Switchable linear-cavity nanotube-mode-locking fiber laser emitting picosecond or femtosecond pulses

Switchable linear-cavity nanotube-mode-locking fiber laser emitting picosecond or femtosecond pulses

Optics Communications 335 (2015) 262–265 Contents lists available at ScienceDirect Optics Communications journal homepage: www.elsevier.com/locate/o...

1MB Sizes 16 Downloads 108 Views

Optics Communications 335 (2015) 262–265

Contents lists available at ScienceDirect

Optics Communications journal homepage: www.elsevier.com/locate/optcom

Switchable linear-cavity nanotube-mode-locking fiber laser emitting picosecond or femtosecond pulses Xiankun Yao n State Key Laboratory of Transient Optics and Photonics, Xi'an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi'an 710119, China

art ic l e i nf o

a b s t r a c t

Article history: Received 4 July 2014 Received in revised form 29 August 2014 Accepted 12 September 2014 Available online 26 September 2014

A switchable linear-cavity nanotube-mode-locking fiber laser is proposed by exploiting fiber Bragg grating for the first time to the author's best knowledge. The proposed all-fiber laser can deliver either picosecond or femtosecond pulses by controlling the polarization controllers. The durations of picosecond and femtosecond pulses are 16.4 ps and 838 fs with the central wavelengths of 1529.5 and 1560 nm, respectively. The femtosecond pulse has symmetrical spectrum sidebands, but the picosecond pulse almost has no spectral sidebands. Our work provides a simple, low-cost, dual-scale pulse source for practical applications. & 2014 Elsevier B.V. All rights reserved.

Keywords: Mode-locked lasers Switchable lasers Ultrafast lasers Carbon nanotube

1. Introduction Passively mode-locking fiber lasers have been extensively investigated, since they offer a perfect platform to generate ultrashort/ ultrafast pulses [1–6] and study nonlinear phenomenon [7–12]. In past decades, various mode-locking techniques have been proposed in fiber laser, such as nonlinear optical loop mirrors [13], nonlinear polarization rotation [14–16], semiconductor saturable absorber mirrors (SESAMs) [7], graphite [17], carbon nanotubes [18,19, charcoal [20], graphene [21–24], graphene-nanotube mixtures [25], and topological insulator [21,26]. Among them, single-wall carbon nanotubes (SWNTs) have attracted considerable interest for their intrinsic advantages of the ultrashort recovery time, broad operation bandwidth, and polarization insensitivity [27–30]. By using SWNTs, pulses from tens of femtosecond to picosecond have been realized in fiber laser [18,19,29]. The multi-wavelength mode-locking fiber lasers with various configurations have been investigated based on nanotubes [30,31]. In addition, the harmonically passive mode-locking of fiber lasers with a SWNT were demonstrated recently [32]. With the dispersion design of laser cavity, fiber lasers can emit conventional soliton [3,33,34], dispersion-managed soliton [35], selfsimilar pulse [1], dissipative soliton [36,37], and soliton molecule [38,39]. Generally, mode-locking fiber lasers are realized by the ring [7,27,28,40,41] or figure-eight [13,42] configuration. However, fiber lasers with linear cavity are often employed for its simple design by employing chirped fiber Bragg grating (CFBG) or SESAM without circulator [43,44]. What's more, with incorporating CFBG into a linear

cavity, the spatial hole burning can be reduced due to the lack of a fixed cavity length [45]. As a reliable, compact component, FBG is usually utilized for wavelength selectivity in fiber laser or communication system [46–50]. By utilizing FBGs, switchable fiber laser has been proposed by various means [51–54]. For example, a switchable mode-locking operation has been reported by changing the intracavity loss with a tunable attenuator [54]. However, the previous works mainly concentrated on separately emitting picosecond or femtosecond pulses [51–54]. So far, the generation of picosecond or femtosecond pulses in a linear-cavity fiber laser based on SWNTs has not been reported. In this paper, we report a switchable linear-cavity nanotubemode-locked fiber laser, which can deliver picosecond or femtosecond pulses by adjusting the polarization controllers. A FBG with central wavelength of 1529.5 nm is inserted in the cavity. By appropriately setting the polarization controllers (PCs), the switchable mode-locking operation with pulse duration of 838 fs or 16.4 ps is observed in the fiber laser. The femtosecond operation spectrum, with a central wavelength of 1560 nm, has clearly symmetrical sidebands at both sides of the spectrum. However, the picosecond operation centered at 1529.5 nm almost has no spectral sideband, which can be attributed to the spectral filtering effect caused by FBG. Compared with constructing independent picosecond laser and femtosecond laser, the proposed scheme significantly reduces the cost and is attractive for ultrafast optics.

