Open Access
Add to BibSonomyAbstract
We demonstrate the frequency stabilization of a high output power, erbium silica fiber laser by utilizing a 13C2H2 (acetylene) absorption line at 1538.8 nm and a H13C14N (hydrogen cyanide) absorption line at 1549.73 nm. We introduced a novel short ring cavity configuration and pump power feedback control to suppress the intensity noise of the laser output, which is caused by the relaxation oscillation of erbium ions. As a result, we succeeded in simultaneously obtaining a stable single-frequency oscillation with an output power of over 290 mW, a linewidth of 5 kHz, and a low relative intensity noise (RIN) of −120 dB/Hz. The frequency stabilities reached 2.8 × 10−11 and 6.9 × 10−11 for an integration time of 1 s with a 13C2H2 and a H13C14N absorption line, respectively.
© 2016 Optical Society of America
1. Introduction
A frequency-stabilized laser operating in the 1.5 μm region with a high output power, a high optical signal-to-noise ratio (OSNR), a narrow linewidth, and a low intensity noise is very attractive for use in the fields of coherent optical communication, optical metrology, and high-resolution interferometric measurement. This laser plays an important role in ultra-multilevel quadrature amplitude modulation (QAM) coherent optical transmission, where optical amplitude and phase are multi-level modulated simultaneously with a high OSNR [1]. It is also helpful in improving the performance of the delivery systems of an optical frequency standard and a highly accurate optical clock via an optical fiber network [2], and of a precise optical seismometer by employing a laser interferometer [3]. Several types of frequency-stabilized lasers operating in this wavelength band have already been reported, and they use atomic and molecular absorption lines as frequency references [4–6]. Of the many available lasers, fiber lasers are particularly attractive candidates for such applications because their linewidths are narrow due to the long cavity length. For example, the oscillation frequencies of erbium fiber lasers with a linewidth of several kHz have been stabilized to a C2H2 (acetylene) absorption line [7–9]. However, the output power of these lasers, which employ a silica-based erbium-doped fiber (EDF), has been limited to several tens of mW. This is attributed to concentration quenching in silica EDF [10], self-pulsation induced by the relaxation oscillation of erbium ions [10], and multi-mode oscillation resulting from a long cavity configuration, especially in a high pump power condition.
Intensive efforts have been made to increase the output power by employing a heavily Er/Yb-codoped phosphate glass fiber (EYDF) as a short-length, high-gain medium [11] without quenching, or using a master-oscillator power amplifier (MOPA) scheme [12]. A clad pumping scheme has also been employed to excite a double-clad EYDF with an extremely high pump power by adopting a multimode pump source [13]. However, the reliability of phosphate glass fiber including its splicing to silica fibers has not been guaranteed for telecommunication applications where long-term reliability is strongly required. In addition, the laser configuration becomes complicated with a MOPA system or clad pumping, while the use of MOPA inevitably degrades the OSNR because of amplified spontaneous emission (ASE) noise from the amplifiers.
To increase the output power of an erbium fiber laser, we developed a continuous wave (CW) erbium-doped silica fiber ring laser (EFRL) with a short ring cavity configuration [14], which can emit a high power, stable single-frequency output even under a high pump power condition. In addition, by using a pump power feedback control to suppress the relaxation oscillation of the erbium ions [15], we demonstrated a 1538.8 nm, 160 mW output 13C2H2 frequency-stabilized laser with a linewidth of 5 kHz and a relative intensity noise (RIN) of −130 dB/Hz [16].
In this paper, by applying a bidirectional pumping scheme to our EFRL, we have successfully increased its output power to over 290 mW, and demonstrated its 13C2H2 frequency stabilization. To the best of our knowledge, this is the highest output power yet reported for a frequency-stabilized CW fiber laser employing a core pumped silica EDF without a MOPA system. In addition, we also demonstrated a frequency-stabilized EFRL by utilizing a hydrogen cyanide (H13C14N) absorption line as an optical frequency reference [17].
2. Configuration and output characteristics of 13C2H2 frequency-stabilized CW EFRL
Figure 1(a) shows the configuration of our single-frequency EFRL. It consists of two 1.48 μm laser diodes (LD1, LD2), polarization-maintained erbium-doped fiber (PM-EDF), a 4-port PM optical circulator, a 1.2 GHz narrow band PM-fiber Bragg grating (PM-FBG) filter with a reflection center wavelength of around 1538.8 nm, a PM wavelength division multiplexing (WDM) coupler, a PM optical isolator, two photo detectors (PDs) and two feedback circuits. The reflection spectrum of the PM-FGB is shown in Fig. 1(b). The 4-port PM circulator acts simultaneously as an isolator, a polarizer, and a WDM coupler that can couple the 1.48 μm pump signal of LD1 to the EDF. The 1.48 μm pump signal of LD2 is coupled to the EDF through the WDM coupler and the PM-FBG with a coupling ratio of approximately 70%. To obtain a high output power without concentration quenching, we adopted a highly Al co-doped EDF with high Er (4500 ppm) and Al (12 wt%) concentrations. The EDF was wound around a drum-type piezoelectric transducer (PZT) to tune the oscillation frequency by changing the voltage applied to the PZT with a tuning ratio of 5.2 MHz/V. The FBG filter was laid on a multi-layer PZT (MLP). By applying a voltage to the MLP, the center wavelength of the FBG was tuned with a tuning ratio of 1 GHz/V. In addition, its center wavelength was feedback controlled to suppress mode hopping by tracking the oscillation longitudinal mode using feedback circuit-1 [8]. The maximum frequency tuning range of the EFRL was approximately 125 GHz (1 nm), which was determined by the maximum allowable voltage applied to the MLP of 150 V. We employed active pump power control to suppress the relaxation oscillation of the erbium ions. Figure 2 shows the configuration of our relaxation oscillation control circuit. Here, a relaxation oscillation component is extracted and negatively fed back to the pump power of 1.48 μm LD1 to suppress the oscillation. This was achieved with feedback circuit-2, which consisted of low-pass and high-pass filters. The bandwidth of this feedback system was optimized to 100 Hz-1 MHz. A CW signal with a low intensity noise is output from one end of the FBG followed by an isolator.
Fig. 1 (a) Configuration of high power, single-frequency EFRL, (b) reflection spectrum of installed FBG filter.
Figure 3 shows the configuration of the 13C2H2 frequency stabilization circuit of the EFRL. In this circuit, we employed a phase sensitive detection (PSD) circuit consisting of a LiNbO3 (LN) phase modulator, a 13C2H2 cell (3 Torr), a PD, a double balanced mixer (DBM), and a feedback circuit based on proportional and integral (PI) control. Figure 4(a) shows the 13C2H2 absorption lines observed for a long span. Among these lines, we selected the P(10) linear absorption line with a center wavelength of 1538.80 nm and a spectral width of 500 MHz as shown in Fig. 4(b). In the PSD circuit, to detect frequency fluctuations from the 13C2H2-P(10) absorption peak, an optical beam from the EFRL was phase-modulated by using an LN phase modulator driven at a modulation frequency of 10 MHz with a modulation width of 110 MHz. The DBM generated a voltage error signal that was proportional to the frequency deviation, and was fed back to the PZT in the laser cavity through a PI circuit with a bandwidth of 100 Hz.
Fig. 4 13C2H2 absorption lines observed for (a) a long span, and (b) a P(10) linear absorption line.
Figure 5 shows the laser output power characteristics as a function of pump power for different EDF lengths. Two pump LDs operated at a pump power exceeding 490 mW, while only pump LD1 operated at a pump power of less than 490 mW. At these EDF lengths, we obtained single-frequency oscillation even with high pump powers. With an EDF length of over 2.5 m, multi-mode oscillation occurred, where a few longitudinal modes appeared simultaneously. Thus, we chose an EDF length of 2 m and obtained an output power of over 290 mW from a maximum pump power of 980 mW. This is the highest power yet obtained in a silica-based EFRL with a core pumping scheme. Under this condition, the total cavity length was approximately 2.5 m, which corresponds to a free spectral range of 80 MHz, and the slope efficiency was 30.2%.
Figures 6(a) and 6(b) show the optical spectra of the 13C2H2 frequency-stabilized EFRL measured with a 0.01 nm resolution bandwidth for 200 nm and 5 nm spans, respectively. The OSNR was as high as 80 dB. A high output power, a narrow linewidth, and low intensity noise characteristics are very important when the present laser is employed for a multi-level coherent transmission system or precise interferometric measurements. Figure 7 shows the heterodyne beat spectrum between the present frequency-stabilized EFRL and a reference single-frequency laser diode with a linewidth of 100 kHz. Only one beat signal is seen in the 300 MHz span, which indicates that this EFRL operates in a single-frequency oscillation. The side-mode suppression ratio (SMSR) was over 60 dB as seen in Fig. 7.
Fig. 6 Optical spectrum of 13C2H2 frequency-stabilized CW EFRL (0.01 nm resolution bandwidth) with 200 and 5 nm spans, respectively.
Figure 8 shows the spectral profile of the 13C2H2-P(10) linear absorption line and the corresponding 1st derivative signal obtained with the PSD circuit. The slope and the signal-to-noise ratio (SNR) of the derivative signal are 36.7 V/GHz and approximately 88 dB, respectively. This signal was used for feedback control of the laser frequency. The short-term stability of a frequency-stabilized laser can be estimated from the Q value of an absorption line, the SNR of the derivative signal, the feedback bandwidth of B, and the integration time τ, which is given by σ(τ) = 0.2/[Q × (SNR) × B1/2 × τ1/2] [18]. Here, Q is defined as Q = ν0/Δν, where ν0 and Δν represent absorption frequency and absorption linewidth, respectively. In our case, the Q value, SNR and B were 3.88 × 105, 88 dB and 100 Hz, respectively. Therefore, the estimated stability is 2.1 × 10−11τ -1/2. We evaluated the frequency stability of this laser using the Allan deviation [19] obtained from the beat note signal between two identical frequency-stabilized lasers. The result is shown in Fig. 9, where we measured the beat frequency using a frequency counter with a gate time of 1 s. This figure also shows the estimated stability curve as described above. For τ = 1 s, the Allan deviation was 2.8 × 10−11. For τ = 100 s, it was 8.8 × 10−12. In our laser, the short-term frequency stability was almost the same as the estimated value. However, the long-term stability was degraded. This may be due to optical intensity fluctuation in the phase sensitive detection circuit and the slow offset-voltage fluctuation of the operational amplifier in the PI-feedback circuit.
We also measured the linewidth of the frequency-stabilized EFRL using a delayed self-heterodyne detection method with a 50 km-delay fiber. Figures 10(a) and 10(b) show the delayed self-heterodyne spectra of the laser plotted on linear and log scales, respectively. From Fig. 10(a), we estimated the laser linewidth to be approximately 5 kHz. This linewidth is the same as that of the original EFRL under a free running condition. This result indicates that the frequency-stabilized EFRL has an output beam without any additional modulation components during the course of frequency stabilization.
Fig. 10 Delayed self-heterodyne spectra of 13C2H2 frequency-stabilized CW EFRL plotted on (a) linear and (b) log scale.
The relative intensity noise (RIN) spectrum from 100 kHz to 1 MHz is shown in Fig. 11. The output power fluctuation of the laser within a 5 MHz bandwidth is also shown in Figs. 12(a) and 12(b). This fluctuation originates from the relaxation oscillation of erbium ions. These figures compare the output characteristics with and without pump power feedback control. We successfully suppressed the RIN peak at around 425 kHz. As a result, the RIN level was improved to less than −120 dB/Hz and the output power fluctuation was reduced from 5% to 0.8%.
Fig. 12 (a) Power fluctuation (DC~5 MHz) of laser output of 13C2H2 frequency-stabilized CW EFRL, and (b) enlarged view.
3. Output characteristics of H13C14N frequency-stabilized CW EFRL
H13C14N gas has many absorption lines in the 1530-1560 nm wavelength region [17]. Here, we fabricated a single-frequency EFRL emitting at around 1550 nm, and stabilized its frequency to the 1549.73 nm H13C14N-P(10) linear absorption line. The laser configuration was the same as that of the 13C2H2 frequency-stabilized EFRL shown in Figs. 1-3 except that it included a PM-FBG filter with a center wavelength of around 1549.73 nm and an H13C14N gas cell in the PSD circuit.
Figure 13 shows the output power characteristics of a 1550 nm, H13C14N frequency-stabilized CW EFRL as a function of pump power. We obtained an optical output power of 295 mW with a pump power of 980 mW. The slope efficiency was 30.7%. Figures 14(a) and 14(b) show optical spectra of this laser measured with a 0.01 nm resolution at spans of 200 and 5 nm, respectively. The OSNR was as high as 82 dB.
Fig. 13 Output power characteristics of H13C14N frequency-stabilized CW EFRL as a function of pump power.
Fig. 14 Optical spectrum of H13C14N frequency-stabilized CW EFRL (0.01 nm resolution bandwidth) with 200 and 5 nm spans, respectively.
