WorksPhysics and Astronomy
Pfaffian and minor formulas in number-conserved pairing theory
Abstract
In this work, I present closed-form formulas for the norm and many-body density matrices between wave functions with exact particle numbers in pairing theory. The expressions, written as sums of Pfaffians and minors, apply to both even and odd particle-number systems and to pair condensates as well as broken-pair configurations. This formalism facilitates calculations using the generator coordinate method and symmetry restoration techniques, such as angular momentum projection. I apply these formulas to Ge and Ge using the angular-momentum-projected number-conserved BCS theory. The calculations describe low-lying collective bands and provide an interpretation of their intrinsic structures.
Introduction
Pairing theories, such as the Bardeen-Cooper-Schrieffer (BCS) theory and the Hartree-Fock-Bogoliubov (HFB) approach, have been widely used in theoretical studies of quantum many-body systems. While BCS and HFB approaches break U(1) symmetry, this violation has a minimal impact in large systems treated within the grand canonical ensemble. However, for finite-size systems, such as atomic nuclei, ultra-cold atoms, and ultrasmall superconducting grains, particle-number conservation is crucial for accurately capturing intrinsic properties and quantum phase transitions.
To restore particle-number conservation, one common approach is particle number projection, which involves numerical integration over the gauge angle [1, 2]. Alternatively, methods that directly conserve particle number offer a more straightforward description, for example, shape evolution and quantum phase transitions in rare-earth nuclei have been successfully described using number-conserved pair condensates in a unified framework [3, 4]. An -pair condensate is defined as:
where is the number of pairs, is the vacuum state, and is the Cooper pair creation operator defined in a canonical basis of dimension as
where is the pair coefficient. The energy of the pair condensate can be computed based on recursive formulas derived from the generalized Wick theorem and the commutation relations between Cooper pairs [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]. Nevertheless, computations in number-conserved pairing theory are generally complex.
Angular-momentum projection is required when the BCS or HFB wave function does not have good angular momentum. This is done by integration over the Euler angles or by linear algebra projection [16, 17]. For instance, for angular momentum projection onto a pair condensate, the Hamiltonian and norm matrices can be written as
written as
where is a Wigner- matrix element and is the rotation operator.
This work presents formulas for evaluating norms and matrix elements of many-body operators between wave functions with exact particle numbers, valid for both even and odd particle numbers. The formulas include configurations with broken pairs and matrix elements between different pair condensates in different canonical bases. They can therefore be used for configuration mixing in the generator coordinate method.
Formalism
Starting from Eq. (1), I extend the formalism to a more general broken-pair state with particles in a canonical basis:
where , and is the number of unpaired particles, typically referred to as the seniority number in nuclear physics [18, 19]. Eq. (4) applies to both even and odd particle numbers. The canonical basis can be expressed as a unitary transformation of an orthonormal reference basis:
where is a unitary matrix. The rotation operator transforms the canonical basis as
where is the transpose of the Wigner- matrix for single-particle states. Using Eqs. (2), (4), and (6), the norm between two different broken-pair states can be written as
where , and is the generalized Kronecker delta: () if are distinct and form an even (odd) permutation of , and otherwise. In obtaining the second line, the sum over involving the generalized Kronecker delta and the product of matrix elements of yields a minor of . denotes the minor, i.e., the determinant of the submatrix formed by selecting rows and columns of . This minor represents the overlap between the Slater determinants formed from the selected canonical orbitals of the states and . The summation runs over all -element combinations from , reflecting the pair correlation. Eq. (7) and related formulas can also be derived using Grassmann algebra [20, 21]. The details of this alternative derivation are omitted here for brevity.
The norm can be expressed in an alternative form. To derive this, I rewrite the Cooper pair creation operator in Eq. (2) as
where the matrix , and
The equivalence of Eqs. (2) and (8) can be obtained from Refs. [22, 23]. Under rotation, the Cooper pair creation operator becomes
where . The norm is then expressed as
where represents an -element combination from , and the summation runs over all such combinations. Each combination must satisfy
The term denotes the sign of the permutation , and denotes the Pfaffian of the submatrix of whose rows and columns are indexed by . For , Eq. (11) reduces to the Pfaffian expression for the overlap between pair condensates [24]. It is worth mentioning that in the pioneering work of Ref. [25], Robledo introduced a Pfaffian formula to resolve the sign problem in HFB wave functions. Following this approach, substantial efforts have been devoted to improving HFB calculations using Pfaffian formulas, as demonstrated in Refs. [24, 26, 27, 28, 29].
