Open Access
Add to BibSonomy
Abstract
A high peak-power and narrow-linewidth all-fiber Raman pulsed laser operating around 1.65 µm is introduced. A 1541 nm laser seed is modulated into pulse trains, which will be used as the Raman pump laser, by driving a reflective semiconductor optical amplifier (RSOA) with a continuous periodic square-wave voltage. A homemade high peak-power 1541 nm pulsed laser is employed to modulate and amplify a 1653.7 nm distributed feedback laser (DFB) seed synchronously in a segment of the 52-meter-long highly germania-doped fiber (HGDF). The repetition-rate and the pulse-width of the 1653.7 nm pulsed laser are 100 kHz and 31 ns, respectively. The peak power is estimated to be as high as about 30.85 W, and a 3-dB linewidth as narrow as less than 0.08 nm is achieved when the average power of 1541 nm pump is 3.1 W. The wavelength of Raman pulsed laser can be tuned from 1652.0 nm to 1654.0 nm continuously with an optical signal-to-noise ratio (OSNR) of more than 35 dB.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Methane (CH4) is a kind of greenhouse gas, as well as the main component of natural gas, biogas, coal-mine gas, etc. However, it can cause very dangerous asphyxia or even explosion when the CH4 concentration in the air reaches a certain level. Thus, it is important to monitor the ambient CH4 concentration in some particular places such as natural gas pipelines. Compared to the traditional sensors, optical sensors have the merits of high-precision, fast-response and durability in detecting the CH4 concentration. To date, several optical methods have been developed to measure CH4. In 2013, Chen et al., combined a Fabry-Perot quantum cascaded (QC) laser operating at 7.5 µm and a liquid nitrogen (LN) cooled HgCdTe mid-infrared detector to monitor the CH4 concentration [1]. The QC laser was driven by a microsecond pulsed current and a precision of 2 ppm was obtained with a pulse-to-pulse normalization technique. Nevertheless, this system is not suitable for long-distance remote sensing due to the high loss atmospheric transmission at 7.5 µm. In addition, the LN cooled detector needs the frequent supplement of LN as LN volatilizes easily at room temperature, which limits the long-term maintenance-free operation of this system outdoors. In 2015, Bin et al., presented a near-infrared CH4 detection system based on tunable diode laser absorption spectroscopy (TDLAS) [2]. They used a DFB seed which had the central wavelength of 1.654 µm as the optical source to meet the CH4’s absorption band in 1.65 µm waveband. Two InGaAs photodetectors which can operate normally at room temperature were employed to convert the 1.654 µm light into the reference signal and the absorption signal. Although absorption bands for H2O and CO2 also exist in 1.65 µm waveband, their absorption intensities are 4 order smaller than that of CH4. Therefore, these two molecules hardly cause any effect on the CH4 detection.
It is well-known that a high-power and narrow-linewidth laser can enable the ability of long-distance and high-precision gas detection. However, to the best of our knowledge, the output power of the commercial narrow-linewidth 1.65 µm laser diode (LD) is less than 20 mW. Hence it is very useful to improve the laser output power. Unfortunately, the usual rare-earth doped fibers such as erbium-doped fiber (EDF) and thulium-doped fiber (TDF) are unable to provide effective stimulated emission gain around 1.65 µm. Stimulated Raman scattering (SRS) is an important nonlinear phenomenon which can provide an extra optical gain that the rare-earth doped fibers cannot. In 1972, R. H. Stolen et al., demonstrated the first Raman gain in glass optical fibers [3]. They used an Xe laser (526 nm) to pump a segment of 5.9-meter-long single-mode cladded glass fiber and signal at 535.3 nm was observed. In the dense wavelength division multiplexing (DWDM) long-haul optical communication system, the Raman amplifiers can meet the needs of amplifying the optical carrier signals covering the S-band and increase the channel capacity greatly [4]. During the past decades, researchers have developed many high-power Raman lasers which are pumped by different kinds of high-power ytterbium-doped fiber lasers (YDFLs) [5–8]. For example, in 2018, V. Balaswamy et al., made a cascaded Raman laser system to transfer the lasing wavelength from 1170 nm to 1480 nm [9]. In this letter, we demonstrate a high peak-power and narrow-linewidth Raman pulsed fiber laser in 1.65 µm waveband by employing a homemade high peak-power 1541 nm pulsed laser to amplify a commercial 1653 nm continuous-wave (CW) DFB seed. The peak power of the Raman pulsed laser can be amplified to 30.85 W and the OSNR of over 35 dB is achieved.
