betaFeedback

WorksPhysics and Astronomy

Intercomponent vortex interaction and dynamics in a uniform Bose quantum liquid

Abstract

We investigate the interaction and dynamics of vortices belonging to different components of a binary uniform quantum liquid. Using the coupled extended Gross-Pitaevskii equations with logarithmic beyond-mean-field corrections, we derive the asymptotic interaction potential between well-separated vortices. The interaction potential retains the asymptotic form s−2s^{-2} as a function of the intervortex distance s, but its coefficient depends on the background density through the quantum fluctuation. As a result, the interaction between vortices changes from attractive to repulsive at a critical density. Real-time simulations demonstrate a corresponding reversal of orbital motion of a vortex pair from clockwise to counterclockwise, consistent with their Magnus response.

Introduction

Bose-Einstein condensates (BECs) of ultracold atoms provide a highly controllable platform through tunable interatomic interactions [1, 2]. Within the mean-field approximation [3, 4], weakly interacting BECs are well described by the Gross-Pitaevskii (GP) equation. In binary liquid of Bose mixtures, however, the Lee-Huang-Yang (LHY) [5] correction can stabilize a droplet against collapse and support a self-bound droplet [6, 7, 8, 9, 10, 11, 12]. In a two-dimensional (2D) model, the fluctuation-corrected energy exhibits a logarithmic density dependence, leading to a minimum of the energy at a finite density. Experimental observations of droplets in atomic mixtures [13, 14, 15] and dipolar gases [16, 17] have stimulated growing interest in quantum liquids. Their collective and topological excitations, such as vortices [18, 19, 20], provide a useful probe of how the equation of state influences superfluid behavior beyond the equilibrium properties of the liquid.

Quantized vortices are particularly sensitive to the density of the surrounding fluid. A vortex combines phase winding with a density depletion, and its structure and dynamics are governed by the energetic cost of this deformation. Theoretical studies have established the existence and stability of 2D vortex droplets [21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33] and demonstrated that background density affects vortex structure and stability in uniform Bose quantum liquids [34]. Binary systems introduce an additional degree of freedom because phase winding can reside in either component independently. A vortex in one component therefore perturbs the density of the other, providing a mechanism for interaction between spatially separated vortices belonging to different components.

For two-component condensates coupled through their densities, this mechanism produces an interaction distinct from that between vortices in a single component. Eto et al. [35] derived the asymptotic intercomponent vortex interaction, which scales as ln⁡(s/ξ)/s2\ln(s/\xi)/s^2, where ss denotes the intervortex separation and ξ\xi is a short-distance cutoff. The sign of interaction is determined by the intercomponent coupling and is independent of the vortex circulations. Subsequent studies examined the interaction at shorter separation distances and related vortex trajectories to the balance between the interaction force and the Magnus force [36]. In circular confinement, boundary-induced flows were shown to modify the vortex dynamics [37]. These studies established a framework relating intervortex interaction to vortex motion.

In a binary quantum liquid, the LHY correction introduces a density dependence into the response that mediates the intercomponent vortex interaction. Changing the background density modifies the relative energetic costs of total-density and relative density deformations. As a result, the density disturbance induced by a vortex in the other component can change character, suggesting that the intercomponent vortex interaction may be controlled through the background density even when the microscopic coupling parameters remain fixed. A central question is whether this mechanism can reverse the interaction force and the resulting vortex motion, and whether a critical density exists at which the leading logarithmic interaction vanishes.

In this work, we investigate the interaction and dynamics of two initially separated half-quantized vortices belonging to different components of a 2D binary quantum liquid. Using the coupled extended GP equations with the logarithmic LHY correction, we derive the asymptotic intervortex interaction potential and identify a density-dependent coefficient that changes sign with the background density. This leads to a critical density separating attractive and repulsive regimes. Real-time simulations show a corresponding reversal of orbital motion: for two vortices with positive unit circulation, clockwise motion at low density changes to counterclockwise motion at high density. At the predicted critical density, the leading asymptotic interaction vanishes, and no resolvable vortex motion is observed over the simulated time interval. These results demonstrate how the fluctuation-modified equation of state controls the intervortex interactions and vortex dynamics in multicomponent quantum liquids.

The paper is organized as follows: § 2 introduces the model, confinement, and numerical method. Section III derives the density-dependent intervortex interaction, identifies the critical density, and compares it with numerical results. We explain the reversal of vortex motion and present their numerical dynamics in § 4. § 5 summarizes the results and discusses their implications.

Model and Numerical Framework

Extended Gross-Pitaevskii model

We consider a 2D homonuclear Bose mixture of atoms having mass mm, with repulsive intracomponent and attractive intercomponent interactions. The two components are coupled through their densities and are not coherently interconverted. Including the leading 2D quantum-fluctuation contribution, the dynamics is described by the coupled extended GP equations [7],

iℏ∂Ψj∂t=[−ℏ22m∇2+V(r)+Λ(nj−nk)+Gnln⁡(nnr)]Ψj,j≠k,(1)i\hbar\frac{\partial\Psi_j}{\partial t}=\left[-\frac{\hbar^2}{2m}\nabla^2+V(\mathbf{r})+\Lambda(n_j-n_k)+Gn\ln\left(\frac{n}{n_r}\right)\right]\Psi_j,\qquad j\ne k, \tag*{(1)}

where nj=∣Ψj∣2n_j=\lvert\Psi_j\rvert^2 is the density of jjth component, n=n1+n2n=n_1+n_2 is the total density, and V(r)V(\mathbf{r}) is the external potential. The coefficient Λ=4πℏ2mln⁡(∣a12∣/a)>0\Lambda=\frac{4\pi\hbar^2}{m\ln(\lvert a_{12}\rvert/a)}>0 is the modified coupling constant, with akja_{kj} being the 2D scattering lengths between component jj and kk, and a=a11=a22a=a_{11}=a_{22}. It determines the energetic cost of a density imbalance between the components. The logarithmic term describes the fluctuation-corrected response of total density with a coefficient G=Λln⁡(∣a12∣/a)>0G=\frac{\Lambda}{\ln(\lvert a_{12}\rvert/a)}>0. The reference density nrn_r in the last term is related to the equilibrium density per component n0n_0 through nr=2en0n_r=2\sqrt{e}n_0, where

n0=e−2γE−3/22πln⁡(∣a12∣/a)a∣a12∣,(2)n_0=\frac{e^{-2\gamma_E-3/2}}{2\pi}\frac{\ln(\lvert a_{12}\rvert/a)}{a\lvert a_{12}\rvert}, \tag*{(2)}

with γE\gamma_E being the Euler’s constant.

