nn_z\Lambda}^K|Nn_z\Lambda\Sigma\rangle. \tag*{(6)}

In the case of reflection asymmetry (β3≠0\beta_3\ne0) the wave function has a mixed parity and can be decomposed as FK=∑πsp=±1FK(πsp)=FK(+)+FK(−)\mathcal{F}_K=\sum_{\pi_{\mathrm{sp}}=\pm1}\mathcal{F}_K^{(\pi_{\mathrm{sp}})}=\mathcal{F}_K^{(+)}+\mathcal{F}_K^{(-)}, with the s.p. parity given by πsp=(−1)N=±1\pi_{\mathrm{sp}}=(-1)^N=\pm1. The action of the s.p. parity operator π^sp\hat{\pi}_{\mathrm{sp}} gives π^spFK=FK(+)−FK(−)\hat{\pi}_{\mathrm{sp}}\mathcal{F}_K=\mathcal{F}_K^{(+)}-\mathcal{F}_K^{(-)}, and for the parity-projected parts one has π^spFK(±)=±FK(±)\hat{\pi}_{\mathrm{sp}}\mathcal{F}_K^{(\pm)}=\pm\mathcal{F}_K^{(\pm)}. In our approach the projection is made with respect to the experimentally assigned good parity πb\pi^b of the bandhead s.p. state (see below). It is clear that in the presence of octupole deformation each s.p. orbital is characterized by an average (expectation) value of the parity determined as [43]

⟨π^sp⟩=∑NnzΛ(−1)N∣CNnzΛN∣2(7)\langle\hat{\pi}_{\mathrm{sp}}\rangle= \sum_{N n_z \Lambda} (-1)^N \left|C^{N}_{N n_z \Lambda}\right|^2 \tag*{(7)}

with the expansion coefficients calculated in the DSM. The quantity ⟨π^sp⟩\langle\hat{\pi}_{\mathrm{sp}}\rangle takes values in the interval −1≤⟨π^sp⟩≤+1-1 \le\langle\hat{\pi}_{\mathrm{sp}}\rangle\le+1 in dependence on the octupole β3\beta_3 and quadrupole β2\beta_2 deformations entering the DSM.

The quantity NK(πb)=[⟨FK(πb)∣FK(πb)⟩]1/2N^{(\pi^b)}_{K} = \left[\langle\mathcal{F}^{(\pi^b)}_{K} \mid\mathcal{F}^{(\pi^b)}_{K} \rangle\right]^{1/2} in Eq. (5) is a parity-projected normalization factor, whereas PKν′,Kνb=UKν′bUKνb+VKν′bVKνbP^{b}_{K'_\nu,K_\nu} = U^{b}_{K'_\nu}U^{b}_{K_\nu}+V^{b}_{K'_\nu}V^{b}_{K_\nu} involves the BCS occupation factors. The index bb corresponds to the blocked s.p. orbital on which the collective spectrum is built. Since the BCS procedure is performed separately for each (blocked) bandhead orbital, the overlap integrals and the matrix elements between states built on different bandhead orbitals involve the average of both separate occupation factors PKν′,Kνbb′=12(PKν′,Kνb+PKν′,Kνb′)P^{bb'}_{K'_\nu,K_\nu} = \frac{1}{2}\left(P^{b}_{K'_\nu,K_\nu}+P^{b'}_{K'_\nu,K_\nu}\right). The occupation factors UU and VV and the q.p. energies ϵqpK\epsilon^K_{\mathrm{qp}} are obtained by solving the BCS gap equation as done in Ref. [41] with the pairing constant G=GN/PG=G_{N/P} for neutron/proton subsystems of NN protons and ZZ neutrons determined as [19]

GN/P=1N+Z(g0∓g1N−ZN+Z)(8)G_{N/P}=\frac{1}{N+Z}\left(g_0 \mp g_1\frac{N-Z}{N+Z}\right) \tag*{(8)}

Here the pairing parameter g0g_0 is considered to vary between the values g0=17.8 MeVg_0=17.8\ \mathrm{MeV}, used in Ref. [41] for DSM plus BCS calculations in the actinide region, and g0=19.2 MeVg_0=19.2\ \mathrm{MeV}, suggested in Ref. [19] for rare-earth nuclei, while g1g_1 is considered to vary around the value g1=7.4 MeVg_1=7.4\ \mathrm{MeV} used in both references cited above.

Model solution, spectrum and wave functions

The spectrum which corresponds to the Hamiltonian (1) represents QO vibrations and rotations built on a q.p. state with K=KbK=K_b and parity πb\pi^b. It is obtained in two steps. First, the s.p. and q.p. energy levels and wave functions are obtained through a DSM plus BCS calculation performed for fixed β2\beta_2- and β3\beta_3- parameter values of the s.p. Woods-Saxon potential, providing the odd-nucleon energy contribution to the bandhead, ϵqpKb\epsilon^{K_b}_{\mathrm{qp}}, and the Coriolis mixing factors a~\tilde{a}, Eq. (4), in the centrifugal part X~\widetilde{X}, Eq. (3), of the QO Hamiltonian (2). In the second step the collective QO vibration-rotation energies and wave functions are obtained through the solution of the Schrödinger equation for the two-dimensional potential in the collective β~2\widetilde{\beta}_2- and β~3\widetilde{\beta}_3- variables of Hamiltonian (2). In the following we will address in more detail this second step.

In the general case of arbitrary values of the Hamiltonian parameters B2B_2, B3B_3, C2C_2, C3C_3 and d2d_2, d3d_3, the solution of the Schrödinger equation in β~2\widetilde{\beta}_2 and β~3\widetilde{\beta}_3 has to be obtained numerically. A transformation of variables introduces the ellipsoidal “radial” and “angular” coordinates, respectively,

η=[2(d2β~22+d3β~32)d2+d3]12,ϕ=arctan⁡(β~3β~2d3d2)(9)\eta=\left[\frac{2(d_2\widetilde{\beta}_2^2+d_3\widetilde{\beta}_3^2)}{d_2+d_3}\right]^{\frac{1}{2}},\qquad\phi=\arctan\left(\frac{\widetilde{\beta}_3}{\widetilde{\beta}_2}\sqrt{\frac{d_3}{d_2}}\right) \tag*{(9)}

such that

β~2=pηcos⁡ϕ,β~3=qηsin⁡ϕ(10)\widetilde{\beta}_2=p\eta\cos\phi,\qquad\widetilde{\beta}_3=q\eta\sin\phi \tag*{(10)}

with

p=dd2,q=dd3,d=12(d2+d3).(11)p=\sqrt{\frac{d}{d_2}},\qquad q=\sqrt{\frac{d}{d_3}},\qquad d=\frac{1}{2}(d_2+d_3). \tag*{(11)}

An analytical solution for the spectrum of HqoH_{\mathrm{qo}} can be found for a specific set of parameters, when assuming coherent QO oscillations (the so-called coherent QO mode, CQOM) with a frequency ω=C2/B2=C3/B3=C/B\omega=\sqrt{C_2/B_2}=\sqrt{C_3/B_3}=\sqrt{C/B}. Then, the two-dimensional potential in Hamiltonian (2) obtains a shape with an ellipsoidal equipotential bottom in the space of the collective deformation variables β~2\widetilde{\beta}_2 and β~3\widetilde{\beta}_3 [28]. This allows for a separation of the ellipsoidal radial variables and reduction of the problem to the one-dimensional Schrödinger equation for an analytically solvable potential of Davidson type in the radial variable η\eta. The motion with respect to this potential corresponds to a “soft” QO vibration mode without fixed minima in β~2\widetilde{\beta}_2 and β~3\widetilde{\beta}_3. These should not be confused with the fixed β2\beta_2- and β3\beta_3- deformations in the s.p. Woods-Saxon potential of the DSM. The CQOM approach has been successfully applied to QO spectra of even-even and odd mass nuclei [28, 29, 30, 31].

The quadrupole and octupole semiaxes β~2sa\widetilde{\beta}^{\mathrm{sa}}_{2} and β~3sa\widetilde{\beta}^{\mathrm{sa}}_{3} of the ellipsoidal CQOM potential bottom are defined for even-even nuclei as [30]

β~λsa(I)=[2X(I)/dλCλ]1/4,λ=2,3(12)\widetilde{\beta}^{\mathrm{sa}}_{\lambda}(I)=\left[2X(I)/d_{\lambda}C_{\lambda}\right]^{1/4},\qquad\lambda=2,3 \tag*{(12)}

with the centrifugal factor X(I)=[d0+I(I+1)]/2X(I)=[d_0+I(I+1)]/2. For an odd-AA nucleus the expression for the semiaxes takes the form

β~λsa(Iπ,Kb)=[2X~(Iπ,Kb)/dλCλ]1/4,λ=2,3(13)\widetilde{\beta}^{\mathrm{sa}}_{\lambda}(I^\pi,K_b)=\left[2\widetilde{X}(I^\pi,K_b)/d_{\lambda}C_{\lambda}\right]^{1/4},\qquad\lambda=2,3 \tag*{(13)}

with X~(Iπ,Kb)\widetilde{X}(I^\pi,K_b) determined in Eq. (3). Comparing the expressions of X(I)X(I) and X~(Iπ,Kb)\widetilde{X}(I^\pi,K_b) it becomes clear that for the odd-nucleus the semiaxes β~λ=2,3sa(Iπ,Kb)\widetilde{\beta}^{\mathrm{sa}}_{\lambda=2,3}(I^\pi,K_b) in Eq. (13) differ from the semiaxes β~λsa(I)\widetilde{\beta}^{\mathrm{sa}}_{\lambda}(I) (12) of the original even-even core CQOM potential because here the X~(Iπ,Kb)\widetilde{X}(I^\pi,K_b) factor includes the additional term (−Kb2)(-K_b^2) as well as the Coriolis mixing and decoupling contributions from the single nucleon. The CQOM potential semiaxes obey the relation [30]

β~3aβ~2a=12p2−1,(14)\frac{\tilde{\beta}_{3a}}{\tilde{\beta}_{2a}}=\frac{1}{\sqrt{2p^2-1}}, \tag*{(14)}

with pp defined in Eq. (11) determining the relative contribution of the quadrupole and octupole collective modes in the coherent QO motion. We note that the value p=1p=1 corresponds to equal values of both semiaxes, i.e., to a circle form of the CQOM potential bottom. In terms of the coherence assumption concept this means that both the quadrupole β~2\tilde{\beta}_2 and octupole β~3\tilde{\beta}_3 deformation modes enter the collective CQOM motion with the same weight. This case will be discussed later in the paper and is exemplified in Figure 4 and Figure 5.

Three-dimensional wave-function density plots and projected two-dimensional contour plots for the $I^\pi=5/2^+$ and $5/2^-$ states of $^{229}\mathrm{Th}$

Figure 4. The CQOM QO wave-function density ∣ΦnkInq(β~2,β~3)∣2|\Phi_{nkI}^{\mathrm{nq}}(\tilde{\beta}_2,\tilde{\beta}_3)|^2 from Eq. (15) as a function of β~2\tilde{\beta}_2 and β~3\tilde{\beta}_3 for the Iπ=5/2+I^\pi=5/2^+ GS (with k=1k=1) and 5/2−5/2^- state (with k=2k=2) of the yrast quasi-parity-doublet in 229Th^{229}\mathrm{Th}. We use here the DSM+CQOM fit with qS=qR=0.6q_{\mathrm{S}}=q_{\mathrm{R}}=0.6 and the Woods-Saxon DSM deformation parameters (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115). The upper panels represent three-dimensional plots while the lower panels illustrate the corresponding projected two-dimensional contour plots. The CQOM potential bottoms defined by the semiaxes β~2,3sa\tilde{\beta}^{\mathrm{sa}}_{2,3} are shown as red circles.

Projected two-dimensional contour plots for the $I^\pi=15/2^+$ and $15/2^-$ states of $^{229}\mathrm{Th}$

Figure 5. The same as Figure 4 (lower panels), but for the Iπ=15/2+I^\pi=15/2^+ and 15/2−15/2^- yrast quasi-parity-doublet states of 229Th^{229}\mathrm{Th}. Note the considerable increase in the QO semiaxes and the corresponding expansion of the dynamical-deformation peak positions in the space of the collective CQOM variables β~2\tilde{\beta}_2 and β~3\tilde{\beta}_3.

