Considerations in the evaluation and correction of mid-spatial frequency surface features

Surfaces are commonly specified with peak to valley (PV) and root-mean-square (rms) requirements for surface form and roughness, describing surface quality with a few simple numbers. These specs ignore lateral feature sizes between form and roughness (so-called mid-spatial frequencies, or MSFs), however, making them inadequate for many modern optics fabricated with advanced technologies. Specifications sensitive to the lateral feature size, such as such as slope or power-spectral density (PSD), are increasingly employed to fill this void. Having a detailed view of the surface error as a function of lateral feature size can drive fabrication decisions. For example, a large aperture tool tends to correct small features well but performs less well on large features; while subaperture tools tend to do the opposite. Therefore after each fabrication step we want to know how the surface features at different lateral sizes evolved, so that we can optimize the choice of the next fabrication step. A spec like PSD often doesn’t inform the fabricator whether it’s failing because of a localized error (such as edge roll), continuous texture, or an artifact of metrology or computation. So while it may indicate the need for additional fabrication steps, it is not ideal for guiding specifically which fabrication step ought to be undertaken next. We have developed analyses to help determine what surface characteristics are failing spec, and thus optimize the next fabrication step. Finally, we demonstrate an example of how we have applied these techniques to fabricate parts with demanding slope and MSF specifications.

[1]  Richard Leach,et al.  Open questions in surface topography measurement: a roadmap , 2015 .

[2]  G W Forbes Fitting freeform shapes with orthogonal bases. , 2013, Optics express.

[3]  David M. Aikens,et al.  Use of power spectral density (PSD) functions in specifying optics for the National Ignition Facility , 1995, Other Conferences.

[4]  William Kordonski,et al.  Magnetorheological finishing (MRF) in commercial precision optics manufacturing , 1999, Optics + Photonics.

[5]  Paul E. Murphy Methods and challenges in quantifying mid-spatial frequencies , 2008 .

[6]  Mark A. Henesian,et al.  National Ignition Facility wavefront requirements and optical architecture , 2004 .

[7]  David M. Aikens,et al.  Specification and Control of Mid-Spatial Frequency Wavefront Errors in Optical Systems , 2008 .

[8]  R N Youngworth,et al.  Simple Estimates for the Effects of Mid-spatial-Frequency Surface Errors on Image Quality. , 2000, Applied optics.

[9]  John M Tamkin,et al.  Effects of structured mid-spatial frequency surface errors on image performance. , 2010, Applied optics.

[10]  Mark A. Henesian,et al.  NIF optical specifications: the importance of the RMS gradient , 1998, Other Conferences.

[11]  James E. Harvey,et al.  Scattering effects from residual optical fabrication errors , 1995, Other Conferences.

[12]  Hyun Kyoung An,et al.  Parametric smoothing model for visco-elastic polishing tools. , 2010, Optics express.

[13]  Guilin Wang,et al.  Use of wavelet in specifying optics , 2007 .

[14]  Christoph Menke,et al.  Wavelet Methods for the Representation, Analysis and Simulation of Optical Surfaces , 2012 .

[15]  Bryan J. Rice,et al.  Determination of the flare specification and methods to meet the CD control requirements for the 32-nm node using EUVL , 2004, SPIE Advanced Lithography.

[16]  Alan V. Oppenheim,et al.  Discrete-time Signal Processing. Vol.2 , 2001 .