1:
2: \documentstyle[prl,aps,epsfig]{revtex} % 2 col mode
3: \begin{document}
4: \draft % preprint mode
5: % 2 col mode:
6: \twocolumn[\hsize\textwidth\columnwidth\hsize\csname @twocolumnfalse\endcsname
7:
8:
9: \title{Spatial Semiconductor-Resonator Solitons}
10:
11: \author{V. B. Taranenko, C. O. Weiss}
12:
13: \address{Physikalisch-Technische Bundesanstalt \\
14: 38116 Braunschweig, Germany}
15:
16: %\date{\today{}}
17: \maketitle
18:
19: \begin{abstract}
20: We demonstrate experimentally and numerically the existence of
21: spatial solitons in multiple-quantum-well semiconductor
22: microresonators driven by an external coherent optical field. We
23: discuss stability of the semiconductor-resonator solitons over a
24: wide spectral range around the band edge. We demonstrate the
25: manipulation of such solitons: switching solitons on and off by
26: coherent as well as incoherent light; reducing the light power
27: necessary to sustain and switch a soliton, by optical pumping.
28: \end{abstract}
29: \pacs{PACS 42.65.Sf, 42.65.Pc, 47.54.+r} \vskip1pc ]
30: % BEGIN TEXT HERE
31: \section{Introduction}
32: %
33: Spatial resonator solitons theoretically predicted in
34: \cite{tag:1,tag:2,tag:3,tag:4,tag:5} can exist in a variety of
35: nonlinear resonators, such as lasers (vortices), lasers with
36: saturable absorber (bright solitons), parametric oscillators
37: (phase solitons), driven nonlinear resonators (bright/dark
38: solitons). Such resonator solitons can be viewed as self-trapped
39: domains of one field state surrounded by another state of the
40: field. The two different field states can be a high and a low
41: field (bright/dark solitons), positive and negative field (phase
42: solitons), right-hand and left-hand polarized field (polarization
43: solitons) or high and zero field (vortices).
44:
45: Bright solitons in laser resonators with saturable absorber were
46: initially shown to exist in \cite{tag:6}, which was limited to
47: single stationary solitons. Existence of moving solitons and
48: simultaneous existence of large numbers of stationary solitons
49: was shown in \cite{tag:7,tag:8}. Phase solitons in degenerate
50: parametric wave mixing resonators were predicted in \cite{tag:9}
51: and demonstrated in \cite{tag:10}. Theoretically it was shown in
52: \cite{tag:11} that not only 2D spatial resonator solitons exist
53: but that also in 3D such structures can be stable, linking the
54: field of optical solitons with elementary particle physics
55: \cite{tag:12}. Vortex solitons differ from other soliton types in
56: that they possess \emph{structural stability} in addition to
57: \emph{dynamical stability}, the only stabilizing mechanism of the
58: other solitons. The existence of vortices in lasers initially
59: shown in \cite{tag:13} and later also the existence of vortex
60: solitons \cite{tag:14}.
61:
62: Since the resonator solitons are bistable and can be moved around
63: they are suited to carry information. Information can be written
64: in the form of a spatial soliton, somewhere, and then transported
65: around at will; finally being read out somewhere else; possibly
66: in conjunction with other solitons. In this respect the spatial
67: resonator solitons have no counterpart in any other kind of
68: information-carrying elements and lend themselves therefore to
69: operations not feasible with conventional electronic means, such
70: as an all-optical pipeline storage register ("photon buffer") or
71: even processing in the form of cellular automata. Experiments on
72: the manipulation of bright solitons as required for such
73: processing tasks were first carried out on a slow system: laser
74: with (slow) saturable absorber. In particular it was demonstrated
75: how to write and erase solitons and how to move them or localize
76: them. For reviews see \cite{tag:15,tag:16}.
77:
78: \begin{figure}[htbf] \epsfxsize=65mm
79: \centerline{\epsfbox{Fig01.eps}} \vspace{0.7cm}
80: \caption{Schematic of semiconductor microresonator consisting of
81: two plane distributed Bragg reflectors (DBR) and multiple quantum
82: wells (MQW).}
83: \end{figure}
84:
85: In order to be applicable to technical tasks it is mandatory to
86: operate in fast, miniaturized systems. For compatibility and
87: integrability with other information processing equipment it is
88: desirable to use semiconductor systems. We chose the
89: semiconductor microresonator structure (Fig.~1) as commonly used
90: for vertical cavity surface emitting lasers (VCSEL) \cite{tag:17}
91: consisting of multiple quantum well (MQW) structure sandwiched
92: between distributed Bragg reflectors (DBR).
93:
94: The resonator length of such a structure is $\sim$ $\lambda$
95: while the transverse size is typically 5 cm. Therefore such short
96: length and wide area microresonator permits only one longitudinal
97: mode (Fig.~2~(a)) and an enormous number of transverse modes that
98: allow a very large number of spatial solitons to coexist. The
99: resonator is obviously of the plane mirror type, implying
100: frequency degeneracy of all transverse modes and thus allowing
101: arbitrary field patterns to be resonant inside the resonator.
102: This is another prerequirement for existence and manipulability
103: of spatial solitons.
104:
105: The resonator soliton existence is closely linked with the plane
106: wave resonator bistability (Fig.~2~(b)) caused by longitudinal
107: nonlinear effects: the nonlinear changes of the resonator length
108: (due to nonlinear refraction changes) and finesse (due to
109: nonlinear absorption changes) \cite{tag:18}. The longitudinal
110: nonlinear effects combined with transverse nonlinear effects
111: (such as self-focusing) can balance diffraction and form resonator
112: soliton. Generally these nonlinear effects can cooperate or act
113: oppositely, with the consequence of reduced soliton stability in
114: the later case.
