Open Access
Add to BibSonomyAbstract
An optimized beam transformation system (BTS) is proposed to improve the beam quality of laser-diode bars. Through this optimized design, the deterioration of beam quality after the BTS can be significantly reduced. Both the simulation and experimental results demonstrate that the optimized system enables the beam quality of a mini-bar (9 emitters) approximately equal to 5.0 mm × 3.6 mrad in the fast-axis and slow-axis. After beam shaping by the optimized BTS, the laser-diode beam can be coupled into a 100 μm core, 0.15 numerical aperture (NA) fiber with an output power of over 100 W and an electric-optical efficiency of 46.8%.
© 2016 Optical Society of America
1. Introduction
New solid-state laser devices, particularly fiber laser systems, require increasingly high brightness, high efficiency, and single wavelength pump sources. The fiber-coupled laser-diode module is one of the best pump source choices for new solid-state laser devices because of its nearly unlimited beam guiding flexibility, symmetrical energy distribution and high pointing stability [1–9]. Meanwhile, the fiber-coupled laser-diode module can also be widely used in materials processing and biomedical applications [3–6]. In recent years, the output power of fiber-coupled modules has increased remarkably; e.g., nLight has commercialized 100 W of output power from a 105 μm core, 0.15 NA fiber [3]; Directphotonics has realized 150 W of output power from a 105 μm core, 0.15 NA fiber [4]; and Jenoptik has commercialized 85 W of output power from a 105 μm core, 0.15 NA fiber [5]. These fiber-coupled modules all use single emitters as the light source. However, the high output power fiber-coupled system (100 μm core, 0.15 NA) needs more than 16 single emitters and a large number of optical elements; consequently, the size and cost of the fiber-coupled modules is considerable.
In the fiber-coupled system, a mini-bar was selected as the light source (typical values of the mini-bar are shown in Table 1), which can bring several benefits. One major benefit is that the beam shaping process needs only a few optical elements, which results in a simple, high-efficiency, and high-brightness optical system. However, the beam quality of the mini-bar is asymmetric. In the Y-direction (the fast-axis, perpendicular to the junction), the beam quality is nearly diffraction-limited. On the other hand, in the X-direction (the slow-axis, parallel to the junction), the beam quality is poor because of the wide emitting aperture (approximately 100 μm) and the lightless areas between two adjacent emitters [1,2]. To obtain equal beam quality in both axes, S. Yamaguchi proposed a BTS that can interchange the divergence angle of the X-direction and Y-direction before the beams from adjacent emitters overlap [10]. However, in practice, this BTS has some problems that degrade the beam quality.
Table 1. Typical values of the mini-bar
In this work, an optimized BTS is proposed to improve the beam quality of the mini-bar. Compared with a commercial BTS, the optimized design can significantly reduce beam tilting and deterioration of residual divergence angle. After beam shaping, the beam qualities in the fast-axis and slow-axis are approximately equal; consequently, the focused beam can be coupled into a 100 μm core, 0.15 NA fiber with over 100W of output power and an electric-optical efficiency of 46.8%, which is demonstrated by both ZEMAX Software simulation and fiber-coupled experiments. (The electric-optical (E-O) conversion efficiency is defined as the fiber-coupled output power divided by the input electrical power of the whole system.)
2. Beam transformation system
The commercial BTS includes two optical elements. The first element is a fast-axis collimation lens (FAC lens), which is used to reduce the large divergence angle along the fast-axis. After the fast-axis collimation, the laser-diode beam propagates into the tilted cylindrical lens array, which includes several biconvex cylindrical lenses. As shown in Fig. 1(a), each biconvex cylindrical lens has infinite effective focal length (r1 = −r2, where r1 and r2 are the radii of curvature of surface 1 and surface 2, respectively). The lateral facet of each biconvex cylindrical lens and a horizontal plane form an angle of −45° ( ± 45° each can realize the beam twisting, but −45° is selected according to the parameters of the experimental BTS). After the fast-axis collimation, the XY cross section of the laser-diode beam and the meridian line of the biconvex cylindrical lens form an angle of −45° when the laser-diode beam is incident on surface 1. Under ideal conditions (no abaxial ray aberration and no divergence angle), the laser-diode beam will be perfectly focused on the focal plane of surface 1. Finally, the XY cross section of the laser-diode beam and the meridian line form an angle of 45° when the laser-diode beam is projected from surface 2. Thus, the original laser-diode beam is transformed into a beam in which the horizontal and vertical components (beam size and divergence angle) are interchanged, as shown in Figs. 1(b) and 1(c) obtained by ZEMAX simulation.
