We report on a passively mode-locked erbium-doped fiber laser, using a high nonlinear modulation depth saturable absorber mirror, in a Fabry-Perot cavity. A segment of dispersion compensation fiber is added inside the cavity in order to build a high-positive dispersion regime. The setup produced highly chirped pulses with an energy of 1.8 nJ at a repetition rate of 33.5 MHz. Numerical simulations accurately reflect our experimental results and show that pulse-shaping in this laser could be interpreted as producing ‘dissipative solitons’.
©2009 Optical Society of America
Ultrashort pulses from femtosecond lasers are interesting for a wide range of applications, including micromachining, surgical medicine, imaging, optical telecommunications, etc . In recent years, fiber lasers have attracted significant interest with the achievement of high energy femtosecond pulses. In particular, fiber lasers operating in the 1-µm range have shown significant progress due to the exceptional efficiency and gain bandwidth of Yb-doped fibers. Energy per pulse of more than 100 nJ has been obtained in the 1-µm range [2, 3].
Energy scaling in mode-locked fiber lasers depends strongly on the cavity dispersion management, and in particular in the so-called all-normal dispersion regime (ANDi). Several practical cavity designs have been both theoretically and experimentally tested for the few years, leading to a significant improvement in the extracted energy from fiber oscillators [4-8]. Herda et al. have reported the first Yb-doped fiber laser operating in purely normal group velocity dispersion (GVD) regime by using a high modulation depth saturable absorber mirror (SAM) . Pulse shaping was attributed to the combined effects of the normal fiber GVD and SAM nonlinearity. Recently, Chong et al. demonstrated a Yb-doped femtosecond fiber laser without anomalous dispersion component, in which pulse shaping arose from the spectral filtering of a highly-chirped pulse in the laser cavity, leading to the achievement of more than 20 nJ energy . A wide variety of spectral and temporal shapes could be obtained with variation of the filter bandwidth  and have been interpreted in terms of highly chirped ‘dissipative solitons’ in the frame of the cubic-quintic Ginzburg-Landau equation [12-14].
In the 1.5-µm range, such a design remains a challenge since the majority of fibered components generally present an anomalous GVD at such a wavelength. Dispersion compensation fiber (DCF) offers other alternative to implement such highly normal regime. Indeed, due to the adapted index profile and core diameter, such fibers present a positive GVD at 1.5 µm. Zhao et al. have reported on the dynamics of the so-called “gain-guided soliton” (GGS) in a passively mode-locked Er-doped fiber (EDF) laser, made of purely normal dispersion fibers [15-17]. Cabasse et al. have recently demonstrated that the apparent lack of dechirping capabilities for these GGS could be avoided and that these regimes belong to the more general family of dissipative solitons .
In this contribution, we report on a passively mode-locked EDF laser, with large net cavity dispersion using a segment of DCF inside the Fabry-Perot cavity. Passive mode-locking is achieved using a high modulation depth SAM. The laser generates 60 mW of average power, corresponding to pulse energy of 1.8 nJ.
2. Laser setup and experimental results
The experimental setup of the passively mode-locked fiber laser operating in a Fabry-Perot configuration is illustrated in Fig. 1. A highly doped EDF serves as gain medium . It presents unpumped peak absorption of 80 dB/m at 1530 nm. Its GVD has been estimated around +0.061 ps2/m at 1550 nm. The gain fiber is pumped through a multiplexer using standard fiber (single mode Hi1060/D=+8.7 ps/nm/km). A segment of dispersion compensation fiber (DCF) has been added in the cavity in order to build a highly normal dispersion cavity. Its GVD has been estimated around +0.116 ps2/m. Rest of the cavity comprises a 50/50 coupler (SMF28 fiber/D=+17.7 ps/nm/km), coupling lenses and a saturable absorber mirror. The total optical cavity length is about 3.1 meters, leading to a repetition rate of 33.5 MHz. The net positive cavity dispersion is β2=+0.19 ps2 at 1550 nm.
Passively mode-locking is achieved employing a SAM as one of the cavity mirrors. Its epitaxial structure has been made by the LPN laboratory . The device shown schematically in Fig. 2, consists of an active layer of 7 InGaAs/InAlAs quantum-wells (QWs), grown by metal-organic vapour-phase epitaxy upon an InP substrate, and incorporated in a Fabry-Perot microcavity. The back mirror was made by deposition of a thin silver layer and the front mirror consists on a Bragg mirror, which enables the impedance matching, Fig. 2(a). This SAM presents a low intensity reflectivity of 12%, a modulation depth of 37% and a saturation fluence of 17 µJ/cm2 (Fig. 2(b)). Time-resolved pump-probe experiments at 1555 nm showed temporal relaxation around 2.3 ps.
