2024 Volume 33 Issue 10
Article Contents

Huan-Bo Luo, Xin-Feng Zhang, Runhua Li, Yongyao Li, Bin Liu. Bessel vortices in spin-1 Bose–Einstein condensates with Zeeman splitting and spin–orbit coupling[J]. Chinese Physics B, 2024, 33(10): 100304. doi: 10.1088/1674-1056/ad6424
Citation: Huan-Bo Luo, Xin-Feng Zhang, Runhua Li, Yongyao Li, Bin Liu. Bessel vortices in spin-1 Bose–Einstein condensates with Zeeman splitting and spin–orbit coupling[J]. Chinese Physics B, 2024, 33(10): 100304. doi: 10.1088/1674-1056/ad6424

Bessel vortices in spin-1 Bose–Einstein condensates with Zeeman splitting and spin–orbit coupling

  • Received Date: 14/06/2024
    Accepted Date: 16/07/2024
    Available Online: 01/09/2024
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(8)

Article Metrics

Article views(485) PDF downloads(1) Cited by(0)

Access History

Bessel vortices in spin-1 Bose–Einstein condensates with Zeeman splitting and spin–orbit coupling

Abstract: We investigate the ground states of spin–orbit coupled spin-1 Bose–Einstein condensates in the presence of Zeeman splitting. By introducing the generalized momentum operator, the linear version of the system is solved exactly, yielding a set of Bessel vortices. Additionally, based on linear solution and using variational approximation, the solutions for the full nonlinear system and their ground state phase diagrams are derived, including the vortex states with quantum numbers m = 0, 1, as well as mixed states. In this work, mixed states in spin-1 spin–orbit coupling (SOC) BEC are interpreted for the first time as weighted superpositions of three vortex states. Furthermore, the results also indicate that under strong Zeeman splitting, the system cannot form localized states. The variational solutions align well with numerical simulations, showing stable evolution and meeting the criteria for long-term observation in experiments.

1.   Introduction
  • Bose–Einstein condensates (BECs) of atoms serve as versatile platforms for simulating a wide range of phenomena drawn from condensed-matter physics.[1–25] Among these phenomena, the spin–orbit coupling (SOC) plays a basic role in spin Hall effects,[26] topological insulators,[27] spintronic devices,[28] and more.[29–36] The experimental achievement of two-dimensional (2D) SOC[37] in spin-1/2 BECs is indeed a milestone, similar to its counterpart in one-dimensional (1D) case.[38] Recently, three-dimensional (3D) SOC in quantum gas has also been achieved experimentally.[39] Additionally, many schemes have been proposed.[40,41] The Gross–Pitaevskii equations (GPEs) are widely applied in studying the stationary state and dynamic behaviors of SOC BECs.[42] The diverse ground-state phase diagram of SOC spin-1/2 BECs, apart from including stripe, plane-wave, and zero-momentum phases,[43,44] also encompasses vortex states,[45–52] etc. In addition to ground states, solitons arising from the intrinsic nonlinear effects of SOC BECs have also been extensively studied and reported.[53–74]

    In addition to spin-1/2, understanding the effect of spin-1 SOC in 2D BEC is also an active topic.[75] The spin-1 2D BEC system with SOC can support not only plane-wave and standing-wave phases,[76] but also vortex states,[77,78] and these vortex states have recently been proven to be Bessel vortices.[79] Furthermore, besides the equilibrium mode, the collective excitation modes of this system have also been reported recently.[80] In a spin-1/2 SOC BEC, in addition to the vortex state, another typical ground state is the mixed state. However, the mixed state has not been identified and explained well in spin-1 systems. Therefore, obtaining the complete ground state phase diagram for 2D spin-1 SOC BEC is essential for a comprehensive understanding of the entire system.

    Recently, the gap solitons and the dynamics of bright solitons in SOC systems with Zeeman splitting have been studied.[81,82] However, the transition between the higher-order vortex state and the ground state is rarely studied. In our settings, under the action of Zeeman splitting added to the SOC system, higher-order vortex states can become the ground state. In this article, we will investigate the ground state of spin-1 SOC BEC with Zeeman splitting. In the linear case, by introducing the generalized momentum operator, the system can be exactly solved even in the presence of Zeeman splitting, with the solution being a set of Bessel vortices. In the nonlinear case, we construct a variational ansatz based on truncated linear solutions. The variational parameters are determined using the principle of minimum energy. Finally, by analyzing each variational parameter, we can obtain a complete ground state phase diagram. Under strong attractive interactions, the ground state of the system will transition from a vortex state to a mixed state. All variational results are in good agreement with the numerical results, and the stability of these ground states has been checked.

