betaFeedback

WorksPhysics and Astronomy

Saturation of Nuclear Binding from Lattice Hamiltonians

Abstract

There is a conundrum regarding the binding of αα particles in nuclei. On one hand, auxiliary-field Monte Carlo simulations of Hamiltonians on discrete spatial lattices proposed that attractive two-nucleon potentials, alone or together with attractive three-nucleon potentials, yield accurate nuclear binding. On the other hand, such Hamiltonians typically overbind all but the lightest nuclei in continuum-space approaches. We address this puzzle by performing Hartree-Fock computations of the light nuclei 4^4He, 8^8Be, 12^{12}C, and 16^{16}O, and of nuclear and neutron matter using established lattice Hamiltonians. These variational upper bounds for the ground-state energies show that the Hamiltonians with only two-nucleon potentials do not yield accurate binding, in contrast to the results from auxiliary-field Monte Carlo simulations. The case is different for Hamiltonians with three-nucleon potentials although it is the dense packing on the lattice – and not repulsive potentials – that yield a constant binding energy per nucleon.

Introduction.— Nuclear Hamiltonians from effective field theories of quantum chromodynamics [1, 2, 3, 4, 5, 6, 7, 8, 9], the fundamental theory of the strong force, consist of two- and three-body interactions and are rather complicated. There is a scale-dependent balance between attractive and repulsive forces and a complex spin and isospin structure. At next-to-next-to-leading order in the Weinberg power counting, Hamiltonians from chiral effective field theory have about 15 unknown low-energy constants. While such Hamiltonians have been used to predict [10, 11, 12, 13] and understand [14, 15, 16, 17] properties of atomic nuclei, there have also been efforts to understand key elements of nuclear structure using much simpler Hamiltonians [18, 19, 20, 21, 22].

The accurate reproduction of nuclear saturation, i.e., binding energies of about 8 MeV per nucleon and charge radii that scale as A1/3A^{1/3} for a nucleus with mass number AA is a key test of nuclear Hamiltonians. This implies that the equation of state for nuclear matter with equal parts neutrons and protons has a minimum at a density of about 0.16 fm−30.16\ \mathrm{fm}^{-3} (which is approximately the central density of stable medium-mass nuclei) and at an energy of about −16-16 MeV per nucleon. Accurate nuclear saturation is essential for reproducing bulk properties of nuclei, and for this reason many nuclear Hamiltonians are optimized to nuclear matter properties or bulk properties of medium-mass nuclei [5, 23, 24, 25, 26].

In this work we focus on interactions from effective field theories of quantum chromodynamics that are formulated on discrete lattices. At leading order of chiral effective field theory [1, 2] the two-body potential consists of the one-pion exchange and two ss-wave contacts. There are no three-nucleon forces at this order. Specific formulations are presented in Refs. [19, 20]. At leading order of pionless effective field theory, the potential consists of two-body ss-wave contacts and a three-body contact [27]. A specific formulation was given in Ref. [21]. These lattice formulations use two- and three-body contacts that are spin/isospin symmetric and consist of products of densities with nonlocal and local smearing; details are presented below.

Auxiliary-field Monte Carlo simulations of these lattice Hamiltonians found (i) that simple two-body interactions can yield accurate binding energies and radii for nuclei up to oxygen [20], (ii) that nuclear binding and α\alpha-particle clustering is sensitive to the mixture of local and nonlocal smearing in the contact interactions [19], and (iii) that attractive two- and three-body potentials yield essential elements of nuclear binding [21].

These lattice results are somewhat surprising because the two-body lattice interactions of Refs. [19, 20] are very soft (the lattice spacing of a=1.97a=1.97 fm corresponds to a momentum cutoff of π/a≈314\pi/a \approx314 MeV) and overall attractive in nature. The quantitative description of nuclei using Hamiltonians with only attractive two-body forces is at odds with computations that use Hamiltonians formulated in continuum space. There, interactions from chiral effective field theory at leading order do not bind α\alpha-particles into nuclei [9] (at least at momentum cutoffs of 450 and 500 MeV), and phase-shift equivalent two-body Hamiltonians alone generally fail to saturate at accurate densities [28] or overbind medium-mass nuclei [29, 30]. Three-nucleon interactions are critical for an accurate description of nuclear structure [31, 32, 33, 34, 35], and this is particularly so for nucleon-nucleon interactions with a low momentum cutoff [28, 29]. It is the purpose of this Letter to resolve this conundrum.

We note that the lattice computations of Refs. [19, 20, 21] escaped scrutiny so far because alternative ab initio methods [36, 37, 38, 39, 40, 41, 42, 43, 44] employ the harmonic oscillator basis or work in the continuum [45]. However, the recent publication of NuLattice [46], a publicly available Python package for ab initio computations on lattices, allows us now to compare with auxiliary-field Monte Carlo simulations of lattice Hamiltonians.

Lattice Hamiltonians

We employ a discrete three-dimensional spatial lattice as the single-particle basis and use Hamiltonians from Refs. [19, 20, 21]. The papers [19, 20] used a lattice spacing of 1.97 fm and Hamiltonians from chiral effective field theory at leading order. These consist of the kinetic energy TT, the one-pion exchange term VOPEV_{\mathrm{OPE}}, and a short-range term V0V_0. In Ref. [20] the Hamiltonian is

H=T+VOPE+V0.(1)H = T + V_{\mathrm{OPE}} + V_0 . \tag*{(1)}

Expressions for the kinetic energy and the one-pion exchange are given in [link text](End Matter). Here we discuss the short-range potential V0V_0 in more detail and follow Refs. [19, 20].

Using a combined spin and isospin index ii, one defines a smeared annihilation operator on the lattice site n\mathbf{n} as

a~i(n)=ai(n)+sNL∑∣n−n′∣=1ai(n′).(2)\widetilde{a}_i(\mathbf{n}) = a_i(\mathbf{n}) + s_{\mathrm{NL}} \sum_{\lvert\mathbf{n}-\mathbf{n}'\rvert=1} a_i(\mathbf{n}') . \tag*{(2)}

One then defines the smeared density operator as

ρ~(n)=∑i[a~i†(n)a~i(n)+sL∑∣n−n′∣=1a~i†(n′)a~i(n′)].(3)\widetilde{\rho}(\mathbf{n}) = \sum_i \left[ \widetilde{a}_i^\dagger(\mathbf{n})\widetilde{a}_i(\mathbf{n}) + s_{\mathrm{L}} \sum_{\lvert\mathbf{n}-\mathbf{n}'\rvert=1} \widetilde{a}_i^\dagger(\mathbf{n}')\widetilde{a}_i(\mathbf{n}') \right] . \tag*{(3)}

The short-range potential of Ref. [20] is then simply

V0=c02∑n:ρ~(n)2:.(4)V_0 = \frac{c_0}{2} \sum_{\mathbf{n}} :\widetilde{\rho}(\mathbf{n})^2: . \tag*{(4)}

Here, the colons indicate normal ordering. We note that the potential is nonlocal, i.e., it does not act multiplicative, when sNL≠0s_{\mathrm{NL}} \ne0. The parameters are c0=−0.185c_0=-0.185, and sNL=sL=0.08s_{\mathrm{NL}}=s_{\mathrm{L}}=0.08, see Ref. [20]. The potential (4) is in lattice units and needs to be multiplied with ℏ/a=100\hbar/a=100 MeV in actual computations.

The Hamiltonian of Ref. [19] also has the form (1) but employs different short-range two-body potentials. We see from Eq. (3) that ρ~(n)=ρ~(sNL,sL;n)\widetilde{\rho}(\mathbf{n})=\widetilde{\rho}(s_{\mathrm{NL}},s_{\mathrm{L}};\mathbf{n}). Thus one can consider potentials made from local densities (where sNL=0s_{\mathrm{NL}}=0) and from nonlocal densities (where sL=0s_{\mathrm{L}}=0). Furthermore, one can insert spin and isospin operators acting on these densities. The upshot is that Ref. [19] proposed a Hamiltonian

HB=T+VOPE+VL+VNL.(5)H_B = T + V_{\mathrm{OPE}} + V_{\mathrm{L}} + V_{\mathrm{NL}} . \tag*{(5)}

where VLV_{\mathrm{L}} and VNLV_{\mathrm{NL}} denote local and nonlocal short-range potentials, respectively, that also contain spin-isospin operators. A key result of that work was that a purely nonlocal interaction fails to bind α\alpha particles into nuclei while the Hamiltonian HBH_B yields accurate results for light nuclei. Details of the Hamiltonian HBH_B are presented in [link text](End Matter).

Finally, we follow Ref. [21] and use a Hamiltonian from pion-less effective field theory. It has the form

Hπ ⁣ ⁣ ⁣/=T+V0+W.(6)H_{\pi\!\!\!/} = T + V_0 + W . \tag*{(6)}

Here,

W=c33!∑n:ρ~(n)3:.(7)W = \frac{c_3}{3!} \sum_{\mathbf{n}} :\widetilde{\rho}(\mathbf{n})^3: . \tag*{(7)}

is the three-body potential. For this Hamiltonian, the lattice spacing is a=1.32a=1.32 fm (corresponding to a momentum

Nucl.HH from Ref. [20]HBH_B from Ref. [19]Exp.
HFAFMCHFAFMC
4He{}^{4}\mathrm{He}−19.81-19.81−25.4-25.4−32.19-32.19−29.2-29.2−28.3-28.3
8Be{}^{8}\mathrm{Be}−63.17-63.17−51.9-51.9−76.37-76.37−59.7-59.7−56.5-56.5
12C{}^{12}\mathrm{C}−137.28-137.28−83.8-83.8−139.83-139.83−95.0-95.0−92.2-92.2
16O{}^{16}\mathrm{O}−213.37-213.37−128.2-128.2−222.06-222.06−135.4-135.4−127.6-127.6

Table 1. Variational upper bounds from Hartree-Fock (HF) for ground-state energies (in MeV) for various nuclei and the Hamiltonians HH [from Eq. (1)] and HBH_B [from Eq. (5)] compared to the results from auxiliary-field Monte Carlo (AFMC) simulations [20] and [19] that used the same Hamiltonians, respectively, and experiment [47].

cutoff of π/a≈471\pi/a \approx471 MeV), and the constants of the potential are sNL=0.5s_{\mathrm{NL}=0.5}, sL=0.061s_{\mathrm{L}}=0.061, c0=−3.41×10−7c_0=-3.41\times10^{-7} MeV−2^{-2}, and c3=−1.4×10−14c_3=-1.4\times10^{-14} MeV−5^{-5}. (In numerical computations, the last two constants need to be multiplied with three and six powers of ℏ/a=150\hbar/a=150 MeV, respectively.) We note that the two- and three-body potentials are both attractive and that the nonlocal smearing sNLs_{\mathrm{NL}} is about an order of magnitude larger than what was used in Refs. [19, 20]. We also note already at this point that the auxiliary-field Monte Carlo simulations of Ref. [21] used a temporal lattice spacing (i.e., a time step) of at=1/(1000a_t=1/(1000 MeV) in the computations with the Hamiltonian (6) while the Hamiltonians (1) and (5) of Refs. [19] and [20], respectively, were solved with a much larger at=1/(150a_t=1/(150 MeV).

Finite nuclei.—We start with Hartree-Fock computations of finite nuclei based on the Hamiltonian (1) from chiral effective field theory at leading order, using the parameterizations and lattice (L=6L=6) of Ref. [20]. Our computations start from simple localized initial states (i.e., compact clusters of four nucleons on neighboring sites) and solve the Hartree-Fock equations self consistently. Our initial states are too compact, and the Hartree-Fock iterations relax the density. At fixed LL the resulting Hartree-Fock energies are variational upper bounds on the exact ground-state energy, which would require additional correlations beyond the mean-field level to compute.

Our results for the Hamiltonian HH of Ref. [20] and Eq. (1) are summarized in Table I and compared to auxiliary-field Monte Carlo simulations of the same publication. The Hartree-Fock energy of 4He{}^{4}\mathrm{He} is a plausible upper bound. For heavier nuclei, however, the variational bounds from Hartree Fock are significantly lower than the results from auxiliary-field Monte Carlo [20]. The nuclei 8Be{}^{8}\mathrm{Be}, 12C{}^{12}\mathrm{C}, and 16O{}^{16}\mathrm{O} are overbound with 7.9, 11.4, and 13.3 MeV binding energy per particle, respectively. We see no sign of saturation, and we are far from the nuclear binding energies of about 8 MeV per nucleon one observes in nature.

