Maskless lithography

Several attempts have been made to produce maskless lithography via an liquid crystal display. This has led to two diploma work at  Linköping University and Umeå University.  

first  experiment

The first experiments were made with a disassembled pocket calculator displays to verify the method to expose pattern directly on photoresist. In this case a simple contact copying procedure was performed. A pattern in the form of  digital signs from the display were copied directly to the photoresist. It worked fine to prove the method.


This is an image of the result from the contact copying experiment, 13:th November 2001  An ordinary mercury lamp was used and  the display was pressed directly on the  photoresist coated glass.

Second experiment

The second attempt  were done with an transmittive LC display from Kopin. It was a small display with resolution of  ca 290 X 218 pixels. The pixel pitch were circa 15um.  This display was placed in the Gloria 2x reduction UV projector.  Due to very low etendue, exposure times of 2-3 minutes were needed. However very fine patterns were made.  The Gloria projector had a  Rodenstock Apo rodagon N with 2x reduction and 50mm focus. It was the best available optics for a low cost. It had quite good MTF and large aperture. However designed for the whole visible spectrum and in this case only 405nm wavelength were used. It was not possible to do any simulation  becouse of the unknown lens curvatures.  Some data of NA is however avalible for the whole visible spectra..
MTF of  Apo rodagon N 50mm.
lens data page 1
lens data page 2
This image shows the finest lines in the bottom of the picture,  wich are only one pixel wide. Due to the reduction optics the line with is appr 5um wide.

Here is a close up of the kopin display.  Model RS170. Electronic datasheet of the RS170 and a short description of the pin configurations.
The Kopin display was mounted on an aluminum frame wich hold the drive electronics also. In this way the whole assembly could easily be pushed into the gloria UV projector instead of the reticle.
 image of the whole display.

Third experiment

A more complex solution were needed if high numerical aperure and MTF should be reached. This was necesery if short exposure times and high resolution is a goal.  The best choice were a catadioptric design wich incorporated a mirror and lenses. Polarised light were used to be able to link off the  light without obscuration in the light path. The light was circulary polariced in a quarter wave plate to turn the polarisation angle 90 degrees in the second path, and the light could be linked down in a cube beamsplitter.  This also enabled the use of the polarisation beam splitter as an aperture for a focusing laser. Many other issues were also solved in this design. It could be made very compact, high etendue, even a option of a CCD camera, mounted over one polarisation beam splitter and take real time image on the exposure area from a dark field microscope mounted on top.This function was critical if mask alignment could be used in multi layer devices.

The report from VTT Electronics. (pdf)

VTT managed to achive very good MTF for the lithographic lens. The major goal was to be able to assemble the 14 different optical components in just two NC machined block without post adjustment. This made the lens very suited for massproduction.

MTF for the second lens. Take in account that this MTF was measured without the coverglass from the Lcos display wich cause appr. 10 percent lower  MTF value than it actually is. At 300-400 l/mm there is a slight measuring error due to  limitations in the slitwith in the measuring equipment.

The diploma reports

Autofocusering av litografisk lins med astigmatisk metod.(word)


The setup with a Nd:YAG laser for astigmatic focusing at +/-3um.

Microlithography_for_half_tone_gobos (word)
Microlithography for half tone Gobos (pdf)


An image of the prototype, still without the autofocusing and the positioning table.


This LED lamp was used for some of the exposure tests. 91 parallell GaN diodes collimated by a 50mm large condensor lens.

this research system was Developed by VTT Electronics in Finland. Some practical  tests were performed in Jönköping and later in Norrköping at ITN.  The tests were performed without proper focusing but in theory an astigmatic  focusing system could be used, integrated in the system. Also some inventive lightsource  with GaN LED source could replace unplesant ultra high pressure mercury discharge lamps.

This image shows an early attempt to prove the resolution of the system. A line pattern was created via  the Lcosdisplay and blue 405 nm light from an array of GaN LEDs produced this pattern. image.

The first sucessful  exposure on a photoresist coated a glass substrate. The diameter of the glass is 37.5mm.  An image was created from a computer screen.  The display is not  masked  yet.  image.

One of the first test of a full scale rastered image. Note some vertical  lines over the image. These are from dead pixel rows from the display. Still the image lack of uneven exposure and bad focus. image.

Chrome etching with ceriumcompound