Multi-pulses dynamic patterns in a topological insulatormode-locked ytterbium-doped fiber laser

Yan Peiguang a, Lin Rongyong a, Zhang Han b, Wang Zhiteng b,Chen Han a, Ruan Shuangchen a,n

a Shenzhen Key Laboratory of Laser Engineering, Shenzhen University, Shenzhen 518060, Chinab Key Laboratory of Optoelectronic Devices and Systems of Ministry of Education and Guangdong Province, Shenzhen University, Shenzhen 518060, China

a r t i c l e i n f o

Article history:Received 2 August 2014Received in revised form31 August 2014Accepted 3 September 2014Available online 18 September 2014

Keywords:Topological insulatorMode-locked fiber laserMulti-pulsesQuasi-square pulses

a b s t r a c t

Multi-pulse dynamic patterns have been experimentally observed in an Ytterbium-doped fiber laserpassively mode locked by a topological insulator (TI) Bi2Te3 saturable absorber (SA). The fundamentalmode-locking operation with a repetition rate of �1.10 MHz was achieved under a pump power of�160 mW with an appropriate setting of the polarization controller (PC). It was found that througheither changing the pump power or rotating the orientation of intra-cavity PC, several characteristicmodes have been experimentally observed, including disordered multi-pulses, bunch of pulses, andsoliton rains. Simultaneously, quasi-square pulses have also been observed in the laser cavity. Oursystematic study clearly demonstrated that TI could be developed as an effective SA for the generation ofdifferent pulse operation states in a passively mode-locked all-normal-dispersion Ytterbium-dopedfiber laser.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

Passively, mode-locked fiber lasers have attracted much atten-tion as a useful tool for the investigation of multi-soliton nonlineardynamic [1–4]. It has been shown that various mechanisms had beproposed to interpret the formation of multi-solitons in ananomalous dispersive cavity, including the soliton shaping ofdispersive waves, the effective spectral filtering effect, the wave-breaking effect and the soliton peak clamping effect. Indeed,various modes of multi-soliton operation have be observed inprevious experimental, including bound solitons [5,6], disorderedbunched multi-soliton [7,8], soliton rains [9–11], high-order har-monic mode-locking (HML) [12,13].

In compare with the erbium-doped fiber lasers, it was moredifficult to achieve the formation of multi-pulses in the ytterbium-doped fibers due to its relatively larger spectral gain bandwidth in [2],but multi-pulses could be still obtained if the pump power was higherthan some value [14,15]. However, most of the previous experimentalstudies on the multi-soliton operation states were concentrated onfiber lasers passively mode-locked with nonlinear amplifying loopmirror (NALM) [16] and the nonlinear polarization rotation (NPR)technique [12]. The usage of real passive saturable absorber (SA)seemed to be more advantageous for the development of commercial

mode-locked sources because NPR and NALM mode-locked laserwere environmentally more sensitive, and it would be important toinvestigate multi-soliton characteristics in fiber lasers passively mode-locked based on a SA, such as semiconductor saturable absorbers(SESAM) [17], single-walled carbon nanotube (SWNT) [18–21] andgraphene [22–30]. However, SESAMs had a narrow broadbandresponse (a few tens of nanometers) and required complex fabrica-tion process including integration and packaging. Moreover, SWNTsalways resulted in non-saturable loss, and restricted its applicationsdue to its response spectral range that was determined by thediameter and chirality. Although graphene had the broadbandresponse, it had a relatively low damage threshold and low saturationintensity, which may further limit the pulse energy.

Mostly recently, a new type of Dirac nanoscale material named“topological insulator (TI)” [31–36], which has been successfullydemonstrated as an effective saturable absorber material for thepassive Q-switched or mode-locking in fiber lasers or solid-state laserat different wavelengths [37–47]. Those experimental results demon-strated that TI-SA had a broadband response wavelength range ofsaturable absorption, large modulation depth, high damage thresholdand strong saturation intensity. It was demonstrated that introducingproper high nonlinear effect into the laser cavity was favorable forgenerating multi-solitons [48], and TI was experimentally found toshow very large nonlinear refractive index (a value of 10�14 m2/W)[49]. This indicated that TI played an important role on the formationof multi-soliton in mode-locked fiber lasers. Chen et al. have experi-mentally studied the formation of various multi-soliton patterns in a

Contents lists available at ScienceDirect

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

Optics Communications

http://dx.doi.org/10.1016/j.optcom.2014.09.0090030-4018/& 2014 Elsevier B.V. All rights reserved.

n Corresponding author.E-mail address: scruan@szu.edu.cn (R. Shuangchen).

Optics Communications 335 (2015) 65–72

mode-locked Erbium-doped fiber laser based TI:Bi2Te3 SA [50]. How-ever, no previous work reported the multi-pulse in a TI:Bi2Te3 basedmode-locked ytterbium-doped fiber laser, and the correspondingformation behavior of multi-pulse in such lasers still needed to beexplored.

In this paper, we reported on the generation of multi-pulseoperation states from all-normal dispersion ytterbium-doped fiberring laser passively mode-locked by TI:Bi2Te3 SA, which was sand-wiched between two optical fiber connectors through the opticaldeposition approach [51–53]. The fundamental mode-locking opera-tion was achieved under a pump power of �160 mW with anappropriate setting of the polarization controller (PC). By adjustingthe pump power and the orientation of the PC, it has been found thatmulti-pulse operation exhibited several characteristic operationsfrom disordered multi-pulse, bunch of pulses, and soliton rains.Besides the above mode-locking operation states, quasi-squarepulses could be also observed in the cavity under appropriate pumppower and orientation of the PC.