The laser starts in a Q-switched regime for a pump power of 10 mW and self-starts in a pure mode-locking regime when pump power reaches 327 mW. The single-pulse mode-locking operation is sustained up to 750 mW (i.e. max pump power). The measured average output power at maximum pump power is 60 mW, which correspond to pulse energy of 1.8 nJ. The autocorrelation trace measurement shows that the pulse duration is 10.3 ps, assuming a Gaussian shape, Fig. 3(a). The output pulses have been extra-cavity dechirped with a pair of 800 lines/mm free-space gratings down to 528 fs, Fig. 4(a). The corresponding time-bandwidth product is 0.77. It is about 1.7 times higher than the Fourier-Transform (FT) limit (0.44) and indicates that the output pulses suffer from a non-negligible amount of nonlinear chirp. The optical spectrum, Fig. 3(b), shows that pulse duration below 300 fs could be obtained within the FT limit considering an optimization of the nonlinear chirp compensation . To evaluate the quality of the mode-locked pulse train, we performed amplitude noise analyses using the radio-frequency power spectrum obtained with a microwave spectrum analyzer via a high-speed photodetector (8-GHz bandwidth). Figure 4(b) shows the fundamental harmonic centred at 33.5 MHz, with a span of 500 kHz and a resolution bandwidth of 300 Hz. The noise level is lower than 0.1%, which highlights the good stability of the output pulse train that did not suffer from Q-switch mode-locking instabilities.
3. Numerical simulations
In order to obtain more information on the intra-cavity evolution of both the temporal and spectral shape, we numerically simulated our laser cavity using the Non Linear Schrödinger Equation formalism, . Our code describes the propagation of the slowly-varying envelope function A(z,t) (where t is the time and z is the distance along the axis of propagation) along each nonlinear dispersive element. Furthermore, we implement a real saturable absorption function . The parameters used in the simulation for each element are reported in Table 1.
Numerical results close to experiments are obtained, for a small-signal gain coefficient g0=4.4 m-1 and a saturation energy Esat=610 pJ, with an output pulse energy of about 1.77 nJ. A 25 nm width Gaussian profile is assumed for the second Er gain band centred on 1560 nm. Figure 5(a) shows the intra-cavity evolution of the temporal and spectral pulse widths (FWHM values) over one cavity round-trip, while Fig. 5(b) gives the similar evolution for the energy of the single-pulse. Figures 5(c) and 5(d) show the temporal intensity profiles (before and after compression) and the output power spectrum for the pulse extracted by the output coupler and before arriving on the saturable absorber (SA) i.e. for energy of 1.77 nJ. Pulses are always positively chirped inside the cavity with one minimum duration located inside the gain fiber. The most significant insight gained by such numerical computations is to show that gain filtering effects always play a major role in the self-consistency of the pulse over one round-trip for such quasi all-normal cavity dispersion regimes. However we have checked that high modulation depth of the SAM (typ. greater than 30%) is essential for convergence of the regime. These results outline and push forward the power scalability limits of the so-called dissipative soliton regime in erbium cavities as already detailed recently .
In Figs. 5(c) and 5(d) we note that the theoretical widths are larger than the experimental ones. We attribute this discrepancy to our too simple model which implements a constant value for the saturation energy all along the doped fiber . Indeed we recall that within the gain expression of a three-level medium, g(z)=g0/[1+E[z]/Esat], the small signal gain coefficient g0 could be related to the doping concentration and E(z) is the pulse energy while Esat parameter is pump-power dependent but considered here as z-independent. However in a real configuration, pump power is injected from one end of the doped fiber rendering the parameter Esat as z-dependent. We should note that the 100% reflective gold mirror (Fig. 1) should modify the monotonic z-dependence in a complicated way due to reflection of the residual unabsorbed pump power. Additional calculations are under progress to better simulate the gain dependence of the laser cavity. Nonetheless we can consider that Figs 5 agree quite well with our experimental results, both spectrally (essentially in shape) and temporally (both for the chirped and dechirped pulse). It is worth noting that the Esat value of 610 pJ is not critical, as we did not observe a drastic change for the characteristics of the pulse when varying Esat e.g. from 600 pJ to 700 pJ. Indeed for Esat=700 pJ, the spectral width and the energy of the pulse increase and the chirped and dechirped pulse widths decrease of about only 10 %. Within the configuration of Fig. 1 and the parameters of Table 1, we could predict a maximum output energy of 2.4 nJ from the output coupler when Esat≈900 pJ. Above this value pulses become unstable essentially due to a too large pulse spectrum bandwidth. Nevertheless another possibility exists in our case to overcome power scalability limits, i.e. with the same SAM and the same highly normal active fiber but containing a higher doping concentration (and so higher g0 gain value up to 6 m-1). It consists to completely remove anomalous fibers considering WDM and coupler made-up of DCF fibers. Simulations are reported in Figs 6 and they predict that more than 8 nJ output energy per pulse are feasible with such a configuration. Realization of such a cavity is currently under progress and experimental results for such an all-normal fibered laser cavity with a high modulation depth SAM will be reported in a forthcoming paper.
