We demonstrated passively mode-locked ytterbium and erbium doped fiber lasers operating at 1039 and 1560 nm by using a common gold nanorods (GNRs) saturable absorber (SA). The GNRs with designed aspect ratios were mixed with sodium carboxymethylcelluose to form the GNRs SA film. The film had broadband longitudinal SPR (surface plasmon resonance) absorption from 800 to 1800 nm. By inserting the same film into a ytterbium or erbium doped fiber laser cavity pumped by a 980 nm laser diode, stable passively mode-locked laser operation at 1039 or 1560 nm was achieved for a threshold pump power of ~100 or ~70 mW, respectively. The pulse width, the output power, and the repetition rate of the 1039 nm mode-locked laser were 460 ps, 1.47 mW, and 43.5 MHz for a pump power of ~110 mW, respectively. The corresponding output parameters of the 1560 nm mode-locked laser were 2.91 ps, 2 mW, and 35.6 MHz for a pump power of ~74 mW, respectively. Our results showed that one GNRs SA could be used for constructing broadband mode-locked lasers.
© 2015 Optical Society of America
Gold nanorods (GNRs) have attracted much attention due to their unique structural, optical, electrical, and mechanical properties. GNRs have been widely investigated for photonic, optoelectronic, catalytic, and gas- and bio-sensing applications [1–5]. Optical properties of GNRs are governed by the excitation of localized surface plasmon resonance (SPR). The sensitivity of each SPR mode, whose spatial distribution and resonant energy depend on the diameter and length of GNRs, has been employed to construct various devices including sensors, solar cells, and all-optical signal processing devices. Despite their success in many fields, saturable absorption (SA) of GNRs has little considered previously. SA is a property of materials which the absorption of light decreases with increasing light intensity. It can be used for constructing passively Q-switched or mode-locked lasers [6, 7].
Generally, GNRs have two SPR absorption bands. One originates from the transverse SPR absorption of GNRs and the other from the longitudinal SPR absorption of GNRs. In contrast to the transverse SPR absorption, the longitudinal SPR absorption peak can be tuned in a wide wavelength range by varying the aspect ratio of the GNRs. Therefore, Q-switched or mode-locked fiber lasers operating with a wide wavelength range can be achieved by using GNRs as SAs. Recently, several works have been done to investigate the SA of GNRs. Elim et al. observed SA and reverse-SA at longitudinal SPR in GNRs . Lin derived the figure of merit of SPR of GNRs . Lamarre et al. measured the anisotropic nonlinear optical absorption coefficient of GNRs in a silica matrix . In our previous papers, we demonstrated experimentally passively mode-locked erbium-doped fiber lasers operating at 1.56 μm by using GNRs with an average aspect ratio of ~5.5 (GNRs 5.5) as SAs . Very recently, we reported passively mode-locked ytterbium-doped fiber lasers by using GNRs with an average aspect ratio of ~4.5 (GNRs 4.5) as SAs . Despite recent progress in this field, the broadband saturable absorption characteristics of GNRs have not yet been fully exploited for constructing pulsed lasers with a broadband operating wavelength. Especially, mode-locking at several discrete wavelengths (e.g. 1 and 1.5 μm) by using a GNRs SA has not yet been demonstrated, to the best of our knowledge.
In this paper, we demonstrated passively mode-locked ytterbium and erbium doped fiber lasers operating at 1039 and 1560 nm by using a common GNRs SA. GNRs were mixed with sodium carboxymethylcelluose (NaCMC) to form GNRs SA film. By optimizing the aspect ratio of GNRs, we obtained a GNRs SA film that had broadband longitudinal SPR absorption from 800 to 1800 nm. By inserting the same film into a ytterbium or erbium doped fiber laser cavity pumped by a 980 nm laser diode, stable passively mode-locked laser operation at 1039 or 1560 nm was achieved for a threshold pump power of ~100 or 70 mW, respectively.