2. Sample preparation and experimental setup

The layered Bi2Te3 nano-sheets were synthesized by the liquidexfoliation approach, which had been elaborated in Ref. [37], andthe inter-band relaxation time of Bi2Te3 was �0.5 ps [54]. Thecorresponding scanning electron microscope (SEM) image wasshown in Fig. 1(a), it could be obviously seen that the TI nano-material had a quasi-two-dimensional structure with regular hex-agonal shape. The X-ray diffraction (XRD) pattern of TI nanosheets(PDF Card: 15-0863) showed a high [0 0 6] orientation and somecharacteristic peaks [0 1 5 and 0 0 1 5], as shown in Fig. 1(b).

Then Bi2Te3 nano-sheets were dissolved in distilled water andleading to an efficient dispersion through 3 h of ultra-sonication. Theuniform dispersion of Bi2Te3 nano-sheets was suitable for opticaldeposition. In the following, a continuous wave (CW) single modelaser diode (LD) with central wavelength of 975 nm and maximumpower of 500 mW spliced with a standard fiber connector (FC–PC)was used as the laser source for optical deposition, then light exitingfrom the fiber connector end-facet could be directly injected into thedispersion solution. The corresponding experimental setup wasshown in Fig. 2. After 2-min light illumination under a pump powerof 60 mW, Bi2Te3 nano-sheets could be gradually adsorbed anddeposited onto the fiber connector end-facet due to optical trappingforce and heat convention effects. After being dried in ambientenvironment for one day, the fibered topological insulator basedoptical device was fabricated after connecting with another clean anddry fiber connector. The inset of Fig. 2 showed that part of the fiber

connector end-facet, and it could be clearly seen that the fiberconnector core area was fully covered by the TI sample.

The saturable absorption property of the TI:Bi2Te3 sandwichedbetween optical fiber connectors was shown in Fig. 3. The lasersource used was home built mode-locked laser oscillator with arepetition rate of 5 MHz and 150 ps operating at �1064 nm. Thedate obtained from the experiment was then fitted according to

α Ið Þ ¼ α0 1þ IIsat

� ��1

þαns ð1Þ

where I was the input laser intensity, Isatwas the saturationintensity (the intensity with the absorption coefficient of half theinitial value), αðIÞ was the intensity-dependent absorption coeffi-cient, and α0 and αns were the modulation depth and the non-saturable loss, respectively. The results gave a saturation intensityof �24.6 MW∕cm2, modulation depth of �16.2%, and non-saturable loss of �22.8%.

Our fiber laser was schematically shown in Fig. 4. The laser waspumped by a 976 nm pump laser through a wavelength divisionmultiplexer (WDM), and the output of the laser was coupledthrough a 10% fiber optical coupler (OC). A segment of �3.4 m YDFwith absorption of 250 dB/m@975 nm was used as the gainmedium. A polarization independent isolator (ISO) was used toensure the unidirectional operation of the ring laser cavity, and athree-spool polarization controller (PC) was employed to adjustthe cavity birefringence. A bandwidth filter with a central wave-length of 1064 nm and 3 dB bandwidth of �5 nm was insertedinto the cavity. A total length of �180 m HI-1060 Flex single modefiber (SMF) was also incorporated to the cavity. The total cavitylength was �185 m. The monitoring of the output temporal pulsetrains and optical spectrum were performed using a 20-GHzmixed signal oscilloscope (Tektronix MSO72004C) with a 45-GHz

20 30 40 50 60 70 80

Inte

nsit

y(a.

u.)

2θ(degree)

006

015

1010

110

0015

205 1016

0210

1115

JCPDS No.15-0863

Fig. 1. (a) SEM image of the Bi2Te3 nano-sheets. (b) The corresponding XRD diffraction pattern.

FiberLD

Bi2Te3 / distilled waterdispersion

Fig. 2. Schematic diagram of optical deposition for TI:Bi2Te3. Inset: Optical image ofthe fiber connector end-facet after deposition.

Y. Peiguang et al. / Optics Communications 335 (2015) 65–7266

photo-detector (Newport 1014) and an optical spectrum analyzer(OSA, AQ6370B) with a minimum resolution of 0.02 nm, respec-tively. The pump power and output power were monitored by aphotodiode power meter (COHERENT).

3. Experimental result and analyses

3.1. Fundamental mode-locking

The laser threshold pumping power for CW operation was�110 mW. At the pump power of �160 mWwith an average outputpower �1.41 mW, the fundamental mode-locking operation wasachieved by appropriately adjusting the orientation of PC. As shownin Fig. 5(a), the pulse train had a cavity round-trip time of �905.6 ns,corresponding to the fundamental repetition rate of �1.10 MHz andthe total cavity length of �185 m. The Fig. 5(b) showed that the fullwidth at half maximum (FWHM) of the single pulse profile was�1.11 ns. The corresponding optical spectrum was also measured, asshown in Fig. 5(c). It had a central wavelength of �1064.3 nm anda 3-dB spectral bandwidth of �1.38 nm. In addition, the opticalspectrum showed the typical steep spectral edges was a result ofshock wave resonance in all-normal dispersion fiber lasers, indicatingthat the mode-locked pulses had been shaped to dissipative solitons[55]. The corresponding radio frequency (RF) spectrum was mea-sured by a Mixed Domain Oscilloscope (Tektronix MDO4034B-3).Fig. 5(d) showed the fundamental frequency with a signal-to-noiseratio (SNR) of �64 dB, and it was clearly shown that the repetitionrate of the pulse train was �1.10 MHz (the resolution bandwidth(RBW) was 2.50 kHz).

