Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Optically controlled tunable ultra-narrow linewidth fiber laser with Rayleigh backscattering and saturable absorption ring

Open Access Open Access

Abstract

We propose an optically controlled tunable ultra-narrow linewidth fiber laser assisted with the mode selection induced by a saturable absorption interference ring and linewidth narrowing of fiber Rayleigh backscattering (RBS). The interference ring serves as an artificial narrow-band filter, which conduces to the laser operating at a single-frequency state. To realize narrower linewidths, additional single-mode fiber is utilized to accumulate a weak RBS feedback. On basis of inherent wavelength universality of this linewidth-narrowing mechanism, an all-optical technique is employed to enable linear and stable tunability of the laser. Cooperating with a micro-fiber Bragg grating covered by graphene, the lasing wavelength is tuned precisely and reversibly with a sensitivity of 12.4 pm/mW and a linear fitting R2 over 0.997 by changing the power of a controlling beam. During a stability test with the controlling pump power fixed, the long-term free-running power fluctuation is less than 0.5%. The Output laser linewidth is compressed to be ~200 Hz, which is also confirmed by the descending frequency noise spectrum.

© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

1. Introduction

Ultra-narrow linewidth fiber lasers with maneuverable and practicable tunability are tailored to various emerging applications, such as coherent detection, high-resolution spectroscopy, precise sensing and large-capacity optical communication, for their unique attributes of long coherence length, low phase noise and simplicity [1–4]. Compared to gain mediums in semiconductor lasers, erbium ions doped in fibers possess purer particle level distribution, which is beneficial for the direct oscillation of ultra-narrow longitudes. Amidst the absence of any compression mechanism for the generated single longitude, laser linewidths easily reach several to tens of kilohertz order. Therefore, previous studies on the fiber lasers mainly emphasize on longitudinal mode selections generally by enlarging free spectrum ranges (FSRs) or restricting effective gain bandwidths. In order to make FSRs approach to the gain bandwidth, short-cavity structures with lengths of a centimeter order have been proposed [5,6]. For their high integration, it is difficult to insert additional tunable components into those short cavities to regulate their characteristics. In long-length cavities, it is necessary to induce effective mode-selection devices. High finesse filters with bandwidths of megahertz order can be utilized for the SLM operation. Traditional compound-cavity fiber lasers usually suffer from unstable side-mode suppression ratios (SMSRs), because inadequate selections of cavity lengths may result in simultaneous oscillation of multi-longitudes with FSRs corresponding to different sub-cavities [7,8]. Slow saturable absorbers have a narrow-band filtering effect by generating an artificial grating because of the longitudinal index refraction modulation, and mode hoping in ring lasers can be well restrained [9–11]. Previous methods for narrow-linewidth outputs are mainly employed in longitude selections. Nevertheless, they pose to be ineffective towards the spectral compression of the generated single longitude with linewidths lower than kilohertz order. To obtain a narrower laser linewidth on basis of SLM operations, the Rayleigh backscattering (RBS) has been experimentally studied in fibers, where the RBS acts as a ultra-weak feedback with a linewidth compression factor [12,13]. Fibers can be regarded as random-distributed cascaded scattering sections, and laser circularly propagating inside the cavity can be compressed continually.

On basis of the mode selection mechanisms and RBS feedback, we ought to identify an efficient method to control the laser longitude. Some high-precision measurements, for example, optical frequency domain reflectometry, have a strict requirement of the laser tuning linearity. Several mature wavelength-tuning methods, including mechanical stress, piezoelectric components, and acoustic-optic modulation, have been reported [14–18]. However, all-optical methods of tunable lasers are still worth exploring because of their unique superiorities for simple implementation, high stability and remote control with ease. Fiber Bragg grating (FBG) is one of the most frequently-used tunable components with a low insertion loss in long laser cavities [19–21]. The pure fiber structure and small size make FBGs easily integrated with other fiber devices with low insertion loss and maintains the simplicity of all-fiber laser systems. Its optic performances can be conveniently regulated by various physical fields. To enhance the modulation capacity of FBGs responding to the external light fields, their surfaces have to be modified by special materials with sensitive optical properties. Due to the fast photo-induced self-heating of graphene, the passband of FBGs can be linearly tuned by a controlling beam [22], which provides narrow-linewidth lasers with a light-enabled tunability. Commercial light sources with fiber pigtails can be directly coupled into FBGs, which has no need of other devices for alignments.