115:
116: \begin{figure}[htbf]
117: \epsfxsize=85mm \centerline{\epsfbox{Fig02.eps}} \vspace{0.7cm}
118: \caption{Semiconductor microresonator reflectance spectrum (a)
119: and typical bistability loop in reflection (b). Arrows mark the
120: driving field that is detuned from the resonator resonance and
121: the pump field that is tuned to be coupled into resonator through
122: one of short-wavelength interference notches of the resonator
123: reflectance spectrum.}
124: \end{figure}
125:
126: In the present paper we demonstrate experimentally and
127: numerically existence of bright and dark spatial solitons as well
128: as extended hexagonal patterns in MQW-semiconductor
129: microresonators at room temperature. We discuss stability of the
130: semiconductor-resonator solitons over a wide spectral range
131: around the band edge. We demonstrate the manipulation of such
132: solitons in view of technical application: switching solitons on
133: and off by coherent as well as incoherent light; reducing the
134: light power necessary to sustain and switch a soliton, by optical
135: pumping.
136:
137: \section{Model and numerical analysis}
138: %
139: We consider phenomenological model of a driven wide area
140: MQW-semiconductor microresonator similar to \cite{tag:19,tag:20}.
141: The optical field $E$ inside the resonator is described in
142: mean-field approximation \cite{tag:21}. The driving incident field
143: {$E_{\rm in}$} assumed to be a stationary plane wave. Nonlinear
144: absorption and refractive index changes induced by the
145: intracavity field in the vicinity of the MQW-structure band edge
146: are assumed to be proportional to the carrier density $N$
147: (normalized to the saturation carrier density). The equation of
148: motion for $N$ involves nonresonant (to avoid saturation effect)
149: pumping $P$, carrier recombination and diffusion. The resulting
150: coupled equations describing the spatio-temporal dynamics of $E$
151: and $N$ have the form:
152: \begin{eqnarray}
153: \cr {\partial E}/{\partial t}=E_{\rm in}-\sqrt{T}E\{[1+C{\rm
154: Im}(\alpha)(1-N)]+\cr{+i(\theta-C{\rm
155: Re}(\alpha)N-\nabla^{2}_{\bot})\}}\,, \cr\cr{\partial N}/{\partial
156: t}=P-\gamma[N-|E|^2(1-N)-d\nabla^{2}_{\bot}N]\,\,,
157: \end{eqnarray}
158: where $C$ is the saturable absorption scaled to the resonator DBR
159: transmission $T$ ($T$ is assumed to be small since the DBR
160: reflectivity is typically $\geq$ 0.995). ${\rm Im}(\alpha)(1-N)$
161: and ${\rm Re}(\alpha)N$ describe the absorptive and refractive
162: nonlinearities, respectively, $\theta$ is the detuning of the
163: driving field from the resonator resonance, $\gamma$ is the photon
164: lifetime in the resonator scaled to the carrier recombination
165: time, $d$ is the diffusion coefficient scaled to the diffraction
166: coefficient and $\nabla^{2}_{\bot}$ is the transverse Laplacian.
167:
168: Linear effects in the resonator are spreading of light by
169: diffraction and diffusion (terms with $\nabla^{2}_{\bot}$ in
170: (1)). The material nonlinearity that can balance this linear
171: spreading can do this in various ways. It has a real (refractive)
172: and imaginary (dissipative) part and can act longitudinally and
173: transversely. The nonlinear changes of the resonator finesse (due
174: to nonlinear absorption change) and length (due to nonlinear
175: refractive index change) constitute longitudinal nonlinear
176: effects, also known under the name \emph{nonlinear resonance}
177: \cite{tag:22}. The transverse effect of the nonlinear refractive
178: index can be self-focusing (favorable for bright and unfavorable
179: for dark solitons) and self-defocusing (favorable for dark and
180: unfavorable for bright solitons). Absorption (or gain) saturation
181: (bleaching) leading to nonlinear gain guiding in laser parlance,
182: a transverse effect. Longitudinal and transverse effects can work
183: oppositely, or cooperate.
184:
185: There are two main external control parameters: the driving field
186: intensity $|E|^2$ and the resonator detuning $\theta$. Then for
187: driving intensities not quite sufficient for reaching the
188: resonance condition for the whole resonator area, the system
189: "chooses" to distribute the light intensity in the resonator in
190: isolated spots where the intensity is then high/low enough to
191: reach the resonance condition thus forming bright/dark patterns.
192: Instead of saying "the system chooses" one would more
193: mathematically express this by describing it as a modulational
194: instability (MI). The detuned plane wave field without spatial
195: structure with intensity insufficient to reach the resonance
196: condition is unstable against structured solutions. According to
197: our numerical solutions of (1) a large number of such structured
198: solutions coexist and are stable (see e.g. patterns in Fig.~7).
199:
200: Fig.~3 shows the existence domains (in coordinates $\theta$ -
201: $|E|^2$) of MI, dark spatial solitons and plane-wave bistability
202: calculated for the case of purely dispersive (defocusing)
203: nonlinearity. The dark soliton structure (inset in Fig.~3) can be
204: interpreted as a small circular switching front. A switching
205: front connects two stable states: the high transmission and the
206: low transmission state. Such a front can in 2D surround a domain
207: of one state. When this domain is comparable in diameter to the
208: "thickness" of the front, then each piece of the front interacts
209: with the piece on the opposite side of the circular small domain,
210: which can lead, particularly if the system is not far from a
211: modulational instability (see Fig.~3), to a stabilisation of the
212: diameter of the small domain. In which case the small domain
213: becomes an isolated self-trapped structure or a dissipative
214: resonator soliton.
215:
216: \begin{figure}[htbf]
217: \epsfxsize=60mm \centerline{\epsfbox{Fig03.eps}} \vspace{0.9cm}
218: \caption{Numerical solutions of (1) for unpumped ($P=0$) and
219: dispersive/defocusing (${\rm Im}(\alpha)=0$) case. Area limited
220: by dashed lines is optical bistability domain for plane waves.
