Silicon waveguide with four zero-dispersion

wavelengths and its application in on-chip

octave-spanning supercontinuum generation
Lin Zhang,1,2,* Qiang Lin,3,4 Yang Yue,1 Yan Yan,1 Raymond G. Beausoleil,5
and Alan E. Willner1
1
Department of Electrical Engineering, University of Southern California, Los Angeles, California 90089, USA
2
Current address: Microphotonics Center and Department of Material Science and Engineering, Massachusetts
Institute of Technology, Cambridge, Massachusetts 02139, USA
3
Department of Electrical and Computer Engineering, University of Rochester, Rochester, New York 14627, USA
4
Institute of Optics, University of Rochester, Rochester, New York 14627, USA
5
HP Laboratories, Palo Alto, California 94304, USA
*linzhang@mit.edu

Abstract: We propose a novel silicon waveguide that exhibits four zero-

dispersion wavelengths for the first time, to the best of our knowledge, with
a flattened dispersion over a 670-nm bandwidth. This holds a great potential
for exploration of new nonlinear effects and achievement of ultra-
broadband signal processing on a silicon chip. As an example, we show that
an octave-spanning supercontinuum assisted by dispersive wave generation
can be obtained in silicon, over a wavelength range from 1217 to 2451 nm,
almost from bandgap wavelength to half-bandgap wavelength. Input pulse
is greatly compressed to 10 fs.
©2012 Optical Society of America
OCIS codes: (130.4310) Nonlinear; (190.4390) Nonlinear optics, integrated optics; (190.5530)
Pulse propagation and temporal solitons; (320.5520) Pulse compression; (320.6629)
Supercontinuum generation.

References and links

1. G. P. Agrawal, Nonlinear Fiber Optics, 3rd ed. (Academic, 2001).
2. A. Ferrando, E. Silvestre, J. J. Miret, and P. Andrés, “Nearly zero ultraflattened dispersion in photonic crystal
fibers,” Opt. Lett. 25(11), 790–792 (2000).
3. F. Poletti, V. Finazzi, T. M. Monro, N. G. R. Broderick, V. Tse, and D. J. Richardson, “Inverse design and
fabrication tolerances of ultra-flattened dispersion holey fibers,” Opt. Express 13(10), 3728–3736 (2005).
4. D. J. J. Hu, P. P. Shum, C. Lu, and G. Ren, “Dispersion-flattened polarization-maintaining photonic crystal fiber
for nonlinear applications,” Opt. Commun. 282(20), 4072–4076 (2009).
5. H. Xu, J. Wu, K. Xu, Y. Dai, C. Xu, and J. Lin, “Ultra-flattened chromatic dispersion control for circular
photonic crystal fibers,” J. Opt. A, Pure Appl. Opt. 13(5), 055405 (2011).
6. W. H. Reeves, D. V. Skryabin, F. Biancalana, J. C. Knight, P. St. J. Russell, F. G. Omenetto, A. Efimov, and A.
J. Taylor, “Transformation and control of ultra-short pulses in dispersion-engineered photonic crystal fibres,”
Nature 424(6948), 511–515 (2003).
7. K. Saitoh and M. Koshiba, “Highly nonlinear dispersion-flattened photonic crystal fibers for supercontinuum
generation in a telecommunication window,” Opt. Express 12(10), 2027–2032 (2004).
8. M. H. Frosz, P. Falk, and O. Bang, “The role of the second zero-dispersion wavelength in generation of
supercontinua and bright-bright soliton-pairs across the zero-dispersion wavelength,” Opt. Express 13(16), 6181–
6192 (2005).
9. W.-Q. Zhang, S. Afshar V, and T. M. Monro, “A genetic algorithm based approach to fiber design for high
coherence and large bandwidth supercontinuum generation,” Opt. Express 17(21), 19311–19327 (2009).
10. S. Stark, F. Biancalana, A. Podlipensky, and P. St. J. Russell, “Nonlinear wavelength conversion in photonic
crystal fibers with three zero-dispersion points,” Phys. Rev. A 83(2), 023808 (2011).
11. J. I. Dadap, N. C. Panoiu, X. Chen, I.-W. Hsieh, X. Liu, C.-Y. Chou, E. Dulkeith, S. J. McNab, F. Xia, W. M. J.
Green, L. Sekaric, Y. A. Vlasov, and R. M. Osgood, Jr., “Nonlinear-optical phase modification in dispersion-
engineered Si photonic wires,” Opt. Express 16(2), 1280–1299 (2008).
12. Q. Lin, O. J. Painter, and G. P. Agrawal, “Nonlinear optical phenomena in silicon waveguides: modeling and
applications,” Opt. Express 15(25), 16604–16644 (2007).
13. X. Chen, N. C. Panoiu, and R. M. Osgood, Jr., “Theory of Raman-mediated pulsed amplification in silicon-wire
waveguides,” IEEE J. Quantum Electron. 42(2), 160–170 (2006).
14. J. Leuthold, C. Koos, and W. Freude, “Nonlinear silicon photonics,” Nat. Photonics 4(8), 535–544 (2010).