In this work, we cooperate a saturable absorption interference ring (SAIR) fabricated by erbium-doped fiber (EDF) to achieve the elementary mode selection. For the periodical distribution of EDF’s index refraction caused by the interference between two lights propagated reversely, an auto-tracking equivalent filter with a bandwidth of ~5 MHz is generated. We add 100 m single-mode fiber (SMF) to accumulate the RBS feedback for the main cavity. The RBS feedback can be induced to further narrow down the laser linewidths to be ~200 Hz. On basis of the narrow-linewidth performances, the output wavelength is optically tuned with a FBG covered by commercial single-layer graphene that is separately pumped by a controlling light. Due to the wavelength-insensitivity of the RBS, such a light-induced tunable fiber laser poses both of ultra-narrow linewidths at all wavelength positions and linear all-optical tuning performance.

2. Experimental setup and principle

Figure 1(a) schematically shows the experimental setup of the all-optical tunable ultra-narrow linewidth fiber laser. In this experiment, we use a double-pump structure (980 nm pump 1 and 2) and 1 m high-doped single-mode EDF (Peak absorption: 80 ± 8 dB/m@1530 nm, NA: 0.2; Er80-8/125, LIEKKITM) to provide the gain condition. The lasing wavelength is tuned by a single-layer graphene-coated micro-FBG (GCMFBG) with a bandwidth of 0.2 nm shown in the above red box. The grating region of FBG is 12 mm. Figure 1(b) is the photograph of the whole structure, and the red shadow is the graphene-coating region. The graphene interacts with superficial evanescent wave outside the FBG. Therefore, the uniform FBG is etched in hydrofluoric acid to minish the cladding diameter to be ~16 um, and the periodical distribution of refraction index in core is not influenced. To ensure enough tuning efficiency and moderate insertion loss, the etched FBG is partly covered by graphene with an interaction length of ~6 mm, resulting in a reflectivity at central wavelength of ~25%. The additional insertion loss is caused by the mode-mismatch of the etched cladding and absorption of graphene. The 980 nm controlling pump (pump 3) as a controlling light is injected into the GCMFBG. The nonlinear responses of graphene, such as four-wave mixing and Kerr effect need to be pumped by ultrafast lasers, which has been employed in all-optical signal modulations [23,24]. For the wavelength tuning, we mainly emphasis on the linearity and stability. Thus, continuous-wave light is used to achieve graphene’s static controls. When carriers of graphene absorb pump photos, they easily jump up to the conduction band for the zero band gap, as depicted in Fig. 1(c). Then they spontaneously descend to a lower energy level by phonon radiations to achieve the self-heating [23]. Due to the high thermal conduction of graphene, the micro-FBG is heated to achieve the spectral wavelength-shift, when we increase the controlling light power, as shown in Fig. 1(d). The change of graphene’s refraction index also directly increases the effective index refraction of the GCMFBG, which is another factor to change the reflected central wavelength. Due to graphene’s partial coverage on the surface of MFBG, only the graphene-coated grating region can be effectively heated, resulting in a higher refraction index. There, the original periodical refraction index of the FBG is influenced, which generates some unwanted spectral side-lopes. Their peak intensity is lower than that of the central reflected wavelength by 6 dB. Meanwhile, the central wavelength still represents a linear relationship with the injected power, as shown in the red box in Fig. 1(d).

 figure: Fig. 1

Fig. 1 (a) Experimental setup of the optically tunable ultra-narrow linewidth fiber laser; (b) photograph of the graphene-coated micro-FBG (GCMFBG), and the red shadow part is the graphene-coated region. (c) Phonon relaxation process of graphene; (d) red-shifted reflection spectra of the GCMFBG with controlling light powers of 26.8 mW, 64.3 mW, and 101.5 mW.

