Numerical analysis and optimization of 2.8 μm lightly-erbium-doped fluoride fiber laser based on cascaded transition

Abstract

The Er<sup>3+</sup>-doped ZBLAN fiber laser is a promising approach for 2.8 μm mid-infrared (MIR) laser. The long lifetime of the lower-laser-level <sup>4</sup>I<sub>13/2</sub> often results in serious self-terminating effect which harms the laser power and efficiency significantly, especially for the active fiber with low dopant concentration which is preferred for weak thermal issues but can not depopulate the lower-laser-level effectively via the up-conversion process. The 1.6 μm lasing (<sup>4</sup>I<sub>13/2</sub>→<sup>4</sup>I<sub>15/2</sub>) in Er-ZBLAN fiber could deplete the population on <sup>4</sup>I<sub>13/2</sub>. Therefore, cascaded 2.8 μm and 1.6 μm lasing in Er<sup>3+</sup>-doped ZBLAN fiber provides a promising solution to the self-termination effects on laser power scaling. Moreover, the <sup>4</sup>I<sub>13/2</sub>→<sup>4</sup>I<sub>15/2</sub> 1.6 μm laser also has some overlap with the <sup>4</sup>I<sub>13/2</sub>→<sup>4</sup>I<sub>9/2</sub> excited state absorption (ESA) spectrum. The ions on the <sup>4</sup>I<sub>9/2</sub> level would then relax to the upper-laser-level of 2.8 μm lasing (<sup>4</sup>I<sub>11/2</sub>), and results in enhanced laser efficiency. In general, the 1.6 μm cascaded lasing in 2.8 μm Er-ZBLAN fiber laser involves both lasing and ESA. The two processes have different spectra and different influences on the 2.8 μm laser gain. Therefore, there should exist an optimal wavelength of the 1.6 μm laser, which would balance the two processes, ensure the lower-laser-level depopulation while maximizing the ions recycling. For this consideration, we developed a comprehensive numerical model of cascaded 2.8 μm and 1.6 μm lasers based on Er-ZBLAN fiber. After verifying the validity of the numerical model by using the previous experimental results, the influence of MIR and 1.6 μm lasing wavelengths on the power and conversion efficiency of 2.8 μm laser were investigated thoroughly. The results show that a suitable trade-off between the two processes can be reached with the cascaded lasing wavelength of 1610 nm, for the optimized 2.8 μm laser power/efficiency. Moreover, the influence of 1.6 μm laser cavity feedbacks on the power/efficiency behavior of the 2.8 μm laser is investigated. It is found that very low feedback at 1.6 μm, even only 4% as provided by the Fresnel reflection of the fiber facet, could enable the efficient 1.6 μm laser generation and improve the efficiency of 2.8 μm laser significantly.