Fig. 1 The commercial beam transformation system (BTS): (a) Beam shaping principle of the cylindrical lens tilted at −45° and the beam profiles at different positions. (b) The simulated beam profile which is obtained at a distance of 5 mm from the fast-axis collimation lens (FAC lens). (c) The simulated beam profile which is obtained at a distance of 5 mm from the cylindrical lens array.
In the experiment, BTS365 (manufactured by LIMO GmbH; product model number: MOD000683) was applied to shape the laser-diode beam; BTS365 is designed for 976 nm, and its other key parameters are shown in Table 2. However, after beam shaping, there are two major problems, which degrade the beam quality of the mini-bar. The first problem is that the laser-diode beam didn’t perfectly interchange beam size along the X-direction and Y-direction when projected from BTS365. As shown in Figs. 2(a) and 2(b), both the simulated results obtained by ZEMAX simulation and the experimental results captured by the charge-coupled device (CCD) demonstrate that the laser-diode beam was tilted by a certain angle after beam shaping. This phenomenon resulted in an increase of the beam size along the X-direction and a reduction of the focusability of the laser-diode beam (when using a focusing lens with the same effective focal length (EFL), a larger beam size results in a bigger numerical aperture [11–13]).
Table 2. Key parameters of BTS365 (Material: S-TIH53, best performed at 976 ± 4 nm).
Fig. 2 Beam profiles after passing through BTS365: (a) The simulated beam profile which is obtained at a distance of 5 mm from BTS365. (b) The experimental beam profile which is captured at a distance of 5 mm from BTS365.
The uncollimated divergence angle in the slow-axis is the main reason for the tilted laser-diode beam after BTS365. As shown in Fig. 3(a), an uncollimated ray in the slow-axis can be divided into two components: one is perpendicular to the meridian plane of the biconvex cylindrical lens and the other one is parallel to the sagittal plane. As the cylindrical lens can only focus a ray that is parallel to the meridian plane, even if there is no ray aberration, the vertical component deviates the rays from the ideal position, which can be demonstrated by the distorted spot at the focus of surface 1. When projected from Surface 2, the rays which asymmetrically deviated along the X-direction didn’t form an exactly angle of 45° with the meridian line. As a result, the laser-diode beam was similar to be tilted by a certain angle along the Z-direction.
Fig. 3 Simulation analysis of the tilted angle of the laser-diode beam after BTS365: (a) Propagation of the laser-diode beam in the biconvex cylindrical lens. (b) The tilted angle of the laser-diode beam as a function of the slow-axis divergence angle. (c) The tilted angle of the laser-diode beam as a function of the radius of curvature.
Two parameters determine the tilted angle (α) of the laser-diode beam after BTS365: (1) the divergence angle (θ) of the laser-diode beam along the slow-axis; (2) length (L) of the biconvex cylindrical lens. As shown in Fig. 3(b) (L = 1.65 mm), when θ = 1°, diameter of the laser-diode beam along the Y-direction is 0.2 mm and α = 2.4°. However, when θ = 7°, diameter of the laser-diode beam along the Y-direction becomes 0.8 mm and α increases to 12.5°. Therefore, a larger divergence angle results in a more severe beam tilting and a larger beam size along the Y-direction after BTS365.
Length (L) of the biconvex cylindrical lens is determined by the radius of curvature (r1) of Surface 1 (under ideal condition, L = 2 × r1/(n − 1), where n is the refractive index of the biconvex cylindrical lens), the influence of r1 on the tilted angle (α) can be seen in Fig. 3(c) (θ = 7°). With a larger radius of curvature, the laser-diode beam propagated over a longer distance. Consequently, the deviation caused by the vertical components was more evident.