Download Full Size | PDF

Compared with fixed-wavelength lasers, the SLM operation condition of tunable lasers are much stricter. Although it is possible to achieve a SLM operation by carefully adjusting the cavity loss only with the gain restraint of the FBG, owing to homogenous-broaden energy levels of erbium-ion. Nevertheless, inevitable changes of the spectral shape including increased insertion loss, slightly changed bandwidth and unwanted spectral side-lopes, will seriously change the inner-cavity gain-loss balance. Therefore, it cannot ensure a stable SLM operation within the FBG’s passband window when the laser is continuously tuned. To solve this problem, the laser filtered by the GCMFBG enters into the SAIR to achieve the mode-reselection and restrict the mode hopping. As shown in the left red box of Fig. 1(a), the SAIR is consisted of a 2 × 2 3-dB optical coupler (OC1), 5 m un-pumped low-doped EDF (Absorption: 5.4–7.1 dB/m@1531 nm, NA: 0.21-0.24; M-5(980/125), Fibercore) and a polarization controller (PC). After entering into the interference ring from port 1, the light is split to be two beams that reversely propagate. PC 1 is used to optimize the double-beam interference. Owing to the erbium-ion’s millisecond-order relaxation, the interference maintains periodical spatial burning holes inside EDF’s core. According to the Kramer-Kronig integral relationship [25], the refraction index is approximately proportional to the absorption coefficient when the dispersion effect is neglected in narrow-linewidth lasers. Due to the saturable absorption mechanism, the longitudinal index refraction of the EDF is periodically modulated as the formula:

n(λ,z)neff(λ)¯1+I0Isatcos(4πneff(λ)¯λz),
where I0 is the amplitude of the standing wave, Isat is the saturable intensity of EDF, and neff(λ) is the effective index refraction. The EDF can be regarded as a self-build grating with a period of λ/2neff. According to the coupled-mode theory [26], the equivalent bandwidth can be theoretically estimated by the expression:
Δf(λ,Δn)Δncδλ2neff¯2(λ)Δn2+λ2L2EDF,
where Δn is the modulated amplitude of the refraction index (less than 3 × 10−7 [27]), c is the light speed in vacuum, LEDF is the length of EDF, and δ, defined as the dimension compensation parameter of mode coupling, equals 1 m. In our experiment, LEDF, neff, and λ are selected as 5 m, 1.45 and 1.55 μm, respectively, which generates an equivalent narrow filter with a gain bandwidth theoretically less than ~5 MHz. The selection of 5 m EDF is mainly based on the consideration of a trade off between the filter bandwidth and the pump efficiency. Because, the strong absorption of the longer EDF has to also be taken into consideration. The basic principle of the length optimization is to select the un-pumped EDF as long as possible on the premise of a sufficient gain condition for the laser oscillation.

As shown at the bottom-right corner of Fig. 1(a), the feedback structure (within the red box) consists of 100 m SMF for RBS, a variable optical attenuation (VOA) and a Faraday rotating mirror (FRM). Longer fiber (hundreds of meter) easily stimulates unfavourable Brillouin scattering that will submerge the weak Rayleigh light. To accumulate sufficient RBS feedback and enhance the Brillouin threshold, the light from port 1 of circulator 2 enters into the RBS fiber with a length of 100 m. It is reflected by the FRM to stimulate the RBS light for the next circulation. This RBS fiber can be regarded as countless cascaded scattering cross sections where scattered light will be accumulated as a weak feedback to compress the laser linewidth. The backscattered light is injected into the main-cavity from the port 3. Thus, the signal laser and RBS light share the same optical path, which simplifies the cavity structure. The VOA behind the RBS fiber is used to suppress the side longitudes by optimizing the intensity proportion between the signal laser and RBS. The output is achieved by OC 2 with a couple ratio of 8:2, and the total length of the main cavity is ~14 m.

3. Experimental results and discussion

We set output powers of pump 1 and 2 as 260 mW without RBS fiber. The output radio frequency (RF) spectrum is shown in Fig. 2(a), which confirms that no beat-frequency signal appears within a frequency span of 150 MHz. To observe the SMSR and linewidths, we also measure the RF spectra using the delay self-heterodyne system (DSHS) with a 100 MHz acousto-optical modulator as shown as the inset of Fig. 2(a). Owing to the narrow-band filtering effect induced by the SAIR, the laser can operate in a SLM state with a cavity length less than 40 m. Thus, the 14 m cavity length only allows single longitude oscillation. We keep the pump power unchanged and add the RBS fiber into this cavity. The inset of Fig. 2(b) shows the RF spectrum measured by the DSHS, and the side modes occurs and rapidly decay within the frequency span of less than 9 MHz, approaching the previously estimated bandwidth of the SAIR. As the previous discussion, the Rayleigh fiber has random scattering sections, which generates multiple laser reflections along this fiber, leading to a mode selection effect. If we regard the FRM as the last scattering point, the light at the FRM equivalently has a highest scattering intensity, resulting in unwanted side modes, as shown in Fig. 2(b). However, due to the weak RBS intensity, the SMSR just reaches 20 dB.

 figure: Fig. 2

Fig. 2 (a) RF spectrum without Rayleigh feedback directly measured by the PD and the corresponding beat-frequency signal measured by the DSHS; (b) RF spectrum with Rayleigh fiber directly measured by PD and the corresponding RF spectrum of the beat-frequency signal.