We rescale the position r=r~λ\mathbf{r}=\tilde{\mathbf{r}}\lambda, time t=στt=\sigma\tau, and the wave function Ψj(r,t)=nrΨ~j(r~,τ)\Psi_j(\mathbf{r},t)=\sqrt{n_r}\tilde{\Psi}_j(\tilde{\mathbf{r}},\tau), where λ=ℏ/2mGnr\lambda=\hbar/\sqrt{2mGn_r} and σ=ℏ/(Gnr)\sigma=\hbar/(Gn_r). Writing Eq. (1) in a dimensionless form,

i∂Ψ~j∂τ=[−∇~2+V~(r~)+2β(n~j−n~k)+n~ln⁡n~]Ψ~j,(3)i\frac{\partial\tilde{\Psi}_j}{\partial\tau}=\left[-\tilde{\nabla}^{2}+\tilde{V}(\tilde{\mathbf{r}})+2\beta(\tilde{n}_j-\tilde{n}_k)+\tilde{n}\ln\tilde{n}\right]\tilde{\Psi}_j, \tag*{(3)}

where

β=Λ2G=12ln⁡(∣a12∣a)(4)\beta=\frac{\Lambda}{2G}=\frac{1}{2}\ln\left(\frac{\lvert a_{12}\rvert}{a}\right) \tag*{(4)}

is the scaled interaction coefficient and V~=V/(Gnr)\tilde{V}=V/(Gn_r) is the scaled potential. We omit the tildes from the dimensionless quantities below.

The associated dimensionless energy functional is defined as

E[Ψ]=Ekin[Ψ]+Eint[Ψ]+Etrap[Ψ].(5)E[\Psi]=E_{\mathrm{kin}}[\Psi]+E_{\mathrm{int}}[\Psi]+E_{\mathrm{trap}}[\Psi]. \tag*{(5)}

Here, the kinetic, contact interaction, and potential energies are given by,

Ekin[Ψ]=∫d2r∑j=12∣∇Ψj∣2,(6)E_{\mathrm{kin}}[\Psi]=\int d^2r\sum_{j=1}^{2}\lvert\nabla\Psi_j\rvert^2, \tag*{(6)}
Eint[Ψ]=∫d2r[β(n1−n2)2+n22(ln⁡n−12)],(7)E_{\mathrm{int}}[\Psi]=\int d^2r\left[\beta(n_1-n_2)^2+\frac{n^2}{2}\left(\ln n-\frac{1}{2}\right)\right], \tag*{(7)}

and

Etrap[Ψ]=∫d2r V(r)n,(8)E_{\mathrm{trap}}[\Psi]=\int d^2r\,V(\mathbf{r})n, \tag*{(8)}

respectively. In Eq. (7), the first interaction term penalizes local density imbalance between the components and the second one combines the residual mean-field interaction and the LHY correction to the total-density energy. Each component norm Nj=∫njd2rN_j=\int n_jd^2r is conserved in real time, and the total number of atoms can be calculated as nrλ2(N1+N2)n_r\lambda^2(N_1+N_2).

Homogeneous state and density control

In the absence of an external potential, a balanced homogeneous solution is

Ψ1=Ψ2=nˉ2e−iμ(nˉ)τ,(9)\Psi_1=\Psi_2=\sqrt{\frac{\bar{n}}{2}}e^{-i\mu(\bar{n})\tau}, \tag*{(9)}

with the background density nˉ\bar{n} and chemical potential μ(nˉ)=nˉln⁡nˉ\mu(\bar{n})=\bar{n}\ln\bar{n}. The associated pressure and total-density stiffness are

P(nˉ)=nˉ22(ln⁡nˉ+12),(10)P(\bar{n})=\frac{\bar{n}^{2}}{2}\left(\ln\bar{n}+\frac{1}{2}\right), \tag*{(10)}
ge(nˉ)=dμdn∣nˉ=1+ln⁡nˉ.(11)g_{e}(\bar{n})=\left.\frac{d\mu}{dn}\right|_{\bar{n}}=1+\ln\bar{n}. \tag*{(11)}

Thus, the zero-pressure bulk density is neq=e−1/2≃0.60653n_{\mathrm{eq}}=e^{-1/2}\simeq0.60653, whereas the positive compressibility requires nˉ>e−1\bar{n}>e^{-1}.

To explore a broader density range, we confine the mixture in a circular box,

V(r)={0r<Rb,Vbr≥Rb,(12)V(\mathbf{r})= \begin{cases} 0 & r<R_{b},\\ V_{b} & r\ge R_{b}, \end{cases} \tag*{(12)}

where r=x2+y2r=\sqrt{x^{2}+y^{2}}, Vb=10V_{b}=10 is the height and Rb=126R_{b}=126 is the radius of the potential. The background density is varied through the total norm N=N1+N2N=N_{1}+N_{2}, keeping the populations balanced and the barrier parameters fixed. For a nearly uniform liquid filling the box, nˉ≃N/(πRb2)\bar{n}\simeq N/(\pi R_{b}^{2}) provides an estimate. The relevant density for the vortex theory is the measured interior plateau, away from cores and the boundary layer. An operational definition is

nˉ=1∣B∣∫Bn(r)d2r,(13)\bar{n}=\frac{1}{|B|}\int_{B}n(r)d^{2}r, \tag*{(13)}

where BB is a bulk sampling region. In contrast to an unconfined self-bound droplet, the box wall supplies the pressure required to reach densities larger than the equilibrium value.

Preparation and propagation

We solve Eq. (3) numerically using the split-operator Fourier pseudospectral method [38]. The numerical grid consists of 2048×20482048\times2048 points in a domain of size (320)2(320)^{2}, giving dimensionless spacings Δx=Δy=0.156\Delta x=\Delta y=0.156. The time step is dt=0.004dt=0.004. Fourier propagation imposes periodicity on the computational square, so the density must be negligible at its edges. The circular wall provides the physical confinement inside that square.