2. Fabrication and chacterization of gold nanorods (GNRs) saturable absorber
GNRs were synthesized through a seed-mediated growth method [11–14]. The seed solution for GNRs was prepared as reported previously [11, 12]. To prepare the growth solution, 20 mL of 0.15 M hexadecyltrimethyl ammonium bromide (CTAB), 10 mL of 0.2 M 5-bromosalicylic acid, 1 mL of 4 mM AgNO3 aqueous solution and 1 mL of 0.1 mM ascorbic acid were dissolved in a flask, and then the solution was vigorously stirred for 30s until it became colorless. Finally, 1mL seed solution was injected into the growth solution to initiate the growth of GNRs. By varying the amounts of reagents, we could synthesize GNRs with average aspect ratios between 2 and 5.5. The longitudinal SPR absorption peak could be tuned from 730 to 1460 nm by varying the average aspect ratio from 2 to 5.5 . The final solution was kept for 48h and no precipitation was found. GNRs SA film was formed by casting the GNRs -NaCMC solution onto a flat substrate, and then followed by a slow drying at room temperature.
In our previous work, we reported mode-locking at 1 or 1.56 μm wavelength by using GNRs 4.5 or GNRs 5.5, respectively [11, 12]. Figures 1(a) and 1(b) show the transmission electron microscopy (TEM) images of GNRs 4.5 and GNRs 5.5 (the scale bar is 200 nm). The dimensions of GNRs 4.5 and GNRs 5.5 are (10 ± 2 nm) × (38 ± 4 nm) and (13 ± 2 nm) × (60 ± 4 nm), respectively. Only a small fraction of spherical nanoparticles exists in the as-synthesized GNRs sample. The insets of Figs. 1(a) and 1(b) show the aspect ratios distribution of the above GNRs. The average aspect ratio for GNRs 4.5 or GNRs 5.5 is 4.5 or 5.5, respectively. The absorption spectra of the above GNRs are shown in Fig. 1(d). The blue dot dash curve represents the absorption spectrum of GNRs 4.5 and the red dot curve represents the absorption spectrum of GNRs 5.5. There are two absorption peaks for each GNRs. The transverse SPR absorption peaks are almost fixed at 532 nm for both of them. Interestingly, the longitudinal SPR absorption peaks can be tuned in a wide wavelength range by varying the aspect ratio of the GNRs. GNRs 4.5 has an obviously absorption peak at 1145 nm and GNRs 5.5 has an absorption peak at 1460 nm. And, both GNRs 4.5 and GNRs 5.5 have broadband longitudinal SPR absorption from 800 nm to 1800 nm, which imply that passively mode-locked ytterbium and erbium doped fiber lasers operating at 1 and 1.56 μm wavelength might be achieved by using a GNRs 4.5 or GNRs 5.5 film. In our previously work, we tried to realize mode-locking at 1 and 1.56 μm by using a film, GNRs 4.5 or GNRs 5.5. However, we could not achieve them. To explain the above experimental results, we measured the dependence of the transmission ratio of GNRs 4.5 or GNRs 5.5 on the pump power density by using 1 and 1.56 μm mode-locked fiber lasers. Figure 2 shows the schematic setup of the power-dependent transmittance measurements. In our experiments, we used two kinds of home-made pulsed fiber lasers as pump laser sources. One has a center wavelength of 1039 nm, a spectral width of ~1.2 nm, a repetition rate of 43.5 MHz, and a pulse duration of 380 ps. The other has a center wavelength of 1560 nm, a spectral width of ~3.1 nm, a repetition rate of 35.6 MHz, and a pulse duration of 760 fs. The incident pump power could be varied by tuning the tunable attenuator. The 3dB coupler was used to split the pulsed laser into 50% reference signal and 50% input light to GNRs film. The power meter was used to measure the output power. Then we obtained the power-dependent transmittance of the GNRs film. By fitting the experiment data with the equation α(I) = αs/(1 + I/Is) + αns (where α(I) is the absorption coefficient, αs and αns are the saturable and nonsaturable absorption components, and I and Is are input and saturable intensities, respectively), the modulation depth (ΔT), non-bleachable loss (αns) and saturation intensity (Is) were determined, as shown in Fig. 3and Table 1. In Figs. 3(a) and 3(b), we can see that GNRs 4.5 has relatively lower saturation intensity at 1039 nm (~0.13 MW/cm2) than GNRs 5.5 (~4.1 MW/cm2), but have higher saturation intensity at 1560 nm (~19.9 MW/cm2) than GNRs 5.5 (~1.05 MW/cm2). Generally, the lower the saturation intensity is, the easier the mode locking becomes. Therefore, it is difficult to realize mode-locking at 1560 nm by using GNRs 4.5. Similarly, it is also difficult to realize mode-locking at 1039 nm by using GNRs 5.5. This might be the main reason why we could not realized mode-locking at 1039 and 1560 nm by using a film, GNRs 4.5 or GNRs 5.5.
