Open Access
Add to BibSonomy
Abstract
Two long-period fiber gratings (LPFGs) used to separately suppress the stimulated-Raman-scattering (SRS) in the seed and amplifier of kW-level continuous-wave (CW) MOPA fiber laser are developed in this paper. A process that combines constant-low-temperature and dynamic-high-temperature annealing was employed to reduce the thermal slopes of 10/130 µm (diameter of core/cladding fiber) and 14/250 LPFGs, used in the seed and amplifier respectively, from 0.48 °C/W to 0.04 °C/W and from 0.53 °C/W to 0.038 °C/W. We also proposed a reduced-sensitivity packaging method to effectively reduce the influence of axial-stress, bending, and environmental temperature on LPFGs. Further, we established a kW-level CW MOPA system to test SRS suppression performance of the LPFGs. Experimental results demonstrated that the SRS suppression ratios of the 10/130 and 14/250 LPFGs exceed 97.0% and 99.6%, respectively.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Fiber lasers are widely used in many fields including industrial processing [1,2], biomedicine [3,4], and communications [5] because of their compact structure, high efficiency, and high spatial beam quality [6]. With the increasing demand for high power, stimulated Raman scattering (SRS) has become a major limiting factor for the development of fiber lasers [7]. To solve this problem, researchers have recently been focusing on using fiber gratings with special structures to suppress SRS in fiber lasers.
One method is to use a chirped and tilted fiber Bragg grating (CTFBG), which is a short-period fiber grating having an angle between the fiber axis and grating plane. This allows the forward-propagating core modes, which originally transmit only in the core, to be coupled to the backward-propagating core and cladding modes. The light of a specific wavelength can be coupled from the core to the cladding by controlling the period of the CTFBG. In 2017, Wang et al. [8,9] were the first to fabricate a CTFBG with the ability to suppress SRS and apply it to a high-power continuous wave (CW) fiber laser [10,11]. In 2019, Jiao et al. [12] enhanced the power-carrying load of the CTFBG to 1 kW, which further promoted the development of this method. However, the CTFBG is a reflective fiber grating, which can couple some signal and SRS light to the backward-propagating core modes. Although the power of this backward-propagating light is low, for some high stability and high-power fiber lasers, according to Antipov et al. [13], this light (the backward-propagating laser power could be watt-level when the power of the signal laser is over 3 kW) should not be ignored because it may drastically reduce the mode instability power threshold and even stimulate other nonlinear effects such as four wave mixing, further worsening the linewidth and stability of the system. Hence, when the SRS light of the high-power fiber lasers is supposed to be suppressed commendably, a novel device without backward-propagating light should be developed. Moreover, CTFBG can couple a part of light to the radiation modes which will cause the overheating in the coating of CTFBG in the incident direction of the laser [12]. Thus, it is necessary to make a cladding power stripper (CPS) on the grating. This will not only increase the fabrication difficulty of the grating, but also may shorten the lifetime of the grating by damage of the cladding of the fiber induced in the process of making CPS.
Another option to suppress SRS is using long-period fiber grating (LPFG), which is a transmission fiber grating with a submillimeter period. This kind of fiber grating can couple the forward-propagating core modes to the forward-propagating cladding modes. Therefore, the light of a specific wavelength can also be coupled from the core to the cladding by controlling the period of the LPFG. This allows the LPFG to effectively avoid the potential instability to the fiber laser caused by backward light produced by the CTFBG, which must be considered in the utilization of CTFBG to suppress SRS in high-power laser. Further, by controlling the period of LPFG, we can couple the SRS light from the core to the low-order cladding modes, so that this light can continues to transmit along the fiber cladding until stripped by the laser’s own CPS. This means that the LPFG can avoid the requirement to fabricate an additional stripper on the grating surface, which effectively reduces the complexity of the fabrication and prolongs the lifetime of the grating. So far, LPFGs have only been able to suppress SRS in pulsed fiber lasers, which have an average power of only milliwatts [14,15]. Research on SRS suppression in high-power CW fiber lasers with LPFGs is still lacking for two main reasons. First, inscription generates a large number of hydroxyl compounds in the LPFGs; meanwhile, there are still a large number of unreacted hydrogen molecules in LPFGs after the inscription. At high power, these residual components cause significant heating due to their strong infrared absorption, which greatly limits the power-carrying load of the LPFGs. Second, the long period of the LPFGs means that it is very sensitive to variations in the axial stress, bending, and temperature [16–18]. Consequently, the cladding-mode resonances of the LPFGs are easily affected by the external environment and arbitrarily drift, which diminishes the SRS suppression ratio.
