AVS2001 Session AS-MoM: Quantitative Analysis and Data Interpretation I: SIMS

Monday, October 29, 2001 9:40 AM in Room 134

Monday Morning

Start	Invited?	Item
9:40 AM		AS-MoM-1 TOF-SIMS Characterization of Additives in Polymer Materials: Influence of Primary Ion Bombardment Conditions R. Kersting, B. Hagenhoff (TASCON GmbH, Germany) Additives are an essential part of polymer formulations because they decisively influence the chemical behavior of daily polymer products. Whereas many analytical techniques exist to analyze additives in the polymer bulk, only a few techniques are able to obtain chemical information on the polymer surface, among these Time-of-Flight Secondary Ion Mass Spectrometry. Many studies have proven the general usefulness of TOF-SIMS analysis for the characterization of polymer materials and polymer additives. However, systematic investigations on the optimum analytical conditions for a sensitive detection of additives from polymer surfaces are still sparse. We therefore investigated the influence of the primary ion bombardment parameters on the emission behavior of model samples. The samples were prepared by spin coating the antioxidant Irganox 1010 in different concentrations onto additive free LDPE substrates. The secondary ion parameters yield, disappearance cross-section and secondary ion emission efficiency (yield per damaged area) were determined for primary ion bombardment with Ga⁺, Cs⁺, and SF₅⁺. The primary ion energies were varied between 4 and 10 keV for SF₅⁺ and Cs⁺ and between 5 and 25 keV for Ga⁺ bombardment. From these experiments optimum analytical conditions for the analysis of thin additive coatings on polymer materials were deduced. The results will be applied to real world polymer surfaces with respect to additive quantification and additive degradation processes.
10:00 AM		AS-MoM-2 ToF-SIMS Surface Chemical State Imaging of Biomaterials B.T. Wickes, D.G. Castner (University of Washington) New developments in molecular biology and materials science are now being applied to designing molecular specificity and recognition into the surface of biomaterials. These novel surfaces are envisioned to have a well-defined array of recognition sites designed to interact specifically with proteins and cells. The development of surface analysis techniques that will provide detailed chemical state information at high spatial resolution is required to investigate the presentation of these recognition sites on a biomaterial surface. Static ToF-SIMS has the potential to provide detailed information about the chemical surface structure of biomaterials with a spatial resolution of 1 micron. However, static ToF-SIMS images containing a full mass spectrum at each pixel can be complex to analyze. The challenge is to develop a method for determining all chemical species present in the surface region along with their location and concentration without a priori knowledge about the sample. Image analysis can be considered to be a 3 step process of denoising, component identification, and image reconstruction. A gold surface patterned with 2 micron lines of ethylene glycol thiol molecules separated by 2 micron lines of fluorinated thiol molecules was used as a model system to evaluate different imaging processing strategies. Effectiveness of various denoising methods (wavelet, boxcar, median, etc.) was evaluated in terms of removing speckle noise, preserving image features, and the time required for denoising. Principal component analysis was used to identify the combination of mass fragments that provided the best image contrast.
10:20 AM		AS-MoM-3 Application of Spectrum Imaging and Multivariate Statistical Analysis to Time-of-Flight Secondary Ion Mass Spectrometry¹ J.A. Ohlhausen, M.R. Keenan, D.E. Peebles, P.G. Kotula (Sandia National Laboratories) The combination of spectrum imaging and multivariate spectral image analysis for information extraction have recently been applied as powerful new phase mapping tools in a variety of materials characterization techniques. In this process, an image is collected by acquiring an entire spectrum for each pixel of the image. Subsequently, multivariate spectral image analysis is applied to the image data to obtain the primary component spectra from each significant phase present in the image. Finally, the distribution of each phase may be obtained from the initial image and primary component spectra to provide a detailed phase mapping of the surface composition. No subjective guesses as to the identity of phases believed to be present are required to complete the phase identification and mapping process, and much of the spectral noise is removed from the image. Instead, the robust statistical process is able to unambiguously identify all of the spectral features uniquely associated with each distinct phase throughout the image, providing information well beyond that contained in a series of traditional image maps. In this work, this new approach is applied to Time-of-Flight Secondary Ion Mass Spectrometry (TOF SIMS) images, illustrating the power and advantages of precise phase identification and imaging for a number of representative sample materials.
