We have studied the output characteristics of the EESLD and found that, its emitting region is very elongated and the aperture constraints in the fast and slow axis directions are different. The far field of the output beam shows a Gaussian elliptical distribution, with inconsistent beam quality levels and divergence angles in both directions. In the fast axis direction, the divergence angle is typically 40°~50°, the beam quality is close to the diffraction limit and the energy distribution is consistent with a single-Gaussian-mode. In the slow axis direction, the divergence angle is generally 10°~15°, and the width of the luminous area is often greater than the minimum width of the fundamental mode, showing a horizontal multi-mode [Shang et al. 2010].
In this work, a blue EESLD (NUBM 08) at wavelength 450 nm with 3.5 W output power is selected. We first measure the basic luminous parameters of NUBM 08. To fast axis direction, the divergence angle and the waist radius in are 46° and 10.2 µm. To slow axis direction, the divergence angle and the waist radius in are 10° and 2.3 µm. There is a certain aberration between fast and slow axis direction.
According to the emission characteristics of the beam in the fast axis direction, with the help of the optical software ZEMAX, an even-order aspheric lens is used. After selecting the initial structure of the aspherical lens, the REAY operand is used to control the deflection of the beam for different incident radii, adjusting the weights to optimize the surface parameters and achieve a uniform distribution of energy [Coluccelli 2010].
The surface parameters could be determined after selecting emergent light spot diameter as 3 mm, the aspheric lenses have a refractive index of 1.595. The operand instructions in ZEMAX concluded that the slow-axis homogenizing process completed with a plano-concave negative lens with a focal length of -20 mm and a plano-convex positive lens with a focal length of + 100 mm.
In order to determine the effectiveness of the shaping scheme for highly divergent EESLD beam spot, the optical system of the flattop beam shaper was tested with optics design software ZEMAX. As shown in Fig. 2(a), ZEMAX simulates the beam propagation process of the laser diode NUBM 08 in the shaping system. Figure 2(b) shows the light spot diagram very close to the light-emitting surface of the EESLD. Due to the divergence angle of the fast and slow axes, the light beam spreads rapidly thereafter. It can be seen from the image that the energy concentrates in the center of the beam with an unevenly distribution. The emergent light spot is rectangular when all the lenses are placed confocal, owing to the special structure of the EESLD light-emitting surface. After the light passes through the shaping system, as shown in Fig. 2(c), the light spot at 1.2 meters presents a very regular square, and the size of the light spot is 2.97×2.99 mm2. The uniformity of the spot energy distribution is greatly improved. Figure 2(d) shows the state of the beam without shaping system. Expanding the detector area 500 times in Fig. 2(d) to capture diffused light spot. The spot diagram at 1.8 m is shown in Fig. 2(e), and the size of the light spot is 2.98×3.01 mm2. It can be concluded that the spot size does not spread compared with the spot size at 1.2 m. The shaping system in this paper is supposed to be very effective in compressing the divergence angle of the outgoing beam of the EESLD.
Comparing the beam intensity distribution in fast-axis and slow-axis direction, respectively, and analyzing the energy distribution of the EESLD after propagating in the shaping system. Figure 3(a) and Fig. 3(b) are cross-sectional views of the light intensity before shaping in the fast axis and slow axis directions, respectively. The beam in the fast axis direction exhibits a standard single-Gaussian-mode. The slow-axis beam has a multi-mode distribution, which can be understood as the superposition of multiple Gaussian fundamental mode beams at the FWHM. After the shaping system, the energy distribution of the fast-axis beam is significantly improved compared to the unshaped beam uniformity. As shown in Fig. 3(c), the energy fluctuation at the edge is attributed to the diffraction effect at the edge of the optical system. As shown in Fig. 3(d), the uniformity of the energy distribution in the slow axis direction is also improved.
According to the Bessel formula, the parameter V for evaluating the uniformity of light intensity,
$$V=1 - \frac{{\sqrt {{{\sum\limits_{{i=1}}^{N} {{{\left( {{E_i} - {E_0}} \right)}^2}} } \mathord{\left/ {\vphantom {{\sum\limits_{{i=1}}^{N} {{{\left( {{E_i} - {E_0}} \right)}^2}} } N}} \right. \kern-0pt} N}} }}{{{E_0}}}$$
5
Where Ei is the light intensity value of the flattop area, E0 is the average energy value of the flattop site, and N is the number of sampling points in the flattop area. The light intensity uniformity V describes the deviation of the overall light intensity in the profile from the average light intensity. When V > 0.9, it can be considered that the beam homogenization work is practical. The shaping scheme adopted in this paper has a calculated uniformity V = 98.6%, which shows that the homogenization scheme designed in this paper is reasonable.
For investigating uniform pump is beneficial for fiber coupling, shaped pump source coupling into the fiber is simulated with ZEMAX, as shown in Fig. 4(a). The coupling model consists of a focusing mirror with a focal length of 20 mm and a piece of fiber S460-HP with a length of 20 cm. Here, single-mode fiber S460-HP with core/ cladding parameters of 3/125 µm and a numerical aperture (NA) of 0.13 was used to estimate the pump-launching efficiency. An annular surface is placed in order to prevent light from entering the cladding structure. By arranging a detector at each end of the fiber, the beam spot energy distribution on the two detectors is shown in Fig. 4(b) and Fig. 4(c). Varying the EESLD output power to test the coupling efficiency. Figure 5 shows the typical pump-launching coupling properties. According to the EESLD output characteristics, the maximum outgoing power is 3.5 W and the efficiency is 58.35%.