In this study, we designed and fabricated two kinds of high power LPFGs to suppress SRS in the seed and amplifier of a kW-level CW MOPA fiber laser, respectively. To address the heating phenomenon of LPFGs at high power, we proposed an annealing process that combines a constant-low-temperature and dynamic-high-temperature to reduce the thermal slope of the LPFGs. A reduced-sensitivity packaging method was also developed to effectively reduce the influences of the axial stress, bending, and temperature on the LPFGs. Thereby, a 100 W-level 10/130 µm (diameter of core/cladding fiber) LPFG and 1 kW-level 14/250 LPFG employed in the seed and amplifier respectively, were developed. Further, we established a kW-level CW MOPA fiber laser to test the SRS suppression ratio of the LPFGs. To the best of our knowledge, it is a novel method of simultaneous suppression of SRS in both the seed and amplifier of a kW-level CW fiber laser.
2. Design and simulation of the high-power LPFGs
In contrast to the fiber Bragg grating (FBG), which has a submicron-level period, LPFGs usually have a submillimeter-level period. Thus, forward-propagating core modes, which originally transmit only in the core, can be coupled to the forward-propagating cladding modes, as shown in Fig. 1(a). According to the mode coupling theory [19,20], the light of a specific wavelength is coupled from the core to the cladding and rapidly attenuates when the forward-propagating core modes and cladding modes of the LPFG can satisfy the phase matching condition:
where ${\beta _{co}}$ is the propagation constant of the core mode, $\beta _{cl}^i$ is the propagation constant of the ith cladding mode, and $\Lambda $ is the period of the LPFG. This results in a series of loss peaks in the transmission spectrum of the LPFG, as shown in Fig. 1(b). It can further be observed clearly that LPFG has no reflection peak, which means that LPFG is a pure transmission grating, thus, it does not reflect signal or Raman light. The propagation constants of the core mode and cladding modes are Then, the resonance wavelength can be expressed as where $n_{eff}^{co}$ is the effective index of the core mode and $n_{eff}^{cl, i}$ is the effective index of the ith cladding mode. Thus, we can filter the SRS from the signal light by controlling the parameters of the LPFG.
Fig. 1. (a) Schematic diagram of the structure of a LPFG and (b) simulated transmission spectrum of a LPFG. The period is 490 µm, the period number is 40, and the index modulation amplitude is 0.00022.
The output spectrum of a 1080 nm fiber laser was simulated, as shown in Fig. 2(a). The central wavelength of the SRS was about 1136 nm and the bandwidth at 20 dB was 30 nm. To estimate the SRS suppression performance, we simulated LPFGs with different parameters using the LMA-GDF-10/130 and LMA-GDF-14/250 (a few-mode fiber used in kW-level fiber lasers with the high power carrying-load as well as the good output beam quality) double-clad fibers (Nufern Inc.), respectively. For the 10/130 LPFG, we simulated the resonance when the LP01 and LP09 modes were coupled. And the resonance of 14/250 LPFG when the LP01 and LP015 modes coupled was also simulated simultaneously. Figures 2(b) and 2(c) show the simulated transmission spectra of the two kinds of LPFGs with different periods. The results indicated that an increasing period will cause the cladding resonances wavelength move towards longer wavelengths. The SRS in the 1080 nm laser can be completely spanned by cladding resonance when the period of the 10/130 LPFG is 490 µm and the period of the 14/250 LPFG is 485 µm.
Fig. 2. Simulation results: (a) SRS spectrum of the 1080 nm fiber laser and transmission spectra of (b) 10/130 and (c) 14/250 LPFGs with different periods.
Figures 3(a) and 3(b) show the simulated transmission spectra of the 10/130 and 14/250 LPFGs with different period numbers (i.e., the length of the LPFG $L = N \times \Lambda $, where N is the number of periods). Figures 3(c) and 3(d) show the influence of the period numbers on the extinction and full width at half maximum (FWHM) of the 10/130 and 14/250 LPFGs. The simulation results indicate that the increasing numbers of period can clearly increase the extinction and decrease the FWHM of the LPFGs before over-coupling.
Fig. 3. Simulation results: transmission spectra of (a) 10/130 LPFGs with different period numbers at a period of 490 µm and index modulation amplitude of 0.00022, (b) 14/250 LPFGs with different period numbers at a period of 485 µm and index modulation amplitude of 0.00042, and influence of the period number on the extinction and FWHM of the (c) 10/130 and (d) 14/250 LPFGs.
Figures 4(a) and 4(b) present the simulated transmission spectra with different index modulation amplitudes of the 10/130 and 14/250 LPFGs. Figures 4(c) and 4(d) show the influence of the index modulation amplitude on the extinction and FWHM of the two LPFGs. The extinction increased and FWHM decreased as the index modulation amplitude increased before over-coupling. The value of the index modulation amplitude depends on the fiber’s photosensitivity and exposure time of the UV light during lithography. Thus, based on the above simulation results, we decided that the lithographic parameters of the 10/130 LPFG should be a period of 490 µm and 40 periods and those of the 14/250 LPFG should be a period of 485 µm and 40 periods, so that the fabricated LPFGs can effectively suppress the SRS stimulated by 1080 nm fiber laser.
