1390

OPTICS LETTERS / Vol. 32, No. 11 / June 1, 2007

Surface superlattice gap solitons

Y. J. He,1 W. H. Chen,1 H. Z. Wang,1,* and Boris A. Malomed2
1

State Key Laboratory of Optoelectronic Materials and Technologies, Zhongshan (Sun Yat-Sen) University, Guangzhou 510275, China 2 Department of Interdisciplinary Studies, School of Electrical Engineering, Faculty of Engineering, Tel Aviv University, Tel Aviv 69978, Israel *Corresponding author: stswhz@zsu.edu.cn

Received January 3, 2007; revised March 1, 2007; accepted March 15, 2007; posted March 27, 2007 (Doc. ID 78662); published April 25, 2007 We demonstrate that specic surface superlattice gap solitons can be supported at an interface between a one-dimensional photonic superlattice and a uniform medium with saturable nonlinearity. The solitons are stable in the semi-innite gap but do not exist in the rst gap. With the decrease of the power, the solitons jump from the surface site to the next one, and they may continue the motion into the lattices, which offers potential applications for the routing of optical signals. 2007 Optical Society of America OCIS codes: 190.5530, 190.4360, 060.1810.

Surface waves propagate at interfaces between media with different values of the refractive index. In optics, surface waves have been observed, in particular, at interfaces of linear and photorefractive media [1]. Recently, surface-wave solitons at interfaces of uniform media and periodic waveguide arrays have attracted a great deal of attention [25]. Discrete surface solitons have also been observed in recent experiments [6]. Photonic and optical lattices are ideal testbeds for the studies of optical and matter-wave solitons, since the depth, period, and structure of the lattice may be readily altered in experiment. In particular, photonic lattices can be easily created and manipulated in photorefractive crystals [7,8]. Recently, it has been demonstrated that double-period optical lattices, alias superlattices, may serve as an efcient tool for manipulating matter-wave gap solitons [9,10]. Superlattices more general than double-period ones can be created by imposing a long-period modulation on a regular periodic lattice. In particular, Bragg supergratings created in this way open up extra-narrow bandgaps in the corresponding spectrum, where new types of gap solitons can be predicted [11]. In some cases, optical superlattices may be considered as regular lattices hosting a periodic chain of defects. It was recently demonstrated that the interaction of solitons with defects in optical lattices gives rise to new dynamic phenomena [12]. In this Letter, we demonstrate that stable surface superlattice gap solitons (SSGSs) can be formed at the interface between superlattices and uniform media with focusing saturable nonlinearity. Such solitons exist in the semi-innite (SI) gap but are absent in the rst gap. A noteworthy dynamic effect is that, with the decrease of its power, the soliton detaches itself from the surface: it jumps from the surface lattice site to the next one, and it may thus start progressive motion into the bulk of the lattice. The beam propagation along z in a onedimensional photonic superlattice with self-focusing saturable nonlinearity obeys the following normalized equation for amplitude q of the electromagnetic wave [7,12,13]:
0146-9592/07/111390-3/$15.00

q i z +

E0 1 + IL x + q 2

x2

q = 0,

where IL is the intensity prole of the SI lattices, IL = 0 at x / 2, and IL x = I0 sin2 x + /2 /d1 + 1 sin2 x 2

+ /2 /d2

at x / 2. In the general case of incommensurate periods d1 and d2, lattice (2) is quasi-periodic [14]. However, in this Letter we focus on the double-period lattice (such as, e.g., in [9]), with d1 = 2d2. The relative and total strengths of the superlattice are controlled by parameters (which takes values between 0 and 1) and I0, respectively. Further, x in Eq. (1) is the transverse coordinate measured in units of d1 / , 2 2 and E0 is the bias eld in units of 2 / k0n4d1 33 , e where k0 = 2 / 0 ( 0 is the wavelength in vacuum), ne is the unperturbed refractive index, and 33, the electro-optic coefcient of the crystal, is the wavenumber. We take typical parameters: d1 = 30 m, d2 = 15 m, 0 = 0.5 m, ne = 2.3, and 33 = 280 pm/ V. Other parameters are I0 = 3 and E0 = 6 [7,12]. In a photorefractive crystal, the effective superlattice potential given by Eq. (2) can be induced optically, illuminating the sample with a broad laser beam passed through a properly designed amplitude mask. First, it is necessary to nd the systems linear spectrum. To this end, we substitute q x , z = f x exp ikx i z in the linearized version of Eq. (1), where is the propagation constant, k is the Bloch wavenumber, and f x is a periodic function. Next, we search for stationary solitons as q x , z = f x exp i z , where f x is a real function satisfying the equation
2

f
2

E0 1 + IL x + f2

f + f = 0.