We repeat the Hartree-Fock computations for the Hamiltonian HBH_B of Eq. (5) with the parameterization and lattice (L=6L=6) of Ref. [19]. Results are also shown in Table I and compared to the auxiliary-field Monte Carlo simulations of that publication. The latter are all above the variational upper bounds from Hartree Fock. The results of Table I show that the auxiliary-field Monte Carlo simulations of Refs. [19, 20], which tuned the Hamiltonians to reproduce experiment, do not accurately solve the Hamiltonians presented in those papers [48] and that the corresponding Hamiltonians do not produce accurate nuclear saturation.

Nuclear matter.— For our nuclear matter computations we consider lattices of different extent LL. The kinetic energy is a one-body operator and can be represented by a matrix of dimension D=4L3D = 4L^3. In a first step we compute the eigenvalues TkT_k and eigenstates ∣Tk⟩\lvert T_k\rangle with k=1,…,Dk = 1,\ldots,D of the kinetic energy and store them in order of increasing eigenvalues. The eigenstates are translationally invariant and can therefore serve in computations of homogeneous nuclear matter.

The eigenvalues TkT_k come in sets of degenerate numbers that reflect the cubic symmetry of the lattice. On sufficiently large lattices, the ordered eigenvalues change after eigenstate numbers 4, 28, 76, 108, 132, and so on [49, 50, 51]. In nuclear physics parlance, these are closed-shell configurations on a cubic lattice. We then consider nucleon numbers AA that correspond to these closed-shell systems and construct the one-body density matrix

n^=∑k=1A∣Tk⟩⟨Tk∣.(8)\hat{n} = \sum_{k=1}^{A} \lvert T_k\rangle\langle T_k\rvert. \tag*{(8)}

Its matrix elements in the spatial lattice basis are nqp=⟨p∣n^∣q⟩n_{q}^{p} = \langle p\lvert\hat{n}\rvert q\rangle. The energy expectation value for the AA fermion state with this density matrix is

E=∑pqϵqpnqp+12∑pqrsVqsprnqpnsr+16∑pqrstuWsuvpqrnqpnsrnvu.(9)E = \sum_{pq}\epsilon_{q}^{p}n_{q}^{p} + \frac{1}{2}\sum_{pqrs}V_{qs}^{pr}n_{q}^{p}n_{s}^{r} + \frac{1}{6}\sum_{pqrstu}W_{suv}^{pqr}n_{q}^{p}n_{s}^{r}n_{v}^{u}. \tag*{(9)}

where ϵqp\epsilon_{q}^{p}, VqsprV_{qs}^{pr}, and WsuvpqrW_{suv}^{pqr} are the matrix elements of the one-, two-, and three-body terms in the Hamiltonian (if the latter are present). These closed-shell configurations are solutions of the Hartree-Fock equations.

Figure 1 shows the energy per nucleon as a function of the density ρ=A/(La)3\rho= A/(La)^3 for closed-shell configurations on lattices with extent L=5L = 5, 6, and 7. Results are shown for the Hamiltonian HH of Eq. (1) and HBH_B of Eq. (5). We see that the data points from different LL fall onto relatively smooth curves for each Hamiltonian. For the Hamiltonian HH nuclear matter saturates at a much higher density and binding energy than the empirical saturation point, while Hamiltonian HBH_B does not saturate, reaching its lowest energy when the lattice is fully occupied. The top xx axis marks the number of nucleons per lattice site. The equations of state in Figure 1 indicate that finite nuclei will be overbound, and this is consistent with the results shown in Table I.

Image not extracted

Figure 1. Energy per nucleon (E/AE/A) in symmetric nuclear matter as a function of density ρ\rho from mean-field expectation values computed on lattices with extent L=5L = 5, 6, and 7 with the Hamiltonians of Ref. [20] (full symbols) and of Ref. [19] (open symbols). The region −16.5 MeV≤E/A≤−15.5 MeV-16.5\ \mathrm{MeV} \le E/A \le-15.5\ \mathrm{MeV} and 0.15 fm−3≤ρ≤0.17 fm−30.15\ \mathrm{fm}^{-3} \le\rho\le0.17\ \mathrm{fm}^{-3} around the empirical saturation point is shown as a gray rectangle. The top xx axis shows the density in nucleons per lattice site.

“Why you should trust us [52].”— First, we benchmarked the two-nucleon interaction of the Hamiltonian (1) in two-nucleon systems [53]. Second, we benchmarked the NuLattice Hartree-Fock code with a modification of the Hartree-Fock workhorse that is used to compute reference states for coupled-cluster computations in the harmonic-oscillator basis [39]. Third, NuLattice is publicly available [46, 54], and the computations that produce the results of this Letter can be replicated and verified. Fourth, neglecting the small contribution of the one-pion exchange, the analytical and numerical Hartree-Fock energies in a fully occupied lattice agree with each other for the Hamiltonians of this Letter.

Saturation on lattices.— We finally turn to Ref. [21], where auxiliary-field Monte Carlo computations seem accurate (see below) and examine how attractive two-body and three-body potentials can yield reasonable saturation on lattices. (Further examples are the most recent lattice Hamiltonians of Refs. [55, 56] which also employ attractive three-nucleon forces.) This is in contrast to continuum formulations where, for soft nuclear interactions, repulsive three-nucleon forces yield saturation [28, 57, 58, 59]. The basic mechanism is simple: In the Thomas-Fermi approximation the kinetic energy scales as ρ5/3\rho^{5/3}, while short-range two-body and three-body potential energies scale as ρ2\rho^2 and ρ3\rho^3, respectively. Thus, an attractive two-body potential alone overwhelms the kinetic energy at sufficiently high density ρ\rho and leads to a collapse of the system. A repulsive three-body force stabilizes the system.

We perform mean-field computations of neutron matter and present results in Table II. The number of three-body matrix elements WsuvpqrW_{suv}^{pqr} is about 1 GB per lattice site for the potential (7), which is prohibitively large. However, as nuclear matter is homogeneous, we only need to compute the expectation value of the three-nucleon potential for a

ends_mid=1 single lattice site (and multiply the result with L3L^3).

LLNNρ\rhoE/NE/NE/NE/N Ref. [21]
7140.0183.82.7
7380.0485.65.1
7540.0687.57.1
6380.0768.07.6
7660.0849.38.9
6540.1111.611.3
5380.1313.613.4
6660.1314.013.8

Table 2. Energy per particle E/NE/N (in MeV) of neutron matter at a density ρ\rho (in fm−3^{-3}) computed with Hartree Fock using lattices of extent LL and number of neutrons NN, and compared to the auxiliary-field Monte Carlo simulations [21].

Our results for neutron matter are close to (and above) those from auxiliary-field Monte Carlo simulations [21]. Assuming that the latter are accurate, we infer that the correlation energy, i.e., the difference between the ground-state energy and the Hartree-Fock energy, is relatively small (about 0.2 MeV per neutron) at ρ=0.13\rho= 0.13 fm−3^{-3} for an interaction with a momentum cutoff of 471 MeV. While it is not easy to compare potentials with different regulators and cutoffs, we note that the correlation energies by Alp et al. [60] at ρ=0.13\rho= 0.13 fm−3^{-3} are 0.5 MeV per neutron for the very soft interaction 1.8/2.0 (EM) of Ref. [28] (whose cutoffs are about 360 and 400 MeV for the nucleon-nucleon and the three-nucleon interaction, respectively) and 1.7 MeV per neutron for the harder interaction Δ\DeltaNNLOGO(450) of Ref. [24] (whose cutoff is 450 MeV).

Mean-field results for the equation of state of symmetric nuclear matter are shown in Figure 2. We see that the saturation point is close to the empirical one already at the Hartree-Fock level and that shell effects (which cause the scatter) are significant. For comparison we also show the results from the auxiliary-field Monte Carlo, labeled as “AFMC 2025” and “AFMC 2026”, from Refs. [61] and [56], respectively. (These differ slightly from each other because of different lattice extents LL and infinite-time extrapolations.) We note that the Hartree-Fock results are only variational upper bounds for computations with identical LL and AA. Apparently, the correlation energy is small. For a comparison, the correlation energies from Ref. [60] (at the saturation density ρ=0.16\rho= 0.16 fm−3^{-3}) are 3.9 MeV per nucleon for the interaction 1.8/2.0 (EM) and 6.7 MeV per nucleon for Δ\DeltaNNLOGO(450).

Image not extracted

Figure 2. Energy per nucleon E/AE/A in symmetric nuclear matter as a function of density ρ\rho from mean-field expectation values computed on lattices with extent LL using the Hamiltonian of Ref. [21] [See Eq. (6)]. The empirical saturation point is shown as a gray rectangle. The auxiliary-field Monte Carlo results “AFMC 2025” and “AFMC 2026” are from Refs. [61] and [56], respectively.

To understand the saturation mechanism we computed the expectation values of the kinetic energy per nucleon ⟨T⟩/A\langle T\rangle/A and the positive expectation values −⟨V0⟩/A-\langle V_0\rangle/A and −⟨W⟩/A-\langle W\rangle/A on lattices with L=4L = 4, 5, 6 and show the results in Figure 3. (The kinetic energy vanishes for the lowest density in each lattice system and is not shown on the log plot.) The kinetic energy per nucleon is approximately proportional to ρ2/3\rho^{2/3} for sufficiently large densities, as expected from Thomas-Fermi theory [62]. At low densities, the magnitude of the attractive two- and three-body potentials increase as ρ\rho and ρ2\rho^2, respectively, and this is much faster than the kinetic energy. However, the potential energies peak at densities of about one nucleon per lattice site and then decrease in magnitude. This is due to effects of the finite lattice on the nonlocal interaction.

Image not extracted

Figure 3. Positive expectation values of the kinetic energy TT, the two-body potential (−V0-V_0) and the three-body interaction (−W-W) energy per nucleon (E/AE/A) in symmetric nuclear matter as a function of density ρ\rho from mean-field expectation values computed on lattices with extent L=4L = 4, 5, and 6 using the Hamiltonian of Ref. [21]. The top xx axis shows the density in nucleons per lattice site.

We recall that a nonlocal potential is not multiplicative on the lattice and really moves nucleons to neighboring lattice sites. As the lattice becomes densely packed, Pauli blocking prevents nucleons from moving, reducing the attraction from the interaction. Apparently, this transition happens when a density of one nucleon per lattice site is reached. Thus, saturation is obtained from the dominance of kinetic over potential energies due to a dense packing of nucleons. It is a lattice artifact.

Summary and discussion.— We scrutinized the claims of Refs. [19, 20, 21] that simple lattice Hamiltonians with purely attractive potentials yield nuclear saturation. Our Hartree-Fock computations of the finite nuclei and of symmetric nuclear matter showed that the Hamiltonians from Refs. [19, 20] do not saturate properly and that the auxiliary field Monte Carlo simulations reported in those works did not accurately solve the corresponding Hamiltonians. We also showed that the attractive two-body and three-nucleon potentials of Ref. [21] approximately yield a physical saturation point at the Hartree-Fock level. This saturation comes from the dominance of the kinetic energy over potential energies once occupations become saturated on the lattice; it is a lattice artifact. With view on Refs. [19, 20] we conclude that an ab initio understanding of binding and clustering of α\alpha particles in atomic nuclei [63, 64] remains an open challenge.

Acknowledgments

We are grateful to Serdar Elhatisari for benchmarks of two-nucleon systems and many helpful discussions. We also thank Dick Furnstahl, Dean Lee, Witek Nazarewicz, and Achim Schwenk for useful discussions. We thank participants of the INT Program “Nuclear Hamiltonians for Advancing Nuclear Physics and Beyond” for questions about the continuum limit. This work was supported by the U.S. Department of Energy, Office of Science, Office of Nuclear Physics, under award Nos. DE-FG02-96ER40963 and DE-SC0026198; by the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research and Office of Nuclear Physics, Scientific Discovery through Advanced Computing (SciDAC) program (SciDAC-5 NUCLEI); and by the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U.S. Department of Energy. This research used resources of the Oak Ridge Leadership Computing Facility at the Oak Ridge National Laboratory, which is supported by the Advanced Scientific Computing Research programs in the Office of Science of the U.S. Department of Energy under Contract No. DE-AC05-00OR22725.

Data Availability

The data supporting this study may be reproduced using the published update to NuLattice [54].

