Topic tags: 

Scaling up

With construction on the European Extremely Large Telescope due to start this year and a number of laser fusion research projects underway, the need for large optics is growing, as Greg Blackman finds out

Astronomy is all about the big questions: How was our planet formed? Are we alone in the Universe? These are the sorts of things that go through a star-gazer’s mind on a cloudless night. To look up at the night sky is to look back in time, as the light reaching us originated millions of years ago – the Andromeda Galaxy, one of the most distant celestial bodies visible to the naked eye, is 2.5 million light-years away. Therefore, clues to the formation of planets like ours can be seen in distant galaxies.

It’s been more than 400 years since Galileo Galilei turned his telescope skywards (2009 was the International Year of Astronomy marking the 400th anniversary) and, since then, ground-based telescopes have reached impressive dimensions and resolving power. This year construction of the European Southern Observatory’s (ESO) European Extremely Large Telescope (E-ELT) is due to commence and, at a proposed diameter of 42 metres, it will be the largest optical telescope in the world. On completion in 2018 it will gather more light than all of the existing 8-10m class telescopes currently in operation put together. The questions the E-ELT hopes to address, along with the other two extremely large telescope projects currently underway – the Thirty Meter Telescope (TMT) to be sited in Mauna Kea, Hawaii and the Giant Magellan Telescope (GMT) to be built in Las Campanas, Chile, both also due for completion in 2018 – are no less impressive in scale and include the search for Earth-like planets where life could exist, measuring the properties of the first objects in the Universe and the evolution of galaxies, and probing the nature of dark matter.

The primary mirror of the E-ELT will be made up from 906 segments, each 1.45m corner to corner. Each segment will be an off-axis asphere and will be positioned in a mosaic to form the total aspheric primary mirror. Seven of the segments will be polished at the Technium OpTIC (Opto-electronics Technology and Incubation Centre) facility in St Asaph, north Wales, using UK-based Zeeko’s high-precision optics polishing machines, (the facility houses various Zeeko equipment, including 2m, 1.2m and 600mm machines).

Working with such large optics presents some distinct challenges. Their physical size means they are heavy and difficult to move around. Zeeko’s larger polishing machines have an interferometer from 4D Technology mounted above the chamber, to allow measurements to be taken without moving the optic during the polishing process. ‘Smaller optics can be transported easily to a separate interferometer in a production line,’ comments Richard Freeman, managing director at Zeeko, ‘but this isn’t possible with larger optics – they are large and heavy, and also very valuable.’ Zeeko provides small-tool polishing machines to manufacture complex shapes. A large constant radius lap can be used to produce a spherical optic, but a complex or freeform optic can’t be made in that way. Instead, a small tool a few millimetres in diameter is passed across the surface in a CNC-controlled way to ascribe a shape to the optic.

Dr Helmut Kessler, general manager of Europe at optical component manufacturer CVI Melles Griot, states that optics of this size can be worth several thousand pounds. ‘Generally, the larger an optic, the more expensive it is – and components such as windows for high-power lasers will be made from a very high grade of substrate material, increasing the value further. Therefore, end users will try and keep the size of the optic to a minimum,’ he says. Whereas a standard catalogue optic usually has around 85 per cent clear aperture, larger optics often have 90-95 per cent clear aperture to save material, which Kessler says is very much a tightening on standard catalogue tolerances.

At CVI Melles Griot’s site in the Isle of Man, optics up to about 1m are being polished, but generally the site polishes and coats optics up to 0.5m for large laser systems. Within the next few months, the site is looking to expand to polish and coat even larger substrates.

Making measurements

UK company, Optical Surfaces, manufactures optics up to 600mm in diameter, although it has produced parts in the order of 1.2m in diameter. The company has recently invested in a new coating facility, which will allow it to coat substrates up to 350mm in diameter. The facility will allow it to produce coated optics for customers and will also be used as an in-house test bed for conducting R&D work on coating larger optics.