2. Experimental setup
Figure 1 shows the schematic diagrams for the proposed all-fiber 1.65 µm Raman pulsed laser. We choose 1541 nm light as the Raman pump as it can provide Raman gain in 1.65 µm waveband. Figure 1(a) shows a 1541 nm CW fiber laser constructed with a linear cavity. A pair of high-reflection fiber Bragg grating (HR-FBG-1) is employed to determine the resonant wavelength. The central wavelength, reflectivity and reflection bandwidth of the HR-FBG-1 are 1541 nm, >99.7% and 0.98 nm, respectively. The gain fiber is the single-mode erbium-doped fiber (SM-EDF; OFS, GP980) whose nominal peak absorption near 980 nm is 8.74 dB/m and a segment of about 3.3-meter-long SM-EDF is used in this linear cavity. A 980/1550 nm wavelength-division multiplexing (WDM1) coupler is used to connect a 976 nm LD for pumping the SM-EDF. 70% laser power is spilt out from the resonant cavity through an optical coupler (OC) with 30:70 power ratio.
Fig. 1. (a) The configuration of the 1541 nm CW fiber laser as the Raman pump seed (976 LD: 976 nm laser diode, WMD1: 980/1550 nm wavelength-division multiplexing coupler, HR-FBG-1: high-reflection fiber Bragg grating, SM-EDF: single-mode erbium-doped fiber, OC: optical coupler). (b) The main schematic diagrams for the all-fiber 1.65 µm Raman pulsed laser (Seed: 1541 nm laser, RF: radio frequency, RSOA: reflective semiconductor optical amplifier, CIR: optical circulator, ISO1: 1541 nm optical isolator, DC-EYDF, double-cladding erbium/ytterbium co-doped fiber, 915 LD: 915 nm laser diode, HP-ISO: 1541 nm high-power isolator, WDM2: 1541/1653 nm wavelength-division multiplexing coupler, HGDF: highly germania-doped fiber, ISO2: 1653 nm optical isolator, 1653 DFB: 1653 nm distributed feedback laser as the seed).
Figure 1(b) shows the pulse modulation and the amplification, including a three-stage amplifier of the 1541 nm laser seed and the Raman amplifier for the 1653 nm DFB seed. First, the 1541 nm laser seed from Fig. 1(a) is injected into a RSOA (CIP-SOA-RL-OEC-1550-1.55µm) through an optical circulator (CIR) and modulated into pulse trains via the RSOA, which is driven by a radio frequency (RF) signal generator (Agilent 81110A). Then the pulse trains are sent to two similar forward-pumped amplifiers including the 1541 nm optical isolators (ISO1), 976 nm LDs, WMD1s and SM-EDF. The lengths of SM-EDF in the first stage and second stage amplifiers are 4.7 meter and 7.1 meter, respectively. After the pre-amplification, the pulse trains are input into a 1541 nm HR-FBG-2, which has a 3-dB bandwidth of 2 nm and the unwanted amplified spontaneous emission (ASE), which appears during the process of previous amplification, can be filtered out. Next, we use a segment of about 3-meter-long double-cladding erbium/ytterbium co-doped fiber (DC-EYDF), which is pumped by a high-power 915 nm LD to further improve the pulse energy of the 1541 nm laser. A 1541 nm high-power isolator (HP-ISO) is used to avoid any light reflection and consequent damage to the previous amplifiers. It can be seen that in all the three stages of amplification for the 1541 nm laser seed, we choose the forward-pumping configuration rather than back-pumping or bi-directional pumping configuration. Although back-pumping and bi-directional pumping configurations have higher conversion efficiency, the backward-pumping configuration has the worst OSNR performance and the bi-directional pumping configuration increases the complexity and cost of the whole system. On the other hand, in these two kinds of pumping configurations, permanent damage to our 976/915 nm LDs (without pump protectors such as isolators) may occur if a very high-energy laser pulse is generated accidentally and propagates into the LDs. Therefore, we choose the forward-pumping configuration to be simple, cost-effective and stable.