2. Experimental setup n

Corresponding author. E-mail address: [email protected]

http://dx.doi.org/10.1016/j.optcom.2014.09.036 0030-4018/& 2014 Elsevier B.V. All rights reserved.

The configuration of the proposed fiber laser is schematically shown in Fig. 1(a). The linear cavity is constructed with a fiber-based

X. Yao / Optics Communications 335 (2015) 262–265

mirror and a circulator, whose port1 and port3 are connected to form the other cavity mirror. The circulator in contrast to coupler takes minimal cavity loss, which is desirable to act as a cavity mirror. In the

Reflectivity

1.0

0.5

0.0 1529.0

1529.5

1530.0

Wavelength (nm)

Fig. 1. (a) Laser setup. LD—laser diode; WDM—wavelength-division multiplexer; EDF—erbium-doped fiber; OC—optical coupler; PC—polarization controller; PI-ISO —polarization-insensitive isolator; SWNTs—single-wall carbon nanotubes; FBG— fiber Bragg grating; CR—circulator; fs—femtosecond;ps—picosecond. (b) The reflection spectrum of the FBG.

263

linear cavity, a 7 m EDF with absorption of 6 dB/m at 980 nm is employed as the gain medium. A 980 nm laser diode (LD) provides pump with a 980/1550 nm wavelength-division-multiplexer (WDM). The 10% port of optical coupler (OC) provides the laser output. The polarization-insensitive isolator (PI-ISO) external to the cavity can prohibit the reflected light from disturbing the signals in fiber laser. The packaged SWNT-polyvinyl alcohol polymer is incorporated into the cavity to generate ultrashort pulses, which is fabricated as shown in Ref. [30]. A FBG centered at  1529.5 nm is inserted into the cavity adjacent to the circulator, and its reflected spectrum is shown in Fig. 1(b). The reflectivity and the 3 dB bandwidth of the FBG are  99.1% and  0.5 nm, respectively. The bandwidth of the FBG is a crucial parameter in the experiment which determines the spectral bandwidth of the picosecond pulse. The PC1 is used to optimize the mode-locking conditions, and PC2 can be utilized to control the loss between FBG and circulator. When we tighten the PC2 paddle and change its orientation, the SMF in PC2 is twisted and squeezed, which leads to increasing the fiber loss. The other fibers in the linear cavity are the standard single mode fiber (SMF) with the length of 14.1 m. The dispersion parameter of EDF and SMF are -9 ps/(nm km) and 17 ps/(nm km), respectively. An optical spectrum analyzer, a commercial autocorrelator, a radio-frequency (RF) analyzer, a 6 GHz oscilloscope, and a 10 GHz photodetector are employed to monitor the laser output.

1.0 ο

Exper. 2 Sech fit

Intensity (a.u)

0.8

0.6

1.54× 838 fs 0.4

0.2

0.0

-4

-2

2

0

4

Delay time (ps) 1.0

205.6 ns Intensity (a.u)

0.8 0.6 0.4 0.2 0.0 0

300

600

900

Time (ns) Fig. 2. The femtosecond operation at 1560 nm. (a) Optical spectrum; (b) AC traces; (c) fundamental RF spectrum; and (d) oscilloscope trace.

X. Yao / Optics Communications 335 (2015) 262–265

3. Experimental results and analysis

-30 -45

1.0

1.00

ο

Reflectivity

Intensity (dBm)