3.2. Disordered multi-pulses

By carefully tuning the orientation of PC, pulse splitting wasfound to occur owing to the peak power-limiting effect [3]. Thecorresponding pump power and average output power were�185 mW and �2.24 mW, respectively. Disordered multi-pulseshad been observed at this pump power, in which pulses were

randomly distributed in the cavity. Fig. 6(a) showed a typicalexample, where two individual pulses coexist in the cavity. Thesetwo pulses were unequally separated. It was worth mentioningthat this operation state was steady and static. Another type ofdisordered multi-pulse state could be also observed by adjustingthe PC, as shown in Fig. 6(c). More and more new pulses occupiedeither the entire or part of the space within the cavity. It was foundthat the pulses randomly moved with respect to each other in thecavity. In order to understand its origin, the corresponding opticalspectra were measured as shown in Fig. 6(b) (corresponding to thepulse trains of Fig. 6(a) and (d)) (corresponding to the pulse trainsof Fig. 6(c)). It was clearly seen that there were CW spectralcomponent appeared on the optical spectrum. The existence ofCW spectral component also played an important role in theformation of the disordered multiple pulses. The CW componentcould introduce a kind of global pulse interaction forces that linkwith all the pulses.

Since different pulses were located at different wavelengths,the pulse trains would have different group velocities andround-trip times because the cavity dispersion was nonzero. Inother words, the distribution of these pulses along the cavitybecame disordered and the pulses were in constant motion withrespect to each other [56]. In addition, as could be also clearlyseen in Fig. 6(a)–(d), the CW component was represented by theappearance of background noise, just liked the report in Refs.[57,58]. Increasing the strength of the CW component, theinstability of the pulse trains became more obvious, and conse-quently the noise background disappeared from the oscilloscopetrace as shown in Fig. 6(c). It would be formed as a stableirregular distribution pattern without the noise background inthe cavity (as shown in Fig. 6(a)).

3.3. Bunch of pulses

To form the state of bunch of pulses, the pump power needed tobe increased to �190 mW with an output power of �2.33 mW.Fig. 7(a) showed the oscilloscope trace in which two optical pulseswere grouped in a tight packet, and the bunch repeated with afundamental repetition rate of �1.10 MHz. The corresponding opticalspectrum was shown in Fig. 7(d). In compare with Fig. 6(d), it wasclearly shown that the CW spectral component still appeared on thespectrum, but the strength decreased. In addition, we found anoticeable phenomenon that the number of pulses in the bunchwould be increased from 2 to 6 by fixing the orientation of the PC butfurther increasing the pump power slightly. There were 3, 4, 5 and6 pulses coexisting in the bunch at the pump power of �198 mW,�204 mW, �209 mW and �213 mW, respectively. The correspond-ing output power was �2.49 mW, �2.67 mW, �2.75 mW and�2.83 mW, respectively. Fig. 7(b) and (c) showed the correspondingoscilloscope traces at a pump power of �198 mW and �213 mW.On the other hand, an inverse progress has also been explored: thepulse number could be decreased one by one by slightly decreasingthe pump power. There were two pulses coexisting in the bunch ifthe pump power was decreased down to �186 mW. Eventually, thefundamental mode-locking state could be maintained at a pumppower of �155 mW, which was below �160 mW due to hysteresisphenomena [59,60].

3.4. Soliton rains

If the pump power was further increased by adjusting the PC inan appropriate position, an interesting dynamic pattern statecould be obtained in the cavity, which was called soliton rain.The typical case with a pump power of �288 mW was given inFig. 8(a), and the corresponding output power was �6.16 mW. Itcould be found that some small drifting pulses drifted (right to

0 50 100 150 200 250 300 350

0.20

0.25

0.30

0.35

0.40

0.45

Isat ~24.6 MW/cm2

ns ~ 22.8%

0 ~ 16.2%

Experimental dataNonlinear fitting

Abs

orpt

ion

Input Intensity MW/cm2

a

a

Fig. 3. Nonlinear absorbance of the TI:Bi2Te3.

PCSMF

OC

Filter

PI-ISO

YDF

TI-SA

WDM976 LD

Fig. 4. Schematic diagram of the fiber ring laser.

