The Manufacture of Random Phase Plates and Phase Zone Plates

Document Sample
The Manufacture of Random Phase Plates and Phase Zone Plates Powered By Docstoc
					                                                                                              Laser Science and Development – Vulcan

The Manufacture of Random Phase Plates and Phase Zone Plates
T Boland, D Pepler, T B Winstone, C Danson
Central Laser Facility, CLRC Rutherford Appleton Laboratory, Chilton, Didcot, Oxon, OX11 0QX, UK
Main contact email address:
Introduction                                                            The plate is developed in Universal developer3) for 45 seconds,
Vulcan has used Random Phase Plates (RPPs)1) and Phase Zone             then rinsed with water for a further 45 seconds and left to dry in
Plates (PZPs)2) for a considerable period of time as part of a          a fume cupboard.
suite of focal spot smoothing and manipulation techniques.
During this period of time the manufacturing processes have
altered due to availability issues with some of the original
chemicals used. Here we present the current manufacturing
process for the masks and the phase plates.
The new manufacturing method uses a visible photoresist to
coat a quartz plate and a UV sensitive resist on a glass plate.
Figure 1 below shows schematically how a design for either a
RPP or PZP is transferred from an acetate template onto a
quartz mask, and then from the quartz mask to the resist coated

Mask Generation
The design of the mask is computer generated using the
PostScript graphics language. The image of the mask is created
on to an acetate sheet using a high resolution Linotronic laser
printer that has a resolution of 2540dpi, having a smallest
feature of 10 microns. The uncoated quartz plate is cleaned by
drag wiping with acetone and the dust removed using a                     Figure 2. Exposure of optical mask.
compressed air blower.       Approximately 5ml of positive
photoresist (Shipley S1813 SP-15)3) is applied to the central
region of the plate, which is then spun at 1500rpm for                  Phase Plate Manufacture
60 seconds. The plate is then placed in a preheated oven
(115°C) and baked for 4 minutes.                                        The phase plate is made using a glass or quartz substrate and
                                                                        coating it with UV photo-resist. The thin film depositions on
When the plate has cooled it is ready to be exposed. The plate          phase plates manufactured at RAL are produced using the
is placed resist side down onto the printed side of the acetate         Microlithographical Chemical Corporation’s PMMA A11
mask and exposed in a UV light box for 20 seconds as shown in           positive radiation sensitive resist3) and thinned using a standard
Figure 2.                                                               photo resist thinner.

    Figure 1. Showing how a pattern is transferred from acetate to quartz mask, then from the quartz mask to the glass plate.

                                                                  185                  Central Laser Facility Annual Report 2001/2002
Laser Science and Development – Vulcan

The depth of the resist is of importance as it determines the           The three most common errors, why they occur and how we test
operating wavelength of the plate. It can be determined by              for them are:
varying the spin speed and the dilution of the resist. The two
main concentrations used are 60/40 and 50/50.                           •     The photoresist layer is not the correct thickness. This
                                                                              occurs if the spin speed or resist concentration is not
A number of test plates were spun using these concentrations at               correct or even if the room temperature is different from
different spin speeds. The resist thicknesses were then                       that when the tests were carried out.
measured using an interference microscope. The results are
shown in Figure 3.                                                      •     The 0 and π phase areas are not exactly the same. This
                                                                              can be due to an incorrect mask design or an exposure
Generally phase plates used on Vulcan are manufactured for use                problem during the manufacture. The mask is fully
at the principal and second harmonics (1053nm and 527nm).                     checked in advance through numerical analysis and the
From the graph it is possible to optimize the spin speed and                  areas compared following manufacture.
concentration to produce the required thickness. For instance:
600 rpm at 60/40 concentration for 1053nm, and 900 rpm at               •     The phase transitions do not have an exact edge definition.
50/50 concentration for 527nm. Intermediate thicknesses, e.g.                 This is usually due to the mask not being fully in contact
for a plate to be used at 640nm, can be obtained by varying the               with the UV resist and becomes evident when the resist
concentration of PMMA to Thinner.                                             thickness is checked in the interference microscope.

The plate is baked for 20 minutes in an oven, preheated to              Accurate control of the production process and the use of fresh
160°C, then left to cool slowly to prevent the resist cracking.         chemicals for all procedures can minimize these errors.

For exposure the plate is placed resist side up and the mask is         References
placed on top with resist side down so that the coated sides of         1. Y Kato, K Mima, N Miyanaga, S Arinaga, Y Kitagawa, M
mask and plate are in contact. The plate is then exposed beneath            Nakatsuka and C Yamanaka, 'Random Phasing of high-
two mercury UV lamps, (253nm) positioned 10cm above the                     power lasers for uniform target acceleration and plasma-
plates. Exposure time is twenty hours for green (527nm) plates              instability supression,' Phys Rev Lett. 53, 1057-1060
and twenty-two hours for infrared (1053nm) plates.                          (1984)
Once the plate has been exposed it is developed in a mix of             2.    RM Stevenson, MJ Norman, TH Bett, DA Pepler, CN
50% Hexone (MIBK) and 50% Isopropyl Alcohol (IPA). The                        Danson and IN Ross, 'Binary Phase zone plate arrays for
plate is developed for thirty seconds whilst being gently                     the generation of uniform focal profiles,' Opt. Lett. 19,
agitated then washed in 100% IPA for the same amount of time.                 363-365 (1994)
Discussion                                                              3.    Chestech Ltd, 79 Somers Road, Rugby. CV22 7DG.
There are several fabrication errors that affect the quality of the
phase plate, and hence its operational performance. These lead
to a central spike being evident in the far-field. For small errors
it is possible to overcome this by placing the target slightly out
of the focal plane to achieve a smooth profile.

         Layer Thickness/nm                                                                         Concentration 60/40
                                                                                                    Concentration 60/40
            Layer Thickness (nm)                                                                    Concentration 50/50
                                                                                                    Concentration 50/50
                   500       600       700        800       900        1000     1100      1200      1300      1400      1500

                                                                                              Spin Speed inSpin Speed/rpm

        Figure 3. Results of spinning test plates as described above.

Central Laser Facility Annual Report 2001/2002                        186

Shared By:
Description: The Manufacture of Random Phase Plates and Phase Zone Plates