10:40 AM		AS-MoM-4 Multivariate Statistical Approaches for Distinguishing Between Chemical and Topographical Features in TOF-SIMS Images B.J. Tyler (University of Utah) Despite its many strengths, TOF-SIMS imaging presents the analyst with several formidable challenges. One of the most notable of these challenges is differentiating between chemical and topographical effects. The intensity of ion signals depends not only on the surface composition but also upon the surface height and inclination (topography) and the material beneath the surface (matrix). These effects can be particularly dramatic on insulating samples where topography may result in a distortion of the electric field. In many cases, the intensity variations due to the structure of the sample can obscure features associated with surface chemistry. We have been involved in quantifying the effects of topography on TOF-SIMS images and exploring multivariate statistical methods that can be used to deconvolve chemical and topographic effects. Images of surfaces with strong topographic features, including fibers, spherical particles, and trenches will be presented. The influence of these topographic features on the absolute and relative peak intensities has been explored on conducting and insulating surfaces. We have found that when images are generated by rastering the ion beam, topography can cause severe distortions in the image. For example, spikes on the surface can appear as thin stripes in the image. Particles can create field lines that result in repressed ion emission causing a halo surrounding the particles. Multivariate statistics can help reduce some but not all of these effects on the images. Results will be presented using principle components analysis and mixture models to process images with confounding chemical and topographical features.
11:00 AM	Invited	AS-MoM-5 Data Interpretation and Quantitative Analysis - A Global Approach in Static SIMS I.S. Gilmore, M.P. Seah (National Physical Laboratory, UK) Static SIMS spectra are rich in information but their complexity is an acknowledged barrier to the wider take up of the technique in industry. To identify an unknown material using static SIMS, an analyst needs to compare the measured spectrum with those available in spectral libraries. However, existing spectrometers use a wide variety of instrumental geometries, primary ion species and operating energies. Consequently, data from different laboratories differ significantly and data in handbooks and libraries are only broadly comparable. Additionally, existing libraries (1250 spectra) are small compared to those used in organic mass spectrometry (356000 spectra) and are tiny in comparison to the industrial need. A strategy of methods is required to interpret and quantify spectra of unknown materials. For those instances when reliable spectra are contained in libraries, multivariate and artificial neural network approaches can give accurate identification and quantitative information. These methods will be discussed and compared. However, if the material is not in the library as is generally the case, an entirely new method, known as G-SIMS or gentle-SIMS¹ is fruitful. This method gives relatively simple spectra whose peaks are directly related to the molecules present and their main constituents. These spectra may be interpreted without a library. These approaches may be combined to form a global strategy for the analysis of a wide range of surfaces. ¹I S Gilmore and M P Seah, Appl. Surf. Sci. 161 (2000) 465.
11:40 AM		AS-MoM-7 Secondary Ion Mass Spectrometry Analysis of Wafer Contamination Resulting from Gloved Hands W.R. Morinville, C. Blackmer (Micron Technology, Inc.) It is well known that wafer-handling protocols to avoid contamination are a mandatory part of semiconductor fabrication. One of the most significant pieces of the standard "cleanroom" suit is the glove worn to eliminate contamination from human skin, of which mobile ions from fingerprint oils are a major contributor. Once the glove is in place, it is often considered unnecessary to use appropriate tools when the wafers are to be submitted for testing or analysis. This paper will show that wafer handling, even with fabrication recommended gloves in place, still significantly contaminates the wafer. The study began with a prime Si wafer that was divided into sections for experimentation. One piece of the wafer was not touched and would become the control sample. The remaining pieces were touched with bare hands along with various gloves that were all approved for use in a fabrication environment. An intriguing observation was that fingerprint-like marks were visible in all cases where the wafer had been touched. Secondary Ion Mass Spectrometry was then utilized to analyze the fingerprinted areas. Mobile ion contamination was found to be associated with many of the gloves in the study. It was also verified that the contamination was coming from the glove itself and not diffusing through the glove from the hands.