Y. Peiguang et al. / Optics Communications 335 (2015) 65–72 67

left) toward the so-called condensed phase, collided with it, andspontaneously and randomly arose from the noisy backgroundagain after the collision. This behavior was similar to that reportedin [9–11]. The condensed state span was a single pulse with theFWHM of �5.96 ns, and the whole process could repeat in a quasi-stationary state while keeping the experimental conditionsunchanged. The corresponding optical spectrum had a centerwavelength of �1062.79 nm and a 3-dB bandwidth of

�0.82 nm, as given in Fig. 8(b). In this operation state, we alsofound three noticeable phenomena. The first was that the con-densed phase would disappear with an appropriate orientation ofthe PC. This typical case with the same pump power of �288 mWwas given in Fig. 8(c), the condensed phase disappeared butdrifting pulses existed, and the drifting direction was unchanged.However, compared with media 1, the drift speed of the driftingpulses here was much slower. The second phenomenon was that

-4 -2 0 2 4

0.00

0.05

0.10

Inte

nsity

(a.u

)

Time ( s) Time ( s)-0.005 0.000 0.005

0.00

0.02

0.04

Inte

nsity

(a.u

)1055 1060 1065 1070

-80

-60

-40

-20

Inte

nsity

(dB

m)

Wavelength(nm)0.8 1.0 1.2 1.4 1.6 1.8

-120

-100

-80

-60

-40

RF

Pow

er (d

Bm

)

Frequency (MHz)

~64 dB

~1.10 MHzSpan: 1.00 MHzRBW: 2.50 kHz

Fig. 5. (a) The fundamental mode-locking pulse train, (b) single pulse width, (c) the corresponding spectrum, (d) the fundamental frequency signal and the inset ofcorresponding RF spectrum with a span range of 50 MHz.

-3 -2 -1 0 1 2 3

0.00

0.02

0.04

0.06

Inte

nsity

(a.u

)

round-trip time

1055 1060 1065 1070-80

-60

-40

-20

0

Inte

nsity

(dB

m)

Wavelength(nm)

-3 -2 -1 0 1 2 3

0.00

0.02

0.04

round-trip time

Inte

nsity

(a.u

)

1060 1065 1070-80

-60

-40

-20

0

Inte

nsity

(dB

m)

Wavelength(nm)Time ( s)

Time ( s)

Fig. 6. (a) The oscilloscope trace of disordered multi-pulses and (b) their optical spectrum. (c) Another type of disordered multi-pulses (d) their optical spectrum.

Y. Peiguang et al. / Optics Communications 335 (2015) 65–7268

the number of drifting pulses could be controlled depending onthe orientation of the PC and the pumping power, as shown inFig. 8(d). The condensed phase existed and the drifting directionwas also unchanged. However, the number of drifting pulses wasonly one, and the drifting pulses drifted much faster than theabove two operation states. In this states, the pump power was�293 mW with the output power of �6.50 mW. The last phe-nomenon was that the condensed phase could be changed from asingle pulse to a bunch by adjusting the PC carefully with thepump power of 293 mW. Fig. 8(e) showed this case, two pulsescould bunch together, and the two pulses had the same peakpowers and stable relative positions in the bunch. The number ofdrifting pulses was also only one, and the drift speed of thedrifting pulses here (right to left) was the same as the thirdphenomenon.

3.5. Quasi-square pulse

In our experiment, besides the fundamental mode-locking, a trainof mode-locked quasi-square pulse could be also observed at thepump power of �164mW with an appropriate PC setting. Fig. 9(a) showed the typical oscilloscope trace under the pump power of�180mW. As exhibited in the inset (left) of Fig. 9(a), indicating thatthe pulse had a quasi-square profile structure, with a FWHM of�8.22 ns. The corresponding optical spectrum of the quasi-squarepulse was shown in the inset (right) of Fig. 9(a), it could be seen thatweak CW spectral component appeared in the spectrum. The evolu-tion of quasi-square pulse had also been experimentally investigated.The duration of quasi-square pulse could be changed by increasing thepump power while keeping the PCs unchanged. The evolution ofsingle pulse width with respect to pump power was given in Fig. 9(b).At the pump powers of 170 mW, 220mW, 260mW, 300mW and320mW, the corresponding pulse durations were �7.76 ns,�12.98 ns, �21.57 ns, �27.60 ns and �33.74 ns, respectively. It wasclearly seen that the output pulse duration became broader with theincrease of pump power. The corresponding evolution of the optical

spectra exhibited that the spectral intensities increased slightly, andthe weak CW spectral component would disappear with the increaseof pump power, as shown in Fig. 9(c). In order to better show thecharacteristics of quasi-square pulses, Fig. 9(d) showed the variationsof quasi-square pulse duration and output power with respect to thepump power. We found that both of them monotonically increasewith the pump power. The narrowest pulse was �5.92 ns at thepump power of �164mW, while the maximum pulse duration was�33.74 ns at the pump power of �320mW. The largest outputpower was �7.99 mW at the pump power of �320mW. The outputpulse energy with respect to the pump power was given in Fig. 9(e).The pulse energy also increased with the pump power, and the largestoutput pulse energy was 7.23 nJ under the pump power of �320mW.The above behavior was similar to the dissipative soliton resonancewhich reported in [61,62]. The theory of dissipative soliton resonanceindicated that, if the parameters were chosen closer to the resonancepoint, the pulse energy could be enhanced to an infinitely large valueand the pulse duration could increase with the pump power, whereasthe pulse amplitude converged to a given plateau value. Severalexperimental observations of dissipative soliton resonance in fiberlasers have been reported [63–65]. The square-profile dissipativesolitons whose energy could increase to a very large value withoutpulse breaking were obtained in their experiment. However, the quasi-square pulses here would be suffered pulse break up at a higher pumppower. In addition, the pump was increased, the pulse durations wereincreased and the pulse energy was also increased almost 6 times, butthe pulse amplitude did not change or change very little.