Fig. 4. Simulation results: transmission spectra of (a) 10/130 LPFGs with different index modulation amplitudes at a period of 490 µm and 40 periods, (b) 14/250 LPFGs with different index modulation amplitudes at a period of 485 µm and 40 periods, and influence of the index modulation amplitudes on the extinction and FWHM of the (c) 10/130 and (d) 14/250 LPFGs.
3. Inscription of the high-power LPFGs
The point-by-point scanning method was used to write the LPFGs, and the inscription system is shown in Fig. 5. An argon ion laser (Innova 90C FreD Ion Laser, Coherent Inc.) was used as the inscription light source and produced a CW ultraviolet laser with a wavelength of 244 nm and output power of 100 mW. The laser spot was compressed by dual UV cylindrical lens (${f_{Lens1}} = {f_{Lens2}} = 250mm$). The axial meridian of lens 1 was parallel to the x-axis to compress the laser spot longitudinally, which drastically increased the UV laser power density at the fiber core. Although the size of the laser spot was greatly reduced in the longitudinal direction, the size in the transverse direction was still large at about 0.9 mm, which caused a periodic overlap during the inscription process and degraded the LPFGs performance. To solve this problem, we added another cylindrical lens 50 mm behind lens 1. The axial meridian of lens 2 was parallel to the y-axis to compress the laser spot in the transverse direction. The size of the laser spot in the transverse direction was about 200 µm; this was far less than the periods of the LPFGs at 490 µm (10/130) and 485 µm (14/250), so the periodic overlap was effectively avoided. Fibers (LMA-GDF-10/130-M and LMA-GDF-14/250-HP-M, Nufern Inc.), which had previously been processed by high-pressure (13 MPa) and low-temperature (40 °C) hydrogenation for 18 days, were immobilized on the six-dimensional translation stage and placed at the focus of lens 1 to write the LPFGs.
To monitor the spectral parameters of the LPFGs, an online measurement system was introduced. The light emitted by the amplified spontaneous emission (ASE) source (VASS-1060-B-13-GF, Connect Fiber Optics Corporation) was coupled to the LPFGs through the circulator and mode field adapter (MFA) and then coupled to the optical spectrum analyzer (OSA; AQ6730D, Yokogawa Corporation) by another MFA so that the spectral curve could be measured in real-time. This was used to calculate the extinction, central wavelength, and bandwidth of the LPFGs.
Our initial lithographic parameters for the 10/130 LPFG were a period of 490 µm, 40 periods, and an exposure time of 45 s for each period; the lithographic parameters for the 14/250 LPFG were a period of 485 µm, 40 periods, and an exposure time of 120 s for each period. The transmission spectra of the LPFGs after heat treatment are shown as the solid blue line in Figs. 6(a) and 6(b). Strong side-lobes were observed at the edge of the cladding resonance of the LPFGs; these were caused by the sudden change in the refractive index of the fiber at both ends of the LPFGs as it were written [21,22]. These side-lobes will increase the insertion loss of the signal light.
To solve this problem, we controlled the exposure time for each period to achieve Gaussian apodization of the LPFGs. The Gaussian function is expressed by the following equation:
Fig. 7. Simulation results of the transmission spectra of the (a) 10/130 and (b) 14/250 LPFGs before and after the apodization.
The exposure times of each period of the 10/130 and 14/250 LPFGs are shown in Figs. 8(a) and 8(c). In order to ensure the suppression ratio of the LPFGs, we increased the period numbers from 40 to 55 for the 10/130 LPFG and from 40 to 50 for the 14/250 LPFG. The exposure times t were also increased from 45 s to 60 s for the 10/130 LPFG and from 120 s to 150 s for the 14/250 LPFG. The spectra of the 10/130 and 14/250 LPFGs after apodization are shown as solid red lines in Figs. 8(b) and 8(d). The side-lobes on both sides of the cladding resonances were greatly reduced from 1 dB to 0.2 dB for the 10/130 LPFG and from 3 dB to 0.3 dB for the 14/250 LPFG, effectively decrease the anticipated insertion loss of the LPFGs.
Fig. 8. Schematic diagrams of Gaussian apodization of the (a) 10/130 and (c) 14/250 LPFGs; and the transmission spectra of the (b) 10/130 and (d) 14/250 LPFGs after apodization.
4. Heat treatment of high-power LPFGs
LPFGs are currently not being applied to suppressing SRS in high-power CW fiber lasers because of tremendous heats on it induced by high energy CW laser. The generated heat can be attributed to the enormous amounts of Ge-OH and Si-OH inside the LPFG produced by photo-induced reactions in the hydrogenated fiber during lithography [23]. Additionally, after lithography, tremendous amounts of unreacted hydrogen molecules remain inside the fiber. Consequently, as the high-power laser is transmitted though the grating, these residues generate abundant heat due to the strong infrared absorption, which limits the carried power of the LPFGs. Figures 9(a) and 9(b) show thermal images (acquired by a Fluke Ti400 Infrared Imager) of the 10/130 and 14/250 LPFGs without any annealing. When the signal power rose to 25 W, the temperature of the 10/130 LPFG reached 34 °C, and its corresponding thermal slope was 0.48 °C/W at a room temperature of 22 °C. Similarly, for the 14/250 LPFG, the grating temperature reached 47.5 °C when the signal power rose to 48 W and the corresponding thermal slope was 0.53 °C/W.