Next, I present a formula for the matrix element of a many-body operator between broken-pair states. An -body operator is defined as
Using Eqs. (2), (4), (6), and (13), the matrix element of the -body operator is expressed in a compact form:
In obtaining the second line, I have used the identity:
Similar to the norm formulas, an alternative expression for the matrix element of the -body operator is presented. Using Eqs. (4), (6), (8) and (10), the matrix element can be written as
where the set is defined as
and the sets and are defined as
Eqs. (16) and (11) are equivalent to Eqs. (14) and (7), respectively.
As a consistency check, I consider the -pair condensate defined in Eq. (1), in a doubly degenerate Hartree–Fock basis with the pairing Hamiltonian:
Using Eqs. (7) and (14), the norm and Hamiltonian matrix element are
For degenerate single-particle states with constant pairing strength ( and ), the pair coefficients become uniform, and the norm in Eq. (20) reaches its maximum value for a normalized pair: . This value is smaller than the norm of an -boson condensate due to the Pauli principle, and approaches in the limit . Using Eq. (21), the energy in this case is , which recovers the result obtained from other approaches, such as quasi-spin theory [30].
Application of PNBCS to 70Ge and 69Ge
One of the applications of the present formalism is the study of collective states using the angular-momentum-projected number-conserved BCS (PNBCS) method [3, 4, 31]. In this work, I consider both the even-even nucleus and the odd-mass nucleus in the space above the core, using the JUN45 effective interaction [32]. The PNBCS calculation is performed in three steps. First, I solve the self-consistent Hartree–Fock (HF) equation with the Kramers time-reversal symmetry for under the JUN45 interaction, without enforcing any other constraints like shapes or orientations. The solutions are called “intrinsic” HF states. One usually obtains multiple distinct HF solutions. Then, for a given HF basis I consider the pair condensate in Eq. (1), where the Cooper pair coefficients are determined by solving the number-conserved BCS (NBCS) equation [10, 11, 31]. Finally, the energies and wave functions are obtained by angular momentum and parity projection, i.e., solving the Hill–Wheeler equation.
For , the projected space mixes two pair-condensate structures obtained from an axially symmetric oblate HF solution with and a triaxial solution with . The deformation parameters are calculated following the convention of Ref. [4]. The Hill–Wheeler equation can be written as
where and label the two intrinsic HF configurations, and is the parity. For comparison, full-configuration-interaction (FCI) shell-model calculations are performed using BIGSTICK [33, 34], with the same interaction and single-particle space. To calculate electric-quadrupole transition strengths , I use the effective charges and .
Figure Figure 1 compares the PNBCS and FCI spectra of with experiment [35], and Table I lists the corresponding values. Here only the ground, , and bands are presented. The collective bands are identified using relatively larger values. I exemplify this with the ground-band state. Ref. [35] reports that the state at 4.431 MeV feeds the ground-band state with W.u., while the yrast state at 4.204 MeV feeds the with only W.u. Thus the former is assigned as the ground-band state in this work. Similarly, in FCI the state at 4.597 MeV feeds the ground-band with W.u., while the yrast at 3.927 MeV feeds it with only W.u. The PNBCS calculation shows that the ground and bands arise predominantly from the triaxial HF configuration, whereas the band is mainly associated with the oblate HF configuration. It is interesting to see that while the -band state is MeV higher than the ground-band , the oblate HF solution lies 0.08 MeV lower than the triaxial one. The ground-band state gains more energy from angular-momentum projection and configuration mixing.
Figure 1. Excitation energies of the ground, , and bands in from experiment [35], the FCI, and the PNBCS. The experimental ground band shows the state with the larger to and its candidate; the experimental spin assignment remains uncertain. The experimental band is adopted from [35]. But the collective interpretation of the , , and states is uncertain, as discussed in the text.
The experimental band shows odd–even staggering with close – and – pairs, indicating -soft character. The FCI and PNBCS spectra instead show the opposite staggering regularity with –, –, and – pairs; the staggering is much weaker in FCI. Nevertheless, the FCI and PNBCS calculations show similar patterns of in-band values, with stronger transitions from odd-spin states and suppressed ones from even-spin states (see Table I). For the transition, Ref. [35] adopts the mixing ratio measured by Robinson et al. [37], corresponding to predominantly transition. The later
measurement of Sugawara et al. [38] also regarded it as a dipole transition. However, the experimental result of Dostmohes et al. [39], listed as an alternative in Ref. [35], gives and implies predominantly transition. Thus the adopted W.u. remains uncertain. Using the calculated and values together with the experimental energy of 0.744 MeV, I obtain and 1.06 for FCI and PNBCS, respectively. The calculations use the free-nucleon orbital factors and spin factors with a quenching factor of 0.7. Both calculations predict comparable and contributions.
