DOC PREVIEW
UW-Madison G 777 - Tomographic Spectral Imaging with Multivariate Statistical Analysis: Comprehensive 3D Microanalysis

This preview shows page 1-2-3-4 out of 13 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 13 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 13 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 13 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 13 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 13 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Tomographic Spectral Imaging with MultivariateStatistical Analysis: Comprehensive 3D MicroanalysisPaul G. Kotula,* Michael R. Keenan, and Joseph R. MichaelSandia National Laboratories, Materials Characterization Department, P.O. Box 5800,MS0886, Albuquerque, NM 87185-0886, USAAbstract: A comprehensive three-dimensional ~3D! microanalysis procedure using a combined scanningelectron microscope ~SEM!/focused ion beam ~FIB! system equipped with an energy-dispersive X-ray spectrom-eter ~EDS! has been developed. The FIB system was used first to prepare a site-specific region for X-raymicroanalysis followed by the acquisition of an electron-beam generated X-ray spectr al image. A small sectionof material was then removed by the FIB, followed by the acquisition of another X-ray spectral image. Thisserial sectioning procedure was repeated 10–12 times to sample a volume of material. The series of two-spatial-dimension spectral images were then concatenated into a single data set consisting of a series of volumeelements or voxels each with an entire X-ray spectrum. This four-dimensional ~three real space and one spectraldimension! spectral image was then comprehensively analyzed with Sandia’s automated X-ray spectral imageanalysis software. This technique was applied to a simple Cu-Ag eutectic and a more complicated localizedcorrosion study where the powerful site-specific comprehensive analysis capability of tomographic spectralimaging ~TSI! combined with multivariate statistical analysis is demonstrated.Key words: tomography, spectral imaging, multivariate statistical analysis, multivariate curve resolution, 3Dchemical analysis, tomographic spectral imaging, serial sectioning, 3D microanalysisINTRODUCTIONChemical analysis is typically performed at points, lines, orover areas in images. This would include single-spectrumacquisitions, line profiles ~or spectrum lines!, chemical maps~Cosslett & Duncumb, 1956!, or spectral images ~an imagewhere each pixel contains an entire spectrum!~Legge &Hammond, 1979!. Recently, however, interest has been grow-ing in extending microanalysis to the third spatial dimen-sion through three-dimensional ~3D! at om probe techniques~see, e.g., references in Miller, 1997! or various direct tomo-graphic ~Patkin & Morrison, 1982; Rüdenauer, 1982, 1993;Sharonov et al., 1994; Marschallinger, 1998; Saadi et al.,1998; Dunn & Hull, 1999; Takanashi et al., 2000; Hull et al.,2001; Dunn et al., 2002; Vekemans et al., 2004! or computedtomographic approaches ~Schofield & Lefevre, 1992;Schofield, 1995; Möbus & Inkson, 2001; Midgley & Wey-land, 2003; Möbus et al., 2003!. Each of the techniquesreferenced above has a combination of useful analyticalsignal, specimen preparation requirements/limitations, andrelevant resolution/total volume sampled as well as vari-ous degrees of experimental and computational complexity.Table 1 shows some of the direct tomographic methods andtheir relevant resolutions. With the exception of atom-probe techniques, in which each atom from the specimen is,in principle, identified, and the more recent 3D confocalX-ray fluorescence analysis ~Vekemans et al., 2004!, onlyrudimentary acquisition and analysis techniques were ap-plied. For example, elemental maps ~spectroscopic images!were acquired from known elements and then rendered in3D images ~Marschallinger, 1998!. Spectral imaging ap-proaches ~full spectra from each spatial element! have tradi-tionally been limited to smaller numbers of spectra andcovering lines or areas. This was due primarily to the lack ofthe ability to acquire the data in three dimensions or thelack of computational ability to analyze the data ~e.g.,reconstr uct spectra from points/regions, map chemical sig-nals, or perform more sophisticated data analyses!.Inactu-ality, methods for the acquisition of extremely large spectralimages are commercially available and their comprehensiveand unbiased analysis, based on multivariate statistical analy-sis ~MSA!, has been developed ~Kotula & Keenan, 2003;Kotula et al., 2003a! including for 3D spectral images ~Kotulaet al., 2003b, 2004!. In contrast, the method described byVekemans ~Vekemans et al., 2004! for the analysis of 3Dconfocal micro X-ray fluorescence spectral images presumesall of the peak shapes are known a priori so that theproblem can be reduced from thousands of var iables ~chan-nels! to 10 or fewer variables ~now elements! prior toReceived February 5, 2005; accepted October 18, 2005.*Corresponding author. E-mail: [email protected]. Microanal. 12, 36–48, 2006DOI: 10.1017/S1431927606060193MicroscopyANDMicroanalysis© MICROSCOPY SOCIETY OF AMERICA 2006multivariate statistical analysis ~principal components analy-sis! and clustering ~k-means!.In this work, we demonstrate an analytical geometryfor tomographic or three-spatial-dimension spectr al imag-ing using a combined focused ion beam ~FIB!/scanningelectron microscope ~SEM!–energy-dispersive X-ray spec-trometer ~EDS! system with the FIB for serial sectioningand the SEM-EDS for X-ray spectral imaging. We furtherdemonstrate the use of MSA methods for rapid andcomprehensive analysis of the very large resultant four-dimensional ~4D! data sets. The techniques described hereinare more generally applicable to other tomographic methods~e.g., metallography, microtomy, tilt-series reconstructions,etc.! as well other analytical techniques ~e.g., time-of-flightsecondary-ion mass spectrometry ~TOF -SIMS!, X-ray phot o-electron spectroscopy ~XPS!, X-ray fluorescence ~XRF!,particle-induced X-ray emission ~PIXE!,etc.! where tomo-graphic spectral imaging ~TSI! data is or could be acquired,and peak identities, shapes ~including families of peaks!,positions, and relative intensities may not be known a priori.MATERIALS AND METHODSThe method used here for sectioning the specimen is simi-lar to that described by Sakamoto et al. ~1998! and Dunnand Hull ~1999!. Ser ial sectioning was performed with a FEIDB-235, FIB/SEM, equipped with an ultrathin window EDScontrolled by a Thermo NORAN Vantage Digital Imagingwith Spectral Imaging system. The ion column is cofocalwith the field emission SEM and at an angle of 528 withrespect to the same. The EDS has a take-off angle of 358 andis at a 458 azimuthal angle with respect to the plane of theion and electron columns. For the serial sectioning, thesample, initially untilted with respect to the elect ron beam,is tilted 528 toward the ion column. The ion beam is thennormal to the sample


View Full Document

UW-Madison G 777 - Tomographic Spectral Imaging with Multivariate Statistical Analysis: Comprehensive 3D Microanalysis

Documents in this Course
Load more
Download Tomographic Spectral Imaging with Multivariate Statistical Analysis: Comprehensive 3D Microanalysis
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Tomographic Spectral Imaging with Multivariate Statistical Analysis: Comprehensive 3D Microanalysis and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Tomographic Spectral Imaging with Multivariate Statistical Analysis: Comprehensive 3D Microanalysis 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?