betaFeedback

WorksPhysics and Astronomy

Optical vortex bullets with large angular momenta

Abstract

We investigate three-dimensional spatiotemporal optical vortex solitons (light bullets with a toroidal shape) carrying higher-order orbital angular momentum in Kerr media with anomalous group-velocity dispersion and a transversely localized Gaussian profile of the linear refractive index. Stationary vortex bullets with topological charges (winding numbers) 1≤ℓ≤41\leq \ell \leq 4 are obtained as numerical solutions of the three-dimensional nonlinear Schrödinger equation, and their existence, structure and stability are systematically analyzed. In terms of atomic Bose-Einstein condensates, these solutions represent matter-wave solitons with embedded vorticity. We examine the dependence of the solution families on the propagation constant, the depth and width of the guiding profile, and ℓ\ell . It is found that the increase of ℓ\ell requires one to use a stronger or wider waveguiding potential to sustain localized states satisfying the Vakhitov-Kolokolov (VK) stability criterion. VK-complying and non-complying branches of vortex-soliton families are separated by a degenerate Townes branch. Direct simulations reveal that, although the guiding profile substantially suppresses the collapse and extends the propagation distance of vortex bullets, their robustness is ultimately limited by azimuthal instabilities, whose growth is enhanced for values of ℓ\ell . These results demonstrate that engineered refractive-index landscapes significantly enlarge the stability domain of spatiotemporal vortex bullets and provide guidelines for supporting optical bullets carrying large orbital angular momenta.

Introduction

Spatiotemporal optical solitons [1], commonly referred to as light bullets, are multidimensional nonlinear wave packets that keep their spatial and temporal self-trapping through the balance between the diffraction, group-velocity dispersion (GVD), and nonlinear self-focusing. Since the theoretical prediction by Silberberg in 1990 [2], light bullets have been attracting much interest due to their fundamental properties, as well as potential applications to optical telecommunications, laser physics, propagation of ultra-strong light pulses, and other fields [3]. Nowadays it is a field of intense research concerning the production of such bullets [4] as well as their existence and stability in a number of systems, as crystals with impurities [5], optical fibers [6], lattices and fiber arrays [7, 8] or other types of designed media [9, 10]. The creation and evolution of the spatiotemporal solitons are governed by the multidimensional nonlinear Schrödinger equation (NLSE) [11], which provide the basis of the fundamental model for the propagation of slowly varying envelopes of light pulses and beams.

It is well known that the optical bullets are unstable in self-focusing materials with anomalous GVD [12, 13]: above a certain threshold energy the (2+1)-dimensional [(2+1)D] bullet undergoes catastrophic collapse (blow-up) after a finite propagation distance, while (3+1)D bullets are still more vulnerable to the collapse instability, as the respective threshold energy is zero [11, 14]. To prevent the blow-up and achieve stable propagation of nonlinear optical wave packets in two or three spatial dimensions, various mechanisms have been explored [15], including the incorporation of higher-order terms in the nonlinear refractive index, such as saturable [16, 17] or cubic–quintic nonlinearity [18, 19], as well as nonlocality [20, 21, 22, 23] and external potentials[24], among other effects. In particular, optical bullets in materials with waveguiding linear-index profiles can be stabilized through the trapping provided by the effective potential representing the guiding profile [25, 26, 27], including ingenious systems like moiré lattices [28] or twisted waveguide arrays [29].

Endowing these nonlinear wave packets with a topological charge (alias winding number), which represents a phase singularity in the wave field [30, 31], leads to the formation of vortex light bullets [32, 33] carrying the orbital angular momentum [34, 35], in addition to being spatiotemporally self-trapped[36]. However, the ring-shaped intensity distribution in the vortex bullets is highly susceptible to the azimuthal modulational instability, causing the vortex to break up into several fundamental (zero-vorticity) solitons. Despite their theoretical appeal and promising properties, the experimental realization of stable optical bullets—and especially of vortex bullets—is highly challenging [37]. Precise control over the input beam and the balance of competing nonlinearities are critical for achieving the sustained propagation [38, 39, 40]. In the case of discrete vortex bullets, the possibility of the existence of transient stable states in waveguiding arrays was experimentally demonstrated [41]. The possibility of producing stable bullets in 2D Su-Schrieffer-Heeger lattices with disclination defects was also demonstrated recently [42]. When an external index-guiding profile is used, the possibility of producing stable fundamental (zero-vorticity) optical bullets is well known [25], although it was only recently theoretically demonstrated that such a profile can support stable optical vortices with winding number 1, that propagate stably over distances much larger than the underlying diffraction length and survive mutual collisions [43]. Conditions under which the stable propagation is possible require that the index-guiding profile must be sufficiently strong, provided that the propagation constant belongs to a specific range of values.

In the present paper we aim to produce stable vortex bullets with relatively large values of the winding number. It is well known that increasing angular momentum normally enhances the azimuthal instability, leading to fragmentation of stationary vortex states into a set of fundamental solitons, thus making the stabilization of multidimensional vortex solitons one of long-standing challenges in nonlinear optics. In this study, we address a system similar to that considered in Ref. [43], which includes the focusing Kerr nonlinearity and anomalous GVD in the presence of a Gaussian-shaped linear-index guiding pattern in the transversal plane.

The adopted model is introduced in the next section. The results for the bullet families and their stability are reported in section III. The paper is concluded by section 4.