+ The solitons total power is P = f2 x dx. Solving Eq. (3) by dint of the shooting method, we construct families of the soliton solutions, which depend on , peri-

2007 Optical Society of America

June 1, 2007 / Vol. 32, No. 11 / OPTICS LETTERS

1391

rather spontaneously jump from the surface to the next site or may start progressive motion into the lattices [Figs. 3(e)3(h)]. This phenomenon originates from the weakened connement of solitons because the SSGSs are trapped at the lower refractive index regions, and their propagation constants are close to the band edge. Long simulations conrm that the motion regime is stable, continuing indenitely long (a possibility of the steady motion of the soliton in the same model with a uniform single-period lattice as demonstrated in Ref. [13]). The possibility to control the nal position of the soliton by varying its power
Fig. 1. (a) Lattice intensity prole with I0 = 3 and = 0.3. (b) Linear spectrum of Eq. (1) in the plane of E0 , ; the gray areas are Bloch bands. (c) Solitons total power versus for E0 = 6, with the solid and dashed segments pertaining to stable and unstable solutions, respectively. (d) Perturbation growth rate corresponding to (c).

ods d1 and d2, and lattice-modulation parameters I0 and . To examine the stability of SSGSs, we search for the perturbed solutions to Eq. (1) in the form q x , z = f x + h x , z + ie x , z exp i z , where h x , z and e x , z are real and imaginary parts of a perturbation eigenmode, respectively, with a complex instability growth rate . By linearizing Eq. (1), we arrive at coupled equations for the eigenmodes: h = 2e / x2 e + E0e / 1 + IL + f2 and e = 2h / x2 + h E0h 1 / 1 + IL + f2 2f2 / 1 + IL + f2 2 . These equations are solved numerically to get the perturbation growth rate Re [as shown in Fig. 1(d)]. The robustness of perturbed SSGSs was additionally tested in direct simulations of Eq. (1) with initial conditions obtained by multiplying the soliton with 1 + 1,2 x , where 1,2 x is a Gaussian random function with 1,2 = 0, the mean size of the perturbation amounting to 10% of the solitons amplitude. Figures 1(a) and 1(b) show typical examples of the intensity prole in the SI double-periodic lattice, with E0 = 6, I0 = 3 and = 0.3 in Eqs. (1) and (2), and the corresponding bandgap spectrum. Figure 1(c) presents a numerically found family of the SSGS solutions, in terms of the solitons power versus the propagation constant. The alternating stable and unstable regions are identied by nding the perturbation growth rate Re , as additionally illustrated by Fig. 1(d). Some properties of SSGSs are similar to those of surface defect solitons with a positive defect [15], including the existence of solitons in the SI gap with multiple stability regions and the decrease of solitons power with the increase of . In Fig. 2, we show examples of the propagation of stable and weakly unstable SSGSs in the SI gap up to z = 200, which is tantamount, roughly, to 200 diffraction lengths. The peak position of the soliton with 2.25 and 2.25 2.45 is located right at the surface or at the next lattice site, respectively, while, at larger , the soliton exists in the depth of the lattices, no longer being surface states; see an example in Fig. 2(h) for = 2.6. In Figs. 3(a)3(d) we demonstrate that, when SSGSs are unstable, they are not destroyed but

Fig. 2. (Color online) SSGSs in the SI gap for = 0.3. (a), (d) = 2.15 and (b), (e) = 2.25: stable solitons centered at the surface site of the lattice; (c), (f) = 2.4 and (g), (j) = 2.45: stable solitons centered at the next site; (h), (k) = 2.6 and (i), (l) = 1.8: a stable and a weakly unstable soliton, respectively, in the depth of the lattice and the surface.

Fig. 3. (Color online) Spontaneous shift of a surface soliton, due to its instability, to a stable position at the next = 2.3, (b), (d) = 0.7, = 2.35. site: (a), (c) = 0.3, Instability-induced transition of the soliton from the position pinned at the second site near the surface to a state of steady motion into the lattice: (e), (g) = 0.1, = 2.6, (f), (h) = 0.3, = 2.5.

