High-power fiber lasers combine advanced laser technologies to create laser systems with excellent beam quality, high electrical efficiency, and large output powers. They are currently widely used for marking parts, cutting, drilling, welding, and medical surgery, with additional uses currently being developed for weapons systems. If we could further scale up fiber lasers' power, they could be used to manipulate a greater variety of materials, including often hard-to-work glass and ceramics. In military applications, the range of operation could be increased and the engagement time shortened.
Further power scaling is primarily limited by fiber nonlinearities, which occur in the presence of intense light. The light interacts strongly with the medium such that the light is scattered and the refractive index of the material is altered.1 As the transmitted mode (the elemental spatial pattern that light takes when it propagates in a waveguide) travels through the core of the optical fiber, it spreads through a slightly larger volume that includes the inner edge of the fiber cladding that encompasses the mode area.
The fundamental solution to fiber nonlinearities is to increase mode areas in fibers while maintaining single-mode operation. This gives rise to a high-quality Gaussian-like beam that enables tight focussing onto distant targets. However, increasing the physical dimensions of waveguide areas in fibers results in multiple modes being supported. One solution is to design fibers that suppress higher-order modes. When conventional waveguide dimensions are increased, all modes are increasingly guided away from the core/cladding boundary and into the center of the core.2 Coupling a mode out to suppress it is difficult, which severely limits scalability. However, in an all-solid photonic bandgap fiber, modes are guided due to the anti-resonance of cladding photonic crystal lattice. This provides strongly mode-dependent guidance and leads to very high differential losses among modes. The all-solid nature of the fiber also means that it can be easily spliced to other fibers.
We have performed extensive theoretical studies to optimize designs at desired coil diameters for a negligible fundamental-mode loss and large higher-order-mode losses using a finite element method (FEM) mode solver.3 For core diameters of ∼50μm, differential mode losses of over 40dB between fundamental and higher-order modes can be achieved with sufficient tolerance for fabrication. Our optimized design operates in the third bandgap at ∼1050nm. We have demonstrated, for the first time, that all-solid photonic bandgap fibers with core diameters of ∼50μm can be made with > 30dB higher-order-mode suppressions, with strong potential for further mode area scaling.
We fabricated a number of fibers and thoroughly characterized two of them. The nodes are high-index circular structures in the cladding, which are critical for the formation of photonic bandgaps there. For Fiber 1, the nodes were made from a graded index preform with germanium-doped core, with a relative peak refractive index contrast between core and cladding, Δ, of 1.72%. The nodes of Fiber 2 were made from a step-index preform with a germanium-doped core, with peak Δ of 1.52%. Fibers 1 and 2 have core diameters of 55.1μm and 49.1μm, respectively. The effective mode area of Fiber 1 was simulated to be ∼920μm2 at a wavelength of 1050nm and a coil diameter of 50cm. The cross-section of Fiber 2 is shown in Figure 1, alongside the measured loss in loose coil.
Figure 1. (a) Cross-sectional photos of our fabricated ‘Fiber 2’ and (b) the measured loss of loosely coiled Fiber 2.
We measured the bend-dependent loss for both fibers at various coil diameters. For both, we observed very little bend-dependent loss at the center of the third bandgap (∼1.05μm) for coil diameters between 30cm and 50cm. By monitoring the near-field output pattern while the launch beam moved across the front face of the fiber, we were able to qualitatively test the output of 2m of Fiber 1 coiled at 30cm diameters (see Figure 2). No higher order modes were observed during this process, indicating a fairly robust single-mode operation.
Figure 2. Near-field pattern output of 2m of our fabricated ‘Fiber 1’ coiled at 30cm while launch beam is moved slowly across the center of the fiber (sequence moves from left to right).
More quantitative mode analysis was performed with an S2 setup, using an external cavity diode laser (rapidly tunable from 1020nm to 1085nm), a wave meter, and a CCD camera.4 Input and output polarizers were used and aligned to the birefringence axis of the fiber. Weak birefringence existed in the fibers due to a slight asymmetry introduced in their fabrication. S2 measurements were performed on 6m of Fiber 2, coiled at various coil diameters, and efforts were made to ensure that the launch condition was not altered while the coil diameter was varied. The measured mode images and contents are shown in Figure 3. The observed (higher-order) LP11 and LP02 mode contents were below −30dB for coil diameters of less than 50cm. We also measured the polarization extinction ratio of Fiber 2 as 14dB in this nominally non-polarization-maintaining fiber.4
Figure 3. (a) Resolved higher order modes LP01, LP11 and LP02 in Fiber 2 at coil diameters of 30cm and (b) measured relative LP11 and LP02 mode contents at various coiling diameters.
In summary, we have succeeded in designing and fabricating fibers that maintain single-mode output beam quality, despite increased mode area: a development critical for focusing intense light onto distant objects. We are now researching how to use these designs to fabricate ytterbium-doped all-solid photonic bandgap fibers with further mode-area scaling. It was recently observed that, when combined with thermal effects, even a very low power in higher-order modes can initiate mode instability. Fibers with strong higher-order-mode suppression are critical for further power scaling of single mode fiber lasers to beyond kilowatt levels.
This work is supported by the Joint Technology Office Multidisciplinary Research Initiative grant (contract W911NF-10-1-0423).
Liang Dong, Fanting Kong, Devon Mcclane, Guancheng Gu
Liang Dong is an associate professor at Clemson University and has served in senior management roles for several companies. During the 20 years he has spent working in photonics, he has given a large number of invited talks, been granted more than 20 patents, and published more than 200 papers.
1. R. W. Boyd, Nonlinear Optics , Academic Press, 1992.
2. J. Senior, Optical Fiber Communications: Principles and Practice , Prentice Hall, 2008.
3. K. Saitoh, M. Koshiba, Full-vectorial imaginary-distance beam propagation method based on a finite element scheme: application to photonic crystal fibers, IEEE J. Quantum Electron. 38(7), p. 927-933, 2002.
4. L. Dong, K. Saitoh, F. Kong, P. Foy, T. Hawkins, D. McClane, G. Gu, Proc. SPIE 8547, p. 8547-18, 2012.