The Model

We start from the (3+1)D NLSE with the Kerr nonlinear term,

i∂Ψ~∂z~+12k0∇~⊥2Ψ~−β22∂2Ψ~∂T~2+k0[ΔnV(r~)+n2∣Ψ~∣2]Ψ~=0.(1)i\frac{\partial\widetilde{\Psi}}{\partial\widetilde{z}}+\frac{1}{2k_0}\widetilde{\nabla}_{\perp}^{2}\widetilde{\Psi}-\frac{\beta_2}{2}\frac{\partial^2\widetilde{\Psi}}{\partial\widetilde{T}^2}+k_0\left[\Delta nV(\widetilde{r})+n_2|\widetilde{\Psi}|^2\right]\widetilde{\Psi}=0. \tag*{(1)}

where z~\widetilde{z} is the propagation distance, and Laplacian ∇~⊥2\widetilde{\nabla}_{\perp}^{2} is the diffraction operator acting in the transversal spatial plane, while k0=ω/ck_0=\omega/c, where ω\omega is the carrier frequency and cc the speed of light in vacuum. Further, T~=t−z/vg\widetilde{T}=t-z/v_g, which includes the conventional time tt, is the temporal coordinate in the reference frame traveling with the group velocity vgv_g, β2<0\beta_2<0 is the coefficient of the anomalous GVD, and n2>0n_2>0 is the coefficient of the focusing Kerr nonlinearity. The axially symmetric index profile is represented by function V(r~)V(\widetilde{r}) with amplitude Δn\Delta n. By means of rescaling,

r=k02Δnr~,z=k0Δnz~,r=k_0\sqrt{2\Delta n}\widetilde{r},\qquad z=k_0\Delta n\widetilde{z},
T=(−2k0Δnβ2)1/2T~,Ψ=(n2Δn)1/2Ψ~,(2)T=\left(-\frac{2k_0\Delta n}{\beta_2}\right)^{1/2}\widetilde{T},\qquad\Psi=\left(\frac{n_2}{\Delta n}\right)^{1/2}\widetilde{\Psi}, \tag*{(2)}

Eq. (1) is cast in the scaled form:

i∂Ψ∂z+∇⊥2Ψ+∂2Ψ∂T2+[V(r)+∣Ψ∣2]Ψ=0.(3)i\frac{\partial\Psi}{\partial z}+\nabla_{\perp}^{2}\Psi+\frac{\partial^2\Psi}{\partial T^2}+\left[V(r)+|\Psi|^2\right]\Psi=0. \tag*{(3)}

The shape of the waveguiding index is fixed as the Gaussian,

V(r)=γexp⁡(−qr2),(4)V(r)=\gamma\exp\left(-qr^2\right), \tag*{(4)}

with amplitude γ\gamma and width w=1/qw=1/\sqrt{q}.

It is relevant to mention that Eq. (3) with zz replaced by tt, and TT replaced by zz, is identical to the Gross-Pitaevskii equation for the self-attractive atomic Bose-Einstein condensate. In that case, potential given by Eq. (4), with γ>0\gamma>0 and qq large enough, represent the “cigar-shaped” trapping potential for atoms [44]. In that case, the vortex optical bullets are tantamount to 3D matter-wave solitons with embedded vorticity.

Stationary vortex states, with integer winding number (topological charge) ℓ≥0\ell\geq0 and propagation constant β\beta, are sought for, in terms of the cylindrical coordinates (r,z,ϕ)(r,z,\phi), as

Ψ(r,z,ϕ;T)=ψ(r,T)eiℓϕei(βz−ωt),(5)\Psi(r,z,\phi;T)=\psi(r,T)e^{i\ell\phi}e^{i(\beta z-\omega t)}, \tag*{(5)}

where ψ(r,T)\psi(r,T) is a real function, determined by the following equation in the (r,T)(r,T) plane:

−βψ+∂2ψ∂T2+∂2ψ∂r2+1r∂ψ∂r−ℓ2r2ψ+[V(r)+ψ2]ψ=0.(6)-\beta\psi+\frac{\partial^2\psi}{\partial T^2}+\frac{\partial^2\psi}{\partial r^2}+\frac{1}{r}\frac{\partial\psi}{\partial r}-\frac{\ell^2}{r^2}\psi+\left[V(r)+\psi^2\right]\psi=0. \tag*{(6)}

Solutions of Eq. 6 are characterized by their propagation constant β\beta and the value of the optical energy (alias norm), NN,

N=2π∫−∞∞∫0∞∣ψ∣2r dr dT,(7)N=2\pi\int_{-\infty}^{\infty}\int_{0}^{\infty}|\psi|^2r\,dr\,dT, \tag*{(7)}

which is a dynamical invariant of Eq. (3). Below, we find solutions of Eq. (6) and identify their stability for increasing values of vorticity ℓ\ell, in the parameter plane of (γ,q)(\gamma,q).

Solution Families and Stability

For numerical solution of Eq.(6), variables (r,T)(r,T) and real function ψ\psi were discretized using the finite-differences scheme [45], and the resulting algebraic problem was solved by iterations, by dint of the globally convergent Newton method. Localized solutions are subject to conditions ψ→0\psi\to0 as r→∞r\to\infty and T→±∞T\to\pm\infty. At r→0r\to0, the asymptotic form is obvious, ψ→rℓ\psi\to r^\ell.

