Daniel J Graham oral presentation (FN2-Tue2-1-1)
A Sample and A Sputter Beam Enter a Bar...Dealing with sputter rate differences is no joke
University of Washington - Department of Bioengineering, 3946 W. Stevens Way NE, WA 98195 Seattle, United States
Cluster ion beams have opened up new opportunities for time-of-flight secondary ion mass spectrometry (ToF-SIMS) depth profiling. Argon cluster sources have been shown to sputter most all organic materials with minimal damage and excellent depth resolution. While depth profiling single component systems is fairly straight forward, working with multicomponent materials significantly increases the complexity of the system. It has been seen that most materials sputter at different rates. This can result in significantly distorted 3D depth profiles if one does not account for the various material sputter rates when reconstructing the data. This can be fairly straight forward in a simple layered system, however properly adjusting for sputter rates with different materials in more complex geometries becomes problematic. Good results have been obtained using external calibration methods such as AFM to track the sample topography and then adjust the SIMS data to fit the measured shape of the sample. Unfortunately this experiment is not trivial and requires precise alignment of samples and image registration. This process also scales the data uniformly, which may not be accurate if materials with different sputter rates are located on top of each other. In this work we present results of an alternate methodology which generates an estimate of the 3D reconstructed data by accounting for materials with different sputter rates by identifying the components with in the system and applying a sputter rate correction factor during the data reconstruction. Results will be presented from a brute force correction of data from a simple layered polymer system and from using multivariate analysis to identify different components in a layered polymer system with a controlled pattern at the interface, and then adjusting the 3D reconstruction to account for the varying sputter rates for each component. These results present an engineering approach to a best estimate of the 3D profile of multicomponent systems.