Despite the missing single (β~2,β~3)(\tilde{\beta}_2,\tilde{\beta}_3)-minimum in the CQOM potential, the collective QO states of the system are still characterized by the so-called dynamical deformations determined by the density maxima of the QO vibration wave function. Explicitly, the CQOM QO vibration wave function is given by [28, 30]

ΦnkIπqo(η,ϕ)=ψnkI(η)φkπqo(ϕ),(15)\Phi_{nkI}^{\pi_{\mathrm{qo}}}(\eta,\phi)=\psi_{nk}^{I}(\eta)\varphi_k^{\pi_{\mathrm{qo}}}(\phi), \tag*{(15)}

where the radial part

ψnkI(η)=2cΓ(n+1)Γ(n+2s+1)e−cη2/2(cη2)sLn2s(cη2)(16)\psi_{nk}^{I}(\eta)=\sqrt{\frac{2c\Gamma(n+1)}{\Gamma(n+2s+1)}}e^{-c\eta^2/2}(c\eta^2)^sL_n^{2s}(c\eta^2) \tag*{(16)}

involves generalized Laguerre polynomials in the variable η\eta, with s=(1/2)k2+bX~(I,K)s=(1/2)\sqrt{k^2+b\tilde{X}(I,K)} and c=BC/ℏc=\sqrt{BC}/\hbar, the latter having the meaning of a reduced QO oscillator frequency, and Γ(z)\Gamma(z) denotes the Gamma function. The angular part in the variable ϕ\phi appears with a positive or negative parity πqo\pi_{\mathrm{qo}} of the collective QO mode as follows

φk+(ϕ)=2πcos⁡(kϕ),k=1,3,5,…;(17)\varphi_k^{+}(\phi)=\sqrt{\frac{2}{\pi}}\cos(k\phi),\qquad k=1,3,5,\ldots; \tag*{(17)}
φk−(ϕ)=2πsin⁡(kϕ),k=2,4,6,….(18)\varphi_k^{-}(\phi)=\sqrt{\frac{2}{\pi}}\sin(k\phi),\qquad k=2,4,6,\ldots. \tag*{(18)}

The maxima of the density ∣ΦnkIπqo∣2→∣ΦnkIπqo(β~2,β~3)∣2|\Phi_{nkI}^{\pi_{\mathrm{qo}}}|^2\to|\Phi_{nkI}^{\pi_{\mathrm{qo}}}(\tilde{\beta}_2,\tilde{\beta}_3)|^2 calculated in the (β~2,β~3)(\tilde{\beta}_2,\tilde{\beta}_3) space pin down the dynamical deformations [30]. Strictly speaking, the dynamical deformation is defined by the expectation value of the square of the corresponding multipole (deformation) operator in the CQOM state, but considering the density maximum is enough to locate its position in the (β~2,β~3)(\tilde{\beta}_2,\tilde{\beta}_3) space. The positions of these maxima are situated outside of the potential bottom ellipse and move further out with increasing angular momentum. They essentially characterize the collective dynamical behaviour of the nucleus in the presence of a coherent mode. This will be illustrated in § 3 B for the present model application in 229Th^{229}\mathrm{Th}. It will be seen that the CQOM dynamical deformations appearing in the overall collective spectrum of the nucleus are reasonably correlated with the intrinsic Woods-Saxon DSM QO deformations.

We should stress here, however, that the dynamical QO deformations in CQOM do not need to ultimately coincide or even to be close to the fixed Woods-Saxon deformations β2\beta_2 and β3\beta_3 of the DSM. Imposing artificially such a constraint would deprive the overall algorithm of the capability to incorporate the individual (separate) dynamic properties of the collective and s.p. degrees of freedom (carried by the available data) and, therefore, of the possibility to plausibly reproduce the interaction between them. The present model formalism does not put a constraint on both potentials but rather leaves them to independently feel, as much as possible, the corresponding physical conditions which govern the nuclear collective and intrinsic motions and their very fine interplay. As it will be seen in the following § 2 E, in the case of 229mTh^{229m}\mathrm{Th}, the DSM deformations β2\beta_2 and β3\beta_3 determine the hyperfine (from the nuclear point of view) conditions for the appearance of the Kπ=3/2+K^\pi=3/2^+ isomer while the dynamical CQOM deformations in the (β~2,β~3)(\tilde{\beta}_2,\tilde{\beta}_3)-space reflect the conditions imposed by the overall collective spectrum which complement the microscopic isomer-formation mechanism.

By taking the analytical CQOM solution together with the result of the DSM plus BCS calculation the QO core plus particle spectrum built on the given q.p. bandhead state is obtained in the form [27]

Enktot(Iπ,Kb)=ϵqpKb+ℏω[2n+1+k2+bX~(I,Kb)].(19)E_{nk}^{\mathrm{tot}}(I^\pi,K_b)=\epsilon_{\mathrm{qp}}^{K_b}+\hbar\omega\left[2n+1+\sqrt{k^2+b\tilde{X}(I,K_b)}\right]. \tag*{(19)}

Here b=2B/(ℏ2d)b=2B/(\hbar^2d) has the meaning of a reduced inertia parameter, while n=0,1,2,…n=0,1,2,\ldots and k=1,2,3,…k=1,2,3,\ldots stand for the radial and angular QO oscillation quantum numbers, respectively, with kk odd (even) for the even-parity (odd-parity) states of the core [28, 29]. The levels of the total QO core plus particle system, determined by the given nn and pair of k(+)k^{(+)} and k(−)k^{(-)} values for the states with Iπ=+I^{\pi=+} and Iπ=−I^{\pi=-}, respectively, form a split doublet with respect to the parity, called a quasi-parity-doublet [29, 31].

The corresponding wave functions can be constructed in three steps. First, the quadrupole-octupole vibration wave function of the CQOM is calculated according to Eq. (15). Second, we can construct the unperturbed QO core plus particle wave function [31, 42]:

ΨnkIMKπb(η,ϕ,θ)=1NK(πb)2I+116π2ΦnkIπb(η,ϕ)[DMKI(θ)FK(πb)+ππb(−1)I+KDM−KI(θ)F−K(πb)],(20)\Psi_{nkIMK}^{\pi_b}(\eta,\phi,\theta)=\frac{1}{N_K^{(\pi_b)}}\sqrt{\frac{2I+1}{16\pi^2}}\Phi_{nkI}^{\pi_b}(\eta,\phi)\left[D^I_{MK}(\theta)\mathcal{F}^{(\pi_b)}_K+\pi\pi_b(-1)^{I+K}D^I_{M-K}(\theta)\mathcal{F}^{(\pi_b)}_{-K}\right], \tag*{(20)}

where DMKI(θ)D^I_{MK}(\theta) are the rotation (Wigner) functions and ΦnkIπb(η,ϕ)\Phi_{nkI}^{\pi_b}(\eta,\phi) are the QO vibration functions (15) with πqo=πb\pi_{\mathrm{qo}}=\pi_b. In Eq. (20) the relevant part FK(πb)=FK(+)\mathcal{F}^{(\pi_b)}_K=\mathcal{F}^{(+)}_K or FK(−)\mathcal{F}^{(-)}_K of the s.p. wave function FK\mathcal{F}_K given in Eq. (6) is taken by projecting the latter with respect to the experimentally assigned bandhead parity πb=+\pi^b=+ or −-, thus providing a good parity of the total core-plus-particle wave function.

Finally, the Coriolis perturbed wave function Ψ~nkIMKbπ,πb\widetilde{\Psi}^{\pi,\pi_b}_{nkIMK_b} corresponding to Hamiltonian (1) with the spectrum (19) is obtained in the first order of perturbation theory and has the form

Ψ~nkIMKbπ,πb=1N~IπKb[Ψ~nkIMKbπ,πb+A∑ν≠bCKν,KbIπΨ~nkIMKνπ,πb],(21)\widetilde{\Psi}^{\pi,\pi_b}_{nkIMK_b}=\frac{1}{\widetilde{N}_{I\pi K_b}}\left[\widetilde{\Psi}^{\pi,\pi_b}_{nkIMK_b}+A\sum_{\nu\ne b}C^{I\pi}_{K_\nu,K_b}\widetilde{\Psi}^{\pi,\pi_b}_{nkIMK_\nu}\right], \tag*{(21)}

where Kν=Kb±1,12K_\nu=K_b\pm1,\frac{1}{2}, the expansion coefficients read

CKν,KbIπ=α~Kν,Kb(π,πb)(I)ϵKν(π)−ϵKb(π),(22)C^{I\pi}_{K_\nu,K_b}=\frac{\widetilde{\alpha}^{(\pi,\pi_b)}_{K_\nu,K_b}(I)}{\epsilon^{(\pi)}_{K_\nu}-\epsilon^{(\pi)}_{K_b}}, \tag*{(22)}

while the normalization factor is given by

N~IπKb2=⟨Ψ~nkIMKbπ,πb|Ψ~nkIMKbπ,πb⟩=1+2A∑ν≠bKν=Kb±12CKν,KbIπδKν,KbPKν,KbbNKν(πb)NKb(πb)⟨FKν(πb)|FKb(πb)⟩+A2∑ν1,ν2≠bKν1,Kν2=Kb±1,12CKν1,KbIπCKν2,KbIπδKν1Kν2PKν1Kν2bNKν1(πb)NKν2(πb)⟨FKν1(πb)|FKν2(πb)⟩.(23)\begin{aligned} \widetilde{N}^{2}_{I\pi K_b}=\left\langle\widetilde{\Psi}^{\pi,\pi_b}_{nkIMK_b}\middle|\widetilde{\Psi}^{\pi,\pi_b}_{nkIMK_b}\right\rangle\\ &=1+2A\sum_{\nu\ne b}^{K_\nu=K_b\pm\frac{1}{2}}C^{I\pi}_{K_\nu,K_b}\delta_{K_\nu,K_b}\frac{P^{b}_{K_\nu,K_b}}{N^{(\pi_b)}_{K_\nu}N^{(\pi_b)}_{K_b}}\left\langle F_{K_\nu}(\pi_b)\middle|F_{K_b}(\pi_b)\right\rangle\\ &+A^{2}\sum_{\substack{\nu_1,\nu_2\ne b\\K_{\nu_1},K_{\nu_2}=K_b\pm1,\frac{1}{2}}}C^{I\pi}_{K_{\nu_1},K_b}C^{I\pi}_{K_{\nu_2},K_b}\delta_{K_{\nu_1}K_{\nu_2}}\frac{P^{b}_{K_{\nu_1}K_{\nu_2}}}{N^{(\pi_b)}_{K_{\nu_1}}N^{(\pi_b)}_{K_{\nu_2}}}\left\langle F_{K_{\nu_1}}(\pi_b)\middle|F_{K_{\nu_2}}(\pi_b)\right\rangle. \tag*{(23)} \end{aligned}

Electric and magnetic transition rates

Expressions for the reduced B(E1)B(E1)-, B(E2)B(E2)- and B(E3)B(E3)- probabilities for transitions between states with energies given by Eq. (19) and Coriolis perturbed wave function given by Eq. (21) are derived by using the electric transition operators in the general form

Qμ(Eλ)=2λ+14π(4−3δλ,1)Q^λ0∑νDμνλ,Q_\mu(E\lambda)=\sqrt{\frac{2\lambda+1}{4\pi(4-3\delta_{\lambda,1})}}\widehat{Q}_{\lambda0}\sum_\nu D^\lambda_{\mu\nu},
λ=1,2,3,μ=0,±1,…,±λ,(24)\lambda=1,2,3,\quad\mu=0,\pm1,\ldots,\pm\lambda, \tag*{(24)}

with the explicit form of the operators Q^λ0\widehat{Q}_{\lambda0} given by Eqs. (31)–(33) in [30].