221: Grey shaded area is modulational instability domain. Dark shaded
222: area is domain of stability for dark solitons. Inset is dark
223: soliton in 3D representation. Parameters: $C=10$, $T=0.005$,
224: ${\rm Re}(\alpha)=-1$, $d=0.01$.}
225: \end{figure}
226:
227: \section{Experimental technique}
228: %
229: Fig.~4 shows the optical arrangement for the
230: semiconductor-resonator soliton experiments, and in particular for
231: their switching on or off. The semiconductor microresonators
232: consisting of MQW (GaAs/AlGaAs or GaInAs/GaPAs) structures
233: sandwiched between high-reflectivity ($\geq$ 0.995) DBRs (Fig.~1)
234: operate at room temperature. The microresonator structures were
235: grown on the GaAs substrates by molecular beam epitaxy technique
236: that makes possible high quality MQW structures with small radial
237: layer thickness variation. The best sample tested in our
238: experiments has only $\sim$ 0.3 nm/mm variation of the resonator
239: resonance wavelength over the sample cross section.
240:
241: The driving light beam was generated by either a tunable (in the
242: range 750-950 nm) Ti:Sa laser or a single-mode laser diode, both
243: operating in continuous wave regime. For experimental convenience
244: to limit thermal effects we perform the whole experiments within
245: a few microseconds, by admitting the light through an
246: acousto-optical modulator. The laser beam of suitable wavelength
247: is focused on the microresonator surface in the light spot of
248: $\sim$ 50 $\mu$m in diameter, thus providing a quite large Fresnel
249: number ($\geq$ 100).
250:
251: \begin{figure}[htbf]
252: \epsfxsize=85mm \centerline{\epsfbox{Fig04.eps}} \vspace{0.9cm}
253: \caption{Experimental setup. Laser: Ti:Sa (or diode) laser, AOM:
254: acousto-optic modulator, $\lambda$/2: halfwave plate, PBS:
255: polarization beam spliters, EOM: electro-optical amplitude
256: modulators, BE: beam expander, PZT: piezo-electric transducer, P:
257: polarizer, L: lenses, BS: beam splitters, PD: photodiode. }
258: \end{figure}
259:
260: Part of laser light is split away from the driving beam and then
261: superimposed with the main beam by a Mach-Zehnder interferometer
262: arrangement, to serve as the address beam. The address beam is
263: sharply focused and directed to some particular location in the
264: illuminated. The switching light is opened only for a few
265: nanoseconds using an electro-optic modulator. For the case of
266: incoherent switching the polarization of the address beam is
267: perpendicular to that of the main beam to avoid interference. For
268: the case of coherent switching the polarizations should be
269: parallel and a phase control of the switching field is thus
270: always needed: for switching on as well as switching off a
271: soliton. One of the interferometer mirrors can be moved by a
272: piezo-electric element to control the phase difference between
273: the driving light and the address light.
274:
275: Optical pumping of the MQW-structures was done by a multi-mode
276: laser diode or a single-mode Ti:Sa laser. To couple the pump
277: light into the microresonator the laser wavelength was tuned into
278: the high transmission spectral window of the microresonator
279: reflectance spectrum as shown in Fig.~2~(a).
280:
281: The observation is done in reflection by a CCD camera combined
282: with a fast shutter (another electro-optic modulator), which
283: permits to take nanosecond snapshots at a given time of the
284: illuminated area on the resonator sample. Recording movies on
285: this nanosecond time scale is also possible. To follow intensity
286: in time in certain points (e.g. at the location of a soliton) a
287: fast photodiode can be imaged onto arbitrary locations within the
288: illuminated area.
289:
290: \section{Experimental results and discussions}
291: %
292: To find the most stable resonator solitons for applications one
293: can play with the nonlinear (absorptive/dispersive) response by
294: choice of the driving field wavelength, with the resonator
295: detuning, and finally with the carrier population inversion by
296: the pumping. We recall that all nonlinearities change their sign
297: at transparency i.e. at the point where in the valence- and the
298: conduction band populations are equal. Going from below
299: transparency (absorption) to above transparency (population
300: inversion, producing light amplification), nonlinear absorption
301: changes to nonlinear gain, self-focusing changes to
302: self-defocusing and vice versa, and decrease of resonator length
303: with intensity changes to increase (and vice versa). The
304: population of the bands can be controlled by pumping ($P$ in (1))
305: i.e. transferring electrons from the valence band to the
306: conduction band. This can be done by optical excitation
307: \cite{tag:23}, with radiation of wavelength shorter than the band
308: edge wavelength or - if the structure is suited to support
309: electrical currents (i.e. if it is a real VCSEL-structure) - by
310: electrical excitation.
311:
312: \subsection{Below bandgap hexagons and dark solitons}
313: %
314: Working well bellow the bandgap when the driving field wavelength
315: is $\sim$ 30 nm longer than the band edge wavelength we observed
316: spontaneous formation of hexagonal patterns (Fig.~5). The hexagon
317: period scales linearly with $\theta^{-1/2}$ \cite{tag:24}
318: indicating that they are formed by the tilted-wave mechanism
319: \cite{tag:25} that is the basic mechanism for resonator hexagon
320: formation \cite{tag:26}. Dark-spot hexagon (Fig.~5~(a)) converts
321: in to bright-spot hexagon (Fig.~5~(b)) when the driving intensity
322: increases. Experiment shows that individual spots of these
323: patterns can not be switched on/off independently from other
324: spots as it is expected for strongly correlated spot structure
325: (or coherent hexagons).
326:
327: \begin{figure}[htbf]
328: \epsfxsize=50mm \centerline{\epsfbox{Fig05.eps}} \vspace{0.7cm}
329: \caption{(a) Bright (dark in reflection) and (b) dark (bright in
330: reflection) hexagonal patterns for the dispersive/defocusing
331: case. Driving light intensity incrieses from (a) to (b). }
332: \end{figure}
333:
334: However when we further increased the driving intensity we found
335: that the bright spots in such hexagonal patterns can be switched
336: independently by the addressed focused optical (incoherent) pulses
337: \cite{tag:24}. Fig.~6 shows the experimental results. Fig.~6~(a)
338: shows the hexagonal pattern formed. The focused light pulse can
339: be aimed at individual bright spots such as the ones marked "1"
340: or "2". Fig.~6~(b) shows that after the switching pulse aimed at
341: "1" spot "1" is off. Fig.~6~(c) shows the same for spot "2". We
342: remark that in these experiments we speak of true logic
343: switching: the spots remain switched off after the switching
344: pulse (if the energy of the pulse is sufficient, otherwise the
345: bright spot reappears after the switching pulse). These
346: observations of local switching indicate that these hexagonal
347: patterns are not coherent pattern: the individual spots are
348: rather independent, even at this dense packing where the spot
349: distance is about the spot size.
350:
351: \begin{figure}[htbf]
352: \epsfxsize=70mm \centerline{\epsfbox{Fig06.eps}} \vspace{0.7cm}
353: \caption{Switching-off of individual spots of incoherent
354: hexagonal structure with addressed pulses located at different
355: places (marked 1 and 2) of the pattern. }
356: \end{figure}
357:
358: These experimental findings can be understood in the frame of the
359: model (1) \cite{tag:27}. Fig.~7 shows the bistable plane wave
360: characteristic of the semiconductor resonator for conditions
361: roughly corresponding to the experimental conditions. At the
362: intensities marked (a) to (d) patterned solutions exist.
363:
364: \begin{figure}[htbf]
365: \epsfxsize=60mm \centerline{\epsfbox{Fig07.eps}} \vspace{0.7cm}
366: \caption{Numerical solution of (1) for intracavity light
367: intensity as function of incident intensity: homogeneous solution
368: (dashed line marks modulationaly unstable part of the curve) and
369: patterns (a-d). Shaded area marks existence range for dark-spot
370: hexagons. Parameters: $C=10$, $T=0.005$, ${\rm Im}(\alpha)=0.1$,
371: ${\rm Re}(\alpha)=-1$, $\theta=-10.3$, $d=0.01$. }
372: \end{figure}
373:
374: The pattern period in Fig.~7~(a) corresponds precisely to the
375: detuning in the following way. When the driving field is detuned,
376: the resonance condition of the resonator cannot be fulfilled by
377: plane waves travelling exactly perpendicularly to the mirror
378: plane. However, the resonance condition can be fulfilled if the
379: wave plane is somewhat inclined with respect to the mirror plane
380: (the tilted wave solution \cite{tag:25}). The nonlinear system
381: chooses therefore to support resonant, tilted waves. Fig.~7~(a) is
382: precisely the superposition of 6 tilted waves that support each
383: other by (nonlinear) 4-wave-mixing. The pattern period
384: corresponds to the resonator detuning as in the experiment for
385: structures Fig.~5~(a). Thus the pattern formation in Fig.~7~(a) is
386: mostly a linear process. In this pattern the bright spots are not
387: independent. Individual spots cannot be switched as in the
388: experiment Fig.~5.
389:
390: On the high intensity pattern Fig.~7~(d) the pattern period is
391: remarkably different from Fig.~7~(a) even though the detuning is
392: the same. This is indication that the internal detuning is
393: smaller and means that the resonator length is nonlinearly
394: changed by the intensity-dependent refractive index (the nonlinear
395: resonance). From the ratio of the pattern periods of Fig.~7~(a)
396: and (d) one sees that the nonlinear change of detuning is about
397: half of the external detuning. That means the nonlinear detuning
398: is by no means a small effect. This in turn indicates that by
399: spatial variation of the resonator field intensity the detuning
400: can vary substantially in the resonator cross section. In other
401: words, the resonator has at the higher intensity a rather wide
402: freedom to (self-consistently) arrange its field structure. One
403: can expect that this would allow a large number of possible
404: stable patterns between which the system can choose.
405:
406: \begin{figure}[htbf]
407: \epsfxsize=70mm \centerline{\epsfbox{Fig08.eps}} \vspace{0.7cm}
408: \caption{Stable hexagonal arrangements of dark patial solitons:
409: (a) without defects, (b) with single-soliton defect, (c) with
410: triple-soliton defect.}
411: \end{figure}
412:
413:
414: Fig.~8 shows that the high intensity conditions of Fig.~7~(d)
415: allow reproducing the experimental findings on switching
416: individual bright spots. Fig.~8~(a) is the regular hexagonal
417: pattern, Fig.~8~(b) shows one bright spot switched off as a
418: stable solution and Fig.~8~(c) shows a triple of bright spots
419: switched off as a stable solution, just as observed in the
420: experiments \cite{tag:27}.
421:
422: Thus while Fig.~7~(a) is a completely coherent space filling
423: pattern, Fig.~7~(d) is really a cluster of (densest packed)
424: individual dark solitons. The increase of intensity from (a) to
425: (d) allows the transition from the extended patterns to the
426: localized structures by the increased nonlinearity, which gives
427: the system an additional internal degree of freedom. We note that
428: the transition from the coherent low intensity pattern to the
429: incoherent higher intensity structure proceeds through
430: stripe-patterns as shown in Fig.~7~(b) \cite{tag:27}. For the
431: intensity of Fig.~7~(c) the individual spots are still not
432: independent as in the experiment Fig.~5~(b).
433:
434: \subsection{Near bandgap bright and dark solitons}
435: %
436: Working at wavelengths close to the band edge we found both
437: bright and dark solitons, as well as collections of several of
438: the bright and dark solitons, several solitons existing at the
439: same time (Figs~9,~10). Fig.~10 demonstrates that shape and size
440: of bright spots are independent on the shape and intensity of the
441: driving beam that is a distinctive feature of spatial solitons.
442:
443: \begin{figure}[htbf]
444: \epsfxsize=65mm \centerline{\epsfbox{Fig09.eps}} \vspace{0.7cm}
445: \caption{Solitons in the semiconductor microresonator: (a)
446: switched area without solitons (for completeness); (b) dark
447: soliton in switched area; (c) bright soliton on unswitched
448: background; (d) 2 bright solitons; (e), (f) 2,3 dark solitons.}
449: \end{figure}
450:
451: \begin{figure}[htbf]
452: \epsfxsize=75mm \centerline{\epsfbox{Fig10.eps}} \vspace{0.7cm}
453: \caption{Switched domain (a) and dark solitons (b)-(d) in an
454: illuminating area of elliptical shape.}
455: \end{figure}
456:
457: Nonlinearity of the MQW structure near the band edge is
458: predominantly absorptive. Therefore in the first approximation we
459: can neglect the refractive part of the complex nonlinearity in
460: the model equations (1) and describe the nonlinear structure as a
461: saturable absorber. Numerical simulations for this case (Fig.~11)
462: confirm existence of both bright and dark resonator solitons as
463: they are observed in the experiment (Figs~9,~10). We can contrast
464: these resonator spatial solitons with the propagating (in a bulk
465: nonlinear material) spatial solitons: the latter can not be
466: supported by a saturable absorber.
467:
468: \begin{figure}[htbf]
469: \epsfxsize=60mm \centerline{\epsfbox{Fig11.eps}} \vspace{0.7cm}
470: \caption{Numerical solutions of (1) for unpumped ($P=0$) and
471: absorptive (${\rm Re}(\alpha)=0$) case. Area limited by dashed
472: lines is optical bistability domain for plane waves. Shaded areas
473: are domains of existence of bright (grey area) and dark (black
474: area) solitons. Insets are bright (left) and dark (right)
475: soliton. Parameters: $C=20$, $T=0.005$, ${\rm Im}(\alpha)=1$,
476: $d=0.01$.}
477: \end{figure}
478:
479: Fig.~12~(c) and (d) show details of the spontaneous formation of
480: the bright soliton (as in Fig.~9~(c)). As discussed in
481: \cite{tag:28} temperature effects lead in this case to a slow
482: formation of solitons, associated with the shift of the band edge
483: by temperature \cite{tag:29}. Fig.~12~(c) gives the incident and
484: reflected intensity at the location of the bright soliton and
485: Fig.~12~(d) gives the reflectivity on a diameter of the
486: illuminated area. At 1.3 $\mu$s (arrow) the resonator switches to
487: high transmission (low reflection). The switched area then
488: contracts relatively slowly to the stable structure Fig.~12~(d),
489: which is existing after $t$ $\approx$ 3.0 $\mu$s.
490:
491: After the resonator switches to low reflection its internal field
492: and with it the dissipation is high. A rising temperature
493: decreases the band gap energy \cite{tag:29} and therefore shifts
494: the bistable resonator characteristic towards higher intensity.
495: Thus the basin of attraction for solitons which is located near
496: the plane wave switch-off intensity (see locations of the
497: existence domains for the bright solitons and the plane wave
498: bistability in Fig.~11) is shifted to the incident intensity,
499: whereupon a soliton can form. Evidently for different parameters
500: the shift can be substantially larger or smaller than the width of
501: bistability loop, in which case no stable soliton can appear. We
502: note that in absence of thermal effects (good heat-sinking of
503: sample) solitons would not appear spontaneously but would have to
504: be switched on by local pulsed light injection.
505:
506: \begin{figure}[htbf]
507: \epsfxsize=70mm \centerline{\epsfbox{Fig12.eps}} \vspace{0.7cm}
508: \caption{Comparison of bright soliton formation above bandgap
509: (left) and below bandgap (right). Reflectivity on a diameter of
510: the illuminated area as a function of time (b), (d). Intensity of
511: incident (dotted) and reflected (solid) light, at the center of
512: the soliton as a function of time (a), (c). Arrows mark the
513: switch-on and -off.}
514: \end{figure}
515:
516: Fig.~13~(a) shows the incoherent switch-on of bright soliton,
517: where the perpendicularly polarized switching pulse ($\sim$ 10 ns)
518: is applied at $t$ = 4 $\mu$s. As apparent, a soliton forms after
519: this incoherent light pulse. The slow formation of the soliton is
520: apparent in Fig.~13~(a) (using roughly the time from $t$ = 4 to
521: $t$ = 4.5 $\mu$s).
522:
523: It should be emphasized that this thermal effect is not
524: instrumental for switching a soliton on. However, it allows
525: switching a soliton off incoherently \cite{tag:30}. This is shown
526: in Fig.~13~(b) where the driving light is initially raised to a
527: level at which a soliton forms spontaneously (note again the slow
528: soliton formation due to the thermal effect). The incoherent
529: switching pulse is then applied which leads to disappearance of
530: the soliton.
531:
532: \begin{figure}[htbf]
533: \epsfxsize=60mm \centerline{\epsfbox{Fig13.eps}} \vspace{0.7cm}
534: \caption{Recording of incoherent switching-on (a) and
535: switching-off (b) of a soliton. Snapshot pictures show unswitched
536: state (left), circular switched domain (center). and a soliton
537: (right). Dotted trace: incident intensity.}
538: \end{figure}
539:
540: The soliton can thus be switched on and also off by an incoherent
541: pulse. The reason for the latter is thermal. Initially the
542: material is "cold". A switching pulse leads then to the creation
543: of a soliton. Dissipation in the material raises the temperature
544: and the soliton is slowly formed. At the raised temperature the
545: band edge (and with it the bistability characteristic) and the
546: existence range of solitons is shifted so that a new pulse brings
547: the system out of the range of existence of solitons.
548: Consequently the soliton is switched off.
549:
550: Thus switching on a soliton is possible incoherently with the
551: "cold" material and switching off with the "heated" material.
552: When the driving intensity is chosen to be slightly below the
553: spontaneous switching threshold the nonlinear resonator is cold.
554: An incoherent pulse increases illumination locally and can switch
555: the soliton on that causes local heating of the resonator. Another
556: incoherent pulse aimed into the heated area can then switch the
557: soliton off and thereby cool the resonator to its initial state,
558: so that the soliton could be switched on/off again.
559:
560: \subsection{Above bandgap bright solitons}
561: %
562: At excitation above bandgap bright solitons form analogously to
563: the below/near bandgap case \cite{tag:31}. Fig.~14 shows the
564: bright solitons as observed in the reflected light with their
565: characteristic concentric rings with the same appearance as the
566: bright solitons below band gap (Fig.~9~(c)).
567:
568: \begin{figure}[htbf]
569: \epsfxsize=50mm \centerline{\epsfbox{Fig14.eps}} \vspace{0.7cm}
570: \caption{Bright soliton above bandgap in 3D representation: view
571: from above (a), from below (b).}
572: \end{figure}
573:
574: Fig.~12 compares the dynamics of the bright soliton formation for
575: excitation above bandgap (left) and below bandgap (right).
576: Difference between these two cases can be understood from the
577: model. From (1) we obtain the reflected intensity as a function of
578: incident intensity for wavelengths above (${\rm Re}(\alpha)>0$),
579: as well as below the bandgap (${\rm Re}(\alpha)<0$), for plane
580: waves (Fig.~15). One sees that the bistability range is large
581: below and small above the bandgap. Solving (1) numerically the
582: typical bright soliton (top of Fig.~15) is found coexisting with
583: homogeneous intensity solutions in the shaded regions of
584: Fig.~15~(a), (b).
585:
586: After switching on the resonator below band gap (Fig.~15~(b)), the
587: intensity in the resonator is high and with it the thermal
588: dissipation. The temperature consequently rises, which shifts the
589: band gap \cite{tag:29} and with it the bistability characteristic,
590: so that the switch-off intensity, close to which the stable
591: solitons exist, becomes close to the incident intensity. Then the
592: resonator is in the basin of attraction for the solitons and the
593: soliton forms as observed in Fig.~12~(c), (d).
594:
595: \begin{figure}[htbf]
596: \epsfxsize=85mm \centerline{\epsfbox{Fig15.eps}} \vspace{0.7cm}
597: \caption{Steady-state plane wave solution of (1) above bandgap
598: (a): ${\rm Re}(\alpha)=0.05$; and below band gap (b): ${\rm
599: Re}(\alpha)=-0.05$. Other parameters: $C=30$, ${\rm
600: Im}(\alpha)=0.99$, $\theta=-3$, $P=0$, $d=0.1$. The soliton
601: solution shown exists for incident intensities corresponding to
602: the shaded areas, in coexistence with homogeneous solutions. For a
603: temperature increase the characteristics together with soliton
604: existence ranges shift to higher incident intensities. Reflected
605: and incident intensities normalized to the same value.}
606: \end{figure}
607:
608: \begin{figure}[htbf]
609: \epsfxsize=80mm \centerline{\epsfbox{Fig16.eps}} \vspace{0.7cm}
610: \caption{Bistability characteristics (reflected light intensity
611: versus incident light intensity) measured at the center of an
612: illuminated area of 100 $\mu$m diameter, for above band gap (a)
613: and below band gap (b) excitation.}
614: \end{figure}
615:
616: Above the band gap Fig.~12~(a), (b) show that the soliton is
617: switched on "immediately" without the slow thermal process. Fig.~
618: 15~(a) shows why. The plane wave characteristic of the resonator
619: above band gap is either bistable but very narrow, or even
620: monostable (due to the contribution of the self-focusing reactive
621: nonlinearity \cite{tag:32}) but still with bistability between the
622: soliton state (not plane wave) and the unswitched state. In this
623: case the electronic switching leads directly into the basis of
624: attraction for solitons and the switch-on of the soliton is
625: purely electronic and fast. The width of the bistability
626: characteristics observed experimentally (Fig.~16) scale in
627: agreement with Fig.~15.
628:
629: Nonetheless, also above bandgap there is strong dissipation after
630: the switch-on. The associated temperature rise influences and can
631: even destabilize the soliton. The effect can be seen in
632: Fig.~12~(a). Over a time of a few $\mu$s after the soliton
633: switch-on the soliton weakens (reflectivity increases slowly)
634: presumably by the rise of temperature and the associated shift of
635: the band gap. At 6.5 $\mu$s the soliton switches off although the
636: illumination has not yet dropped.
637:
638: Thus, while the dissipation does not hinder the fast switch-on of
639: the soliton, it finally destabilizes the soliton. After the
640: soliton is switched off, the material cools and the band gap
641: shifts back so that the soliton could switch on again.
642:
643: \begin{figure}[htbf]
644: \epsfxsize=60mm \centerline{\epsfbox{Fig17.eps}} \vspace{0.7cm}
645: \caption{Recording of coherent switching-on (a) and switching-off
646: (b) of a bright soliton. Heavy arrows mark the application of
647: switching pulses. Dotted traces: incident intensity. The insets
648: show intensity snapshots, namely unswitched state (left) and
649: soliton (right).}
650: \end{figure}
651:
652: Fig.~17 shows the soliton coherent switching observations
653: \cite{tag:33}. Fig.~17~(a) shows switching a soliton on. The
654: driving light intensity is chosen slightly below the spontaneous
655: switching threshold. At t $\approx$ 1.2 $\mu$s the addressing
656: pulse is applied. It is in phase with the driving light, as
657: visible from the constructive interference. A bright soliton
658: results, showing up in the intensity time trace as a strong
659: reduction of the reflected intensity. Fig.~17~(b) shows switching
660: a soliton off. The driving light is increased to a level where a
661: soliton is formed spontaneously. The addressing pulse is then
662: applied in counterphase to the driving light, as visible from the
663: destructive interference. The soliton then disappears, showing up
664: in the intensity time trace as reversion of the reflected
665: intensity to the incident intensity value. The Fig.~17 insets
666: show 2D snapshots before and after the switching pulses for
667: clarity.
668:
669: \subsection{Optical pumping}
670: %
671: The thermal effects discussed above result from the local heating
672: caused by the high intracavity intensity at the bright soliton
673: location. They limit the switching speed of solitons and they
674: will also limit the speed at which solitons could be moved
675: around, limiting applications. The picture is that a soliton
676: carries with it a temperature profile, so that the temperature
677: becomes a dynamic and spatial variable influencing the soliton
678: stability.
679:
680: As opposed, a spatially uniform heating will not cause such
681: problems, as it shifts parameters but does not constitute a
682: variable in the system. The largely unwanted heating effects are
683: directly proportional to the light intensity sustaining a
684: soliton. For this reason and quite generally it is desirable to
685: reduce the light intensities required for sustaining solitons.
686:
687: Conceptually this can be expected if part of the power sustaining
688: a resonator soliton could be provided incoherently to the driving
689: field, e.g. by means of optical pumping. Optical pumping of MQW
690: structure generates carriers and allows converting from
691: absorption to gain. In the last case the semiconductor
692: microresonator operates as VCSEL \cite{tag:17}. To couple pump
693: light into the microresonator the pump laser was tuned in one of
694: short-wavelength transparency windows in the reflection spectrum
695: of the microresonator as shown in Fig.~2~(a).
696:
697: When the pumping was below the transparency point and the driving
698: laser wavelength was set near the semiconductor MQW structure
699: band edge the bright and dark solitons formed (Fig.~18~(a), (b))
700: \cite{tag:34} similar to the unpumped case Figs 9, 10.
701:
702: \begin{figure}[htbf]
703: \epsfxsize=70mm \centerline{\epsfbox{Fig18.eps}} \vspace{0.7cm}
704: \caption{Intensity snapshots of structures observed in reflection
705: from pumped (below transparency) semiconductor microresonator
706: illuminated near resonance showing bright (a) and dark (b)
707: soliton. The illuminating beam from the laser diode has an
708: elliptical shape (c).}
709: \end{figure}
710:
711: Analysis of (1) shows that increase of the pump intensity leads to
712: shrinking of the resonator solitons' existence domain and
713: shifting towards low intensity of the light sustaining the
714: solitons. Such reduction of the sustaining light intensity was
715: observed experimentally in \cite{tag:23}. That is why soliton
716: switchings in the pumped case are fast \cite{tag:23} and not
717: mediated by thermal effects as for soliton formation without
718: pumping (Fig.~12~(c), Fig.~13).
719:
720: When the pump intensity approaches the transparency point of the
721: semiconductor material, the resonator solitons' domain of
722: existence disappears. It reappears above the transparency point.
723: In the experiment we have quite strong contribution of the
724: imaginary part (absorption/gain) of the complex nonlinearity at
725: the working wavelength (near band edge). Therefore the
726: transparency point is very close to the lasing threshold so that
727: inversion without lasing is difficult to realize.
728:
729: \begin{figure}[htbf]
730: \epsfxsize=70mm \centerline{\epsfbox{Fig19.eps}} \vspace{0.7cm}
731: \caption{Intensity snapshots of typical beam structures at
732: optical pump intensities slightly above lasing threshold (pump
733: increases from (a) to (c)).}
734: \end{figure}
735:
736: Slightly above threshold we observe in presence of illumination
737: structures (Fig.~19) reminiscent of the solitons in electrically
738: pumped resonators \cite{tag:35}. We note that optical as opposed
739: to electrical pumping allows more homogeneous pumping conditions
740: \cite{tag:36}. This suggests that optically pumped resonators
741: lend themselves more readily for localization and motion control
742: of solitons then electrically pumped ones.
743:
744: \begin{figure}[htbf]
745: \epsfxsize=70mm \centerline{\epsfbox{Fig20.eps}} \vspace{0.7cm}
746: \caption{Results of numerical simulations of below bandgap
747: (purely dispersive) solitons using the model (1) for pumped above
748: the transparency point microresonator ($P=2$). Inset is bright
749: soliton. Shaded area is domain of existence of resonator bright
750: solitons. Area limited by dashed lines is optical bistability
751: domain for plane waves. Other parameters same as in Fig.~3.}
752: \end{figure}
753:
754: There is difference between below bandgap resonator solitons in
755: pumped and unpumped cases. The nonlinear resonance mechanism of
756: soliton formation \cite{tag:22} requires a defocusing
757: nonlinearity below transparency (Fig.~3) and a focusing
758: nonlinearity above transparency. Transverse nonlinear
759: self-focusing effect is generally furthering soliton formation.
760: Fig.~20 shows typical examples of calculated resonator solitons
761: for pumped semiconductor microresonator. Bright solitons have a
762: large existence range in the pumped case (Fig.~20), dark solitons
763: exist, though with smaller range of stability, in the unpumped
764: case (Fig.~3).
765:
766: Thus optically pumped semiconductor resonators are well suited
767: for sustaining solitons below bandgap: (i) background light
768: intensity necessary to sustain and switch resonator solitons is
769: substantially reduced by the pumping and therefore destabilizing
770: thermal effects are minimized, (ii) above the transparency point
771: only the dispersive part of semiconductor nonlinearity stabilizes
772: a soliton, therefore the domain of existence of below bandgap
773: (purely dispersive) bright solitons can be quite large.
774:
775: \section{Conclusion}
776: %
777: In conclusion, we have shown experimentally and numerically
778: existence of spatial solitons in driven semiconductor
779: microresonators over a wide spectral range around the bandgap:
780: below bandgap hexagons and dark solitons, near bandgap bright and
781: dark solitons, above bandgap bright solitons, optically pumped
782: below bandgap bright solitons. We have demonstrated the
783: manipulation of such solitons: switching them on and off by
784: coherent as well as incoherent light; reducing the light power
785: necessary to sustain and switch a soliton, by optical pumping.\\
786:
787: Acknowledgment\\ This work was supported by Deutsche
788: Forschungsgemeinschaft under grant We743/12-1.
789:
790: %\bibliographystyle{empty}
791: \begin{thebibliography}{99}
792: \vspace*{-0.7cm}
793: \bibitem{tag:1} D.W. Mc Laughlin, J.V. Moloney, A.C. Newell: Phys. Rev. Lett. \textbf{51}, 75 (1983)
794:
795: \bibitem{tag:2} N.N. Rosanov, G.V. Khodova: Opt. Spectrosk. \textbf{65}, 1375 (1988)
796:
797: \bibitem{tag:3} M. Tlidi, P. Mandel, R. Lefever: Phys. Rev. Lett. \textbf{73}, 640 (1994)
798:
799: \bibitem{tag:4} G.S. Donald, W.J. Firth: J. Opt. Soc. Am. B \textbf{73}, 1328 (1990)
800:
801: \bibitem{tag:5} M. Brambilla, L.A. Lugiato, M. Stefani: Europhys. Lett. \textbf{34}, 109 (1996)
802:
803: \bibitem{tag:6} V.Yu Bazhenov, V.B. Taranenko, M.V. Vasnetsov: `Transverse optical effects in bistable active cavity with nonlinear absorber
804: on bacteriorhodopsin'. In: \emph{Proc. SPIE}, \textbf{1840},
805: (1992) pp. 183-193
806:
807: \bibitem{tag:7} K. Staliunas, V.B. Taranenko, G. Slekys, R. Viselga, C.O. Weiss: Phys. Rev. A \textbf{57}, 599 (1998)
808:
809: \bibitem{tag:8} G. Slekys, K. Staliunas, C.O. Weiss: Opt. Comm. \textbf{149}, 113 (1998)
810:
811: \bibitem{tag:9} K. Staliunas, V.J. Sanchez-Morcillo: Phys. Rev. A \textbf{57}, 1454 (1998)
812:
813: \bibitem{tag:10} V.B. Taranenko, K. Staliunas, C.O. Weiss: Phys. Rev. Lett. \textbf{81}, 2236 (1998)
814:
815: \bibitem{tag:11} K. Staliunas: Phys. Rev. Lett. \textbf{81}, 81 (1998)
816:
817: \bibitem{tag:12} C. Rebbi, G. Soliani: \emph{Solitons and Particles}, (World Scientific, 1984)
818:
819: \bibitem{tag:13} C. Tamm: Phys. Phys. Rev. A \textbf{38}, 5960 (1988)
820:
821: \bibitem{tag:14} K. Staliunas, C.O. Weiss, G. Slekys: `Optical vortices in lasers'. In: \emph{Optical Vortices}. ed.
822: by M. Vasnetsov and K. Staliunas (Nova Science Publishers, 1999)
823: pp. 125-182
824:
825: \bibitem{tag:15} C.O. Weiss, M. Vaupel, K. Staliunas, G. Slekys, V.B. Taranenko: Appl. Phys. B \textbf{68}, 151 (1999)
826:
827: \bibitem{tag:16} \emph{Soliton Driven Photonics}, ed. by A.D. Boardman, A.P. Sukhorukov (Kluver Academic Publishers, 2001)
828:
829: \bibitem{tag:17} T.E. Sale: \emph{Vertical Cavity Surface Emitting Lasers}, (Wiley, 1995)
830:
831: \bibitem{tag:18} H.M. Gibbs: \emph{Optical Bistability - Controlling Light with Light}, (Academic Press, 1985)
832:
833: \bibitem{tag:19} L. Spinelli, G. Tissoni, M. Brambilla, F. Prati, L.A. Lugiato: Phys. Rev. A \textbf{58}, 2542 (1998)
834:
835: \bibitem{tag:20} D. Michaelis, U. Peschel, F. Lederer: Phys. Rev. A \textbf{56}, R3366 (1997)
836:
837: \bibitem{tag:21} L.A. Lugiato, R. Lefever: Phys. Rev. Lett. \textbf{58}, 2209 (1987)
838:
839: \bibitem{tag:22} K. Staliunas, V.J. Sanchez-Morcillo: Opt. Comm. \textbf{139}, 306 (1997)
840:
841: \bibitem{tag:23} V.B. Taranenko, C.O. Weiss, W. Stolz: Opt. Lett. \textbf{26}, 1574 (2001)
842:
843: \bibitem{tag:24} V.B. Taranenko, I. Ganne, R. Kuszelewicz, C. O. Weiss: Phys. Rev. A \textbf{61}, 063818 (2000)
844:
845: \bibitem{tag:25} P.K. Jacobsen, J.V. Moloney, A.C. Newell, R. Indik: Phys. Rev. A \textbf{45}, 8129 (1992)
846:
847: \bibitem{tag:26} W.J. Firth, A.J. Scroggie: Europhys. Lett. \textbf{26}, 521 (1994)
848:
849: \bibitem{tag:27} V.B. Taranenko, C.O. Weiss, B. Sch\"apers: Phys. Rev. A \textbf{65}, 013812 (2002)
850:
851: \bibitem{tag:28} V.B. Taranenko, I. Ganne, R. Kuszelewicz, C.O. Weiss: Appl. Phys. B \textbf{72}, 377 (2001)
852:
853: \bibitem{tag:29} T. Rossler, R.A. Indik, G.K. Harkness, J.V. Moloney, C.Z. Ning: Phys. Rev. A \textbf{58}, 3279 (1998)
854:
855: \bibitem{tag:30} V.B. Taranenko, C.O. Weiss: Appl. Phys. B \textbf{72}, 893 (2001)
856:
857: \bibitem{tag:31} V.B. Taranenko, C.O. Weiss, W. Stolz: J. Opt. Soc. Am. B \textbf{19}, 8129 (1992)
858:
859: \bibitem{tag:32} S.H. Park, J.F. Morhange, A.D. Jeffery, R.A. Morgan, A. Chavez-Pirson, H.M. Gibbs, S.W. Koch, N. Peyghambarian, M.
860: Derstine, A.C. Gossard, J.H. English, W. Weidmann: Appl. Phys.
861: Lett. \textbf{52}, 1201 (1988)
862:
863: \bibitem{tag:33} V.B. Taranenko, F.-J. Ahlers, K. Pierz: Appl. Phys. B
864: (2002) in print
865:
866: \bibitem{tag:34} V.B. Taranenko, C.O. Weiss: `Spatial solitons in an optically pumped semiconductor microresonator'. nlin.PS/0204048 (2002)
867:
868: \bibitem{tag:35} Report as given in www.pianos-int.org
869:
870: \bibitem{tag:36} W.J. Alford, T.D. Raymond, A.A. Allerman: J. Opt. Soc. Am. B \textbf{19}, 663 (1992)
871:
872:
873:
874: \end{thebibliography}
875:
876: \end{document}
877: