Open Access
Add to BibSonomy
Abstract
We report on the mode field adaption of an active thulium-doped fiber by using the thermally-diffused expanded-core technique. The fiber core diffusion is analyzed by splice transmission measurements and visually from side view images. The obtained heating parameters are used to build a thulium-doped fiber laser emitting at 2036nm that is core-pumped by an erbium:ytterbium fiber laser. By allowing the fiber cores to diffuse, the mode fields of the active and passive fibers are adapted for both the signal and pump wavelength. The adaptation of the mode fields increases the slope efficiency from 66.1% to 75.0%. The obtained slope efficiency is close to the stoke efficiency of 77.0%. By comparing the results with a fiber laser simulation, the slope efficiency of 75.0% is verified to be the maximum slope efficiency taking the active fiber length into account.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Thulium-doped fiber lasers (TDFL) emitting at 2 µm are used for various applications such as LIDAR, optical communication and imaging. Many applications benefit from low SWaP (size, weight and power) fiber laser architectures that can work in harsh environment at different temperatures. All-fiber lasers can address these points by the substitution of free-space optics with alignment-free fiber components to optimize laser efficiency, housing footprint and costs. For achieving reliable laser operation at different temperatures, attention has to be paid to the pump scheme [1]. There are two efficient pump schemes to realize TDFLs by using either the 3H6 → 3H4 transition with diode lasers or the 3H6 → 3F4 in-band transition with fiber lasers. Since the spectral bandwidth of the 3H6 → 3H4 transition is narrow compared to the temperature depended shift of the center wavelength of pump diode lasers, additional measures such as wavelength stabilization or temperature stabilization are required for this pump scheme. Both approaches have disadvantages: Peltier elements for temperature stabilization have limited efficiency and require additional electronics. Moreover, wavelength stabilized VBG diode laser can only be used at limited temperatures [2] between 20°C till 40°C. In contrast, the 3H6 → 3F4 transition has a broad spectral bandwidth and allows pumping with available erbium:ytterbium-codoped fiber lasers (EYDFL) without additional measures. Using EYDFLs as pump source for TDFLs is therefore a suitable approach for low SWaP setups.
Many investigations report on TDFLs emitting around 2 µm that are pumped with EYDFLs. In 1994T. Yamamoto et al. [3] described a TDFL pumped by an EYDFL at 1570 nm. The slope efficiency was 71% at 1900nm (Stokes efficiency is 82.6%). Further, D. Y. Shen et al. [4] reported in 2006 about a TDFL built with free-space optics pumped at 1565 nm. Depending on the fiber length, the slope efficiencies were 72% at 1977nm and 69% at 1991nm (Stoke efficiencies about 79%). Recently, A. V. Kir’yanov and E. M. Sholokhov [5] reported on an all-fiber TDFL that is pumped by an EYDFL and amplifier chain at 1567 nm. They reached an overall efficiency of about 80%. Unfortunately, it is not clear from this paper what the signal wavelength was (1940nm in the graph or 1947nm in the text) and they did not mention the slope efficiency. Therefore, no statement can be deduced on how close the slope efficiency to the Stokes efficiency was.
In this investigation, we optimize the slope efficiency of a core-pumped TDFL emitting at 2036nm by mode field adaption between the active and passive fibers with the thermally expanded core (TEC) technique. The obtained mode-field adaptation increases the slope efficiency from 66.1% to 75.0%, which is close to the Stokes efficiency of 77.0%. By comparison with a fiber laser simulation given in [1], this is the highest obtainable slope efficiency for this particular setup. To the best of our knowledge, this is the highest slope efficiency reported from TDFL above 2 µm that is pumped by an EYDFL. Additionally, to the best of our knowledge, this is the first time experimentally shown that the TEC technique is used with a thulium-doped fiber.
2. Preliminary consideration
There are different methods for the fabrication of mode field adapters. The thermally-diffused expanded-core (TEC) technique [6] is originally used to enlarge the mode field diameter (MFD) of an optical fiber to increase the tolerance to lateral and longitudinal misalignment [6,7]. However, it can also be used to adapt MFDs of two different optical fibers that are fusion spliced together [8]. In the TEC technique, the dissimilar fibers are heated up along the fiber splice to force the diffusion of the dopants inside the optical fiber. Because of the concentration gradient between the doped fiber core and the undoped fiber cladding, the dopants diffuse from the core to the cladding region in lateral direction (diffusion in axial direction is commonly neglected [7]). Since it is assumed that the refractive index change in glass is proportional to the doping concentration [7], the fiber core diameter increases while the numerical aperture decreases. As a consequence, the MFD increases with increasing diffusion of dopants.
Many investigations have been carried out using the TEC technique to adapt the mode field for optimizing the fiber splice transmission. In 2013 X. Zhou et al. [8] reported on mode field adaption between a single-mode fiber (SMF) and large mode area (LMA) fiber by using the TEC technique. They reduced the splice loss of a SMF-LMA-SMF chain from 6.4 dB to 0.4 dB. Furthermore, the TEC technique was also used for the fabrication of pump combiners [9] for signal transmission. In 2018 F. Fanlong et al. [10] reported on a mode field adapter fabricated by the TEC technique. They achieved a transmission efficiency of 92% at 1064 nm with 100 W handling power between a SMF and LMA fiber.
In this investigation the mode fields of the active thulium-doped fiber and the passive fiber are adapted. The core diameters are 4.5 µm and 6 µm and the numerical apertures are 0.25 and 0.21 for the active and passive fiber, respectively. Further data of the active fiber can be found in Ref. [1]. Assuming a step-index profile and the validity of the Marcuse Formula [11], the MFD at the signal wavelength for the active and passive fiber are 6.6 µm and 7.8 µm, respectively. Furthermore, the MFD at the pump wavelength for the active and passive fiber are 5.2 µm and 6.4 µm, respectively. Using the overlap factor between Gaussian beams [11], the splice transmission is 97.2% and 95.7% for the signal and pump wavelength, respectively. Therefore, the core of the active fiber has to expand more compared to the passive fiber to match both mode fields for increasing the splice transmission. However, the dopants have a significant effect on the diffusion rate. The active fiber core contains primarily Al2O3 whereas the passive fiber contains mainly GeO2. Investigations [12] indicated that Al2O3 has a significantly higher diffusion coefficient and stronger nonlinearity (dependency of the diffusion coefficient from dopant concentration) compared to GeO2 in silica. Therefore, the mode fields of both fibers can be adapted because the active fiber core will diffuse and expand significantly faster than the passive fiber core.
Assuming Gaussian mode fields, the splice transmission T between two single-mode step-index fibers can be calculated by
where w1 and w2 are the mode field radii of both fibers, respectively. When a step-index fiber core diffuses, the refractive index profile of the fiber core tends toward a Gaussian refractive index profile. Since the Gaussian beam approximation is still adequate for fibers with Gaussian refractive index profiles [7], we use Eq. (1) for estimating the transmission for both wavelengths (2036nm and 1567 nm) after splicing with the TEC technique. The mode field radius of the active fiber after diffusion can be calculated by [7] where rco,D is the radius of the diffused fiber core. Moreover, the V number remains constant during the heat treatment [7]. Since the active fiber core diffuses significantly faster than the passive fiber core (because of the mentioned diffusion coefficients for Al2O3 and GeO2 at higher temperatures), it can be assumed that only the active fiber core diffuses during the mode field adaption. Taking Eqs. (1) and (2) into account the splice transmission can be optimized for a diffused fiber core diameter of 4.9 µm (initial core diameter is 4.5 µm) with about 99.9% for both signal and pump wavelengths.3. Experiment
For the adaption of the mode fields between the active and passive fiber, a laser source at 2036nm or at 1567 nm is launched into the core of the passive fiber. Then both active and passive fibers are clamped in close proximity to each other and cold spliced together. A cold splice refers here to a fusion splice in which the heating energy is reduced to avoid diffusion effects of the fiber cores but still enable a fiber joint. This can be achieved with all common commercial splicers by a power calibration procedure given from the supplier and subsequent reduction of the delivered heat. After cold splicing, the splice is heated up along the fiber axis. In parallel the transmitted laser signal in the active fiber is measured with a photodiode. After the photodiode current is converted with a transimpedance amplifier, the photodiode voltage is measured over time. Figure 1 shows normalized transmitted power measured over the heating time for the signal wavelength (2036nm) and pump wavelength (1567 nm). The heating time at zero seconds represent the time where the cold splice is finished whereas the time after represent the heating phase to force diffusion of the fiber cores. Measurements for the transmitted power are acquired at a fixed interval of 8 s. The measured power for the signal and pump wavelength after the cold splice is 97.8% and 96.0%, respectively. Afterwards, the transmitted power increases for both wavelengths followed by a decrease. This can be explained by the difference in diffusion rates [12], which results in a different change of MFDs. Therefore, the MFDs of the active and passive fiber become more similar until both equalize. At this point the transmitted power is maximized. After that both MFDs diverge because of the different diffusion rate leading to a decrease of the transmitted power.
Since the normalized transmitted power is a relative measurement, the obtained measurement values are not equal to the splice transmission. However, by comparing the measurement values after the cold splice with the calculated values (difference of 0.6% and 0.3% for the signal and pump wavelength, respectively), it can be assumed that the splice connection is almost loss-free at 27 s for both the signal and pump wavelength.
In addition, Fig. 2 shows side images of both fibers during the heating phase. The active fiber is on the left side while the passive fiber is on the right side. In the first image after the cold splice, a clear transition between the active and passive fiber is visible. Afterwards, the core of the active fiber is expanded whereas the core of the passive fiber is maintained. As already mentioned, this can be explained by the difference in the diffusion coefficients [12]. Since Al2O3 of the active fiber core has a significantly higher diffusion coefficient compared GeO2 of the passive fiber core, the active fiber core will expand considerably faster than the passive fiber core. At a heating time of 24 s, no transition between the active and passive fiber is visible anymore. An expanded passive fiber core can be seen at 48 s. Afterwards, the contrast of both fiber cores reduces as the dopants distribute increasingly in the fiber cladding.
Furthermore, a TDFL [1] is built by using a heat time of 27 s for splicing the active fiber to passive fibers with a fiber Bragg grating pair (RHR = 99% and RLR = 10%). Figure 3 shows schematically the setup that includes an in-house built EYDFL to core-pump a TDFL. The optimized fiber splices are marked with red crosses. The thulium-doped fiber (TDF) has a core diameter of 4.5 µm and a numerical aperture of 0.25 and the passive fiber has a core diameter of 6 µm and a numerical aperture of 0.21.
Figure 4 shows the measured signal power of the TDFL at 2036nm against the pump power at 1567 nm. The graph includes the measurement with and without mode field adaption. The obtained values are compared with the simulation values given in [1]. By considering an active fiber length of 2 m, the maximum obtainable slope efficiency is reached with 75.0% at this fiber length [1]. Therefore, it can be assumed that the mode field adaption enables a nearly lossless connection between the active and passive fiber. Through using the TEC technique for mode field adaption between the active and passive fiber, the slope efficiency is improved from 66.1% to 75.0%. Moreover, the signal power is increased from 1.73 W to 1.98 W.
4. Conclusion
This investigation shows that mode field adaption by using the TEC technique with active and passive fibers can improve the slope efficiency and output power of TDFLs. By using the TEC technique, a TDFL at 2036nm is realized with a slope efficiency of 75% that is close to the Stokes efficiency of 77% considering the pump wavelength at 1567 nm. By comparing the results with a fiber laser simulation, the slope efficiency of 75% is verified to be the maximum slope efficiency taking the active fiber length of 2 m into account.
Funding
German Ministry of Defense.
Acknowledgments
We recognize the support of the mechanical workshop of the IOSB and Artur Schander, who fabricated special opto-mechanical components for the experimental setup.
Disclosures
The authors declare no conflicts of interest.
Data availability
The results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. D. Panitzek, C. Romano, M. Eichhorn, et al., “Temperature investigation of low SWaP thulium-doped fiber lasers,” Opt. Express 32(2), 1890 (2024). [CrossRef]
2. S. R. Karlsen, R. K. Price, M. Reynolds, et al., “100-W 105-µm 0.15NA fiber coupled laser diode module,” in M. S. Zediker, ed. (2009), p. 71980T.
3. T. Yamamoto, Y. Miyajima, and T. Komukai, “1.9 µm Tm-doped silica fibre laser pumped at 1.57 µm,” Electron. Lett. 30(3), 220–221 (1994). [CrossRef]
4. D. Y. Shen, J. K. Sahu, and W. A. Clarkson, “High-power widely tunable Tm:fibre lasers pumped by an Er,Yb co-doped fibre laser at 1.6 µm,” Opt. Express 14(13), 6084 (2006). [CrossRef]
5. A. V. Kir’yanov and E. M. Sholokhov, “35 W highly effective Ytterbium-Erbium-Thulium tandem all-fiber 1.947 µm laser system at 975 nm diode pumping,” Laser Phys. Lett. 20(10), 105102 (2023). [CrossRef]
6. K. Shiraishi, Y. Aizawa, and S. Kawakami, “Beam expanding fiber using thermal diffusion of the dopant,” J. Lightwave Technol. 8(8), 1151–1161 (1990). [CrossRef]
7. G. S. Kliros and N. Tsironikos, “Variational analysis of propagation characteristics in thermally diffused expanded core fibers,” Optik 116(8), 365–374 (2005). [CrossRef]
8. X. Zhou, Z. Chen, H. Zhou, et al., “Mode-field adaptor between large-mode-area fiber and single-mode fiber based on fiber tapering and thermally expanded core technique,” Appl. Opt. 53(22), 5053 (2014). [CrossRef]
9. K. Zhao, Z. Chen, X. Zhou, et al., eds. (2015), p. 953211.
10. F. D. Fanlong Dong, X. Z. Xinhai Zhang, and F. S. Feng Song, “Thermal expanded core technique applied to high power fiber mode field adapter,” Chin. Opt. Lett. 16(3), 030602 (2018). [CrossRef]
11. D. Marcuse, “Loss analysis of single-mode fiber splices,” Bell Syst. Tech. J. 56(5), 703–718 (1977). [CrossRef]
12. J. Kobelke, K. Schuster, J. Bierlich, et al., “Germania and alumina dopant diffusion and viscous flow effects at preparation of doped optical fibers,” AEEE 15(1), 101–106 (2017). [CrossRef]