Another problem caused by BTS365 is the deterioration of the residual divergence angle. FAC365 (manufactured by LIMO GmbH), which has the same parameters as that of the FAC lens used in BTS365, was employed to compare its collimation ability with BTS365. After the collimation of FAC365, the beam size along the Y-direction was approximately 0.5 mm. Considering the small radius of curvature (r1) of Surface 1, some marginal rays are not in the ideal paraxial area, and rays from each emitter will be focused into a fuzzy spot at point F (the diameter of the spot was approximately 5.0 μm when r1 = 0.37 mm) because of abaxial ray aberration (mainly spherical aberration). When rays are projected from Surface 2, abaxial ray aberration will further degrade the residual divergence angle.
As shown in Fig. 4(a), the simulation results indicate that the residual divergence angle after FAC365 is 2.7 mrad (enclosed 90% energy), and the residual divergence angle after BTS365 becomes 5.5 mrad (enclosed 90% energy). In the experimental results, the residual divergence angle is slightly larger than the simulation results because of the alignment error (all measurements of the residual divergence angle were performed by the setup based on the Gaussian beam propagating theory [1,2], as shown in Fig. 4(b); all measurements of the beam size were performed by the setup shown in Fig. 4(c)). The conclusion is that the residual divergence angle after BTS365 doubled compared with that after FAC365 (5.7 mrad compared with 2.8 mrad).
Fig. 4 Comparison of the residual divergence angle and schematic of the experimental setups for beam quality measurement: (a) The residual divergence angle along the fast-axis after BTS365 and FAC365 (two curves: the simulation results, two far-field beam profiles: the experimental results captured by CCD @ EFL = 300 mm). (b) The experimental setup for the residual divergence measurement. (c) The experimental setup for the beam size measurement.
3. Optimized design and experimental results
Based on the analysis above, three methods can eliminate the influence of ray aberration, as shown in Table 3. Method 1: a mini-bar with a small divergence angle along the fast-axis has a thick epitaxial layer (because beam quality of the fast-axis is nearly diffraction-limited), this results in a large residual divergence angle after the FAC lens [1,2]; Method 2: Surface 1 with a larger radius of curvature causes the laser-diode beam propagating over a longer distance in the BTS. Consequently, beam tilting is more severe; Method 3: the FAC lens with a small EFL can decrease the beam size along the fast-axis before the laser-diode beam incidents on Surface 1, abaxial ray aberration can be reduced because more rays are paraxial. Thus, FAC150 (manufactured by LIMO GmbH) was employed in the optimized system; this unit has an EFL = 150 μm, which is the smallest in the market.
Table 3. Improvement methods for ray aberration.
As shown in Table 4, three methods can suppress beam tilting after the BTS. Method 1: a mini-bar with a small divergence angle (< 1°) along the slow-axis has a complex structure, a low wall-plug efficiency, and a low reliability; Method 2: if the laser-diode beam is collimated along the slow-axis before beam shaping by the BTS (according to the width of each emitter, the EFL of the slow-axis collimation lens should be several millimeters), a longer optical path causes an increase of the beam size along the fast-axis and leads to more severe abaxial ray aberration; Method 3: reducing the radius of curvature of Surface 1 can be effective in suppressing beam tilting provided that the parameters of the biconvex cylindrical lens meet the requirements given by Eq. (1) in order to control energy loss:
where T shown in Fig. 5(a) is the thickness of the biconvex cylindrical lens; r1 is the radius of curvature of Surface 1; dY and dX are the beam size along the Y-direction and X-direction, respectively. Based on the parameters in Table 1, dY is approximately 0.17 mm and dX is approximately 0.24 mm after the collimation of FAC150. Thus, the optimized parameters of the biconvex lens can be calculated (r1 = 0.18 mm, L = 0.76 mm, conic constant = −0.30). However, considering the pitch limitation (based on the optimized parameters, the maximum thickness (0.36 mm) of the biconvex cylindrical lens divided by a factor of √2 is smaller than the pitch (0.5mm) between two adjacent emitters), a group of plates (material: Schott BK7 borosilicate glass, thickness: 0.35 mm, length: 0.76 mm) are inserted among the biconvex cylindrical lenses; this can make the alignment much easier, as shown in Fig. 5(a). The incident facet of each plate is coated with a high-reflectivity film (reflectivity > 99.9% @ 976 nm); thus, the transmitted power can be used to indicate whether the biconvex cylindrical lens array is in the optimal place.Table 4. Improvement methods for beam tilting.
Fig. 5 Schematic of the optimized BTS: (a) The optimized cylindrical lens array. (b) The laser-diode beam propagates in the plane perpendicular to the junction. (c) The laser-diode beam propagates in the plane parallel to the junction.
