1:
2: \documentstyle{article}
3:
4: \begin{document}
5:
6: \title{Stable three-dimensional spinning optical solitons supported by competing
7: quadratic and cubic nonlinearities}
8: \author{}
9: \maketitle
10:
11: \begin{center}
12: D. Mihalache$^1$, D. Mazilu$^1$, L.-C. Crasovan$^1$, I. Towers$^2$, B. A.
13: Malomed$^2$, A. V. Buryak$^3$, L. Torner$^4$, and F. Lederer$^5$\\[0pt]
14:
15: \bigskip
16:
17: $^1$Department of Theoretical Physics, Institute of Atomic Physics, PO Box
18: MG-6, Bucharest, Romania\\[0pt]
19:
20: $^2$Faculty of Engineering, Tel Aviv University, Tel Aviv 69978, Israel
21: \\[0pt]
22:
23: $^3$School of Mathematics and Statistics, University of New South Wales at
24: ADFA, Canberra, ACT 2600, Australia\\[0pt]
25:
26: $^4$Department of Signal Theory and Communications, Universitat Politecnica
27: de Catalunya, ES 08034 Barcelona, Spain\\[0pt]
28:
29: $^5$Institute of Solid State Theory and Theoretical Optics,
30: Friedrich-Schiller University Jena, Max-Wien-Platz 1, D-07743, Jena, Germany
31: \\[0pt]
32:
33: \bigskip \bigskip
34:
35: {\large {Abstract} }
36: \end{center}
37:
38: We show that the quadratic interaction of fundamental and second harmonics
39: in a bulk dispersive medium, combined with self-defocusing cubic
40: nonlinearity, give rise to completely localized spatiotemporal solitons
41: (vortex tori) with vorticity $s=1$. There is no threshold necessary for the
42: existence of these solitons. They are found to be stable if their energy
43: exceeds a certain critical value, so that the stability domain occupies
44: about 10 {\%} of the existence region of the solitons. On the contrary to
45: spatial vortex solitons in the same model, the spatiotemporal ones with $s=2$
46: are never stable. These results might open the way for experimental
47: observation of spinning three-dimensional solitons in optical media.
48:
49: \newpage
50:
51: \section{Introduction}
52:
53: Solitons, i.e., self-trapped light beams or pulses that are supported by a
54: balance between diffraction and/or dispersion and nonlinearity, are
55: prominent objects in nonlinear optics \cite{George}. Optical spatiotemporal
56: solitons (STS) \cite{KanRub}, alias superspikes \cite{MBA} or light bullets
57: \cite{Yaron}, were predicted in many works \cite{KanRub}~-~\cite{greek}.
58: They result from the simultaneous balance of diffraction and group-velocity
59: dispersion (GVD) by self-focusing. Although they cannot be stable in the
60: uniform self-focusing Kerr ($\chi ^{(3)}$) medium \cite{collapse}, stability
61: can be achieved in saturable \cite{MBA,AS,Edmundson}, quadratically
62: nonlinear ($\chi ^{(2)}$) \cite{KanRub,quadr1,quadr2,quadr3}, and
63: graded-index Kerr media \cite{Agrawal}. STS can also be found in
64: off-resonance two-level systems \cite{Igor}, in self-induced-transparency
65: media \cite{Miriam}, as well as in engineered tandem structures
66: incorporating quadratically nonlinear slices \cite{tandem}.
67:
68: While a fully localized ``light bullet'' in three dimensions (3D) has not
69: yet been found in an experiment, two-dimensional (2D) STS in a bulk $\chi
70: ^{(2)}$ medium were observed \ in Ref. \cite{Wise}. That work reported the
71: formation of pulses in quadratic media, which overcome diffraction in one
72: transverse spatial dimension and GVD in the longitudinal direction. However,
73: such experiments were performed by means of the tilted-pulse technique,
74: which employs highly elliptical beams; therefore, diffraction is negligible
75: in the remaining transverse spatial dimension.
76:
77: Optical vortex solitons constitute another class of self-supporting objects,
78: that have attracted much attention because of possible applications to the
79: all-optical processing of information, or to guiding and trapping of atoms.
80: The concepts of a multidimensional optical soliton and of an optical vortex
81: may be combined, giving rise to {\it spinning} (vortex) solitons. Starting
82: with the seminal works \cite{vortex}, both delocalized (``dark'') and
83: localized (``bright'') optical vortices were investigated in various 2D
84: environments \cite{DiTrapp,Sammut,2DvortexYuri,2DvortexSegev}. In the 3D
85: case, the bright spinning solitons take the shape of a torus (``doughnut'')
86: \cite{Anton,new}.
87:
88: For bright vortex solitons, stability is a major concern, as, unlike their
89: zero-spin counterparts, the spinning solitons are apt to be destabilized by
90: azimuthal perturbations. For 2D models with $\chi ^{(2)}$ nonlinearities, an
91: azimuthal instability was discovered by simulations \cite{unstable1} and
92: observed experimentally \cite{experiment}. As a result, a soliton with spin $
93: 1$ splits into three or two fragments in the form of separating zero-spin
94: soliton. Numerical simulations of the 3D spinning STS in the $\chi ^{(2)}$
95: model also demonstrates splitting into moving zero-spin solitons \cite{new}.
96:
97: Nevertheless, the $\chi ^{(2)}$ nonlinearity acting in combination with the
98: self-{\em defocusing} Kerr [$\chi _{-}^{(3)}$, where we use the subscript
99: ``minus'' to stress the self-repulsion] nonlinearity gives rise to stable
100: spinning (ring-shaped) 2D solitons with spin $s=1$ and $2$ \cite{Sammut}.