We consider that mode-locking at 1 and 1.56 μm wavelength can be realized by using mixed GNRs with designed aspect ratios as SAs. Based on the above consideration, we fabricated mixed GNRs by mixing GNRs 4.5 with GNRs 5.5 at equal proportion (~50%). Fig. 1(c) shows the TEM image of the mixed GNRs (1 mL GNRs 4.5 and 1 mL GNRs 5.5). The dimension of the mixed GNRs is (10 ± 3 nm) × (50 ± 5 nm) (the scale bar is 200 nm). The inset in Fig. 1(c) shows the aspect ratios distribution of the mixed GNRs. The aspect ratios were in a range of 2.5 to 8.5, and the main average aspect ratios are 4.5 and 5.5. The black solid curve in Fig. 1(d) shows the absorption spectrum of the mixed GNRs. The transverse SPR absorption peak is fixed at 532 nm and the broadband absorption from 800 nm to 1800 nm is induced by the overlapping of longitudinal SPR absorption bands of GNRs 4.5 and GNRs 5.5. We also measured the dependence of the transmission ratio of the mixed GNRs on the pump power density by using 1 and 1.56 μm mode-locked fiber lasers, as shown in Figs. 3(c) and 3(d). Its modulation depth (ΔT), non-bleachable loss (αns) and saturation intensity (Is) were determined and included into Table 1. The saturation intensity of the mixed GNRs SA are about 0.17 MW/cm2 at 1039 nm and 5.1 MW/cm2 at 1560 nm, respectively. As we know, the SA of gold nanostructures originates from SPR absorption. In the case of GNRs, the saturation absorption at 1 or 1.56 μm is mainly related to the longitudinal SPR absorption. The bleach of the longitudinal SPR modes occurring at moderate intensities is an SA process that results in an increase of optical transmission [8, 15]. As we can see from Table 1, the saturation intensity (Is) at 1039 nm of the mixed GNRs SA decreases to 0.17 MW/cm2 compare to that (4.1 MW/cm2) of GNRs 5.5 owing to the contribution of GNRs 4.5. Similarly, the saturation intensity (Is) at 1560 nm of the mixed GNRs SA decreases to 5.1 MW/cm2 compare to that (19.9 MW/cm2) of GNRs 4.5 owing to the contribution of GNRs 5.5. In addition, the modulation depth (ΔT) at 1039 or 1560 nm of the mixed GNRs SA becomes lower than that of GNRs 4.5 or GNRs 5.5. As above mentioned, the lower the saturation intensity is, the easier the mode locking becomes. The above results showed that the mixed GNRs SA film might be used to induce mode-locking at 1 and 1.56 μm wavelength. We also measured different parts of the same GNRs film, and obtained similar results in our experiments. The results showed that the dimension and spatial distribution of the GNRs were not the main factors to influence the nonlinear SA property of GNRs. The main factors to influence the nonlinear SA property of GNRs were the concentrations and the aspect ratios distribution of GNRs.