Finally, the amplified high-power 1541 nm pulse trains are employed to backward pump a segment of about 52-meter-long HGDF through a 1541/1653 nm wavelength-division multiplexing coupler (WDM2). The core diameter, cladding diameter and GeO2 doping concentration of our HGDF are 4 µm, 125 µm and 75 mol.%, respectively. Compared to the normal highly nonlinear fiber (HNLF), HGDF has the same small core diameter but much higher GeO2 doping concentration, which enables stronger Raman gain coefficient than that of the HNLF [10,11]. The single-frequency 1653 nm DFB seed has the maximum output power of about 12 mW and a side-mode suppression ratio (SMSR) of more than 54.5 dB. The WDM2 between the HGDF and the 1653 nm optical isolator (ISO2) is employed to spill the residual 1541 nm pump laser. The fiber connector type of the output end is FC/APC type, which has an oblique angle of 80. All optical fiber devices are spliced together to be compact and robust and minimize the transmission loss.
3. Results and discussion
Figure 2(a) shows the output spectrum of the 1541 nm laser seed [corresponding to Fig. 1(a)] at 20 mW, which is measured by an optical spectrum analyzer (OSA; Agilent 86142B) with a resolution of 0.06 nm. The central wavelength of this seed is located at 1541.24 nm and the OSNR of over 45 dB is achieved. The inset in Fig. 2(a) shows the corresponding RF spectrum, which is measured by a RF spectrum analyzer (Agilent E4447A) with a resolution of 100 kHz. The seed operates in the multi-longitudinal-mode (MLM) state as many beat frequency signals appear in this RF domain. The longitudinal mode interval can be estimated to be about 19.7 MHz from the beat frequency signals, corresponding to approximately 5.22-meter-long resonant cavity. This laser seed of MLM operation will broaden the linewidth, and consequently increase the Brillouin threshold in the HGDF. This can increase the Raman amplification efficiency by avoiding the gain competition between the Brillouin and Raman effects. The RSOA is driven with the continuous periodic square-wave voltage and the 1541 nm laser seed is modulated into pulse trains with the repetition-rate of 100 kHz and the pulse-width of 100 ns. Figure 2(b) shows the spectrum of the pulsed laser after the first stage amplifier when the power of 976 nm LD is 179 mW (maximum output power). Note that in the first stage amplifier, the input power of the pulsed laser is very small (less than 0.1 mW) and thus it suffices to use a relatively low power (as compared to that for the second stage amplifier) 976 nm LD to amplify the laser. The average power of the pulsed laser is amplified to 7.19 mW. However, the OSNR of the pulsed laser decreases to 32 dB and obvious ASE in 1530 nm waveband is observed.
Fig. 2. (a) The spectrum of the 1541 nm laser seed (Inset: the corresponding RF spectrum). (b) The spectrum of the 1541 nm pulsed laser after the first stage amplifier.
Figure 3(a) shows the spectrum of the pulsed laser after the second stage amplifier when the power of 976 nm LD is 300 mW (maximum output power). The average power of the pulsed laser is amplified to 57.9 mW. Though the OSNR maintains almost the same level, an additional ASE appears near the 1560 nm waveband. Therefore, a HR-FBG-2 combining a CIR is used to filter out the unwanted ASE for the better amplification in the third stage amplifier. In the Fig. 3(b), the black and the red curves represent the reflection and the transmission spectra of the pulsed laser after it goes through the HR-FBG-2, respectively. Though the average power of the reflected pulsed laser decreases to 41.3 mW, the ASE is significantly reduced in the C-band at the same time.
Fig. 3. (a) The spectrum of the 1541 nm pulsed laser after the second stage amplifier. (b) The spectra of the 1541 nm pulsed laser after going through the HR-FBG-2 (reflection and transmission).