The experimental band in Figure 1 is adopted from Ref. [35]. The state at 3.059 MeV decays through the at 2.157 MeV to the excited , which was confirmed in Ref. [40]. However, the measured W.u. [35] is much smaller than the FCI and PNBCS values of 10.59 and 13.81 W.u., respectively. This weak transition does not support a simple collective interpretation for the . The extension of this band to the and at 4.104 and 5.436 MeV was not confirmed in Ref. [40]. Among the 31 FCI states below 6 MeV, the state at 4.661 MeV has the largest W.u. to the -band state at 2.898 MeV. The PNBCS state at 5.422 MeV feeds its -band state with 13.96 W.u. Both calculated states lie above the proposed experimental . Further experimental study is required.
| Band | Ref. [35] | Ref. [36] | FCI | PNBCS | |
| ground | 20.65 | 25.93 | |||
| 33.82 | 36.79 | ||||
| 39.60 | 46.33 | ||||
| 26.03 | 47.70 | ||||
| 23.08 | 45.95 | ||||
| 41.57 | 48.31 | ||||
| 10.15 | 13.53 | ||||
| 5.89 | 3.38 | ||||
| 21.02 | 25.22 | ||||
| 23.70 | 25.37 | ||||
| 13.59 | 19.93 | ||||
| 0.03 | 0.34 | ||||
| 24.51 | 31.38 | ||||
| 13.02 | 19.40 | ||||
| 17.07 | 19.67 | ||||
| 0.59 | 0.92 | ||||
| 10.12 | 32.78 | ||||
| 3.50 | 19.91 | ||||
| 11.38 | 11.43 | ||||
| 10.59 | 13.81 | ||||
| 4.22 | 13.96 | ||||
| to ground | 0.66 | 0.49 | |||
| 31.72 | 37.63 | ||||
| 1.35 | 1.02 | ||||
| to | 1.67 | 0.63 | |||
| to ground | 22.84 | 11.56 | |||
| 0.05 | 0.27 | ||||
| ground to | 0.63 | 0.81 | |||
| to | 0.70 | 0.01 |
Table 1. values in Weisskopf units (W.u.) in . transitions are listed for the ground and bands, and both and for the band. The experimental data are taken from Ref. [35] and the recent measurement of Ref. [36]. The calculations use effective charges and . The adopted value depends on conflicting experimental E2/M1 mixing ratios; see the discussion in the text.
For , the many-body basis is constructed by adding one neutron to the two pair condensates of . These pair condensates are obtained by NBCS optimization in the bases of two triaxial HF solutions with and , respectively, which describe the shape coexistence phenomenon in well [31]. Thus the many-body basis for corresponds to in Eq. (4). This nucleus tests the capability of PNBCS to describe low-lying collective states
| Band | Expt. | FCI | PNBCS | |
| 14.31 | 12.64 | |||
| 18.76 | 13.83 | |||
| 18.77 | 39.30 | |||
| 15.14 | 41.07 | |||
| 21.59 | ||||
| 6.48 | 11.24 | |||
| 13.00 | 20.27 | |||
| 10.29 | 22.73 | |||
| 6.32 | 30.28 | |||
| 10.20 | 12.14 | |||
| 12.11 | 12.41 | |||
| 18.50 | 14.53 | |||
| 8.86 | 10.14 | |||
| 12.22 | ||||
| 15.39 | 17.47 | |||
| 15.17 | 20.49 | |||
| 11.61 | 19.30 | |||
| 17.52 | 16.53 | |||
| 15.23 | 18.12 | |||
| 11.87 | ||||
| (2) | 18.39 | 23.00 | ||
| 23.38 | 27.68 | |||
| 27.85 | 23.44 | |||
| 18.72 | 35.79 |
Table 2. (W.u.) in . The experimental data are taken from Ref. [41].
in odd-mass nuclei. Figure 2 compares the calculated collective bands in with experiment [41]. Both FCI and PNBCS identify six low-lying bands: the , , , , first , and second bands. No experimental levels are assigned to the second band because the transition data are insufficient. For the band, both calculations predict the energy close to the experimental value of 3.605 MeV. The measured supports the assignment of the state to the band (see Table II). In the experimental band, the states at 2.149 and 2.258 MeV decay to the state by E2 transitions. Their values are unknown, so they are shown as alternative candidates. For the experimental band, the measured supports the assignment of the state. The available data do not establish higher-spin candidates for these experimental bands.
Figure 2. Six selected collective bands in Ge: the , , , , first , and second bands. No experimental level is assigned to the second band. The two experimental states in the band are alternative candidates. Experimental data are from Ref. [41].
The experimental, FCI, and PNBCS results agree well for the lower-lying excitation energies and values (see Table II). Discrepancies appear at higher levels. In FCI, the values support separate extensions of the two bands. In PNBCS, the state at 3.898 MeV feeds the states of the first and second bands with and 23.44 W.u., respectively. The larger value supports its assignment to the second band. For the band, the FCI states at 4.882 and 4.859 MeV decay to the state with and 4.38 W.u., respectively. The state with the larger is shown as a tentative band candidate in Figure 2. PNBCS gives larger in-band values than FCI for some higher-spin states. Configuration mixing beyond the adopted configuration space might redistribute the strength among nearby states. The model space can be enlarged by including more intrinsic pair structures and/or higher-seniority configurations.
To examine the structures of the low-lying collective bands in Ge, I label the two intrinsic pair condensates of Ge as even-even cores 1 and 2, in the order given above. For each core , I couple the unpaired neutron to that pair condensate and project onto the chosen . The subspace spanned by the projected states is denoted by . For a normalized PNBCS eigenstate of Ge, , I calculate
where is the projection operator onto the subspace . Thus estimates how well can be represented by configurations constructed from intrinsic core . Note that the two core subspaces and are generally nonorthogonal, so need not equal unity.
Summary
The calculation shows a change of core overlaps with increasing spin in the band. The and states have larger overlaps with core 1, with and , respectively. becomes at and at . The similar feature appears in the and bands. For the band, changes from at the bandhead to at . For the band, it changes from at the bandhead to at . The lower-spin positive-parity states have comparable overlaps with both core spaces. For example, for the bandheads of the , first , and second bands, , , and , respectively. These states are not fully described using either core alone, indicating that mixing of the two core configurations is important.
To summarize, this work presents closed-form formulas for the norm and matrix elements of many-body operators between number-conserved pairing states. The bra and ket can contain unpaired particles and have different canonical bases. The expressions are derived using the generalized Kronecker delta and written in terms of minors and Pfaffians. The formalism is applied to the study of low-lying collective bands in and using PNBCS with the JUN45 effective interaction, demonstrating its applicability to systems with and without unpaired particles. For , the calculated ground and bands arise from the triaxial intrinsic pair condensate, while the band corresponds to the axially symmetric oblate structure. For , the calculated low-spin negative-parity states can be described mainly by coupling the unpaired neutron to the intrinsic even-even core with , while mixing with configurations based on the other core with is important for the positive-parity bandheads.
Acknowledgments
The author thanks Prof. Y. Lu and Y. X. Yu for checking some of the test calculations, and Prof. Calvin W. Johnson, Prof. Y. M. Zhao, and Prof. C. Qi for valuable discussions. This research was supported by the National Natural Science Foundation of China under Grant Nos. 12322506, 12535009, and 12075169, and by the Fundamental Research Funds for the Central Universities.
References
- [1]P. Ring and P. Schuck, The nuclear many-body problem (Springer Science & Business Media, 2004).
- [2]K. Dietrich, H. J. Mang, and J. H. Pradal, Phys. Rev. 135, B22 (1964).
- [3]S. T. Guo, Y. X. Yu, C. W. Johnson, S. Pittel, Y. Lu, H. Jiang, Y. Lei, G. J. Fu, and Z. Z. Ren, Phys. Rev. C 112, L011301 (2025).
- [4]S. T. Guo, Y. X. Yu, C. W. Johnson, S. Pittel, H. Jiang, Z. Z. Ren, and G. J. Fu, Phys. Rev. C 112, 054323 (2025).
- [5]I. Talmi, Simple models of complex nuclei: the shell model and the interacting boson model (CRC Press, 1993).
- [6]I. Talmi, Nucl. Phys. A 172, 1 (1971).
- [7]J.-Q. Chen, B.-Q. Chen, and A. Klein, Nucl. Phys. A 554, 61 (1993).
- [8]J.-Q. Chen, Nucl. Phys. A 562, 218 (1993).
- [9]N. Sandulescu, B. Errea, and J. Dukelsky, Phys. Rev. C 80, 044335 (2009).
- [10]L. Y. Jia, Phys. Rev. C 96, 034313 (2017).
- [11]L. Y. Jia, Phys. Rev. C 99, 014302 (2019).
- [12]B. C. He, L. Li, Y. A. Luo, Y. Zhang, F. Pan, and J. P. Draayer, Phys. Rev. C 102, 024304 (2020).
- [13]Y. Lei, Y. Lu, and Y. M. Zhao, Chin. Phys. C 45, 054103 (2021).
- [14]Y. Lu, Y. Lei, C. W. Johnson, and J. J. Shen, Phys. Rev. C 105, 034317 (2022).
- [15]C. Ma, X. Yin, and Y. M. Zhao, Phys. Rev. C 108, 034308 (2023).
- [16]C. W. Johnson and K. D. O’Mara, Phys. Rev. C 96, 064304 (2017).
- [17]C. W. Johnson and C. Jiao, J. Phys. G: Nucl. Part. Phys. 46, 015101 (2019).
- [18]G. Racah, Phys. Rev. 62, 438 (1942).
- [19]G. Racah, Phys. Rev. 63, 367 (1943).
- [20]J. W. Negele and H. Orland, Quantum Many-Particle Systems (Addison-Wesley, Redwood City, 1988).
- [21]T. G. Stryzhak, Grassmann Algebra and Determinant Theory (Ibidem, Stuttgart, 2010).
- [22]D. C. Youla, Canad. J. Math. 13, 694 (1961).
- [23]B. Zumino, J. Math. Phys. 3, 1055 (1962).
- [24]T. Mizusaki, M. Oi, and N. Shimizu, Phys. Lett. B 779, 237 (2018).
- [25]L. M. Robledo, Phys. Rev. C 79, 021302 (2009).
- [26]L. M. Robledo, Phys. Rev. C 84, 014307 (2011).
- [27]G. F. Bertsch and L. M. Robledo, Phys. Rev. Lett. 108, 042505 (2012).
- [28]Z.-C. Gao, Q.-L. Hu, and Y. S. Chen, Phys. Lett. B 732, 360 (2014).
- [29]B. G. Carlsson and J. Rotureau, Phys. Rev. Lett. 126, 172501 (2021).
- [30]M. Ichimura, On the Quasispin Formalism (Pergamon, Oxford, 1968).
- [31]Y. X. Yu, Y. Lu, G. J. Fu, C. W. Johnson, and Z. Z. Ren, Phys. Rev. C 106, 044309 (2022).
- [32]M. Homma, T. Otsuka, T. Mizusaki, and M. Hjorth-Jensen, Phys. Rev. C 80, 064323 (2009).
- [33]C. W. Johnson, W. E. Ormand, and P. G. Krastrev, Comput. Phys. Commun. 184, 2761 (2013).
- [34]C. W. Johnson, W. E. Ormand, K. S. McElvain, and H. Shan, arXiv preprint arXiv:1801.08432 (2018).arxiv.org/abs/1801.08432
- [35]G. Gürdal and E. A. McCutchan, Nucl. Data Sheets 136, 1 (2016).
- [36]T. M. Kowalewski, A. D. Ayangeakaa, N. Sensharma, R. W. F. Janssens, Y. M. Wang, Q. B. Chen, J. M. Allmond, C. M. Campbell, S. Carmichael, M. P. Carpenter, P. Copp, C. Cousins, M. Devlin, U. Garg, C. Müller-Gatermann, T. J. Gray, D. J. Hartley, J. Heery, J. Henderson, H. Jayatissa, S. R. Johnson, S. P. Kisyov, F. G. Kondev, T. Lauritsen, S. Nandi, R. Rathod, W. Reviol, M. Rocchini, E. Rubino, R. Russell, A. Saracino, D. Seweryniak, M. Siciliano, and C. Y. Wu, Phys. Rev. C 112, 034332 (2025).
- [37]R. L. Robinson, H. J. Kim, R. O. Sayer, J. C. Wells, Jr., R. M. Ronningen, and J. H. Hamilton, Phys. Rev. C 16, 2268 (1977).
- [38]M. Sugawara, Y. Toh, M. Oshima, M. Koizumi, A. Kimura, A. Osa, Y. Hatsukawa, H. Kusakari, J. Goto, M. Honma, M. Hasegawa, and K. Kaneko, Phys. Rev. C 81, 024309 (2010).
- [39]G. A. Dostemsova, D. K. Kaipov, and Yu. G. Kosyak, Izv. Akad. Nauk SSSR, Ser. Fiz. 51, 1928 (1987).
- [40]R. A. Haring-Kaye, S. I. Morrow, J. Döring, S. L. Tabor, K. Q. Le, P. R. P. Allegro, P. C. Bender, R. M. Elder, N. H. Medina, J. R. B. Oliveira, and V. Tripathi, Phys. Rev. C 97, 024308 (2018).
- [41]C. D. Nesaraja, Nucl. Data Sheets 207, 1 (2026).