1392

OPTICS LETTERS / Vol. 32, No. 11 / June 1, 2007

may nd potential applications to light-signal routing devices. Finally, we show how modulation rate affects the 0.1 and 0.7, the superproperties of SSGSs. At lattice is approximatively tantamount to the singleperiod one, with period d2 and d1, respectively; there0.7, where the fore we focus on the region of 0.1 double-period lattice demonstrates its specic features. Stability and instability regions for the SSGSs in the plane of / are displayed in Fig. 4(a). For 0.3 0.7, the stable regions I, II, and III correspond to the solitons pinned, respectively, at the surface, at the next site, and deep inside the lattices, 0.2 only stable solitons pinned at while for 0.1 the second site were found. On the other hand, the instability-induced jump of the soliton from the surface to the next site, and its transition into the mobile state, as shown in Figs. 3 and 4(a), resemble known properties of the staggered discrete surface solitons in waveguide arrays [2,6] but do not occur with surface defect solitons in ordinary (single-period) lattices [15]. Note that for = 0.4 solitons do not stagger, which shows the strongest connement of solitons. In addition, the solitons power decreases with the increase of (or ) for xed (or ) [Fig. 4(b)]. To summarize, we have demonstrated that stable gap solitons can be formed at the interface between

optical superlattices and uniform media with focusing saturable nonlinearity. Their essential properties are remarkably different from those of surface gap solitons in ordinary lattices with a defect [15], such as the shift of unstable solitons from the surface site to the next one and the onset of steady inward motion with a decrease of power. Potential applications to the design of light-signal routing devices were proposed. Such superlattices can also support the linear modes localized at or near the interface. This work was supported by the National Natural Science Foundation of China (10674183), the National 973 Project of China (2004CB719804), and the Ph.D. Degrees Foundation of Ministry of Education of China (20060558068). References
1. M. Cronin-Golomb, Opt. Lett. 20, 2075 (1995). 2. K. G. Makris, S. Suntsov, D. N. Christodoulides, and G. I. Stegeman, Opt. Lett. 30, 2466 (2005). 3. Y. V. Kartashov, V. A. Vysloukh, and L. Torner, Phys. Rev. Lett. 96, 073901 (2006). 4. M. I. Molina, I. L. Garanovich, A. A. Sukhorukov, and Y. S. Kivshar, Opt. Lett. 31, 2332 (2006). 5. Y. V. Kartashov, V. A. Vysloukh, D. Mihalache, and L. Torner, Opt. Lett. 31, 2329 (2006). 6. E. Smirnov, M. Stepic, C. E. Rter, D. Kip, and V. Shandarov, Opt. Lett. 31, 2338 (2006). 7. J. W. Fleischer, M. Segev, N. K. Efremidis, and D. N. Christodoulides, Nature 422, 147 (2003). 8. Z. Chen and K. McCarthy, Opt. Lett. 27, 2019 (2002). 9. P. J. Y. Louis, E. A. Ostrovskaya, and Y. S. Kivshar, Phys. Rev. A 71, 023612 (2005). 10. M. A. Porter, P. G. Kevrekidis, R. Carretero-Gonzlez, and D. J. Frantzeskakis, Phys. Lett. A 352, 210 (2006). 11. K. Yagasaki, I. M. Merhasin, B. A. Malomed, T. Wagenknecht, and A. R. Champneys, Europhys. Lett. 74, 1006 (2006). 12. J. Yang and Z. Chen, Phys. Rev. E 73, 026609 (2006). 13. B. A. Malomed, T. Mayteevarunyoo, E. A. Ostrovskaya, and Y. S. Kivshar, Phys. Rev. E 71, 056616 (2005). 14. H. Sakaguchi and B. A. Malomed, Phys. Rev. E 74, 026601 (2006). 15. W. H. Chen, Y. J. He, and H. Z. Wang, Opt. Express 14, 11271 (2006).

Fig. 4. Areas of the solitons stability (gray) and instability (white) in the SI gap. At the points marked by dots (at 0.7), it has been checked that the instability re0.3 sults in the jump of the soliton to the next site; see Figs. 0.3) show the tran3(c) and 3(d). The squares (at 0.1 sition of the solitons pinned the second site to steady motion across the lattice. (b) Solitons power versus at xed propagation constant .