The most convenient way to measure the beam quality of a laser-diode beam is to characterize the beam parameter product (BPP) [1,2]; this is defined as θ × W (where θ is the far-field half-divergence angle and W is the waist radius of the laser-diode beam). To couple the beam of the mini-bar into a 100 μm core, 0.15 NA fiber with high efficiency, the beam quality of the laser-diode beam before focusing should meet the requirements given by Eq. (2):
where BPPY is the BPP of the laser-diode beam in the Y-direction, BPPX is the BPP of the laser-diode beam in the X-direction, and dfiber and NA denote the diameter and numerical aperture of the fiber, respectively. The factor of √2 accounts for the geometry of coupling a square beam into a round fiber, as the BPP of the laser-diode beam must be less than BPPfiber divided by a factor of √2.From the experimental results in Figs. 6(a) and 6(b), the beam size and divergence angle along the X-direction after the optimized cylindrical lens array (at position A) are 4.7 mm and 7.7 mrad (the calculated BPP is 9.1 mm × mrad), the laser-diode beam cannot be coupled into the 100 μm core, 0.15 NA fiber. Considering that the lightless area still exists between two adjacent emitters, the FAC lens array was applied to perform the re-collimation and further reduce the residual divergence angle along the X-direction, as shown in Figs. 5(b) and 5(c). To achieve equal beam size and divergence angle in both axes without beam expanding or compressing, the EFL of the slow-axis collimation lens (SAC lens) should be 40 mm. As shown in Figs. 6(a) and 6(b), the beam size and divergence angle in the Y-direction after the SAC lens (at position B) are 4.9 mm and 3.6 mrad. The ideal divergence angle after the SAC lens is 2.5mrad; however, because of the pointing error of each emitter, 3.6 mrad was achieved in the experiment. Finally, the EFL of the FAC lens array should be 47.6 mm to match the divergence angle in the Y-direction after the SAC lens. Thus, equal beam quality in both axes was achieved by means of the beam shaping of the optimized BTS, as shown in Fig. 6(c) (at position C).
Fig. 6 Beam quality in different positions: (a) The near-field beam profiles (enclosed 90% energy). (b) The far-field beam profiles @ EFL = 300 mm (enclosed 90% energy). (c) BPPX and BPPY.
After beam shaping with the optimized BTS, the laser-diode beam passed through the aspherical lens (EFL = 18.75 mm, manufactured by Edmund Optics, product model number: F49-110). The focal spot and the numerical aperture of the focal spot are shown in Fig. 7(a) and 7(b). As it can be seen, the spot dimensions along the Y-direction and X-direction are 67.5 μm × 70.1 μm (both are smaller than the diameter of the fiber core divided by a factor of √2), respectively; and the NA along the Y-direction and X-direction are 0.13 × 0.13, respectively. Hence, the beam can be effectively coupled into the 100 μm core, 0.15 NA fiber.
Fig. 7 Fiber-coupled experimental results: (a) The focal spot of the laser-diode beam (enclosed 95% energy). (b) The numerical aperture of the focal spot (enclosed 95% energy). (c) The output power and fiber-coupled efficiency.
At the coolant temperature of 20°C and CW operating mode, the results in terms of the fiber-coupled output power and fiber-coupled efficiency are shown in Fig. 7(c) (all the results were tested from 10 to 80 A). The measured maximum electric-optical (E-O) conversion efficiency was 53.5% while the fiber-coupled output power was 30.4 W. Corresponding to the maximum injection current, the measured E-O conversion efficiency was 46.8%. The optical-optical (O-O) conversion efficiency is defined as the ratio of the output power of the mini-bar to the fiber-coupled output power. Although the O-O conversion efficiency slightly decreased with the increasing injection current, which could be due to the increasing of the slow-axis divergence angle [1–6], the O-O conversion efficiency of 78.3% was achieved at an operating current of 80 A, corresponding to the fiber-coupled output power of 57.2 W. Considering that the polarization degree of the laser-diode beam is usually greater than 95%, more than 100 W of fiber-coupled output power could be achieved by using the polarization-combining method (two mini-bars with different polarization states).