An initially vortex-free state is relaxed in imaginary time at fixed, equal component norms. A vortex is imprinted in each component by multiplying the corresponding wave function by the phase factor eiΘj(r)e^{i\Theta_{j}(\mathbf{r})}, where

Θj(r)=wjtan⁡−1(y−yjx−xj),(j=1,2)(14)\Theta_{j}(\mathbf{r})=w_{j}\tan^{-1}\left(\frac{y-y_{j}}{x-x_{j}}\right),\qquad(j=1,2) \tag*{(14)}

with wj=±1w_{j}=\pm1 and (xj,yj)(x_{j},y_{j}) denoting the circulation sign and the xx-yy coordinates of the vortex in the jjth component, respectively. We label the vortices such that y1>y2y_{1}>y_{2}. The two vortices are displaced symmetrically from the center of the box along the yy-axis by an inter-vortex distance ss, with initial core positions r1,2=(0,±s/2)\mathbf{r}_{1,2}=(0,\pm s/2). A short imaginary-time relaxation reduces the excitation energy introduced by imprinting.

Asymptotic intervortex interaction

Vortex profiles and density response

We label a vortex in component 1 by "(w1,0)(w_{1},0)" and one in component 2 by "(0,w2)(0,w_{2})" with circulation direction wj=±1w_{j}=\pm1. Each component has an independent U(1) phase. Around a (1,0)(1,0) vortex, the phase of component 1 changes by 2π2\pi and that of component 2 does not change. The common phase (Θ1+Θ2)/2(\Theta_{1}+\Theta_{2})/2 therefore winds by π\pi. For equal far-field densities, the resulting mass circulation is half that of a vortex winding both components, hence the term half-quantized vortex. The phase winding in the occupied component itself is an integer. Figure 1 shows the density profiles of the individual components and the total density. The density of the winding component vanishes at the vortex center, whereas the non-winding component retains a finite density.

Figure 1. Relaxed density profiles along x=0x = 0 for vortices at (0,±50)(0, \pm50), with β=1\beta= 1. The left and right columns show the individual and total densities, respectively, for (a) nˉ=0.612\bar{n} = 0.612, (b) nˉ=2.718\bar{n} = 2.718, and (c) nˉ=3.5\bar{n} = 3.5. The insets in (b) and (c) enlarge the density of the non-winding component near the vortex core at (0,50)(0, 50). The non-winding component remains nonzero at the vortex center and exhibits a transition from depletion to enhancement as the background density increases. Coordinates and densities are normalized by λ\lambda and nrn_r, respectively.

For an isolated (w1,0)(w_{1},0) vortex in a uniform background, we write the stationary fields as

Ψ1(w1,0)=nˉ2 u(r)eiw1θ,Ψ2(w1,0)=nˉ2 v(r),(15)\Psi_{1}^{(w_{1},0)}=\sqrt{\frac{\bar{n}}{2}}\,u(r)e^{iw_{1}\theta},\qquad \Psi_{2}^{(w_{1},0)}=\sqrt{\frac{\bar{n}}{2}}\,v(r), \tag*{(15)}

where rr and θ\theta are polar coordinates. The real functions satisfy u(0)=0u(0)=0, v′(0)=0v'(0)=0, and (u,v)→(1,1)(u,v)\to(1,1) as r→∞r\to\infty, with the prime denotes differentiation with respect to rr. The corresponding (0,w2)(0,w_{2}) configuration is obtained by exchanging the component indices.

By linearizing the stationary extended GP equation with V(r)=0V(\mathbf{r})=0,

μ(nˉ)Ψj=[−∇2+nln⁡n+2β(nj−nk)]Ψj,j≠k(16)\mu(\bar{n})\Psi_{j}=\left[-\nabla^{2}+n\ln n+2\beta(n_{j}-n_{k})\right]\Psi_{j},\qquad j\ne k \tag*{(16)}

around the uniform background state and retaining only the leading density perturbations, we obtain two associated healing lengths

ξ+−2=2nˉ(1+ln⁡nˉ)=2nˉge(nˉ),ξ−−2=4βnˉ.(17)\xi_{+}^{-2}=2\bar{n}(1+\ln\bar{n})=2\bar{n}g_{e}(\bar{n}),\qquad \xi_{-}^{-2}=4\beta\bar{n}. \tag*{(17)}

The healing lengths ξ+\xi_{+} and ξ−\xi_{-} characterizes the recovery of total-density and relative-density perturbations, respectively. For r≫max⁡(ξ+,ξ−)r\gg\max(\xi_{+},\xi_{-}), the vortex profiles approach their asymptotic forms,

u(r)=1−ξ+2+ξ−22r2+O(r−4),(18)u(r) = 1 - \frac{\xi_{+}^{2}+\xi_{-}^{2}}{2r^{2}} + O(r^{-4}), \tag*{(18)}
v(r)=1−ξ+2−ξ−22r2+O(r−4).(19)v(r) = 1 - \frac{\xi_{+}^{2}-\xi_{-}^{2}}{2r^{2}} + O(r^{-4}). \tag*{(19)}

The detailed derivation of Eq. (18) is given in the Appendix A. In particular, the far-field perturbation of the unwound component is

δn2(r)=−nˉ2r2(ξ+2−ξ−2)+O(r−4).(20)\delta n_{2}(r) = -\frac{\bar{n}}{2r^{2}}(\xi_{+}^{2}-\xi_{-}^{2}) + O(r^{-4}). \tag*{(20)}

Thus, the far-field density response of the unwound component changes sign at ge(nˉ)=2βg_e(\bar{n}) = 2\beta. It corresponds to a depletion for ge(nˉ)<2βg_e(\bar{n}) < 2\beta and an enhancement for ge(nˉ)>2βg_e(\bar{n}) > 2\beta, as illustrated in the Figure 1. At lower density, total-density depletion is less costly than component imbalance, so both components can deplete. At higher density, preserving the total density becomes energetically more favorable than maintaining equal component densities, and the unwound component compensates for the vortex-induced depletion.