Hence, we proposed two different annealing methods to eliminate dissociative hydrogen molecules and hydroxyl to improve the carried power of the LPFGs. First, the LPFGs were annealed at a constant and relatively low temperature. The gratings were placed at 60 °C for 30 days to remove residual dissociative hydrogen molecules. Then, the gratings were annealed at dynamic high temperatures. As shown in Figs. 10(a) and 10(c), the gratings were slowly heated to 300°C at intervals of 50°C from room temperature, and remained at 300°C for several minutes before being gradually cooled to room temperature. The high-temperature annealing was to reduce the hydroxyl concentration in the grating, for which we give our conjecture in a previous paper [12].
Fig. 10. Temperature variation for high-temperature annealing of the (a) 10/130 LPFG and (c) 14/250 LPFG, and thermal images of the (b) 10/130 LPFG and (d) 14/250 LPFG after annealing.
High temperature breaks up the hydroxyl to form water molecules which will be accelerated to diffuse away from the gratings by the thermal energy. The purpose of the stepwise heating and cooling was to prevent residual thermal stress inside the gratings from the drastic temperature changes. Moreover, due to the longer exposure time of the 14/250 LPFG, the concentration of hydroxyl compounds is much higher than that of the 10/130 LPFG. Therefore, we doubled the high-temperature annealing time of the 14/250 LPFG, as Fig. 10(c) shows, to reduce the thermal slope. Figures 10(b) and 10(d) show thermal images of the 10/130 and 14/250 LPFGs after annealing. When the signal power was increased to 99.8 W, the temperature of the 10/130 LPFG was 26 °C, and the thermal slope was improved to 0.04 °C/W. And the temperature of 14/250 LPFG was 57 °C when the signal power reached 915 W which means the thermal slope was only 0.038 °C/W.
Figures 11(a) and 11(b) show the spectra of the 10/130 and 14/250 LPFGs at room temperature. The extinction and FWHM of 10/130 LPFG was 26 dB (99.7%) and 34 nm, and the corresponding values of the 14/250 LPFG were 26.5 dB (99.8%) and 38 nm. It can be observed that the SRS spectra of the 1080 nm fiber laser is completely spanned by the spectra of the LPFGs. However, it should be considered that the heating phenomenon generated by high-power laser may cause the drift of the cladding resonances of LPFGs, which would lead to a decrease of the SRS suppression ratio. Therefore, we heated the 10/130 LPFG to 30 °C and the 14/250 LPFG to 60 °C to simulate the heating phenomenon produced by high-power laser. The results are shown in Figs. 11(c) and 11(d). The central wavelength of the 10/130 LPFG shows almost no change and the extinction at 1136 nm remains at 26 dB. Slight differently, for the 14/250 LPFG, the central wavelength shifted about 1 nm and the extinction at 1136 nm marginally reduced to 25 dB. However, this has almost no influence on the SRS suppression performance because the SRS wavelength is still completely spanned. This is also demonstrated by our subsequent experiments. The experimental results clearly indicate that the proposed heat treatment method optimized the thermal slopes of the LPFGs and confirmed the feasibility of applying the LPFGs to suppress SRS in the seed and amplifier of kW-level MOPA fiber laser.
Fig. 11. Spectra of the (a) 10/130 and (b) 14/250 LPFGs after annealing operated at room temperature (22 °C), (c) 10/130 LPFG operated at 30 °C and (d) 14/250 LPFG operated at 60 °C.
5. Reduced-sensitivity packaging method of LPFGs
Another crucial issue with using LPFG to suppress SRS in high-power CW fiber lasers is that its period is easily changed due to its high sensitivity to the environment. Consequently, the SRS suppression ratio of an LPFG is dramatically decreased by unstable cladding-mode resonances induced by random variations in the period.
First, we analyzed the influence of the axial stress on the central wavelength of the LPFGs. One end of the LPFG was fixed, and the other end was connected a spring tension meter that imposed axial tension on the LPFG. The results are shown in Fig. 12(a). The central wavelengths of the LPFGs clearly moved towards the long wavelength direction as the axial stress increased. When the axial stress reached 6 N the wavelength of the 10/130 LPFG drifted approximately 2.3 nm and that of the 14/250 LPFG 1.2 nm. Second, we measured the influence of bending on LPFGs. We made several metallic arcuate belts with different curvatures and fixed LPFGs to the belts to match their curvatures. Then, we measured the changes in wavelength with increasing curvature, as shown in Fig. 12(b). As the curvature increased, the central wavelengths of the LPFGs drifted rapidly in the short wavelength direction. When the bending curvature reached 3.3 m−1, the central wavelength of the 10/130 LPFG drifted approximately 57 nm. Similarly, the central wavelength of the 14/250 LPFG drifted approximately 58 nm at a bending curvature of 2.0 m−1. Finally, we measured the influence of the environmental temperature by placing LPFGs in electric heating incubator with different temperatures. Figure 12(c) shows that the central wavelengths of the LPFGs drifted in the long wavelength direction with increasing temperature. When the temperature reached 80 °C, the central wavelength of the 10/130 LPFG drifted approximately 1.2 nm and the central wavelength of 14/250 LPFG drifted approximately 2.5 nm.