3. Construction of wideband mode-locked fiber lasers
Figure 4 shows a schematic configuration of the mode-locked ring cavity fiber laser. The pump source we used was a 980 nm laser diode. The pump light was launched into the laser cavity through a wavelength division multiplexer (WDM) coupler. A 20 cm-long ytterbium or 30 cm-long erbium doped fiber was used as the gain medium. An optical isolator was used to force the unidirectional operation. The polarization controller (PC) was utilized for optimization of the laser cavity to induce mode-locking operation. A 10 dB WDM coupler was utilized to output mode-locked lasers. The same GNRs SA film was used as a mode-locking element. The output lasers were analyzed by using an optical spectrum analyzer, a power meter, an autocorrelator, and an oscilloscope together with a photodetector.
4. Experimental results and discussions
In the case of mode-locked ytterbium doped fiber lasers (YDFLs), stable mode-locked laser oscillation was obtained when the pump power exceeded the threshold of ~100 mW. The mode-locked laser was operated at the central wavelength of 1039 nm with a spectral width of 0.9 nm for a pump power of ~110 mW. The spectrum of the mode-locked YDFL was measured by an optical spectrum analyzer with a resolution of 0.01 nm, as shown in Fig. 5(a).The spectrum exhibits steep edges which is the typical feature of dissipative-soliton laser in the all-normal dispersion regime. Figures 5(b) and 5(c) show the pulse train and single pulse profile of the above mode-locked YDFL, which was measured by using a high-speed digital oscilloscope (Tektronix, DPO70604C, 6 GHz bandwidth) with a photodetector (EOT, ET350, 12 GHz bandwidth). The interval between two adjacent pulses is about ~23 ns, corresponding to a repetition rate of ~43.5 MHz. The measured pulse duration is ~460 ps. The time-bandwidth product (TBP) is ~119.6, indicating that the pulse have large chirps in the cavity [16–19]. Compared to other all-normal dispersion mode-locked fiber lasers based on other SAs (SWCNT or graphene), the pulse durations are also the time scale of few hundred picoseconds. Although all-normal dispersion mode-locked fiber lasers have been realized by using different kinds of SAs, it is also important to find some new SAs for ultrafast mode-locked laser generation. Figure 5(d) shows the output power of the mode-locked fiber laser as a function of the pump power. The output power increases linearly from ~1.31 mW to ~1.67 mW, corresponding to a slope efficiency of ~1.67%. In addition, the mode-locked laser became unstable when the pump power is larger than ~121 mW. The SA film might be damaged due to the photo-thermal effects of GNRs. In the future, we will try to increase the damage threshold of the film by using the evanescent effect .
The same GNRs SA film was also used to induce mode-locking in the erbium doped fiber laser (EDFL). The ring-cavity laser was similar to that shown in Fig. 4. Stable mode-locked laser oscillation was obtained when the pump power exceeded the threshold of ~70 mW. The mode-locked laser spectrum for a pump power of ~74 mW was shown in Fig. 6(a).The spectral width at half-maximum is about ~1.5 nm. Obviously Kelly bands due to the interference of dispersion waves are also observed, which is the feature of the solitary operation of the laser [21–23]. Output pulse trains of the above mode-locked laser was measured by using a digital oscilloscope, as shown in Fig. 6(b). The time interval between two adjacent pulses is about ~28.1 ns, corresponding to a repetition rate of ~35.6 MHz, which coincides with the above calculated value by using the length of the laser cavity (~5.61 m). Figure 6(c) shows a single pulse profile of the above mode-locked fiber laser measured by using an autocorrelator (Alinar-lab HAC-200). The pulse duration is ~2.91 ps. The TBP is calculated as ~0.523, which is slightly larger than the transform-limit value of 0.315. Figure 6(d) shows the output power of the mode-locked laser as a function of the pump power of the 980 nm laser. The mode-locking threshold is ~70 mW and the maximum average output power is about ~2.33 mW. The optical-to-optical efficiency is about ~4.9%. The mode-locked laser became unstable when the pump power higher than ~80 mW. Future work will be done for increasing the damage threshold of the GNRs SA by using the evanescent effect.
In addition, to verify the effects of the mixed GNRs on mode-locking, we placed a bare NaCMC film in the cavity. Figure 7 shows the absorption spectrum of bare NaCMC film. The NaCMC only contributes a small background for the whole range from 400 to 1800 nm. With increasing the pump power to over 300 mW, only continuous wave laser operation was observed and no pulsed operation was observed. These results confirmed that the above mode-locked fiber lasers was induced by the mixed GNRs SA film.
In summary, we have experimentally obtained all-fiber passively mode-locked YDFL and EDFL using the same GNRs as SA. The GNRs had broadband absorption from ~800 nm to ~1800 nm, which was caused by the longitudinal SPR absorption of the GNRs. Stable mode-locked operation at ~1039 or 1560 nm was achieved for a threshold pump power of ~100 or 70 mW. Our experimental results show that the designed GNRs film is an effective SA for broadband mode-locked fiber lasers.
This work was supported by the NSFC (grants 51072065, 61178073, 60908031, 60908001, 61378004, and 61077033), the Program for NCET in University (No: NCET-08-0243), the Opened Fund of the State Key Laboratory on Integrated Optoelectronics, and Tsinghua National Laboratory for Information Science and Technology(TNList)Cross-discipline Foundation.
References and links
1. M. C. Daniel and D. Astruc, “Gold Nanoparticles: Assembly, Supramolecular Chemistry, Quantum-Size-Related Properties, and Applications toward Biology, Catalysis, and Nanotechnology,” Chem. Rev. 104(1), 293–346 (2004). [CrossRef] [PubMed]
2. S. Maity, J. R. Bochinski, and L. I. Clarke, “Metal nanoparticles acting as light-activated heating elements within composite materials,” Adv. Mater. 22, 5259–5270 (2012).
3. K. H. Kim, A. Husakou, and J. Herrmann, “Linear and nonlinear optical characteristics of composites containing metal nanoparticles with different sizes and shapes,” Opt. Express 18(7), 7488–7496 (2010). [CrossRef] [PubMed]
4. H. B. Liao, R. F. Xiao, J. S. Fu, P. Yu, G. K. L. Wong, and P. Sheng, “Large third-order optical nonlinearity in Au:SiO2 composite films near the percolation threshold,” Appl. Phys. Lett. 70(1), 1 (1997). [CrossRef]
5. A. Kumar, J. Prakash, D. S. Mehta, A. M. Biradar, and W. Haase, “Enhanced photoluminescence in gold nanoparticles doped ferroelectric liquid crystals,” Appl. Phys. Lett. 95(2), 023117 (2009). [CrossRef]
6. K. H. Kim, U. Griebner, and J. Herrmann, “Theory of passive mode locking of solid-state lasers using metal nanocomposites as slow saturable absorbers,” Opt. Lett. 37(9), 1490–1492 (2012). [CrossRef] [PubMed]
7. K. H. Kim, U. Griebner, and J. Herrmann, “Theory of passive mode locking of semiconductor disk lasers in the blue spectral range by metal nanocomposites,” Opt. Express 20, 16174–16179 (2012).