#157567 - $15.00 USD Received 3 Nov 2011; revised 25 Dec 2011; accepted 25 Dec 2011; published 11 Jan 2012
(C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1685
15. B. J. Eggleton, B. Luther-Davies, and K. Richardson, “Chalcogenide photonics,” Nat. Photonics 5, 141 (2011).
16. M. R. Lamont, C. M. de Sterke, and B. J. Eggleton, “Dispersion engineering of highly nonlinear As2S3
waveguides for parametric gain and wavelength conversion,” Opt. Express 15(15), 9458–9463 (2007).
17. X. Liu, W. M. J. Green, X. Chen, I.-W. Hsieh, J. I. Dadap, Y. A. Vlasov, and R. M. Osgood, Jr., “Conformal
dielectric overlayers for engineering dispersion and effective nonlinearity of silicon nanophotonic wires,” Opt.
Lett. 33(24), 2889–2891 (2008).
18. L. Zhang, Y. Yue, Y. Xiao-Li, J. Wang, R. G. Beausoleil, and A. E. Willner, “Flat and low dispersion in highly
nonlinear slot waveguides,” Opt. Express 18(12), 13187–13193 (2010).
19. S. Mas, J. Caraquitena, J. V. Galán, P. Sanchis, and J. Martí, “Tailoring the dispersion behavior of silicon
nanophotonic slot waveguides,” Opt. Express 18(20), 20839–20844 (2010).
20. L. Zhang, Y. Yue, R. G. Beausoleil, and A. E. Willner, “Flattened dispersion in silicon slot waveguides,” Opt.
Express 18(19), 20529–20534 (2010).
21. D. D. Hudson, S. A. Dekker, E. C. Mägi, A. C. Judge, S. D. Jackson, E. Li, J. S. Sanghera, L. B. Shaw, I. D.
Aggarwal, and B. J. Eggleton, “Octave spanning supercontinuum in an As2S3 taper using ultralow pump pulse
energy,” Opt. Lett. 36(7), 1122–1124 (2011).
22. L. Zhang, Y. Yan, Y. Yue, Q. Lin, O. Painter, R. G. Beausoleil, and A. E. Willner, “On-chip two-octave
supercontinuum generation by enhancing self-steepening of optical pulses,” Opt. Express 19(12), 11584–11590
(2011).
23. R. Halir, Y. Okawachi, J. S. Levy, M. A. Foster, M. Lipson, and A. L. Gaeta, “Octave-spanning supercontinuum
generation in CMOS-compatible silicon nitride waveguides,” in CLEO - Laser Applications to Photonic
Applications 2011, paper PDPA6 (2011).
24. Ö. Boyraz, P. Koonath, V. Raghunathan, and B. Jalali, “All optical switching and continuum generation in
silicon waveguides,” Opt. Express 12(17), 4094–4102 (2004).