Fig. 12. Curves of the central wavelength of LPFGs with varying (a) axial stress, (b) bending curvature, and (c) environmental temperature.
The above experimental results indicate that the axial stress, bending, and environmental temperature can induce a drift in the grating’s wavelength, which destabilizes the cladding-mode resonance. Thus, we should eliminate the influence of these factors as much as possible to improve the SRS suppression ratio of the LPFGs. We developed a novel packaging structure that can effectively decrease the environmental sensitivity of the LPFGs, as shown in Fig. 13. The structure has two main merits. First, both ends of the LPFGs are fixed by epoxy resin, which not only reduces the impact of axial stress on the LPFGs but also effectively prevents bending. Second, cold water (20 °C) is passed outside the glass tube through the structure to take away the heat generated by the LPFGs and prevent the influence of the external environmental temperature. The glass tube (wall thickness is 1 mm) is mainly to prevent the LPFGs from being contaminated by dust and water.
We used COMSOL to analyze the effectiveness of this structure. Initially, we analyzed the deformation of the LPFGs before and after packaging under an imposed axial tensile stress. The stress was loaded as indicated by the yellow arrow shown in Fig. 13. One end of the LPFG was fixed and a stress was imposed on the other end. Due to the elasticity of fiber, it will deform and elongate under an axial stress if not packaged. We selected a section line along the x-direction indicated by the white arrow shown in Fig. 13, and calculated the axial deformation of the fiber caused by the applied stress. The result is shown by the blue solid line in Fig. 14. When a 6 N stress was composed on the LPFG, the fiber deformed and extended. It can be observed that the grating area (light orange area in Fig. 14) extended about 20 µm leading to the drift of the resonance and diminishing of the suppression ratio of the LPFG. Therefore, we used epoxy resin to fix both ends of the LPFG, and the epoxied regions are expressed by the light green area in Fig. 14. Then we applied 6 N stress to LPFG again, and the deformation result is shown by the solid red line in Fig. 14. Although the deformation of the fiber beyond the epoxied regions is essentially unchanged, in the grating area between these regions the displacement caused by the stress is only 0.3 µm, which indicates that the package method can effectively reduce the sensitivity of the LPFG to axial stress.
Fig. 14. Simulation results: deformation of the LPFG before and after packaging under an axial stress of 6 N. The light orange area represents the grating and the light green area the epoxied regions.
Then we selected another line section along the y-direction as indicated by the black arrow shown in Fig. 13, to calculate the thermal distribution of the structure. We heated the air outside the package to 80 °C as the red arrow shown in Fig. 13 to determine how efficiently the package thermally isolates the fiber from the ambient temperature. The solid blue line in Fig. 15(a) shows the temperature distribution (along the y-axis in Fig. 13) of the whole structure before packaging structure working. When the whole structure reached the thermal equilibria, the temperature of the LPFG (indicated by the light red area), the air around the LPFG (indicated by the white area), the glass tube (indicated by the light purple area), the cooling water (not working in this simulation and indicated by the light yellow area) and the packaging housing (indicated by the gray area) were the same as that of the external environment (80 °C). However, when the packaging structure are working continuously, the temperature distribution is totally different as the solid red line shown in Fig. 15(a). It is clear that the temperature of the LPFG area can consistently stay at about 22 °C when the external environment temperature of the whole device is heated to 80 °C. The simulation results indicate that the flowing cooling water can effectively dissipate the heat transferred from the environment. Furthermore, we heated the LPFG to 60 °C (simulating the heat phenomenon caused by the kW-level CW laser) with the external environment heated to 80 °C simultaneously, to simulate temperature distribution of the severe working conditions and explore the heat dissipation effect of the structure on the LPFG, as the Fig. 15(b) shown. It can be observed that when the package structure is not working, after the thermal balance is reached, the grating temperature increases from 60 °C to 80 °C and stay at 80 °C which is same as the ambient temperature, as the solid blue line shown in Fig. 15(b). However, when the structure is working, the temperature distribution of the structure has been greatly improved, as the solid red line shown in Fig. 15(b). The flowing cold water not only insulates the ambient temperature, but also effectively dissipates the heat generated by the grating itself to prevent the tremendous heat accumulation which can easily destroy LPFG. Moreover, it also demonstrates that the glass tube could not adversely affect the heat dissipation of the grating.