Intervortex interaction force

For well-separated cores at r1,2=(0,±s/2)\mathbf{r}_{1,2} = (0, \pm s/2), we introduce the local polar coordinates ρj=∣r−rj∣\rho_j = |\mathbf{r} - \mathbf{r}_j| and an azimuthal angle θj\theta_j measured counterclockwise from the positive xx direction about the jjth vortex core (j=1,2j = 1, 2), and use the product ansatz

Ψ1(w1,w2)≃nˉ2u(ρ1)v(ρ2)eiw1θ1,(21)\Psi_{1}^{(w_1,w_2)} \simeq\sqrt{\frac{\bar{n}}{2}}u(\rho_1)v(\rho_2)e^{iw_1\theta_1}, \tag*{(21)}
Ψ2(w1,w2)≃nˉ2v(ρ1)u(ρ2)eiw2θ2.(22)\Psi_{2}^{(w_1,w_2)} \simeq\sqrt{\frac{\bar{n}}{2}}v(\rho_1)u(\rho_2)e^{iw_2\theta_2}. \tag*{(22)}

At fixed background chemical potential μ(nˉ)\mu(\bar{n}), the appropriate functional is the grand potential η=E−μ(nˉ)N\eta= E - \mu(\bar{n})N. Subtracting the two isolated vortex contributions and adding back the uniform reference gives the pair interaction as

U(s/2)=∫d2r (δK+δW),(23)U(s/2) = \int d^{2}r\,(\delta K + \delta W), \tag*{(23)}

with δK=K[Ψ(w1,w2)]−K[Ψ(w1,0)]−K[Ψ(0,w2)]\delta K = \mathcal{K}[\Psi^{(w_1,w_2)}] - \mathcal{K}[\Psi^{(w_1,0)}] - \mathcal{K}[\Psi^{(0,w_2)}] and δW=W[Ψ(w1,w2)]−W[Ψ(w1,0)]−W[Ψ(0,w2)]+W0\delta W = \mathcal{W}[\Psi^{(w_1,w_2)}] - \mathcal{W}[\Psi^{(w_1,0)}] - \mathcal{W}[\Psi^{(0,w_2)}] + W_0 being the differences in kinetic and local grand-potential densities, respectively. Here, W0W_0 denotes the local grand-potential density of the uniform, vortex-free background with n1=n2=nˉ/2n_1 = n_2 = \bar{n}/2.

By substituting Eqs. (18) and (20) in Eq. (23), the individual self-energy contributions cancel out. The leading surviving terms

δK≃−nˉ(ξ+2−ξ−2)ρ12ρ22+O(r−6),(24)\delta K \simeq-\frac{\bar{n}(\xi_{+}^{2}-\xi_{-}^{2})}{\rho_{1}^{2}\rho_{2}^{2}} + O(r^{-6}), \tag*{(24)}
δW≃nˉ(ξ+2−ξ−2)2ρ12ρ22+O(r−6).(25)\delta W \simeq\frac{\bar{n}(\xi_{+}^{2}-\xi_{-}^{2})}{2\rho_{1}^{2}\rho_{2}^{2}} + O(r^{-6}). \tag*{(25)}

produce the leading interaction

U(s/2)=−π[1ge(nˉ)−12β]ln⁡[s/(2ξ)]s2.(26)U(s/2) = -\pi\left[\frac{1}{g_e(\bar{n})} - \frac{1}{2\beta}\right]\frac{\ln[s/(2\xi)]}{s^2}. \tag*{(26)}

Here, we have used the relation nˉ(ξ+2−ξ−2)=12ge(nˉ)−14β\bar{n}(\xi_+^2-\xi_-^2)=\frac{1}{2g_e(\bar{n})}-\frac{1}{4\beta} and ξ=ξ(nˉ,β)\xi=\xi(\bar{n},\beta) is an effective short-distance cutoff accounting for matching the asymptotic profiles to the nonlinear core structure, such that s/2≫ξs/2 \gg\xi. For the detailed calculation of Eqs. (24)–(26), see Appendix B. The logarithmic spatial dependence also occurs in ordinary two-component BECs. The distinctive feature here is that its coefficient becomes density dependent through the LHY correction and can change sign as the background density is varied.

The asymptotic force between the two vortices is obtained by differentiating Eq. (26) with respect to the separation ss as

Fs=−∂U∂s≃−2π[1ge(nˉ)−12β]ln⁡[s/(2ξ)]−1/2s3.(27)F_s=-\frac{\partial U}{\partial s}\simeq-2\pi\left[\frac{1}{g_e(\bar{n})}-\frac{1}{2\beta}\right]\frac{\ln[s/(2\xi)]-1/2}{s^3}. \tag*{(27)}

Thus, in the asymptotic regime, the leading interaction is attractive for ge(nˉ)<2βg_e(\bar{n}) < 2\beta and repulsive for ge(nˉ)>2βg_e(\bar{n}) > 2\beta. The leading interaction vanishes for ge(nˉc)=2βg_e(\bar{n}_c) = 2\beta, giving the critical background density,

nˉc=e2β−1.(28)\bar{n}_c=e^{2\beta-1}. \tag*{(28)}

This sign reversal originates from the competition between the total-density stiffness (ge(nˉ)g_e(\bar{n})) and the relative-density stiffness (2β2\beta). At low density, where ge(nˉ)<2βg_e(\bar{n}) < 2\beta, total-density depletion is less costly than relative-density distortion, allowing both components to deplete around a vortex and resulting in an attractive interaction. In contrast, at high density, where ge(nˉ)>2βg_e(\bar{n}) > 2\beta, preserving the total density becomes energetically more favorable, and the unwound component compensates for the vortex-induced depletion, leading to repulsion. Figure 2 illustrates the resulting transition from attractive to repulsive intervortex interaction.

Figure 2. Intervortex potential in Eq. (26) as a function of half separation s/2s/2 for β=1\beta= 1 and ξ=ξ−\xi= \xi_-. The interaction is attractive for nˉ<e\bar{n} < e, vanishes at nˉ=e\bar{n} = e, and becomes repulsive for nˉ>e\bar{n} > e. The inset enlarges the positive branch of the potential.

It should be noted that the asymptotic interaction is independent of the circulation signs. Reversing either wjw_j leaves the static interaction unchanged because the two components are coupled only through densities n1n_1 and n2n_2. Therefore, all combinations of circulation signs share the same asymptotic interaction given by Eq. (26).

