Current technology cuts solar Si wafers by a wire saw process, resulting in 50% 'kerf' loss when machining silicon from a boule or brick into a wafer. We want to develop a kerf-free laser wafering technology that promises to eliminate such wasteful wire saw processes and achieve up to a ten-fold decrease in the g/W{sub p} (grams/peak watt) polysilicon usage from the starting polysilicon material. Compared to today's technology, this will also reduce costs ({approx}20%), embodied energy, and green-house gas GHG emissions ({approx}50%). We will use short pulse laser illumination sharply focused by a solid immersion lens to produce subsurface damage in silicon such that wafers can be mechanically cleaved from a boule or brick. For this concept to succeed, we will need to develop optics, lasers, cleaving, and high throughput processing technologies capable of producing wafers with thicknesses < 50 {micro}m with high throughput (< 10 sec./wafer). Wafer thickness scaling is the 'Moore's Law' of silicon solar. Our concept will allow solar manufacturers to skip entire generations of scaling and achieve grid parity with commercial electricity rates. Yet, this idea is largely untested and a simple demonstration is needed to provide credibility for a larger scale research and development program. The purpose of this project is to lay the groundwork to demonstrate the feasibility of laser wafering. First, to design and procure on optic train suitable for producing subsurface damage in silicon with the required damage and stress profile to promote lateral cleavage of silicon. Second, to use an existing laser to produce subsurface damage in silicon, and third, to characterize the damage using scanning electron microscopy and confocal Raman spectroscopy mapping.
Sandia National Laboratories has developed a means of manufacturing high precision aspheric lenslet arrays turned on-center. An innovative chucking and indexing mechanism was designed and implemented which allows the part to be indexed in two orthogonal directions parallel to the spindle face. This system was designed to meet a need for center to center positioning of 2μm and form error of λ/10. The part utilizes scribed orthogonal sets of grooves that locate the part on the chuck. The averaging of the grooves increases the repeatability of the system. The part is moved an integral number of grooves across the chuck by means of a vacuum chuck on a tool post that is mated to the part and holds the part while the chuck repositions to receive the part. The current setup is designed to create as many as 169 lenslets distributed over a 3mm square area while holding a true position tolerance of 1μm for all lenslets.
The LIGA microfabrication technique offers a unique method for fabricating 3-dimensional photonic lattices based on the Iowa State "logpile" structure. These structures represent the [111] orientation of the [100] logpile structures previously demonstrated by Sandia National Laboratories, The novelty to this approach is the single step process that does not require any alignment. The mask and substrate are fixed to one another and exposed twice from different angles using a synchrotron light source. The first exposure patterns the resist at an angle of 45 degrees normal to the substrate with a rotation of 8 degrees. The second exposure requires a 180 degree rotation about the normal of the mask and substrate. The resulting pattern is a vertically oriented logpile pattern that is rotated slightly off axis. The exposed PMMA is developed in a single step to produce an inverse lattice structure. This mold is filled with electroplated gold and stripped away to create a usable gold photonic crystal. Tilted logpiles demonstrate band characteristics very similar to those observed from [100] logpiles. Reflectivity tests show a band edge around 5 μm and compare well with numerical simulations.
The potential for electrochromic (EC) materials to be incorporated into a Fabry-Perot (FP) filter to allow modest amounts of tuning was evaluated by both experimental methods and modeling. A combination of chemical vapor deposition (CVD), physical vapor deposition (PVD), and electrochemical methods was used to produce an ECFP film stack consisting of an EC WO{sub 3}/Ta{sub 2}O{sub 5}/NiO{sub x}H{sub y} film stack (with indium-tin-oxide electrodes) sandwiched between two Si{sub 3}N{sub 4}/SiO{sub 2} dielectric reflector stacks. A process to produce a NiO{sub x}H{sub y} charge storage layer that freed the EC stack from dependence on atmospheric humidity and allowed construction of this complex EC-FP stack was developed. The refractive index (n) and extinction coefficient (k) for each layer in the EC-FP film stack was measured between 300 and 1700 nm. A prototype EC-FP filter was produced that had a transmission at 500 nm of 36%, and a FWHM of 10 nm. A general modeling approach that takes into account the desired pass band location, pass band width, required transmission and EC optical constants in order to estimate the maximum tuning from an EC-FP filter was developed. Modeling shows that minor thickness changes in the prototype stack developed in this project should yield a filter with a transmission at 600 nm of 33% and a FWHM of 9.6 nm, which could be tuned to 598 nm with a FWHM of 12.1 nm and a transmission of 16%. Additional modeling shows that if the EC WO{sub 3} absorption centers were optimized, then a shift from 600 nm to 598 nm could be made with a FWHM of 11.3 nm and a transmission of 20%. If (at 600 nm) the FWHM is decreased to 1 nm and transmission maintained at a reasonable level (e.g. 30%), only fractions of a nm of tuning would be possible with the film stack considered in this study. These tradeoffs may improve at other wavelengths or with EC materials different than those considered here. Finally, based on our limited investigation and material set, the severe absorption associated with the refractive index change suggests that incorporating EC materials into phase correcting spatial light modulators (SLMS) would allow for only negligible phase correction before transmission losses became too severe. However, we would like to emphasize that other EC materials may allow sufficient phase correction with limited absorption, which could make this approach attractive.
The ''Design and Manufacturing of Complex Optics'' LDRD sought to develop new advanced methods for the design and manufacturing of very complex optical systems. The project team developed methods for including manufacturability into optical designs and also researched extensions of manufacturing techniques to meet the challenging needs of aspherical, 3D, multi-level lenslet arrays on non-planar surfaces. In order to confirm the applicability of the developed techniques, the team chose the Dragonfly Eye optic as a testbed. This optic has arrays of aspherical micro-lenslets on both the exterior and the interior of a 4mm diameter hemispherical shell. Manufacturing of the dragonfly eye required new methods of plunge milling aspherical optics and the development of a method to create the milling tools using focused ion beam milling. The team showed the ability to create aspherical concave milling tools which will have great significance to the optical industry. A prototype dragonfly eye exterior was created during the research, and the methods of including manufacturability in the optical design process were shown to be successful as well.
In order to optically vary the magnification of an imaging system, continuous mechanical zoom lenses require multiple optical elements and use fine mechanical motion to precisely adjust the separations between individual or groups of lenses. By incorporating active elements into the optical design, we have designed and demonstrated imaging systems that are capable of variable optical magnification with no macroscopic moving parts. Changing the effective focal length and magnification of an imaging system can be accomplished by adeptly positioning two or more active optics in the optical design and appropriately adjusting the optical power of those elements. In this application, the active optics (e.g. liquid crystal spatial light modulators or deformable mirrors) serve as variable focal-length lenses. Unfortunately, the range over which currently available devices can operate (i.e. their dynamic range) is relatively small. Therefore, the key to this concept is to create large changes in the effective focal length of the system with very small changes in the focal lengths of individual elements by leveraging the optical power of conventional optical elements surrounding the active optics. By appropriately designing the optical system, these variable focal-length lenses can provide the flexibility necessary to change the overall system focal length, and therefore magnification, that is normally accomplished with mechanical motion in conventional zoom lenses.