The expression for the B(M1)B(M1) reduced transition probability was obtained by using the standard core plus particle magnetic dipole (M1M1) operator (e.g. see Eq. (3.61) in Ref. [33]) written as

M^1=34πμN[gR(I^−j^)+gss^+gll^],(25)\widehat{M}_{1}=\sqrt{\frac{3}{4\pi}}\mu_N\left[g_R(\widehat{I}-\widehat{j})+g_s\widehat{s}+g_l\widehat{l}\right], \tag*{(25)}

after taking it in the intrinsic frame. The operators s^\widehat{s} and l^\widehat{l} in Eq. (25) correspond to the s.p. spin and orbital momenta and j^=l^+s^\widehat{j}=\widehat{l}+\widehat{s}. The quantities gsg_s and glg_l are the spin and orbital gyromagnetic factors, respectively, and gRg_R is the collective gyromagnetic factor. The orbital factor is gl=0 (1)g_l=0\ (1) for neutrons (protons), while the spin factor is taken as gs=qsgsfreeg_s=q_sg_s^{\mathrm{free}}, with gsfree=−3.826 (5.586)g_s^{\mathrm{free}}=-3.826\ (5.586) for neutrons (protons) [33]. The quantity qsq_s is an attenuation factor usually supposed to be qs=0.6−0.7q_s=0.6-0.7, taking into account spin-polarization effects [44]. The collective gyromagnetic factor gRg_R is often associated with the ratio gR=Z/(Z+N)g_R=Z/(Z+N), with ZZ and NN being the proton and neutron numbers, respectively, adopted on the basis of the liquid-drop-model [45]. However, it is known that in most deformed nuclei gRg_R is lowered with respect to this ratio by 20%–30% or more [46, 47], with the attenuation being explained by the influence of the pairing interaction on the collective moment of inertia [48, 49, 50]. Therefore, in Ref. [37] we have introduced the relevant quenching factor qRq_R such that gR=qRZ/(Z+N)g_R=q_RZ/(Z+N), showing that on the basis of several earlier theoretical and experimental analyses, it can be taken for 229^{229}Th as low as qR∼0.6q_R\sim0.6. Below it will be seen that both attenuation factors qsq_s and qRq_R play an important role in the model prediction of the B(M1)B(M1) transition rates and magnetic moments and their consideration with further slightly lower values may shed more light on the 229^{229}Th formation mechanism.

The following common form of the expressions for both types (TT) of the electric (T=ET=E) and magnetic (T=MT=M) transition with multipolarity λ\lambda between initial (i) and final (f) states was derived [27]

B(Tλ;πibiIiπiKi→πfbfIfπfKf)=RδπfπiTλ[(1+πfπi(−1)δT,E)/2]×1N~IfπfKf2N~IiπiKi2[δKfKiCIiKiKfIfπfPKfKibfbiMKfKiπbfπbiNKf(πbf)NKi(πbi)\begin{aligned} B(T\lambda;\pi_i^{b_i}I_i\pi_iK_i\to\pi_f^{b_f}I_f\pi_fK_f) =R^{T\lambda}_{\delta\pi_f\pi_i}\left[(1+\pi_f\pi_i(-1)^{\delta_{T,E}})/2\right] \\ &\times\frac{1}{\widetilde{N}^{2}_{I_f\pi_fK_f}\widetilde{N}^{2}_{I_i\pi_iK_i}}\left[\delta_{K_fK_i}C^{I_f\pi_f}_{I_iK_iK_f}\frac{P^{b_fb_i}_{K_fK_i}M^{\pi_b^f\pi_b^i}_{K_fK_i}}{N^{(\pi_b^f)}_{K_f}N^{(\pi_b^i)}_{K_i}} \right. \end{aligned}
+ACIiKfλ0Ifπf∑ν≠fKν=Kf±1,12δKfKνCKν,KiIiπiPKfKνbfbiMKfKνπbfπbiNKf(πbf)NKν(πbi)\left. +A C^{I_f\pi_f}_{I_iK_f\lambda0}\sum_{\substack{\nu\ne f\\K_\nu=K_f\pm1,\frac{1}{2}}}\delta_{K_fK_\nu}C^{I_i\pi_i}_{K_\nu,K_i}\frac{P^{b_fb_i}_{K_fK_\nu}M^{\pi_b^f\pi_b^i}_{K_fK_\nu}}{N^{(\pi_b^f)}_{K_f}N^{(\pi_b^i)}_{K_\nu}} \right.
+ACIiKiλ0Ifπf∑ν≠fKν=Kf±1,12δKνKiCKν,KfIfπfPKνKibfbiMKνKiπbfπbiNKν(πbf)NKi(πbi)\left. +A C^{I_f\pi_f}_{I_iK_i\lambda0}\sum_{\substack{\nu\ne f\\K_\nu=K_f\pm1,\frac{1}{2}}}\delta_{K_\nu K_i}C^{I_f\pi_f}_{K_\nu,K_f}\frac{P^{b_fb_i}_{K_\nu K_i}M^{\pi_b^f\pi_b^i}_{K_\nu K_i}}{N^{(\pi_b^f)}_{K_\nu}N^{(\pi_b^i)}_{K_i}} \right.
+A2∑ν′,ν′′≠fKν′,Kν′′=Kf±1,12∑ν≠iKν=Ki±1,12δKν′KνCKν′,KfIfπfCKν,KiIiπiPKν′KνbfbiMKν′KνπbfπbiNKν′(πbf)NKν(πbi)],(26)\left. +A^{2}\sum_{\substack{\nu',\nu''\ne f\\K_{\nu'},K_{\nu''}=K_f\pm1,\frac{1}{2}}}\sum_{\substack{\nu\ne i\\K_\nu=K_i\pm1,\frac{1}{2}}}\delta_{K_{\nu'}K_{\nu}}C^{I_f\pi_f}_{K_{\nu'},K_f}C^{I_i\pi_i}_{K_{\nu},K_i}\frac{P^{b_fb_i}_{K_{\nu'}K_{\nu}}M^{\pi_b^f\pi_b^i}_{K_{\nu'}K_{\nu}}}{N^{(\pi_b^f)}_{K_{\nu'}}N^{(\pi_b^i)}_{K_{\nu}}} \right], \tag*{(26)}

where the factor

Rλ=EλTλ=2λ+14π(4−3δλ,1)RδπbinikiIi→πbfnfkfIf2(27)R^{T\lambda}_{\lambda=E\lambda}=\frac{2\lambda+1}{4\pi(4-3\delta_{\lambda,1})}R^{2}_{\delta\pi_b^i n_i k_i I_i\to\pi_b^f n_f k_f I_f} \tag*{(27)}

involves integrals on the radial and angular variables in CQOM (see Eqs. (35)–(41) and Appendixes B and C in Ref. [30]) and

RT1=M1=34πμN2(28)R^{T1=M1}=\frac{3}{4\pi}\mu_N^2 \tag*{(28)}

involves the nuclear magneton μN\mu_N. Also here

MKfKiπbπf={⟨FKf(πb)∣FKi(πi)⟩,for T=E(gl−gR)KiδKfKi⟨FKf(πb)∣FKi(πi)⟩+(gs−gl)⟨FKf(πb)∣s^0∣FKi(πi)⟩,for T=M,(29)M^{\pi_b\pi_f}_{K_fK_i}= \begin{cases} \langle\mathcal{F}^{(\pi_b)}_{K_f}|\mathcal{F}^{(\pi_i)}_{K_i}\rangle, & \text{for }T=E\\ (g_l-g_R)K_i\delta_{K_fK_i}\langle\mathcal{F}^{(\pi_b)}_{K_f}|\mathcal{F}^{(\pi_i)}_{K_i}\rangle\\ \qquad+(g_s-g_l)\langle\mathcal{F}^{(\pi_b)}_{K_f}|\hat{s}_0|\mathcal{F}^{(\pi_i)}_{K_i}\rangle, & \text{for }T=M, \end{cases} \tag*{(29)}

where s^0\hat{s}_0 is the zz component of the spin operator in spherical representation. The factors CI1K1λμI2K2C^{I_2K_2}_{I_1K_1\lambda\mu} in Eq. (26) are Clebsch-Gordan coefficients. The integrals in Eq. (27) depend on the model parameters $c$, defined below Eq. (16), and $p$, {xref}Eq. (11), both determining the electric transition probabilities {citep}[30]`.

The reduced transition probability expression (26) contains first-order and second-order KK-mixing effects. First-order mixing terms practically contribute with nonzero values only in the cases Ki/f=Kν=1/2K_{i/f}=K_\nu=1/2, i.e., when a Ki/f=1/2K_{i/f}=1/2 bandhead state is mixed with another Kν=1/2K_\nu=1/2 state present in the considered range of admixing orbitals. A second-order mixing effect connects states with ΔK=1,2\Delta K=1,2 and allows different combinations of ∣Ki−Kf∣≤2|K_i-K_f|\leq2 which provide respective nonzero contribution of the Coriolis mixing to the transition probability. In this way the present formalism provides nonzero transition probabilities between states with different KK-values despite the axial symmetry assumed in both CQOM and DSM parts of Hamiltonian (1). We stress that although often disregarded in the literature, it is only through the Coriolis mixing that the M1M1 and E2E2 isomer decay channels for 229mTh^{229\mathrm{m}}\mathrm{Th} are rendered possible within the model discussed here.

Magnetic moment

The described model formalism allows us to obtain the magnetic-dipole moment in any state of the quasi-parity doublet spectrum characterized by the Coriolis perturbed wave function Ψ~IMKb\widetilde{\Psi}_{IMK_b} (21). The magnetic moment is determined by the matrix element μ=4π3⟨Ψ~IIKb∣M^10∣Ψ~IIKb⟩\mu=\sqrt{\frac{4\pi}{3}}\langle\widetilde{\Psi}_{IIK_b}|\hat{M}_{10}|\widetilde{\Psi}_{IIK_b}\rangle, where M^10\hat{M}_{10} is the zeroth spherical tensor component of the operator M^1\hat{M}_1, Eq. (25), taken after transformation into the intrinsic frame (see Chapter 9 of Ref. [47]). Thus we obtain the following expression for the magnetic moment in a state with collective angular momentum II and parity π\pi built on a q.p. bandhead state with K=KbK=K_b and π=πb\pi=\pi^b:

μ=μNgRI+1I+11N~IπKb2[KbMKbKbπbNKb(πb)+2AKb∑ν≠bKν=Kb=12δKνKbCKνKbIπPKνKbbMKνKbπbNKν(πb)NKb(πb)+A2∑ν1,ν2≠bKν1,Kν2=Kν=Kb±1,12δKν1Kν2CKν1KbIπCKν2KbIπPKν1Kν2bMKν2Kν1πbNKν1(πb)NKν2(πb)](30)\mu=\mu_Ng_RI+\frac{1}{I+1}\frac{1}{\widetilde{N}_{I\pi K_b}^2}\left[K_b\frac{M^{\pi_b}_{K_bK_b}}{N^{(\pi_b)}_{K_b}}+2AK_b\sum_{\substack{\nu\ne b\\K_\nu=K_b=\frac{1}{2}}}\delta_{K_\nu K_b}C^{I\pi}_{K_\nu K_b}\frac{P^{b}_{K_\nu K_b}M^{\pi_b}_{K_\nu K_b}}{N^{(\pi_b)}_{K_\nu}N^{(\pi_b)}_{K_b}}+A^2\sum_{\substack{\nu_1,\nu_2\ne b\\K_{\nu_1},K_{\nu_2}=K_\nu\\=K_b\pm1,\frac{1}{2}}}\delta_{K_{\nu_1}K_{\nu_2}}C^{I\pi}_{K_{\nu_1}K_b}C^{I\pi}_{K_{\nu_2}K_b}\frac{P^b_{K_{\nu_1}K_{\nu_2}}M^{\pi_b}_{K_{\nu_2}K_{\nu_1}}}{N^{(\pi_b)}_{K_{\nu_1}}N^{(\pi_b)}_{K_{\nu_2}}}\right] \tag*{(30)}

with MKμKνπb≡MKμKνπb,πbM^{\pi_b}_{K_\mu K_\nu}\equiv M^{\pi_b,\pi_b}_{K_\mu K_\nu} being defined in Eq. (29) (T=MT=M) and all other quantities being already defined above. We note that the complete expression would involve an additional decoupling term applying for the case of Kb=1/2K_b=1/2 appearing after transforming the M^1\hat{M}_1 operator (25) into the intrinsic frame [47]. Here we do not take it into account in Eq. (30), since in the present application of the model to 229Th^{229}\mathrm{Th} no K=1/2K=1/2 bandheads appear. The second and the third term in the brackets of Eq. (30) take into account the influence of the Coriolis mixing on the magnetic moment. In fact, also the second term only applies for Kb=1/2K_b=1/2, but we keep it for consistency with the B(Eλ)B(E\lambda) and B(M1)B(M1) expressions (26). Thus, in the present application of the model only the third term is important for the Coriolis mixing in the magnetic moment.