Examples of the solutions for vorticity values from ℓ=1\ell=1 to 4 are presented in Figure 1. In all cases parameter values are fixed to γ=4\gamma=4, q=1/2q=1/2 and β=1.0\beta=1.0. Field profiles in the radial and temporal directions are plotted in panels

Figure 1. Examples of optical bullets with embedded vorticity ranging from ℓ=1\ell= 1 to 4, for γ=4\gamma= 4, q=1/2q = 1/2 and β=1.0\beta= 1.0. (a) Radial spatial profiles in cross-section T=0T = 0 (to which the maximum amplitude belongs). (b) Temporal profiles in cross-section r=rmax⁡r = r_{\max}, where the maximum amplitude is attained. (c-f) 3D toroidal shapes of vortex bullets with winding numbers ℓ=1\ell= 1, 2, 3 and 4, respectively, represented by means of isosurfaces ψ=0.25\psi= 0.25.

(a) and (b), while 3D toroidal shapes of the vortex bullets are displayed by means of isosurfaces of ψ\psi in panels (c-f). A natural feature is the steep increase of the spatial width of the central hole as ℓ\ell increases. Also, the amplitude increases with ℓ\ell, attaining a maximum at ℓ=3\ell= 3, slightly decreasing and further saturating at larger values of ℓ\ell. The temporal width in Figure 1(b) initially decreases with the increase of ℓ\ell up to 3, and then saturates.

Families of the vortex-bullet solutions with vorticities ℓ=1\ell= 1, 2, 3 and 4 are represented in Figure 2 by means of dependences N(β)N(\beta), for q=1/2q = 1/2 and different values of γ\gamma [see Eq. (4)]. These curves identify families meeting the celebrated Vakhitov-Kolokolov (VK) criterion, which is the necessary (although not sufficient) stability criterion for solitons [11, 43, 46]. It is observed that larger values of γ\gamma (stronger guiding profiles) are required to secure the validity of the VK criterion for the families belonging to higher values of ℓ\ell.

The norm of vortex bullets versus the propagation constant

Figure 2. The norm of vortex bullets vs. the propagation constant for different values of the guiding-profile’s strength γ\gamma, fixed width parameter q=1/2q = 1/2, see Eq. (4), and four values of the winding number, from ℓ=1\ell= 1 to 4. Points marked with letters correspond to the modes whose propagation is displayed in Figure 6 (the letters match in both figures).

In the case of the weak guiding channel, which corresponds to small values of γ\gamma in Eq. (4), the N(β)N(\beta) dependence is close to that for strongly unstable 3D solitons in free space, which is determined by the obvious scaling relation, N∼β−1/2N \sim\beta^{-1/2}. It explains the divergence of N(β)N(\beta) at β→0\beta\to0 for (γ=2,ℓ=1)(\gamma= 2, \ell= 1), (γ=4,ℓ=2)(\gamma= 4, \ell= 2), and (γ=10,ℓ=4)(\gamma= 10, \ell= 4) in Figure 2 (in this limit, the solitons are very broad, with the size scaling as 1/β1/\sqrt{\beta}, while their amplitude is small and scales as β\sqrt{\beta}). On the other hand, the curves N(β)N(\beta) which originate, at N=0N = 0, from eigenvalues βEV>0\beta_{\mathrm{EV}} > 0 for larger γ\gamma (e.g., γ=4\gamma= 4 for ℓ=1\ell= 1 and γ=30\gamma= 30 for ℓ=4\ell= 4) bifurcate from the respective eigenmodes which exist at β=βEV\beta= \beta_{\mathrm{EV}} in the linearized version of Eq. (6). Intermediate values of γ\gamma, at which the bullet branches originate, at β=0\beta= 0, from finite values N>0N > 0 in Figure 2 (e.g., γ=6\gamma= 6 and δ=2\delta= 2) correspond to 3D solitons, at β=0\beta= 0, with slowly decaying tails.

This intermediate behavior, i.e., the family of nonlinear solutions originating at a finite value of NN, is quite unusual and is an effect of the waveguide combined with the nonlinearity as none of these factors can separately provide an explanation for it. To check that the effect is not an artifact produced by a coarse step in varying the value of β\beta, as it approaches zero, or by a too narrow calculation domain, we have carried out a careful analysis of the solutions at values of β\beta close to zero. The region of β→0\beta\to0 was explored in several stages, scanning values of β\beta on different scales. The results are summarized in Figure 3.

Four-panel plots of norm, temporal field profiles, field widths, and maximum field amplitude

Figure 3. (a) The norm vs. low values of the propagation constant for the family of solutions with ℓ=1\ell= 1, γ=3.3\gamma= 3.3, q=0.5q = 0.5. (b) Temporal field profiles in the cross-section r=rmax⁡r = r_{\max}, corresponding to the field maximum. (c) Widths vs. β\beta of the field in the spatial (wrw_r) and temporal (wTw_T) directions in the cross-sections corresponding to the field maximum. (d) The maximum field amplitude vs. β\beta.