Dr Aris Kouris, sales manager at Optical Surfaces, attributes the company’s ability to manufacture high-precision optics to the interferometric test equipment, which allows greater flexibility in the procedure. Interferometers can establish how close the substrate is to a particular surface shape to fractions of a wave (up to lambda over 20 at 633nm peak-to-valley for demanding applications).

4D Technology’s Dynamic Interferometry technology allows optical metrology to be made under difficult conditions, such as vibration or air turbulence (as is the case with Zeeko’s large polishing systems in which the interferometer is mounted on a ‘humming’ machine). ‘Laser interferometers traditionally need very quiet environments to operate, but 4D has developed a technique for making accurate readings in noisy environments,’ says Chip Ragan, director of marketing at 4D Technology.

Measurements of telescope mirror segments, for instance, generally have to be taken without the benefit of vibration isolation or environment control due to the optic’s size. It is common for 4D’s interferometers to be mounted on towers looking down on telescope primary mirrors. ‘The interferometer must be placed at the radius of curvature and therefore there are often tens of metres of air between the instrument and the mirror,’ explains Ragan. ‘Any airflow in that cavity shows up as an index change and a measurement error.’

4D has developed a system whereby all the data is captured very quickly, in tens of microseconds. The reading is fast enough that any environment disturbances can be averaged out easily with multiple readings, leaving an accurate measurement of the surface shape.

There are challenges with taking an interferometric reading that covers the entire surface area of large optics. During the initial stages of polishing, the substrate can deviate substantially from the intended shape. These deviations can often cause problems for interferometers, in that the fringe density returning to the receptor is too high – there are too many fringes over a given area. According to Ragan, one way to address this is to have an interferometer system that has a very high quality optical path providing high spatial resolution, maximising the fringe density that can be resolved. This is typically used with a computer generated hologram (CGH) in between the interferometer and the test part, a technique specifically for aspheric optics. The hologram takes the aspheric wavefront and reshapes it to bring it back to a spherical wavefront that the interferometer is able to measure. Each hologram has to be custom-designed for each mirror.

A PhaseCam interferometer from 4D Technology mounted on a polishing system. Image courtesy of Technium OpTIC.

The other option for making measurements over a large complex optic is stitching interferometry. Here, the interferometer moved across the surface to capture a series of snapshots, which can then be stitched together to give an overall measurement. The interferometer is either moved in an arc that reflects the radius of curvature or the optic is rotated to get snapshots that span from the centre to the outer edges. Zeeko provides stitching software to handle both examples.

A further method for measuring large optics is swing-arm profilometry, whereby the optic is rotated and a probe moved across the surface to generate a point-cloud measurement. This technique is part of the battery of metrology tests being used at the Technium OpTIC facility to test the E-ELT mirror segments.

Once the individual segments have been produced, the next challenge is getting those segments lined up to make one giant optic. 4D Technology has developed a multiwavelength interferometer that allows the user to measure the steps between the mirror segments and bring them into phase. The instrument contains multiple lasers within the system across a certain wavelength band, with at least one of them being tuneable. Two measurements are taken at different wavelengths and combined to give a synthetic wavelength that’s longer than either of the constituent wavelengths. ‘This longer wavelength can measure larger steps, millimetres in scale. This is not a normal application for laser interferometry, which is typically limited to hundreds of nanometres,’ explains Ragan.

Optics for laser fusion research

Stephen Mounsey’s article on research-class lasers in the last issue of Electro Optics looked at some of the largest of the large lasers currently under development. The National Ignition Facility’s (NIF) laser fusion project at Lawrence Livermore National Laboratory has delivered a one megajoule laser shot, a previously unheard-of amount of energy for a group of lasers to produce – and one that brings the ultimate goal of fusion ignition one step closer. The optics used in the system include KDP (potassium dihydrogen phosphate) and deuterated KD*P crystals of around 420 x 420mm, weighing up to 700kg – which, according to Dr Andrew Robertson, senior vice president, business development at Gooch and Housego, in terms of frequency doubling crystals, don’t come much larger.

Gooch and Housego’s Cleveland site manufactures the KDP/ KD*P crystals, which can take up to two years to grow, although the company, in conjunction with Lawrence Livermore, has accelerated this process. The crystals are diamond-turned, which involves passing a diamond piece over the surface of the material very quickly, paring off surface layers. They are coated with a sol-gel layer, firstly because of the size of the crystals and the problematic nature of conventional thin film adhesion on KDP/KD*P, but also because sol-gels have a high damage threshold in controlled environmental conditions, which is a requirement for optics used in these powerful laser systems.

‘In all these large, high-power laser systems, any errors in the wavefront caused by the optics will result in hotspots in the laser beam, which could damage optics further down the line,’ explains Peter Mackay at Gooch and Housego. ‘Every single component that goes into these systems is specified to a very high quality level for aberrations in the wavefront.’ He adds that special attention must be paid to macro-roughness requirements, which require specialised test equipment.

‘Generally, the larger the optic the more care that has to be taken in the tolerances and the manufacturing methods used,’ continues Robertson. ‘It’s going to be a very expensive piece of glass, whatever its use, and typically the tolerances would have to be tighter than a standard optic from a catalogue.’

The PACA2M magnetron sputtering machine, from Cilas, can coat optics measuring up to 2 x 2m.

The coating process for large optics must be such that there’s no gradient in the coating – the percentage reflectivity must be the same at the edges of the optic as at the centre. Kessler of CVI Melles Griot comments: ‘To get a coating gradient on a one-inch optic is virtually impossible using a standard coating chamber. With larger optics, however, the chamber must be set up to ensure that there is an even distribution of coating material to avoid gradients, which would distort the beam.’

The French company Cilas has developed PACA2M, an 18m-long magnetron sputtering machine, which can deposit optical coatings on parts measuring up to 2x2m, up to 40cm thick and weighing up to 1.5 tonnes. The machine has been used to silver coat the stainless steel laser reflectors positioned around the laser barrels in the French Commission for Atomic Energy’s (CEA) Laser Mégajoule (LMJ) project in Bordeaux, France. These reflect the light from the flash lamps used to pump the lasers.

The PACA2M machine has been designed to deposit material and create optical functions over different substrates, and Cilas is involved in coating mirrors for space telescopes on satellites – the resistant coatings produced with magnetron sputtering are ideally suited to these telescopes as they can withstand the severe environment found in space – and for ground-based telescopes.

‘Coating optics, as opposed to glass for buildings, requires high uniformity in the process over several layers of material,’ says Gilles Borsoni, Marseille branch general manager. ‘The optical function gained by the coating has to be equal over the entire surface of the substrate. To obtain this, an even thickness of coating needs to be deposited.’

Cilas is working very closely with the French laboratory, Institut Fresnel, a specialist in optical thin-fi lm coatings, which has developed a new class of optical monitoring system that is part of the PACA2M machine. This improves the level of control and allows the thickness of the layer to be corrected during deposition to even out any deviations in the process.

The 2.5m magnetrons used in the PACA2M machine (to coat a 2m substrate uniformly requires a larger magnetron) were developed in conjunction with Alliance Concept, a French company specialising in magnetron sputtering, supported by POPSud/Optitec, the French competitiveness cluster in optics and photonics, and funded by the French government.

‘There are certainly a number of research projects under development that are pushing the optics manufacturing community to greater capacity in terms of large optics – up to 1.5m at reasonably high volumes,’ says Gooch and Housego’s Mackay.

‘It’s one of those areas where within the next few years, if laser fusion shows the potential to work, there will be a lot of interest in capabilities in manufacturing large optics,’ adds Robertson, also of Gooch and Housego. ‘There are half a dozen other laser fusion projects in the planning stages that could get the go-ahead if the NIF project shows promise.’