Fig. 15. Simulation results: the radial temperature distribution inside the structure before and after its working when the environment temperature is 80 °C and (a) the grating is not heated, (b) the grating is heated to 60 °C. The light red area represents the LPFG, white areas represent the air, light purple areas represent the glass tube, light yellow areas represent the cooling water and the light gray areas represent the packaging housing.
Figure 16(a) shows the changes in the central wavelength curves of the LPFGs with increasing axial stress after packaged. When the axial stress was 6 N, the central wavelength of the 10/130 LPFG redshifted approximately 0.07 nm and that of the 14/250 LPFG drifted approximately 0.04 nm. These drifts are much less than those before packaged shown in the Fig. 12. Further, these results indicate that the proposed packaging method can effectively reduce the influence of axial stress on the LPFGs. Figure 16(b) shows the curves of the central wavelength with increasing environmental temperature after packaging when the gratings are not heated. The central wavelength of the 10/130 LPFG redshifted approximately 0.025 nm when the environmental temperature was 80 °C, and the 14/250 LPFG central wavelength redshifted approximately 0.04 nm. These results clearly demonstrate that the packaging method can effectively mitigate the influence of environment temperature. Figure 16(c) shows the curves corresponding to those of Fig. 16(b) after packaging when the environmental temperature is increased, in which the 10/130 LPFG was heated to 26 °C and the 14/250 LPFG was heated to 60 °C. It can be observed that when the 10/130 LPFG was heated to 26 °C, the central wavelength was essentially unchanged, and it further slightly redshifted by approximately 0.025 nm when the ambient temperature was increased to 80 °C. Moreover, when the 14/250 LPFG was heated to 60 °C, the central wavelength drifted approximately 1.1 nm (from 1136.41 nm as shown in Fig. 16(b) to 1137.53 nm as shown in Fig. 16(c)). Similarly, the central wavelength of the 14/250 LPFG was further drifted approximately 0.04 nm when the ambient temperature was increased to 80 °C. This little change of the central wavelength has almost no adversely influence on the SRS suppression performance of the 14/250 LPFG, which is also demonstrated in the subsequent experiments. The above results once again demonstrate that the packaging method can effectively reduce the impact of ambient temperature on LPFGs and the structure can effectively dissipate the heat generated by the gratings themselves to stop the hazardous heat accumulation.
Fig. 16. Curves of the change in central wavelengths for the LPFGs after packaged with (a) increasing axial stress, and increasing environmental temperature when (b) the LPFGs are not heated, (c) the 10/130 LPFG is heated to 26 °C and 14/250 LPFG is heated to 60 °C.
6. SRS suppression experiments
After solving the two key issues, we tested the SRS suppression ratio of our LPFGs. We first established a 100W CW oscillator as the evaluation system of 10/130 LPFG as shown in Fig. 17. The system had a single-end-pump structure and 976 nm pump source (pump power of 150 W) to realize a 1080 nm laser output (output power of 99.8 W). The linear laser cavity consisted of a pair of FBGs with a central reflective wavelength of 1080 nm and gain fiber with core and cladding diameters of 10 and 130 µm, respectively (LAM-YDF-10/130-M, Nufern Inc.). The reflectivity of the highly reflective (HR) FBG was greater than 99.5%, and that of the output coupling (OC) FBG was about 10%. The length of the gain fiber was about 12 m. The LPFG was connected to the OC and cladding power stripper (CPS), and the temperature of the LPFG was monitored in real time by an infrared thermal imager. Most of the laser output was reflected by an infrared mirror (reflectivity of >99.8%) and then transmitted to the power meter. The others are spectroscopically measured to characterize the SRS suppression ratio of the LPFG. The position of the detector was fixed before and after the LPFG was inserted into the system.
Before we tested the SRS suppression ratio of the LPFG, we measured the insertion loss, which was approximately 0.23 dB (5.2%). Figure 18(a) shows the output spectra of the evaluation system without the LPFG. When the laser output power reached 99.8 W (pump power 150 W), the difference between the spectral peaks of the SRS and signal laser was about 15 dB. Figure 18(b) shows the output spectra of the evaluation system with the LPFG. The difference between the spectral peaks of the SRS and signal laser changed to about 30 dB at a laser output power of 88.5 W (pump power 150 W). The SRS spectral signal was almost invisible. The yellow solid curve in Fig. 18(c) shows the difference between the original output spectrum of the system (pump power of 150 W) and that with the LPFG, which indicates the SRS suppression performance of the LPFG. The yellow curve was fitted to remove the effects of noise (solid blue curve in Fig. 18(c)), which indicated that the SRS suppression ratio was over 15 dB (97.0%). The reason that the SRS suppression ratio of the 10/130 LPFG (15 dB) developed in this paper is lower than that of the CTFBG (23 dB) developed in our previous paper [12] is that the output power of the oscillator used in this paper is much lower, so the intensity of the SRS is only about 15 dB. Furthermore, the SRS intensity is essentially zero after inserting the LPFG, which suggests that the suppression ratio of the 10/130 LPFG developed in this paper actually exceeds 15 dB.