101: Models of this type for spatial [(2+1)-dimensional] solitons are well known
102: \cite{comp1,comp2}. The stability of the spinning solitons in the $\chi
103: ^{(2)}:\chi _{-}^{(3)}$ model may be realized as a result of competition
104: between the self-focusing and self-defocusing nonlinearities. This
105: understanding is further supported by the fact that stable spinning solitons
106: of the same type have also been found in another optical model displaying
107: both focusing and defocusing nonlinearities, viz., the one based on the
108: cubic-quintic (CQ) nonlinear Schr\"{o}dinger (NLS) equation. In addition to
109: optics, the same equations have been investigated in the contexts of
110: Bose-Einstein condensates (BECs) \cite{Abdullaev} and Langmuir waves in
111: plasmas \cite{plasma} (however, in the former case, the quintic nonlinearity
112: arises from three-body interactions, which also give rise to losses by
113: recombination of BEC constituents into different species, thus making the
114: quintic nonlinear coefficient a complex one).
115:
116: In the first direct simulations of 2D solitons with spin $1$ in the CQ
117: model, reported in the pioneer work \cite{Q}, it was found that they are
118: robust, provided that their energy is not too small \cite{Q}. Later
119: analysis, based on the computation of linear-stability eigenvalues,
120: demonstrated that some of the spinning 2D solitons considered in Ref. \cite
121: {Q} are subject to a weak azimuthal instability. Nonetheless, in another
122: part of their existence region, where they have a very large energy, the
123: solitons with spin $s=1$ and $s=2$ were confirmed to be stable in the 2D CQ
124: model \cite{IsaacPLA} (see also Ref. \cite{Berezhiani} for the stability
125: investigation of the solitons with spin $s=1$). Stable 2D vortex solitons in
126: the CQ model can self-trap from Gaussian inputs with an embedded vorticity
127: \cite{Michinel}. Notice that all the solitons with $s\geq 3$ have been
128: demonstrated to be unstable in the CQ model \cite{IsaacPLA}.
129:
130: A challenging issue is the search for physically relevant models in which
131: {\it stable} 3D spinning solitons exist. In fact, the only previously known
132: model which could support stable 3D vortex solitons was the Skyrme model
133: (see reviews \cite{reviews}). Very recently, we have found stable 3D
134: spinning STS in the CQ model, which could again be construed as a result of
135: the competition between self-focusing and self-defocusing \cite{PRL}. Direct
136: simulations of the 3D CQ model \cite{new2} demonstrated that 3D spinning
137: solitons with moderate energies were unstable against azimuthal
138: perturbations, while the ones with very large energies, i.e., broad
139: ``doughnuts'' with a small hole in the center, were robust under
140: propagation. However, a consistent stability analysis makes it necessary to
141: compute eigenvalues of small perturbations. By calculating the instability
142: growth rates, in Ref. \cite{PRL} it was rigorously shown that sufficiently
143: broad STS with spin $s=1$ are stable, the stability region occupying $
144: \approx 20\%$ of their existence region, while all the STS with $s\geq 2$
145: are unstable.
146:
147: The aim of this paper is to show that the existence of stable spinning 3D
148: solitons is a more generic fact, which is not limited to the CQ model
149: considered in Ref. \cite{PRL}. To this end, we will analyze the existence
150: and stability of spinning STS solitons in the 3D version of the
151: above-mentioned $\chi ^{(2)}:\chi _{-}^{(3)}$ model with the self-defocusing
152: cubic term. In section 2, the model is formulated, and general results
153: concerning the existence of 3D spinning STS in it, with different values of
154: the spin, are displayed. Fundamental results for the stability of the
155: spinning solitons, based on eigenvalues found from equations linearized
156: around the soliton solutions, are presented in section 3. Direct simulations
157: of the solitons' stability within the framework of the full nonlinear
158: equations are displayed in section 4, and section 5 concludes the work.
159:
160: \section{The model and spinning solitons}
161:
162: The scaled equations describing the reversible generation of the second
163: harmonic (SH) from a single fundamental-frequency (FF) component $u$, in the
164: presence of the self-defocusing cubic nonlinearity, dispersion and
165: diffraction in the (3+1)-dimensional geometry, are well known \cite
166: {quadr1,quadr2,Sammut,comp1,comp2}:
167: \begin{eqnarray}
168: &&i\frac{\partial u}{\partial Z}+\frac{1}{2}\left( \frac{\partial ^{2}u} {
169: \partial X^{2}}+\frac{\partial ^{2}u}{\partial Y^{2}}+\frac{\partial ^{2}u} {
170: \partial T^{2}}\right) +u^{\ast }\,v\,-(|u|^{2}+2|v|^{2})u\,=0, \nonumber \\
171: &&i\frac{\partial v}{\partial Z}+\frac{1}{4}\left( \frac{\partial ^{2}v} {
172: \partial X^{2}}+\frac{\partial ^{2}v}{\partial Y^{2}}+\sigma \frac{\partial
173: ^{2}v}{\partial T^{2}}\right) -\beta v+u^{2}\,-2(2|u|^{2}+|v|^{2})v\,=0.
174: \label{scaled}
175: \end{eqnarray}
176: Here, $T$, $X$, $Y$ and $Z$ are the normalized reduced time, transverse
177: spatial coordinates, and propagation distance, $u$ and $v$ are envelopes of
178: the FF and SH fields, and $\beta $ is a phase mismatch between the FF and SH
179: waves. In particular, the variables used in Eqs. (\ref{scaled}) are related
180: to their counterparts (to be denoted by tildes) in Ref. \cite{Sammut} as
181: follows: $u\equiv 2\widetilde{u}$, $v\equiv \widetilde{w}$, $\sqrt{2}\left(
182: x,y\right) \equiv \left( \widetilde{x},\widetilde{y}\right) $.
183:
184: Equations (\ref{scaled}) assume different GVD coefficients at the two
185: harmonics, $\sigma $ being their ratio \cite{quadr1}, but neglects the
186: Poynting-vector walkoff between the harmonics, and assumes that the temporal
187: group-velocity mismatch between them \cite{quadr3,walking,PO} has been
188: compensated. On the other hand, in the case $\sigma =1$ the model possesses
189: an additional spatiotemporal spherical symmetry \cite{quadr1,quadr2}. Below,
190: we will display results for the case $\sigma =1$, assuming that the
191: group-velocity mismatch may be neglected in this case too.