Download Full Size | PDF

To improve the SMSR, we use the VOA to adjust the inner-cavity gain-loss balance and the intensity proportional between RBS light and signal laser. Because the SAIR has already limited longitudes within a narrower frequency range. The mode hopping can be equally reduced at a cost of additional cavity loss, which largely reduces difficulties of the subsequent SLM operation. Figure 3(a) shows the RF spectral evolution under with attenuation increased and pump power unchanged. Because the total cavity is enlarged, the output power will be decreased. To compare the side-mode intensity, these curves are all measured with the same power injected into the photodetector (PD). We use another VOA behind output port of the laser to adjust the power for measurement. Side modes are gradually suppressed with enhanced attenuation. When the attenuation value of the VOA approaches ~3 dB, side-modes are almost submerged, as shown as pink and black curves. With the attenuation increased, the relative intensity of the RBS to the reflected light of the FRM will be further enlarged. The RBS distributed feedback will be more preponderant in the back-propagated light, which is in favour of the generation of the SLM. Owing to the ultra-narrow filtering effect and the random RBS feedback, the main mode can be overwhelming in the mode competition process. The main mode is obviously dominate among longitudes within this frequency range. Thus, with the increased attenuation of the VOA, the competition under the broadening gain of EDF will further weaken the side mode and select the dominated mode. Owing to these factors, including self-induced filtering effects of SAIR, RBS random feedbacks and homogenous-broaden gain of the EDF, the compressed laser achieves a robust SLM operation. To observe the phase stability, we test the time-domain beat signal of the compressed laser within a timescale of 100 ms. Figure 3(b) just shows arbitrarily intercepted 230 periods, and the inset is the partial enlargement. The sine-shaped curve with constant frequency and stable amplitude indicates a very low phase noise, and meanwhile there is also no mode-hopping appearing.

 figure: Fig. 3

Fig. 3 (a) Output RF spectra with Rayleigh fiber with the increment of the attenuation value; (b) time domain of the beat-frequency signal with a SLM operation.

Download Full Size | PDF

We further optimize the attenuation as 3 dB and keep previous powers of pump 1 and 2 unchanged. Utilizing the light-matter performance between graphene and evanescent wave, the lasing wavelength can be tuned conveniently by adjusting the pump 3 (controlling pump). The circulator 1 with a work passband at C band has a remarkable filtering effect for the 980 nm controlling pump laser. Thus, the controlling pump laser cannot enter into the laser cavity, which just controls the lasing wavelength. Figure 4(a) shows tuning processes with the controlling pump power varied from 0 mW to 159 mW, and the spectral contrast reaches 65 dB. The corresponding RF spectrum is shown in Fig. 4(b), where the output laser always keeps a stable SLM operation during the tuning process. The relationship between the wavelength and controlling pump power is shown in Fig. 4(c), where the fitting R2 reaches 0.99742, whose value is enough to promise the availability of this optical-controlled method in engineering applications, especially for OFDR systems. The high tuning linearity also makes it very easy to demarcate output wavelengths. The inset shows the precise tuning in a smaller scale, where the linearity still maintains in a good condition with the step of ~0.02 nm.

 figure: Fig. 4

Fig. 4 (a) Output spectra of the ultra-narrow linewidth laser with the increment of the pump 3; (b) color-corresponded RF spectra during the tuning process; (c) peak wavelengths with changed powers of pump 3 and the linear fitting curve; (d) long-term stability test of the wavelength and output power with a fixed power of pump 3.

