Linewidth uniformity in Lloyd’s mirror interference lithography systems
This is a schematic of a Lloyd’s mirror IL system. The mirror and substrate to be exposed are rigidly locked together at a angle. A spatially filtered laser source located in the plane illuminates the substrate (represented by the circular area parallel to the plane) and mirror resulting in a standing wave pattern that can be recorded in a photoresist.
These two plots show how (a) average dose and (b) dose modulation vary across the exposure area for the example Lloyd’s mirror configuration described in the text. The dose is normalized by setting the maximum dose to 1.
These two plots show the variation in duty cycle (linewidth/period) for a positive resist exposed on Lloyd’s mirror system described in the text at a spatial period of 200 nm. (a) This plot shows the uniformity at a relatively low dose ( at ); the linewidth variation mirrors the variation in average dose shown in Fig. 2(a). (b) The plot shows the uniformity at a higher dose ( at ); here the effects of varying dose modulation become more apparent.
(a) This figure shows the linewidth uniformity for the same dose, , as was used in Fig. 3(b), except that the spatial filter has been moved closer to the substrate, increasing the variation in average dose. This offsets the effect of the dose modulation variation in the -direction and reduces the linewidth variation in the -direction. (b) This figure shows variation in duty cycle along the -axis for the two cases (the dashed line is for a 2 m pinhole distance, the solid for a 1.5 m distance).
This micrograph shows a typical sample of the type produced using the double-exposure technique; in this case, Sumitomo PFI-88 resist was exposed with a period of 500 nm. This is a section of a larger image produced by stitching together multiple individual micrographs.
Samples of Sumitomo PFI-88 photoresist were exposed using the double-exposure technique at a total of ten different dose values. For each sample, 20 SEM images were made and stitched together. The linewidth was then extracted from each image in the areas where lines developed properly. The numbers next to each curve in the graph are the normalized exposure dose, that is, the exposure dose divided by the value of (clipping dose) that results from fitting the linewidth measurements using a binary-resist model. The left end of each curve represents the minimum dose modulation required for lines to develop properly for that dose.
Article metrics loading...
Full text loading...