192:
193: We look for stationary solutions to Eqs. (1) in the form $u=U(r,T)\exp
194: \left( i\kappa Z+is\theta \right) $, $v=V(r,T)\exp \left[ 2\left( i\kappa
195: Z+is\theta \right) \right] $, where $\theta $ is the polar angle in the
196: plane $\left( x,y\right) $, $\kappa $ is a wave number shift, and the
197: integer $s$ is the above-mentioned spin. The amplitudes $U$ and $V$ may be
198: taken real, obeying the equations
199: \begin{eqnarray}
200: &&\frac{1}{2}\left( \frac{\partial ^{2}U}{\partial r^{2}}+\frac{1}{r}\frac{
201: \partial U}{\partial r}-\frac{s^{2}}{r^{2}}U+\frac{\partial ^{2}U}{\partial
202: T^{2}}\right) -\kappa U+UV-(U^{2}+2V^{2})U=0, \nonumber \\
203: &&\frac{1}{4}\left( \frac{\partial ^{2}V}{\partial r^{2}}+\frac{1}{r}\frac{
204: \partial V}{\partial r}-\frac{4s^{2}}{r^{2}}V+\sigma \frac{\partial ^{2}V}{
205: \partial T^{2}}\right) -(2\kappa +\beta )V+U^{2}-2(2U^{2}+V^{2})V=0.
206: \label{stat}
207: \end{eqnarray}
208:
209: Dynamical equations (\ref{scaled}) conserve the total energy
210: \begin{equation}
211: E~=~\int \int \int (\left| u\right| ^{2}+\left| v\right| ^{2})dXdYdT\equiv
212: E_{u}+E_{v}, \label{E}
213: \end{equation}
214: Hamiltonian
215: \begin{eqnarray}
216: H &=&\frac{1}{2}\int \int \int \left\{ \left[
217: (|u_{X}|^{2}+|u_{Y}|^{2}+|u_{T}|^{2})+\frac{1}{4}(|v_{X}|^{2}+|v_{Y}|^{2}+
218: \sigma |v_{T}|^{2})\right] \right. \nonumber \\
219: &&+\left. \left[ \beta |v|^{2}-(u^{\ast 2}v+u^{2}v^{\ast
220: })+(|u|^{4}+4|u|^{2}|v|^{2}+|v|^{4})\right] \vphantom { \left[
221: (|u_{X}|^{2}+|u_{Y}|^{2}+|u_{T}|^{2})+ \frac{1}{4}(|v_{X}|^{2}+|v_{Y}|^{2}+
222: \sigma |v_{T}|^{2})\right] }\right\} dXdYdT\,, \label{inv2}
223: \end{eqnarray}
224: momentum (equal to zero for the solutions considered), and longitudinal
225: component of the orbital angular momentum $L$ \cite{AA}. The following
226: relations between $L$, $H$ and $E$ for a stationary spinning STS follow from
227: Eqs. (\ref{stat}): $L=sE $, and
228: \begin{equation}
229: H=-\frac{1}{3}\kappa E+\frac{1}{3}\beta E_{v}-\frac{1}{3}\int \int \int
230: \left( |u|^{4}+4|u|^{2}|v|^{2}+|v|^{4}\right) dXdYdT. \label{H}
231: \end{equation}
232:
233: We have numerically found one-parameter families of stationary 3D spinning
234: solitons which have the shape of a doughnut with a hole (supported by a
235: phase dislocation) in the center. To this end, we solved numerically the
236: coupled system of equations (\ref{stat}) using a standard band-matrix
237: algorithm \cite{D-B} to deal with the corresponding two-point boundary-value
238: problem. We will display results for $\sigma =1$; however, we have also
239: found that the STS exist for all $\sigma \geq 0$, cf. Ref. \cite{quadr2},
240: where nonspinning STS were studied in detail for $\sigma \neq 1$.
241:
242: In Fig. 1 we summarize the output of extensive numerical calculations aimed
243: to detect the domains of existence and stability of spinning STS. The
244: continuous lines border the existence domain, and the dashed line
245: constitutes a boundary between stable and unstable regions in the parameter
246: plane $(\beta ,\kappa )$. The way the stability boundary has been found will
247: be explained in detail in the following sections.
248:
249: Shapes of three representative doughnut-like {\it stable} STS are plotted in
250: Fig. 2 for a fixed value of the net energy [see Eq. (\ref{E})], $E=12000$.
251: We see that, with the increase of the mismatch $\beta $, the energy of the
252: FF component of the spinning soliton increases, similar to the case of
253: nonspinning solitons in pure $\chi ^{(2)}$ media \cite{PO,SHT,TS,Yuri}.
254:
255: Below, we present systematic results which characterize spinning STS in the
256: case of the zero phase matching, $\beta =0$. In Fig. 3 we plot the curves
257: $\kappa =\kappa (E)$ and $H=H(E)$ for {\em both} nonspinning and spinning STS
258: in this case. The full and dashed lines in Fig. 3 correspond to stable and
259: unstable branches according to results presented below. The $s=0$ solitons
260: are stable according to the known Vakhitov-Kolokolov (VK) criterion, which
261: states that the fundamental ($s=0$) soliton branch undergoes a stability
262: change at the point $dE/d\kappa =0$ \cite{VK}.
263:
264: A feature shared by the nonspinning and spinning solitons, as it is evident
265: in Fig. 3, is the absence of any finite threshold for their existence. This
266: is a drastic difference from the recently studied STS in the CQ model, where
267: well-defined thresholds were found for zero and nonzero values of the spin
268: \cite{Anton,PRL}.