Light propagation in the linear cavity has two different routes for two wavelengths, which is tunable depending on the loss between the FBG and the circulator. When the PC-induced loss is negligible, lights around 1560 nm can circulate in the cavity between the fiber-based mirror and the circulator due to the dominating gain of EDF. In this case, the FBG can be regarded as a section of SMF because the lasing wavelength is beyond the reflection spectrum of the FBG. The length and net dispersion of the cavity are 21.1 m and -0.45 ps2, respectively. When the PC-induced loss is strong, the light centered at 1529.5 nm is directly reflected by the FBG and can not reach the circulator. In the case, the light oscillates between the fiber-based mirror and the FBG. The FBG acts as a high-reflective mirror and narrowband spectral filter which dominates the formation of picosecond solitons. The length of the cavity is 18.6 m. In our experiment, the continuous wave state can be achieved at the pump power of 12 mW. With relaxing the PC2 and adjusting the polarization state of the PC1, self-started passive mode locking at 1560 nm can be obtained at the pump power of 33 mW. Fig. 2 (a) shows the typical output spectrum. The clear Kelly sidebands can be seen on the spectrum, which is originated form the constructive interference between the solitons and dispersive waves [3,18,55]. That is the typical characteristics of conventional

solitons. The optical spectrum is centered at 1560 nm, and the 3 dB bandwidth is 3.3 nm. Fig. 2(b) illustrates the autocorrelation (AC) trace. If a sech2 profile is assumed for fitting, the full width at half maximum (FWHM) of the autocorrelation trace is 1.29 ps, and the pulse duration is estimated as 838 fs. The time-bandwidth product (TBP) is calculated as 0.347, which indicates that the pulse is almost chirp-free. The corresponding RF spectrum is shown in Fig. 2(c) that the fundamental repetition rate is 4.863266 MHz (The measured result of RF analyzer can be accurate to 1 Hz). The stable mode-locking operation is confirmed by a high signal/noise ratio (SNR) of 50 dB. As demonstrated in Fig. 2(d), the oscilloscope trace shows that the separation of adjacent pulses is 205.6 ns corresponding to the cavity length of 21.1 m. When the loss imposed by PC2 is strong, only the spectral band near 1529.5 nm have output. The light centered at 1529.9 nm is reflected by the FBG. In this case, the FBG plays two roles of a spectral filter achieving wavelength selection and a grating providing large anomalous dispersion [48]. Through carefully adjusting the PC1 state, the single-wavelength mode-locking operation was easily achieved, as shown in Fig. 3(a)the corresponding output spectrum of solitons at a pump power about 33 mW. The spectrum is centered at 1529.5 nm, which corresponds to the reflection wavelength of FBG. We note that the spectrum has a bandwidth 0.16 nm much smaller than the bandwidth of FBG. From the inset of Fig. 3(a), we can see that the top regions of the two spectra

0.8

0.95

-60 1529 1530 0.90 Wavelength (nm)

Intensity (a.u.)

264

Exper. 2 Sech fit

0.6

1.54×16.4 ps 0.4

0.2

0.0

-40

-20

0 20 Delay time (ps)

40

1 .0

181.5 ns

Intensity (a.u)

0.8 0.6 0.4 0.2 0.0 0

200

400

600

800

Time (ns) Fig. 3. The picosecond operation at 1529.5 nm. (a) Optical spectrum, the inset is the reflection spectrum of the FBG and optical spectrum of picosecond operation in detail; (b) AC trace; (c) fundamental RF spectrum; and (d) oscilloscope trace.