25. L. Yin, Q. Lin, and G. P. Agrawal, “Soliton fission and supercontinuum generation in silicon waveguides,” Opt.
Lett. 32(4), 391–393 (2007).
26. I.-W. Hsieh, X. Chen, X. Liu, J. I. Dadap, N. C. Panoiu, C.-Y. Chou, F. Xia, W. M. Green, Y. A. Vlasov, and R.
M. Osgood, “Supercontinuum generation in silicon photonic wires,” Opt. Express 15(23), 15242–15249 (2007).
27. B. Kuyken, X. Liu, R. M. Osgood, Y. A. Vlasov, R. Baets, G. Roelkens, and W. M. Green, “Generation of a
telecom-to-mid-infrared spanning supercontinuum using silicon-on-insulator wire waveguides,” in CLEO - Laser
Applications to Photonic Applications 2011, paper CTuS1 (2011).
28. J. M. Dudley, G. Genty, and S. Coen, “Supercontinuum generation in photonic crystal fiber,” Rev. Mod. Phys.
78(4), 1135–1184 (2006).
29. D. V. Skryabin and A. V. Gorbach, “Colloquium: Looking at a soliton through the prism of optical
supercontinuum,” Rev. Mod. Phys. 82(2), 1287–1299 (2010).
30. P. Koonath, D. R. Solli, and B. Jalali, “Limiting nature of continuum generation in silicon,” Appl. Phys. Lett.
93(9), 091114 (2008).
31. Q. Lin, “Generalized nonlinear envelope equation with high-order dispersion of nonlinearity” (to be published).
32. G. Genty, P. Kinsler, B. Kibler, and J. M. Dudley, “Nonlinear envelope equation modeling of sub-cycle
dynamics and harmonic generation in nonlinear waveguides,” Opt. Express 15(9), 5382–5387 (2007).
33. R. Sun, P. Dong, N.-N. Feng, C.-Y. Hong, J. Michel, M. Lipson, and L. Kimerling, “Horizontal single and
multiple slot waveguides: optical transmission at λ = 1550 nm,” Opt. Express 15(26), 17967–17972 (2007).
34. A. D. Bristow, N. Rotenberg, and H. M. van Driel, “Two-photon absorption and Kerr coefficients of silicon for
850-2200 nm,” Appl. Phys. Lett. 90(19), 191104 (2007).
35. Q. Lin, J. Zhang, G. Piredda, R. W. Boyd, P. M. Fauchet, and G. P. Agrawal, “Dispersion of silicon
nonlinearities in the near-infrared region,” Appl. Phys. Lett. 91(2), 021111 (2007).
36. S. Afshar V and T. M. Monro, “A full vectorial model for pulse propagation in emerging waveguides with
subwavelength structures part I: Kerr nonlinearity,” Opt. Express 17(4), 2298–2318 (2009).