End Matter

Details about the Hamiltonians.— The Hamiltonians employed in this work are described in detail in Refs. [19, 20, 21]. We list the relevant expressions here for completeness. In what follows

a⃗†(n)≡[a1†(n),a2†(n),a3†(n),a4†(n)](10)\vec{a}^{\dagger}(\mathbf{n}) \equiv\left[a_{1}^{\dagger}(\mathbf{n}),a_{2}^{\dagger}(\mathbf{n}),a_{3}^{\dagger}(\mathbf{n}),a_{4}^{\dagger}(\mathbf{n})\right] \tag*{(10)}

is a four component spin-isospin row vector. The kinetic energy is

T=−ℏ22ma2∑n[496a⃗†(n)a(n)−32∑∣n−n′∣=1a⃗†(n′)a(n)+320∑∣n−n′∣=2a⃗†(n′)a(n)−190∑∣n−n′∣=3a⃗†(n′)a(n)].(11)T=-\frac{\hbar^{2}}{2ma^{2}}\sum_{\mathbf{n}}\left[\frac{49}{6}\vec{a}^{\dagger}(\mathbf{n})a(\mathbf{n})-\frac{3}{2}\sum_{\lvert\mathbf{n}-\mathbf{n}'\rvert=1}\vec{a}^{\dagger}(\mathbf{n}')a(\mathbf{n})+\frac{3}{20}\sum_{\lvert\mathbf{n}-\mathbf{n}'\rvert=2}\vec{a}^{\dagger}(\mathbf{n}')a(\mathbf{n})-\frac{1}{90}\sum_{\lvert\mathbf{n}-\mathbf{n}'\rvert=3}\vec{a}^{\dagger}(\mathbf{n}')a(\mathbf{n})\right]. \tag*{(11)}

The one-pion exchange potential is

VOPE=−gA28πfπ2∑n′,n,S′,S,I:ρS′,I(n′)fS′,S(n′−n)ρS,I(n):.(12)V_{\mathrm{OPE}}=-\frac{g_{A}^{2}}{8\pi f_{\pi}^{2}}\sum_{\mathbf{n}',\mathbf{n},S',S,I}:\rho_{S',I}(\mathbf{n}')f_{S',S}(\mathbf{n}'-\mathbf{n})\rho_{S,I}(\mathbf{n}): . \tag*{(12)}

Here the spin-isospin density ρS,I\rho_{S,I} is defined for spin and isospin indices S=x,y,zS=x,y,z and I=x,y,zI=x,y,z, respectively. Using the Pauli spin matrices σS\sigma_{S} and isospin matrices τI\tau_{I} we have

ρS,I(n)=a†(n)[σS⊗τI]a(n),(13)\rho_{S,I}(\mathbf{n})=a^{\dagger}(\mathbf{n})[\sigma_{S}\otimes\tau_{I}]a(\mathbf{n}) , \tag*{(13)}

and fS′,Sf_{S',S} is given by

fS′,S(n′−n)=1L3∑qe−iq(n′−n)−bπq2qS′qSq2+mπ2.(14)f_{S',S}(\mathbf{n}'-\mathbf{n})=\frac{1}{L^{3}}\sum_{\mathbf{q}}\frac{e^{-i\mathbf{q}(\mathbf{n}'-\mathbf{n})-b_{\pi}\mathbf{q}^{2}}q_{S'}q_{S}}{\mathbf{q}^{2}+m_{\pi}^{2}} . \tag*{(14)}

Here qS\mathbf{q}_{S} is the SS component of the momentum transfer. It is an integer multiplied by 2π/L2\pi/L. bπb_{\pi} acts like a cutoff for the interaction and uses the value bπ=0.7b_{\pi}=0.7. The constants used are the pion mass mπ=134.98m_{\pi}=134.98 MeV, the axial-vector coupling constant gA=1.287g_{A}=1.287, and the pion decay constant fπ=92.2f_{\pi}=92.2 MeV. The potential (12) is in lattice units and needs to be multiplied with three powers of ℏ/a=100\hbar/a=100 MeV in actual computations.

In addition to the kinetic energy and the one-pion exchange, the Hamiltonian HBH_{B} of Ref. [19] consists of a combination of local and nonlocal potentials, see Eq. (5). The local potential is

VL=cL2∑n:ρL(n)ρL(n):+cI,L2∑I,n:ρI,L(n)ρI,L(n):+cS,L2∑S,n:ρS,L(n)ρS,L(n):+cS,I,L2∑I,n:ρS,I,L(n)ρS,I,L(n):.(15)\begin{aligned} V_{L}={}&\frac{c_{L}}{2}\sum_{\mathbf{n}}:\rho_{L}(\mathbf{n})\rho_{L}(\mathbf{n}):\\ &+\frac{c_{I,L}}{2}\sum_{I,\mathbf{n}}:\rho_{I,L}(\mathbf{n})\rho_{I,L}(\mathbf{n}):\\ &+\frac{c_{S,L}}{2}\sum_{S,\mathbf{n}}:\rho_{S,L}(\mathbf{n})\rho_{S,L}(\mathbf{n}):\\ &+\frac{c_{S,I,L}}{2}\sum_{I,\mathbf{n}}:\rho_{S,I,L}(\mathbf{n})\rho_{S,I,L}(\mathbf{n}): . \tag*{(15)} \end{aligned}

Here we used the smeared local densities

ρL(n)=ρ(n)+sL∑∣m−n∣=1ρ(m)\rho_{L}(\mathbf{n})=\rho(\mathbf{n})+s_{L}\sum_{\lvert\mathbf{m}-\mathbf{n}\rvert=1}\rho(\mathbf{m})
ρI,L(n)=ρI(n)+sL∑∣m−n∣=1ρI(m)\rho_{I,L}(\mathbf{n})=\rho_{I}(\mathbf{n})+s_{L}\sum_{\lvert\mathbf{m}-\mathbf{n}\rvert=1}\rho_{I}(\mathbf{m})
ρS,L(n)=ρS(n)+sL∑∣m−n∣=1ρS(m)\rho_{S,L}(\mathbf{n})=\rho_{S}(\mathbf{n})+s_{L}\sum_{\lvert\mathbf{m}-\mathbf{n}\rvert=1}\rho_{S}(\mathbf{m})
ρS,I,L(n)=ρS,I(n)+sL∑∣m−n∣=1ρS,I(m),(16)\rho_{S,I,L}(\mathbf{n})=\rho_{S,I}(\mathbf{n})+s_{L}\sum_{\lvert\mathbf{m}-\mathbf{n}\rvert=1}\rho_{S,I}(\mathbf{m}) , \tag*{(16)}

which are defined in terms of the local densities

ρ(n)=a†(n)a(n)\rho(\mathbf{n})=a^{\dagger}(\mathbf{n})a(\mathbf{n})
ρI(n)=a†(n)[τI]a(n)\rho_{I}(\mathbf{n})=a^{\dagger}(\mathbf{n})[\tau_{I}]a(\mathbf{n})
ρS(n)=a†(n)[σS]a(n),(17)\rho_{S}(\mathbf{n})=a^{\dagger}(\mathbf{n})[\sigma_{S}]a(\mathbf{n}) , \tag*{(17)}

and ρS,I(n)\rho_{S,I}(\mathbf{n}) was already defined in Eq. (13).

We turn to the nonlocal potential. Similarly to Eq. (10) we combine the four components of the smeared operator (2) into a row vector

a~†(n)≡[a~1†(n),a~2†(n),a~3†(n),a~4†(n)].(18)\tilde{a}^{\dagger}(\mathbf{n})\equiv\left[\tilde{a}_{1}^{\dagger}(\mathbf{n}),\tilde{a}_{2}^{\dagger}(\mathbf{n}),\tilde{a}_{3}^{\dagger}(\mathbf{n}),\tilde{a}_{4}^{\dagger}(\mathbf{n})\right] . \tag*{(18)}

One defines the nonlocal density

ρNL(n)=a~†(n)a~(n),(19)\rho_{NL}(\mathbf{n})=\tilde{a}^{\dagger}(\mathbf{n})\tilde{a}(\mathbf{n}) , \tag*{(19)}

and the nonlocal isospin density

ρI,NL(n)=a~†(n)[τI]a~(n).(20)\rho_{I,NL}(\mathbf{n})=\tilde{a}^{\dagger}(\mathbf{n})[\tau_{I}]\tilde{a}(\mathbf{n}) . \tag*{(20)}

The nonlocal potential then becomes

VNL=cNL2∑n:ρNL(n)ρNL(n):+cI,NL2∑I,n:ρI,NL(n)ρI,NL(n):.(21)\begin{aligned} V_{NL}={}&\frac{c_{NL}}{2}\sum_{\mathbf{n}}:\rho_{NL}(\mathbf{n})\rho_{NL}(\mathbf{n}):\\ &+\frac{c_{I,NL}}{2}\sum_{I,\mathbf{n}}:\rho_{I,NL}(\mathbf{n})\rho_{I,NL}(\mathbf{n}): . \tag*{(21)} \end{aligned}

We see that the local potential VLV_{L} only has local smearing, and that the nonlocal potential VNLV_{NL} only has nonlocal smearing. According to Ref. [19] the parameters of these potentials are sNL=0.077s_{NL}=0.077, sL=0.81s_{L}=0.81, cNL=−0.1171c_{NL}=-0.1171, cI,NL=0.02607c_{I,NL}=0.02607, cL=−0.01013c_{L}=-0.01013, and cS,L=cI,L=cS,I,L=−cL/3c_{S,L}=c_{I,L}=c_{S,I,L}=-c_{L}/3. They were “determined by fitting to the low-energy nucleon-nucleon phase shifts, the observed deuteron energy, and the low-energy alpha-alpha SS-wave phase shifts.” The results shown in Table I indicate that the alpha-alpha SS-wave phase shifts were not computed accurately. We also note that the significant local smearing sL=0.81s_{L}=0.81 implies that two nucleons interact strongly even when they are two lattice sites (i.e., almost 4 fm) apart. In numerical implementations one needs to multiply the couplings cNLc_{NL}, cI,NLc_{I,NL}, and cLc_{L} by one power of ℏ/a=100\hbar/a=100 MeV.

VV⟨V⟩/A\langle V\rangle/A⟨T+V⟩/A\langle T+V\rangle/A
V0V_0−76.80-76.80−33.31-33.31
V0+VOPEV_0+V_{\mathrm{OPE}}−75.55-75.55−32.06-32.06

Table 3. Kinetic and potential energies per nucleon (in MeV) for a fully filled lattice for the potentials of the Hamiltonian (1) from [20].

Analytical expressions for the fully occupied lattice.

From Eq. (11) we immediately see that the kinetic energy per particle is

⟨T⟩A=4912ma2.(22)\frac{\langle T\rangle}{A}=\frac{49}{12ma^2}. \tag*{(22)}

A somewhat lengthy derivation [see Supplemental Material for details] yields the expectation value of the two-body potential (4) on a fully occupied lattice with spin-isospin degeneracy gg. The result is

⟨V0⟩c0A=12g⟨:ρ~ 2(n):⟩=[g−12+6gsL+3(6g−1)sL2](1+6sNL2)2−3sLsNL2(8+17sLsNL2).(23)\begin{aligned} \frac{\langle V_0\rangle}{c_0 A} &=\frac{1}{2g}\langle:\widetilde{\rho}^{\,2}(\mathbf{n}):\rangle\\ &=\left[\frac{g-1}{2}+6gs_L+3(6g-1)s_L^2\right](1+6s_{NL}^2)^2\\ &\quad-3s_Ls_{NL}^2(8+17s_Ls_{NL}^2). \tag*{(23)} \end{aligned}

Similarly, we obtain the following expression for the three-body potential.

⟨W⟩c3A=13!g⟨:ρ~ 3(n):⟩=(1+6sNL2)3{16(g−1)(g−2)+3g(g−1)sL+3g(6g−1)sL2+2[1+9g(2g−1)]sL3}−(1+6sNL2)[24(g−1)sLsNL2+24(6g−1)sL2sNL2+51gsL2sNL4+102(3g−1)sL3sNL4+8sL2sNL4(27+28sLsNL)].(24)\begin{aligned} \frac{\langle W\rangle}{c_3 A} &=\frac{1}{3!g}\langle:\widetilde{\rho}^{\,3}(\mathbf{n}):\rangle\\ &=(1+6s_{NL}^2)^3\left\{\frac{1}{6}(g-1)(g-2)+3g(g-1)s_L\right.\\ &\quad\left.+3g(6g-1)s_L^2+2[1+9g(2g-1)]s_L^3\right\}\\ &\quad-(1+6s_{NL}^2)\left[24(g-1)s_Ls_{NL}^2+24(6g-1)s_L^2s_{NL}^2\right.\\ &\quad\left.+51gs_L^2s_{NL}^4+102(3g-1)s_L^3s_{NL}^4\right.\\ &\quad\left.+8s_L^2s_{NL}^4(27+28s_Ls_{NL})\right]. \tag*{(24)} \end{aligned}

To verify our nuclear matter results, we compute the energy expectation value of a completely filled lattice, i.e., for A=4L3A=4L^3. Results are presented in Table III for the Hamiltonian (1). The expectation value ⟨T+V0⟩/A\langle T+V_0\rangle/A is close to the full data points shown at the maximum density in Figure 1. We also see that the one-pion exchange acts repulsive and contributes only about +1.25+1.25 MeV per nucleon.