X. Yao / Optics Communications 335 (2015) 262–265

correspond to each other very well. The spectral bandwidth of the picosecond mode-locked pulse is significantly affected by the reflection characteristic of FBG with the pulse running a large number of rounds in the cavity. The spectrum almost has no sidebands, which is attributed to spectral filtering effect of FBG. When the pulse circulating in the cavity, the FBG cuts off the spectrum and suppresses the generation of sidebands. However, the output pulse is conventional soliton as well for the netnegative-dispersion cavity. The AC trace of the solitons is shown in Fig. 3(b). If a Sech2 temporal profile is assumed for fitting, the FWHM of the AC trace is 25.25 ps, and the pulse duration is estimated as 16.4 ps. Here, the pulse duration is much larger than 838 fs in Fig. 2(b), which can be attributed to the spectral filtering effect and the large dispersion induced by the FBG. The TBP is given by 0.337, indicating that the pulse is almost chirp-free. The stable mode locking operation is conformed by RF spectrum as illustrated in Fig. 3(c). The signal/noise ratio (SNR) is about  56 dB, indicating that a stable mode locking is achieved. The fundamental repetition rate is 5.509066 MHz. As shown in Fig. 3(d), the round-trip time is  181.5 ns, corresponding to the cavity length of 18.6 m. In our experiment, by appropriately adjusting the PCs, the switchable picosecond or femtosecond solitons at different wavelength is achieved. When the loss imposed by PC2 is negligible, the pulse duration of the mode-locking is 838 fs. The central wavelength is 1560 nm, which is determined by the gain profile of EDF and the transmission spectrum of SWNTs. The solitons possess typical characteristics of conventional solitons that the spectral sidebands are symmetrically distributed at both sides of the spectrum. When the loss induced by PC2 is strong, only the wavelength band selected by FBG can circulated in the cavity. As a consequence, the picosecond solitons at 1529.5 nm are achieved, the pulse duration is 16.4 ps, and the sidebands are suppressed by FBG. However, the picosecond pulse is less stable than the femtosecond pulse, because the autocorrelator trace in Fig. 2 (b) is much better fitted by sech2 function than Fig. 3(b). The FBG we used has a very narrow bandwidth of 0.5 nm, which corresponds to strong spectral filtering effect. So the picosecond pulse power is rather weak even without any attenuator in the cavity. When the FBG is stretched by the stretching device, the wavelength shift induced by the variable FBG period happened. The ratio of the picosecond and femtosecond soliton durations is about 19.58. The calculated TBPs are 0.337 and 0.347 respectively, which are slightly larger than the transform limit (0.315) of Sech2—shape pulses. The difference of the fundamental repetition rate between the two mode-locking operations is 0.6458 MHz, which can be attributed to the difference of the effective cavity lengths of 1.5 m. 4. Conclusions A switchable linear-cavity nanotube-mode-locking all-fiber laser is investigated experimentally for the first time to author's best knowledge. It can deliver either picosecond or femtosecond pulses by controlling the polarization controllers of laser cavity. The switchable mode-locking operation with pulse duration of 16.4 ps or 838 fs is generated experimentally. The femtosecond operation spectrum, with a central wavelength of 1560 nm, has clearly symmetrical sidebands at both sides of the spectrum. However, the picosecond operation centered at 1529.5 almost has no spectral sideband, which can be attributed to the spectral filtering effect caused by FBG. The ratio of the picosecond and femtosecond solitons durations is about 19.58. Both of the

265

operations are almost chirp-free. The scheme we proposed may have important application by providing dual-scale pulse sources in the future.