Download Full Size | PDF

Different from stress- and electric-based fine tuning methods assisted by well-designed high-precision mechanical displacement systems and expensive electrical driving setups, this graphene-integrated device can be conveniently realized by an inexpensive infrared continuous-wave light source with an output power of dozens to hundreds of milliwatts, which also reduces the complexity of the tuning setups and operation cost. This tunable device possesses a considerable spectral response to the light power variation of less than 1 mW. Owing to graphene’s wideband absorption performance, this tunable device has no special consideration for the wavelength of the controlling light. The photo-thermal effect of graphene is resulted from photo-excited interactions between carriers and phonon, as the previous discussion. Thus, the generated Joule thermal should have a strict macroscopical correspondence with the controlling light power. Therefore, the tuning precision is mainly limited by the power fluctuation of the controlling pump and the operation stability of this cavity, which can be facilitated by lower surrounding vibration noise. The experimental data points and fitting curves during the round-trip variation of powers are overlapped, which means an invertible and repetitive tuning performance. This laser can achieve wavelength scanning with easiness because of the linearly power-dependent property. The tuning slope in this experiment is 12.4 pm/mW, whose value positively depends on the evanescent wave intensity exposed on the graphene and the coating region length. Compared with temperature-controlled tuning by directly heating, photo-thermal induced index change has a faster response time. Because of the zero-bandgap structure of graphene, it absorbing photos can fast generates sufficient Joule thermal by phonon relaxation. Graphene’s ultra-high thermal conduction coefficient (5300 W (m K)−1) also makes it effectively transfer heat to the fiber device with a high speed [28]. In our experiment, the graphene can response to the intensity change of the controlling light at microsecond order scale with easiness. Nevertheless, tuning speed of the GCMFBG is still limited by thermal conduction processes from graphene to the FBG. To obtain a higher tuning sensitivity and faster tuning response, we can select etched FBGs with smaller diameters to enhance the evanescent wave intensity and expedite the thermal conduction speed. However, in this case, due to a strong absorption of graphene with a millimeter-order interaction length, larger insertion loss and serious spectral distortion both have to be considered. As shown in Fig. 4(d), the long-term cavity stability is tested. We arbitrarily set one controlling pump power as 81.9 mW unchanged. In this case, the wavelength is tuned at ~1552.68 nm. Without frequency-locked setups, long-term fluctuation ranges of free-running wavelength and power are less than ± 0.005 nm and ± 0.5%, respectively, indicating a very robust operation.

Compared with the reflected light from the FRM, the Rayleigh scattering within one round trip is weak within a single laser cycle. However, the RBS light has a narrower frequency spectrum, which will be positively superimposed to gradually modify the laser linewidth with multiple laser circulations until it reaches dynamical equilibrium. To confirm the ultra-narrow linewidth property when the laser is tuned, Fig. 5(a), 5(c) and 5(e) plot RF spectra of DSHS signals with and without RBS fiber under the controlling pump power of 0 mW, 41.8 mW, and 81.9 mW. Linewidth narrowing is very obvious through an intuitionistic comparison of red and black curves. All measured RF spectra of beat signals can be well fitted by Lorentz curves, and their fitting R2 values are over 0.988. To avoid Gaussian broadening near the central frequency induced by 1/f noise, linewidths are estimated by the Lorentz data at 20 dB below peak values. 20-dB Lorentz linewidths of three wavelength channels without RBS feedback are deduced to 16 kHz, 18.8 kHz and 20.6 kHz, corresponding to estimated laser linewidths of ~800 Hz, ~940 Hz, ~1.03 kHz. Figure 5(b), (d) and (f) more clearly show the linewidths with RBS feedback within a smaller frequency span. With the Rayleigh feedback, the output linewidths can be narrowed to be ~200 Hz, ~245 Hz and ~255 Hz, corresponding to compression ratios of 4, 3.7 and 4.04, respectively. The linewidth narrowing effect is very remarkable at each wavelength channels and does not depend on operating wavelengths. In theory, to measure the exact linewidth, we should use the delay fiber length over 1000 km to achieve incoherent self-mixing frequency. However, due to the limitation of output powers and serious broaden induced by 1/f noise, it is impossible to accurately extract the Lorentz linewidth from the Gauss-dominant RF spectrum. Our experimental results are obtained by the 1.67 kHz theoretical precision of the DSHS with 60 km delay fiber. It is more beneficial for white noise to generate the Lorentzian line-shape with longer delay fiber, and insufficient delay fiber inversely leads to oscillation broadening in the self-heterodyne spectra [29,30]. Besides, the frequency spectrum obtained by this DSHS method is the Lorentzian spectrum convoluted with the Guassian spectrum due to the unavoidable 1/f frequency noise induced by 60 km long-length delay fiber. We directly fit the original experiment data by the Lorentz curve. Therefore, the fitting results also include the Guassion part. The real linewidth does not exceed the measured results. Owing to the well Lorentz-shaped spectra, these results can be regarded as the conservative characterization of natural laser linewidth to demonstrate the linewidth narrowing law in optical-controlled processes.

 figure: Fig. 5

Fig. 5 (a), (b), (c), (d), (e), and (f) Linewidth measurements and Lorentz fitting curves with versus without the RBS feedback under arbitrarily selected controlling pump powers of 0 mW, 41.8 mW and 81.9 mW;