We used a 165 mm long commercially available HCN gas cell filled with H13C14N gas at a pressure of 1 Torr. Figures 15(a) and 15(b) show the absorption lines of H13C14N gas for a broad span and a P(10) linear absorption line, respectively. The linewidth of the P(10) absorption line was 500 MHz. Here, we optimized the phase modulation width in the PSD circuit to 175 MHz, with which we obtained the maximum SNR of the 1st derivative signal. Figure 16 shows the spectral profile of the H13C14N-P(10) linear absorption line and the corresponding 1st derivative signal. The slope and the SNR of the derivative signal were 18.5 V/GHz and approximately 82 dB, respectively. The estimated short-term stability of this laser was 4.1 × 10−11τ -1/2. The slope and the SNR of the derivative signal of the H13C14N-P(10) absorption line were degraded compared with those obtained with the 13C2H2-P(10) absorption line. This was because the frequency modulation (FM)-amplitude modulation (AM) conversion coefficient of the H13C14N-P(10) absorption line, which was estimated from the slope of the absorption curve, was small, in comparison with that of the 13C2H2 absorption. This was due to the relatively weak H13C14N absorption depth in Fig. 15 compared with that in Fig. 4.
Fig. 15 H13C14N absorption lines observed for a broad span (a), and a specific P(10) linear absorption line (b).
Fig. 16 H13C14N-P(10) absorption line and its first derivative signal obtained with the PSD circuit.
Figure 17 shows the Allan deviation of the frequency fluctuation of this laser. This figure also shows the estimated stability curve. The Allan deviation was 6.9 × 10−11 for τ = 1 s, and 1.0 × 10−11 for τ = 100 s, which were 1.1~2.5 times larger than those of the 13C2H2 frequency-stabilized EFRL. This is attributed to the low SNR voltage error signal that originated from the relatively weak H13C14N absorption. Figures 18(a) and 18(b) show the delayed self-heterodyne spectra of the frequency-stabilized EFRL plotted on linear and log scales, respectively. From these results, we estimate the linewidth of this laser to be approximately 5 kHz. This is the same as that of the free running EFRL, and so again there was no spectral broadening associated with the present frequency stabilization.
Fig. 18 Delayed self-heterodyne spectra of H13C14N frequency-stabilized CW EFRL plotted on (a) linear and (b) log scale.
The RIN spectrum and the output power stability are shown in Figs. 19 and 20, respectively. By employing a pump power control, the RIN level was reduced to less than −120 dB/Hz, and the output power stability improved. The output power fluctuation of this laser was only 0.8% (DC~5 MHz).
Fig. 20 (a) Output power fluctuation (DC~5 MHz) of H13C14N frequency-stabilized CW EFRL, (b) expanded view.
4. Conclusion
We described a 13C2H2, and an H13C14N frequency-stabilized EFRL that employ a bidirectional pumping scheme and pump power feedback control as well as a short cavity configuration. With this configuration, we realized a frequency-stabilized EFRL with an output power of over 290 mW, a linewidth of 5 kHz and a RIN of less than −120 dB/Hz. The frequency stability of the 13C2H2 frequency-stabilized laser reached 2.8 × 10−11 for τ = 1 s and 8.8 × 10−12 for τ = 100 s. With an H13C14N absorption line, the frequency stabilities for τ = 1 s and τ = 100 s were 6.9 × 10−11 and 1.0 × 10−11, respectively. These lasers are expected to constitute an attractive light source for multi-level coherent transmission with higher-order multiplicity and for various precise interferometric measurements.
Acknowledgments
This work was supported by a JSPS Grant-in Aid for Scientific Research, Grant number 26706016.
References and links
1. S. Beppu, K. Kasai, M. Yoshida, and M. Nakazawa, “2048 QAM (66 Gbit/s) single-carrier coherent optical transmission over 150 km with a potential SE of 15.3 bit/s/Hz,” Opt. Express 23(4), 4960–4969 (2015). [CrossRef] [PubMed]
2. S. M. Foreman, K. W. Holman, D. D. Hudson, D. J. Jones, and J. Ye, “Remote transfer of ultrastable frequency references via fiber networks,” Rev. Sci. Instrum. 78(2), 021101 (2007). [CrossRef] [PubMed]
3. A. Araya, K. Sekiya, and Y. Shindo, “Laser-interferometric broadband seismometer for ocean borehole observations, Feasibility of giant fiber-optic gyroscopes,” in Proceedings of the Underwater Technology and Workshop on Scientific Use of Submarine Cables and Related Technologies (2007), pp. 245–248.