Fig. 18. Output spectra of the oscillator (a) without and (b) with the 10/130 LPFG; (c) difference between the output spectra of the system with and without the 10/130 LPFG.
Additionally, we built a MOPA system to evaluate the suppression ratio of 14/250 LPFG, as shown in Fig. 19. The 10/130 oscillator (shown in Fig. 17) with the 10/130 LPFG was used as the seed laser, and its SRS light was suppressed perfectly. The amplifier was constituted by a bidirectional pump structure (pump power 1200 W). The gain fiber of the amplifier is LMA-YDF-14/250-HP, which was chosen to be 20 m for adequate total pump absorption. In order to significantly evaluate the SRS suppression performance of the 14/250 LPFG, a 20 m LMA-GDF-14/250-HP-M fiber was spliced between the combiner and CPS to lower the Raman threshold. The amplified laser was output through a CPS and QBH. We inserted the 14/250 LPFG between the CPS and additional passive fiber, and used the same method as before to measure the suppression ratio.
The insertion loss of the 14/250 LPFG is approximately 0.32 dB (7.1%). Figure 20(a) shows the output spectra of the MOPA system without the 14/250 LPFG. When the laser output power reached 910 W (pump power 1200 W), the difference between the spectral peaks of the SRS and signal laser was approximately 17 dB. When the 14/250 LPFG was inserted into the system, the difference between the spectral peaks of the SRS and signal laser increased to approximately 41 dB at a laser output power of 805 W (pump power 1200 W) as Fig. 20(b) shows. Thus, the SRS suppression ratio of the 14/250 LPFG is 24 dB (99.6%) which can be concluded from Fig. 20(c).
Fig. 20. Output spectra of the MOPA system (a) without and (b) with the 14/250 LPFG; (c) difference between the output spectra of the system with and without the 14/250 LPFG.
7. Conclusion
In this study, we developed two high power LPFGs to suppress SRS in the seed and amplifier of a kW-level MOPA fiber laser, respectively. To solve the crucial issue of tremendous heat generation in the LPFGs, we developed a novel heat treatment method that combines annealing at a constant-low-temperature and dynamic-high-temperature to reduce the thermal slopes of the LPFGs. The thermal slope of the 10/130 LPFG used in the seed was reduced from 0.48 °C/W to 0.04 °C/W, and that of the 14/250 LPFG used in the amplifier was reduced from 0.53 °C/W to 0.038 °C/W which enabled the LPFGs to deliver a kilowatt level high-power laser. Another crucial issue is the high environmental sensitivity of the LPFGs, which causes cladding-mode resonances instability and drastically diminishes the SRS suppression ratio. To address this issue, we developed a reduced-sensitivity packaging method that effectively reduces the influence of the axial stress, bending, and environmental temperature on the LPFGs. In addition, we built a kW-level MOPA system to evaluate the LPFGs, the experimental results of which demonstrated SRS suppression ratios of the 10/130 and 14/250 LPFGs of over 15 dB (97.0%) and 24 dB (99.6%) measured separately. These results convincingly indicate that the proposed method will dramatically improve the output spectra of high-power fiber lasers. Further, to suppress SRS in higher power fiber lasers, larger mode area LPFGs should be developed in future.
Funding
Jiangsu Provincial Key Research and Development Program (BE2019114); Pre-research Foundation of Equipment Development Department (61404140105).
Disclosures
The authors declare no conflicts of interest.