Download Full Size | PDF

Above narrowing phenomenon of linewidths can be also confirmed from the frequency noise spectra obtained by the interference system in reference [31]. As shown in Fig. 6(a), the frequency noise curve has an obvious monolithic decline with the RBS feedback. The frequency noise has higher values in the low-frequency range, which is seriously influenced by the inherent RF noise of the drive signal source and the vibration of the fiber interference arms and experimental platform. As shown as the two dashed green lines, the estimated linewidths and compression ratio are well coincided with the previous results of the beat frequency spectra measured by the previous DSHS methods. Linewidth measurements obtained under different powers of controlling pump 3 are shown in Fig. 6(b). Considering the great influence on 60 km fiber delay of the DSHS because of temperature fluctuations and vibration noise, the linewidth is varied just within the range from ~200 Hz to ~330 Hz, indicating a maintained ultra-narrow linewidth performance when the laser is tuned.

 figure: Fig. 6

Fig. 6 (a) Frequency noise spectra of the laser with and without RBS feedback; (b) linewidth measurements with the RBS feedback under different controlling pump powers.

Download Full Size | PDF

4. Conclusion

In this paper, we have built an optically controlled tunable ultra-narrow linewidth fiber laser with a SAIR and fiber RBS feedback. Owing to the periodical index refraction modulation inside EDF’s core, the gain bandwidth of the laser cavity can be theoretically restricted within ~5 MHz, which is contributed to the mode selection and enhancement of long-term stabilities. Cooperated with the RBS feedback of the SMF, the output linewidth is low to be ~200 Hz. The laser tunability is achieved by the photo-induced self-heating of graphene that attaches on the surface of the micro-FBG. The output laser has a reversible power-tuning slope of 12.4 pm/mW with a linear fitting R2 of 0.99742 and maintains a SLM operation in the tuning process. The linewidth narrowing phenomenon at different wavelength channels is experimentally demonstrated. Taking advantage of the ultra-narrow linewidth performance with linear and precise light-controlled tunability, this fiber laser can afford potentials for high-precision sensing, all-optical signal processing, and coherent signal detection.

Funding

Key Research and Development Project of Ministry of Science and Technology (2016YFC0801200), the Natural Science Foundation of China (NSFC) (61635004), the Postdoctoral Science Foundation of Chongqing (Xm2017047), the Fundamental Research Funds for the Central Universities (106112017CDJXY120004), and the Science Fund for Distinguished Young Scholars of Chongqing (CSTC2014JCYJJQ40002).

References

1. Y. Koshikiya, X. Fan, and F. Ito, “Long range and cm-level spatial resolution measurement using coherent optical frequency domain reflectometry with SSB-SC modulator and narrow linewidth fiber laser,” J. Lightwave Technol. 26(18), 3287–3294 (2008). [CrossRef]  

2. A. Cygan, D. Lisak, P. Morzyński, M. Bober, M. Zawada, E. Pazderski, and R. Ciuryło, “Cavity mode-width spectroscopy with widely tunable ultra narrow laser,” Opt. Express 21(24), 29744–29754 (2013). [CrossRef]   [PubMed]  

3. J. Li, J. Gan, Z. Zhang, X. Heng, C. Yang, Q. Qian, S. Xu, and Z. Yang, “High spatial resolution distributed fiber strain sensor based on phase-OFDR,” Opt. Express 25(22), 27913–27922 (2017). [CrossRef]   [PubMed]  

4. H. Ishii, N. Fujiwara, K. Watanabe, S. Kanazawa, M. Itoh, H. Takenouchi, Y. Miyamoto, K. Kasai, and M. Nakazawa, “Narrow linewidth tunable DFB laser array integrated with optical feedback planar lightwave circuit (PLC),” IEEE J. Sel. Top. Quantum Electron. 23(6), 1501007 (2017). [CrossRef]  

5. S. Mo, X. Huang, S. Xu, C. Li, C. Yang, Z. Feng, W. Zhang, D. Chen, and Z. Yang, “600-Hz linewidth short-linear-cavity fiber laser,” Opt. Lett. 39(20), 5818–5821 (2014). [CrossRef]   [PubMed]  

6. W. Fan, J. Gan, Z. Zhang, X. Wei, S. Xu, and Z. Yang, “Narrow linewidth single frequency microfiber laser,” Opt. Lett. 37(20), 4323–4325 (2012). [CrossRef]   [PubMed]  