Figure 3(a) shows the region of small β\beta on the logarithmic scale, corresponding to the solution family with ℓ=1\ell= 1, γ=3.3\gamma= 3.3. Different examples are presented in Figure 3(b), showing tails that decay to zero in the calculation domain, which extends to the interval of [−300,300][-300, 300] in both rr and TT (though a smaller domain is shown in the figure for a better clarity), to secure a negligible effect of the boundaries on the solution. A fine mesh of 3000 ×\times 3000 points in both coordinates was used for the numerical computations. This mesh was later made finer, including 5000 ×\times 5000 and even 7000 ×\times 7000 points for the sake of comparison, to make it sure that the discretization is suitable to produce correct results. In Figure 3(c) the field size, calculated as the full width at half maximum (FWHM), is shown for both spatial and temporal profiles. The width initially increases, as β\beta varies from larger values toward β=0\beta= 0 [cf. the examples displayed in Figure 1(b)]. This happens mainly for the temporal profile, as the waveguide helps to keep the spatial width at lower values. Further, the width saturates and converges to a finite value at β→0\beta\to0. Similarly, the maximum amplitude decreases, converging to a nonzero value [Figure 3(d)].

Figure 2 suggests that there might exist a special value of γ\gamma (close to γ=8\gamma= 8 at ℓ=2\ell= 2, γ=16\gamma= 16 at ℓ=3\ell= 3, and to γ=21\gamma= 21 at ℓ=4\ell= 4), which corresponds to a separatrix between the VK-complying and non-complying branches, with N(β)=constN(\beta) = \mathrm{const}. Such a degenerate branch would represent a family of Townes solitons [11]. Being formally neutral with respect to the VK criterion, Townes solitons are subject to a weakly growing instability, that eventually leads to the onset of the collapse [11]. In reality, detailed numerical results demonstrate that NN always decreases at sufficiently large values of β\beta. What is actually found is a value of γ\gamma marking the transition from monotonously-decreasing N(β)N(\beta) curves to ones with a maximum of N(β)N(\beta), that separates regions with dN/dβ<0\mathrm{d}N/\mathrm{d}\beta< 0 and dN/dβ>0\mathrm{d}N/\mathrm{d}\beta> 0. This value, which is, e.g., γ∼3.14\gamma\sim3.14 for ℓ=1\ell= 1 in Figure 2, has an almost null slope for small values of β\beta, resembling a branch of Townes solitons, but it ultimately rises to a higher finite value of NN at β→0\beta\to0. Further increasing γ\gamma, a value is attained at which the bifurcation from the linear mode commences.

The VK criterion is not sufficient for the stability and, additionally, the fact that the optical bullets are endowed with angular momentum implies the possibility of azimuthal instabilities, which the VK criterion cannot address. It is then necessary to check the stability by means of numerical simulations of the propagation. For ℓ=1\ell= 1 it was already found that the stability can be maintained for the propagation distances larger than the respective Rayleigh (diffraction) length, zR=k0w02/2z_R = k_0 w_0^2/2, where w0w_0 is the spatial width of the optical bullet. For instance, the profiles displayed in Figure 1 have w0<10w_0 < 10, with zR∼50z_R \sim50. Simulations were carried out applying the beam propagation method (BPM) to the 3D equation [Eq. (3)] and using different vortex bullets as initial conditions. For this purpose, a finite-difference scheme was used to discretize the transverse variables (x,y)(x,y) and the temporal one (TT). The equation kernel was applied in two steps, following the usual splitting operator techniques [47], to obtain the propagated field along the step dz\mathrm{d}z. First, the Laplace operator is applied, using an alternating-direction implicit (ADI) method, using transparent boundary conditions at all domain boundaries [48]. As this algorithm is not unconditionally numerically-stable in three dimensions, an improved version of the ADI algorithm was used [49]. Second, the potential V(r)V(r) and nonlinear term were applied, using the fourth-order Runge–Kutta algorithm [47]. The propagation simulations in this work were performed using a mesh of 300 ×\times 300 ×\times 300 points and a step size dz=0.002\mathrm{d}z = 0.002.

Typical examples of the instability development of the toroidal vortex bullets with winding numbers ℓ=2,3\ell= 2, 3 and 4, initiated by the above-mentioned azimuthal instability, are presented in Figure 4, in the form of 2D images representing the plane of the vortex ring. It is seen that the instability which splits the axially symmetric structures develops much faster for ℓ=3\ell= 3 and 4 than for ℓ=2\ell= 2. To carry out a more quantitative study of this instability, we have performed the simulations for optical bullets with different values of the parameters, namely the charge ℓ\ell, index strength, γ\gamma, and index-profile width qq, maintaining constant values of other two parameters in all the three cases. The propagation constant was fixed as β=3\beta= 3, as for such a value there exists a solution (bullet) for a wide range of parameters, and also because the azimuthal instability sets in after a relatively short propagation distance. The results are summarized in Figure 5. The instability growth rate represented in panels (a-c) is evaluated as the inverse of the distance at which the vortex breaks, taking, as a breaking criterion, the position at which the maximum amplitude of the field changes by 10% against the initial one.

Development of the splitting instability for toroidal vortex bullets

Figure 4. The development of the splitting (azimuthal) instability for the toroidal vortex bullets with winding numbers ℓ=2\ell= 2, 3, and 4. In all cases, the guiding-profile strength is γ=4\gamma= 4. (a) ℓ=2\ell= 2, q=0.16q = 0.16, β=1.0\beta= 1.0; (b) ℓ=3\ell= 3, q=0.08q = 0.08, β=1.0\beta= 1.0; (c) ℓ=4\ell= 4, q=0.04q = 0.04, β=1.5\beta= 1.5. The panels show the plane (x,y)(x,y) with T=0T = 0, where the field maximum is located.