Under strong pumping, the positive cavity feedback of the laserbecame strong. In the case, the bright pulses formed in a mode-lockedfiber laser would have the tendency of bunching together. As shown inFig. 9(a) and (b), it is clearly that the non-uniform modulationappeared on top of the pulse. Although the pulse duration of thesingle pulse is equal to 1 ns, it is also uncertain that the structure withsuch duration can be accurately measured on the top of the pulsebunch by ordinary speed oscilloscope and fast photo-detector if thespacing of the pulses was much closer. It may be clearly observed if

-1 0 1 20.00

0.03

0.06

0.09

Inte

nsity

(a.u

)-0.02 0.00 0.02

Inte

nsit

y (a

.u)

Time ( s)

P=190 mW

-1 0 1 20.00

0.03

0.06

0.09

-0.02 0.00 0.02 0.04

Inte

nsit

y (a

.u)

Time ( s)

Inte

nsity

(a.u

)

P=198 mW

-1 0 1 2

0.00

0.03

0.06

0.09

Inte

nsity

(a.u

)

P=213 mW

-0.1 0.0 0.1

Inte

nsity

(a.u

)

Time ( s)

1060 1065 1070-80

-60

-40

-20

0

Inte

nsity

(dB

m)

Wavelength(nm)Time ( s)

Time ( s) Time ( s)

Fig. 7. (a) The bunch of pulse at the pump power of �190 mW. (b) �198 mW. (c) �213 mW. (d) The optical spectrum.

Y. Peiguang et al. / Optics Communications 335 (2015) 65–72 69

the pulses are detected by a much higher speed oscilloscope and amuch faster photo-detector. Based on these, we believe this was not asingle pulse, but rather the plenty of pulses tightly bound together andtheir number predictably increased with pump rate.

In the all-normal-dispersion cavity, we incorporated the extra fiberof �185m into the cavity, which meant introducing into moreadditional linear dispersion and nonlinearity, and it is required forthe extreme pulse broadening [66,67]. Thus it could explain why thefundamental mode-locking regime here was larger than 1 ns.

Because we are devoted to the experimental research of multi-pulse dynamic patterns, some pulse compression mechanisms werenot applied in the cavity. If pulse compression mechanisms wereapplied, we need shorter cavity lengths or standard compressionformats, such as photonic crystal fibers or bulk gratings. However, themain purposes of these above measures are to decrease lineardispersion and nonlinearity, which are not conducive to obtainmulti-pulses. It was demonstrated that introducing proper high non-linear effect into the laser cavity was favorable for generating multi-pulses [48]. Due to the highly nonlinear effect of TI SA, it has beenobtained one kind of multi-pulse operation states in an erbium-dopedfiber laser with a microfiber-based TI SA in Ref. [68]. By properly

adjusting the cavity parameters, the different dual-wavelength har-monic mode locking (HML) orders were achieved there. However,besides the highly nonlinear effect of TI SA, the long cavity and theinsertion of the 5-nm narrow bandwidth filter also played animportant role in the formation of the multi-pulses in our experiment.The �185m long laser cavity here introduced into more additionallinear dispersion and nonlinearity, which leaded to higher single pulseenergy in the mode-locking state with a certain pump powercompared with the case using a short laser cavity. Therefore, in thelong laser cavity, the multi-pulses would be relatively easy to obtaindue to the peak power clamping effect [3] and soliton energyquantization [69]. In fact, compared with erbium-doped fiber laser, itwas difficult for ytterbium-doped fiber ring laser to show the gain-limiting effect due to the broader gain bandwidth of ytterbium and thelarger gain saturation power. Hence the insertion of the 5-nm narrowbandwidth filter played an important role in the formation of themulti-pulses [2]. The single-pulse energy would be restricted whenthe gain bandwidth was limited by a bandwidth filter, which wouldinduce the pulse to break under the higher pump powers. Addition-ally, the effective gain bandwidth of the laser also depended on theartificial fiber birefringence filter [70–72]. Usually, the cavity

-1 0 1 2

0.00

0.02

0.04

Inte

nsity

(a.u

)

Drifting

pulsesCondensed

phase

round-trip time

Direction

1055 1060 1065 1070-80

-60

-40

-20

0

Inte

nsity

(dB

m)

Wavelength(nm)

-0.5 0.0 0.5 1.0

0.00

0.02

0.04

Direction

Drifting

pulses

round-trip time

Inte

nsity

(a.u

)

-1 0 1 2

0.00

0.02

0.04

0.06

Drifting

pulses

Direction

Condensed

phase

round-trip time

Inte

nsity

(a.u

)

-1 0 1 2

0.00

0.02

0.04

0.06

Drifting

pulses

Direction

Condensed

phase

round-trip time

Inte

nsity

(a.u

)

Time ( s)

Time ( s)Time ( s)

Time ( s)

Fig. 8. (a) Temporal trace of soliton rains. (b) Corresponding spectrum of soliton rains. (c) First temporal trace of noticeable phenomenon. (d) Second temporal trace ofnoticeable phenomenon. (e) Third temporal trace of noticeable phenomenon.

Y. Peiguang et al. / Optics Communications 335 (2015) 65–7270

birefringence-induced artificial birefringence filtering effect couldnormally be ignored due to the large filtering bandwidth in theweakly birefringent fiber cavities. However, strong cavity birefringencecould be introduced into our fiber laser, meaning that the artificialbirefringence filter effect of the laser cavity was no longer ignorable.The strength of the cavity birefringence could be changed by tuningthe intra-cavity PCs, resulting in the change of the effective cavitytransmission bandwidth, which could be viewed as variation in thebandwidth of the gain in the laser. Hence the effective gain bandwidthof the laser could be determined by both the 5-nm bandwidth filterand the artificial fiber birefringence filter. As a result, differentoperation states could be obtained by changing the pump powerand the PCs.