The interaction potential is numerically estimated by constrained minimization of Eq. (5), with β=1\beta= 1 and μ(nˉ)=nˉln⁡nˉ\mu(\bar{n})=\bar{n}\ln\bar{n}. For each vortex separation ss, we obtain the relaxed density profiles using the limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) algorithm [40]. Writing the condensate fields as Ψj(r)=\Psi_j(\mathbf{r})= Aj(r)eiwjθj(r)A_j(r)e^{iw_j\theta_j(r)}, we optimize the real amplitudes AjA_j while holding the phase profiles fixed to maintain the prescribed vortex positions and circulations. The algorithm uses gradients and information from recent iterations to approximate the local curvature of the grand potential, enabling efficient minimization without storing the full Hessian matrix. The numerical interaction potential is then calculated as

Real-time density snapshots of vortex motion

Figure 4. Real-time density snapshots of (w1,0)(w_1,0) and (0,w2)(0,w_2) vortices having same circulation (wj=1w_j=1) obtained by the extended GP eq. (3) with β=1\beta=1. (a–c) show clockwise motion for nˉ=0.612\bar{n}=0.612 at τ=0\tau=0, 800, and 1600; panels (d–f) show counter-clockwise motion for nˉ=3.5\bar{n}=3.5 at τ=0\tau=0, 1600, and 3200. Circle and square represent component 1 and 2, respectively, and the insets enlarge the central region. The initial spacing is set to be s=15s=15. See the Supplemental Material [39] for animations of the dynamics in (a)-(c) and (d)-(f). Other conditions are the same as those in Figure 1.

Unum(s)=η12(s)−η1(r1)−η2(r2)+η0,(29)U_{\mathrm{num}}(s)=\eta_{12}(s)-\eta_1(r_1)-\eta_2(r_2)+\eta_0, \tag*{(29)}

where η12(s)\eta_{12}(s), ηj(rj)\eta_j(r_j), and η0\eta_0 denote the minimized grand potentials of the vortex pair, an isolated vortex in component jj at rjr_j, and the vortex-free background, respectively. All configurations are relaxed independently on the same computational grid at the same chemical potential and with identical boundary conditions. This subtraction removes the individual vortex contributions and restores the uniform background reference, isolating the interaction between the two vortices. Figure 3 compares the numerical intervortex interaction potential with the analytical asymptotic prediction in Eq. (26) for nˉ=0.612\bar{n}=0.612 and nˉ=3.5\bar{n}=3.5.

Figure 3. Comparison between the asymptotic interaction potential U(sd/2)U(s_d/2) in Eq. (26) and numerical results obtained by pinned-core GP minimization for β=1\beta= 1 and (a) nˉ=0.612\bar{n} = 0.612 with (b) nˉ=3.5\bar{n} = 3.5. The solid lines represent the analytical result with ξ\xi fitted to the numerical results in the asymptotic regime.

The analytical expression captures the numerical behavior in the asymptotic regime, 5<s/2≲105<s/2\lesssim10. The only fitting parameter is the short-range cutoff distance ξ\xi, which is determined by fitting Eq. (26) to the numerical interaction potential in the asymptotic regime. At nˉ=0.612\bar{n}=0.612 [Figure 3(a)], the numerical curve exhibits a sudden deviation from the asymptotic prediction as the separation decreases below s/2≃5s/2\simeq5. This behavior is consistent with substantial overlap and deformation of the vortex-induced density depletions, for which the well-separated approximation breaks down. In contrast, at nˉ=3.5\bar{n}=3.5 [Figure 3(b)], the numerical potential instead remains smooth because the non-winding component fills the density depletion at the vortex core of the winding component, reducing the total-density depletion and allowing the two vortex cores to remain distinct even at smaller separations.

Reversal of vortex motion

Magnus response to the interaction

The density-dependent sign reversal of the intervortex interaction potential leads to a reversal of vortex motion. The motion of well-separated vortices in a locally uniform fluid can be described by the Magnus relation. Treating each vortex core as a massless point vortex located at rj\mathbf{r}_j, and neglecting any imposed background flow, the interaction force Fj\mathbf{F}_j is balanced by the Magnus force

Fj+2πwjnj,bgz^×drjdt=0,(30)\mathbf{F}_j + 2\pi w_j n_{j,\mathrm{bg}} \hat{\mathbf{z}} \times\frac{d\mathbf{r}_j}{dt} = 0, \tag*{(30)}

where nj,bg=nˉ/2n_{j,\mathrm{bg}} = \bar{n}/2 is the background density of the jjth component [36]. Equation (27) then gives

drjdt=z^×Fj2πwjnj,bg,(31)\frac{d\mathbf{r}_j}{dt} = \frac{\hat{\mathbf{z}} \times\mathbf{F}_j}{2\pi w_j n_{j,\mathrm{bg}}}, \tag*{(31)}

which indicates that the vortex velocity is perpendicular to the force acting on it. The direction is obtained by rotating Fj\mathbf{F}_j by ±π/2\pm\pi/2 for wj=±1w_j = \pm1. For two vortices with wj=1w_j = 1 at (0,±s/2)(0,\pm s/2), F1=Fsy^\mathbf{F}_1 = F_s\hat{\mathbf{y}} and F2=−Fsy^\mathbf{F}_2 = -F_s\hat{\mathbf{y}}, and therefore attraction drives the upper and lower cores towards +x+x and −x-x directions, respectively, resulting in clockwise motion. Repulsion reverses these velocities and gives counter-clockwise motion. A radial interaction force therefore produces transverse orbital motion; attraction does not necessarily imply that the vortex cores move towards one another in this non-dissipative point-vortex limit. Taking counter-clockwise angular velocity as positive, Eqs. (27) and (31) give

Ωj=−4nˉ[1ge(nˉ)−12β]ln⁡[s/(2ξ)]−1/2s4,(32)\Omega_j = -\frac{4}{\bar{n}}\left[\frac{1}{g_e(\bar{n})}-\frac{1}{2\beta}\right]\frac{\ln[s/(2\xi)]-1/2}{s^4}, \tag*{(32)}

which implies a reversal of the orbital motion at ge(nˉ)=2βg_e(\bar{n}) = 2\beta. Opposite circulation signs leave the static potential unchanged but reverse the Magnus response, predicting the translation of the vortex pair. The direction of the translation also changes its sign at ge(nˉ)=2βg_e(\bar{n}) = 2\beta. The numerical results presented below are restricted to the case of equal positive circulations, w1=w2=1w_1 = w_2 = 1.

