Scatterometry is emerging as a prominent metrology technique for lithography. Not only does scatterometry produce
line profile information such as sidewall angle and height along with line width, but the speed and nondestructive
nature of scatterometry accommodates in-line process applications. Scatterometry systems employ reflectometry or
ellipsometry to acquire spectra resulting from the interaction of the input radiation and a symmetrical grating array.
The systems may use fixed wavelengths or a range of wavelengths. The output spectral data is dependent on the
material and physical properties of the grating array and surrounding (subsurface, film stack) material layers. Typical
scatterometry draws on mathematically modeled spectra from known optical and physical parameters such as the
grating pitch and the index of refraction and absorption coefficient functions of the film stack materials. The optical
properties of the materials in the film stack are of particular interest and critical to scatterometry. Material vendors
typically supply constants associated with the optical dispersion models of resists and anti-reflective coatings used in
lithography. These constants are most often based on a Cauchy model for optical dispersion, a very simple model.
However, the optical properties of the photoresist or other coatings may not fit well to a Cauchy model or they may
change during process baking, exposure or just from aging. To make an accurate scatterometry model for patterned
photoresist, the material characteristics must also be modeled. Using these parameters, an accurate picture of the
lithographic materials can be generated. These methods can be applied to both dry and immersion lithography.
As immersion lithography gains a foothold in the manufacturing line, many initial processes will use standard dry
photoresist with the application of an immersion topcoat to protect the final lens element of the lithography tool, and
to reduce defects formed from substances leaching out of the photoresist. Although the goal for an immersion topcoat
is to be neutral to the resist process in terms of profiles, process windows, and CD control, many topcoats are not
completely benign. Topcoat induced resist thinning is a common but unwelcome attribute. In this paper we discuss the
use of scatterometry to characterize topcoat induced thickness changes, and use this technique to evaluate several
commercially available products. We will also demonstrate the ability of scatterometry to accurately determine resist
profile changes as a result of focal changes, topcoat interactions, and airborne contamination. Measurement stability
results are also shown, and correlation to CD-SEM and cross-section SEM are provided as a reference metrology.
|