References [1] F.Ö. Ilday, et al., Phys. Rev. Lett. 92 (21) (2004) 213902; B. Oktem, et al., Nat. Photon. 4 (5) (2010) 307. [2] Y. Nozaki, et al., Opt. Lett. 37 (24) (2012) 5079. [3] X. Liu, Phys. Rev. A 84 (2) (2011) 023835. [4] N. Akhmediev, et al., Phys. Lett. A 372 (17) (2008) 3124; N. Akhmediev, et al., J. Opt. Soc. Am. B 15 (2) (1998) 515. [5] L. Wang, et al., Laser Phys. Lett. 7 (1) (2010) 63; L. Wang, et al., Laser Phys. Lett. 8 (5) (2011) 376. [6] M.E. Fermann, et al., Nat. Photonics 7 (2013) 868; M.E. Fermann, Opt. Lett. 23 (1) (1998) 52. [7] D. Mao, et al., Opt. Lett. 37 (13) (2012) 2619; D. Mao, et al., Opt. Lett. 38 (16) (2013) 3190. [8] T. Brabec, et al., Rev. Mod. Phys. 72 (2) (2000) 545. [9] X.M. Liu, et al., Opt. Express 18 (9) (2010) 8847–8852. [10] D. Mao, et al., Sci. Rep. 3 (2013) 3223. [11] W.H. Renninger, et al., Nat. Commun. 4 (2013) 1719; W.H. Renninger, et al., Phys. Rev. A 77 (2) (2008) 023814. [12] X. Liu, et al., Opt. Express 17 (10) (2009) 8506. [13] L. Yun, et al., Opt. Express 20 (19) (2012) 20992; L. Yun, et al., IEEE Photon. J. 4 (2) (2012) 512. [14] D. Mao, et al., Opt. Express 18 (22) (2010) 23024; D. Mao, et al., Opt. Express 19 (5) (2011) 3996; X.M. Liu, et al., Laser Phys. 18 (11) (2008) 1357. [15] L. Duan, et al., Opt. Express 20 (1) (2012) 265. [16] S. Kobtsev, et al., Opt. Express 17 (23) (2009) 20707. [17] G.R. Lin, et al., Laser Phys. Lett. 8 (12) (2011) 880; Y.H. Lin, et al., Laser Phys. Lett. 9 (5) (2012) 398. [18] C. Zeng, et al., Opt. Express 21 (16) (2013) 18937; L. Yun, et al., Opt. Express 22 (5) (2014) 5442. [19] N. Nishizawa, et al., Opt. Express 19 (22) (2011) 21874. [20] Y.H. Lin, et al., Opt. Express 21 (21) (2013) 25184; Y.H. Lin, et al., Laser Phys. Lett. 10 (5) (2013) 055105. [21] F. Bonaccorso, et al., Opt. Mater. Express 4 (1) (2014) 63. [22] A. Martinez, et al., Nat. Photonics 7 (11) (2013) 842. [23] Y.F. Song, et al., Opt. Express 20 (24) (2012) 27283. [24] Y.H. Lin, et al., Opt. Express 21 (14) (2013) 16763. [25] Y. Cui, et al., Opt. Express 21 (16) (2013) 18969; Y. Cui, et al., Laser Phys. Lett. 11 (5) (2014) 055106. [26] J. Sotor, et al., Appl. Phys. Lett. 104 (25) (2014) 251112. [27] J.C. Chiu, et al., Opt. Express 18 (4) (2010) 3592. [28] K.N. Cheng, et al., Laser Phys. 23 (4) (2013) 045105. [29] A. Schmidt, et al., Opt. Express 20 (5) (2012) 5313. [30] X. Liu, et al., Sci. Rep. 3 (2013) 2718. [31] S.J. Tan, et al., Laser Phys. 23 (5) (2013) 055101. [32] K.N. Cheng, et al., IEEE Photon. J. 4 (5) (2012) 1542. [33] L. Wang, et al., Opt. Express 19 (8) (2011) 7616. [34] X.M. Liu, Phys. Rev. A 84 (5) (2011) 053828. [35] S.K. Turitsyn, et al., Phys. Rep. 521 (4) (2012) 135. [36] X. Liu, Phys. Rev. A 81 (2) (2010) 023811. [37] P. Grelu, et al., Nat. Photonics 6 (2) (2012) 84; P. Grelu, et al., Opt. Express 12 (14) (2004) 3184. [38] X.M. Liu, Phys. Rev. A 82 (6) (2010) 063834. [39] V. Tsatourian, et al., Sci. Rep. 3 (2013) 3154. [40] X.M. Liu, Opt. Express 17 (12) (2009) 9549. [41] X.H. Li, et al., Opt. Lett. 35 (19) (2010) 3249. [42] H.B. Sun, et al., Laser Phys. 20 (2) (2010) 522. [43] M.A. Ummy, et al., Opt. Express 19 (4) (2011) 3202. [44] M. Bravo, et al., Opt. Lett. 38 (9) (2013) 1542. [45] M.E. Fermann, et al., Opt. Lett. 20 (2) (1995) 172. [46] R. Kashyap, Fiber Bragg Gratings, Academic Press, San Diego, 1999. [47] X.M. Liu, IEEE Photon. Technol. Lett. 19 (9) (2007) 632. [48] D. Han, et al., Opt. Express 20 (24) (2012) 27045; D. Han, et al., Opt. Lett. 39 (6) (2014) 1565. [49] H. Tsuda, Opt. Lett. 35 (14) (2010) 2349. [50] C.S. Kim, et al., Opt. Express 15 (7) (2007) 3702. [51] X. Liu, Phys. Rev. A 77 (4) (2008) 043818; X.M. Liu, et al., Opt. Lett. 30 (17) (2005) 2257; X. Liu, Laser Phys. 20 (4) (2010) 842. [52] D.E. Ceballos-Herrera, et al., Opt. Laser Technol. 43 (4) (2011) 825. [53] X.M. Liu, et al., Laser Phys. Lett. 5 (12) (2008) 904; X. Liu, et al., Opt. Express 13 (1) (2005) 142. [54] X. Zhao, et al., Opt. Express 19 (2) (2011) 1168. [55] Y.H. Lin, Laser Phys. Lett. 10 (4) (2013) 045109.