Numerical dynamics and the reversal density

Figure 4 shows the real-time evolution obtained by solving the extended GP equation numerically, where (w1,0)(w_1,0) and (0,w2)(0,w_2) vortices with wj=1w_j = 1 are initially located at (0,±7.5)(0,\pm7.5). At the lower background density, nˉ=0.612\bar{n} = 0.612, the vortex cores undergo clockwise orbital motion [Figs. 4(a)–4(c)]. At the higher density, nˉ=3.5\bar{n} = 3.5, the direction of rotation reverses, and the vortices exhibit substantially slower counterclockwise motion [Figs. 4(d)–4(f)]. These observations are consistent with the Magnus response described in § 4 A, which predicts clockwise and counterclockwise motion for the attractive and repulsive intervortex interactions, respectively. Figure 5 quantifies the density dependence of the initial angular velocity, extracted from the short-time vortex trajectories. Counterclockwise rotation is taken as positive, so that Ω<0\Omega< 0 and Ω>0\Omega> 0 correspond to clockwise and counterclockwise motion, respectively. The analytical prediction captures the strong reduction in the magnitude of the clockwise angular velocity as the background density approaches the reversal point. The inset enlarges the small angular velocities near and above the crossing. On the higher-density side, the numerical angular velocity becomes positive and increases over the range shown, while remaining substantially smaller in magnitude than the clockwise angular velocity at low density. The numerical reversal occurs near nˉ≃2.928\bar{n} \simeq2.928, whereas the leading asymptotic theory predicts a vanishing interaction coefficient at nˉc=e\bar{n}_c = e. Near nˉc\bar{n}_c, the leading contribution is suppressed, making the dynamics more sensitive to subleading interaction terms, finite vortex separation, and possible confinement-induced motion. Such effects may shift the numerical reversal point away from the asymptotic prediction.

Figure 5. Initial angular velocity versus total background density nˉ\bar{n} with s≃15s \simeq15 and β=1\beta= 1. The line and dots represent the analytical cutoff fit and numerical extended GP results, respectively. Negative and positive values denote clockwise and counter-clockwise motion. The inset resolves the difference between the analytical zero at nˉ=e\bar{n} = e and the numerical crossing labeled nˉ=2.928\bar{n} = 2.928.

Conclusions

We have investigated the interaction and dynamics of vortices belonging to different components in a 2D Bose quantum liquid. Using the logarithmic extended-GP model, we derived the density-dependent asymptotic interaction between vortices. Its leading coefficient is set by the difference between the inverse total-density and relative-density stiffnesses. As the background density increases, the sign of the leading coefficient changes at a critical density nˉc=e2β−1\bar{n}_{c}=e^{2\beta-1}, leading to a reversal of the intervortex interaction from attractive to repulsive. Real-time simulations revealed a corresponding reversal of vortex motion. For two vortices with equal positive circulation, clockwise orbital motion at low density changes to counterclockwise motion at high density. This behavior was explained by the Magnus response to the density-dependent interaction force.

These results demonstrate that interactions between topological defects and their motion can be controlled by the equation of state in multicomponent quantum liquids. Future work could investigate the short-range interaction and the coupling of vortex motion to compressible and surface excitations, which become increasingly important near the interaction-reversal density.

Acknowledgment

We acknowledge support from the National Science and Technology Council (NSTC), Taiwan, under the Grants No. NSTC-115-2112-M-001-035-MY3 and No. NSTC-115-2119-M-001-009, and from Academia Sinica under Grant AS-CDA-113-M04 and for support from TG 1.2 of NCTS at NTU. H.S. was supported by JSPS KAK-ENHI Grants No. JP23K03276 and No. JP26K00638.

Data Availability

The data supporting the findings of this article are not publicly available at the time of publication due to technical limitations. The data are available from the authors upon reasonable request.

Far-field vortex profiles

Substituting (15) into (16), gives

−u′′−u′r+ur2+[nln⁡n−μ(nˉ)+βnˉ(u2−v2)]u=0,(33)-u^{\prime\prime}-\frac{u^{\prime}}{r}+\frac{u}{r^{2}}+\left[n\ln n-\mu(\bar{n})+\beta\bar{n}(u^{2}-v^{2})\right]u=0, \tag*{(33)}
−v′′−v′r+[nln⁡n−μ(nˉ)−βnˉ(u2−v2)]v=0,(34)-v^{\prime\prime}-\frac{v^{\prime}}{r}+\left[n\ln n-\mu(\bar{n})-\beta\bar{n}(u^{2}-v^{2})\right]v=0, \tag*{(34)}

where n=nˉ(u2+v2)/2n=\bar{n}(u^{2}+v^{2})/2 and μ(nˉ)=nˉln⁡nˉ\mu(\bar{n})=\bar{n}\ln\bar{n}. Setting u=1+a1u=1+a_{1}, v=1+a2v=1+a_{2}, and introducing a±=a1±a2a_{\pm}=a_{1}\pm a_{2}, the linearized equations decouple into

[−d2dr2−1rddr+2nˉge(nˉ)]a+=−1r2+O(r−4),(35)\left[-\frac{d^{2}}{dr^{2}}-\frac{1}{r}\frac{d}{dr}+2\bar{n}g_{e}(\bar{n})\right]a_{+}=-\frac{1}{r^{2}}+O(r^{-4}), \tag*{(35)}
[−d2dr2−1rddr+4βnˉ]a−=−1r2+O(r−4).(36)\left[-\frac{d^{2}}{dr^{2}}-\frac{1}{r}\frac{d}{dr}+4\beta\bar{n}\right]a_{-}=-\frac{1}{r^{2}}+O(r^{-4}). \tag*{(36)}

Since ∇r2r−2=4r−4\nabla_{r}^{2}r^{-2}=4r^{-4} away from the origin, the leading terms are a±=−ξ±2/r2+O(r−4)a_{\pm}=-\xi_{\pm}^{2}/r^{2}+O(r^{-4}). Taking their sum and difference gives eq:18. Since the centrifugal term is independent of the sign of the circulation, reversing the circulation leaves the density tails unchanged.