At last, we use a high-power 915 nm LD to forward pump a segment of the DC-EYDF to increase the laser power further. Figure 4(a) shows the final output power of the amplified pulsed laser as a function of the 915 nm pump power. The maximum average power of the pulsed laser can be boosted up to 3.1 W when the 915 nm pump power is 10.8 W, and the slope efficiency is calculated to be about 30.7%. Figure 4(b) shows the spectra of the pulsed laser for different average powers of the 1541 nm laser. The OSNR of the amplified pulsed laser decreases from 40 dB to 24 dB approximately with the growth of the output power. This phenomenon can be explained as below: When the 915 nm pump power increases, it causes stronger population inversion in the DC-EYDF. However, the inverted population in the starting portion of the DC-EYDF cannot be consumed efficiently as the initial input power of 1541 nm laser is not large enough. Therefore, more spontaneous emission will appear and be amplified along with the increment of the residual inverted population. In addition, the 1541 nm laser is modulated with a low repetition-rate square-wave, which gives the ASE an extra chance to become stronger as the 1541 nm laser disappears and reappears periodically in the DC-EYDF. These two main mechanisms result in the deterioration of the OSNR as the output power increases.
Fig. 4. (a) The average output power of the 1541 nm pulsed laser as the launched pump power increases. (b) The spectra of the amplified 1541 nm pulsed laser at different output powers.
For the Raman amplification of the 1653 nm CW laser, the output power of the DFB seed is fixed at 12 mW by adjusting the injected current while the central wavelength is located at 1653.7 nm. When the 1541 nm pump pulse is turned off, the 1653.7 nm laser power at the output end is measured to be only 2.82 mW and the main transmission loss is caused by the splicing loss between the HGDF and the SMF28e fiber. The 1653.7 nm laser will be modulated and amplified into pulse trains synchronously in the HGDF, which is pumped by the high-power 1541 nm pump pulse. Figs. 5(a) and 5(b) show the single-pulse waveforms of the 1541 nm and the 1653.7 nm pulsed lasers which are recorded by an oscilloscope (Tektronix, TBS1104), respectively. Compared to the 1541 nm pump pulse, the 3-dB pulse-width of 1653.7 nm pulsed laser is estimated to be about 31 ns. The inset in Fig. 5(b) shows the oscilloscope trace of the pulse trains. The repetition-rate of the pulse trains is about 100 kHz, corresponding to the period of 10 µs.
Fig. 5. (a) The single-pulse waveform of the 1541 nm pulsed pump after the third stage amplifier. (b) The single-pulse waveform of the 1653.7 nm pulsed laser (Inset: the corresponding pulse trains).
Figure 6(a) shows the output spectra of the 1653.7 nm laser for different average powers of the 1653.7 nm laser (due to different 1541 nm pump powers). Additional optical frequency components around 1653.7 nm arise and broaden gradually as the output power increases. Nevertheless, a 3-dB linewidth of less than 0.08 nm and an OSNR of at least 35 dB are achieved when the average power of 1653.7 nm laser reaches 98.45 mW. Figure 6(b) shows the output power of 1653.7 nm laser as a function of the 1541 nm pump power. The 1653.7 nm laser power increases quickly when the pump power is below 1.5 W but the increment speed becomes slow when the pump power exceeds 1.5 W. This slow-down phenomenon can be explained as below: as the average power of 1541 nm pump increases further after 1.5 W, the OSNR of 1541 nm pump worsens quickly (i.e., more energy will go to the noise) and consequently the pump power really at 1541 nm does not increase linearly any more. The maximum average power of 1653.7 nm laser (black square) is measured to be 98.45 mW and the peak power (black triangle) is estimated to be about 30.85 W. We replace the HGDF with a segment of SMF28e fiber of the same length to make a comparison. Almost no Raman amplification in the SMF28e fiber (red point) is observed as the output power increases very slowly (from 9.53 mW to 9.87 mW). When the 1541 nm pump power is fixed at 3.1 W, we try to tune the central wavelength of 1653 nm DFB seed by changing the injected current and the controlled temperature. Figure 7 shows the tuning spectra every other 0.5 nm from 1652.0 nm to 1654.0 nm, exhibiting the good consistency of OSNR.
Fig. 6. (a) The spectra of the 1653.7 nm pulsed laser at different output powers. (b) The average and peak output power of the 1653.7 nm pulsed laser as the launched 1541 nm pump power increases.
Fig. 7. The output spectra of the Raman pulsed laser at different lasing wavelengths by changing the injected current and the controlled temperature for the DFB Raman seed.