4. Conclusions
In summary, the factors that degrade the beam quality of the laser-diode bars after the commercial BTS have been analyzed. Based on the analysis results, an optimized design is proposed to collimate and equalize the highly divergent elliptical beam of a mini-bar with 9 emitters. After beam shaping with the optimized BTS, the beam size and residual divergence angle of the mini-bar in the X-direction and Y-direction are almost equal: 5.0 mm × 3.6 mrad. The focused beam can be coupled into a 100 μm core, 0.15 NA fiber with a fiber-coupled output power of over 100 W and an electric-optical efficiency of 46.8%. Hence, this novel design is suitable for the application of high power laser-diode.
Funding
Research Foundation of China Academy of Engineering Physics (Grants No. 2015-909PY-4).
Acknowledgment
The author would like to thank all interested persons at the laser-diode Department of Institute of Applied Electronics in CAEP, especially Linhui Guo and Hualing Wu, for their helpful discussions and insights .
References and links
1. H.-G. Treusch and R. Pandey, “High-power diode laser arrays,” in High-Power Laser Handbook, H. Injeyan and G. D. Goodno, ed. (Academic, 2011), pp. 133–159.
2. U. Brauch, P. Loosen, and H. Opower, “High-power diode lasers for direct applications,” in High-Power Diode Lasers: Fundamentals, Technology, Applications, R. Diehl, ed. (Academic, 2000), pp. 303–368.
3. K. Price, S. Karlsen, P. Leisher, and R. Martinsen, “High brightness fiber coupled pump laser development,” Proc. SPIE 7583, 842102 (2010).
4. S. Heinemann, H. Fritsche, B. Kruschke, T. Schmidt, and W. Gries, “Compact high brightness diode laser emitting 500 W from a 100 μm fiber,” Proc. SPIE 8605, 86050Q (2013). [CrossRef]
5. D. Schröder, E. Werner, A. Franke, L. Wagner, G. Bonati, F. Dörfel, H. Ziemer, A. Liem, and T. Gabler, “Roadmap to low cost high brightness diode laser power out of the fiber,” Proc. SPIE 7583, 758309 (2010). [CrossRef]
6. D. G. Matthews, K. Kleine, V. Krause, A. Koesters, D. Duennwald, and S. Pflueger, “A 15 kW fiber-coupled diode laser for pumping applications,” Proc. SPIE 8241, 824103 (2012). [CrossRef]
7. S. Yamaguchi, T. Kobayashi, Y. Saito, and K. Chiba, “Collimation of emissions from a high-power multistripe laser-diode bar with multiprism array coupling and focusing to a small spot,” Opt. Lett. 20(8), 898–900 (1995). [CrossRef] [PubMed]
8. S. H. Ghasemi, M. R. Hantehzadeh, J. Sabbaghzadeh, D. Dorranian, M. Lafooti, V. Vatani, R. Rezaei-Nasirabad, A. Hemmati, A. A. Amidian, and S. A. Alavian, “Beam shaping design for coupling high power diode laser stack to fiber,” Appl. Opt. 50(18), 2927–2930 (2011). [CrossRef] [PubMed]
9. J. Yu, L. Guo, H. Wu, Z. Wang, H. Tan, S. Gao, D. Wu, and K. Zhang, “High brightness beam shaping and fiber coupling of laser-diode bars,” Appl. Opt. 54(11), 3513–3516 (2015). [CrossRef] [PubMed]
10. S. Yamaguchi, M. Daimon, K. Chiba, T. Kobayashi, and Y. Saito, “Optical path rotating device used with linear array laser diode and laser appartus applied therewith,” United States patent 5513201 (April 30, 1996).
11. W. A. Clarkson and D. C. Hanna, “Two-mirror beam-shaping technique for high-power diode bars,” Opt. Lett. 21(6), 375–377 (1996). [CrossRef] [PubMed]
12. R. K. Huang, B. Chann, J. Burgess, M. Kaiman, R. Overman, J. D. Glenn, and P. Tayebati, “Direct diode lasers with comparable beam quality to fiber, CO2, and solid state lasers,” Proc. SPIE 8241, 824102 (2012). [CrossRef]
13. N. Coluccelli, “Nonsequential modeling of laser diode stacks using Zemax: simulation, optimization, and experimental validation,” Appl. Opt. 49(22), 4237–4245 (2010). [CrossRef] [PubMed]