    The rest of this paper is structured as follows. In Section 2, the theoretical model is introduced. In Section 3, the linear solution is constructed by means of Bessel function. Based on the linear vortex solutions, the nonlinear solution is constructed by means of the variational approximation in Section 4. The vortex state and mixed state obtained through variational methods are discussed in Sections 5 and 6, respectively. Finally, the main results of the paper are summarized in Section 7.

2.   Model and its reductions
  • We consider a quasi-(2D) spin-1 7Li BEC with Rashba-type spin–orbit coupling in the Zeeman field. The spinor wave function, Ψ = (Ψ1,Ψ0,Ψ−1)T, of this system can be governed by Gross–Pitaevskii (GP) equations in the dimensionless form

    where the spin–orbit (SO) coupling is chosen as VSO = iFy∂x–iFx∂y and F = (Fx,Fy,Fz) is a vector of spin-1 matrices with

    The characteristic length, energy, and time are defined by l = 2 μm, ε = ℏ2/(ml2) = 2.4 × 10−31 J and τ = ℏ/ε = 0.44 ms, where m = 1.16 × 10−26 kg is the mass of 7Li. The coefficients c0 > 0 and c2 > 0 are the strengths of the particle-attraction and spin-attraction contact interactions, respectively. kz represents the strength of the linear Zeeman effect. n = Ψ†Ψ is the particle density and S = Ψ†FΨ is the local spin vector. Here, c0 = 1 is fixed by means of rescaling. Thus, equation (1) includes only two free parameters kz and c2. In the absence of the Zeeman field in z direction, i.e., kz = 0, equation (1) reduces to the case of Ref. [79], in which the vortex ground-state phase transition between the stationary vortex solutions, stability, and the unit Bloch vector textures were discussed. In this study, we will provide a detailed description of the vortex states and mixed states for the SOC spin-1 BECs in the presence of the Zeeman splitting.

    Stationary solutions of Eq. (1) with chemical potential μ have the form

    where the stationary functions ψ(x, y) = (ψ1,ψ0,ψ−1)T. Furthermore, equation (1) can support vortex solutions given by

    where (r,θ) are the polar coordinates and R0,±1(r) are three radial wave functions. m serves as a quantum number to index the vortex state. In the following, we will seek vortex solutions with the from of Eq. (4) form in both linear and nonlinear cases.

3.   Exact vortex states of the linear version
  • The linear version of Eq. (1) can be considered as a special case where c0 = 0 and c2 = 0. By substituting Eq. (3) into the linear version of Eq. (1), we can obtain the stationary Schrödinger equation, i.e., H^ψ=μψ, where the Hamiltonian H^ has the form

    Solving this equation directly can be challenging, so we introduce a generalized momentum operator P^, which is defined by

    It is easy to check the Hamiltonian H^ and the generalized momentum operator P^ commute, i.e., [H^,P^]=0, which indicates that H^ and P^ share the same eigenstates. By solving the eigenvalue equation P^ψ=kψ with real k, we can obtain a set of eigenstates in the form (4) with

    where Jm denotes the Bessel function of the first kind with integer number m. kr is the radial momentum and kz can be considered as momentum in z component. Thus the total momentum should satisfy k=kr2+kz2. The solution (4) with Eq. (7) presents a set of the Bessel vortices with winding numbers −(m+1), −m, and −(m–1) in the ψ1, ψ0, and ψ−1 components, respectively. The similar exact solutions of the SO coupled binary linearlized GP equations were recently reported in Refs. [59,79,83].

