Файл: Staliunas K., Sanchez-Morcillo V.J. (eds.) Transverse Patterns in Nonlinear Optical Resonators(ST.pdf

ВУЗ: Не указан

Категория: Не указан

Дисциплина: Не указана

Добавлен: 28.06.2024

Просмотров: 708

Скачиваний: 0

ВНИМАНИЕ! Если данный файл нарушает Ваши авторские права, то обязательно сообщите нам.

7.4 Experimental Demonstration of the “Restless” Vortex

113

We recall that the coe cient G12 vanishes for phase-insensitive (helical) modes (6.3), and is nonzero for phase-sensitive (flower) modes (6.12).

7.4.2 Phase-Insensitive Modes

In a class A laser, the frequencies of phase-insensitive modes are not a ected by the nonlinear coupling. The vortices rotate around the optical axis with a frequency exactly equal to the di erence of the eigenfrequencies of the two corresponding modes. In the degenerate case the vortices stop rotating. We show below that for a class B laser, the vortices behave di erently.

We look for solutions of (7.17) in the form fi(t) = ni exp (iΩit). The coe cients of the population inversion become

dii = fifi ,

 

γ

 

 

 

 

 

(7.18a)

dij = −fifj

 

 

 

 

,

 

 

(7.18b)

 

 

 

 

 

γ + i (Ωi j )

 

 

and the mode beat frequency ∆Ω = Ωi j

obeys the equation

1 ∆Ω γ2

G12

+ 2

+ ∆Ω2

G12

+ 1 = 2γ (D0 1) , (7.19)

 

ω

 

G11

 

 

 

 

G11

 

where ∆ω = ωi − ωj is the mode degeneracy here 2, or, in other words, the di erence of the eigenfrequencies of the modes.

Fig. 7.8. The mode beat frequency ∆Ω versus the mode frequency detuning ∆ω for di erent values of γ. The mode-coupling coe cients correspond to the beating of the TEM01 and TEM01 modes (a circling vortex)

A family of curves of ∆Ω versus ∆ω for di erent values of γ is plotted in Fig. 7.8. The curves clearly show the mode-pushing phenomenon in the case of a class B laser. The pushing is weaker in the intermediate case between a class A and a class B laser (γ ≈ 1) and grows with decreasing γ. When

2 Do not confuse this ∆ω with ∆ω = ω − a 2 used previously in this chapter.


114 7 The Restless Vortex

γ <

2 (D0 1)

,

(7.20)

G11/G12 + 2

 

 

 

bistability appears. In this case the beat frequency is never zero, which means that a stationary pattern consisting of helical modes is never generated.

The frequency of the self-induced mode beat in the case ∆ω = 0 is given

by

∆Ω2

=

2γ (D0 1) − γ2 (G11/G12 + 2)

,

(7.21)

G11/G12 + 1

0

 

 

 

 

 

 

 

which is proportional to the frequency of the relaxation oscillations in the limit γ 1.

7.4.3 Phase-Sensitive Modes

Mode pushing also occurs for the modes which frequency-pull for class A lasers. For a class B laser, the relation between the mode beat frequency and the detuning is modified with respect to the class A laser, as shown in Fig. 7.9. The mode-locking region shrinks with decreasing γ and bistability appears, similarly to the case of no mode-locking.

Fig. 7.9. The mode beat frequency ∆Ω versus the mode frequency detuning ∆ω for di erent values of γ in the mode-pulling case. The modecoupling coe cients correspond to the TEM10 and TEM01 Gauss–Hermite modes, and their locking to the helical TEM01 mode

A numerical analysis leads to the conclusion that the transition between mode locking and unlocking occurs at a mode beat frequency proportional to the relaxation oscillation frequency: the pattern either oscillates at a frequency larger than the relaxation frequency or locks to a stationary pattern. A pattern consisting of nonlocking modes is never at rest for a class B laser.

The mode locking was investigated experimentally by studying the vortex behavior in a CO2 laser [3]. Figure 7.10 shows the frequency of the spatial


References 115

oscillations of the laser pattern as the frequency mismatch of the modes was varied (by varying the astigmatism of the resonator). The experiment shows the predicted hysteresis in the locking and unlocking of the vortex. The experiment also shows that the smallest possible frequency of the oscillations is proportional to the relaxation oscillation frequency.

Fig. 7.10. Locking and unlocking of the “doughnut” in a CO2 laser. The pin position controls the astigmatism of the resonator, and correspondingly the frequency detuning between the Gauss–Hermite modes

References

1.J. Lega, J.V. Moloney and A.C. Newell, Universal description of laser dynamics near threshold, Physica D 83, 478 (1995). 105

2.K. Staliunas and C.O. Weiss, Nonstationary vortex lattices in large-aperture class B lasers, J. Opt. Soc. Am. B 12, 1142 (1995). 108, 110

3.G. Slekys, K. Staliunas, M.F.H. Tarroja and C.O. Weiss, Cooperative frequency locking and tristability in a class-B laser, Appl. Phys. B 59, 11 (1994). 114


8 Domains and Spatial Solitons

This chapter is devoted to the description of the general properties of a particular class of solutions of extended nonlinear systems, namely spatial solitons, or localized structures. In contrast to extended patterns, such as rolls or hexagons, spatial solitons are solutions where the field is inhomogeneous only in a localized region of the space and is homogeneous in the rest of the space.