One can easily check that in the case of missing Coriolis mixing Eq. (30) appears in the usual form of the particle-rotor expression, e.g. Eq. (3.62) in Ref. [33], in which the intrinsic gyromagnetic ratio gKg_{K} is

gKb=1Kb1[NKb(πb)]2⟨FKb(πb)∣gs⋅Σ+gl⋅Λ∣FKb(πb)⟩.(31)g_{K_b}=\frac{1}{K_b}\frac{1}{[N^{(\pi_b)}_{K_b}]^2}\langle\mathcal{F}^{(\pi_b)}_{K_b}|g_s\cdot\boldsymbol{\Sigma}+g_l\cdot\boldsymbol{\Lambda}|\mathcal{F}^{(\pi_b)}_{K_b}\rangle. \tag*{(31)}

Equation (31) still takes into account the circumstance that in the case of nonzero octupole deformation we have to apply the projected and renormalized s.p. wave function as explained below Eq. (20). In the case of missing octupole deformation (nonmixed s.p. wave function), Eq. (31) reduces to the standard “reflection-symmetric” expression (3.63) in Ref. [33]. In this way the present model expression for the magnetic-dipole moment in Eq. (30) is consistent with the relevant limiting cases.

Model application in 229Th^{229}\mathrm{Th}

The CQOM plus DSM-BCS model framework described above contains a number of parameters that are determined according to the physical conditions which govern the structure and dynamics of the nucleus 229Th^{229}\mathrm{Th} and to the available experimental data. These parameters are the two already discussed Woods-Saxon DSM QO deformations β2\beta_2 and β3\beta_3, the five CQOM parameters, namely the QO oscillator frequency ω\omega, the reduced inertia factor bb in Eq. (19), the parameter d0d_0 in Eq. (3) and the parameters cc and pp from Eqs. (16) and (11), respectively, entering Eq. (27), the Coriolis mixing constant AA, and the two pairing parameters g0g_0 and g1g_1 entering Eq. (8). As it will be detailed below, the first two (Woods-Saxon DSM QO deformation) parameters are determined in a region of the deformation space providing ultimate DSM conditions for the formation of the 229m^{229\mathrm{m}}Th isomer. The pairing constants are fixed for the overall study to values in a range typical for the adjacent regions of nuclei. Finally, the five CQOM parameters and the Coriolis mixing constant are adjusted in a fitting procedure to quantitatively reproduce the positive- and negative-parity levels of 229^{229}Th with energy below 400 keV as well as the available experimental data on transition rates and magnetic moments at each particular Woods-Saxon DSM QO deformation. As explained in § 2 B, the five CQOM parameters also determine the shape of the collective potential and the corresponding dynamical deformations in the (β~2,β~3)(\widetilde{\beta}_2,\widetilde{\beta}_3)-space. The rather fine parameter determination procedure described here is based on the following physical assumptions:

  1. The considered part of the spectrum consists of two quasi-parity-doublets: an yrast one, based on the Kb=5/2+K_b = 5/2^+ GS corresponding to the 5/2[633] s.p. orbital and a nonyrast quasi-parity-doublet, built on the isomeric Kb=3/2+K_b = 3/2^+ state corresponding to the 3/2[631] orbital. Both orbitals are very close to each other providing a quasidegeneracy of the GS and IS. This condition primary depends on the choice of the quadrupole (β2\beta_2) and octupole (β3\beta_3) deformation parameters in DSM and on the BCS pairing contribution in the q.p. energy of both states.

  1. Both quasi-parity-doublets correspond to coherent QO vibrations and rotations with the same radial-oscillation quantum number n=0n = 0, the lowest possible angular-oscillation number k(+)=1k^{(+)} = 1 for the positive-parity sequences and one of the few lowest possible k(−)=2,4,6k^{(-)} = 2, 4, 6 values for the negative-parity states [see Eq. (19) and the text below it]. Hereinafter we consider only the lowest k(−)=2k^{(-)} = 2 value in the both quasi-parity-doublets. This suggests completely identical QO vibration modes superposed on both GS and IS. The vibration modes alone obviously do not cause any mutual displacement of the two quasi-parity-doublets, but the term Kb2K_b^2 in the centrifugal expression X~(Iπ,Kb)\widetilde{X}(I^\pi,K_b) in Eq. (3) does. It directly mixes the collective energy with the bandhead and down-shifts the Kb=5/2+K_b = 5/2^+ level sequences with respect to the Kb=3/2+K_b = 3/2^+ ones. This term affects the mutual displacement of IS and GS and, therefore, plays a role in the finally observed quasidegeneracy effect.

  1. The Coriolis mixing affects the total spectrum and the IS-GS displacement as well through the corresponding perturbation sum in Eq. (3). As realized in [37], the mixing directly affects the Ib,Kb=5/2+I_b,K_b = 5/2^+ GS which gets an admixture from the I=5/2+I = 5/2^+ state of the IS-based band, whereas the Ib,Kb=3/2+I_b,K_b = 3/2^+ IS remains unmixed due to the missing I=3/2+I = 3/2^+ counterpart in the yrast (GS) band. The corresponding effect of the Coriolis mixing in the GS is that it lowers the value of the GS magnetic moment. On the other hand it raises the B(M1)B(M1) and B(E2)B(E2) transition probabilities.

The assumptions above sketch the mechanism which may lead to the formation of a quasidegenerate pair of 5/2+ GS and 3/2+ IS in 229^{229}Th. We see that the very fine interplay between the involved collective and s.p. degrees of freedom is directly governed by the Woods-Saxon DSM QO deformations β2\beta_2 and β3\beta_3, the pairing strength determined by the parameters g0g_0 and g1g_1 in Eq. (8) and the Coriolis mixing strength determined by the parameter AA in (3). The remaining CQOM parameters ω\omega, bb, d0d_0, cc and pp influence the isomer energy through the overall fit of the energy spectrum, transition rates and magnetic moments. Within the above physical mechanism the 3/2+ IS of 229^{229}Th appears as an essentially s.p., i.e., microscopic, effect, the energy and electromagnetic properties of which, however, are formed under the influence of the collective dynamics of the nucleus.

In [27] the above algorithm was applied through several steps, including the choice of β2\beta_2 and β3\beta_3 in DSM based on information available for neighbouring even-even nuclei (see the beginning of next section), tuning of the pairing constants in BCS to reach a rough proximity of GS and IS and subsequent fine adjustment of the collective CQOM parameters together with the KK-mixing constant AA to obtain overall model description and predictions. It was demonstrated that at the expense of a minor deterioration of the agreement between the overall theoretical and experimental spectrum, one can exactly reproduce the IS energy of about 8 eV. Of course, such a refinement is of little practical significance since it is beyond the genuine accuracy provided by any nuclear structure model.

Few comments regarding the results in the next section should be given here in advance. We remark that some model parameters are not completely independent regarding particular physical observables. Thus, the change in the IS-GS displacement due to variation in the DSM QO deformations could be compensated by variations in the pairing constants or the KK-mixing constant AA. Therefore, one of the important issues to be clarified is the extent to which the different model parameters are correlated in the problem and how we can constrain them to reach most unambiguously the correct solution. Our numerical study showed that if we fix the pairing parameters in Eq. (8) to the values of g0=18.805g_0 = 18.805 MeV and g1=7.389g_1 = 7.389 MeV, which were tuned in the model description in Ref. [27], the further analysis and drawn conclusions also apply for the pairing strengths adopted in Refs. [19, 41]. Therefore, hereinafter we use the above fixed g0g_0 and g1g_1 parameter values while directing our study to the examination of the QO deformation space of the DSM. Another point is that in Ref. [37] the IS and GS magnetic moments were predicted without taking their experimental values into the model adjustment procedure. In the present work we include the magnetic moments into the fitting procedure by considering all observables in the fit analysis (energies, transition rates and magnetic moments) on the same footing. We will also investigate to what extent the gyromagnetic quenching factors qsq_s and qRq_R can be reasonably varied for the reproduction of the GS and IS magnetic moments. This analysis aims to reduce the arbitrariness in the model predictions for the 229Th{}^{229}\mathrm{Th} IS properties.

Numerical results and discussion

Determination of the deformed shell model deformation space

The Woods-Saxon DSM shape parameters β2\beta_2 and β3\beta_3 represent a basic input of our model and their values are decisive for the model predictions. Hereafter under “QO deformations and/or parameters” we will understand these two quantities unless otherwise specified. In Ref. [27] the quadrupole-deformation parameter β2\beta_2 was chosen by varying it between the experimental values 0.230 and 0.244 available for the neighbouring even-even nuclei 228Th{}^{228}\mathrm{Th} and 230Th{}^{230}\mathrm{Th}, respectively [51]. Simultaneously the octupole-deformation parameter β3\beta_3 was varied to obtain the GS and IS orbitals very close to each other, with leading 5/2[633]5/2[633] and 3/2[631]3/2[631] components in the respective s.p. wave-function expansions given in Eq. (6), and with positive average values of the parity ⟨πsp⟩>0\langle\pi_{\mathrm{sp}}\rangle> 0 from Eq. (7) in both s.p. states. We note that the chosen interval for the octupole deformation was at that time solely relying on model estimates. However, in the meantime we have found out that this range is also supported by an independent microscopic result. In Ref. [52] self-consistent relativistic Hartree-Bogoliubov model calculations with the universal energy density functional DD-PC1 [53] predict a rather deep total-energy minima for β3\beta_{3} between 0.1 and 0.2 in the neighboring even-even nuclei 228^{228}Th and 230^{230}Th.

In this study we identify the (β2,β3)(\beta_{2},\beta_{3}) deformation space which could provide a relevant model description of the 229m^{229\mathrm{m}}Th isomer similar to the one obtained in Refs. [27, 37] which had considered the values β2=0.240\beta_{2}=0.240 and β3=0.115\beta_{3}=0.115. To this end we have performed DSM calculations on a grid in the ranges 0.2≤β2≤0.260.2 \le\beta_{2} \le0.26 and 0≤β3≤0.150 \le\beta_{3} \le0.15 which are supposed to include the QO deformations physically relevant for a nucleus in the mass region of 229^{229}Th. At each point of the grid we obtain the KK-value and the average parity ⟨πsp⟩\langle\pi_{\mathrm{sp}}\rangle for the last occupied s.p. orbital, which is supposed to determine the GS and for the next (first) non-occupied orbital, candidate for the IS. Note that the calculation does not involve the collective (CQOM) part of the model and the only entering parameters are the two Woods-Saxon DSM deformations.

The result of this calculation is shown in Figure 1. Figure 1 (a) and Figure 1 (b) present in color coding the (β2,β3)(\beta_{2},\beta_{3}) areas

K-values for the GS and IS single-particle orbitals and respective average parities

Figure 1. KK-values for the (a) GS and (b) IS s.p. orbitals and the respective average parities (c), (d) (⟨πsp⟩\langle\pi_{\mathrm{sp}}\rangle) appearing in the DSM within the space of quadrupole and octupole deformations. The regions of relevant deformations providing the correct KGS=5/2K_{\mathrm{GS}}=5/2 and KIS=3/2K_{\mathrm{IS}}=3/2 values with ⟨πsp⟩>0\langle\pi_{\mathrm{sp}}\rangle>0 are delimitated by thick contour lines. The thinner black line in plot (a) maps the relevant border of the KIS=3/2K_{\mathrm{IS}}=3/2 region from plot (b).

in which different KK values appear for the GS and IS orbitals, respectively. For the GS orbital the K=5/2K = 5/2 value appears in two (yellow) regions, while for the IS the K=3/2K = 3/2 value appears in one narrow (blue) region. The intersection of the K=5/2K = 5/2 GS and K=3/2K = 3/2 IS subspaces coincides with the blue K=3/2K = 3/2 region for the IS and depicts the (β2,β3)(\beta_2,\beta_3)-region in which the DSM provides the required 5/2[633]5/2[633] and 3/2[631]3/2[631] orbitals for the GS and IS, respectively. Furthermore, considering the information from Figure 1 (c) and 1 (d), one can identify in a similar way the regions with positive and negative average values of the parity in the GS and IS orbitals, respectively. By retaining only the ⟨πsp⟩>0\langle\pi_{\mathrm{sp}}\rangle> 0 areas for both orbitals, one ends up with a rather limited (β2,β3)(\beta_2,\beta_3)-region given by the thick triangle contour in the four plots. This region includes all QO deformations from the considered space which are relevant within the DSM regarding the current experimental information and theoretical interpretation of the Kπ=3/2+K^\pi= 3/2^+ isomer in 229^{229}Th. Hereinafter we call this region our “model deformation space”.

Based on the above result, we can draw the following important conclusion. Considering the long-adopted KK-value and parity of the 229m^{229\mathrm{m}}Th IS, our DSM prediction shows that this isomer can only exist at essentially nonzero octupole deformation of the s.p. potential. More precisely, one can say that the coexistence of the Kπ=3/2+K^\pi= 3/2^+ IS together with the Kπ=5/2+K^\pi= 5/2^+ GS requires the presence of nonzero octupole deformation, as seen from Figure 1(a). In fact our more extended calculations in the QO deformation grid show that for β2<0.2\beta_2 < 0.2 the (yellow) range of coexisting Kπ=5/2+K^\pi= 5/2^+ GS and Kπ=3/2+K^\pi= 3/2^+ IS orbitals goes down and further reaches the β3=0\beta_3 = 0 line. However, this occurs around β2∼0.1\beta_2 \sim0.1, which is far beyond the deformation limits typical for this mass region. Thus, we can conclude that the octupole deformation appears to be of a crucial importance for the formation of the 229m^{229\mathrm{m}}Th isomer according to the present knowledge on the corresponding GS and IS angular momenta and parities.

Furthermore, we note that the deformation values (β2=0.240,β3=0.115)(\beta_2 = 0.240,\beta_3 = 0.115) used in Refs. [27, 37] appear close to the lowest vertex of the investigated DSM deformation space. The relatively small area of this space suggests a reasonable degree of arbitrariness in the model conditions imposed in the studies of Refs. [27] and [37] regarding the choice of QO deformation. Moreover, the deformation region determined in this work appears to be consistent with the corresponding areas of the QO minima in the energy surfaces of 228^{228}Th and 230^{230}Th obtained in the relativistic Hartree-Bogoliubov model calculations [52]. Nevertheless, the precise determination and prediction of the 229m^{229\mathrm{m}}Th isomer properties as well as the deeper understanding of the mechanism governing its formation requires a more detailed examination of the model descriptions obtained for various deformations fixed in the outlined model space. In the following our study is focused on this task.

A direct consequence of the location of the model

space at nonzero octupole deformation is that the GS and IS s.p. orbitals provided by the DSM always appear with mixed parity which has to be projected in the total core plus particle wave function, as seen in Eq. (20), and implemented in the model procedure applied in [27]. The average parity ⟨πsp⟩\langle\pi_{\mathrm{sp}}\rangle of both orbitals calculated from Eq. (7) as a function of the QO deformations within the model space is illustrated in Figure 2. The results show that while in the GS the quantity ⟨πsp⟩\langle\pi_{\mathrm{sp}}\rangle varies within the limits 0.37−0.460.37 - 0.46, in the IS the parity mixing is even much stronger with ⟨πsp⟩\langle\pi_{\mathrm{sp}}\rangle varying between 00 and 0.140.14. The black side of the triangle in Figure 2(b) corresponds to the (left) border of the space where the average parity of the IS turns from positive to negative values. This result shows that the mechanism governing the formation of the isomer is even more complicated due to the fine parity-mixed structure of the s.p. wave functions and the accordingly applied projection procedure.

Image not extracted

Figure 2. Average parity ⟨πsp⟩\langle\pi_{\mathrm{sp}}\rangle in the (a) GS and (b) IS s.p. orbitals appearing in DSM within the model-defined QO deformation space.

Other important quantities delivered by the DSM are the s.p. and q.p. energies for the IS orbital determined with respect to the corresponding energies in the GS orbital, E3/2+sp=Esp3/2+−Esp5/2+E^{\mathrm{sp}}_{3/2^+} = E^{3/2^+}_{\mathrm{sp}} - E^{5/2^+}_{\mathrm{sp}} and E3/2+qp=Eqp3/2+−Eqp5/2+E^{\mathrm{qp}}_{3/2^+} = E^{3/2^+}_{\mathrm{qp}} - E^{5/2^+}_{\mathrm{qp}}. The lowering of the q.p. energy with respect to the s.p. energy can be controlled through additional tuning of the pairing constants as shown in Ref. [27]. However, as argued at the end of Sec. II E, here we use the g0g_0 and g1g_1 values fixed in [27] focusing our analysis on the deformation dependencies. In Figure 3 both E3/2+spE^{\mathrm{sp}}_{3/2^+} and E3/2+qpE^{\mathrm{qp}}_{3/2^+} for the IS are plotted as functions of β2\beta_2 and β3\beta_3. As expected, they show an identical dependence but with different nominal values. In addition, along the right side of the triangle the q.p. and s.p. content of the isomer energy goes to zero, i.e. the two orbitals, 5/2[633]5/2[633] and 3/2[631]3/2[631] mutually degenerate. This is an important limit of the model deformation space. In fact the black lines in Figure 3 correspond to the crossing of both orbitals when leaving the model space to enter the blue area with K=3/2K=3/2 in Figure 1(a) and the lower K=5/2K=5/2 (yellow) area in Figure 1(b), a situation in which the GS appears with Kπ=3/2+K^\pi=3/2^+ and the IS obtains Kπ=5/2+K^\pi=5/2^+. The proximity to this line from the model space interior determines the degree of the q.p. quasidegeneracy effect. For the pair of QO deformations (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) adopted in Refs. [27, 37], the 3/2+3/2^+ q.p. energy yields E3/2+qp=2.196E^{\mathrm{qp}}_{3/2^+}=2.196 keV. We note that this is not the final IS energy in which additional contributions take a part as explained in § 2 E. The upper side of the triangle corresponds to the crossing of the Kπ=3/2+K^\pi=3/2^+ orbital with a Kπ=7/2−K^\pi=7/2^- orbital with leading component 7/2[743]7/2[743] [see red area in Figure 1(b)]. It is not of a particular interest from the isomer-formation point of view.

Image not extracted

Figure 3. (a) S.p. and (b) q.p. energy (in keV) of the 3/2+3/2^{+} isomer orbital with respect to the 5/2+5/2^{+} GS orbital appearing in DSM within the model-defined space of QO deformations. The q.p. energy is obtained with pairing parameters g0=18.805g_{0}=18.805 MeV and g1=7.389g_{1}=7.389 MeV used in Eq. (8). See text for further explanations.

Coherent quadrupole-octupole model fits in the deformed shell model deformation space

At this point we are ready to examine the behaviour of the model description and prediction for the physical observables of interest within the DSM deformation space. We are especially interested in the corresponding behaviour of the B(M1)B(M1) and B(E2)B(E2) IS transition rates and of the IS and GS magnetic moments μGS\mu_{\mathrm{GS}} and μIS\mu_{\mathrm{IS}}. To this end we have performed full model fits by adjusting the five CQOM parameters, ω\omega, bb, d0d_0, cc, pp and the Coriolis mixing constant AA with respect to the experimental quasi-parity-doublet spectrum, the available transition rates and magnetic moments at each point of the deformation space grid with the pairing constants fixed as described above. Thus, for each pair of Woods-Saxon DSM QO deformations we obtain the full spectroscopic description of the nucleus storing the quantities of interest for our systematic analysis presented below.

The five adjusted CQOM parameters show a smooth behavior along the deformation space with values consistent with those obtained in Ref. [27]. We therefore refrain from addressing further numerical details here. We note however that the parameter pp defined in Eq. (11) is close to unity p≈1p\approx1 throughout the entire DSM (β2,β3)(\beta_2,\beta_3)-model space. This parameter indicates the relative contribution of the quadrupole and octupole modes in the CQOM. Thus, according to Eq. (14) all parameter fits lead to a practically circular bottom of the CQOM potential with β~2,3sa≈β~2,3sa\tilde{\beta}_{2,3}^{\mathrm{sa}}\approx\tilde{\beta}_{2,3}^{\mathrm{sa}}, showing that the model describes the collective quasi-parity-doublet structure of the 229^{229}Th spectrum with equal weights of the quadrupole β~2\tilde{\beta}_2 and octupole β~3\tilde{\beta}_3 deformation modes. Considering the DSM deformation parameters β2=0.240\beta_2=0.240 and β3=0.115\beta_3=0.115 used in Ref. [27], we obtain for the 5/2+5/2^+ GS and 3/2+3/2^+ IS states values between 0.12 and 0.17 for the two equal β~2,3sa\tilde{\beta}_{2,3}^{\mathrm{sa}} semiaxes (which define the circle radius of the CQOM potential bottom).

Inspection of the odd-nucleon contribution to β~2,3sa\tilde{\beta}_{2,3}^{\mathrm{sa}} determined in Eq. (13) through Eq. (3) at (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) shows that in the case of 229^{229}Th the Coriolis mixing causes a negligible decrease in β~2,3sa\tilde{\beta}_{2,3}^{\mathrm{sa}} compared with the core case, Eq. (12), while a considerable decrease is caused by the term (−Kb2)(-K_b^2). Thus while in the core+particle case the GS β~2,3sa=0.122\tilde{\beta}_{2,3}^{\mathrm{sa}}=0.122, for the core only [without the term (−Kb2)(-K_b^2)] these values become 0.149. Similarly, in the isomeric state the core+particle β~2,3sa=0.121\tilde{\beta}_{2,3}^{\mathrm{sa}}=0.121, while in the core case the semiaxes values rise to 0.132. Figure 4 illustrates the CQOM QO wave-function densities ∣ΦnKπqo(β~2,β~3)∣2|\Phi_{nK}^{\pi\mathrm{qo}}(\tilde{\beta}_2,\tilde{\beta}_3)|^2 from Eq. (15) for the Iπ=5/2+I^\pi=5/2^+ GS and its negative-parity counterpart 5/2−5/2^{-} obtained with the parameters of the CQOM fit at (β2,β3)=(0.240,0.115)(\beta_{2},\beta_{3})=(0.240,0.115). For simplicity we have only taken into account the (−Kb2)(-K_{b}^{2}) term in X^(Iπ,Kb)\hat{X}(I^{\pi},K_{b}) dropping the negligible Coriolis mixing term. The wave function density was calculated for the quenching parameter set (qS,qR)=(0.6,0.6)(q_{\mathrm{S}},q_{\mathrm{R}})=(0.6,0.6). The dynamical deformations are indicated by the positions of the density maxima in the Figure. We see that in the two states these deformations appear outside of the potential bottom circle. Furthermore, the dynamical deformation parameters are not coinciding with the DSM (β2,β3)(\beta_{2},\beta_{3}) parameters, confirming the discussion in Sec. II B on the distinction between dynamical deformation parameters in the CQOM and the DSM deformation parameters. We also note that the CQOM potential bottom semiaxes β~2,3sa\widetilde{\beta}_{2,3}^{\mathrm{sa}} (the red circles) do not change between the positive- and negative-parity counterparts in the quasi-parity-doublet since the Coriolis mixing term only mixes states with the same (bandhead) parity. This situation, however, would be different in a spectrum build on the Kb=1/2K_{b}=1/2 bandhead (which is not present for the case of 229^{229}Th) where the decoupling term in X^(Iπ,Kb)\hat{X}(I^{\pi},K_{b}) in Eq. (3) would act in opposite directions on the semiaxes lengths of the opposite-parity counterparts.

We have checked that the density plots for the Iπ=3/2+I^{\pi}=3/2^{+} IS and its 3/2−3/2^{-} quasi-parity-doublet counterpart (not given here) look very similar to those in Figure 4. To investigate the effect of higher angular momentum values we show the CQOM QO wave-function densities for the I=15/2±I=15/2^{\pm} states of the yrast quasi-parity-doublet in Figure 5. The CQOM potential semiaxes β~2,3sa\widetilde{\beta}_{2,3}^{\mathrm{sa}} and the corresponding dynamical deformations considerably increase with angular momentum, reaching at I=15/2I=15/2 values larger than 0.2. This shows that the dynamical deformation is responsible for the higher-energy part of the spectrum which otherwise would not be felt by the s.p. potential. We may conclude that the model algorithm rather carefully takes into account also the influence of the collective dynamics at the higher angular momenta, which reflects on the overall deformation characteristics of the CQOM potential.

Finally, special attention is given below to the behaviour of the Coriolis mixing constant AA which, as already mentioned, plays an important role in the formation of the IS energy and electromagnetic properties. The calculations were repeated for four pairs of (qS,qR)(q_S,q_R) values of the quenching factors considered for the spin and collective gyromagnetic rates. As the analysis of magnetic moments made in Ref. [37] suggests the need of rather strong attenuation of the latter, here we consider qSq_S and qRq_R with slightly lower values compared with the lowest pair (qS,qR)=(0.6,0.6)(q_S,q_R)=(0.6,0.6) considered in Ref. [37]. Thus, in the present calculations the gyromagnetic quenching factors were allowed to be as small as (qS,qR)=(0.55,0.45)(q_S,q_R)=(0.55,0.45). Although the experimental value of the isomeric energy is obviously out of reach for the model accuracy, we also consider its theoretical prediction E(3/2+)ISE(3/2^+)_{\mathrm{IS}} in order to see how the model fits “feel” its tiny energy scale as well as to assess accordingly the relevance of the overall model description for the different deformations within the model space.

Energy description

The primary quantity providing overall information about the relevance of the model descriptions in the different points of the deformation space is the root-mean-square (RMS) deviation between the theoretical and experimental energy levels. Its behaviour as a function of the DSM QO deformation for calculations made with quenching factors (qS,qR)=(0.6,0.6)(q_S,q_R)=(0.6,0.6) is illustrated in Figure 6. We indicate with an open star the location of the deformations (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115), adopted in Refs. [27, 37]. The RMS value obtained in this point is about 34 keV which is the same as the value obtained in the original model fits performed in [27], although now the experimental values of the magnetic moments $\mu_{\mathrm{GS}} and μIS\mu_{\mathrm{IS}} are included in the fits. We note that the RMS factor is close to this value over a larger area of the deformation model space, demonstrating the stability of the model solutions with the variation of QO deformations. The upper part of the space with large β2\beta_2 and β3\beta_3 values, however, appears unfavoured. We have verified that in all regions of the space with the RMS close to 34 keV the description of the overall energy spectrum and the available B(M1)B(M1) and B(E2)B(E2) transition rates is of similar accuracy as the one reported in Ref. [27], with the obtained CQOM parameter values being close to those in Ref. [27] (see Figure 1 and Table 1 therein). We notice that in the upper-left parts of the plot some lower RMS deviations are obtained as low as ≈30\approx30 keV, however, for these deformations the model predictions for the isomer energy are less favourable as analyzed below. In addition, we found (barely visible in Figure 6) that towards the line of the 5/2+[633]5/2^{+}[633]–3/2+[631]3/2^{+}[631] degeneracy the RMS factor sharply increases. As discussed below in relation to the particular observables, this is the result of the strong increase of the Coriolis KK-mixing interaction which largely exceeds the perturbation theory limitation and puts a constraint on the model description valid close to the 5/2+5/2^{+}–3/2+3/2^{+} orbital crossing.

Energy RMS values in keV for the GS and IS bands

Figure 6. Energy RMS values in keV for the GS (yrast) and IS (excited) bands together obtained by the model fit on a grid within the model-defined QO deformation space. We use qS=qR=0.6q_S = q_R = 0.6. The open star indicates the location of the deformations (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) adopted in Refs. [27, 37].

In Figure 7 the theoretical isomer energy values EISE_{\mathrm{IS}} obtained by the model fits on the DSM QO space grid are presented for two sets of gyromagnetic quenching values (qs,qR)=(0.6,0.6)(q_s,q_R)=(0.6,0.6) and (0.55,0.45)(0.55,0.45). We find that for the first set the value obtained at the pair of deformations (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) is EIS≈1E_{\mathrm{IS}}\approx1 keV, whereas for the second set it is EIS≈0.3E_{\mathrm{IS}}\approx0.3 keV. For the second set we choose to demonstrate the result for one more pair of deformations from our grid (β2,β3)=(0.2398,0.1108)(\beta_2,\beta_3)=(0.2398,0.1108), further on denoted for simplicity by the rounded values (0.240,0.111)(0.240,0.111), situated closer to the degeneracy line. There we have EIS≈0.040E_{\mathrm{IS}}\approx0.040 keV already approaching the scale of the experimental value. We note that for this pair of deformations the 3/2+3/2^{+} q.p. energy yields E3/2+qp=0.188E^{\mathrm{qp}}_{3/2^{+}}=0.188 keV. All EISE_{\mathrm{IS}} values shown in Figure 7 are obtained in the fitting procedure on the same footing without particular refinement. As already mentioned, one can easily achieve the exact experimental value of 0.008 keV through a very fine tuning of model parameters (e.g. the KK-mixing AA), with a minimal deterioration of the description in the remaining energy levels (see also Ref. [27]). The plots in Figure 7 show that in the large areas of the model space the fits provide reasonable values of EISE_{\mathrm{IS}} which could be renormalized to the experiment in this manner. However, we also see that in the upper-left parts of the plots the EISE_{\mathrm{IS}} considerably increases up to 7-8 keV giving an indication that at these deformations the remoteness of the 5/2[633]5/2[633] and 3/2[631]3/2[631] orbitals (see Figure 3) already makes it difficult for the model mechanism to achieve quasidegeneracy. Also, we notice a thin stripe with large EISE_{\mathrm{IS}} values along the line of degeneracy which obviously indicates the limitation of the perturbation theory. Concluding this part, our analysis of the RMS factor and isomer energy values outlines certain limits of reliability of the present model application and favours the lower vertex of the model space around the deformation set used in Ref. [27] in reasonable proximity to the 5/2+5/2^{+}–3/2+3/2^{+} orbitals crossing line.

Isomer energy maps for $^{229}$Th for two combinations of $q_S$ and $q_R$

Figure 7. Isomer energy EISE_{\mathrm{IS}} obtained by the model fits on a model-defined QO deformation space grid for two different combinations of qSq_S and qRq_R. The open star indicates the location of the deformations (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) adopted in Refs. [27, 37]. The circle indicates the set (β2,β3)=(0.240,0.111)(\beta_2,\beta_3)=(0.240,0.111) situated closer to the degeneracy line.

D B(M1) and B(E2) isomer transition rates

The results obtained for the isomeric B(M1;3/2IS+→5/2GS+)B(M1;3/2^{+}_{\mathrm{IS}}\rightarrow5/2^{+}_{\mathrm{GS}}) and B(E2;3/2IS+→5/2GS+)B(E2;3/2^{+}_{\mathrm{IS}}\rightarrow5/2^{+}_{\mathrm{GS}}) transition rates are illustrated in Figure 8 and Figure 9, respectively, for the four sets of quenching factors. As in the energy analysis above, here we mark with an open star the value obtained by the model at the pair of QO deformations (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115), adopted in Refs. [27, 37], for which the original model predictions for the transition rates and magnetic moments were made. Also, we examine the model predictions towards the line of 5/2+5/2^{+}–3/2+3/2^{+} orbitals degeneracy by considering in the graphs of (qs,qR)=(0.55,0.45)(q_s,q_R)=(0.55,0.45) the pair of deformations (β2,β3)=(0.240,0.111)(\beta_2,\beta_3)=(0.240,0.111) indicated by the open circle. Inspecting Figure 8, first we observe that the overall behaviour of the B(M1;3/2IS+→5/2GS+)B(M1;3/2^{+}_{\mathrm{IS}}\rightarrow5/2^{+}_{\mathrm{GS}}) transition value shows an increase with the approaching of the degeneracy line. This is due to the circumstance that with the decreasing distance between both orbitals 5/2+5/2^{+} and 3/2+3/2^{+}, the KK-mixing effect generated by the perturbation element in Eq. (5) sharply increases and this leads to the increase in the connecting transition rates. It should be noted, however, that in the model procedure this increase is counterbalanced by the adjustable parameter AA which drops accordingly thus preventing a deterioration of the model description due to the excessive mixing force. This will be discussed in more detail in the following (see Figure 13 and related text below). Keeping in mind this clarification, we notice in Figure 8(d) that while the quenching of the gyromagnetic factors leads to a reduction of B(M1)ISB(M1)_{\mathrm{IS}} to 0.005 W.u. at (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) (an effect already addressed in Ref. [37]), the shift of the deformation towards the degeneration line returns the value back to 0.007 W.u., i.e. in the original range of the prediction made in Ref. [27].

Image not extracted

Figure 8. B(M1)B(M1) isomer transition values obtained by the model fits on a grid within the model-defined QO deformation space at four different combinations of qSq_{\mathrm{S}} and qRq_{\mathrm{R}}. The open star indicates the location of the deformations (β2,β3)=(0.240,0.115)(\beta_{2},\beta_{3})=(0.240,0.115) adopted in Refs. [27, 37]. The circle in panel (d) indicates the set (β2,β3)=(0.240,0.111)(\beta_{2},\beta_{3})=(0.240,0.111) situated closer to the degeneracy line.

B(E2) isomer transition values on the QO deformation space grid

Figure 9. B(E2)B(E2) isomer transition values obtained by the model fits on the model-defined QO deformation space grid at four different combinations of qSq_S and qRq_R. The open star indicates the location of the deformations (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) adopted in Refs. [27, 37]. The circle in panel (d) indicates the set (β2,β3)=(0.240,0.111)(\beta_2,\beta_3)=(0.240,0.111) situated closer to the degeneracy line.

Values of the $K$-mixing constant $A$ on the QO deformation space grid

Figure 13. Values of the KK-mixing constant AA obtained by the model fits on the model-defined QO deformation space grid at two different combinations of qsq_s and qRq_R. The open star and the circle indicate the same sets of deformations as shown in Figure 10.

A similar behaviour of the B(E2;3/2IS+→5/2GS+)B(E2;3/2^{+}_{\mathrm{IS}}\rightarrow5/2^{+}_{\mathrm{GS}}) transition rate in the model deformation space is observed in Figure 9. We note that here all obtained B(E2)ISB(E2)_{\mathrm{IS}} values exceed the previous prediction [27] but stay in the range of the correction suggested in Ref. [37].

E Ground-state and isomer magnetic moments

The calculated GS magnetic moment μGS\mu_{\mathrm{GS}} for the four sets of quenching factors (qs,qR)(q_s,q_R) is illustrated in Figure 10. The overall model behaviour of this quantity is such that it decreases both with the attenuation of the gyromagnetic factors and with the approaching of the 5/2+5/2^{+}–3/2+3/2^{+} orbital-degeneracy line. We see that for (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) it drops to 0.45 μN\mu_N, while with the shift of the deformations to the point (0.240,0.111)(0.240,0.111) it reaches the value of 0.43 μN\mu_N. The latter result is obviously due to the increasing Coriolis mixing which reduces the value of μGS\mu_{\mathrm{GS}} as has been shown already in Ref. [37]. However, it appears that for model conditions considered physically reasonable, this is still not sufficient to reproduce the newer experimental value of 0.360(7) μN\mu_N [38], although the model reproduces fairly well the earlier measured value of 0.46(4) μN\mu_N [39].

GS magnetic moment values obtained by the model fits on a grid within the model-defined QO deformation space at four different combinations of $q_s$ and $q_R$

Figure 10. GS magnetic moment values obtained by the model fits on a grid within the model-defined QO deformation space at four different combinations of qsq_s and qRq_R. The open star and the circle indicate the same sets of deformations as shown in Figure 8 and Figure 9.

It is instructive to check here also the “bare” values of μGS\mu_{\mathrm{GS}}, i.e. those obtained from the pure s.p. wave function without including the Coriolis mixing. Therefore, in Figure 11 we show the analog of Figure 10(a) (with qs=qR=0.6q_s=q_R=0.6) in which μGS\mu_{\mathrm{GS}} is calculated in the absence of Coriolis mixing with the KK-mixing constant A=0A=0. In this case Eq. (30) reduces to the terms in its first line with the second term involving the expression of Eq. (31). Here we first see that μGS\mu_{\mathrm{GS}} appears with considerably larger values in the limits μGS=0.55\mu_{\mathrm{GS}}=0.55–0.60 μN0.60\,\mu_N which also show different behaviour in the DSM QO space compared with the Coriolis-mixing case. This result does not depend on the model-parameters fit and illustrates the genuine contribution of the QO deformation for the formation of the GS magnetic moment of 229Th{}^{229}\mathrm{Th}. Comparing both plots we see that in the pure s.p. case without Coriolis mixing, the lowest μGS=0.55 μN\mu_{\mathrm{GS}}=0.55\,\mu_N value appears in the left-upper vertex of the triangle model space, whereas in the Coriolis-mixing case the low values (lower than the pure s.p. ones), appear in the lower vertex of the space.