7. Z. Hu, F. Li, Z. Pan, and W. Tan, “Wavelength-tunable narrow-Linewidth semiconductor fiber-ring Laser,” IEEE Photonics Technol. Lett. 12(8), 977–979 (2000). [CrossRef]  

8. C. C. Lee, Y. K. Chen, and S. K. Liaw, “Single-longitudinal-mode fiber laser with a passive multiple-ring cavity and its application for video transmission,” Opt. Lett. 23(5), 358–360 (1998). [CrossRef]   [PubMed]  

9. Z. Meng, G. Stewart, and G. Whitenett, “Stable single-mode operation of a narrow-linewidth, linearly polarized, erbium-fiber ring laser using a saturable absorber,” J. Lightwave Technol. 24(5), 2179–2183 (2006). [CrossRef]  

10. Y. Liu, M. Zhang, J. Zhang, and Y. Wang, “Single-longitudinal-mode triple-ring Brillouin fiber laser with a saturable absorber ring resonator,” J. Lightwave Technol. 35(9), 1744–1749 (2017). [CrossRef]  

11. B. Fischer, J. L. Zyskind, J. W. Sulhoff, and D. J. Digiovanni, “Nonlinear wave mixing and induced gratings in erbium-doped fiber amplifiers,” Opt. Lett. 18(24), 2108–2110 (1993). [CrossRef]   [PubMed]  

12. T. Zhu, X. Bao, L. Chen, H. Liang, and Y. Dong, “Experimental study on stimulated Rayleigh scattering in optical fibers,” Opt. Express 18(22), 22958–22963 (2010). [CrossRef]   [PubMed]  

13. T. Zhu, S. Huang, L. Shi, W. Huang, M. Liu, and K. Chiang, “Rayleigh backscattering: a method to highly compress laser,” Chin. Sci. Bull. 59(33), 4631–4636 (2014). [CrossRef]  

14. N. Libatique, L. Wang, and R. Jain, “Single-longitudinal-mode tunable WDM-channel-selectable fiber laser,” Opt. Express 10(25), 1503–1507 (2002). [CrossRef]   [PubMed]  

15. Y. W. Song, S. A. Havstad, D. Starodubov, Y. Xie, A. E. Willner, and J. Feinberg, “40-nm-Wide tunable fiber ring laser with single-mode operation using a highly stretchable FBG,” IEEE Photonics Technol. Lett. 13(11), 1167–1169 (2001). [CrossRef]  

16. J. Zhang, C. Yue, G. Schim, W. Clements, and J. Lit, “Stable single-mode compound-ring erbium-doped fiber laser,” IEEE Photonics Technol. Lett. 14(1), 104–109 (1996).

17. N. Yan, X. Han, P. Chang, L. Huang, F. Gao, X. Yu, W. Zhang, Z. Zhang, G. Zhang, and J. Xu, “Tunable dual-wavelength fiber laser with unique gain system based on in-fiber acousto-optic Mach-Zehnder interferometer,” Opt. Express 25(22), 27609–27615 (2017). [CrossRef]   [PubMed]  

18. M. Delgado-Pinar, J. Mora, A. Díez, J. Cruz, and M. Andrés, “Wavelength-switchable fiber laser using acoustic waves,” IEEE Photonics Technol. Lett. 17(3), 552–554 (2005). [CrossRef]  

19. X. Cheng, P. Shum, C. Tse, J. Zhou, M. Tang, W. Tan, R. Wu, and J. Zhang, “Single-longitudinal-lode erbium-doped fiber ring laser based on high finesse fiber Bragg grating Fabry–Pérot etalon,” IEEE Photonics Technol. Lett. 20(12), 976–978 (2008). [CrossRef]  

20. T. Feng, D. Ding, F. Yan, Z. Zhao, H. Su, and X. S. Yao, “Widely tunable single-/dual-wavelength fiber lasers with ultra-narrow linewidth and high OSNR using high quality passive subring cavity and novel tuning method,” Opt. Express 24(17), 19760–19768 (2016). [CrossRef]   [PubMed]  

21. N. Halip, M. Bakar, A. Latif, S. Yasin, M. Zulkifli, K. Sharif, N. Omar, A. Mansoor, H. Rashid, and M. Mahdi, “A narrow linewidth tunable single longitudinal mode Ga-EDF fiber laser,” Opt. Laser Technol. 101, 413–418 (2018). [CrossRef]  