4. Conclusion
In summary, we have demonstrated a high peak-power and narrow-linewidth all-fiber Raman pulsed laser at 1653.7 nm. We have chosen a RSOA which is driven with the continuous periodic square-wave voltage to modulate the 1541 nm laser seed into pulse trains as the Raman pump laser. By combining a homemade high peak-power 1541 nm pump pulse and a segment of the 52-meter-long HGDF, the CW 1653 nm DFB seed is modulated and amplified into pulse trains synchronously. The repetition-rate and the pulse-width of our 1653.7 nm pulsed laser are 100 kHz and 31 ns, respectively. The maximum average output power can be amplified to 98.45 mW and the peak power is estimated to be about 30.85 W. Meanwhile, the 3-dB linewidth is measured to be less than 0.08 nm. The central wavelength of the Raman pulsed laser can be tuned from 1652.0 nm to 1654.0 nm continuously and the OSNR of more than 35 dB is maintained all the time when the 1541 nm pump power is fixed at 3.1 W. The performance of the Raman pulsed laser can be improved further by optimizing the amplification of the 1541 nm pump and minimizing the splicing loss between the HGDF and the SMF28e fiber.
Funding
The work is partially supported by the National Key Research and Development Program of China (No. 2018YFC1407503); National Natural Science Foundation of China (11621101, 91833303); Fundamental Research Funds for the Central Universities.
Disclosures
The authors declare that there are no conflicts of interest related to this article.
References
1. C. Chen, R. W. Newcomb, and Y. Wang, “A trace methane gas sensor using mid-infrared quantum cascaded laser at 7.5 µm,” Appl. Phys. B 113(4), 491–501 (2013). [CrossRef]
2. B. Li, C. Zheng, H. Liu, Q. He, W. Ye, Y. Zhang, J. Pan, and Yi. Wang, “Development and measurement of a near-infrared CH4 detection system using 1.654 µm wavelength-modulated diode laser and open reflective gas sensing probe,” Sens. Actuators, B 225, 188–198 (2016). [CrossRef]
3. R. H. Stolen and E. P. Ippen, “Raman gain in glass optical waveguides,” Appl. Phys. Lett. 22(6), 276–278 (1973). [CrossRef]
4. M. N. Islam, “Raman amplifiers for telecommunications,” IEEE J. Sel. Top. Quantum Electron. 8(3), 548–559 (2002). [CrossRef]
5. Y. Feng, L. R. Taylor, and D. B. Calia, “150 W highly-efficient Raman fiber laser,” Opt. Express 17(26), 23678–23683 (2009). [CrossRef]
6. L. Zhang, C. Liu, H. Jiang, Y. Qi, B. He, J. Zhou, X. Gu, and Y. Feng, “Kilowatt Ytterbium-Raman fiber laser,” Opt. Express 22(15), 18483–18489 (2014). [CrossRef]
7. Q. Xiao, P. Yan, D. Li, J. Sun, X. Wang, Y. Huang, and M. Gong, “Bidirectional pumped high power Raman fiber laser,” Opt. Express 24(6), 6758–6768 (2016). [CrossRef]
8. H. Zhang, H. Xiao, P. Zhou, X. Wang, and X. Xu, “119-W Monolithic Single-Mode 1173-nm Raman Fiber Laser,” IEEE Photonics J. 5(5), 1501706 (2013). [CrossRef]
9. V. Balaswamy, S. Arun, G. Chayran, and V. R. Supradeepa, “All passive architecture for high efficiency cascaded Raman conversion,” Opt. Express 26(3), 3046–3053 (2018). [CrossRef]
10. E. M. Dianov, I. A. Bufetov, V. M. Mashinsky, A. V. Shubin, O. I. Medvedkov, A. E. Rakitin, M. A. Mel’kumov, V. F. Khopin, and A. N. Gur’yanov, “Raman fibre lasers based on heavily GeO2-doped fibres,” IEEE J. Sel. Top. Quantum Electron. 35(5), 435–441 (2005). [CrossRef]
11. E. M. Dianov and V. M. Mashinsky, “Germania-based core optical fibers,” J. Lightwave Technol. 23(11), 3500–3508 (2005). [CrossRef]