4. M. Ohtsu and E. Ikegami, “Frequency stabilization of 1.5 μm DFB laser using internal second harmonic generation and atomic 87Rb line,” Electron. Lett. 25(1), 22–23 (1989). [CrossRef]
5. T. Yanagawa, S. Saito, and Y. Yamamoto, “Frequency stabilization of 1.5-μm InGaAsP distributed feedback laser to NH3 absorption lines,” Appl. Phys. Lett. 45(8), 826–828 (1984). [CrossRef]
6. M. de Labachelerie, K. Nakagawa, Y. Awaji, and M. Ohtsu, “High-frequency-stability laser at 1.5 μm using Doppler-free molecular lines,” Opt. Lett. 20(6), 572–574 (1995). [CrossRef] [PubMed]
7. M. Têtu, F. Ouellette, C. Latrasse, R. Larose, D. Stepanov, and M. A. Duguay, “Simple frequency tuning technique for locking a single mode erbium-doped fiber laser to the center of molecular resonances,” Electron. Lett. 30(10), 791–793 (1994). [CrossRef]
8. K. Kasai, A. Suzuki, M. Yoshida, and M. Nakazawa, “Performance improvement of an acetylene (C2H2) frequency-stabilized fiber laser,” IEICE Electron. Express 3(22), 487–492 (2006). [CrossRef]
9. K. Knabe, S. Wu, J. Lim, K. A. Tillman, P. S. Light, F. Couny, N. Wheeler, R. Thapa, A. M. Jones, J. W. Nicholson, B. R. Washburn, F. Benabid, and K. L. Corwin, “10 kHz accuracy of an optical frequency reference based on 12C2H2-filled large-core kagome photonic crystal fibers,” Opt. Express 17(18), 16017–16026 (2009). [CrossRef] [PubMed]
10. F. Sanchez, P. L. Boudec, and P. L. Francois, “Effects of ion pairs on the dynamics of erbium-doped fiber lasers,” Phys. Rev. A 48(3), 2220–2229 (1993). [CrossRef] [PubMed]
11. C. Spiegelberg, J. Geng, Y. Hu, Y. Kaneda, S. Jiang, and N. Peyghambarian, “Low-noise narrow-linewidth fiber laser at 1550 nm (June 2003),” J. Lightwave Technol. 22(1), 57–62 (2004). [CrossRef]
12. A. Fujisaki, S. Matsushita, K. Kasai, M. Yoshida, T. Hirooka, and M. Nakazawa, “An 11.6 W output, 6 kHz linewidth, single-polarization EDFA-MOPA system with a 13C2H2 frequency stabilized fiber laser,” Opt. Express 23(2), 1081–1087 (2015). [CrossRef] [PubMed]
13. T. Qiu, L. Li, A. Schülzgen, V. Temyanko, T. Luo, S. Jiang, A. Mafi, J. Moloney, and N. Peyghambarian, “Generation of 9.3-W multimode and 4-W single-mode output from 7-cm short fiber lasers,” IEEE Photonics Technol. Lett. 16(12), 2592–2594 (2004). [CrossRef]
14. S. Masuda, S. Niki, and M. Nakazawa, “Environmentally stable, simple passively mode-locked fiber ring laser using a four-port circulator,” Opt. Express 17(8), 6613–6622 (2009). [CrossRef] [PubMed]
15. G. A. Ball, C. Holton, G. Hull-Allen, and W. Morey, “60 mW 1.5 μm single-frequency low-noise fiber laser MOPA,” IEEE Photonics Technol. Lett. 6(2), 192–194 (1994). [CrossRef]
16. K. Kasai, A. Fujisaki, M. Yoshida, T. Hirooka, M. Nakazawa, and S. Masuda, “A 160 mW output, 5 kHz linewidth frequency-stabilized erbium silica fiber laser with a short cavity configuration,” in Proceedings of the Conf. on Lasers and Electro-Optics (CLEO) (2014), pp. SW1N.4. [CrossRef]
17. H. Sasada and K. Yamada, “Calibration lines of HCN in the 1.5-μm region,” Appl. Opt. 29(24), 3535–3547 (1990). [CrossRef] [PubMed]
18. J. Vanier and L. G. Bernier, “On the signal-to-noise ratio and short-term stability of passive rubidium frequency standards,” IEEE Trans. Instrum. Meas. 30(4), 277–282 (1981). [CrossRef]
19. D. W. Allan, “Statistics of atomic frequency standards,” Proc. IEEE 54(2), 221–230 (1966). [CrossRef]