    The state given by Eq. (4) with Eq. (7) is also the eigenstate of H^, with the respective chemical potential

    It includes the kinetic energy, SO-coupling energy and the energy induced by the Zeeman splitting. The chemical potential is a quadratic function of k, and always possesses a minimum value. Depending on the range of the total momentum k determined by kz (i.e., k=kr2+kz2≥kz), we analyze the minimum values of the chemical potential in different cases. For 0 ≤ kz < 1, the chemical potential reaches its minimum value μmin=−(1+kz2)/2 at k = 1, where the radial momentum is given by kr=k2−kz2=1−kz2. In the case of kz ≥ 1, the chemical potential achieves its minimum value μmin = −kz at k = kz with kr = 0. The summary of the distribution of the minimum values of the chemical potential is as follows:

    The analysis above suggests that kz = 1 is a critical point for a phase transition, wherein the ground state of the system transitions from vortex state to a uniform state. The uniform state can only exist in the case m = 1 with ψ1 = ψ0 = 0 and ψ−1 = 1. Note that all the vortex states are degenerate with respect to the excitation number m, but μ, given by Eq. (8), does not depend on m.

    The linear Bessel vortices indicate that particles propagate radially with momentum kr. Naturally, the norm integral (number of particles) for this linear state (7) in the free space diverges, which means that such states require an infinite number of particles and therefore cannot be realized in real world. As shown below, taking into account the attractive nonlinearity in Eq. (1) makes it possible to form localized state with a finite number of particles.

4.   Variational method for ground state in nonlinear system
  • In the nonlinear case, we will employ both variational approximation (VA) and imaginary-time evolution method to solve the system simultaneously.

    The attractive nonlinearity chops off the slowly decaying tails of the Bessel wave function, which make its integral norm diverging. To take this effect into account in the framework of the VA, we adopt an ansatz based on wave functions (4) and (7), multiplied by truncation factor Asech(ar), with amplitude A and inverse width a

    where bm represents the weight of the eigenstate in the superposition state with quantum number m, and it satisfies the condition that ∑m=−MMbm2=1. Here M = 1 is an empirical choice, thus there are only three weights, i.e., b−1, b0, and b1. Similar truncation function Asech(ar) has already been applied in SO coupled systems in the absence of Zeeman splitting. The total momentum still satisfies k=kr2+kz2. The normalization coefficient A can be determined, as a function of a, b−1, b0, b1, and kr, by imposing the normalization condition, 〈ψ|ψ〉 = 1.

    In order to determine the parameters a, b−1, b0, b1, and kr, we need to minimize the total energy

    The specific expression for the total energy E can be obtained by substituting Eq. (10) into Eq. (11). The minimum energy value E and the corresponding variational parameters (a,b−1,b0,b1,kr) can be numerically obtained by means of the simplex search method.[84]

    Based on the distribution characteristics of variational parameters in the (kz, c2) plane, the phase diagram of the system’s ground state is summarized in Fig. 1. We can find that the localized ground states of the system mainly include vortex states with quantum number m = ±1, and mixed states. Specifically, when both kz and c2 are small, the ground state of the system is a vortex state with m = 0, as indicated by the blue region in Fig. 1. However, when kz remains small and c2 increases, the ground state of the system becomes a mixed state, as illustrated by the green region in Fig. 1. When the Zeeman splitting strength is within a certain range, approximately 0.4 < kz < 1.0, the ground state of the system is a vortex state with m = 1, corresponding to the yellow region in Fig. 1. Furthermore, for strong Zeeman splitting kz, the system will be unable to form localized ground states. In conclusion, we considered the limit of kz → 0, where the symmetric SOC term dominates, indicating a ground state that respects both spin symmetry and spatial rotational symmetry (vortex state with m = 0). In the limit of very large kz, i.e., kz ≥ 1, only the Zeeman term takes effect, indicating that the ground state should be populated in ψ−1 and uniformly distributed in real space. In the middle range, the ground state is determined by the interplay of these two terms, resulting in a vortex state with m = 1. In the following, we will discuss in detail the characteristics of vortex states and mixed states by fixing c2 = 0 and c2 = 0.3, respectively.

5.   Vortex states as ground states in the absence of spin interactions
  • Firstly, we consider the ground state of the system in the absence of spin–spin interactions. Figures 2(a)–2(d) illustrate the dependence of variational parameters on kz at c2 = 0. In Fig. 2(a), it is evident that, for kz < 0.4, the weights b0 = 1, and b1 = b−1 = 0. This signifies that the system’s ground state corresponds to a vortex state with m = 0. Conversely, for kz > 0.4, the weights become b1 = 1, and b0 = b−1 = 0. In this region, the ground state of the system is also a vortex state, but with a quantum number of m = 1. Therefore, kz = 0.4 is the ground state phase transition point. From Figs. 2(b)–2(d), it can be observed that for kz > 1.05, a, A, and kr all become 0. The result implies that the uniform state cannot form localized states in free space.

    To validate the accuracy of the variational results, we numerically solved Eq. (1) using the imaginary-time evolution method. By setting c2 = 0, we obtained the system’s ground states for Zeeman field strengths at kz = 0 and 0.7, respectively. The wavefunctions for a three-component system can be expressed as

    where |ψj| and Θj are amplitude and phase of the wavefunction. The distributions of the amplitude and phase for kz = 0 and 0.7 are shown in Figs. 3(a1)–3(a3) and 3(c1)–3(c3). However, the system cannot support a localized ground state under strong Zeeman splitting (kz > 1.05), and such phenomenon also observed in spin-1/2 systems.[48]

    To compare the variational method with numerical method, the distributions of the amplitude and phase for kz = 0 and 0.7 obtained through the variational method are displayed in Figs. 3(b1)–3(b3) and 3(d1)–3(d3). Here the variational parameters in Eq. (10) calculated by variational method are a = 0.1621, b0 = 1, b1 = b−1 = 0, kr = 1.0023, A = 0.1426 and a = 0.2054, b1 = 1, b0=b−1 = 0, kr = 0.7129, A = 0.1358 for kz = 0 and kz = 0.7, respectively.

    Overall, the results obtained from the variational method are essentially consistent with those from numerical method, including both the amplitude distributions and phase distributions. Therefore, below we will only use the variational results to analyze the characteristics of the solutions. From Figs. 3(b1)–3(b3), it can be observed that for kz = 0, the distributions of Θ1, Θ0, and Θ−1 have winding numbers of 1, 0, and −1, respectively. This implies that the ground state of the system corresponds to a vortex state with m = 0. Furthermore, the distributions of the amplitudes |ψ1|, |ψ0|, and |ψ−1| exhibit concentric ring patterns for each. Similarly, from Figs. 3(d1)–3(d3), it can be seen that for kz = 0.7, the distributions of Θ1, Θ0, and Θ−1 have winding numbers of 2, 1, and 0, respectively. This indicates that the ground state of the system corresponds to a vortex state with m = 1. the distributions of the amplitudes |ψ1|, |ψ0|, and |ψ−1| also exhibit concentric ring patterns for each. Additionally, we find that under the action of a stronger Zeeman field (kz = 0.7), the atoms are predominantly populated in the ψ−1 component, followed by the ψ0 component, with the least population in the ψ−1 component.

    Figure 4 compares the profiles of the wave function at y = 0 obtained through numerical method and the variational method for two different sets of parameters kz = 0 and kz = 0.7. Note that, as per Eq. (10), the imaginary component of the wave function is absent, i.e., Im(ψ0,±1) = 0. One can see that for kz = 0, numerical wave functions are nearly identical to the variational counterparts, while for kz = 0.7, there is a slight mismatch at peaks and valleys as shown in Fig. 4(f). Since the variational method is semi-analytical, mismatch always exists. Even the mismatch in Fig. 4(f) is within the tolerance range. Providing more complex ansatz may reduce mismatch. These results indicate that the variational method based on the ansatz (10) exhibits high accuracy and reliability, with the accuracy increasing as kz decreases.

6.   Emergence of mixed states in the presence of spin interactions
  • In the previous section, we mainly discussed the ground state of the system in the absence of spin and spin interactions, i.e., when c2 = 0, and demonstrated that the ground state corresponds to vortex states with m = 0 or 1. In the case of c2 ≠ 0, mixed states rather than vortex states can also serve as the ground state of the system. Here, we choose c2 = 0.3 as a typical example to analyze the ground state of the system under different values of kz.

    The variational parameters in the ansatz (10) for c2 = 0.3 are presented in Fig. 5. We can find that the weights b0, b1, and b−1 are all non-zero for kz < 0.42, indicating that under this condition, the ground state of the system is a mixed state. Specifically, at kz = 0, b12=b−12≈0.2, and b02≈0.6. This indicates that in the mixed state, the vortex states with m = 1 and m = −1 have equal weights, while the vortex state with m = 0 predominates. As kz increases, b1 and b0 gradually decrease, while b1 increases. Until kz = 0.42, b−1 and b0 approach 0, and b1 becomes 1. This implies that kz = 0.42 is a ground state phase transition point, where the system’s ground state transitions from a mixed state to a vortex state with m = 1.

    As typical examples, the distributions of the amplitude and phase obtained through variational method for kz = 0 and kz = 0.3 are respectively shown in Figs. 6(b1)–6(b3) and 3(d1)–3(d3). Here the variational parameters in Eq. (10) calculated by variational method are a = 0.1694, b0 = 0.7815, b1 = b−1 = 0.4412, kr = 1.0034, A = 0.1464 for kz = 0, and a = 0.1717, b1 = 0.6111, b0 = 0.7315, b−1 = 0.3024, kr = 0.7129, A = 0.1440 for kz = 0.3. On the other hand, we also employed the imaginary time evolution method to determine the ground state of the system under the parameters c2 = 0.3, kz = 0, and c2 = 0.3, kz = 0.3. The corresponding distributions of the amplitude and phase are shown in Figs. 6(a1)–6(a3) and 6(c1)–6(c3). By comparing distributions of the amplitude and phase obtained through variational method and numerical method, we can see that these two patterns are very similar. By observing the distributions of amplitude, we can find that the rotational symmetry of a mixed state is broken compared to a vortex state. Additionally, we can also find that the phase of a vortex state changes gradually along the azimuthal direction, while in a mixed state, it changes gradually along the radial direction. In addition, we can draw a conclusion that in the absence of Zeeman splitting (kz = 0), the component density distributions of ψ1 and ψ−1 are mirror symmetric about x = 0, i.e., ψ1(x,y) = ψ−1(−x,y). This symmetry holds for both vortex states and mixed states, as shown in Figs. 3(a1)–3(a3) and 6(a1)–6(a3). However, in the presence of Zeeman splitting, this symmetry is broken.

    Figure 7 compares the profiles of the wave function at y = 0 obtained through numerical method and the variational method for two different sets of parameters kz = 0 and kz = 0.3. Note that, as per Eq. (10), the imaginary component of the wave function is absent, i.e., Im(ψ0,±1) = 0. One can see that for both kz = 0 and kz = 0.3, numerical wave functions are nearly identical to the variational counterparts. This result implies that variational method also exhibit a considerable level of accuracy for solving mixed states. Furthermore, this also indicates that the ansatz we provided in Eq. (10) is reasonable, and mixed states are superpositions formed by three vortex states. It is relevant to mention that mixed states in spin-1 SOC BEC have already been reported in Ref. [77], but the composition of these mixed states has not been explained yet.

    To confirm that the ground states identified by the above analyses are indeed stable modes, their stability was verified through real-time simulations of their perturbed evolution. White noise, with maximum intensity reaching 10% of the wave function’s amplitude, was added to the ground-state wave functions. Subsequently, real-time evolution was applied to these perturbed wave functions. The results of numerical simulations, presented in Figs. 8, showcase a vortex state with parameters kz = 0.0 and c2 = 0.0, and a mixed state with parameters kz = 0.0 and c2 = 0.3. At t = 0, the wave functions exhibit numerous noise points. However, as time progresses, by t = 200, the noise points on the wave functions disappear. In the subsequent evolution, the phases of the wave functions rotate around the vortex centers, while the amplitude distribution remains unchanged. These results fully verify the stability of the ground states in all cases.

7.   Conclusions
  • In summary, we investigated the ground state of vortex formation in spin-1 BECs under the combined effects of spin–orbit coupling and Zeeman splitting. The linear version of the system can be solved exactly by introducing the generalized momentum operator, resulting in a set of Bessel vortices. Additionally, based on these Bessel vortices and using variational approximation, we have derived solutions for the full nonlinear system. Through an analysis of the variational parameters, we have summarized the phase diagrams of the system under different c2 and kz parameters. These phases include vortex states with quantum numbers m equaling 0 and 1, as well as mixed states. It is worth noting that under strong Zeeman splitting, specifically when kz approaches or exceeds 1, localized states cannot exist. Our results demonstrate that the variational solutions are in good agreement with numerical simulations and exhibit stable evolution, meeting the requirements for long-time observation in experiments.

Figure (8)  Reference (84)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return