Figure 5. (a-c) The instability growth rate as a function of different system’s parameters: topological charge (a), index strength (b), and the profile’s width w=1/qw = 1/\sqrt{q} (c), with fixed values of other parameters, as indicated in each panel. The propagation constant is β=3\beta= 3 in all cases. Dashed thin lines connecting the points are plotted as a guide for the eye. (d) The maximum field amplitude plotted against the propagation distance zz for different values of γ\gamma and qq. (e) The maximum field amplitude plotted against the propagation distance zz for different values of qq.

In Figure 5(a) the instability growth rate is plotted vs. the winding number, confirming the fast growth of the instability as ℓ\ell increases. On the other hand, at a particular value of the winding number, the instability growth rate quickly decreases as the waveguide strength γ\gamma increases, as shown in Figure 5(b), approaching zero at large values of γ\gamma, implying slow development of the instability development, which can lead to practical suppression of the instability in some cases [see also 5(d), where the maximum amplitude of the field is plotted against propagation constant]. Finally, in Figure 5(c) the instability growth rate is plotted against the index-profile’s width. The maximum field amplitude vs. the propagation distance for the corresponding points is shown in Figure 5(e). In this case, the increase of the profile’s width (i.e., decrease of qq) indeed delays the onset of the azimuthal instability. As a summary, we conclude that stronger guiding, provided by the deeper index-modulation profile, helps to suppress the spontaneous destruction of the vortex ring. Lower values of the propagation constant β\beta also produce much weaker azimuthal instability (at least for relatively small winding numbers ℓ\ell).

In Figure 6 we display systematic results, obtained by plotting the evolution of the field’s amplitude vs. the propagation distance zz. The states belonging to the VK-unstable branches, such as those labelled A, J and H (for ℓ=1\ell= 1, 3 and 4, respectively) quickly suffer the collapse, attaining high values of the amplitude (which is followed by rapid decay, which may be an effect depending on the numerical scheme). On the other hand, the VK-stable states (belonging to the branches with dN/dβ>0dN/d\beta> 0) do not collapse and tend to slowly spread out in the temporal direction, while their spatial size remains confined by the guiding profile. The spreading rate depends on the guiding strength γ\gamma, propagation constant β\beta, and mainly on the vorticity. In particular, for ℓ=1\ell= 1 the amplitude maintains a nearly constant value for a long propagation

Figure 6. The amplitude of the field in numerical simulations of the propagating vortex bullet, scaled by the input value at z=0z = 0, vs. the propagation distance for different stationary states marked in Figure 2. Letters label curves corresponding to the modes marked in Figure 2.

distance (curve B in Figure 6). The vortex bullets decay essentially faster for larger values of ℓ\ell (compare curve B for ℓ=1\ell= 1 to D, E, F for ℓ=2\ell= 2, and curves I and K for ℓ=3\ell= 3 and 4, respectively). For a fixed vorticity, the effective stability range increases with the increase of the guiding-profile strength γ\gamma (see, e.g., curves D, E and F in Figure 6 for ℓ=2\ell= 2 and γ=8\gamma= 8, 10 and 12, respectively), which is similar to the effect securing the VK criterion by larger values of γ\gamma in Figure 2. As one can see in Figure 2, γ=4\gamma= 4 is sufficient for ℓ=1\ell= 1, while it is necessary to have γ≳8\gamma\gtrsim8 for ℓ=2\ell= 2, γ≳16\gamma\gtrsim16 for ℓ=3\ell= 3, and γ≳24\gamma\gtrsim24 for ℓ=4\ell= 4. On the other hand, the effective stability does not strongly depend on the propagation constant β\beta, although the decay is slightly faster for larger β\beta (compare, for instance, curves G and E).

The strong instability of the vortex bullets for vorticities ℓ≥2\ell\ge2 suggests one to seek additional stabilization scenarios in this case. In particular, it is interesting to consider wider guiding profiles, as it is then necessary to apply a higher power for balancing the spatial diffraction by the Kerr nonlinearity; in turn, the stronger self-attraction may enhance the confinement in the temporal direction. This fact is essential, as the bullet’s spreading occurs in such temporal direction.

The effect of the variation of the width parameter qq is presented in Figure 7, for a fixed guiding strength, γ=4\gamma= 4, and the same vorticities as above, from ℓ=1\ell= 1 to 4. As expected, the VK criterion holds, making the vortex bullets more robust for smaller values of qq, wider guiding profiles, while larger qq strongly destabilize the modes by making dN/dβ<0dN/d\beta< 0. The interpretation of the solution branches with N(β)N(\beta) which diverge at β→0\beta\to0, or originate at N=0N = 0 from the linear modes with β=βEV>0\beta= \beta_{\mathrm{EV}} > 0, and terminate, at β=0\beta= 0, at finite values of NN is the same as outlined above in the connection to Figure 2.

Figure 7. The norm of the vortex bullets vs. the propagation constant for different values of the width parameter qq and a fixed strength of the guiding profile, γ=4\gamma= 4, in Eq. (4). Each panel corresponds to a particular value of the winding number ℓ\ell. Points labeled by letters correspond to the propagating modes shown in Figure 8.

