Pattern distortion in nanoimprint lithography using UV-curable polymer stamps

Quantification of pattern distortion in nanoimprint lithography (NIL) is required when applying it to specific applications, especially those with tight tolerances. We present a systematic study on full wafer NIL distortion using soft stamps made of different carrier foils and UV-curable polymer structure layers. These errors are evaluated by overlay patterning using NIL and optical lithography on 4-in. wafers over a distance of 80 mm. Potential causes for pattern distortion and possible correction methods are discussed in terms of stamp composition and environmental impact. Pattern distortion along axes causing dimensional change is stamp dependent, and stiffer stamps show less pattern dimensional change than the softer ones. In the best case, the minimum variation is 4 parts per million (ppm), and in the worst case, 252 ppm with a softer stamp. Stamp flatness and uniform contact during imprinting are important in reducing high-order pattern distortion. A maximum dimensional variation of 32 ppm in a batch run demonstrates good pattern repeatability. Long-term dimensional stability can be affected by relative humidity, with variations on the order of 100 ppm. • Nanoimprint distortion is evaluated by overlay patterning using nanoimprint and optical lithography. • Linear dimensional change along the axes is shown in nanoimprint and is found to be stamp-dependent. • Stamps made from OP-PS381 show less distortion in UV nanoimprint than stamps made from OP-PS91. • Good repeatability is demonstrated in a batch run but not over long period.