22. X. Gan, Y. Wang, F. Zhang, C. Zhao, B. Jiang, L. Fang, D. Li, H. Wu, Z. Ren, and J. Zhao, “Graphene-controlled fiber Bragg grating and enabled optical bistability,” Opt. Lett. 41(3), 603–606 (2016). [CrossRef]   [PubMed]  

23. S. Yu, X. Wu, Y. Wang, X. Guo, and L. Tong, “2D Materials for optical modulation: challenges and opportunities,” Adv. Mater. 29(14), 1606128 (2017). [CrossRef]   [PubMed]  

24. J. Chen, B. Zheng, G. Shao, S. Ge, F. Xu, and Y. Lu, “An all-optical modulator based on a stereo graphene–microfiber structure,” Light Sci. Appl. 4(12), e360 (2015). [CrossRef]  

25. Y. S. Thirstrun and B. Phlsdottir, “Pump-induced refractive index modulation and dispersions in Er3+-doped fibers,” J. Lightwave Technol. 14(5), 732–738 (1996). [CrossRef]  

26. T. Erdogan, “Fiber grating spectra,” J. Lightwave Technol. 15(8), 1277–1294 (1997). [CrossRef]  

27. E. Desurvire, Erbium-Doped Fiber Amplifiers: Principles and Applications (John Wiley& Sons, 1994).

28. A. A. Balandin, S. Ghosh, W. Bao, I. Calizo, D. Teweldebrhan, F. Miao, and C. N. Lau, “Superior thermal conductivity of single-layer graphene,” Nano Lett. 8(3), 902–907 (2008). [CrossRef]   [PubMed]  

29. M. Chen, Z. Meng, J. Wang, and W. Chen, “Ultra-narrow linewidth measurement based on Voigt profile fitting,” Opt. Express 23(5), 6803–6808 (2015). [CrossRef]   [PubMed]  

30. L. E. Richer, H. I. S. Kruger, and P. A. Mcgrath, “Linewidth determination from self-heterodyne measurement with sucoherence delay times,” Quan. Electron. Lett. 22(11), 2070–2074 (1986). [CrossRef]  

31. O. Llopis, P. H. Merrer, H. Brahimi, K. Saleh, and P. Lacroix, “Phase noise measurement of a narrow linewidth CW laser using delay line approaches,” Opt. Lett. 36(14), 2713–2715 (2011). [CrossRef]   [PubMed]  

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (6)

Fig. 1
Fig. 1 (a) Experimental setup of the optically tunable ultra-narrow linewidth fiber laser; (b) photograph of the graphene-coated micro-FBG (GCMFBG), and the red shadow part is the graphene-coated region. (c) Phonon relaxation process of graphene; (d) red-shifted reflection spectra of the GCMFBG with controlling light powers of 26.8 mW, 64.3 mW, and 101.5 mW.
Fig. 2
Fig. 2 (a) RF spectrum without Rayleigh feedback directly measured by the PD and the corresponding beat-frequency signal measured by the DSHS; (b) RF spectrum with Rayleigh fiber directly measured by PD and the corresponding RF spectrum of the beat-frequency signal.
Fig. 3
Fig. 3 (a) Output RF spectra with Rayleigh fiber with the increment of the attenuation value; (b) time domain of the beat-frequency signal with a SLM operation.
Fig. 4
Fig. 4 (a) Output spectra of the ultra-narrow linewidth laser with the increment of the pump 3; (b) color-corresponded RF spectra during the tuning process; (c) peak wavelengths with changed powers of pump 3 and the linear fitting curve; (d) long-term stability test of the wavelength and output power with a fixed power of pump 3.
Fig. 5
Fig. 5 (a), (b), (c), (d), (e), and (f) Linewidth measurements and Lorentz fitting curves with versus without the RBS feedback under arbitrarily selected controlling pump powers of 0 mW, 41.8 mW and 81.9 mW;
Fig. 6
Fig. 6 (a) Frequency noise spectra of the laser with and without RBS feedback; (b) linewidth measurements with the RBS feedback under different controlling pump powers.

Equations (2)

Equations on this page are rendered with MathJax. Learn more.

n ( λ , z ) n e f f ( λ ) ¯ 1 + I 0 I s a t cos ( 4 π n e f f ( λ ) ¯ λ z ) ,
Δ f ( λ , Δ n ) Δ n c δ λ 2 n e f f ¯ 2 ( λ ) Δ n 2 + λ 2 L 2 E D F ,
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.