Superfluid response of bosonic fluids in composite optical potentials: angular dependence and Leggett’s bounds
Daniel Pérez-Cruz, Grigori E. Astrakharchik, Pietro Massignan · 2026
Read Superfluid response of bosonic fluids in composite optical potentials: angular dependence and Leggett’s bounds by Daniel Pérez-Cruz, Grigori E. Astrakharchik, Pietro Massignan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-29603.txt.
Introduction
Pérez-Cruz and co-authors study the superfluid response of a dilute bosonic fluid in composite light potentials. The potentials include triangular, Kagomé, and quasiperiodic patterns made from laser beams. They first find when the response must be fully isotropic. Then they derive Leggett's upper and lower bounds for the superfluid fraction. The formulas show the best direction to measure each bound. The bounds hold in the perturbative regime. Numerical solutions confirm the analytical results. The work maps how response direction matters in these traps.
Catalog record
| CREATOR | Daniel Pérez-Cruz, Grigori E. Astrakharchik, Pietro Massignan |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25905 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.29603
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Superfluid response of bosonic fluids in composite optical potentials: angular dependence and Leggett’s bounds Daniel Pérez-Cruz, Grigori E. Astrakharchik, Pietro Massignan Abstract We study the superfluid response of a dilute bosonic fluid in the presence of two-dimensional composite potentials (such as triangular, Kagomé and quasiperiodic potentials, or superlattices), which may be obtained for example by superposing multiple laser beams. We first find a sufficient condition for the external potential to yield a fully isotropic superfluid response. Then, we derive analytical expressions for Leggett’s upper and lower bounds to the superfluid fraction (valid in the perturbative regime) that allow us to find the optimal direction along which each bound should be measured. Finally, we solve the problem numerically, and we confirm our analytical findings. Superfluidity is one of the most striking effects appearing in quantum many-body physics. It is found in a wide variety of systems, which encompass liquid helium, where it was first discovered 17 , ultracold bosons 5 , strongly-interacting fermions 18 and even very hot neutron stars 2 , and it is tightly linked to superconductivity of solid-state materials. The extreme versatility of ultracold atom platforms made it possible to study and characterize in depth this crucial phenomenon 26 . The two-fluid model, first introduced by Gorter and Casimir in Ref. 14 and later formalized within Landau’s hydrodynamic formulation 19 ; 35 ; 36 ; 34 , predicts that a Galilean-invariant bosonic quantum fluid in its ground state must be fully superfluid. When Galilean invariance is broken, however, a normal component can emerge even at zero temperature. This phenomenon has been studied in a variety of systems, such as condensates in optical lattices 33 ; 6 ; 25 ; 29 , where the Mott insulators are a prime example 8 ; 15 , supersolids 32 ; 27 ; 3 ; 4 , spin-orbit coupled condensates 22 ; 12 or disordered systems 10 ; 11 ; 25 ; 23 . Currently, intense research is devoted to characterizing the degree of superfluidity present in a system. This information is encoded in the so-called superfluid fraction tensor. A significant breakthrough in the field occurred when Anthony Leggett proposed an upper and a lower bound for this quantity 20 ; 21 . The upper bound is variational and therefore rigorous, whilst the lower bound comes from a heuristic argument and applies only to weakly-interacting systems. Recent work has explored their applicability in bosonic 25 ; 6 ; 4 ; 29 ; 33 and fermionic systems 24 , finding Leggett’s bounds to be of great use as a practical and accurate estimator of the superfluid fraction. In this work, we demonstrate that many widely-employed composite optical potentials which feature a discrete rotational invariance (such as square, triangular, Kagomé and certain quasicrystals, or superlattices) nonetheless yield a fully-isotropic superfluid response. Leggett’s bounds have instead a strong angular dependence, and here we identify the optimal directions along which the bounds are tightest. Finally, we confirm our analytical predictions by comparing them to numerical studies of relevant experimental configurations. Superfluid fraction and perturbation theory. — To study the superfluid fraction of a dilute Bose-Einstein condensate (BEC) in the presence of an external static potential V(\mathbf{r}) , we consider an additional weak potential that drags the normal component of the condensate with velocity \mathbf{v}_{0} . According to linear response theory, a weak drag couples only to the normal fraction, while the part of the system unaffected by the drag constitutes the superfluid fraction 9 . More specifically, given a system of N particles of mass m , its superfluid fraction is a tensor defined through the response along the direction \alpha to a perturbation moving with velocity v_{\beta} along \beta : f_{\alpha\beta}=1-\lim_{v\rightarrow 0}\langle\hat{P}_{\alpha}\rangle/(Nmv_{\beta}), (1) where \langle\hat{P}_{\alpha}\rangle is the mean total momentum along \alpha . In the frame co-moving with the dragging potential, the Gross-Pitaevskii equation (GPE) reads \frac{(-i\hbar\nabla-m\mathbf{v}_{0})^{2}}{2m}\psi(\mathbf{r})+V(\mathbf{r})\psi(\mathbf{r})+g|\psi|^{2}\psi(\mathbf{r})=\mu\psi(\mathbf{r}), (2) where g>0 parametrizes the inter-particle repulsion, and \mu is the chemical potential. In a uniform system of density n_{0} , we have \mu_{0}=gn_{0} and a healing length \xi=\hbar^{2}/\sqrt{2m\mu_{0}} . In Fourier space, the co-moving GPE reads \sum_{\mathbf{l}}\bigg[(\varepsilon_{\mathbf{l}-\mathbf{k}_{0}}-\mu)\delta_{\mathbf{k},\mathbf{l}}+\frac{V(\mathbf{k}-\mathbf{l})}{\sqrt{\mathcal{V}}}+\tilde{W}(\mathbf{k},\mathbf{l})\bigg]\psi_{\mathbf{k}_{0}}(\mathbf{l})=0, (3) with f(\mathbf{k})=\mathcal{V}^{-1/2}\int d\mathbf{r}f(\mathbf{r})e^{-i\mathbf{k}\mathbf{r}} the Fourier transform, \tilde{W}(\mathbf{k},\mathbf{l})=(g/\mathcal{V})\sum_{\mathbf{p}}\psi_{\mathbf{k}_{0}}^{*}(\mathbf{p}\!-\!\mathbf{k})\psi_{\mathbf{k}_{0}}(\mathbf{p}\!-\!\mathbf{l}) the interaction term, \varepsilon_{\mathbf{k}}=\hbar^{2}k^{2}/(2m) and \mathcal{V} the system volume. The superfluid fraction then follows from the total momentum \langle\hat{\mathbf{P}}\rangle=\frac{\hbar}{2i}\int\text{d}\mathbf{r}\;\left[\psi_{\mathbf{k}_{0}}^{*}(\mathbf{r})\,\nabla\psi_{\mathbf{k}_{0}}(\mathbf{r})-\psi_{\mathbf{k}_{0}}(\mathbf{r})\,\nabla\psi_{\mathbf{k}_{0}}^{*}(\mathbf{r})\right]. (4) This equation can be solved perturbatively in both the external potential and the drag wavevector \mathbf{k}_{0}=m\mathbf{v}_{0}/\hbar 10 ; 11 ; 1 . To linear order in both quantities the solution is \psi_{\mathbf{k}_{0}}({\bf k})=\sqrt{N/\mathcal{V}}+\delta\psi_{{\bf k}_{0}}({\bf k}) , with \displaystyle\delta\psi_{\mathbf{k}_{0}}(\mathbf{k})=-\sqrt{\frac{N}{\mathcal{V}}}\frac{V(\mathbf{k})(1-\delta_{\mathbf{k},\mathbf{0}})}{(\varepsilon_{k}+2\mu_{0})}\left(1+\frac{{\mathbf{k}}\cdot{\mathbf{k}}_{0}}{\epsilon_{k}}\right). (5) Substituting Eq. ( 5 ) in Eq. ( 4 ) yields \langle\hat{\mathbf{P}}\rangle=4\sum_{\mathbf{k}}\hat{\mathbf{k}}(\hat{\mathbf{k}}\cdot(m\mathbf{v}_{0}))|\delta\psi_{{\bf k}_{0}=0}(\mathbf{k})|^{2}, (6) an expression correct up to second order in V and first order in \mathbf{k}_{0} . Here \hat{\mathbf{k}}=\mathbf{k}/k denotes a unit vector with components \hat{k}_{i} . Thus, the superfluid fraction tensor ( 1 ) can be perturbatively expressed as 10 ; 11 ; 1 f_{ij}=\delta_{ij}-\frac{4}{\mathcal{V}}\sum_{\mathbf{k}}\frac{|V(\mathbf{k})|^{2}}{(\varepsilon_{k}+2\mu_{0})^{2}}\,\hat{k}_{i}\hat{k}_{j}, (7) a result that can also be derived within the formalism of Bogoliubov theory 13 . Table 1: Examples of potentials formed by M shells with N_{l} travelling waves per shell. Here, q_{l} and V_{l} denote the wavevector and amplitude of peaks in the l -th shell, and \varphi is the golden ratio. Structure M N_{l} Parameters Ref. Square 1 4 V_{1} – Superlattice 2 4 {V_{1},V_{2}} – Triangular/Hexagonal 1 6 V_{1} – Kagomé 2 {6,6} V_{2}=V_{1}/2 , q_{2}=2q_{1} 16 Five-fold quasicrystal 2 {10,10} V_{2}=V_{1} , q_{2}=\varphi q_{1} 31 Superfluid fraction in composite potentials — Superposing multiple laser beams in regular arrangements permits the generation of a large variety of exotic optical potentials. In the following, we consider the general family of two-dimensional (2D) potentials which may be expressed as V(\mathbf{r})=\sum_{l=1}^{M}\frac{V_{l}}{2}\sum_{j=1}^{N_{l}}\exp\left(i\mathbf{q}_{lj}\cdot\mathbf{r}\right)=\sum_{l=1}^{M}V_{l}\sum_{j=1}^{N_{l}/2}\cos\left(\mathbf{q}_{lj}\cdot\mathbf{r}\right) (8) with \mathbf{q}_{lj}=q_{l}\left[\cos(2\pi j/N_{l}),\sin(2\pi j/N_{l})\right] . Here V_{l} denotes the strength of the potential in each “shell”, and all N_{l} need to be even for the potential to be real-valued. Their power spectrum consists of sharp peaks forming M nested regular polygons, each having N_{l} edges: |V(\mathbf{k})|^{2}=\frac{\mathcal{V}}{4}\sum_{l=1}^{M}V_{l}^{2}\sum_{j=1}^{N_{l}}\delta_{\mathbf{k},\mathbf{q}_{lj}}. (9) Notably, this family of potentials covers a wide range of experimentally relevant configurations (see Table 1 for a few examples, and Fig. 1 for a specific case with two shells). For details on how the optical potentials generated by different configurations of laser beams can be expressed in this form, see End Matter. Figure 1: Composite optical potentials. Left: Typical momentum structure. The example shows a case with two shells ( M=2 ), with the inner and outer shells generating square and triangular/hexagonal lattices, respectively ( N_{1}=4,\,N_{2}=6 ). Right: Resulting potential in real space. Figure 2: Ground state and superfluid response in composite potentials. Top and bottom rows correspond respectively to a BEC (modeled by the GPE, with interaction strength gN=14400\hbar^{2}/m ) in a Kagomé potential and a five-fold quasicrystal (see definitions in Table 1 ). From left to right, the panels show: the external potential V(\mathbf{r}) , the ground state density, the momentum density (circles indicate peaks in the Fourier transform of the external potential, with yellow peaks being four times stronger than the red ones), and the superfluid fraction [dashed lines correspond to the perturbative results, Eqs. ( 7 , 16 , 17 )]. In 2D the superfluid fraction is a 2*2 tensor, and the response along the direction \hat{n}=(\cos\phi,\sin\phi) is f(\phi)=f_{xx}\cos^{2}\phi+f_{yy}\sin^{2}\phi+(f_{xy}+f_{yx})\sin\phi\cos\phi. (10) Since \frac{\partial