Figure 8 displays results of the propagation-stability test for representative examples of the vortex bullets, corresponding to dots L, Q, R, S, U marked in Figure 7, and additional examples defined in the caption to Figure 8. It is seen that, at least for ℓ=1\ell= 1, it is possible to find an optimum value of qq which secures the effectively stable prop-

Propagation-stability tests for vortex bullets

Figure 8. The same as in Figure 6, but for the vortex bullets which correspond to dots marked by letters L, Q, R, S, and U in Figure 7, and additional ones, viz., M (ℓ=1\ell= 1, γ=2\gamma= 2, q=0.25q = 0.25, β=0.08\beta= 0.08), N (ℓ=1\ell= 1, γ=14\gamma= 14, q=2q = 2, β=0.1\beta= 0.1), P (ℓ=2\ell= 2, γ=6\gamma= 6, q=0.25q = 0.25, β=0.6\beta= 0.6), and T (ℓ=3\ell= 3, γ=4\gamma= 4, q=0.01q = 0.01, β=2.7\beta= 2.7).

agation over a desired distance. For ℓ=2\ell= 2 or 3, a reasonably large distance of the robust propagation, amounting to a few Rayleigh lengths, can be provided too. An example is curve T, which corresponds to small q=0.01q = 0.01, i.e., a wide guiding profile with width w=1/q=10w = 1/\sqrt{q} = 10. The stable propagation is harder to be supported for higher values of ℓ\ell, at which the above-mentioned azimuthal modulational instability of the toroidal vortex bullets quickly kicks in. In Figure 8 this is shown, e.g., by curves S or U, while in the above-mentioned quasi-stable case of curve T, the azimuthal instability only commences at very large values of zz.

Simulations of collisions between effectively stable vortex bullets in the temporal direction demonstrate that the collisions are elastic, which is an additional argument in favor of the robustness of these modes. As an example, the collision between identical bullets with vorticity ℓ=2\ell= 2, which correspond to symbol R in Figure 7 and Figure 8, is displayed in Figure 9. In this case, even though the slow decay of individual bullets has started prior to the collision, the bullets perfectly persist and bounce back elastically. The colliding bullets in the TT direction are generated introducing a linear phase exp⁡[±i(sin⁡θ)T]\exp[\pm i(\sin\theta)T], which introduces an angle θ\theta between the propagation and time (TT) directions. For this particular simulation, the angle was chosen as θ=π/24\theta= \pi/24, which results in the collision velocity vT=0.26v_T = 0.26 along the TT direction. A similar result is demonstrated by the colliding vortex bullets with ℓ=3\ell= 3, if their width parameter is reduced to q=0.01q = 0.01.

Six-panel simulation of an elastic collision between two quasi-stable toroidal vortex bullets

Figure 9. The simulated elastic collision between two quasi-stable toroidal vortex bullets with ℓ=2\ell= 2, which correspond to symbol R in Figure 7 and Figure 8. The parameters are γ=4\gamma= 4, q=0.04q = 0.04, β=2\beta= 2.

In Figure 10 we present the same simulation but for colliding bullets with opposite charges, ℓ=±2\ell= \pm2, which implies zero total angular momentum (the collision with “hidden vorticity”). As it was found in other systems[50, 51, 52], outcomes of the collisions may be completely different from the case of the collision between bullets with identical vorticities. As seen in Figure 10, collisions between bullets with opposite vorticities are destructive.

Six-panel simulation of a collision between oppositely charged vortex bullets

Figure 10. The same simulation as in Figure 9 but for opposite charged bullets (ℓ=±2\ell= \pm2 respectively). Parameters: γ=4\gamma= 4, q=0.04q = 0.04, β=2\beta= 2.

Conclusion

To address the known problem of the stabilization of multidimensional solitons with embedded vorticity, we have proposed the 3D nonlinear-Schrödinger model, which includes the anomalous GVD, cubic self-attractive nonlinearity, and a guiding index structure in the transverse plane, in the form of an axisymmetric Gaussian. The same model applies to the self-attractive atomic Bose-Einstein condensate. The numerical analysis has demonstrated that the value of the vorticity (ℓ=1,2,3,4\ell= 1, 2, 3, 4) is the main factor affecting the (in)stability of the 3D vortex bullets, along with the strength and width of the guiding profile. Generally, the instability is quickly enhanced with the increase of ℓ\ell. The propagation of the bullets is more robust for the solutions satisfying the VK (Vakhitov-Kolokolov) criterion. A degenerate family of the Townes solitons is found at the boundary between VK-stable and unstable solutions. For a given vorticity, it is possible to find an operational regime, with sufficiently large strength and width of the guiding structure, which provide effectively stable propagation of the bullets over a required distance. Collisions between the effectively stable vortex bullets, in the temporal direction, are fully elastic.

Acknowledgements

This publication is part of the R&D&i project PID2023-146884NB-I00 funded by Ministerio de Ciencia e Innovación. This work was also supported by grant ED431B 2021/22 (Xunta de Galicia).