Image not extracted

Figure 11. The “bare” s.p. GS magnetic moment values obtained without Coriolis mixing (A=0A=0) on a grid within the model-defined QO deformation space with qs=qR=0.6q_s=q_R=0.6. The open star indicates the location of the deformations (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) adopted in Refs. [27, 37]. The full star indicates the grid point which provides the lowest value μGS=0.55 μN\mu_{\mathrm{GS}}=0.55\,\mu_N.

The above result leads us to the following conclusions. The Coriolis effect causes a decrease of the nuclear magnetic moment in the 229Th{}^{229}\mathrm{Th} GS throughout the model deformation space. It plays a considerable role in our approach for fixing the GS magnetic moment through the overall model fits, although this is still not enough to reproduce the latest adopted experimental value. The appearance of essentially lower μGS\mu_{\mathrm{GS}}-values obtained through the adjusted KK-mixing constant AA compared with the corresponding pure s.p. μGS\mu_{\mathrm{GS}}-values shows that the increase of the model-controlled Coriolis-mixing towards the 5/2+5/2^{+}–3/2+3/2^{+} orbitals-degeneracy line essentially determines the behaviour of the GS magnetic moment and dominates over the corresponding effect of changing QO deformation on the pure s.p. μGS\mu_{\mathrm{GS}}-values. This conclusion suggests that no considerably different result can be reached through further variation of deformations parameters in the DSM QO space.

Figure 12 shows the calculated IS magnetic moment μIS\mu_{\mathrm{IS}} for the four sets of gyromagnetic quenching factors considered (qs,qRq_s,q_R). We note its relatively flat behaviour as a function of the QO deformation, with a slight increase towards the degeneracy line. Since μIS\mu_{\mathrm{IS}} is practically not affected by the KK-mixing effect, we may claim that this dependence can be considered as the bare effect of the changing structure of the s.p. wave functions along the deformation space. We see that in all plots of Figure 12 the lowest value of μIS\mu_{\mathrm{IS}} appears in the left upper corner of the model space similarly to the “bare” (s.p.) μGS\mu_{\mathrm{GS}} case (Figure 11). For example in the case of qs=qR=0.6q_s=q_R=0.6, Figure 12(a), the corresponding lowest value μIS=−0.36 μN\mu_{\mathrm{IS}}=-0.36\ \mu_N comes closer to the experimental result. However, the model fits have shown that the lower corner provides better predictions for μGS\mu_{\mathrm{GS}}, and we keep our attention on this region. Besides, for all considered quenching factors (qs,qR)(q_s,q_R), the theoretical μIS\mu_{\mathrm{IS}} values appearing in Figure 12 enter the error bars of the recent experimental value of −0.37(6) μN-0.37(6)\ \mu_N reported in Ref. [16, 17].

Four panels showing IS magnetic moment values over the model-defined QO deformation space grid

Figure 12. IS magnetic moment values obtained by the model fits on the model-defined QO deformation space grid at four different combinations of qsq_s and qRq_R. The open star and the circle indicate the same sets of deformations as shown in Figure 10.

The results presented so far already reveal important details and relations characterizing the model mechanism upon which the 229mTh^{229m}\mathrm{Th} isomer is formed and its spectroscopic properties develop. Obviously the proximity of the 5/2[633]5/2[633] and 3/2[631]3/2[631] s.p. orbitals plays a major role providing the overall condition for the appearance of a low-lying excitation. Now this is clearly quantified by all above plots. On the other hand, it is also clear that the appearance of the isomer can not be due only to the orbital quasidegeneracy. The reason is that at the distance of few eV the mixing between the two orbitals becomes very large and pushes all related observables in unphysical regions of magnitude. In this case the perturbation terms in the centrifugal part, Eq. (3), of the Hamiltonian as well as in the Coriolis perturbed wave function, Eq. (21), collapse in a singularity. The model prevents this situation mostly through the K2K^2 term and KK-mixing constant AA in Eq. (3), which allow us to properly situate both GS and IS, i.e., to obtain the IS energy value as small as necessary, by keeping the 3/2[631]3/2[631]–5/2[633]5/2[633] orbital distance aside from the degeneracy line. In this respect one can say that the physically adequate QO deformations are slightly aside from this line.

The model mechanism feature described above can be seen by following the behaviour of the Coriolis mixing constant AA adjusted at each grid point in the deformation space. We investigate this in Figure 13 for two sets of quenching factors, (qs,qR)=(0.6,0.6)(q_s,q_R)=(0.6,0.6) and (0.55,0.45)(0.55,0.45). The obtained values of AA range from zero to approximately 0.50.5 keV. Towards the degeneracy line, where the KK-mixing matrix element in Eq. (5) connecting the two orbitals sharply increases as they approach each other, the adjustment algorithm strongly reduces the value of AA. In this way the model “feels” the growing magnitude of the Coriolis mixing and tries to compensate its excessive effect on the considered observables through the parameter AA keeping them in physically meaningful ranges. Providing this balancing role of the parameter AA and having in mind all so far obtained model patterns for the spectroscopic observables in 229Th^{229}\mathrm{Th} we can be rather confident in the consistency of the analysis made and the reliability of the QO deformation region outlined.

Finally, it is interesting to identify the degree of spin and collective gyromagnetic factor attenuations required to reproduce in the present model both experimental μGS\mu_{\mathrm{GS}} and μIS\mu_{\mathrm{IS}} values. This is shown in Figure 14, where the values of each of these quantities obtained in the model fits at (β2,β3)=(0.240,0.115)(\beta_2,\beta_3)=(0.240,0.115) are given [Figure 14(a) for μGS\mu_{\mathrm{GS}} and Figure 14(b) for μIS\mu_{\mathrm{IS}}] as functions of the quenching factors qsq_s and qRq_R. The black lines denote the pairs of (qs,qR)(q_s,q_R) values which provide the corresponding μGS\mu_{\mathrm{GS}} and μIS\mu_{\mathrm{IS}} experimental values. The crossing of both lines shows the point at which both magnetic moments are reproduced together. We see that this occurs at qs≈0.52q_s\approx0.52 and a rather low value of qR≈0.22q_R\approx0.22, which corresponds to a quite strong attenuation of the collective gyromagnetic factor. Because this quenching magnitude is hard to justify, we conclude that the agreement between the present theoretical model and the currently adopted experimental value of the GS magnetic moment remains an open issue.

GS and IS magnetic moment values obtained by the model fits on a grid for the spin-gyromagnetic and rotation-gyromagnetic attenuation factors

Figure 14. (a) GS and (b) IS magnetic moment values obtained by the model fits on a grid for the spin-gyromagnetic, qSq_{\mathrm{S}}, and rotation-gyromagnetic, qRq_{\mathrm{R}}, attenuation factors. The black lines show the values that would reproduce the corresponding GS and IS experimental values.

Summary and conclusion

In this work we have thoroughly examined the physical conditions for the formation of the 8 eV isomer of 229^{229}Th according to the model mechanism suggested by our QO vibration-rotation core plus particle approach. First, we have determined the model deformation space encompassing the Woods-Saxon DSM quadrupole and octupole deformations which allow the appearance of the GS and IS with the experimentally adopted KK-values and parities. We were able to clearly identify its borders constrained by the average parity of the isomer and the crossings of the 3/2[631]3/2[631] orbital on which the IS is built with the 5/2[633]5/2[633] orbital of the GS as well as with a 7/2[743]7/2[743] orbital. This space is rather limited and essentially constrains the QO deformation in the s.p. potential within the ranges 0.235≤β2≤0.2550.235 \leq\beta_{2} \leq0.255 and 0.11≤β3≤0.140.11 \leq\beta_{3} \leq0.14. These results lead us to the important conclusion that the appearance of the Kπ=3/2+K^{\pi}=3/2^{+} IS through this mechanism is only possible in the presence of essentially nonzero octupole deformation in the s.p. potential.

Furthermore, we have examined the dependence of the overall DSM+CQOM model description on the QO deformations within the DSM model space. Our analysis of the CQOM fits made in Sec. III B, including the CQOM potential-bottom semiaxes and dynamical deformations, showed that the collective CQOM conditions under which the 229m^{229\mathrm{m}}Th isomer is formed consistently interrelate with the relevant conditions provided by the odd-nucleon degrees of freedom. Under these overall conditions we obtain for all of the considered observables, B(M1;3/2IS+→5/2GS+)B(M1; 3/2^{+}_{\mathrm{IS}} \rightarrow5/2^{+}_{\mathrm{GS}}), B(E2;3/2IS+→5/2GS+)B(E2; 3/2^{+}_{\mathrm{IS}} \rightarrow5/2^{+}_{\mathrm{GS}}), μGS\mu_{\mathrm{GS}}, μIS\mu_{\mathrm{IS}} and E(3/2+)ISE(3/2^{+})_{\mathrm{IS}} a smooth behaviour of the model predictions and descriptions compared with the results obtained for the fixed pair of Woods-Saxon DSM QO deformations considered in our previous works [27, 37], with the only peculiarity appearing close to the line of 5/2+5/2^{+}–3/2+3/2^{+} degeneracy where the mixing of both orbitals exceeds the perturbation theory limits. On the other hand the corresponding behaviour of the model energy RMS factors and Coriolis mixing constant shows that descriptions obtained with values of the above observables essentially deviating from those obtained in [27, 37] are of lower quality and/or violate the perturbation limit. For the remaining descriptions we have verified that in the limits of moderate deviations of the considered observables from the original values in [27, 37], the model is renormalizable, so that through a small variation of model parameters and on the expense of small deteriorations of the RMS factor, we can get very similar model predictions for the different pairs of QO deformations. Using this model feature as well as assuming possible stronger attenuation of the spin and collective gyromagnetic factors, we have outlined rather narrow limits of arbitrariness in which the model values of each of the above quantities can vary by keeping its reasonable physical meaning and predictability.

Our main conclusion is that within the obtained model deformation space the applied QO core plus particle approach provides a rather constrained prediction for the most important 229^{229}Th energy and electromagnetic characteristics related to the formation and manifestation of the 8 eV isomer. This allows us to generally reconfirm the predictions initially made in Ref. [27, 37] and to summarize them with a slight update: The B(M1)B(M1) IS transition remains in the limits 0.006-0.008 W.u. with an open possibility towards lower values such as 0.005 W.u.; the B(E2)B(E2) IS transition may be considered with slightly higher values between 30 and 50 W.u., compared with those in Ref. [27]; the GS magnetic moment allows a limited possibility for variation and remains with a model value around 0.50 μN0.50\ \mu\mathrm{N} possibly getting values as smaller as 0.430.43-0.48 μN0.48\ \mu\mathrm{N} under a stronger assumption for the gyromagnetic factors attenuation, thus covering the old experimental value of Ref [39], but still overestimating the newer one of Ref. [38]; the theoretical IS magnetic moment firmly reproduces the recent experimental value within the uncertainty limits reported in [16, 17] and this is obtained under all considered model conditions; and finally the model values for the isomer energy typically obtained around 1 keV and below suggest that with a small variation of parameters, and with the expense of a minor deterioration of the other energy levels, the model can easily reproduce the experimental value, although this is of little importance due to the overall limitation of the model accuracy in the energy-spectrum description.

We note that for all of the above quoted values the other model observables (energy levels and transition rates), for which experimental data are available, remain described within the accuracy limits reported in Ref. [27]. Thus our analysis suggests that within the above outlined limits of arbitrariness the model predictions could provide reliable estimates for the 229mTh^{229\mathrm{m}}\mathrm{Th} spectroscopic characteristics which could serve to the experiment in further efforts to observe and control the yet elusive radiative isomer transition.

Finally, the results obtained confirm the relevance of the model mechanism emphasizing on the role of the fine interplay between nuclear collective and intrinsic degrees of freedom as a plausible reason for the isomer formation. On this basis we conclude that the same dynamical mechanism may govern also in other nuclei the formation of excitations close to the border of atomic physics energy scale. Such states may exist being not yet observed due to experimental difficulties similar to those encountered in 229mTh^{229\mathrm{m}}\mathrm{Th}. As in this work we give a detailed prescription about the examination and constraining of the physical conditions under which such a phenomenon may emerge, it appears promising to extend the study to other nuclei in the same or other mass regions. In this aspect the close neighbour 231Th^{231}\mathrm{Th} as well as the 235mU^{235\mathrm{m}}\mathrm{U} isomer would be natural candidates for such a study. This could be a subject of future work.

ACKNOWLEDGMENTS

This work is supported by the Bulgarian National Science Fund (BNSF) under Contract No. KP-06-N48/1. AP gratefully acknowledges funding by the EU FET-Open project 664732 (nuClock). This work is part of the ThoriumNuclearClock project that has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement No. 856415).

References

References

  1. [1]B. R. Beck, J. A. Becker, P. Beiersdorfer, G. V. Brown, K. J. Moody, J. B. Wilhelmy, F. S. Porter, C. A. Kilbourne, and R. L. Kelley, Phys. Rev. Lett. 98, 142501 (2007), URL https://link.aps.org/doi/10.1103/PhysRevLett.98.142501.
  2. [2]Improved Value for the Energy Splitting of the Ground-State Doublet in the Nucleus ²²⁹mTh, vol. LLNL-PROC-415170 (Varenna, Italy, 2009).
  3. [3]B. Seiferle et al., Nature (London) 573, 243 (2019).
  4. [4]A. Yamaguchi et al., Phys. Rev. Lett. 123, 222501 (2019).
  5. [5]E. Browne and J. K. Tuli, Nuclear Data Sheets 122, 205 (2014).
  6. [6]E. Peik and Chr. Tamm, Europhys. Lett. 61, 181 (2003), URL https://doi.org/10.1209/epl/i2003-00210-x.
  7. [7]E. Peik and M. Okhapkin, Comptes Rendus Physique 16, 516 (2015), ISSN 1631-0705, the measurement of time / La mesure du temps, URL http://www.sciencedirect.com/science/article/pii/S1631070515000213.
  8. [8]C. J. Campbell, A. G. Radnaev, A. Kuzmich, V. A. Dzuba, V. V. Flambaum, and A. Derevianko, Phys. Rev. Lett. 108, 120802 (2012), URL https://link.aps.org/doi/10.1103/PhysRevLett.108.120802.DOI
  9. [9]V. V. Flambaum, Phys. Rev. Lett. 97, 092502 (2006), URL https://link.aps.org/doi/10.1103/PhysRevLett.97.092502.DOI
  10. [10]J. C. Berengut, V. A. Dzuba, V. V. Flambaum, and S. G. Porsev, Phys. Rev. Lett. 102, 210801 (2009), URL https://link.aps.org/doi/10.1103/PhysRevLett.102.210801.DOI
  11. [11]W. G. Rellergert, D. DeMille, R. R. Greco, M. P. Hehlen, J. R. Torgerson, and E. R. Hudson, Phys. Rev. Lett. 104, 200802 (2010), URL https://link.aps.org/doi/10.1103/PhysRevLett.104.200802.DOI
  12. [12]P. Fadeev, J. C. Berengut, and V. V. Flambaum, Phys. Rev. A 102, 052833 (2020), URL https://journals.aps.org/pra/abstract/10.1103/PhysRevA.102.052833.
  13. [13]E. V. Tk alya, Phys. Rev. Lett. 106, 162501 (2011), URL https://link.aps.org/doi/10.1103/PhysRevLett.106.162501.DOI
  14. [14]L. von der Wense, B. Seiferle, M. Laatiaoui, J. B. Neumayr, H.-J. Maier, H.-F. Wirth, C. Mokry, J. Runke, K. Eberhardt, C. E. Düllmann, et al., Nature 533, 47 (2016), ISSN 0028-0836, article, URL http://dx.doi.org/10.1038/nature17669.
  15. [15]B. Seiferle, L. von der Wense, and P. G. Thirlof, Phys. Rev. Lett. 118, 042501 (2017), URL https://link.aps.org/doi/10.1103/PhysRevLett.118.042501.DOI
  16. [16]J. Thielking, M. V. Okhapkin, P. Glowacki, D. M. Meier, L. von der Wense, B. Seiferle, C. E. Düllmann, P. G. Thirlof, and P. Peik, Nature (London) 556, 321 (2018).
  17. [17]R. A. Müller, A. V. Maiorova, S. Fritzsche, A. V. Volotka, R. Beerwerth, P. Glowacki, J. Thielking, D.-M. Meier, M. Okhapkin, E. Peik, et al., Phys. Rev. A 98, 020503 (2018), URL https://link.aps.org/doi/10.1103/PhysRevA.98.020503.DOI
  18. [18]T. Sikorsky, J. Geist, D. Hengstler, S. Kempf, L. Gastaldo, C. Enss, C. Mokry, J. Runke, C. E. Düllmann, P. Woubrasek, et al., Phys. Rev. Lett. 125, 142503 (2020), URL https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.125.142503.
  19. [19]S. G. Nilsson and I. Ragnarsson, Shapes and Shells in Nuclear Structure (Cambridge University Press, Cambridge, 1995).DOI
  20. [20]K. Gulda, W. Kurcewicz, A. Aas, M. Borge, D. Burke, B. Fogelberg, I. Grant, E. Hagebø, N. Kafrell, J. Kvasil, et al., Nuclear Physics A 703, 45 (2002), ISSN 0375-9474, URL http://www.sciencedirect.com/science/article/pii/S0375947401014567.
  21. [21]E. Ruchowska, W. A. Plociennik, J. Zylicz, H. Mach, J. Kvasil, A. Algora, N. Amzal, T. Bäck, M. G. Borge, R. Boutami, et al., Phys. Rev. C 73, 044326 (2006).
  22. [22]S. V. G., Theory of Complex Nuclei (Pergamon Press, Oxford, 1976).
  23. [23]A. M. Dykhne and E. V. Tkaya, JETP Lett. 67, 251 (1998).
  24. [24]E. V. Tkaya, C. Schneider, J. Jeet, and E. R. Hudson, Phys. Rev. C 92, 054324 (2015), URL https://link.aps.org/doi/10.1103/PhysRevC.92.054324.DOI
  25. [25]G. Alaga, K. Alder, A. Bohr, and B. Mottelson, Kgl. Danske Videnskab. Selskab Mat.-fys. Medd. 29 (No. 9), 1 (1955).
  26. [26]S. G. Nilsson, Kgl. Danske Videnskab. Selskab Mat.-fys. Medd. 29 (No. 16), 1 (1955).
  27. [27]N. Minkov and A. Pálffy, Phys. Rev. Lett. 118, 212501 (2017).
  28. [28]N. Minkov, P. Yotov, S. Drenska, W. Scheid, D. Bonatsos, D. Lenis, and D. Petrellis, Phys. Rev. C 73, 044315 (2006), URL https://link.aps.org/doi/10.1103/PhysRevC.73.044315.DOI
  29. [29]N. Minkov, S. Drenska, P. Yotov, S. Lalkovski, D. Bonatsos, and W. Scheid, Phys. Rev. C 76, 034324 (2007), URL https://link.aps.org/doi/10.1103/PhysRevC.76.034324.DOI
  30. [30]N. Minkov, S. Drenska, M. Strecker, W. Scheid, and H. Lenske, Phys. Rev. C 85, 034306 (2012), URL https://link.aps.org/doi/10.1103/PhysRevC.85.034306.DOI
  31. [31]N. Minkov, S. Drenska, K. Drumev, M. Strecker, H. Lenske, and W. Scheid, Phys. Rev. C 88, 064310 (2013), URL https://link.aps.org/doi/10.1103/PhysRevC.88.064310.DOI
  32. [32]S. Cwiok, J. Dudek, W. Nazarewicz, J. Skalski, and T. Werner, Computer Physics Communications 46, 379 (1987), ISSN 0010-4655, URL http://www.sciencedirect.com/science/article/pii/0010465587900932.
  33. [33]P. Ring and P. Schuck, The Nuclear Many-Body Problem (Springer Verlag, New York, 1980).DOI
  34. [34]J. Jeet, C. Schneider, S. T. Sullivan, W. G. Rellergert, S. Mirzadeh, A. Cassanho, H. P. Jensen, E. V. Tkalya, and E. R. Hudson, Phys. Rev. Lett. 114, 253001 (2015), URL https://link.aps.org/doi/10.1103/PhysRevLett.114.253001.DOI
  35. [35]A. Yamaguchi, M. Kolbe, H. Kaser, T. Reichel, A. Gottwald, and E. Peik, New Journal of Physics 17, 053053 (2015), URL http://stacks.iop.org/1367-2630/17/i=5/a=053053.
  36. [36]J. von der Wense L., On the direct detection of 229mTh (Springer Thesis Series, Springer International Publishing, 2018).DOI
  37. [37]N. Minkov and A. Pálffy, Phys. Rev. Lett. 122, 162502 (2019).
  38. [38]M. S. Safronova, U. I. Safronova, A. G. Radnaev, C. J. Campbell, and A. Kuzmich, Phys. Rev. A 88, 060501(R) (2013), URL https://link.aps.org/doi/10.1103/PhysRevA.88.060501.DOI
  39. [39]S. Gerstenkorn, P. Luc, J. Verges, D. W. Engelkeimer, J. E. Gindler, and F. S. Tomkins, J. Phys. (Paris) 35, 483 (1974).
  40. [40]R. R. Crashman, I. Ahmad, A. M. Friedman, and J. R. Erskine, Rev. Mod. Phys. 49, 833 (1977), URL https://link.aps.org/doi/10.1103/RevModPhys.49.833.
  41. [41]P. Walker and N. Minkov, Physics Letters B 694, 119 (2010), ISSN 0370-2693, URL http://www.sciencedirect.com/science/article/pii/S0370269310011378.
  42. [42]N. Minkov, Physica Scripta T154, 014017 (2013), URL http://stacks.iop.org/1402-4896/2013/i=T154/a=014017.
  43. [43]N. Minkov, S. Drenska, M. Strecker, and W. Scheid, Journal of Physics G: Nuclear and Particle Physics 37, 025103 (2010), URL http://stacks.iop.org/0954-3899/37/i=2/a=025103.
  44. [44]B. R. Mottelson, Selected Topics in the Theory of Collective Phenomena in Nuclei (Int. School of Phys., Varenna, Nordita Publ. no. 78, 1960).
  45. [45]K. Way, Phys. Rev. 55, 963 (1939), URL https://link.aps.org/doi/10.1103/PhysRev.55.963.DOI
  46. [46]E. Bodenstedt, Fortsch. Physik 10, 321 (1962).
  47. [47]J. Eisenberg and W. Greiner, Nuclear Theory: Nuclear Models, Vol. I (North-Holland, Amsterdam, 1970).
  48. [48]S. G. Nilsson and O. Prior, Kgl. Danske Videnskab. Selskab Mat.-fys. Medd. 32 (No. 16), 1 (1961).
  49. [49]O. Prior, F. Boehm, and S. G. Nilsson, Nucl. Phys. A 110, 257 (1968).
  50. [50]W. Greiner, Phys. Rev. Lett. 14, 599 (1965), URL https://link.aps.org/doi/10.1103/PhysRevLett.14.599.DOI
  51. [51]S. Raman, C. W. Nestor, Jr., and P. Tikkanen, At. Data Nucl. Data Tables 78, 1 (2001), ISSN 0092-640X, URL http://www.sciencedirect.com/science/article/pii/S0092640X01908587.
  52. [52]K. Nomura, D. Vretenar, T. Nikšić, and B.-N. Lu, Phys. Rev. C 89, 024312 (2014), URL https://link.aps.org/doi/10.1103/PhysRevC.89.024312.DOI
  53. [53]T. Nikšić, D. Vretenar, and P. Ring, Phys. Rev. C 78, 034318 (2008), URL https://link.aps.org/doi/10.1103/PhysRevC.78.034318.DOI

Paper details

vlnc:W_ttr4u8vh
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