1. Introduction
Chromatic dispersion is one of essential properties of an optical waveguide. It not only shapes
the temporal waveform of an optical pulse traveling along the waveguide, but also acts
together with nonlinearity to affect pulse spectral broadening [1]. In particular, zero-
dispersion wavelength (ZDW) plays a critical role in determining phase matching conditions
and conversion efficiency in nonlinear parametric processes [1]. Silica step-index optical
fibers usually exhibit one ZDW at near infrared wavelengths of interest, while photonic
crystal fibers (PCFs) provide a great opportunity to tailor dispersion profile, being able to
produce three [2,3] or four [4,5] ZDWs and greatly tune their spectral positions. Such a
spectral profile of the dispersion strongly influences nonlinear processes in fibers (e.g.,
supercontinuum generation) and generates rich and unprecedented phenomena [6–10].
Integrated waveguides with high nonlinearity have generated much excitement recently
[11–15]. In particular, flat and low dispersion over a wide wavelength range becomes
desirable for femtosecond pulses or octave-spanning spectral broadening. Dispersion tailoring
with two ZDWs in high-index-contrast waveguides was presented [16–19]. More recently, a

#157567 - $15.00 USD Received 3 Nov 2011; revised 25 Dec 2011; accepted 25 Dec 2011; published 11 Jan 2012
(C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1686
dispersion flattening technique was proposed, producing 20 times better dispersion flatness
and three ZDWs [20]. Dispersion tailoring assists octave-spanning super-continuum
generation in integrated chalcogenide and silicon nitride waveguides [21–23], where index
contrast is lower than in silicon waveguides and waveguide dispersion becomes smaller.
Nevertheless, such supercontinuum generation requires a high peak power of input
femtosecond pulses, which is feasible only in a material platform with two-photon absorption
(TPA) absent or extremely low. In silicon waveguides, strong dispersion and considerable
TPA make it quite challenging to achieve octave-spanning supercontinua [24–27].
We propose a silicon slot waveguide that exhibits four ZDWs in a wavelength range from
1461 and 2074 nm, for the first time to the best of our knowledge. Dispersion values and
ZDW locations can be varied, which provides a great opportunity to explore rich nonlinear
effects such as solitons propagation, modulation instability, and supercontinuum and
frequency comb generation. We show by simulation that the unique dispersion assists octave-
spanning supercontinuum generation from 1217 to 2451 nm on a silicon chip. In this process,
the input pulse is greatly compressed to 10 fs.

Fig. 1. Silicon slot waveguide for flattened dispersion with four ZDWs. A horizontal silica slot
is between two silicon layers. Optical power distributions are shown near the ZDWs.

Fig. 2. (a) Material dispersion in silicon and flattened dispersion with four ZDWs in the
proposed silicon slot waveguide. (b) A close-up view of dispersion and group delay profiles.

2. Dispersion-flattened silicon slot waveguide with four ZDWs

The waveguide has a horizontal silica slot formed between two silicon layers, as shown in
Fig. 1. The waveguide parameters are: width W = 610 nm, upper height Hu = 136 nm, lower
height Hl = 344 nm, and slot height Hs = 40 nm. The buried oxide substrate is 2-µm thick. To
calculate the dispersion of quasi-TM mode (vertically polarized), we obtain its effective index
of refraction as a function of wavelength [20], using a full-vector mode solver, COMSOL,
with material dispersions of Si and SiO2 taken into account.
Figure 2 shows silicon material dispersion and the flattened dispersion of the quasi-TM
mode, calculated as D = -(c/λ)·(d2neff/dλ2). Dispersion varies between −22 and + 20
ps/(nm·km) over a 667-nm bandwidth, from 1435 to 2102 nm. Four ZDWs are found at 1461,
1618, 1889, and 2074 nm, respectively, as shown in Fig. 2(b). The group delay, defined as τ =

#157567 - $15.00 USD Received 3 Nov 2011; revised 25 Dec 2011; accepted 25 Dec 2011; published 11 Jan 2012
(C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1687
(1/c)·(neff-λ·dneff/dλ), has a small variation of 40 fs/cm from 1403 to 2146 nm, which produces
ultra-broadband group-velocity match for femtosecond pulse interactions.
The idea of generating four ZDWs is the following. First, material dispersion in silicon is
always negative at wavelengths of interest (see Fig. 2(a)). Forming a waveguide, we see that,
at short wavelengths close to the Si bandgap wavelength (~1100 nm), material dispersion is
dominant, so overall dispersion is negative. At long wavelengths where the guided mode
approaches cut-off, the dispersion is dominated by waveguide dispersion, which is negative
[18]. Thus, as long as one tailors waveguide dimensions so that, in the middle wavelength
range, waveguide dispersion is positive and stronger than material dispersion, a positive
overall dispersion can be obtained with two ZDWs [18]. Then, we use the anti-crossing effect
caused by mode transition, that is, the guided mode evolves from being more strip-mode-like
to being more slot-mode-like as wavelength increases (see Fig. 1). This induces additional
negative waveguide dispersion [20] and generates another two ZDWs in the middle.

Fig. 3. Dispersion is increased from normal to anomalous dispersion regime as (a) the upper
height decreases or (b) the lower height increases, with a small change in dispersion slope.

One can change the structural parameters in the waveguide around the values given above
to tailor dispersion value and slope. We individually decrease the upper height and increase
the lower height, keeping the others the same. The dispersion profile is moved from normal to
anomalous dispersion regime in Fig. 3, with a dispersion value change of 18.5 and 15.9
ps/(nm·km) per nm, respectively. No significant change in dispersion slope is observed.