References

  1. [1]B. A. Malomed, D. Mihalache, F. Wise, and L. Torner, Spatiotemporal optical solitons, Journal of Optics B: Quantum and Semiclassical Optics 7, R53 (2005).
  2. [2]Y. Silberberg, Collapse of optical pulses, Optics letters 15, 1282 (1990).pubmed.ncbi.nlm.nih.gov/19771066
  3. [3]Y. S. Kivshar and G. P. Agrawal, Optical solitons: from fibers to photonic crystals (Academic press, San Diego, California, 2003).
  4. [4]M. Bock, D. Ueberschaer, G. Steinmeyer, and U. Griebner, Generation of two-cycle light bullets with 2.0 μm energy via self-compression of 5-μm pulses in zinc sulfide, Laser & Photonics Reviews 19, e01102 (2025).
  5. [5]S. V. Sazonov, Two-color bright–dark light bullets at the presence of quasi-resonant impurities, Laser Physics Letters 22, 115201 (2025).DOI
  6. [6]P. Béjot, Space-time multimode helical light bullets in ring-core fibers, J. Opt. Soc. Am. B 42, 1430 (2025).
  7. [7]Z. Chen, Z. Wu, and J. Zeng, Light gap bullets in defocusing media with optical lattices, Chaos, Solitons & Fractals 174, 113785 (2023).
  8. [8]A. B. Aceves, O. V. Shtyrina, A. M. Rubenchik, M. P. Fedoruk, and S. K. Turitsyn, Spatiotemporal optical bullets in two-dimensional fiber arrays and their stability, Phys. Rev. A 91, 033810 (2015).DOI
  9. [9]H. Wang, C. Guo, and S. Fan, Spatiotemporal steering of nondiffracting wave packets, Phys. Rev. Lett. 134, 073803 (2025).DOI
  10. [10]X. Han, S. Li, X. Ge, Z. Man, S. Fu, B. Wang, C. Chen, T. Wan, H. Li, and L. Zhang, Nonlinear propagation characteristics of space-time circular airyprme wave packets in a Kerr-type noninstantaneous media, Optics & Laser Technology 197, 114742 (2026).DOI
  11. [11]G. Fibich, The Nonlinear Schrödinger Equation: Singular Solutions and Optical Collapse, Applied Mathematical Sciences (Springer International Publishing, Switzerland, 2015).DOI
  12. [12]V. I. Bespalov and V. I. Talanov, Filamentary structure of light beams in nonlinear liquids, Soviet Journal of Ex- perimental and Theoretical Physics Letters 3, 307 (1966).
  13. [13]C. Sulem and P.-L. Sulem, The nonlinear Schrödinger equation: self-focusing and wave collapse, Vol. 139 (Springer Science & Business Media, New York, 2007).
  14. [14]D. E. Edmundson, Unstable higher modes of a three-dimensional nonlinear schrödinger equation, Physical Review E 55, 7636 (1997).DOI
  15. [15]B. A. Malomed, Multidimensional solitons (AIP Publishing, Melville, New York, 2022).
  16. [16]N. N. Akhmediev, V. Korneev, and R. F. Nabiev, Modulation instability of the ground state of the nonlinear wave equation: optical machine gun, Optics letters 17, 393 (1992).
  17. [17]D. Edmundson and R. Enns, Robust bistable light bullets, Optics Letters 17, 586 (1992).pubmed.ncbi.nlm.nih.gov/19794566
  18. [18]H. Michinel, J. Campo-Táboas, R. García-Fernández, J. Salgueiro, and M. Quiroga-Teixeiro, Liquid light condensates, Physical Review E 65, 066604 (2002).pubmed.ncbi.nlm.nih.gov/12188848
  19. [19]H. Michinel, M. J. Paz-Alonso, and V. M. Pérez-García, Turning light into a liquid via atomic coherence, Phys. Rev. Lett. 96, 023903 (2006).pubmed.ncbi.nlm.nih.gov/16486577
  20. [20]W. Krolikowski, O. Bang, J. J. Rasmussen, and J. Wyller, Modulational instability in nonlocal nonlinear kerr media, Physical Review E 64, 016612 (2001).DOI
  21. [21]O. Bang, W. Krolikowski, J. Wyller, and J. J. Rasmussen, Collapse arrest and soliton stabilization in nonlocal nonlinear media, Physical Review E 66, 046619 (2002).DOI
  22. [22]D. Mihalache, D. Mazilu, F. Lederer, B. Malomed, Y. V. Kartashov, L.-C. Crasovan, and L. Torner, Three-dimensional spatiotemporal optical solitons in nonlocal nonlinear media, Physical Review E—Statistical, Nonlinear, and Soft Matter Physics 73, 025601 (2006).DOI
  23. [23]A. Paredes, D. N. Olivieri, and H. Michinel, From optics to dark matter: A review on nonlinear schrödinger-poisson systems, Physica D: Nonlinear Phenomena 403, 132301 (2020).DOI
  24. [24]B. B. Baizakov, B. A. Malomed, and M. Salerno, Multidimensional solitons in periodic potentials, Europhysics letters 63, 642 (2003).DOI
  25. [25]D. Mihalache, D. Mazilu, F. Lederer, B. Malomed, Y. V. Kartashov, L.-C. Crasovan, and L. Torner, Stable spatiotemporal solitons in bessel optical lattices, Physical review letters 95, 023902 (2005).DOI
  26. [26]F. Diebel, M. Boguslawski, T. Dadalyan, R. Drampyan, and C. Denz, Controlled soliton formation in tailored bessel photonic lattices, Optics Express 24, 12933 (2016).DOI
  27. [27]N. Aközbek and S. John, Optical solitary waves in two-and three-dimensional nonlinear photonic band-gap structures, Physical Review E 57, 2287 (1998).DOI
  28. [28]Y. V. Kartashov, Light bullets in moiré lattices, Opt. Lett. 47, 4528 (2022).DOI
  29. [29]C. Milián, Y. V. Kartashov, and L. Torner, Robust ultrashort light bullets in strongly twisted waveguide arrays, Phys. Rev. Lett. 123, 133902 (2019).DOI
  30. [30]J. F. Nye and M. V. Berry, Dislocations in wave trains, Proc. Math. Phys. Eng. Sci. 336, 165 (1974).
  31. [31]P. Coullet, L. Gil, and F. Rocca, Optical vortices, Optics Communications 73, 403 (1989).DOI
  32. [32]A. S. Desyatnikov, A. A. Sukhorukov, and Y. S. Kivshar, Azimuthons: spatially modulated vortex solitons, Physical review letters 95, 203904 (2005).pubmed.ncbi.nlm.nih.gov/16384059
  33. [33]A. S. Desyatnikov, Y. S. Kivshar, and L. Torner, Optical vortices and vortex solitons, in Progress in optics, Vol. 47 (Elsevier, 2005) pp. 291–391.
  34. [34]M. Quiroga-Teixeiro and H. Michinel, Stable azimuthal stationary state in quintic nonlinear optical media, J. Opt. Soc. Am. B 14, 2004 (1997).
  35. [35]H. Michinel, J. R. Salgueiro, and M. J. Paz-Alonso, Square vortex solitons with a large angular momentum, Phys. Rev. E 70, 066605 (2004).DOI
  36. [36]Y. I. Salamin, Fields of a bessel-bessel light bullet of arbitrary order in an under-dense plasma, Scientific Reports 8, 11362 (2018).DOI
  37. [37]C. Hang and G. Huang, Ultraslow helical optical bullets and their acceleration in magneto-optically controlled coherent atomic media, Physical Review A—Atomic, Molecular, and Optical Physics 87, 053809 (2013).DOI
  38. [38]M. A. Porras, Upper bound to the orbital angular momentum carried by an ultrashort pulse, Physical Review Letters 122, 123904 (2019).DOI
  39. [39]Z. Chen, M. Segev, D. W. Wilson, R. E. Muller, and P. D. Maker, Self-trapping of an optical vortex by use of the bulk photovlotaic effect, Physical Review Letters 78, 2948 (1997).DOI
  40. [40]H. Zhang, T. Zhou, and C. Dai, Stabilization of higher-order vortex solitons by means of nonlocal nonlinearity, Physical Review A 105, 013520 (2022).DOI
  41. [41]F. Eilenberger, K. Prater, S. Minardi, R. Geiss, U. Röpke, J. Kobelke, K. Schuster, H. Bartelt, S. Nolte, A. Tünnermann, and T. Pertsch, Observation of discrete, vortex light bullets, Phys. Rev. X 3, 041031 (2013).DOI
  42. [42]Z. Zhang, S. K. Ivanov, Y. Li, Y. V. Kartashov, and Y. Zhang, Vortex light bullets guided by disclination defects, Nanophotonics 15, e070206 (2026), https://onlinelibrary.wiley.com/doi/pdf/10.1002/nap2.70206.DOI
  43. [43]C. F. Sánchez, A. Paredes, H. Michinel, B. A. Malomed, and J. R. Salgueiro, Guided vortex bullets, Phys. Rev. Lett. 136, 033802 (2026).DOI
  44. [44]V. Dunjko, V. Lorent, and M. Olshanii, Bosons in cigar-shaped traps: Thomas-fermi regime, tonks-girardeau regime, and in between, Phys. Rev. Lett. 86, 5413 (2001).pubmed.ncbi.nlm.nih.gov/11415264
  45. [45]J. R. Salgueiro, D. Olivieri, and H. Michinel, Computation of linear and nonlinear stationary states of photonic structures using modern iterative solvers, Optical and quantum electronics 39, 239 (2007).DOI
  46. [46]N. G. Vakhitov and A. A. Kolokolov, Stationary solutions of the wave equation in a medium with nonlinearity saturation, Radiophysics and Quantum Electronics 16, 783 (1973).DOI
  47. [47]W. H. Press, S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery, Numerical recipes in C, 2nd ed. (Cambridge University Press, Cambridge, 1992).
  48. [48]G. R. Hadley, Transparent boundary condition for the beam propagation method, IEEE Journal of quantum electronics 28, 363 (1992).DOI
  49. [49]M. J. Chang and L. C. Chow, Improved alternating-direction implicit method for solving transient three-dimensional heat diffusion problems, Numerical Heat Transfer 19, 69 (1991).DOI
  50. [50]J. R. Salgueiro, Vector-vortex solitons in nonlinear photonic crystal fibers, J. Opt. 18, 074004 (2016).DOI
  51. [51]Y. V. Izdebskaya, J. Rebling, A. S. Desyatnikov, and Y. S. Kivshar, Observation of vector solitons with hidden vorticity, Opt. Lett. 37, 767 (2012).
  52. [52]M. Brtka, A. Gammal, and B. A. Malomed, Hidden vorticity in binary Bose-Einstein condensates, Phys. Rev. A 82, 053610 (2010).DOI

Paper details

vlnc:W_gkafev6b
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 16

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