AVS1997 Session AS-TuA: Practical Aspects of Quantification, Data Handling and Standards

Tuesday, October 21, 1997 2:00 PM in Room J2

Tuesday Afternoon

Start	Invited?	Item
2:00 PM	Invited	AS-TuA-1 Practical Quantification in Auger Electron Spectroscopy H.S. Wildman (IBM) The value of applying AES and XPS to semiconductor process or material problems is enhanced if results are reduced to relevant quantitative values such as elemental or chemical composition and thickness. Requirements for such analyses, in order of priority, are timeliness, repeatability, reproducibility, and accuracy. In a typical progression a set of samples is first analyzed to determine the key differences. The measurement of a critical parameter may then be needed at later times, on other instruments, or at other locations, possibly even as a process monitor. For original significance, the repeatability of a measurement depends on its signal to noise. For later process control, the reproducibility depends on instrument calibration and attention to procedure. For proper understanding, the accuracy depends first on the theoretical model and algorithms relating the measurements to physical and chemical properties of the samples, and then on the measurement calibration. When the word "practical" is added to quantitative analysis it often alludes to a real gap between the state of the art possibilities and what is actually done. The gap is due partly to inadequate instrumentation and standards, partly to the skill of the analyst, but mostly to the requirement for timeliness. The gap can be be reduced by incorporating state of the art knowledge into computer automation of data acquisition and data analysis procedures. This paper explores the practical approaches to achieving repeatable quantification. The method of constant relative sensitivity factors will be discussed along with the illusive iterative correction algorithm.
2:40 PM		AS-TuA-3 Quantitative AES: Auger Electron Lineshapes and Intensity for Metals and their Oxides. M.P. Seah, I.S. Gilmore, H.E. Bishop (National Physical Laboratory, United Kingdom); G. Lorang (CNRS, France) Reference Auger electron spectra for 23 known stoichiometric oxides and their parent metals have been recorded as part of a high resolution digital Auger electron database (1). These spectra are at high energy-resolution and have all instrumental contributions removed. For quantitative analysis, peak areas from the spectra may be used after removal of the backgrounds arising from the scattered primary electrons, the inelastically scattered Auger electrons and the secondary electron cascade. For practical work this involves some effort and analysts would prefer to use the differential peak-to-peak height. However, for a given peak area, the differential peak-to-peak intensities of the oxygen Auger electron peaks vary over a factor of 3 (a standard deviation of 30%) for the different oxides. This means that relative sensitivity factors would have to be unique numbers for each oxide for this traditional differential mode. A new differential method is devised, using spectral broadening, which is simple in concept and implementation, which reduces the standard deviation of this scatter to 5%. This method will be explained and illustrated by reference data. The greater precision in measurement allows us to show further that, whilst the LMM peak branching ratios for L_2,3M_2,3M_2,3, L_2,3M_2,3M_4,5 and L_2,3M_4,5M_4,5, for the elements V to Cu, agree with theory, upon oxidation there are small influences of intra-atomic nature in the M_4,5 level which alter these ratios. A simple explanation for this influence by provided which indicates an effective limit to the accuracy of Auger electron measurements made using a single peak. ¹M P Seah and I S Gilmore, J. Vac. Sci. Technol. A14, 1401 (1996).
3:00 PM		AS-TuA-4 Multivariate Image Processing Applications in Surface Analysis R.E. Ericson, S.G. MacKay (3M) Imaging is an important data collection tool in several areas of analytical chemistry, including AES and SIMS surface analysis. In these techniques, images are used to show the spatial distribution of elements or molecular fragments on a surface. A set of species-specific images from the same area of a surface also contains spectral information that relates pixels at the same location in different images. Spectral relationships, some of which may not be apparent from a study of the original images, may be extracted using multivariate statistical analysis. The resulting transformed images have both a structural part with possibly fewer dimensions than the original set, plus a residual part which is mostly noise. Multidimensional scatter diagrams may be used to further refine the data by interactive correlation partitioning (ICP). The end result is a reduced set of low noise principal component images showing phase relationships and containing fewer artifacts than the original images. In AES, multivariate image analysis is a fairly straightforward process because there are often only a half dozen or so acquired images and the spectral relationships are usually simple. In SIMS, spectral relationships are more complex (particularly for organic materials), and many more acquired images must be included in the multivariate image. Data pretreatment is more complicated and computational constraints often necessitate the use of sampling techniques. Emphasis in this presentation will be on applications in AES and TOF-SIMS. The extraction of surface phases, the removal of artifacts and noise, and the use of ICP to refine the images will be illustrated using examples. Implementation issues will be addressed, including problems commonly encountered in data pretreatment and partitioning.
3:20 PM	Invited	AS-TuA-5 Reference Films and Reference Surfaces for Electron Spectroscopy S.W. Gaarenstroom (General Motors R&D Center) The electron spectroscopist requires reference films and reference surfaces for a number of tasks. The most fundamental is to verify the operating performance of the spectrometer, in particular the energy scale and the signal strengths. When sputter depth profiling is performed, the depth scale, depth resolution, and ion beam uniformity can be obtained from measurements on reference films. Finally surfaces can be chemically characterized by comparing the spectra of the surface of interest to the spectra from reference surfaces or films. In this presentation, the current laboratory practices are reviewed and predictions are made for future practices from a user's perspective.
4:00 PM		AS-TuA-7 Tests, Test Specimens and Instrument Performance D.R. Baer, M.H. Engelhard, Y. Liang, A.S. Lea, M.A. Henderson (Pacific Northwest National Laboratory) As part of the setup and operation of a new US Department of Energy user facility, a series of instrument performance checks, calibration schedules and "test" specimens are being used to set up and monitor performance of several surface analysis instruments. Overall capabilities include Auger electron spectroscopy, x-ray photoelectron spectroscopy, and secondary ion mass spectrometry (time of flight and quadrapole based). The performance checks determine how well the instruments perform with time relative to initial instrument evaluations and compare initial set up in the laboratory to operation by newly trained staff. The tests and test specimens check on many features including instrument response functions, spatial resolution, energy resolution, sputter rates and overall instrument performance. Test specimens include a magnification and analysis standard, a multilayer sputter standard and a radiation sensitive insulating specimen. Where possible we have used tests and specimens developed, tested or manufactured by others. Information regarding the tests, test specimens and performance records for four new instruments will be presented. ¹PNNL is a multiprogram laboratory operated by the US Department of Energy by Battelle Memorial Institute under contract DE-AC06-76RLO180.
4:20 PM		AS-TuA-8 Studies of Neutral Particle Distribution in the Laser Desorbed Plumes for Rapid and Quantitative Surface Analysis C. He, C. Becker (SRI International) Recently, we demonstrated direct and complete quantitation of surface elemental analysis (uniform surface analysis) under high sensitivity conditions [1,2]. This was achieved by detecting neutral species ejected from keV ion sputtered analyte surface, followed by high intensity laser photoionization mass spectrometry. To meet the demand for rapid, uniform surface analysis and extended surface area mapping, and to develop an analytical technique for intact nonvolatile molecular analysis, we studied laser ionization mass spectrometry for sensitive and uniform surface analysis featuring laser desorption. We examined quantitatively the flux distributions of species in the desorption plume as a function of laser power density, wavelength, and time delays between desorption and ionization lasers. The concentration distributions of elements in the desorption plume are described by relative desorption factors (RDFs). The experiment results of these studies are presented.
4:40 PM		AS-TuA-9 Correction of Surface Roughness Measurements in SPM Imaging S. Dongmo, P. Vautrot, N. Bonnet, M. Troyon (University of Reims, France) In STM and AFM, when rough surfaces are concerned, the image formation is essentially governed by the non-linear geometrical interaction between the specimen surface and the tip surface. The shape and finite size of the tip are responsible for imperfections of images. Consequently the measured roughnesses are not exact and are always smaller than the true roughness value since the surface is imperfectly probed. We propose a method to predict the true roughness value from the experimental image obtained. For this purpose a large variety of different 3D surfaces which mimic the experimental ones are computed and classified with respect to the shape and density of grains composing the surface. Then by using the mathematical morphology dilation operation which describes the image formation in SPM the corresponding dilated images are computed for different tip geometries. For each category of surfaces and a given tip shape the correction factors between true rms roughness values and experimental ones are obtained allowing one to predict the true roughness value.