Continuum limit.— To address the continuum limit of vanishing lattice spacing a→0a\to0, we focus on low densities. There, the two-body potential dominates (see Figure 3), and at the lowest-density point ρ=4/(aL)3\rho=4/(aL)^3 one finds analytically

⟨V0⟩A=38c0a3(1+6sL)2(1+6sNL)4ρ.(25)\frac{\langle V_0\rangle}{A}=\frac{3}{8}c_0a^3(1+6s_L)^2(1+6s_{NL})^4\rho. \tag*{(25)}

We use this density for renormalization (i.e., for adjusting c0c_0 at ρ=4/(aL)3\rho=4/(aL)^3 such that ⟨V0⟩/A\langle V_0\rangle/A is fixed) and take a→0a\to0, L→∞L\to\infty while keeping the volume (aL)3(aL)^3 fixed. This requires c0∝a−3c_0\propto a^{-3}. Using Eqs. (23) and (22) this then implies that −⟨V0⟩/A∝a−3≫⟨T⟩/A∝a−2-\langle V_0\rangle/A\propto a^{-3}\gg\langle T\rangle/A\propto a^{-2} when a→0a\to0, L→∞L\to\infty and aL=constaL=\mathrm{const} for the fully occupied lattice. Thus, in the continuum limit the two-body interaction (4) leads to a collapse of the system. Alternatively, one might renormalize at low densities by keeping c0c_0 fixed and by scaling sNL∝a−1/2s_{NL}\propto a^{-1/2} and sL∝a−1/2s_L\propto a^{-1/2} as a→0a\to0. For the fully occupied lattice one then again finds from (23) that ⟨V0⟩/A∝a−3\langle V_0\rangle/A\propto a^{-3}, which leads to collapse.

When using nucleons as the degrees of freedom there are limited reasons why one would want to decrease the lattice spacing beyond the breakdown spacing a≈π/Λχa\approx\pi/\Lambda_\chi where Λχ≈0.7\Lambda_\chi\approx0.7 GeV is the chiral breakdown scale [72, 73]. Thus, the continuum limit is not of too much interest in nuclear lattice effective field theory. However, our analysis makes clear that the lattice potential of Ref. [21] cannot simply be used in computations that employ continuum Hamiltonians.

Reference states for nuclei

We start our Hartree-Fock computations for nuclei from compact initial states. For 4He^{4}\mathrm{He}, our initial state is chosen to be four nucleons at the origin n=(0,0,0)\mathbf{n}=(0,0,0), two neutrons with spins up and down and two protons with spins up and down. The lattice Hamiltonians we explore are translationally invariant, so this gives the same energy as defining our initial state at any other site. For 8Be^{8}\mathrm{Be}, we choose the initial state to be four nucleons at n=(0,0,0)\mathbf{n}=(0,0,0) and four nucleons at n=(1,0,0)\mathbf{n}=(1,0,0). Other possible compact initial states are equivalent up to translations and cubic rotations.

For 12C^{12}\mathrm{C}, the obvious choice is a “right-angled” configuration: four nucleons at each of n=(0,0,0)\mathbf{n}=(0,0,0), (1,0,0)(1,0,0), and (0,1,0)(0,1,0). However, other low-lying configurations are also possible. We explore these in Table IV. The right-angled configuration is clearly lower in energy than the linear configuration or the “bent-arm” configuration. In smaller lattices, we occasionally find that the linear configuration gives us the lowest-energy Hartree-Fock state due to large finite-volume effects. This vanishes as we go to large lattices. For 16O^{16}\mathrm{O}, we also find that the “right-angled” configuration, with four nucleons at each of n=(0,0,0)\mathbf{n}=(0,0,0), (1,0,0)(1,0,0), (0,1,0)(0,1,0), and (0,0,1)(0,0,1), gives the lowest Hartree-Fock energy.

Nucl.ConfigurationHHHBH_B
4He^{4}\mathrm{He}(0,0,0)(0,0,0)−19.81-19.81−32.19-32.19
8Be^{8}\mathrm{Be}(0,0,0),(1,0,0)(0,0,0),(1,0,0)−63.17-63.17−76.37-76.37
(0,0,0),(1,0,0),(0,1,0)(0,0,0),(1,0,0),(0,1,0)−137.28-137.28−139.83-139.83
12C^{12}\mathrm{C}(0,0,0),(1,0,0),(2,0,0)(0,0,0),(1,0,0),(2,0,0)−107.64-107.64−128.12-128.12
(0,0,0),(1,0,0),(2,1,0)(0,0,0),(1,0,0),(2,1,0)−101.26-101.26−126.72-126.72
16O^{16}\mathrm{O}(0,0,0),(1,0,0),(0,1,0),(0,0,1)(0,0,0),(1,0,0),(0,1,0),(0,0,1)−213.37-213.37−222.06-222.06
(0,0,0),(1,0,0),(1,1,0),(0,1,0)(0,0,0),(1,0,0),(1,1,0),(0,1,0)−187.22-187.22−212.84-212.84
(0,0,0),(1,0,0),(1,1,0),(2,0,0)(0,0,0),(1,0,0),(1,1,0),(2,0,0)−192.07-192.07−210.72-210.72
(0,0,0),(1,0,0),(2,0,0),(3,0,0)(0,0,0),(1,0,0),(2,0,0),(3,0,0)−151.51-151.51−180.94-180.94

Table IV. Hartree-Fock energies in MeV for different initial states for the Hamiltonians HH [20] and HBH_B [19] for lattice extent L=6L=6. For each configuration, we give the lattice sites at which four nucleons are placed to create the initial state.

Two-nucleon benchmarks

For the Hamiltonian (1) of Ref. [20] we list results from exact diagonalization for the deuteron in Table V and for the two-neutron system in Table VI. When the one-pion exchange is neglected, two neutrons and the deuteron have the same energy, because V0V_0 is spin-isospin symmetric. We see that the one-pion exchange is very weak and contributes little to the ground-state energy.

LLV0+VOPEV_0+V_{\mathrm{OPE}}V0V_0VOPEV_{\mathrm{OPE}}
4−5.04041823824-5.04041823824−5.17260862889-5.17260862889−0.0132647363266-0.0132647363266
5−3.26319416528-3.26319416528−3.42452503725-3.42452503725−0.0139277520259-0.0139277520259
6−2.38697021090-2.38697021090−2.55984789719-2.55984789719−0.0107662557934-0.0107662557934
7−1.88890653391-1.88890653391−2.06746886103-2.06746886103−0.0076071101145-0.0076071101145

Table V. Ground-state energies (in MeV) of the deuteron for the potentials and lattice sizes as indicated, obtained from exact diagonalization. Parameters are from Ref. [20].

LLV0+VOPEV_0+V_{\mathrm{OPE}}VOPEV_{\mathrm{OPE}}
4−5.03104190761-5.03104190761−0.00147403310330-0.00147403310330
5−3.25336204693-3.25336204693−0.00154796654782-0.00154796654782
6−2.37865887117-2.37865887117−0.00119657968867-0.00119657968867
7−1.88190245681-1.88190245681−0.00084529617778-0.00084529617778

Table VI. Same as Table Table V but for the two-neutron system.

Expectation values in the fully occupied lattice

The interaction of Refs. [20] contains a smeared “contact” as the dominant term (the one-pion exchange is very weak). It is based on the smeared creation operator (2) and the smeared density [see Eq. (3)]

ρ~(n)=∑i=1g(a~i†(n)a~i(n)+sL∑∣m−n∣=1a~i†(m)a~i(m)).(26)\widetilde{\rho}(\mathbf{n})=\sum_{i=1}^{g}\left(\widetilde{a}^{\dagger}_{i}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n})+s_L\sum_{\lvert\mathbf{m}-\mathbf{n}\rvert=1}\widetilde{a}^{\dagger}_{i}(\mathbf{m})\widetilde{a}_{i}(\mathbf{m})\right). \tag*{(26)}

Here we introduced the spin-isospin degeneracy gg.

We want to compute the expectation values of : ρ~2(n)\widetilde{\rho}^{2}(\mathbf{n}) : and : ρ~3(n)\widetilde{\rho}^{3}(\mathbf{n}) : in the fully occupied lattice. To do so, we note that the smeared creation operator (2) and its adjoint fulfill the anti-commutation relation

{a~i(n),a~j†(n)}=(1+6sNL2)δij.(27)\left\{\widetilde{a}_{i}(\mathbf{n}),\widetilde{a}^{\dagger}_{j}(\mathbf{n})\right\}=\left(1+6s_{NL}^{2}\right)\delta_{ij}. \tag*{(27)}

The smeared density consists of smeared creation and annihilation operators. Let m\mathbf{m} and m′\mathbf{m}' be nearest neighbors of n\mathbf{n}. We have

{a~i(n),a~j†(m)}=2sNLδij,(28)\left\{\widetilde{a}_{i}(\mathbf{n}),\widetilde{a}^{\dagger}_{j}(\mathbf{m})\right\}=2s_{NL}\delta_{ij}, \tag*{(28)}

and

{a~i(m),a~j†(m′)}=δij{1+6sNL2if m=m′,sNL2if m=−m′,2sNL2else.(29)\left\{\widetilde{a}_{i}(\mathbf{m}),\widetilde{a}^{\dagger}_{j}(\mathbf{m}')\right\} =\delta_{ij} \begin{cases} 1+6s_{NL}^{2} & \text{if }\mathbf{m}=\mathbf{m}',\\ s_{NL}^{2} & \text{if }\mathbf{m}=-\mathbf{m}',\\ 2s_{NL}^{2} & \text{else.} \end{cases} \tag*{(29)}

Here, m=−m′\mathbf{m}=-\mathbf{m}' is understood such that m\mathbf{m} and m′\mathbf{m}' are on opposite sites of n\mathbf{n}. For fixed n\mathbf{n}, there are six possible sites for each m\mathbf{m} and m′\mathbf{m}'. For fixed n\mathbf{n} and m\mathbf{m} the three possible values of the anti-commutator (29) occur with frequency one, one, and four when going from top to bottom on the right-hand-side of Eq. (29).

We are now ready to evaluate the expectation values of :ρ~2(n)::\widetilde{\rho}^{2}(\mathbf{n}): and :ρ~3(n)::\widetilde{\rho}^{3}(\mathbf{n}): in the fully occupied lattice. Starting with Eq. (26) we have

:ρ~2(n):=∑i,j=1ga~i†(n)a~j†(n)a~j(n)a~i(n)+2sL∑m∑i,j=1ga~i†(n)a~j†(m)a~j(m)a~i(n)+sL2∑m,m′∑i,j=1ga~i†(m)a~j†(m′)a~j(m′)a~i(m).(30):\widetilde{\rho}^{2}(\mathbf{n}):=\sum_{i,j=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{n})\widetilde{a}_{j}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n}) +2s_{L}\sum_{\mathbf{m}}\sum_{i,j=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{m})\widetilde{a}_{j}(\mathbf{m})\widetilde{a}_{i}(\mathbf{n}) +s_{L}^{2}\sum_{\mathbf{m},\mathbf{m}'}\sum_{i,j=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{m})\widetilde{a}_{j}^{\dagger}(\mathbf{m}')\widetilde{a}_{j}(\mathbf{m}')\widetilde{a}_{i}(\mathbf{m}) . \tag*{(30)}

Evaluating the expectation value in the fully occupied lattice requires us to perform Wick contractions using the anti-commutation relations (27), (28), and (29). Clearly, if we move the creation operators to the right, they will annihilate the fully occupied lattice. We have