Leading logarithmic interaction

We define q=n1−n2q=n_{1}-n_{2}, δn=n−nˉ\delta n=n-\bar{n}, and the local grand-potential density

W(n,q)=12n2(ln⁡n−12)+βq2−μ(nˉ)n.(37)\mathcal{W}(n,q)=\frac{1}{2}n^{2}\left(\ln n-\frac{1}{2}\right)+\beta q^{2}-\mu(\bar{n})n. \tag*{(37)}

Expanding about (nˉ,0)(\bar{n},0) gives

W(n,q)−W(nˉ,0)=12ge(nˉ)(δn)2+βq2+O((δn)3).(38)\mathcal{W}(n,q)-\mathcal{W}(\bar{n},0)=\frac{1}{2}g_{e}(\bar{n})(\delta n)^{2}+\beta q^{2}+O((\delta n)^{3}). \tag*{(38)}

The linear term cancels because the subtracted chemical potential is that of the chosen background. The isolated vortex tails are

δn(1)=−nˉξ+2ρ12,q(1)=−nˉξ−2ρ12,(39)\delta n^{(1)}=-\frac{\bar{n}\xi_{+}^{2}}{\rho_{1}^{2}},\qquad q^{(1)}=-\frac{\bar{n}\xi_{-}^{2}}{\rho_{1}^{2}}, \tag*{(39)}
δn(2)=−nˉξ+2ρ22,q(2)=+nˉξ−2ρ22.(40)\delta n^{(2)}=-\frac{\bar{n}\xi_{+}^{2}}{\rho_{2}^{2}},\qquad q^{(2)}=+\frac{\bar{n}\xi_{-}^{2}}{\rho_{2}^{2}}. \tag*{(40)}

The quadratic cross term is therefore

δW=ge(nˉ)δn(1)δn(2)+2βq(1)q(2)=nˉ(ξ+2−ξ−2)2ρ12ρ22.(41)\begin{aligned} \delta\mathcal{W}=g_{e}(\bar{n})\delta n^{(1)}\delta n^{(2)}+2\beta q^{(1)}q^{(2)} \\ &=\frac{\bar{n}(\xi_{+}^{2}-\xi_{-}^{2})}{2\rho_{1}^{2}\rho_{2}^{2}}. \tag*{(41)} \end{aligned}

The corresponding phase-gradient contribution is

δK=nˉ2[u2(ρ1)[v2(ρ2)−1]ρ12+u2(ρ2)[v2(ρ1)−1]ρ22].≃−nˉ(ξ+2−ξ−2)ρ12ρ22.(42)\begin{aligned} \delta\mathcal{K}=\frac{\bar{n}}{2}\left[\frac{u^{2}(\rho_{1})[v^{2}(\rho_{2})-1]}{\rho_{1}^{2}}+\frac{u^{2}(\rho_{2})[v^{2}(\rho_{1})-1]}{\rho_{2}^{2}}\right]. \\ &\simeq-\frac{\bar{n}(\xi_{+}^{2}-\xi_{-}^{2})}{\rho_{1}^{2}\rho_{2}^{2}}. \tag*{(42)} \end{aligned}

Amplitude-gradient terms do not change the leading logarithmic coefficient, although core regions can contribute to non-logarithmic remainder.

To evaluate the coefficient, consider

I=∫d2rρ12ρ22.I=\int\frac{\mathrm{d}^{2}r}{\rho_{1}^{2}\rho_{2}^{2}}.

We exclude disks of a matching radius aa around each core and choose a fixed 0<c<1/20<c<1/2. In the annulus a≪ρ1<csa\ll\rho_{1}<cs, the other core distance is of order ss, and the logarithmically divergent contribution is

I=2πs2ln⁡(csa)+O(s−2).(43)I=\frac{2\pi}{s^{2}}\ln\left(\frac{cs}{a}\right)+O(s^{-2}). \tag*{(43)}

The second annulus contributes the same amount. After rescaling r=sx\boldsymbol{r}=s\boldsymbol{x}, the remaining region gives a finite constant times s−2s^{-2}. Thus

I=4πs2ln⁡(sa)+O(s−2).(44)I=\frac{4\pi}{s^{2}}\ln\left(\frac{s}{a}\right)+O(s^{-2}). \tag*{(44)}

Matching to the nonlinear cores determines the constant term and the effective cutoff convention. Combining this result with Eqs. (41) and (42), and using

nˉ(ξ+2−ξ−2)=12ge(nˉ)−14β,(45)\bar{n}(\xi_{+}^{2}-\xi_{-}^{2})=\frac{1}{2g_{e}(\bar{n})}-\frac{1}{4\beta}, \tag*{(45)}

gives Eq. (26). This argument fixes the logarithmic coefficient but does not promote the cutoff-dependent constant to an exact asymptotic result.

