Laser-beam shaping is adopted by some laser manufacturers as a way to provide customers with more engineering control for laser machining to achieve better holes, more precise marking, and cleaner cuts. Some vendors are using fixed optics for beam shaping, but the alternative approach of adaptive optics is gaining momentum. By adding control of the beam-shaping optic surface, these systems can compensate for system aberrations as well as for nonidealities and changes in the input laser-beam profile. In addition, the device is reflective, and so it is achromatic and can be used to shape the intensity profile of lasers with a wide spectral bandwidth. A deformable mirror can also enable high-speed control of the position of the system focus to enable machining complex surfaces.
Unfortunately, most adaptive-optics systems do not meet the laser vendor's requirements. Both micro-electromechanical systems (MEMS)1 and conventional2deformable mirrors are still too costly for consideration in many systems. Most MEMS deformable mirrors also cannot handle the optical-power requirements of laser-machining systems. In addition, some existing deformable mirrors are not robust enough to handle the challenges posed by industrial machining environments.
At Active Optical Systems LLC (AOS),3 we have been developing polymer-membrane deformable mirrors (see Figure 1). Nitrocellulose versions of our mirrors have been available commercially for the last five years for only $1500, but they have only recently been applied to high-peak-power laser-machining applications.
Figure 1.Exploded 3D computer-aided-design rendering of a basic membrane deformable mirror.
To demonstrate the extreme robustness of these nitrocellulose membranes, we used air pressure to bring a half-inch-diameter aluminum-coated membrane to a radius of curvature of 1inch before rupturing. This corresponds to motion of approximately 800μm, which is well beyond the typical electrostatic actuation for comparable devices of 10μm.
We recently began manufacturing membranes using polyimide, which does not absorb water from the air as easily as nitrocellulose, to increase the deformable mirror's frequency response in high-humidity environments.4 We have shown that polyimide membranes are even more robust than nitrocellulose membranes, which tend to fracture easily when pressed and scratch very easily. In contrast, the polyimide membranes are resistant to scratching and can survive reasonable handling. Although these are not the first mirrors made using polyimide,5 they are the first commercially available polyimide deformable mirror on the market.
We built on previously developed infrastructure and processes used for creating covers for photolithography masks in the semiconductor industry. By leveraging these for nitrocellulose membranes, we created a new way of manufacturing low-cost polyimide membranes that were commercially viable for membrane deformable mirrors. By combining them with standard printed-circuit-board material, we were able to make the first electrostatically actuated polyimide-membrane devices that are easily customizable and can be manufactured inexpensively. We have taken these polyimide membrane mirrors from a research curiosity to a commercially available product that is robust and easy to use. A recent polyimide-membrane deformable mirror has demonstrated a 3dB frequency near 2kHz over a range of relative humidity between 30 and 55%.
We silver coated both nitrocellulose and polyimide membranes, with four layers of dielectric enhancement for 99.2% reflectivity, and evaluated them for laser-power handling. We used radiation from a 1060nm-wavelength quality (Q)-switched fiber laser with 10ns pulses at a 20kHz repetition rate to illuminate the mirrors with up to 19W of average output power. We monitored the mirror surfaces with a Shack-Hartmann wavefront sensor during testing to observe thermally induced distortion and any static distortion after termination of the laser illumination. The nitrocellulose membrane performed well at average powers less than 10W with a 7mm-diameter beam, but had some static distortion written into the mirror surface beyond this power level when laser illumination was terminated. At 19W average power, the nitrocellulose membrane exhibited approximately 115nm of rms surface distortion.
In contrast, the polyimide membrane had three times less thermally induced rms wavefront distortion over the test range and did not show any static distortion on the Shack-Hartmann wavefront sensor after termination of laser illumination. At 19W average power, the surface exhibited 45nm rms wavefront distortion. We tested a polyimide membrane with a Q-switched 355nm-wavelength laser and found that it performed well for up to 8W of average power in a 3.6mm-diameter Gaussian beam when coated with an aluminum coating enhanced with two layers of dielectric for 93% reflectivity.
After damage testing, we used a polyimide-membrane deformable mirror to perform beam shaping on the high-peak-power Q-switched 1060nm laser. The control system was able to shape the beam into a variety of shapes using a metric-based adaptive-optics system that is part of commercially available adaptive-optics software (see Figure 2).
Figure 2.Select intensity profiles achieved during beam-shaping experimentation.
We also used a new beam-shaping system that was miniaturized with commercial components to fit onto a 10×12inch2breadboard. This system contained a diode laser, a beam-expansion telescope to illuminate a 1inch-diameter membrane deformable mirror, a beam sampler that enabled illumination of a far-field camera, and a secondary beam line for delivery to a target or an additional Shack-Hartmann wavefront sensor for feedback.
In summary, we have shown that polymer-membrane deformable mirrors are now cost effective and robust enough for beam shaping and laser-beam aberration correction in laser-machining applications. We are continuing to develop these mirrors to improve their packaging, reflectivity, and environmental robustness and to miniaturize the overall beam-shaping and control system.
Justin Mansell, Brian Henderson, Gideon Robertson
Active Optical Systems LLC
Justin Mansell received his PhD from Stanford University in 2002 for developing MEMS adaptive optics for high-power lasers. He is a researcher in adaptive optics, where he has developed novel deformable mirrors, wavefront sensors, and adaptive-optics systems with a focus on enhancing capability and reducing cost for commercialization.
2. J. Mansell, High power plate-type deformable mirror, Mirror Technol. Days, 2007. Conf. presentation.
3. http://www.aos-llc.com. Homepage of Active Optical Systems LLC (AOS).
5. S. Chodimella, Design, fabrication, and validation of an ultra-lightweight membrane mirror, Proc. SPIE
5894, pp. 589416, 2005. doi:10.1117/12.619237