tEchnology DiFFrActiVE optics
➤
International, explains: ‘As far as the interferometer is concerned it is testing a sphere, and if it looks like a good sphere it means we’ve got a good asphere. We essentially make aspheres look like spheres, so they can be tested.’ Diffraction International was formed in 1993, and one of its first projects was developing CGH nulls for the fabrication of corrective optics for the Hubble Space Telescope (HST). Recently, the company has produced nulls for the large primary mirror segments of the James Webb Space Telescope (JWST), still under development. ‘Large optics, such as those for the James Webb Space Telescope, do require larger nulls, but usually not enormous, as they can be set far enough away to test the asphere,’ explains Arnold. There are some exceptions, such as some of the CGH nulls fabricated at the University of Arizona. The fabrication of CGH nulls begins with a ray
trace model using a commercial software package. The CGH substrate is tested and the pattern written on it in the same way that one would write a photomask for making an integrated circuit. ‘The process takes advantage of the
high-accuracy lithography tools used in the microelectronics industry,’ explains Arnold.
innovation in the design of the interferometric test. ‘One of the examples is the Fizeau system that we designed and built to test a 1.5m diameter convex off-axis parabolic mirror, where two CGHs are used to create two aspheric wavefronts, one for the test and the other for the reference,’ he says, adding that the same test has been modified slightly to measure the primary mirror segments of both the European Extremely Large Telescope (EELT) and the Thirty Meter Telescope (TMT).
cgh nulls, such as this one produced at the University of Arizona, combined with an interferometer, are one of the few methods of measuring aspheres to optical tolerances
‘This technology made it possible to produce CGH nulls, but where the advances have been are in the engineering tools used to design and certify the optics in an efficient manner. Now, Diffraction International can go through the whole engineering process in around a week, whereas an engineering project without all these tools would take at least six months.’
Eliminating unwanted diffraction orders As with other diffraction gratings, CGH nulls will split and diffract the light into multiple diffraction orders, only one of which is useful for measuring the asphere. All the other diffraction orders must be excluded to ensure they don’t interfere with the test. ‘It’s a matter of deflecting the unwanted orders so they can’t find their way back into the testing apparatus,’ explains Arnold, which is typically achieved through the design of the CGH – by adding a small decentre or tilt to separate the orders, for example. ‘This is easier to do for a large optic; it’s very hard to do when testing a standard- sized optic positioned close to the interferometer.’ At the University of Arizona, according to Dr
Zhao, linear or power carriers are introduced, which reduce the line spacings of the CGHs, to separate unwanted orders. ‘For a given writing uncertainty, the error of the CGH created wavefront is inversely proportional to the line spacings,’ he says. ‘So the design guideline is that you introduce just enough carriers to separate the unwanted orders, but not more.’ In addition, these unwanted orders cost
photons – not all the input light goes into the test, so a high-power laser is sometimes required to provide enough light intensity. Dr Zhao feels that although the technology
surrounding design and fabrication of CGHs is relatively mature, there is a lot of room for
22 ElEctro optics l MARCH 2011
Eliminating the zero order Moving away from interferometry and returning to diffractive optics for industrial applications, Dr Andreas Hermerschmidt, of German optics manufacturer Holoeye Photonics, says one of the improvements in the field of diffractive optics over the years is in dealing with the undiffracted light. A DOE is typically made up of an optical substrate with a microstructure etched into one of the surfaces. Without the microstructure, the substrate acts as an optical element reflecting or transmitting the original wave, which is known as the zero order or undiffracted light. ‘Generally, DOE manufacturers design their
optics to suppress the zero order,’ states Dr Hermerschmidt. ‘A beam splitter for material processing designed to drill four holes must not have a beam on the optical axis, which would result in a fifth hole. This problem arises in particular for an even number of spots. You want to control the amount of power in the undiffracted order, and doing this requires simulation and getting the fabrication and lithography processes stable, reliable and repeatable.’
limitations There are limits to the complexity of the pattern that can be achieved with diffractive optics, which largely depends on the size of the beam interacting with the optic and the wavelength of light. ‘The resolution is diffraction limited,’ explains Lloyd of Laser Optical Engineering, ‘so the limits on the optic tend to be a function of the laser rather than the design process.’ He adds that the power density to which the optic is exposed has to be kept to a reasonable level to avoid burning out the optic. Lloyd says fabricating DOEs operating at
shorter wavelengths – engineering diffractive optics for UV applications, for instance – would open some potential markets. ‘DOEs can be engineered to operate in the UV, but they are not very efficient and there’s too much noise,’ he says. ‘As a result of UV being such a short wavelength, the features of the optic get much smaller, making it difficult to manufacture. To get down to that size we need to play around with different, more accurate technologies than are currently in use.’ l
www.electrooptics.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40