4. Conclusions

In conclusion, we reported on the generation of multi-pulseoperation states from an all-normal dispersion ytterbium-doped

fiber ring laser by TI:Bi2Te3 SA that was sandwiched betweenoptical fiber connectors through the optical deposition approach.We not only experimentally studied the fundamental mode-locking and multi-pulse dynamic patterns, but also investigatedthe formation of the quasi-square pulses. The quasi-square pulseshad the maximum pulse duration of �33.74 ns and the largestoutput pulse energy of 7.23 nJ at the pump power of �320 mW.Our work has verified that TI might be developed as a saturableabsorber device for the generation of various multiple pulseoperation states in an all-normal dispersion laser.

Acknowledgments

This research was supported by NSFC (61007054&61275144), theNatural science fund of Guangdong province (S2013010012235),the Improvement and Development Project of Shenzhen Key Lab(ZDSY20120612094924467), the Science and technology project ofShenzhen City (JCYJ20120613172042264, JCYJ20130329142040731),Natural Science Foundation of SZU (no. 201221).

-3 -2 -1 0 1 2 30.00

0.04

0.08

0.12

1060 1065

Inte

nsity

(dB

m)

Wavelength(nm)-0.01 0.00 0.01 0.02

Inte

nsity

(a.u

)

Time ( s)

8.22 ns

Inte

nsity

(a.u

)

-0.04 -0.02 0.00 0.02 0.04 0.06

320300

260220

170

Time ( s)

Inte

nsity

(a.u

)

Pum

p Po

wer (m

W)

1060 1065-80

-60

-40

-20

0

Inte

nsity

(dB

m)

Wavelength(nm)

170 mW220 mW260 mW300 mW320 mW

150 180 210 240 270 300 3300

5

10

15

20

25

30

35

Pump power (mW)

Puls

e D

urat

ion

(ns)

0

2

4

6

8

10

12

14

150 180 210 240 270 300 330

2

4

6

8

Pump power (mW)

Time ( s)

Fig. 9. (a) Oscilloscope trace of quasi-square pulses. Insets: Spectrum (right) and a single quasi-square pulse (left). (b) Quasi-square pulses under different pump powers.(c) Corresponding spectra of quasi-square pulses under different pump powers. (d) Quasi-square pulse durations and output powers versus the pump powers. (e) Pulseenergies versus the pump powers.

Y. Peiguang et al. / Optics Communications 335 (2015) 65–72 71

References

[1] P. Grelu, J.M. Soto-Crespo, J. Opt. B: Quantum Semiclassical Opt. 6 (5) (2004)S271.

[2] A. Haboucha, A. Komarov, H. Leblond, F. Sanchez, G. Martel, Opt. Fiber Technol.14 (4) (2008) 262.

[3] D.Y. Tang, L.M. Zhao, B. Zhao, A.Q. Liu, Phys. Rev. A: At. Mol. Opt. Phys. 72(2005) 43816.

[4] P. Grelu, J.M. Soto-Crespo, Temporal soliton “molecules” in mode-lockedlasers: collisions, pulsations, and vibrations, in: Dissipative Solitons: FromOptics to Biology and Medicine, vol. 751, Springer (2008) 1.

[5] B.A. Malomed, Bound solitons in the nonlinear Schrödinger/Ginzburg-Landauequation, in: Large Scale Structures in Nonlinear Physics, vol. 392, Springer(1991) 288.

[6] P. Grelu, F. Belhache, F. Gutty, J. Soto-Crespo, Opt. Lett. 27 (11) (2002) 966.[7] F. Amrani, A. Haboucha, M. Salhi, H. Leblond, A. Komarov, F. Sanchez, Appl.

Phys. B 99 (2010) 107.[8] F. Amrani, M. Salhi, H. Leblond, F. Sanchez, Opt. Commun. 283 (24) (2010)

5224.[9] S. Chouli, P. Grelu, Phys. Rev. A: At. Mol. Opt. Phys. 81 (2010) 63829.[10] C. Bao, X. Xiao, C. Yang, Opt. Lett. 38 (11) (2013) 1875.[11] Q.Y. Ning, H. Liu, X.W. Zheng, W. Yu, A.P. Luo, X.G. Huang, Z.C. Luo, W.C. Xu, S.

H. Xu, Z.M. Yang, Opt. Express 22 (10) (2014) 11900.[12] Z.X. Zhang, L. Zhan, X.X. Yang, S.Y. Luo, Y.X. Xia, Laser Phys. Lett. 4 (8) (2007)

592.[13] F. Amrani, A. Haboucha, M. Salhi, H. Leblond, A. Komarov, P. Grelu, F. Sanchez,

Opt. Lett. 34 (14) (2009) 2120.[14] H. Chen, K. Lin, C. Tsai, H. Wu, C. Wu, C. Chen, Y. Chi, G. Lin, W. Hsieh, Opt. Lett.