Fig. 4. (a) Dispersion slope can be greatly change by varying the slot height. (b) Spacing of
ZDWs is varied by increasing the waveguide width.

Dispersion slope can be greatly tailored by varying the slot height from 32 to 48 nm, as
shown in Fig. 4(a). On the other hand, ZDWs are moved with a spacing change among the
ZDWs as the waveguide width is increased, which causes a slight rotation of the dispersion
profile. The dispersion tailorability by varying the structural parameters gives us a valuable
space for device design, although the dispersion properties in the high-index-contrast
waveguides could be sensitive to fabrication imperfections.

#157567 - $15.00 USD Received 3 Nov 2011; revised 25 Dec 2011; accepted 25 Dec 2011; published 11 Jan 2012
(C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1688
3. Octave-spanning supercontinuum generation in Si waveguides
The flattened dispersion profile with multiple ZDWs can be used to generate rich nonlinear
effects, and one of advantages of using it is to generate octave-spanning supercontinua. We
note that supercontinuum generation in microstructured fiber [28,29] critically relies on its
engineered dispersion. In normal dispersion regime, self-phase modulation (SPM) is mainly
responsible for pulse spectral broadening, while higher-order soliton fission and dispersive
wave generation are identified to be the main reasons for supercontinuum generation in
anomalous dispersion regime [28]. Recently, we showed that, with nearly zero dispersion over
a wide wavelength band, the self-steepening effect in nonlinear pulse propagation can greatly
enhance spectral broadening to produce a two-octave supercontinuum on a chip [22]. To
generate a supercontinuum in silicon, the approaches based on self-steepening and SPM tend
to require very high peak power, which becomes impractical due to TPA in silicon [30]. Then,
operating in the anomalous dispersion regime, one either used femtosecond pulses to reduce
free carrier absorption [25,26] or moved to mid-infrared wavelengths to mitigate TPA [27],
but, without sufficient capability to engineer the dispersion of high-index-contrast Si
waveguides, octave-spanning supercontinuum generation still remains challenging.
Tailoring the dispersion to produce the saddle-shaped dispersion profile (four ZDWs) is
substantially different from what was proposed in [20], in which three ZDWs were obtained
in low-dispersion bandwidth, in terms of the ability to broaden pulse spectrum. This is
because a dispersion profile with three ZDWs has strong anomalous dispersion at the long-
wavelength end of the low-dispersion bandwidth, where dispersive wave can hardly be
generated [28]. In contrast, being able to have the saddle-shaped dispersion, one not only
achieves flat dispersion over an even wider bandwidth but also, more importantly, produces
normal dispersion at both ends of the bandwidth. Assisted by the dispersive wave generation
at the both ends, an octave-spanning supercontinuum is finally obtained in Si.
Here we choose width W = 610 nm, upper height Hu = 132 nm, lower height Hl = 344 nm,
and slot height Hs = 40 nm. The flattened part of the dispersion profile is intentionally moved
into anomalous dispersion regime, as shown in Fig. 3(a). Two ZDWs are located at 1418 and
2108 nm. Two local maxima of dispersion are 80 and 79 ps/(nm·km) at 1540 nm and 1965
nm wavelengths, respectively, and a local minimum of dispersion is 53.7 ps/(nm·km) at 1750
nm. To model supercontinuum generation, we use the following nonlinear envelope equation,
with high-order dispersion of nonlinearity parameters including nonlinear index n2, TPA
coefficient βTPA, and effective mode area Aeff:

∂ α ( −i ) m β m ∂ m  −iγ n  −i   i ∂  * ∂n
∞ ∞ n

 + + i∑ A=∑   1 −
2
m   [ A n ( A )]
 ∂z 2 m = 2 m ! ∂t  n=0 n!  2   ω0 ∂t  ∂t

where A = A(z,t) is the complex amplitude of an optical pulse, and γn is the nth-order
dispersion coefficient of nonlinearity, γn = ω0·∂n[γ(ω)/ω]/∂ωn. We include up to the 6th-order
here. Other terms are defined the same way in [22], and Raman terms are not needed for the
TM-mode [12]. Detailed derivation on the above equation is given in [31]. It was confirmed
that the simulation of pulse propagations using nonlinear envelope equations is quite accurate
[32]. The propagation loss is set to be 7 dB/cm, which may be lower in practice [33], since
more light is confined in lower crystal silicon layer in our case. All order dispersion terms and
carrier dynamics are included as shown in [25], and the free-carrier effects play a negligible
role in the nonlinear processes. We estimate the dispersion of modal distribution by
calculating the mode overlap factor given in [12] for the widely separated wavelengths at
1230, 1810, and 2395 nm. The overlap factor equals 0.745 between 1230 and 1810 nm and
0.971 between 1810 and 2395 nm. We believe that the major results presented here would
hold well when the mode overlap is included. In our simulations, the time window length is
50 ps (∆f = 20 GHz), and the whole bandwidth in the frequency domain is 1000 THz.
We launch a chirp-free hyperbolic secant pulse into the dispersion-flattened silicon slot
waveguide. The pulse has the center wavelength at 1810 nm and a full width at half-

#157567 - $15.00 USD Received 3 Nov 2011; revised 25 Dec 2011; accepted 25 Dec 2011; published 11 Jan 2012
(C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1689
maximum (FWHM) T0 of 120 fs. Its peak power is 62 W. At 1810 nm, we find nonlinear
index n2=7.2×10−18 m2/W and TPA coefficient βTPA=5.3×10−12 m/W, based on the
measurements in [34,35]. The nonlinear coefficient γ of the waveguide is (102+10.8j) /(m·W)
using a full-vector model [36]. The shock time τ equals (1.16+0.33j) fs. At 1810 nm, the 2nd-
order dispersion coefficient β2 = −0.999 ps2/km. Therefore, characteristic dispersion and
nonlinearity lengths LD = T02/|β2| = 14.4 cm and LN = 1/real(γ)P = 0.16 mm, which gives a
soliton number of 30.
Figure 5 shows pulse spectral broadening along the waveguide. At the beginning of the
propagation, SPM causes obvious spectral broadening. Due to a relatively low dispersion over
a wide band and thus small walk-off of frequency components, the pulse is significantly
compressed, which further enhances the spectral broadening. In addition, the pulse transfers
energy to high and low spectral components located around 1230 and 2395 nm, where
dispersion is normal. This is likely to be dispersive wave generation, and group delay matches
for 1300, 1810 and 2240 nm. At a propagation distance of 3.5 mm, the total spectrum
bandwidth at −25 dB level is 1234 nm, from 1217 to 2451 nm, more than one octave.
Accordingly, the compressed pulse has a FWHM of 12.2 fs, as shown in Fig. 5(b). The pulse
waveform has beating patterns at rising and falling edges, which correspond to the dispersive
waves at long and short wavelengths, respectively. With an input pulse FWHM of 60 fs, we
obtain a supercontinuum of 1250 nm at −17 dB level, from 1200 to 2450 nm, when the pulse
propagates 1.6 mm, exhibiting a FWHM reduced to 10.1 fs, as shown in Fig. 5(d). In these
two cases, the pulse root-mean-square width is 155 and 73.6 fs, respectively, due to the pulse
pedestals caused by dispersive waves. No soliton fission is observed, and we expect high
spectral coherence.

Fig. 5. Pulse spectra and waveforms, for input pulse FWHM of 120 fs in (a, b) and 60 fs in (c,
d). The waveforms are captured at 3.5 mm and 1.6 mm, respectively.

4. Conclusion
We present a dispersion tailoring technique that produces four ZDWs in a silicon slot
waveguide. With a flattened dispersion over a 670-nm bandwidth and normal dispersion at
both short- and long-wavelength ends of this bandwidth, one can generate octave-spanning
supercontinua in near-infrared wavelength region using the silicon waveguides, which would
enable numerous chip-scale sub-systems for frequency metrology, sensing, optical coherence
tomography, pulse compression, microscopy and spectroscopy, and telecommunication.
Acknowledgments
This work is supported by HP Labs.

#157567 - $15.00 USD Received 3 Nov 2011; revised 25 Dec 2011; accepted 25 Dec 2011; published 11 Jan 2012
(C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1690