269:
270: \section{Stability eigenvalues of the spinning solitons}
271:
272: Complete understanding of the stability of solitons is provided by direct
273: simulations of the evolution equations (see below) together with the
274: analysis of Eqs. (\ref{scaled}) linearized about the stationary
275: spinning-soliton solution. In this section, we focus on the latter approach,
276: seeking for perturbation eigenmodes in a general form,
277: \begin{eqnarray}
278: &&u(Z,r,T,\theta )-U(r,T)\exp \left[ i(s\theta +\kappa Z)\right] \nonumber
279: \\
280: &=&f(r,T)\exp \left\{ \lambda _{n}Z+i[(s+n)\theta +\kappa Z]\right\}
281: +g^{\ast }(r,T)\exp \left\{ \lambda _{n}^{\ast }Z+i[(s-n)\theta +\kappa
282: Z]\right\} \,, \\
283: &&v(Z,r,T,\theta )-V(r,T)\exp \left[ 2i(s\theta +\kappa Z)\right] \nonumber
284: \\
285: &=&p(r,T)\exp \left\{ \lambda _{n}Z+i[(2s+n)\theta +2\kappa Z]\right\}
286: +q^{\ast }(r,T)\exp \left\{ \lambda _{n}^{\ast }Z+i[(2s-n)\theta +2\kappa
287: Z]\right\} \,,
288: \end{eqnarray}
289: where $n>0$ is an arbitrary integer azimuthal index of the perturbation,
290: $\lambda _{n}$ is the (complex) eigenvalue that needs to be found, and
291: functions $f$, $g$ and $p$, $q$ obey equations
292: \begin{eqnarray}
293: i\lambda _{n}f+\frac{1}{2}\left[ \frac{\partial ^{2}f}{\partial T^{2}}+
294: \frac{\partial ^{2}f}{\partial r^{2}}+r^{-1}\frac{\partial f}{\partial r}
295: -(s+n)^{2}r^{-2}f\right] -\kappa f && \nonumber \\
296: -2\left( U^{2}+V^{2}\right) \,f-\left( U^{2}-V\right) g-\left( 2UV-U\right)
297: p-2UVq &=&0, \\
298: -i\lambda _{n}g+\frac{1}{2}\left[ \frac{\partial ^{2}g}{\partial T^{2}}+
299: \frac{\partial ^{2}g}{\partial r^{2}}+r^{-1}\frac{\partial g}{\partial r}
300: -(s-n)^{2}r^{-2}g\right] -\kappa g && \nonumber \\
301: -2\left( U^{2}+V^{2}\right) \,g-\left( U^{2}-V\right) f-\left( 2UV-U\right)
302: q-2UVp &=&0, \\
303: i\lambda _{n}p+\frac{1}{4}\left[ \sigma \frac{\partial ^{2}p}{\partial T^{2}}
304: +\frac{\partial ^{2}p}{\partial r^{2}}+r^{-1}\frac{\partial p} {\partial r}
305: -(2s+n)^{2}r^{-2}p\right] -(2\kappa +\beta )p && \nonumber \\
306: -4\left( U^{2}+V^{2}\right) \,p-2V^{2}q-2\left( 2UV-U\right) f-4UVg &=&0, \\
307: -i\lambda _{n}q+\frac{1}{4}\left[ \sigma \frac{\partial ^{2}q}{\partial
308: T^{2} }+\frac{\partial ^{2}q}{\partial r^{2}}+r^{-1}\frac{\partial q} {
309: \partial r} -(2s-n)^{2}r^{-2}q\right] -(2\kappa +\beta )q && \nonumber \\
310: -4\left( U^{2}+V^{2}\right) \,q-2V^{2}p-2\left( 2UV-U\right) g-4UVf &=&0,
311: \end{eqnarray}
312: Physical solutions must decay exponentially at $r\rightarrow \infty $. At
313: $r\rightarrow 0$, $f$ and $g$ must vanish as $r^{\left| s\pm n\right| }$,
314: whereas $p$ and $q$ vanish as $r^{\left| 2s\pm n\right| }$.
315:
316: To solve the above equations and find the eigenvalues, we used a known
317: numerical procedure \cite{unstable1,Akh}, which produces results presented
318: in Fig. 4. The most persistent unstable eigenmode is found for value of the
319: azimuthal index $n=2$, for both $s=1$ and $s=2$. As is seen in Fig. 4, the
320: instability of the soliton with $s=1$, accounted for by ${\rm Re\,} \lambda
321: _{2}$, disappears with the increase of $\kappa $ at a stability-change
322: point, $\kappa _{{\rm st}}\approx 0.04572$, and the stability region extends
323: up to $\kappa =\kappa _{{\rm offset}}^{{\rm \ (3D)}}\approx 0.051$,
324: corresponding to the upper continuous line in Fig. 1, i.e., infinitely broad
325: solitons (which implies that the vortex of the dark-soliton type \cite
326: {vortex}, that may be regarded as an infinitely broad spinning soliton, is
327: stable too). The relative width of the stability region is $\left( \kappa _ {
328: {\rm offset}}^{{\rm (3D)}}-\kappa _{{\rm st} }\right) /\kappa _{{\rm offset}
329: }^{{\rm (3D)}}\approx 0.1$. However, {\em no} stability region has been
330: found for the 3D solitons with $s=2$, as well as in the 3D model of the CQ
331: type, and in contrast to the 2D vortex solitons in both the $\chi
332: ^{(2)}:\chi _{-}^{(3)}$ model with the competing quadratic and cubic
333: nonlinearities (the same as considered here) \cite{Sammut}, and 2D CQ model
334: \cite{IsaacPLA}.
335:
336: In the case when the spinning solitons are unstable, their instability is
337: {\it oscillatory}; the corresponding frequency, ${\rm Im}\lambda $ [see
338: Figs. 4 (c) and 4(d)] is found to be, generally, on the same order of
339: magnitude as ${\rm Re}\lambda $ at the maximum-instability point. In the
340: stable region, $\kappa \geq \kappa _{{\rm st}}$, all the eigenvalues are
341: purely imaginary. Oscillatory instabilities of solitons, characterized by
342: complex eigenvalues of the corresponding non-self-adjoint linear operator,
343: are typical to other conservative models of nonlinear optics \cite
344: {osc1,osc2,osc3}.
345:
346: \section{Direct simulations}
347:
348: The above results were checked against direct simulations of Eqs. (\ref
349: {scaled}), carried out by means of the Crank-Nicholson scheme. The
350: corresponding system of nonlinear partial differential equations was solved
351: by means of the Picard iteration method \cite{Ortega}, and the resulting
352: linear system was handled by means of the Gauss-Seidel iterative scheme. For
353: good convergence we needed, typically, five Picard iterations and fifteen
354: Gauss-Seidel iterations. We employed a transverse grid having $121\times
355: 121\times 91$ points, and a typical longitudinal step size was $\Delta Z=0.1$.
