The electrostatic comb finger drive has become an integral design for microsensor and microactuator applications. This paper reports on utilizing the levitation effect of comb fingers to design vertical-to-the-substrate actuation for interferometric applications. For typical polysilicon comb drives with 2 mu m gaps between the stationary and moving fingers, as well as between the microstructures and the substrate, the equilibrium position is nominally 1-2 mu m above the stationary comb fingers. This distance is ideal for many phase shifting interferometric applications. Theoretical calculations of the vertical actuation characteristics are compared with the experimental results, and a general design guideline is derived from these results. The suspension flexure stiffnesses, gravity forces, squeeze film damping, and comb finger thicknesses are parameters investigated which affect the displacement curve of the vertical microactuator. By designing a parallel plate capacitor between the suspended mass and the substrate, in situ position sensing can be used to control the vertical movement, providing a total feedback-controlled system. Fundamentals of various capacitive position sensing techniques are discussed. Experimental verification is carried out by a Zygo distance measurement interferometer.

As EUV lithography progresses from laboratory research to prototype development the realistic performance of manufacturable components becomes a primary concern. Nowhere is this more evident than in the fabrication and implementation of the EUV imaging optics. It is now well understood that the structure of the optical surfaces and the multilayer coatings (ML) that make the surfaces reflective at soft x-ray wavelengths must be specified and fabricated with unprecedented accuracy. Errors in the structure, which include deviations in the surface profile of the substrate and unintentional variations in the multilayer period, cause aberations in the imaging process. When these errors are at very long spatial wavelengths they are treated deterministically, and can be evaluated (and hence corrected) using interferometric methods now under development. However, there are errors in the surface profile at all spatial frequencies. Describing the exact structure of the surfaces of the optics at all spatial scalelengths is an intractable problem. Instead the surface figure errors at mid- and high-spatial frequency (called “roughness” or “finish”) are treated statistically. Within this statistical description, the effect of surface (and multilayer) roughness is to remove intensity from the image (the specular field) and scatter it throughout the image field. This nonspecular scattering is problematic for two reasons: (1) it decreases the useful throughput of the optical system and, (2) it produces a backround halo which reduces the contrast of the image. In this paper we describe a method of relating the nonspecular scattering to the roughness of the optical surfaces in a distributed EUVL imaging system. Our ultimate goal is to develop a robust specification of surface finish that can be used as the guideline for manufacturing EUV optics.

The performance of an Extreme Ultraviolet Lithography (EUVL) imaging optic was characterized by printing resolution test images in resist. While features as small as 0.137µm were successfuly printed, a resolution of 0.175µm better represents the performance of the system over the full 0.9mm2 image field. The contrast of the aerial image was estimated to be approximately 40% or less for the fine features printed. This low contrast value is attributed to a degradation of the modulation transfer function (MTF) due to the presence of scattered light in the image.

Point diffraction interferometry is a common-path technique that is suitable for at-wavelength testing of EUV lithography optics. We have constructed and successfully tested such a device at the Advanced Light Source at Lawrence Berkeley Laboratory. Preliminary results on the characterization of the wavefront produced by a Fresnel zone plate lens are presented. Reference wavefront quality and thermal effects are also discussed.