References
1. I. Mingareev, F. Weirauch, A. Olowinsky, L. Shah, P. Kadwani, and M. Richardson, “Welding of polymers using a 2 µm thulium fiber laser,” Opt. Laser Technol. 44(7), 2095–2099 (2012). [CrossRef]
2. P. Kah, J. Lu, J. Martikainen, and R. Suoranta, “Remote laser welding with high power fiber laser,” Engineering 05(09), 700–706 (2013). [CrossRef]
3. N. M. Fried and K. E. Murray, “High-power thulium fiber laser ablation of urinary tissues at 1.94 µm,” J. Endourol. 19(1), 25–31 (2005). [CrossRef]
4. K. D. Polder and S. Bruce, “Treatment of melasma using a novel 1927-nm fractional thulium fiber laser: a pilot study,” Dermatol. Surg. 38(2 Part 1), 199–206 (2012). [CrossRef]
5. Z. Li, A. M. Heidt, N. Simakov, Y. Jung, J. M. O. Daniel, S. U. Alam, and D. J. Richardson, “Diode-pumped wideband thulium-doped fiber amplifiers for optical communications in the 1800–2050nm window,” Opt. Express 21(22), 26450–26455 (2013). [CrossRef]
6. W. Shi, Q. Fang, X. S. Zhu, R. A. Norwood, and N. Peyghambarian, “Fiber laser and their applications [Invited],” Appl. Opt. 53(28), 6554–6568 (2014). [CrossRef]
7. W. Liu, P. F. Ma, H. B. Lv, J. M. Xu, P. Zhou, and Z. F. Jiang, “General analysis of SRS-limited high-power fiber lasers and design strategy,” Opt. Express 24(23), 26715–26721 (2016). [CrossRef]
8. M. Wang, Y. J. Zhang, Z. F. Wang, J. J. Sun, J. Q. Cao, J. Y. Leng, X. J. Gu, and X. J. Xu, “Fabrication of chirped and tilted fiber Bragg gratings and suppression of stimulated Raman scattering in fiber amplifiers,” Opt. Express 25(2), 1529–1534 (2017). [CrossRef]
9. M. Wang, Z. X. Li, L. Liu, Z. F. Wang, X. J. Gu, and X. J. Xu, “Fabrication of chirped and tilted fiber Bragg gratings on large-mode-area doubled-cladding fibers by phase-mask technique,” Appl. Opt. 57(16), 4376–4380 (2018). [CrossRef]
10. M. Wang, Z. F. Wang, L. Liu, Q. H. Hu, H. Xiao, and X. J. Xu, “Effective suppression of stimulated Raman scattering in half 10 kW tandem pumping fiber lasers using chirped and tilted fiber Bragg gratings,” Photonics Res. 7(2), 167–171 (2019). [CrossRef]
11. M. Wang, L. Liu, Z. F. Wang, X. M. Xi, and X. J. Xu, “Mitigation of stimulated Raman scattering in kilowatt-level diode-pumped fiber amplifiers with chirped and tilted fiber Bragg gratings,” High Power Laser Sci. Eng. 7, e18 (2019). [CrossRef]
12. K. R. Jiao, J. Shu, H. Shen, Z. W. Guan, F. Y. Yang, and R. H. Zhu, “Fabrication of kW-level chirped and tilted fiber Bragg gratings and filtering of stimulated Raman scattering in high-power CW oscillators,” High Power Laser Sci. Eng. 7, e31 (2019). [CrossRef]
13. O. Antipov, M. Kuznetsov, D. Alekseev, and V. Tyrtyshnyy, “Influence of a backward reflection on low-threshold mode instability in Yb3+-doped few-mode fiber amplifiers,” Opt. Express 24(13), 14871–14879 (2016). [CrossRef]
14. F. Jansen, D. Nodop, C. Jauregui, J. Limpert, and A. Tünnermann, “Modeling the inhibition of stimulated Raman scattering in passive and active fibers by lumped spectral filters in high power fiber laser systems,” Opt. Express 17(18), 16255–16265 (2009). [CrossRef]
15. D. Nodop, C. Jauregui, F. Jansen, J. Limpert, and A. Tünnermann, “Suppression of stimulated Raman scattering employing long period gratings in double-clad fiber amplifiers,” Opt. Lett. 35(17), 2982–2984 (2010). [CrossRef]
16. D. D. Yang, P. Q. Zhang, J. H. Zeng, X. S. Wang, B. A. Song, and S. X. Dai, “SRI-immune highly sensitive temperature sensor of long-period fiber gratings in Ge-Sb-Se chalcogenide fibers,” J. Lightwave Technol. 35(18), 3974–3979 (2017). [CrossRef]
17. D. Barrera, J. Madrigal, and S. Sales, “Long period gratings in multicore optical fibers for directional curvature sensor implementation,” J. Lightwave Technol. 36(4), 1063–1068 (2018). [CrossRef]
18. Y. P. Wang, L. M. Xiao, D. N. Wang, and W. Jin, “Highly sensitive long-period fiber grating strain sensor with low temperature sensitivity,” Opt. Lett. 31(23), 3414–3416 (2006). [CrossRef]
19. T. Erdogan, “Fiber grating spectra,” J. Lightwave Technol. 15(8), 1277–1294 (1997). [CrossRef]
20. T. Erdogan, “Cladding-mode resonances in short- and long-period fiber grating filters,” J. Opt. Soc. Am. A 14(8), 1760–1773 (1997). [CrossRef]
21. X. F. Yang, X. Guo, C. Lu, and C. T. Hiang, “Apodized long-period grating with low insertion loss,” Microw. Opt. Techn. Let. 35(4), 283–286 (2002). [CrossRef]
22. F. Y. M. Chan and K. S. Chiang, “Analysis of apodized phase-shifted long-period fiber gratings,” Opt. Commun. 244(1-6), 233–243 (2005). [CrossRef]
23. L. B. Fu, G. Tan, W. J. Xu, H. L. An, X. M. Cui, X. Z. Lin, and H. D. Liu, “Ultraviolet-initiated reactions of H2 with germanosilicate fibers and H2 concentration dependence of the Bragg wavelength of a fiber grating,” Opt. Lett. 25(8), 527–529 (2000). [CrossRef]