8. H. I. Elim, J. Yang, J. Y. Lee, J. Mi, and W. Ji, “Observation of saturable and reverse-saturable absorption at longitudinal surface plasmon resonance in gold nanorods,” Appl. Phys. Lett. 88(8), 083107 (2006). [CrossRef]
9. J. T. Lin, “Nonlinear optical theory and figure of merit of surface plasmon resonance of gold nanorods,” J. Nanophoton. 5(1), 051506 (2011). [CrossRef]
10. J. M. Lamarre, F. Billard, C. H. Kerboua, M. Lequime, S. Roorda, and L. Martinu, “Anisotropic nonlinear optical absorption of gold nanorods in a silica matrix,” Opt. Commun. 281(2), 331–340 (2008). [CrossRef]
11. Z. Kang, Y. Xu, L. Zhang, Z. X. Jia, L. Liu, D. Zhao, Y. Feng, G. S. Qin, and W. P. Qin, “Passively mode-locking induced by gold nanorods in erbium-doped fiber lasers,” Appl. Phys. Lett. 103(4), 041105 (2013). [CrossRef]
12. Z. Kang, Q. Li, X. J. Gao, L. Zhang, Z. X. Jia, Y. Feng, G. S. Qin, and W. P. Qin, “Gold nanorods saturable absorbers for passively mode-locking at 1 μm wavelength,” Laser Phys. Lett. 11(3), 035102 (2014). [CrossRef]
13. X. C. Ye, Y. Z. Gao, J. Chen, D. C. Reifsnyder, C. Zheng, and C. B. Murray, “seeded growth of monodisperse gold nanorods using bromide-free surfactant mixtures,” Nano Lett. 13(5), 2163–2171 (2013). [CrossRef] [PubMed]
14. X. C. Ye, L. H. Jin, H. Caglayan, J. Chen, G. Z. Xing, C. Zheng, V. Doan-Nguyen, Y. J. Kang, N. Engheta, C. R. Kagan, and C. B. Murray, “Improved size-tunable synthesis of monodisperse gold nanorods through the use of aromatic additives,” ACS Nano 6(3), 2804–2817 (2012). [CrossRef] [PubMed]
15. R. Philip, G. R. Kumar, N. Sandhyarani, and T. Pradeep, “Picosecond optical nonlinearity in monolayer-protected gold, silver, and gold-silver alloy nanoclusters,” Phys. Rev. B 62(19), 13160–13166 (2000). [CrossRef]
16. X. H. Li, Y. L. Tang, Z. Y. Yan, Y. Wang, B. Meng, G. Z. Liang, H. D. Sun, X. Yu, Y. Zhang, X. P. Cheng, and Q. J. Wang, “Broadband saturable absorption of grapheme oxide thin film and its application in pulsed fiber lasers,” IEEE J. Sel. Top. Quantum Electron. 20, 1101107 (2014).
17. S. S. Huang, Y. G. Wang, P. G. Yan, J. Q. Zhao, H. Q. Li, and R. Y. Lin, “Tunable and switchable multi-wavelength dissipative soliton generation in a graphene oxide mode-locked Yb-doped fiber laser,” Opt. Express 22(10), 11417–11426 (2014). [CrossRef] [PubMed]
18. X. H. Li, Y. G. Wang, Y. S. Wang, W. Zhao, X. C. Yu, Z. P. Sun, X. Cheng, X. Yu, Y. Zhang, and Q. J. Wang, “Nonlinear absorption of SWNT film and its effects to the operation state of pulsed fiber laser,” Opt. Express 22(14), 17227–17235 (2014). [CrossRef] [PubMed]
19. X. H. Li, Y. G. Wang, Y. S. Wang, Y. Z. Zhang, K. Wu, P. P. Shum, X. Yu, Y. Zhang, and Q. J. Wang, “All-normal dispersion passively mode-locked Yb-doped fiber ring laser based on a graphene oxide saturable absorber,” Laser Phys. Lett. 10(7), 075108 (2013). [CrossRef]
21. Z. P. Sun, D. Popa, T. Hasan, F. Torrisi, F. Q. Wang, E. J. R. Kelleher, J. C. Travers, V. Nicolosi, and A. C. Ferrari, “A stable, wideband tunable, near transform-limited, grapheme-mode-locked, ultrafast laser,” Nano Res. 3(9), 653–660 (2010). [CrossRef]
22. F. X. Kärtner, I. D. Jung, and U. Keller, “Soliton mode-locking with saturable absorbers,” IEEE J. Sel. Top. Quantum Electron. 2(3), 540–556 (1996). [CrossRef]
23. H. Zhang, Q. L. Bao, D. Y. Tang, L. M. Zhao, and K. P. Loh, “Large energy soliton erbium-doped fiber laser with a grapheme-polymer composite mode locker,” Appl. Phys. Lett. 95(14), 141103 (2009). [CrossRef]