The stability of spatial solitons is due to a balance between linear and nonlinear e ects. In transverse nonlinear optics, these balancing e ects are usually the spreading caused by di raction and the compression (self-focusing) caused by a focusing nonlinearity.

Spatial solitons have been theoretically predicted for a variety of nonlinear optical cavities [1]. In this chapter we present some general concepts concerning spatial domains and solitons, and the next three chapters are devoted to the implementation of these ideas in concrete optical systems.

The study of optical solitons is relevant not only from a fundamental viewpoint but also because of their potential applications in information processing technology. Such practical applications will be discussed in Chaps. 9 and 11.

8.1 Subcritical Versus Supercritical Systems

The existence of spatial solitons is closely related to the character of the bifurcation from the nonlasing (trivial) to the lasing regime. It is useful to review here some basic concepts of bifurcation theory [2].

A bifurcation is supercritical if the transition from one solution to another, obtained by varying the control parameter (the pump intensity in the optical case), is continuous, and the solutions connect at the bifurcation point. Equivalently, an infinitesimal variation of the control parameter leads to an infinitesimal change in the amplitude of the solution in the case of a supercritical bifurcation. Examples of a supercritical bifurcation are shown in Figs. 8.1a,c. A supercritical bifurcation corresponds to a phase transition of type II. In contrast, a bifurcation is subcritical if, at the bifurcation point, the amplitudes of the solutions di er by a finite quantity (i.e. they are disconnected). In this case, both solutions may coexist below the bifurcation

K. Stali¯unas and V. J. S´anchez-Morcillo (Eds.):

Transverse Patterns in Nonlinear Optical Resonators, STMP 183, 117–124 (2003)c Springer-Verlag Berlin Heidelberg 2003


118 8 Domains and Spatial Solitons

A

a)

p

A

Domain Soliton

b)

p

space

A

Domain Soliton

c)

p

space

Fig. 8.1. Di erent types of bifurcation in nonlinear systems. (a) Supercritical Hopf bifurcation, (b) subcritical Hopf bifurcation, (c) supercritical pitchfork bifurcation. A is the amplitude of the solution, and is p the pump (criticality) parameter

point, the system showing bistability or hysteresis (Fig. 8.1b, center). The corresponding phase transition is of type I.

Another possible classification of bifurcations takes into account the symmetry of the phase of the emerging solution. When the phase of the solution is invariant (not fixed by the system), a Hopf bifurcation occurs, as shown in the phase diagrams at the left in Figs. 8.1a,b. If two opposite values of the phase are preferred, a pitchfork (static) bifurcation occurs instead, and a real-valued order parameter is obtained.

As discussed in the following sections, some kind of bistability is always needed for the existence of domains and of spatial solitons. The order parameter equations derived in Chaps. 2 and 3 (the Ginzburg–Landau and Swift– Hohenberg equations, either real or complex), which are representative of most nonlinear optical systems, possess a primary bifurcation of supercritical type.

8.2 Mechanisms Allowing Soliton Formation

There are two basic mechanisms that may cause subcriticality, and consequently may lead to soliton formation. Both are related to the existence of

8.2 Mechanisms Allowing Soliton Formation

119

absolute bistability between two di erent extended solutions (which may or not be homogeneous). In order to show this, let us consider the complex Swift–Hohenberg equation, and analyze the properties of its simplest nontrivial solution: a traveling wave.

8.2.1 Supercritical Hopf Bifurcation

Consider the simplest form of the complex Swift–Hohenberg equation, as obtained in Chaps. 2 and 3:

 

∂A

 

2

 

 

 

 

= pA − A |A|2 + i a 2 A − g

a 2

 

A .

(8.1)

 

∂t

 

Consider also a solution of (8.1) in the form of a traveling wave,

 

A = |A| exp(ik0x − iωt) .

 

 

 

(8.2)

The intensity of the traveling wave can be found from (8.1), and plotted in terms of the various parameters. Figure 8.2a shows the usual bifurcation diagram, with the intensity as a function of the criticality parameter p. Another useful representation, shown in Fig. 8.2b, is the dependence of the intensity on the squared wavenumber. It is evident that no bistability is possible here, since the trivial solution is always unstable against a traveling wave when the latter exists. Consequently, solitons cannot be stable under supercritical conditions, and subcriticality of the emerging solution is then required. In order to introduce subcriticality, an external e ect is usually added to the system, such as an intracavity saturable absorber, an intracavity focusing/defocusing material or parametric forcing. The corresponding equation modeling the dynamics is a modified Ginzburg–Landau or Swift–Hohenberg equation, with additional terms describing the subcriticality.

 

A

 

2

 

A

 

2

 

 

 

 

 

 

 

 

 

0

 

0

 

p

k02

k2

Fig. 8.2. Dependence of the intensity on the criticality parameter p and on the squared wavenumber k2, in the case of a supercritical Hopf bifurcation