356: To avoid distortion of the instability development under the action of the
357: periodicity imposed by the Cartesian computational mesh, we added initial
358: perturbations that were mimicking random fluctuations in a real system (cf.
359: Ref. \cite{Edmundson}).
360:
361: To illustrate the evolution of a stable 3D ``bullet'' generated by an input
362: in the form of a completely localized Gaussian pulse with the energy
363: $E_{0}=5986$ [see Eq. (\ref{E})], into which a vortex with $s=1$ was
364: embedded, in Fig. 5 we show the energies of its two components vs. $Z$.
365: Robustness of the spinning STS is attested to by the fact that it can be
366: generated from a Gaussian with a nested vortex, whose shape is far from the
367: soliton's exact form. We see from Fig. 5 that there is a strong reshaping of
368: the input Gaussian, which leads to a redistribution of the energy between
369: the two components; some energy loss occurs, caused by emission of radiation
370: in the course of the formation of the stable STS. Figure 6 shows gray-scale
371: contour plots of the intensity and phase distribution in the FF component,
372: in both the input Gaussian with a nested vortex, and in the emerging
373: spinning STS with the vorticity $s=1$ at $Z=100$, corresponding to the same
374: case which was presented in Fig. 5. No further essential evolution of the
375: soliton was observed in this case at $Z>100$.
376:
377: Typical instabilities of the spinning STS with the spin $s=1$ (in the case
378: when it is unstable) and $s=2$ are illustrated by Figs.~7 through 10. The
379: azimuthal instability breaks the unstable spinning solitons into zero-spin
380: ones, which fly out tangentially relative to the circular crest of the
381: original soliton [similar to what is known about the instability-induced
382: breakup of the (2+1)D spatial vortex solitons \cite{unstable1}]. Thus, the
383: initial internal angular momentum (spin) of the doughnut-shaped spinning
384: soliton is converted into the orbital momentum of the emerging nonspinning
385: fragments.
386:
387: Analyzing a large body of numerical results, we have concluded that the
388: number of the emerging fragments is roughly equal to twice the original spin
389: $s$. The dependence of the number of the fragments on the other parameters
390: is fairly weak.
391:
392: It is noteworthy that, in all the cases displayed in Figs.~7 through 10 (and
393: in many more cases not shown here), the number of the instability-generated
394: fragments is exactly equal to the azimuthal index of the perturbation mode
395: having the largest growth rate. Thus, the full nonlinear evolution of the
396: unstable spinning solitons is in perfect agreement with the stability
397: analysis based on the linearized equations, which was presented in the
398: previous section.
399:
400: \section{Conclusion}
401:
402: In this paper, we have shown that stable bright spatiotemporal spinning
403: solitons (vortex tori), which were recently found in the cubic-quintic model
404: of a dispersive optical medium with competing self-focusing and defocusing
405: nonlinearities \cite{PRL}, are also possible in a model based on the
406: competition between the quadratic and self-defocusing cubic nonlinearities.
407: The solitons are stable, provided that they are broad enough (so that the
408: soliton's energy exceeds a certain critical value, or, in other words, the
409: size of the internal hole is essentially smaller than the overall size of
410: the soliton).
411:
412: In fact, the model with the $\chi ^{(2)}:\chi _{-}^{(3)}$ (quadratic-cubic)
413: nonlinearity may be realized easier in real optical media than the $\chi
414: _{+}^{(3)}:\chi _{-}^{(5)}$ (self-focusing-cubic -- self-defocusing-quintic)
415: one. Possibilities for the experimental implementation of the former model
416: (chiefly, based on the quasi-phase-matching technique) were discussed in
417: Refs. \cite{Sammut,Corney-Bang,Johansen}. Note that such optical media may
418: be used equally well for the experimental generation of both the spatial
419: (2+1)-dimensional solitons (vortex cylinders) considered in Ref. \cite
420: {Sammut} and the 3D spatio-temporal spinning solitons (vortex tori) found in
421: the present work.
422:
423: It is relevant to stress that the amplitude of a beam that can give rise to
424: a stable spinning soliton should not be specifically large: as it is evident
425: from Fig. 2, the necessary power is essentially the same as that which is
426: necessary for the existence of a nonspinning soliton. The difference from
427: the latter case is that the beam generating a stable spinning soliton must
428: be broad (its cross section and temporal width should be large), i.e., its
429: peculiarity is not a large power but rather large total energy.
430:
431: Similar to the cubic-quintic model, only spatiotemporal solitons with spin
432: $s=1$ may be stable in the present system, in contrast with the spatial
433: (2+1)-dimensional solitons, which may be stable in the cases $s=1$ and $s=2$,
434: in models of both types (on the other hand, a difference from the
435: cubic-quintic case is that the existence of STS in the present model is not
436: limited by any energy threshold). These results suggest a conclusion that
437: stable vortex solitons are generic objects, provided that the medium's
438: nonlinearity contains competing elements and the soliton's energy is large
439: enough; in all the known models lacking the nonlinear competition, bright
440: vortex solitons are subject to a strong azimuthal instability.
441:
442: Lastly, one can assume that, very generally speaking, the spinning soliton
443: is not an absolutely stable object, but rather a metastable one. Indeed, the
444: energy of the spinning soliton is larger than that of its zero-spin
445: counterpart, hence it might be possible that a very strong initial
446: perturbation will provoke its rearrangement into a zero-spin soliton, the
447: angular moment being carried away with emitted radiation. In terms of this
448: consideration, it appears that the $s=1$ and $s=0$ solitons are separated by
449: extremely high potential barriers, which make the assumed process
450: practically impossible. To illustrate this point, in Fig. 11 we show the
451: cross sections of the $s=1$ soliton which was very strongly perturbed at the
452: initial point, $Z=0$ (the perturbation is $\approx 30\%$ of the soliton's
453: amplitude), and the result of its evolution at the point $Z=200$. For the
454: same case, the comparison of the distributions of the intensity and phase
455: inside the initial strongly perturbed soliton and the finally established
456: one are shown in Fig. 12 (cf. Fig. 6). As is obvious from Figs. 11 and 12,
457: the soliton was able to completely heal the damage, remaining a truly stable
458: object.
459:
460: \section*{acknowledgments}
461:
462: D. Mihalache, D. Mazilu and L.-C. Crasovan acknowledge support from the
463: Deutsche Forschungsgemeinschaft (DFG) and European Community (Access to
464: Research Infrastructure Action of the Improving Human Potential Program).
465: I.T. and B.A.M. appreciate support from the Binational (US-Israel) Science
466: Foundation through the grant No. 1999459, and a matching support from the
467: Tel Aviv University. L.T. acknowledges support by TIC2000-1010.
468:
469: \newpage
470:
471: \begin{thebibliography}{99}
472: \bibitem{George} G. I. Stegeman, D. N. Christodoulides, and M. Segev, IEEE
473: J. Select. Top. Quant. Electron. {\bf 6}, 1419 (2000).
474:
475: \bibitem{KanRub} A. A. Kanashov and A. M. Rubenchik, Physica D {\bf 4}, 122
476: (1981).
477:
478: \bibitem{MBA} J. T. Manassah, P. L. Baldeck, and R. R. Alfano, Opt. Lett.
479: {\bf 13}, 1090 (1988); J. T. Manassah, {\it ibid.} {\bf 16}, 563 (1991).
480:
481: \bibitem{Yaron} Y. Silberberg, Opt. Lett. {\bf 15}, 1282 (1990).
482:
483: \bibitem{Blagoeva} A. B. Blagoeva {\it et al.}, IEEE J. Quant. Electr. {\bf
484: QE-27}, {\bf \ }2060 (1991).
485:
486: \bibitem{AS} N. Akhmediev and J. M. Soto-Crespo, Phys. Rev. A {\bf 47},
487: 1358 (1993).
488:
489: \bibitem{DarkBullet} Y. Chen and J. Atai, Opt. Lett. {\bf 20}, 133 (1995).
490:
491: \bibitem{collapse} L. Berg{\'{e}}, Phys. Rep. {\bf 303}, 260 (1998).
492:
493: \bibitem{Enns} R. H. Enns {\it et al.}, Opt. Quant. Electron. {\bf 24},
494: S1295 (1992); R. McLeod, K. Wagner, and S. Blair, Phys. Rev. A {\bf 52},
495: 3254 (1995); J. T. Manassah and B. Gross, Laser Phys. {\bf 7}, 9 (1997).
496:
497: \bibitem{Edmundson} D. E. Edmundson, Phys. Rev. E {\bf 55}, 7636 (1997).
498:
499: \bibitem{Skarka} V. Skarka, V. I. Berezhiani, and R. Miklaszewski, Phys.
500: Rev. E {\bf 56}, 1080 (1997).
501:
502: \bibitem{quadr1} K. Hayata and M. Koshiba, Phys. Rev. Lett. {\bf 71}, 3275
503: (1993); B. A. Malomed {\it et al.}, Phys. Rev. E {\bf 56}, 4725 (1997); D.
504: V. Skryabin and W. J. Firth, Opt. Commun. {\bf 148}, 79 (1998); D.
505: Mihalache, D. Mazilu, B. A. Malomed, and L. Torner, {\it ibid.} {\bf 152},
506: 365 (1998).
507:
508: \bibitem{quadr2} D. Mihalache, D. Mazilu, J. D\"{o}rring, and L. Torner,
509: Opt. Commun. {\bf 159}, 129 (1999).
510:
511: \bibitem{quadr3} D. Mihalache, D. Mazilu, B. A. Malomed, and L. Torner,
512: Opt. Commun. {\bf 169}, 341 (1999); D. Mihalache, D. Mazilu, L.-C. Crasovan,
513: L. Torner, B. A. Malomed, and F. Lederer, Phys. Rev. E {\bf 62}, 7340 (2000).
514:
515: \bibitem{Agrawal} S. Raghavan and G. P. Agrawal, Opt. Commun. {\bf 180},
516: 377 (2000).
517:
518: \bibitem{greek} H. E. Nistazakis, D. J. Frantzeskakis, and B. A. Malomed,
519: Phys. Rev. E {\bf 64}, 026604 (2001); C. Polymilis, D. J. Frantzeskakis, A.
520: N. Yannacopoulos, K. Hizanidis, and G. Rowlands, J. OPt. Soc. Am. B {\bf 18}
521: , 75 (2001).
522:
523: \bibitem{Igor} I. V. Mel'nikov, D. Mihalache, and N.-C. Panoiu, Opt.
524: Commun. {\bf 181}, 345 (2000).
525:
526: \bibitem{Miriam} M. Blaauboer, B. A. Malomed, and G. Kurizki, Phys. Rev.
527: Lett. {\bf 84}, 1906 (2000).
528:
529: \bibitem{tandem} L. Torner, S. Carrasco, J. P. Torres, L.-C. Crasovan, and
530: D. Mihalache, Opt. Commun. {\bf 199}, 277 (2001).
531:
532: \bibitem{Wise} X. Liu, L. J. Qian, and F. W. Wise, Phys. Rev. Lett. {\bf 82},
533: 4631 (1999); X. Liu, K. Beckwitt, and F. Wise, Phys. Rev. E {\bf 62}, 1328
534: (2000).
535:
536: \bibitem{vortex} G. A. Swartzlander, Jr., and C. T. Law, Phys. Rev. Lett.
537: {\bf 69}, 2503 (1992); A. W. Snyder, L. Poladian, and D. J. Mitchell, Opt.
538: Lett. {\bf 17}, 789 (1992).
539:
540: \bibitem{DiTrapp} P. Di Trapani {\it et al.}, Phys. Rev. Lett. {\bf 84},
541: 3843 (2000).
542:
543: \bibitem{Sammut} I. Towers, A.V. Buryak, R.A. Sammut, and B.A. Malomed,
544: Phys. Rev. E {\bf 63}, 055601(R) (2001).
545:
546: \bibitem{2DvortexYuri} A. S. Desyatnikov and Yu. S. Kivshar, Phys. Rev.
547: Lett. {\bf 87}, 033901 (2001).
548:
549: \bibitem{2DvortexSegev} Z. H. Musslimani, M. Soljacic, M. Segev, and D. N.
550: Christodoulides, Phys. Rev. E {\bf 63}, 066608 (2001).
551:
552: \bibitem{Anton} A. Desyatnikov, A. Maimistov, and B. Malomed, Phys. Rev. E
553: {\bf 61}, 3107 (2000).
554:
555: \bibitem{new} D. Mihalache, D. Mazilu, L.-C. Crasovan, B. A. Malomed, and
556: F. Lederer, Phys. Rev. E {\bf 62}, R1505 (2000).
557:
558: \bibitem{unstable1} L. Torner and D. V. Petrov, Electr. Lett. {\bf 33}, 608
559: (1997); W. J. Firth and D. V. Skryabin, Phys. Rev. Lett. {\bf 79}, 2450
560: (1997); J. P. Torres, J. M. Soto-Crespo, L. Torner, and D. V. Petrov, J.
561: Opt. Soc. Am. B {\bf 15}, 625 (1998); D. V. Skryabin and W. J. Firth, Phys.
562: Rev. E {\bf 58}, 3916 (1998).
563:
564: \bibitem{experiment} D. V. Petrov {\it et al.}, Opt. Lett. {\bf 23}, 1444
565: (1998).
566:
567: \bibitem{comp1} A. V. Buryak, Yu. S. Kivshar, and S. Trillo, Opt. Lett.
568: {\bf 20}, 1961 (1995), M. A. Karpierz, Opt. Lett. {\bf 20}, 1677 (1995).
569:
570: \bibitem{comp2} O. Bang, J. Opt. Soc. Am. B {\bf 14}, 51 (1997); O. Bang,
571: Yu. S. Kivshar, and A. V. Buryak, Opt. Lett. {\bf 22}, 1680 (1997); L.
572: Berge, O. Bang, J. Juul Rasmussen, and V. K. Mezentsev, Phys. Rev. E {\bf 55},
573: 555 (1997); O. Bang, Yu. S. Kivshar, A. V. Buryak, A. De Rossi, and S.
574: Trillo, Phys. Rev. E {\bf 58}, 5057 (1998).
575:
576: \bibitem{Abdullaev} F. Kh. Abdullaev, A. Gammal, L. Tomio, and T.
577: Frederico, Phys. Rev. A {\bf 63}, 043604 (2001); A. Gammal, T. Frederico, L.
578: Tomio, and F. Kh. Abdullaev, Phys. Lett. A {\bf 267}, 305 (2001).
579:
580: \bibitem{plasma} C. T. Zhou and X. T. He, Physica Scripta, {\bf 50}, 415
581: (1994).
582:
583: \bibitem{Q} M. Quiroga-Teixeiro and H. Michinel, J. Opt. Soc. Am. B {\bf 14},
584: 2004 (1997).
585:
586: \bibitem{IsaacPLA} I. Towers {\it et al.}, Phys. Lett. A {\bf 288}, 292
587: (2001); B. A. Malomed, L.-C. Crasovan, D. Mihalache, Physica D {\bf 161},
588: 187 (2002); R. L. Pego and H. A. Warchall, Los Alamos e-print archive:
589: nlin.PS/0108009.
590:
591: \bibitem{Berezhiani} V. I. Berezhiani, V. Skarka, and N. B. Aleksic, Phys.
592: Rev. E {\bf 64}, 057601 (2001).
593:
594: \bibitem{Michinel} H. Michinel {\it et al.} J. Opt. B: Quantum Semiclass.
595: Opt. {\bf 3}, 314 (2001).
596:
597: \bibitem{reviews} D. O. Riska, Adv. Nucl. Phys. {\bf 22}, 1 (1996); T.
598: Gisiger and M. B. Paranjape, Phys. Rep. {\bf 306}, 110 (1998).
599:
600: \bibitem{PRL} D. Mihalache, D. Mazilu, L.-C. Crasovan, I. Towers, A. V.
601: Buryak, B. A. Malomed, L. Torner, J. P. Torres, and F. Lederer, Phys. Rev.
602: Lett. {\bf 88}, 073902 (2002).
603:
604: \bibitem{new2} D. Mihalache, D. Mazilu, L.-C. Crasovan, B. A. Malomed, and
605: F. Lederer, Phys. Rev. E {\bf 61}, 7142 (2000).
606:
607: \bibitem{walking} L. Torner, D. Mazilu, and D. Mihalache, Phys. Rev. Lett.
608: {\bf 77}, 2455 (1996); C. Etrich, U. Peschel, F. Lederer, and B. A. Malomed,
609: Phys. Rev. E {\bf 55}, 6155 (1997).
610:
611: \bibitem{PO} C. Etrich, F. Lederer, B. A. Malomed, T. Peschel, and U.
612: Peschel, Progress in Optics {\bf 41}, 483 (2000).
613:
614: \bibitem{AA} N. N. Akhmediev and A. Ankiewicz, {\it Solitons, Nonlinear
615: Pulses and Beams} (Chapman and Hall, London, 1997).
616:
617: \bibitem{D-B} {\ G. Dahlquist and {\AA }. Bj{\"{o}}rk, {\it Numerical
618: Methods}, (Prentice Hall, Englewood Cliffs, 1974). }
619:
620: \bibitem{SHT} G. I. Stegeman, D. J. Hagan, and L. Torner, Opt. Quantum
621: Electron. {\bf 28}, 1691 (1996).
622:
623: \bibitem{TS} L. Torner and A. P. Sukhorukov, Opt.\ Photonics News {\bf 13},
624: 26 (2002).
625:
626: \bibitem{Yuri} Yu. S. Kivshar, in {\it Advanced Photonics with Second-Order
627: Optically Nonlinear Processes}, edited by A. D. Boardman, L. Pavlov, and S.
628: Panev (Kluwer Academic, Dordrecht,1998), p. 451.
629:
630: \bibitem{VK} M. G. Vakhitov and A.A. Kolokolov, Radiophys. Quantum El. {\bf
631: 16}, 783 (1973).
632:
633: \bibitem{Akh} J. M. Soto-Crespo, D. R. Heatley, E. M. Wright, and N. N.
634: Akhmediev, Phys. Rev. A {\bf 44}, 636 (1991); J. Atai, Y. Chen, and J. M.
635: Soto-Crespo, Phys. Rev. A {\bf 49}, R3170 (1994).
636:
637: \bibitem{osc1} N. N. Akhmediev, A. Ankiewicz, and H. T. Tran, J. Opt. Soc.
638: Am. B {\bf 10}, 230 (1993).
639:
640: \bibitem{osc2} B. A. Malomed and R. S. Tasgal, Phys. Rev. E {\bf 49}, 5787
641: (1994); I. V. Barashenkov, D. E. Pelinovsky, and E. V. Zemlyanaya, Phys.
642: Rev. Lett. {\bf 80}, 5117 (1998); A. De Rossi, C. Conti, and S. Trillo,
643: Phys. Rev. Lett. {\bf 81}, 85 (1998); J. Sch\"{o}llmann, R. Scheibenzuber,
644: A. S. Kovalev, A. P. Mayer, and A. A. Maradudin, Phys. Rev. E {\bf 59}, 4618
645: (1999).
646:
647: \bibitem{osc3} D. Mihalache, D. Mazilu, and L. Torner, Phys. Rev. Lett.
648: {\bf 81}, 4353 (1998); D. Mihalache, D. Mazilu, and L.-C. Crasovan, Phys.
649: Rev. E {\bf 60}, 7504 (1999).
650:
651: \bibitem{Ortega} J. M. Ortega, W. C. Rheinboldt, {\it Iterative Solution of
652: Nonlinear Equations in Several Variables}, (Academic Press, New York, 1970),
653: p. 182.
654:
655: \bibitem{Corney-Bang} J. F. Corney and O. Bang, Phys. Rev. E {\bf 64},
656: 047601 (2001).
657:
658: \bibitem{Johansen} S. K. Johansen, S. Carrasco, L. Torner, O. Bang, Opt.
659: Commun. {\bf 203}, 393 (2002).
660: \end{thebibliography}
661:
662: \newpage
663:
664: \section*{Figure Captions}
665:
666: Fig. 1. Domains of the existence and stability of spinning STS with spin
667: $s=1$. The upper continuous curve is the existence border, corresponding to
668: infinitely broad (in fact, dark) solitons.
669:
670: Fig. 2. Typical shapes of stable STS with $s=1$ for $E=12000$: (a)
671: $\beta=-0.1$, (b) $\beta=0$, and (c) $\beta=0.2$. The labels FF and SH
672: pertain to the fundamental-frequency and second-harmonic components of the
673: soliton.
674:
675: Fig. 3. The propagation constant $\kappa$ (a) and Hamiltonian $H$ (b) of the
676: three-dimensional solitons, with different values of spin, vs. their energy
677: $E$, in the case of zero phase mismatch, $\beta=0$.
678:
679: Fig. 4. The growth rate of perturbations, ${\rm Re}~ \lambda$, corresponding
680: to different values of the azimuthal index $n$ (indicated by labels near the
681: curves) vs. the soliton's wave number $\kappa$: (a) $s=1$; (b) $s=2$. The
682: imaginary part of the stability eigenvalue, ${\rm Im}~ \lambda$,
683: corresponding to different values of the azimuthal index $n$ (indicated by
684: labels near the curves) vs. the soliton's wave number $\kappa$: (c) $s=1$;
685: (d) $s=2$. Here and in the following plots, $\beta=0$. We stress that, in
686: the case $s=1$, the instability growth rate vanishes at the point
687: $\kappa=\kappa_{{\rm st}}$, see the text, while in the case $s=2$ the growth
688: rate corresponding to $n=2$ remains positive up to the border of the
689: existence region of the solitons. This border is marked in all the panels by
690: vertical arrows.
691:
692: Fig. 5. Evolution of the energy components $E_u$ and $E_v $ of the soliton
693: with $s=1$, as generated by an input configuration in the form of a Gaussian
694: with a nested vortex. Here, the input total energy is $E=5986$.
695:
696: Fig. 6. The formation of the soliton with spin $s=1$ in the same case which
697: corresponds to Fig. 5, shown in terms of the cross section of the fields at
698: $T=0$: (a) the intensity distribution in the initial Gaussian with a nested
699: vortex; (b) its phase field; (c) the intensity distribution of the spinning
700: soliton at $Z=100$; (d) the phase field at $Z=100$.
701:
702: Fig. 7. Isosurface plots illustrating the fragmentation of the $s=1$ soliton
703: with $\kappa=0.01$ into zero-spin ones as a result of the azimuthal
704: instability: (a) $Z=0$; (b) $Z=1000$.
705:
706: Fig. 8. The same as in Fig. 7 in the case $\kappa =0.032$: (a) $Z=0$; (b)
707: $Z=1140$.
708:
709: Fig. 9. The same as in Figs. 7 and 8 in the case of the $s=2$ initial
710: soliton with $\kappa=0.015$: (a) $Z=0$; (b) $Z=900$.
711:
712: Fig. 10. The same as in Fig. 9 in the case $\kappa =0.04$: (a) $Z=0$; (b)
713: $Z=2100$.
714:
715: Fig. 11. Cross sections of an $s=1$ soliton that was strongly perturbed at
716: $Z=0$, and the result of its evolution after having passed the propagation
717: distance $Z=200$.
718:
719: Fig. 12. The recovery of the soliton with spin $s=1$ in the same case as in
720: Fig. 11, shown in terms of the cross section of the fields at $T=0$ (cf.
721: Fig. 6): (a) the intensity distribution in the initial strongly perturbed
722: soliton; (b) its phase field; (c) the intensity distribution of the
723: self-cleared soliton at $Z=100$; (d) the phase field at $Z=100$.
724:
725: \end{document}
726: