Absolute accuracy evaluation of aspherical null testing for EUVL mirrors

Abstract

Interferometric null metrology can produce highly precise figure of aspherical surfaces. However, because the measurement is a direct comparison of the tested surface with the reference wavefront of the null optics, measurement accuracy is equivalent to the quality of the reference null wavefront. Although the asymmetric aberration of the reference null wavefront can be calibrated by rotating the tested surface, it is more difficult to calibrate the rotationally symmetric errors. Especially, the aspherical surface of EUVL mirrors must to be measured with higher accuracy, 0.2~0.3nmRMS. We have developed an aspherical null testing system using a null lens for EUVL mirrors. We analyzed the uncertainty of null lens in each process and estimated the measurement accuracy of aspherical null testing using null lens. If the compensator lens contains only one piece of lens, the measurement accuracy is estimated to be 0.20nmRMS. If the compensator contains two pieces, the measurement accuracy becomes 0.24nmRMS. To verify our estimation, we evaluate a sample lens with aplanatic surfaces that make no spherical aberrations. In this case, we can evaluate the quality of the transmitted wavefront absolutely. The difference between the calculated and the experimental wavefronts is much smaller than our estimation. To this extent, our aspherical testing technique using null lens has been verified to be able to meet the high demanding for EUVL mirror testing.