38 (6) (2013) 845.[15] S. Zhou, D.G. Ouzounov, F.W. Wise, Opt. Lett. 31 (8) (2006) 1041.[16] N.H. Seong, D.Y. Kim, Opt. Lett. 27 (15) (2002) 1321.[17] R. Gumenyuk, O.G. Okhotnikov, Laser Phys. Lett. 10 (5) (2013) 55111.[18] X. Wu, D.Y. Tang, X.N. Luan, Q. Zhang, Opt. Commun. 284 (14) (2011) 3615.[19] S.Y. Set, H. Yaguchi, Y. Tanaka, M. Jablonski, J. Lightwave Technol. 22 (1) (2004)

51.[20] S. Yamashita, Y. Inoue, S. Maruyama, Y. Murakami, H. Yaguchi, M. Jablonski, S.

Y. Set, Opt. Lett. 29 (14) (2004) 1581.[21] X. Liu, D. Han, Z. Sun, C. Zeng, H. Lu, D. Mao, Y. Cui, F. Wang, Sci. Rep. 3 (2718)

(2013).[22] Y.F. Song, L. Li, H. Zhang, D.Y. Shen, D.Y. Tang, K.P. Loh, Opt. Express 21 (8)

(2013) 10010.[23] D. Popa, Z. Sun, F. Torrisi, T. Hasan, F. Wang, A.C. Ferrari, Appl. Phys. Lett. 97

(20) (2010) 203106.[24] Q. Bao, H. Zhang, Y. Wang, Z. Ni, Y. Yan, Z.X. Shen, K.P. Loh, D.Y. Tang, Adv.

Funct. Mater 19 (19) (2009) 3077.[25] Sun Young Choi, Dae Kun Cho, Yong-Won Song, Kyunghwan Oh, Kihong Kim,

Fabian Rotermund, Dong-Il Yeom, Opt. Express 20 (5) (2012) 5652.[26] Evgeni Sorokin Nikolai Tolstik, Irina T. Sorokina, Opt. Express 22 (5) (2014)

5564.[27] P.L. Huang, S.C. Lin, Y.Y. Chao, H.H. Kuo, S.H. Huang, G.R. Lin, L.J. Li, C.Y. Su, W.

H. Cheng., Opt. Express 20 (3) (2012) 2460.[28] J. Liu, S.D. Wu, Q.H. Yang, P. Wang, Opt. Lett. 36 (20) (2011) 4008.[29] Y.C. Meng, S.M. Zhang, X.L. Li, H.F. Li, J. Du, Y.P. Hao, Opt. Express 20 (6) (2012)

6685.[30] Z. Sun, T. Hasan, F. Torrisi, D. Popa, G. Privitera, F. Wang, F. Bonaccorso, D.

M. Basko, A.C. Ferrari, ACS Nano. 4 (2) (2010) 803.[31] Joel E. Moore, Nature 464 (11) (2010) 194.[32] H. Zhang, C.X. Liu, X.L. Qi, X. Dai, Z. Fang, S.C. Zhang, Nat. Phys 5 (2009) 438.[33] Y. Zhang, K. He, C.Z. Chang, C.L. Song, L.L. Wang, X. Chen, J.F. Jia, Z. Fang, X. Dai,

W.Y. Shan, S.Q. Shen, Q. Niu, X.L. Qi, S.C. Zhang, X.C. Ma, Q.K. Xue, Nat. Phys. 6(2010) 584.

[34] N. Bansal, Y.S. Kim, E. Edrey, M. Brahlek, Y. Horibe, K. Iida, M. Tanimura, G.H. Li,T. Feng, H.D. Lee, Thin Solid Films 520 (1) (2011) 224.

[35] L.P. Sun, Z.Q. Lin, J. Peng, J. Weng, Y.Z. Huang, Z.Q. Luo, Sci. Rep. 4 (2014) 4794.[36] S.Q. Chen, C.J. Zhao, Y. Li, H.H. Huang, S.B. Lu, H. Zhang, S.C. Wen, Opt. Mater.

Express 4 (4) (2014) 587.[37] C. Zhao, H. Zhang, X. Qi, Y. Chen, Z. Wang, S. Wen, D. Tang, Appl. Phys. Lett. 101

(21) (2012) 211106.[38] J. Sotor, G. Sobon, K.M. Abramski, Opt. Express 22 (11) (2012) 13244.[39] Z.H. Yu, Y.R. Song, J.R. Tian, Z.Y. Dou, H.Y. Guoyu, K.X. Li, H.W. Li, X.P. Zhang,

Opt. Express 22 (10) (2014) 11508.[40] Z. Luo, Y. Huang, J. Weng, H. Cheng, Z. Lin, B. Xu, Z. Cai, H. Xu, Opt. Express 21

(24) (2013) 29516.[41] Z. Luo, M. Liu, H. Liu, X. Zheng, A. Luo, C. Zhao, H. Zhang, S. Wen, W. Xu, Opt.

Lett. 38 (24) (2013) 5212.[42] Y. Chen, C. Zhao, H. Huang, S. Chen, P. Tang, Z. Wang, S. Lu, H. Zhang, S. Wen,

D. Tang, J. Lightwave Technol. 31 (2013) 2857.[43] Z. Luo, C. Liu, Y. Huang, D. Wu, J. Wu, H. Xu, Z. Cai, Z. Lin, L. Sun, J. Weng,

Quantum Electron. 20 (2014) 902708.[44] C. Zhao, Y. Zou, Y. Chen, Z. Wang, S. Lu, H. Zhang, S. Wen, D. Tang, Opt. Express

20 (25) (2012) 27888.[45] J. Lee, J. Koo, Y. Jhon, J.H. Lee, Opt. Express 22 (5) (2014) 6165.[46] H. Liu, X. Zheng, M. Liu, N. Zhao, A. Luo, Z. Luo, W. Xu, H. Zhang, C. Zhao,

S. Wen, Opt. Express 22 (6) (2014) 6868.[47] Y. Chen, C. Zhao, S. Chen, J. Du, P. Tang, G. Jiang, H. Zhang, S. Wen, D. Tang, IEEE

J. Sel. Top. Quantum Electron. 20 (5) (2013).[48] C.S. Jun, S.Y. Choi, F. Rotermund, B.Y. Kim, D.I. Yeom, Opt. Lett. 37 (11) (2012)

1862.[49] S. Lu, C. Zhao, Y. Zou, S. Chen, Y. Chen, Y. Li, H. Zhang, S. Wen, D. Tang, Opt.

Express 21 (2) (2013) 2072.[50] Y. Chen, M. Wu, P. Tang, S. Chen, J. Du, G. Jiang, Y. Li, C. Zhao, H. Zhang, S. Wen,

Laser Phys. Lett. 11 (5) (2014) 55101.[51] K. Kashiwagi, S. Yamashita, S.Y. Set, Opt. Express 17 (7) (2009) 5711.[52] A. Martinez, K. Fuse, B. Xu, S. Yamashita, Opt. Express 18 (22) (2010) 23054.[53] H. Kim, J. Cho, S. Jang, Y. Song, Appl. Phys. Lett. 98 (2) (2011) 21104.[54] M. Hajlaoui, E. Papalazarou, G. Mauchain, N. Lantz, D. Moisan, Z. Boschetto,

I. Jiang, Y.P. Miotkowski, A. Chen, L. Taleb-Ibrahimi, Perfetti, M. Marsi, NanoLett. 12 (7) (2012) 3532.

[55] C. Lecaplain, J.M. Soto-Crespo, P. Grelu, C. Conti, Opt. Lett. 39 (2) (2014) 263.[56] X. Liu, Phys. Rev. A: At. Mol. Opt. Phys. 84 (2011) 53828.[57] D.Y. Tang, B. Zhao, L.M. Zhao, H.Y. Tam, Phys. Rev. E: Stat. Nonlinear Soft Matter

Phys. 72 (2005) 16616.[58] B. Zhao, D.Y. Tang, P. Shum, W.S. Man, H.Y. Tam, Y.D. Gong, C. Lu, Opt. Commun.

229 (2004) 363.[59] X. Liu, Phys. Rev. A: At. Mol. Opt. Phys. 81 (2010) 23811.[60] A. Komarov, H. Leblond, F. Sanchez, Phys. Rev. A: At. Mol. Opt. Phys. 71 (2005)

53809.[61] W. Chang, A. Ankiewicz, J.M. Soto-Crespo, N. Akhmediev, Phys. Rev. A: At. Mol.

Opt. Phys. 78 (2008) 23830.[62] N. Akhmediev, J.M. Soto-Crespo, P. Grelu, Phys. Lett. A 372 (17) (2008) 3124.[63] X. Wu, D.Y. Tang, H. Zhang, L.M. Zhao, Opt. Express 17 (2009) 5580.[64] Z.C. Luo, W.J. Cao, Z.B. Lin, Z.R. Cai, A.P. Luo, W.C. Xu, Opt. Lett. 37 (2012) 4777.[65] L. Liu, J.H. Liao, Q.Y. Ning, W. Yu, A.P. Luo, S.H. Xu, Z.C. Luo, Z.M. Yang, W.C. Xu,

Opt. Express 21 (2013) 27087.[66] E.J.R. Kelleher1, J.C. Travers, Z. Sun, A.G. Rozhin, A.C. Ferrari, S.V. Popov1, J.

R. Taylor, Appl. Phys. Lett. 95 (2009) 111108.[67] E.J.R. Kelleher, J.C. Travers, E.P. Ippen, Z. Sun, A.C. Ferrari, S.V. Popov, J.R. Taylor,

Opt. Lett. 34 (22) (2009) 3526.[68] M. Liu, N. Zhao, H. Liu, X.W. Zheng, A.P. Luo, Z.C. Luo, W.C. Xu, C.J. Zhao,

H. Zhang, S.C. Wen, IEEE Photonics Technol. Lett. 26 (10) (2014) 983.[69] L.M. Zhao, D.Y. Tang, T.H. Cheng, H.Y. Tam, C. Lu, Opt. Lett. 32 (11) (2007) 1581.[70] H. Zhang, D.Y. Tang, X. Wu, L.M. Zhao, Opt. Express 17 (15) (2009) 12692.[71] A.P. Luo, Z.C. Luo, W.C. Xu, V.V. Dvoyrin, V.M. Mashinsky, E.M. Dianov, Laser

Phys. Lett 8 (8) (2011) 601.[72] X.H. Li, Y.S. Wang, W. Zhao, W. Zhang, Z. Yang, X.H. Hu, H.S. Wang, X.L. Wang,

Y.N. Zhang, Y.K. Gong, Laser Phys. 21 (5) (2011) 940.

Y. Peiguang et al. / Optics Communications 335 (2015) 65–7272