In conclusion, we have reported the generation of ultra-short pulses in a highly positive dispersion regime at 1.55 µm, using a Fabry-Perot configuration. The regime is fully self-starting thanks to a high modulation depth SAM and delivered 1.8 nJ pulse energy. Numerical simulations accurately reflect our experimental results and show that this laser operates in the dissipative soliton regime. In addition a breakthrough in the power scalability of such net normal dispersion Er-doped fiber lasers has been also predicted.
This work was supported by the European Agency and French National Agency of Research (ANR) under contract ANR-06-NSCI-006 (S-Five project).
References and links
1. W. Chang, A. Ankiewicz, J. M. Soto-Crespo, and N. Akhmediev, “dissipative soliton resonances,” Phys. Rev. A 78, 023830 (2008). [CrossRef]
2. B. Ortaç, O. Schmidt, T. Schreiber, J. Limpert, A. Tünnermann, and A. Hideur, “High-energy femtosecond Yb-doped dispersion compensation free fiber laser,” Opt. Express 15, 10725–10732 (2007). [CrossRef] [PubMed]
3. C. Lecaplain, C. Chédot, A. Hideur, B. Ortaç, and J. Limpert, “High-power all-normal-dispersion femtosecond pulse generation from an Yb-doped large-mode-area microstructure fiber laser,” Opt. Lett. 32, 2738–2740 (2007). [CrossRef] [PubMed]
5. T. Schreiber, B. Ortaç, J. Limpert, and A. Tünnermann, “On the study of pulse evolution in ultra-short pulse mode-locked fiber lasers by numerical simulations,” Opt. Express 15, 8252–8262 (2007). [CrossRef] [PubMed]
7. G. Bale, J. N Kutz, A. Chong, W. H. Renninger, and F. W. Wise, “Spectral filtering for high-energy mode-locking in normal dispersion fiber lasers,” J. Opt. Soc. Am. B 251763–1768 (2008). [CrossRef]
9. R. Herda and O. G. Okhotnikov, “Dispersion compensation-free fiber laser mode-locked and stabilized by high-contrast saturable absorber mirror,” IEEE J. Quantum Electron. 40893–899 (2004). [CrossRef]
11. W. H. Chong, F. Renninger, and Wise, “Properties of normal-dispersion femtosecond fiber lasers,” J. Opt. Soc. Am. B 25, 140–148 (2008). [CrossRef]
12. J. D. Moores, “On the Ginzburg-Landau laser mode locking model with fifth-order saturable absorber term,” Opt. Commun. 96, 65–69 (1993). [CrossRef]
13. W. H. Renninger, A. Chong, and F. W. Wise, “Dissipative solitons in normal-dispersion fiber lasers,” Phys. Rev. A 77, 023814 (2008). [CrossRef]
14. N. Akhmediev and A. Ankiewicz, “Dissipative Solitons: From Optics to Biology and Medicine,” N. Akhmediev and A. Ankiewicz, ed., (Springer, Berlin, 2008).
18. G. Cabasse, A. Martel, B. Hideur, J. Ortaç, and Limpert, “Dissipative solitons in a passively mode-locked erbium-doped fiber laser with strong normal dispersion,” Opt. Express 16, 19322–19329 (2008). [CrossRef]
19. J.L. Massoubre, J. Oudar, S. Fantome, G. Pitois, J. Millot, J. Decobert, and Landreau, “All-optical extinction-ratio enhancement of a 160 GHz pulse train by a saturable absorber vertical microcavity,” Opt. Lett. 31, 537–539 (2006). [CrossRef] [PubMed]
20. J. W. Lou, M. Currie, and F. K. Fatemi, “Experimental measurements of solitary pulse characteristics from an all-normal-dispersion Yb-doped fiber laser,” Opt. Express 15, 4960–4965 (2007). [CrossRef] [PubMed]
21. N. N. Akhmediev, A. Ankiewicz, M. J. Lederer, and B. Luther-Davies, “Ultrashort pulses generated by mode-locked lasers with either a slow or a fast saturable-absorber response,” Opt. Lett. 23, 280–282 (1998). [CrossRef]