∑i,j=1ga~i†(n)a~j†(n)a~j(n)a~i(n)⟶g2(1+6sNL2)2(31)\sum_{i,j=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{n)}\widetilde{a}_{j}^{\dagger}(\mathbf{n})\widetilde{a}_{j}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n})\longrightarrow g^{2}\left(1+6s_{NL}^{2}\right)^{2} \tag*{(31)}

and

∑i,j=1ga~i†(n)a~j†(n)a~j(n)a~i(n)⟶−g(1+6sNL2)2(32)\sum_{i,j=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{n})\widetilde{a}_{j}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n})\longrightarrow-g\left(1+6s_{NL}^{2}\right)^{2} \tag*{(32)}

and

∑m∑i,j=1ga~i†(n)a~j†(m)a~j(m)a~i(n)⟶6g2(1+6sNL2)2(33)\sum_{\mathbf{m}}\sum_{i,j=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{m})\widetilde{a}_{j}(\mathbf{m})\widetilde{a}_{i}(\mathbf{n})\longrightarrow6g^{2}\left(1+6s_{NL}^{2}\right)^{2} \tag*{(33)}

and

∑m∑i,j=1ga~i†(n)a~j†(m)a~j(m)a~i(n)⟶−6g(2sNL)2(34)\sum_{\mathbf{m}}\sum_{i,j=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{m})\widetilde{a}_{j}(\mathbf{m})\widetilde{a}_{i}(\mathbf{n})\longrightarrow-6g\left(2s_{NL}\right)^{2} \tag*{(34)}

and

∑m,m′∑i,j=1ga~i†(m)a~j†(m′)a~j(m′)a~i(m)⟶(6g)2(1+6sNL2)2(35)\sum_{\mathbf{m},\mathbf{m}'}\sum_{i,j=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{m})\widetilde{a}_{j}^{\dagger}(\mathbf{m}')\widetilde{a}_{j}(\mathbf{m}')\widetilde{a}_{i}(\mathbf{m})\longrightarrow(6g)^{2}\left(1+6s_{NL}^{2}\right)^{2} \tag*{(35)}

and

∑m,m′∑i,j=1ga~i†(m)a~j†(m′)a~j(m′)a~i(m)⟶−6g[(1+6sNL2)2+sNL4+4(2sNL2)2]=−6g[(1+6sNL2)2+17sNL4].(36)\begin{aligned} \sum_{\mathbf{m},\mathbf{m}'}\sum_{i,j=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{m})\widetilde{a}_{j}^{\dagger}(\mathbf{m}')\widetilde{a}_{j}(\mathbf{m}')\widetilde{a}_{i}(\mathbf{m}) &\longrightarrow-6g\left[\left(1+6s_{NL}^{2}\right)^{2}+s_{NL}^{4}+4(2s_{NL}^{2})^{2}\right]\\ &=-6g\left[\left(1+6s_{NL}^{2}\right)^{2}+17s_{NL}^{4}\right]. \tag*{(36)} \end{aligned}

Thus, we obtain for the expectation value of :ρ~2(n):/2g:\widetilde{\rho}^{2}(\mathbf{n}):/2g per nucleon

12g⟨:ρ~2(n):⟩=g−12(1+6sNL2)2+6sL[g(1+6sNL2)2−4sN2]+3sL2[(6g−1)(1+6sNL2)2−17sNL4].(37)\frac{1}{2g}\langle:\widetilde{\rho}^{2}(\mathbf{n}):\rangle=\frac{g-1}{2}\left(1+6s_{NL}^{2}\right)^{2}+6s_{L}\left[g\left(1+6s_{NL}^{2}\right)^{2}-4s_{N}^{2}\right]+3s_{L}^{2}\left[(6g-1)\left(1+6s_{NL}^{2}\right)^{2}-17s_{NL}^{4}\right]. \tag*{(37)}

This is essentially, i.e., up to the factor c0c_{0}, the expression (23).

We turn to :ρ~3(n)::\widetilde{\rho}^{3}(\mathbf{n}): and expand

:ρ~3(n):=∑i,j,k=1ga~i†(n)a~j†(n)a~k†(n)a~k(n)a~j(n)a~i(n)+3sL∑m∑i,j,ka~i†(n)a~j†(n)a~k†(m)a~k(m)a~j(n)a~i(n)+3sL2∑m,m′∑i,j,ka~i†(n)a~j†(m)a~k†(m′)a~k(m′)a~j(m)a~i(n)+sL3∑l,m,m′∑i,j,ka~i†(m)a~j†(m′)a~k†(l)a~k(l)a~j(m′)a~i(m).(38)\begin{aligned} :\widetilde{\rho}^{3}(\mathbf{n}):={}&\sum_{i,j,k=1}^{g}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{n})\widetilde{a}_{k}^{\dagger}(\mathbf{n})\widetilde{a}_{k}(\mathbf{n})\widetilde{a}_{j}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n})\\ &+3s_{L}\sum_{\mathbf{m}}\sum_{i,j,k}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{n})\widetilde{a}_{k}^{\dagger}(\mathbf{m})\widetilde{a}_{k}(\mathbf{m})\widetilde{a}_{j}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n})\\ &+3s_{L}^{2}\sum_{\mathbf{m},\mathbf{m}'}\sum_{i,j,k}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{m})\widetilde{a}_{k}^{\dagger}(\mathbf{m}')\widetilde{a}_{k}(\mathbf{m}')\widetilde{a}_{j}(\mathbf{m})\widetilde{a}_{i}(\mathbf{n})\\ &+s_{L}^{3}\sum_{\mathbf{l},\mathbf{m},\mathbf{m}'}\sum_{i,j,k}\widetilde{a}_{i}^{\dagger}(\mathbf{m})\widetilde{a}_{j}^{\dagger}(\mathbf{m}')\widetilde{a}_{k}^{\dagger}(\mathbf{l})\widetilde{a}_{k}(\mathbf{l})\widetilde{a}_{j}(\mathbf{m}')\widetilde{a}_{i}(\mathbf{m}). \tag*{(38)} \end{aligned}

Here, it is understood that l\mathbf{l} also is a nearest neighbor of n\mathbf{n}. There are six different ways to fully contract each line on the right-hand side of Eq. (38). The contractions of the first line of the right-hand side of Eq. (38) are standard and yield

g(g−1)(g−2)(1+6sNL2)3.(39)g(g-1)(g-2)\left(1+6s_{NL}^{2}\right)^{3}. \tag*{(39)}

For the second line of the right-hand side of Eq. (38) we find

∑m∑i,j,ka~i†(n)a~j†(n)a~k†(m)a~k(m)a~j(n)a~i(n)⟶6g3(1+6sNL2)3(40)\sum_{\mathbf{m}}\sum_{i,j,k}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{n})\widetilde{a}_{k}^{\dagger}(\mathbf{m})\widetilde{a}_{k}(\mathbf{m})\widetilde{a}_{j}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n})\longrightarrow6g^{3}\left(1+6s_{NL}^{2}\right)^{3} \tag*{(40)}

and

∑m∑i,j,ka~i†(n)a~j†(n)a~k†(m)a~k(m)a~j(n)a~i(n)⟶−6g2(1+6sNL2)3(41)\sum_{\mathbf{m}}\sum_{i,j,k}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{n})\widetilde{a}_{k}^{\dagger}(\mathbf{m})\widetilde{a}_{k}(\mathbf{m})\widetilde{a}_{j}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n})\longrightarrow-6g^{2}\left(1+6s_{NL}^{2}\right)^{3} \tag*{(41)}

and

∑m∑i,j,ka~i†(n)a~j†(n)a~k†(m)a~k(m)a~j(n)a~i(n)⟶−2×6g2(1+6sNL2)(2sNL)2(42)\sum_{\mathbf{m}}\sum_{i,j,k}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{n})\widetilde{a}_{k}^{\dagger}(\mathbf{m})\widetilde{a}_{k}(\mathbf{m})\widetilde{a}_{j}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n})\longrightarrow-2\times6g^{2}\left(1+6s_{NL}^{2}\right)(2s_{NL})^{2} \tag*{(42)}

and

∑m∑i,j,ka~i†(n)a~j†(n)a~k†(m)a~k(m)a~j(n)a~i(n)⟶2×6g(1+6sNL2)(2sNL)2.(43)\sum_{\mathbf{m}}\sum_{i,j,k}\widetilde{a}_{i}^{\dagger}(\mathbf{n})\widetilde{a}_{j}^{\dagger}(\mathbf{n})\widetilde{a}_{k}^{\dagger}(\mathbf{m})\widetilde{a}_{k}(\mathbf{m})\widetilde{a}_{j}(\mathbf{n})\widetilde{a}_{i}(\mathbf{n})\longrightarrow2\times6g\left(1+6s_{NL}^{2}\right)(2s_{NL})^{2}. \tag*{(43)}

The last two contractions come with factors of two, as clearly indicated, because there are two equivalent ways of performing the contractions (but only one is shown). Thus the four different contractions shown above capture the six different ways to perform them.

For the third line of the right-hand side of Eq. (38) we find

∑m,m′∑ijka~i†(n)a~j†(m)a~k†(m′)a~k(m′)a~j(m)a~i(n)→36g3(1+6sNL2)3,(44)\sum_{m,m'}\sum_{ijk}\tilde{a}^{\dagger}_{i}(n)\tilde{a}^{\dagger}_{j}(m)\tilde{a}^{\dagger}_{k}(m')\tilde{a}_{k}(m')\tilde{a}_{j}(m)\tilde{a}_{i}(n) \to36g^{3}\left(1+6s_{NL}^{2}\right)^{3}, \tag*{(44)}

and

∑m,m′∑ijka~i†(n)a~j†(m)a~k†(m′)a~k(m′)a~j(m)a~i(n)→−6g2(1+6sNL2)[(1+6sNL2)2+17sNL4],(45)\sum_{m,m'}\sum_{ijk}\tilde{a}^{\dagger}_{i}(n)\tilde{a}^{\dagger}_{j}(m)\tilde{a}^{\dagger}_{k}(m')\tilde{a}_{k}(m')\tilde{a}_{j}(m)\tilde{a}_{i}(n) \to-6g^{2}\left(1+6s_{NL}^{2}\right)\left[\left(1+6s_{NL}^{2}\right)^{2}+17s_{NL}^{4}\right], \tag*{(45)}

and

∑m,m′∑ijka~i†(n)a~j†(m)a~k†(m′)a~k(m′)a~j(m)a~i(n)→−2×(6g)2(1+6sNL2)(2sNL)2(46)\sum_{m,m'}\sum_{ijk}\tilde{a}^{\dagger}_{i}(n)\tilde{a}^{\dagger}_{j}(m)\tilde{a}^{\dagger}_{k}(m')\tilde{a}_{k}(m')\tilde{a}_{j}(m)\tilde{a}_{i}(n) \to-2\times(6g)^{2}\left(1+6s_{NL}^{2}\right)(2s_{NL})^{2} \tag*{(46)}

and

∑m,m′∑ijka~i†(n)a~j†(m)a~k†(m′)a~k(m′)a~j(m)a~i(n)→2×6g(2sNL)2(1+15sNL2).(47)\sum_{m,m'}\sum_{ijk}\tilde{a}^{\dagger}_{i}(n)\tilde{a}^{\dagger}_{j}(m)\tilde{a}^{\dagger}_{k}(m')\tilde{a}_{k}(m')\tilde{a}_{j}(m)\tilde{a}_{i}(n) \to2\times6g(2s_{NL})^{2}\left(1+15s_{NL}^{2}\right). \tag*{(47)}

As indicated, the last two contractions again come with factors of two.

Finally, for the last line of the right-hand side of Eq. (38) we find

∑l,m,m′∑ijka~i†(m)a~j†(m′)a~k†(l)a~k(l)a~j(m′)a~i(m)→(6g)3(1+6sNL2)3,(48)\sum_{l,m,m'}\sum_{ijk}\tilde{a}^{\dagger}_{i}(m)\tilde{a}^{\dagger}_{j}(m')\tilde{a}^{\dagger}_{k}(l)\tilde{a}_{k}(l)\tilde{a}_{j}(m')\tilde{a}_{i}(m) \to(6g)^{3}\left(1+6s_{NL}^{2}\right)^{3}, \tag*{(48)}

and

∑l,m,m′∑ijka~i†(m)a~j†(m′)a~k†(l)a~k(l)a~j(m′)a~i(m)→−3×(6g)2(1+6sNL2)[(1+6sNL2)2+17sNL4],(49)\sum_{l,m,m'}\sum_{ijk}\tilde{a}^{\dagger}_{i}(m)\tilde{a}^{\dagger}_{j}(m')\tilde{a}^{\dagger}_{k}(l)\tilde{a}_{k}(l)\tilde{a}_{j}(m')\tilde{a}_{i}(m) \to-3\times(6g)^{2}\left(1+6s_{NL}^{2}\right)\left[\left(1+6s_{NL}^{2}\right)^{2}+17s_{NL}^{4}\right], \tag*{(49)}
Image not extracted

Figure 4. Energy per nucleon E/AE/A in symmetric nuclear matter as a function of density ρ\rho from mean-field expectation values computed on lattices with extent LL using the Hamiltonian of Ref. [21] [See Eq. (6)]. The empirical saturation point is shown as a gray rectangle. The top xx axis shows the density in nucleons per lattice site.

and

∑l,m,m′∑ijka~i†(m)a~j†(m′)a~k†(l)a~k(l)a~j(m′)a~i(m)→2×6g[(1+6sNL2)3+3(sNL2)2(1+6sNL2)+8(2sNL2)2(1+6sNL2)+8sNL(2sNL2)2+8(2sNL2)3+4(2sNL2)2(1+6sNL2)+4sNL2(2sNL2)2]=12g[(1+6sNL2)3+51sNL4(1+6sNL2)+112sNL6].(50)\begin{aligned} \sum_{l,m,m'}\sum_{ijk}\tilde{a}^{\dagger}_{i}(m)\tilde{a}^{\dagger}_{j}(m')\tilde{a}^{\dagger}_{k}(l)\tilde{a}_{k}(l)\tilde{a}_{j}(m')\tilde{a}_{i}(m) \to{}& 2\times6g\left[\left(1+6s_{NL}^{2}\right)^{3}+3(s_{NL}^{2})^{2}\left(1+6s_{NL}^{2}\right)\right.\\ &+8(2s_{NL}^{2})^{2}\left(1+6s_{NL}^{2}\right)+8s_{NL}(2s_{NL}^{2})^{2}\\ &\left.+8(2s_{NL}^{2})^{3}+4(2s_{NL}^{2})^{2}\left(1+6s_{NL}^{2}\right)+4s_{NL}^{2}(2s_{NL}^{2})^{2}\right]\\ ={}&12g\left[\left(1+6s_{NL}^{2}\right)^{3}+51s_{NL}^{4}\left(1+6s_{NL}^{2}\right)+112s_{NL}^{6}\right]. \tag*{(50)} \end{aligned}

Here, the prefactors of two and three in the last two contractions are again singled out, and they account for equivalent ways to perform these. In the last contraction (having fixed one of the sites, say mm) there are 36 different possibilities for m′m' and ll. They are all accounted for on the right hand side of the arrow, with the individual factors from the anti-commutator (29) identified in parenthesis. Putting all together we arrive at Eq. (24).

Figure 4 shows the equation of state of the Hamiltonian (6) from Ref. [21] over the full range of density. The analytical results match the numerical ones at the maximum density. The region around the saturation point and the individual energy contributions are shown in Figure 2 and Figure 3 of the main text.

We turn to the potentials VLV_L and VNLV_{NL} of the Hamiltonian HBH_B, see Eqs. (15) and (21), respectively of Ref. [19], and compute their expectation values on a fully filled lattice. In what follows, it will be useful to occasionally separate the two spin indices s=1,2s=1,2 (which stand for the projections 1/21/2 and −1/2-1/2) and the two similarly defined isospin indices i=1,2i=1,2 and introduce a1,1(n)=a1(n)a_{1,1}(\mathbf{n})=a_1(\mathbf{n}), a1,2(n)=a2(n)a_{1,2}(\mathbf{n})=a_2(\mathbf{n}), a2,1(n)=a3(n)a_{2,1}(\mathbf{n})=a_3(\mathbf{n}), and a2,2(n)=a4(n)a_{2,2}(\mathbf{n})=a_4(\mathbf{n}).

We start with the nonlocal potential VNLV_{NL} of Eq. (21). The expectation value from the nonlocal density :ρNL2(n)::\rho_{NL}^2(\mathbf{n}): can simply be obtained by setting sL=0s_L=0 in Eq. (23). Thus, we only need to compute the contribution from the isospin density :ρI,NL2(n)::\rho_{I,NL}^2(\mathbf{n}): of Eq. (20). Because of isospin symmetry of the fully occupied lattice we only need to compute one isospin component and focus on I=zI=z. We have

ρI=z,NL(n)=∑s=12(a~s,1†(n)a~s,1(n)−a~s,2†(n)a~s,2(n)).(51)\rho_{I=z,NL}(\mathbf{n})=\sum_{s=1}^{2}\left(\tilde{a}_{s,1}^{\dagger}(\mathbf{n})\tilde{a}_{s,1}(\mathbf{n})-\tilde{a}_{s,2}^{\dagger}(\mathbf{n})\tilde{a}_{s,2}(\mathbf{n})\right). \tag*{(51)}

Thus,

:ρI=z,NL2(n):=∑s,s′=12[a~s,1†(n)a~s′,1†(n)a~s′,1(n)a~s,1(n)−a~s,1†(n)a~s′,2†(n)a~s′,2(n)a~s,1(n)−a~s,2†(n)a~s′,1†(n)a~s′,1(n)a~s,2(n)+a~s,2†(n)a~s′,2†(n)a~s′,2(n)a~s,2(n)].(52):\rho_{I=z,NL}^{2}(\mathbf{n}):=\sum_{s,s'=1}^{2}\left[\tilde{a}_{s,1}^{\dagger}(\mathbf{n})\tilde{a}_{s',1}^{\dagger}(\mathbf{n})\tilde{a}_{s',1}(\mathbf{n})\tilde{a}_{s,1}(\mathbf{n})-\tilde{a}_{s,1}^{\dagger}(\mathbf{n})\tilde{a}_{s',2}^{\dagger}(\mathbf{n})\tilde{a}_{s',2}(\mathbf{n})\tilde{a}_{s,1}(\mathbf{n})-\tilde{a}_{s,2}^{\dagger}(\mathbf{n})\tilde{a}_{s',1}^{\dagger}(\mathbf{n})\tilde{a}_{s',1}(\mathbf{n})\tilde{a}_{s,2}(\mathbf{n})+\tilde{a}_{s,2}^{\dagger}(\mathbf{n})\tilde{a}_{s',2}^{\dagger}(\mathbf{n})\tilde{a}_{s',2}(\mathbf{n})\tilde{a}_{s,2}(\mathbf{n})\right]. \tag*{(52)}

In the fully occupied lattice, the non-vanishing expectation values come from terms such as

∑s,s′=12a~s,1†(n)a~s′,1†(n)a~s′,1(n)a~s,1(n)⟶4(1+6sNL2)2,(53)\sum_{s,s'=1}^{2}\tilde{a}_{s,1}^{\dagger}(\mathbf{n})\tilde{a}_{s',1}^{\dagger}(\mathbf{n})\tilde{a}_{s',1}(\mathbf{n})\tilde{a}_{s,1}(\mathbf{n})\longrightarrow4(1+6s_{NL}^{2})^{2}, \tag*{(53)}
∑s,s′=12a~s,1†(n)a~s′,1†(n)a~s′,1(n)a~s,1(n)⟶−2(1+6sNL2)2,(54)\sum_{s,s'=1}^{2}\tilde{a}_{s,1}^{\dagger}(\mathbf{n})\tilde{a}_{s',1}^{\dagger}(\mathbf{n})\tilde{a}_{s',1}(\mathbf{n})\tilde{a}_{s,1}(\mathbf{n})\longrightarrow-2(1+6s_{NL}^{2})^{2}, \tag*{(54)}

and

∑s,s′=12a~s,1†(n)a~s′,2†(n)a~s′,2(n)a~s,1(n)⟶4(1+6sNL2)2.(55)\sum_{s,s'=1}^{2}\tilde{a}_{s,1}^{\dagger}(\mathbf{n})\tilde{a}_{s',2}^{\dagger}(\mathbf{n})\tilde{a}_{s',2}(\mathbf{n})\tilde{a}_{s,1}(\mathbf{n})\longrightarrow4(1+6s_{NL}^{2})^{2}. \tag*{(55)}

There are three more analogous contractions from the final two terms in Eq. (52). The sum gives

⟨:ρI=z,NL2(n):⟩=−4(1+6sNL2)2.(56)\left\langle:\rho_{I=z,NL}^{2}(\mathbf{n}):\right\rangle=-4(1+6s_{NL}^{2})^{2}. \tag*{(56)}

This is just for I=zI=z, so we get a factor of 3 from all three isospin II components. Finally, dividing by 4 nucleons per lattice site and accounting for the prefactors in Eq. (21), the complete result is

⟨VNL⟩A=32(cNL−cI,NL)(1+6sNL2)2.(57)\frac{\langle V_{NL}\rangle}{A}=\frac{3}{2}(c_{NL}-c_{I,NL})(1+6s_{NL}^{2})^{2}. \tag*{(57)}

We turn to the local potential VLV_L of Eq. (15). The expectation value of :ρL2(n)::\rho_L^2(\mathbf{n}): can be obtained from Eq. (23) by setting sNL=0s_{NL}=0. For the remaining spin, isospin, and spin-isospin terms we note that we only need to compute one of them because of the spin-isospin symmetry of the fully occupied lattice, and we can again focus on a single component.

Let us take ρS=x,I=x\rho_{S=x,I=x} for example. Then,

ρS=x,I=x(n)=a†(n)[σx⊗τx]a(n)(58)\rho_{S=x,I=x}(\mathbf{n})=a^{\dagger}(\mathbf{n})[\sigma_x\otimes\tau_x]a(\mathbf{n}) \tag*{(58)}
=a1†(n)a4(n)−a2†(n)a3(n)−a3†(n)a2(n)+a4†(n)a1(n)+sL∑∣n−m∣=1[a1†(n)a4(m)−a2†(n)a3(m)−a3†(n)a2(m)+a4†(n)a1(m)].(59)=a_1^{\dagger}(\mathbf{n})a_4(\mathbf{n})-a_2^{\dagger}(\mathbf{n})a_3(\mathbf{n})-a_3^{\dagger}(\mathbf{n})a_2(\mathbf{n})+a_4^{\dagger}(\mathbf{n})a_1(\mathbf{n}) +s_L\sum_{|\mathbf{n}-\mathbf{m}|=1}\left[a_1^{\dagger}(\mathbf{n})a_4(\mathbf{m})-a_2^{\dagger}(\mathbf{n})a_3(\mathbf{m})-a_3^{\dagger}(\mathbf{n})a_2(\mathbf{m})+a_4^{\dagger}(\mathbf{n})a_1(\mathbf{m})\right]. \tag*{(59)}

When computing :ρS=x,I=x2(n)::\rho_{S=x,I=x}^{2}(\mathbf{n}): one gets a lot of terms. However, as we need to match spin-isospin indices when computing the Wick contractions in the fully occupied lattice, only a smaller number of terms survive. These are

:ρS=x,I=x2(n):⟶a1†(n)a4†(n)a1(n)a4(n)+a2†(n)a3†(n)a2(n)a3(n)+2sL∑∣n−m∣=1[a1†(n)a4†(m)a1(m)a4(n)+a2†(n)a3†(m)a2(m)a3(n)]+sL2∑∣n−m∣=1∣n−m′∣=1[a1†(m′)a4†(m)a1(m)a4(m′)+a2†(m′)a3†(m)a2(m)a3(m′)]+h.c.(60):\rho_{S=x,I=x}^{2}(\mathbf{n}):\longrightarrow a_1^{\dagger}(\mathbf{n})a_4^{\dagger}(\mathbf{n})a_1(\mathbf{n})a_4(\mathbf{n})+a_2^{\dagger}(\mathbf{n})a_3^{\dagger}(\mathbf{n})a_2(\mathbf{n})a_3(\mathbf{n}) +2s_L\sum_{|\mathbf{n}-\mathbf{m}|=1}\left[a_1^{\dagger}(\mathbf{n})a_4^{\dagger}(\mathbf{m})a_1(\mathbf{m})a_4(\mathbf{n})+a_2^{\dagger}(\mathbf{n})a_3^{\dagger}(\mathbf{m})a_2(\mathbf{m})a_3(\mathbf{n})\right] +s_L^2\sum_{\substack{|\mathbf{n}-\mathbf{m}|=1\\|\mathbf{n}-\mathbf{m}'|=1}}\left[a_1^{\dagger}(\mathbf{m}')a_4^{\dagger}(\mathbf{m})a_1(\mathbf{m})a_4(\mathbf{m}')+a_2^{\dagger}(\mathbf{m}')a_3^{\dagger}(\mathbf{m})a_2(\mathbf{m})a_3(\mathbf{m}')\right]+\mathrm{h.c.} \tag*{(60)}

The non-vanishing Wick contractions yield

a1†(n)a4†(n)a1(n)a4(n)⟶−1,(61)a_1^{\dagger}(\mathbf{n})a_4^{\dagger}(\mathbf{n})a_1(\mathbf{n})a_4(\mathbf{n})\longrightarrow-1, \tag*{(61)}

and

∑∣n−m∣=1∣n−m′∣=1a1†(m′)a4†(m)a1(m)a4(m′)⟶−6.(62)\sum_{\substack{|\mathbf{n}-\mathbf{m}|=1\\|\mathbf{n}-\mathbf{m}'|=1}}a_1^{\dagger}(\mathbf{m}')a_4^{\dagger}(\mathbf{m})a_1(\mathbf{m})a_4(\mathbf{m}')\longrightarrow-6. \tag*{(62)}

Putting this all together, we get

⟨:ρS=x,I=x2(n):⟩=−4(1+6sL2),(63)\langle:\rho^{2}_{S=x,I=x}(\mathbf{n}):\rangle=-4(1+6s_{L}^{2}), \tag*{(63)}

and the factor of 4 comes from the Hermitian conjugates and (1,4)→(2,3)(1,4)\to(2,3) in Eq. (60).

The final result is

⟨VL⟩A=32cL(1+16sL+46sL2)−32(cS,L+cI,L+3cS,I,L)(1+6sL2).(64)\frac{\langle V_{L}\rangle}{A}=\frac{3}{2}c_{L}(1+16s_{L}+46s_{L}^{2})-\frac{3}{2}(c_{S,L}+c_{I,L}+3c_{S,I,L})(1+6s_{L}^{2}). \tag*{(64)}

Expectation values at lowest densities

The lowest densities we compute are at ρ=4/(aL)3\rho=4/(aL)^{3}, where the zero-momentum state of four nucleons is

∣p=0⟩=L−6∑k,l,m,na1†(k)a2†(l)a3†(m)a4†(n)∣0⟩=L−6∑k,l,m,na~1†(k)a~2†(l)a~3†(m)a~4†(n)(1+6sNL)∣0⟩.(65)\begin{aligned} |p=0\rangle=L^{-6}\sum_{\mathbf{k},l,m,n}a_{1}^{\dagger}(\mathbf{k})a_{2}^{\dagger}(l)a_{3}^{\dagger}(m)a_{4}^{\dagger}(n)|0\rangle\\ &=L^{-6}\sum_{\mathbf{k},l,m,n}\frac{\widetilde{a}_{1}^{\dagger}(\mathbf{k})\widetilde{a}_{2}^{\dagger}(l)\widetilde{a}_{3}^{\dagger}(m)\widetilde{a}_{4}^{\dagger}(n)}{(1+6s_{NL})}|0\rangle. \tag*{(65)} \end{aligned}

Here, the translational invariance of the zero-momentum state allowed us to rewrite it in terms of the a~i†\widetilde{a}_{i}^{\dagger} operators. It is now straightforward (with all we learned in Sec. III) to compute the expectation value (using A=4A=4)

⟨p=0∣V0∣p=0⟩A=3c02L3(1+6sL)2(1+6sNL)4=38c0a3(1+6sL)2(1+6sNL)4ρ.(66)\begin{aligned} \frac{\langle p=0|V_{0}|p=0\rangle}{A}=\frac{3c_{0}}{2L^{3}}(1+6s_{L})^{2}(1+6s_{NL})^{4} \\ &=\frac{3}{8}c_{0}a^{3}(1+6s_{L})^{2}(1+6s_{NL})^{4}\rho. \tag*{(66)} \end{aligned}

This is Eq. (25). The expression is intuitively clear. The nonlocal and local smearing each involves six sites. Nonlocal smearing involves four annihilation and creation operators (hence the power four), while local smearing involves two densities.

Energies from the transfer matrix

The auxiliary field Monte Carlo calculations of Refs. [19, 20, 65, 66, 67, 68, 69, 70] solved the many-body problem by applying the transfer matrix formalism [71] with a discrete temporal lattice spacing at=1/(150 MeV)a_{t}=1/(150\ \mathrm{MeV}). In a nutshell, this works as follows.

One introduces the transfer matrix

M(ϕ)=:e−atH1(ϕ):(67)M(\phi)=:e^{-a_{t}H_{1}(\phi)}: \tag*{(67)}

Here H1H_{1} is a one-body Hamiltonian that depends on auxiliary fields ϕ\phi after a Hubbard Stratonovich transformation of original Hamiltonian HH (that included two- and three-body interactions). We note the occurrence of the temporal lattice spacing ata_{t}. Next, one introduces a partition function via the path integral

Z=∫Dϕe−S(ϕ)⟨ψ∣MLt−1(ϕ)⋯M0(ϕ)∣ψ⟩,(68)Z=\int\mathcal{D}\phi e^{-S(\phi)}\langle\psi|M_{L_{t}-1}(\phi)\cdots M_{0}(\phi)|\psi\rangle, \tag*{(68)}

where ∣ψ⟩|\psi\rangle is an initial state and LtL_{t} temporal steps are being made. (The subscripts on the transfer matrices help us counting the number LtL_{t} of them.) The idea is that the “time evolution” Ltat→∞L_{t}a_{t}\to\infty projects out the ground state. Then the energy is computed as

E=lim⁡Lt→∞log⁡ZatLt.(69)E=\lim_{L_{t}\to\infty}\frac{\log Z}{a_{t}L_{t}}. \tag*{(69)}

Strictly speaking, however, one needs Ltat→∞L_{t}a_{t}\to\infty while keeping ata_{t} sufficiently small. (Sufficiently small means that atE≪1a_{t}E\ll1.) However, Refs. [19, 20, 66, 67, 68, 69, 70] used at=1/(150 MeV)a_{t}=1/(150\ \mathrm{MeV}).

Assuming that ata_{t} is sufficiently small, one can rewrite the partition function Eq. (68) as

Z≈∫Dϕe−S(ϕ)⟨ψ∣1−LtatH1∣ψ⟩,(70)Z\approx\int\mathcal{D}\phi e^{-S(\phi)}\langle\psi|1-L_{t}a_{t}H_{1}|\psi\rangle, \tag*{(70)}

and it is clear that the correct energy is obtained from Eq. (69).

In contrast, if ata_{t} is not sufficiently small then the computed energy

Eat=−log⁡(1−atEtrue)at(71)E_{a_{t}}=-\frac{\log(1-a_{t}E_{\mathrm{true}})}{a_{t}} \tag*{(71)}

is not necessarily close to the true energy EtrueE_{\mathrm{true}}. Indeed, using the Hartree-Fock energies in Table I as the true energies and at=1/(150 MeV)a_{t}=1/(150\ \mathrm{MeV}), one does get the corresponding results from auxiliary field Monte Carlo to a good approximation.

One can also turn this around. Using the temporal spacing at=1/(150 MeV)a_{t}=1/(150\ \mathrm{MeV}) and Eq. (69) one does not really compute the energy of the original Hamiltonian HH but rather that of a different Hamiltonian containing various powers of HH. We finally note that the temporal lattice spacing was decreased to at=1/(1000 MeV)a_{t}=1/(1000\ \mathrm{MeV}) in Ref. [21], and the energy was not computed any more from Eq. (69) but rather from an expectation value of the Hamiltonian, see, e.g., Refs. [26, 61].

References

  1. [1]E. Epelbaum, H.-W. Hammer, and U.-G. Meißner, Modern theory of nuclear forces, Rev. Mod. Phys. 81, 1773 (2009).DOI
  2. [2]R. Machleidt and D. Entem, Chiral effective field theory and nuclear forces, Phys. Rep. 503, 1 (2011).DOI
  3. [3]A. Gezerlis, I. Tews, E. Epelbaum, S. Gandolfi, K. Hebeler, A. Nogga, and A. Schwenk, Quantum Monte Carlo calculations with chiral effective field theory interactions, Phys. Rev. Lett. 111, 032501 (2013).pubmed.ncbi.nlm.nih.gov/23909312
  4. [4]M. Piarulli, L. Girlanda, R. Schiavilla, R. N. Pérez, J. E. Amaro, and E. R. Arriola, Minimally nonlocal nucleon-nucleon potentials with chiral two-pion exchange including Δ resonances, Phys. Rev. C 91, 024003 (2015).DOI
  5. [5]A. Ekström, G. R. Jansen, K. A. Wendt, G. Hagen, T. Papenbrock, B. D. Carlsson, C. Forssén, M. Hjorth-Jensen, P. Navrátil, and W. Nazarewicz, Accurate nuclear radii and binding energies from a chiral interaction, Phys. Rev. C 91, 051301 (2015).DOI
  6. [6]M. Piarulli, L. Girlanda, R. Schiavilla, A. Kievsky, A. Lovato, L. E. Marcucci, S. C. Pieper, M. Viviani, and R. B. Wiringa, Local chiral potentials with Δ-intermediate states and the structure of light nuclei, Phys. Rev. C 94, 054007 (2016).DOI
  7. [7]H. W. Hammer, S. König, and U. van Kolck, Nuclear effective field theory: Status and perspectives, Rev. Mod. Phys. 92, 025004 (2020).DOI
  8. [8]P. Reinert, H. Krebs, and E. Epelbaum, Semilocal momentum-space regularized chiral two-nucleon potentials up to fifth order, Eur. Phys. J. A 54, 86 (2018).DOI
  9. [9]Maris, P., Epelbaum, E., Furnstahl, R. J., Golak, J., Hebeler, K., Hammer, T., Kamada, H., Krebs, H., Meißner, U.-G., Melendez, J. A., Nogga, A., Reinert, P., Roth, R., Skibiński, R., Soloviov, K., Topolnicki, K., Vary, J. P., Volkotrub, H., Witała, J., & Włoch, T. (2021). Light nuclei with semilocal momentum-space regularized chiral interactions up to third order. Physical Review C, 103, 054001.
  10. [10]Hagen, G., Ekström, A., Forssén, C., Jansen, G. R., Nazarewicz, W., Papenbrock, T., Wendt, K. A., Bacca, S., Barnea, N., Carlsson, B., Drischler, C., Hebeler, K., Hjorth-Jensen, M., Morelli, M., Orlandini, G., Schwenk, A., & Simonis, J. (2016). Neutron and weak-charge distributions of the ${}^{48}$Ca nucleus. Nature Physics, 12, 186.
  11. [11]G. Hagen, G. R. Jansen, and T. Papenbrock, Structure of ${}^{78}$Ni from first-principles computations, Phys. Rev. Lett. 117, 172501 (2016).pubmed.ncbi.nlm.nih.gov/27824459
  12. [12]T. D. Morris, J. Simonis, S. R. Stroberg, C. Stumpf, G. Hagen, J. D. Holt, G. R. Jansen, T. Papenbrock, R. Roth, and A. Schwenk, Structure of the lightest tin isotopes, Phys. Rev. Lett. 120, 152503 (2018).DOI
  13. [13]F. Bonaiti, G. Hagen, and T. Papenbrock, Structure of the doubly magic nuclei ²⁰⁸Pb and ²⁶⁶Pb from ab initio computations, arXiv:2508.14217.arXiv
  14. [14]P. Gysbers, G. Hagen, J. D. Holt, G. R. Jansen, T. D. Morris, P. Navrátil, T. Papenbrock, S. Quaglioni, A. Schwenk, S. R. Stroberg, and K. A. Wendt, Discrepancy between experimental and theoretical β-decay rates resolved from first principles, Nat. Phys. 15, 428 (2019).DOI
  15. [15]B. Hu, W. Jiang, T. Miyagi, Z. Sun, A. Ekström, C. Forssén, G. Hagen, J. D. Holt, T. Papenbrock, S. R. Stroberg, and I. Vernon, Ab initio predictions link the neutron skin of ²⁰⁸Pb to nuclear forces, Nat. Phys. 18, 1196 (2022).DOI
  16. [16]Z. H. Sun, A. Ekström, C. Forssén, G. Hagen, G. R. Jansen, and T. Papenbrock, Multiscale physics of atomic nuclei from first principles, Phys. Rev. X 15, 011028 (2025).DOI
  17. [17]C. R. Ding, C. C. Wang, J. M. Yao, H. Hergert, H. Z. Liang, and S. K. Bogner, From spin to pseudospin symmetry: The origin of magic numbers in nuclear structure, Phys. Rev. Lett. 136, 052501 (2026).DOI
  18. [18]König, S., Grießhammer, H. W., Hammer, H.-W., & van Kolck, U. (2016). Effective theory of 3h and 3he. J. Phys. G: Nucl. Part. Phys., 43, 055106.
  19. [19]S. Elhatisari, N. Li, A. Rokash, J. M. Alarcón, D. Du, N. Klein, B.-N. Lu, U.-G. Meißner, E. Epelbaum, H. Krebs, T. A. Lähde, D. Lee, and G. Rupak, Nuclear binding near a quantum phase transition, Phys. Rev. Lett. 117, 132501 (2016).pubmed.ncbi.nlm.nih.gov/27715077
  20. [20]S. Elhatisari, E. Epelbaum, H. Krebs, T. A. Lähde, D. Lee, N. Li, B.-N. Lu, U.-G. Meißner, and G. Rupak, Ab initio calculations of the isotopic dependence of nuclear clustering, Phys. Rev. Lett. 119, 222505 (2017).DOI
  21. [21]B.-N. Lu, N. Li, S. Elhatisari, D. Lee, E. Epelbaum, and U.-G. Meißner, Essential elements for nuclear binding, Phys. Lett. B 797, 134863 (2019).DOI
  22. [22]A. Gnech, B. Fore, A. J. Tropiano, and A. Lovato, Distilling the essential elements of nuclear binding via neural-network quantum states, Phys. Rev. Lett. 133, 142501 (2024).DOI
  23. [23]C. Drischler, K. Hebeler, and A. Schwenk, Chiral interactions up to next-to-next-to-next-to-leading order and nuclear saturation, Phys. Rev. Lett. 122, 042501 (2019).DOI
  24. [24]W. G. Jiang, A. Ekström, C. Forssén, G. Hagen, G. R. Jansen, and T. Papenbrock, Accurate bulk properties of nuclei from A = 2 to ∞ from potentials with Δ isobars, Phys. Rev. C 102, 054301 (2020).DOI
  25. [25]Arthus, P., Hebeler, K., & Schwenk, A. (2024). Neutron-rich nuclei and neutron skins from chiral low-resolution interactions. arXiv:2401.06675 [nucl-th]
  26. [26]S. Elhatisari, L. Bovermann, Y.-Z. Ma, E. Epelbaum, D. Framatino, F. Hildenbrand, M. Kim, Y. Kim, H. Krebs, T. A. Lähde, D. Lee, N. Li, B.-N. Lu, U.-G. Meißner, G. Rupak, Y.-H. Song, and G. Stellin, Wavefunction matching for solving quantum many-body problems, Nature 630, 59 (2024).DOI
  27. [27]P. F. Bedaque and U. van Kolck, Effective field theory for few-nucleon systems, Annual Review of Nuclear and Particle Science 52, 339 (2002), nucl-th/0203055.DOI
  28. [28]K. Hebeler, S. K. Bogner, R. J. Furnstahl, A. Nogga, and A. Schwenk, Improved nuclear matter calculations from chiral low-momentum interactions, Phys. Rev. C 83, 031301 (2011).DOI
  29. [29]Hagen, G., Dean, M. J., Hjorth-Jensen, M., Papenbrock, T., & Schwenk, A. (2007). Benchmark calculations for $^{3}$H, $^{4}$He, $^{16}$O, and Ca with ab initio coupled-cluster theory. Phys. Rev. C, 76, 044305.
  30. [30]Ekström, A., Baardsen, G., Forssén, C., Hagen, G., Hjorth-Jensen, M., Jansen, G. R., Machleidt, R., Nazarewicz, W., Papenbrock, T., Sachi, J., & Wild, S. M. (2013). Optimized chiral nucleon-nucleon interaction at next-to-next-to-leading order. Phys. Rev. Lett., 110, 192502.
  31. [31]J. Carlson, V. R. Pandharipande, and R. B. Wiringa, Three-nucleon interaction in 3-, 4- and ∞-body systems, Nucl. Phys. A 401, 59 (1983).DOI
  32. [32]S. C. Pieper, V. R. Pandharipande, R. B. Wiringa, and J. Carlson, Realistic models of pion-exchange three-nucleon interactions, Phys. Rev. C 64, 014001 (2001).DOI
  33. [33]H.-W. Hammer, A. Nogga, and A. Schwenk, Colloquium: Three-body forces: From cold atoms to nuclei, Rev. Mod. Phys. 85, 197 (2013).DOI
  34. [34]K. Hebeler, J. D. Holt, J. Menéndez, and A. Schwenk, Nuclear forces and their impact on neutron-rich nuclei and neutron-rich matter, Annu. Rev. Nucl. Part. Sci. 65, 457 (2015).DOI
  35. [35]K. Hebeler, Three-nucleon forces: Implementation and applications to atomic nuclei and dense matter, Phys. Rep. 890, 1 (2021).DOI
  36. [36]W. Dickhoff and C. Barbieri, Self-consistent Green’s function method for nuclei and nuclear matter, Prog. Part. Nucl. Phys. 52, 377 (2004).DOI
  37. [37]B. R. Barrett, P. Navrátil, and J. P. Vary, Ab initio no core shell model, Prog. Part. Nucl. Phys. 69, 131 (2013).DOI
  38. [38]V. Somà, C. Barbieri, and T. Duguet, Ab initio Gorkov-Green’s function calculations of open-shell nuclei, Phys. Rev. C 87, 011303 (2013).DOI
  39. [39]G. Hagen, T. Papenbrock, M. Hjorth-Jensen, and D. J. Dean, Coupled-cluster computations of atomic nuclei, Rep. Prog. Phys. 77, 096302 (2014).DOI
  40. [40]H. Hergert, S. K. Bogner, T. D. Morris, A. Schwenk, and K. Tsukiyama, The in-medium similarity renormalization group: A novel ab initio method for nuclei, Phys. Rep. 621, 165 (2016).DOI
  41. [41]K. D. Launey, T. Dytrych, and J. P. Draayer, Symmetry-guided large-scale shell-model theory, Prog. Part. Nucl. Phys. 89, 101 (2016).DOI
  42. [42]S. R. Stroberg, H. Hergert, S. K. Bogner, and J. D. Holt, Nonempirical Interactions for the Nuclear Shell Model: An Update, Annu. Rev. Nucl. Part. Sci. 69, 307 (2019).DOI
  43. [43]A. Tichai, R. Roth, and T. Duguet, Many-body perturbation theories for finite nuclei, Front. Phys. 8, 164 (2020).arxiv.org/abs/2001.10433
  44. [44]M. Heinz, A. Tichai, J. Hoppe, K. Hebeler, and A. Schwenk, In-medium similarity renormalization group with three-body operators, Phys. Rev. C 103, 044318 (2021).DOI
  45. [45]Piarulli, M., Baroni, A., Girlanda, L., Kievsky, A., Lovato, A., Lu, E., Marcucci, L. E., Pieper, S. C., Schiavilla, R., Viviani, M., & Wiringa, R. B. (2017). Light-nuclei spectra from chiral dynamics. In ArXiv e-prints.arXiv
  46. [46]Rothman, M., Johnson-Toth, B., Hagen, G., Heinz, M., & Papenbrock, T. (2026). NuLattice: Ab initio computations of. Eur. Phys. J. A, 62, 28.
  47. [47]Wang, M., Huang, W. J., Kondev, F. G., Audi, G., & Naimi, S. (2021). The AME 2020 atomic mass evaluation (II). Tables, graphs and references. Chin. Phys. C, 45, 030003.
  48. [48]The auxiliary field Monte Carlo calculations of Refs. [19, 20, 65–70] solved the many-body problem by applying the transfer matrix formalism [71] with a discrete temporal lattice spacing $a_t = 1/(150\ \mathrm{MeV})$. Such calculations are not equivalent to calculations with $a_t \to 0$ (see Supplemental Material). Based on our comparison to Refs. [19, 20], the differences are significant.. (n.d.). The auxiliary field Monte Carlo calculations of Refs. [19, 20, 65–70] solved the many-body problem by applying the transfer matrix formalism [71] with a discrete temporal lattice spacing $a_t = 1/(150\ \mathrm{MeV})$. Such calculations are not equivalent to calculations with $a_t \to 0$ (see Supplemental Material). Based on our comparison to Refs. [19, 20], the differences are significant..
  49. [49]S. Gandolfi, A. Y. Illarionov, K. E. Schmidt, F. Pederiva, and S. Fantoni, Quantum Monte Carlo calculation of the equation of state of neutron matter, Phys. Rev. C 79, 054005 (2009).DOI
  50. [50]G. Hagen, T. Papenbrock, A. Ekström, K. A. Wendt, G. Baardsen, S. Gandolfi, M. Hjorth-Jensen, and C. J. Horowitz, Coupled-cluster calculations of nucleonic matter, Phys. Rev. C 89, 014319 (2014).DOI
  51. [51]F. Marino, W. G. Jiang, and S. J. Novario, Diagrammatic ab initio methods for infinite nuclear matter with modern chiral interactions, Phys. Rev. C 110, 054322 (2024).arxiv.org/abs/2407.17098
  52. [52]Colizza, C. (2025). The Anatomy of a Wirecutter Guide. <a href="https://doi.org/https://www.nytimes.com/wirecutter/reviews/anatomy-of-a-guide/" rel="noopener noreferrer" target="_blank">https://doi.org/https://www.nytimes.com/wirecutter/reviews/anatomy-of-a-guide/</a>
  53. [53]Elhatisari, S. (2026). private communications.
  54. [54]M. Rothman, G. Hagen, M. Heinz, and T. Papenbrock, Nulattice v1.2: Saturation of nuclear binding from lattice hamiltonians (2026).DOI
  55. [56]O. Agar, Z. Ren, and S. Elhatisari, From binding and saturation to criticality in nuclear matter with lattice effective field theory, arXiv:2604.09154.arxiv.org/abs/2604.09154
  56. [57]A. Nogga, S. K. Bogner, and A. Schwenk, Low-momentum interaction in few-nucleon systems, Phys. Rev. C 70, 061002 (2004).DOI
  57. [58]S. K. Bogner, R. J. Furnstahl, R. J. Furnstahl, and A. Nogga, Is nuclear matter perturbative with low-momentum interactions?, Nucl. Phys. A 763, 59 (2005), nucl-th/0504043.DOI
  58. [59]S. Bogner, R. Furnstahl, and A. Schwenk, From low-momentum interactions to nuclear structure, Prog. Part. Nucl. Phys. 65, 94 (2010).DOI
  59. [60]F. Alp, Y. Dietz, K. Hebeler, and A. Schwenk, Equation of state and Fermi liquid properties of dense matter based on chiral effective field theory interactions, Phys. Rev. C 112, 055802 (2025).arxiv.org/abs/2504.18259
  60. [61]Z.-W. Niu and B.-N. Lu, Sign-problem-free nuclear quantum Monte Carlo simulation, Phys. Rev. Lett. 135, 222504 (2025).DOI
  61. [62]E. H. Lieb, The stability of matter, Rev. Mod. Phys. 48, 553 (1976).DOI
  62. [63]M. Freer, H. Horiuchi, Y. Kanada-En’yo, D. Lee, and U.-G. Meißner, Microscopic clustering in light nuclei, Rev. Mod. Phys. 90, 035004 (2018).DOI
  63. [64]T. Otsuka, A. Volya, and N. Itagaki, Theoretical studies of α clustering in nuclei and beyond, Eur. Phys. J. A 62, 108 (2026), arXiv:2602.24175 [nucl-th].arxiv.org/abs/2602.24175
  64. [65]E. Epelbaum, H. Krebs, D. Lee, and U.-G. Meißner, Lattice calculations for A = 3, 4, 6, 12 nuclei using chiral effective field theory, Eur. Phys. J. A 45, 335 (2010).DOI
  65. [66]E. Epelbaum, H. Krebs, D. Lee, and U.-G. Meißner, Lattice effective field theory calculations for A = 3, 4, 6, 12 nuclei, Phys. Rev. Lett. 104, 142501 (2010).pubmed.ncbi.nlm.nih.gov/20481934
  66. [67]E. Epelbaum, H. Krebs, D. Lee, and U.-G. Meißner, Ab Initio Calculation of the Hoyle State, Phys. Rev. Lett. 106, 192501 (2011).pubmed.ncbi.nlm.nih.gov/21668146
  67. [68]E. Epelbaum, H. Krebs, T. A. Lähde, D. Lee, and U.-G. Meißner, Structure and rotations of the Hoyle state, Phys. Rev. Lett. 109, 252501 (2012).pubmed.ncbi.nlm.nih.gov/23368455
  68. [69]E. Epelbaum, H. Krebs, T. A. Lähde, D. Lee, and U.-G. Meißner, Viability of carbon-based life as a function of the light quark mass, Phys. Rev. Lett. 110, 112502 (2013).pubmed.ncbi.nlm.nih.gov/25166526
  69. [70]E. Epelbaum, H. Krebs, T. A. Lähde, D. Lee, U.-G. Meißner, and G. Rupak, Ab initio calculation of the spectrum and structure of $^{16}\mathrm{O}$, Phys. Rev. Lett. 112, 102501 (2014).pubmed.ncbi.nlm.nih.gov/24679283
  70. [71]D. Lee, Lattice simulations for few- and many-body systems, Prog. Part. Nucl. Phys. 63, 117 (2009).DOI
  71. [72]Furnstahl, R. J., Klco, N., Phillips, D. R., & Wloslaweski, S. (2015). Quantifying truncation errors in effective field theory. Phys. Rev. C, 92, 024005.
  72. [73]C. Mishra, A. Ekström, G. Hagen, T. Papenbrock, and L. Platter, Two-pion exchange as a leading-order contribution in chiral effective field theory, Phys. Rev. C 106, 024004 (2022).DOI

Paper details

vlnc:W_qhk9e7cg
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 10

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