References

  1. [1]L. Pitaevskii and S. Stringari, Bose-Einstein Condensation and Superfluidity, (Oxford University Press, Oxford, 2016).
  2. [2]C. J. Pethick and H. Smith, Bose-Einstein Condensation in Dilute Gases, 2nd ed. (Cambridge University Press, Cambridge, 2008).
  3. [3]E. P. Gross, Structure of a quantized vortex in boson systems, II Nuovo Cimento (1955-1965) 20, 454 (1961).DOI
  4. [4]L. P. Pitaevskii, Vortex lines in an imperfect Bose gas, J. Exptl. Theoret. Phys. (U.S.S.R.) 40, 646 (1961) [Sov. Phys. JETP 13, 451 (1961)].
  5. [5]T. D. Lee, K. Huang, and C. N. Yang, Eigenvalues and Eigenfunctions of a Bose system of hard spheres and its low-temperature properties, Phys. Rev. 106, 1135 (1957).
  6. [6]D. S. Petrov, Quantum mechanical stabilization of a collapsing Bose-Bose mixture, Phys. Rev. Lett. 115, 155302 (2015).DOI
  7. [7]D. S. Petrov and G. E. Astrakharchik, Ultradilute low-dimensional liquids, Phys. Rev. Lett. 117, 100401 (2016).DOI
  8. [8]D. S. Petrov, Liquid beyond the Van der Waals paradigm, Nat. Phys. 14, 211 (2018).DOI
  9. [9]I. Ferrier-Barbut, Ultradilute quantum droplets, Phys. Today 72(4), 46 (2019).
  10. [10]M. Guo and T. Pfau, A new state of matter of quantum droplets, Front. Phys. 16, 32202 (2021).DOI
  11. [11]Z.-H. Luo, W. Pang, B. Liu, Y.-Y. Li, and B. A. Malomed, A new form of liquid matter: Quantum droplets, Front. Phys. 16, 32201 (2021).arxiv.org/abs/2009.01061
  12. [12]F. Böttcher, J.-N. Schmidt, J. Hertkorn, K. S. H. Ng, S. D. Graham, M. Guo, T. Langen, and T. Pfau, New states of matter with fine-tuned interactions: Quantum droplets and dipolar supersolids, Rep. Prog. Phys. 84, 012403 (2020).DOI
  13. [13]C. R. Cabrera, L. Tanzi, J. Sanz, B. Naylor, P. Thomas, P. Cheiney, and L. Tarruell, Quantum liquid droplets in a mixture of Bose-Einstein condensates, Science 359, 301 (2018).
  14. [14]G. Semeghini, G. Ferioli, L. Masi, C. Mazzinghi, L. Wolswijk, F. Minardi, M. Modugno, G. Modugno, M. Inguscio, and M. Fattori, Self-Bound quantum droplets of atomic mixtures in free space, Phys. Rev. Lett. 120, 235301 (2018).DOI
  15. [15]C. D’Errico, A. Burchianti, M. Prevedelli, L. Salasnich, F. Ancilotto, M. Modugno, F. Minardi, and C. Fort, Observation of quantum droplets in a heteronuclear bosonic mixture, Phys. Rev. Res. 1, 033155 (2019).DOI
  16. [16]I. Ferrier-Barbut, H. Kadau, M. Schmitt, M. Wenzel, and T. Pfau, Observation of quantum droplets in a strongly dipolar Bose gas, Phys. Rev. Lett. 116, 215301 (2016).
  17. [17]M. Schmitt, M. Wenzel, F. Böttcher, I. Ferrier-Barbut, and T. Pfau, Self-bound droplets of a dilute magnetic quantum liquid. Nature 539, 259–262 (2016).DOI
  18. [18]C. F. Barenghi, R. J. Donnelly, and W. Vinen, Quantized vortex dynamics and superfluid turbulence, Vol. 571, (Springer Science and Business Media, Berlin, 2001).DOI
  19. [19]A. L. Fetter, Rotating trapped Bose-Einstein condensates, Rev. Mod. Phys. 81, 647 (2009).
  20. [20]B. P. Anderson, Resource Article: Experiments with vortices in superfluid atomic gases. J. Low Temp. Phys. 161, 574–602 (2010).DOI
  21. [21]Y. Li, Z. Chen, Z. Luo, C. Huang, H. Tan, W. Pang, and B. A. Malomed, Two-dimensional vortex quantum droplets, Phys. Rev. A 98, 063602 (2018).DOI
  22. [22]M. Nilsson Tengstrand, P. Stürmer, E. Karabulut, and S.M. Reimann, Rotating binary Bose-Einstein condensates and vortex clusters in quantum droplets, Phys. Rev. Lett. 123, 160405 (2019).DOI
  23. [23]M. Caldara and F. Ancilotto, Vortices in quantum droplets of heteronuclear Bose mixtures, Phys. Rev. A 105, 063328 (2022).DOI
  24. [24]P. Exami̇lioti and G. M. Kavoulakis, Ground state and rotational properties of two-dimensional self-bound quantum droplets, J. Phys. B: At. Mol. Opt. Phys. 53, 175301 (2020).DOI
  25. [25]X. Zhang, X. Xu, Y. Zheng, Z. Chen, B. Liu, C. Huang, B. A. Malomed, and Y. Li, Semidiscrete quantum droplets and vortices, Phys. Rev. Lett. 123, 133901 (2019).DOI
  26. [26]Z. Lin, X. Xu, Z. Chen, Z. Yan, Z. Mai, and B. Liu, Two dimensional vortex quantum droplets get thick, Communications in Nonlinear Science and Numerical Simulation 93, 105536 (2021).
  27. [27]L. Dong and Y. V. Kartashov, Rotating multidimensional quantum droplets, Phys. Rev. Lett. 126, 244101 (2021).DOI
  28. [28]S. Nikolaou, G. M. Kavoulakis, and M. Ögren, Rotating quantum droplets confined in a harmonic potential, Phys. Rev. A 108, 053309 (2023).DOI
  29. [29]Q. Gu and X. Cui, Self-bound vortex lattice in a rapidly rotating quantum droplet, Phys. Rev. A 108, 063302 (2023).DOI
  30. [30]S. Saqlain, T. Mithun, R. Carretero-González, and P. G. Kevrekidis, Dragging a defect in a droplet Bose-Einstein condensate, Phys. Rev. A 107, 033310 (2023).DOI
  31. [31]S. Nikolaou, G. M. Kavoulakis, and M. Ögren, Rotating quantum droplets confined in an anharmonic potential, Phys. Rev. A 109, 043304 (2024).DOI
  32. [32]D. Liu, Y. Gao, D. Fan, B. A. Malomed, and L. Zhang, Rotating dipole and quadrupole quantum droplets in binary Bose-Einstein condensates, Phys. Rev. Research 6, 033186 (2024).
  33. [33]G. Li, Z. Zhao, R. Zhang, Z. Chen, B. Liu, B. A. Malomed, and Y. Li, Elongated vortex quantum droplets in binary Bose-Einstein condensates, Phys. Rev. A 112, 013318 (2025).arxiv.org/abs/2502.19808
  34. [34]S. Wang, L. Duan, L. Dong, and Z.-Y. Yang, Dynamics of vortices in uniform Bose quantum liquids, Phys. Rev. A 112, 063304 (2025).
  35. [35]M. Eto, K. Kasamatsu, M. Nitta, H. Takeuchi, and M. Tsubota, Interaction of half-quantized vortices in two-component Bose-Einstein condensates, Phys. Rev. A 83, 063603 (2011).DOI
  36. [36]K. Kasamatsu, M. Eto, and M. Nitta, Short-range intervortex interaction and interacting dynamics of half-quantized vortices in two-component Bose-Einstein condensates, Phys. Rev. A 93, 013615 (2016).DOI
  37. [37]J. Han, K. Kasamatsu, and M. Tsubota, Dynamics of two quantized vortices belonging to different components of binary Bose-Einstein condensates in a circular box potential, J. Phys. Soc. Jpn. 91, 024401 (2022).DOI
  38. [38]W. H. Press, S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery, Numerical Recipes, 3rd ed. (Cambridge University Press, Cambridge, UK, 2007).
  39. [39]See Supplemental Material at [URL will be inserted by publisher] for the dynamics of well-separated vortices at different background densities.
  40. [40]D. C. Liu and J. Nocedal, On the limited memory bfgs method for large scale optimization, Math. Program. 45, 503–528 (1989).

Paper details

vlnc:W_zzkdu856
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents