Quantification of Dopants in Silicon using Atom Probe Tomography

The ever-increasing demand for semiconductor devices to shrink into the 3 nm regime without compromising computing power has made characterizing these materials incredibly challenging. In order to maintain device performance, there is a growing need to characterize these materials at device dimensions and accurately quantify and localize dopant atoms [1].<br/><br/>Atom probe tomography (APT) is currently the lone characterization technique with the ability to provide dopant distributions in semiconductor devices at device dimensions and in 3-dimensions. While APT can measure chemical information down to concentrations of 10’s of µg/g (ppm_at), it has also been shown that, for some materials, analysis conditions can result in biased quantitative results [2, 3]. While the use of reference materials to calibrate data analysis is common in many characterization techniques, it is not generally used in APT. However, it has been shown that a standards-based approach to APT analysis can dramatically improve the accuracy of composition measurements [4, 5]. More recently, our work with phosphorous-doped silicon showed similar improvement in the accuracy of dopant quantification in the atom probe. In this work, we used a reference material with a known dose of phosphorous (NIST SRM 2133 [6]). We collected data from this material under a variety of analysis conditions and used that data to construct a calibration curve. We then performed similar analysis on a sample doped with a known concentration of phosphorous. After correcting our measurements using the calibration curve, reduced the relative error from 26% to less than 4% [7].<br/><br/>The reference material used for this work (NIST SRM 2133) was originally developed for use in secondary ion mass spectrometry (SIMS) and has a phosphorous concentration that varies with depth, with the total retained dose being the certified value. This required us to capture the entire doped profile in an atom probe experiment in order to use the results for our calibration curve. However, this introduces complexity both in sample preparation, and in the collection and analysis of the atom probe data. Therefore, we are also working to develop a research grade test material (RGTM) that is better suited for use with atom probe tomography. This material has a blanket film of P-doped Si at a constant composition, eliminating the need to preserve the precise sample surface during FIB sample preparation and speeding up acquisition times, as an entire doped profile no longer needs to be captured [8].<br/><br/>References:<br/>1. Fuechsle, M., et al., Nature Nanotechnology, 2012. <b>7</b>(4): p. 242-246.<br/>2. Morris, R.J.H., et al., Journal of Vacuum Science & Technology B, 2018. <b>36</b>(3): p. 03F130.<br/>3. Meisenkothen, F., et al., Ultramicroscopy, 2015. <b>159</b>: p. 101-111.<br/>4. Gopon, P., et al., Microscopy and Microanalysis, 2022. <b>28</b>(4): p. 1127-1140.<br/>5. Meisenkothen, F., et al., Microscopy and Microanalysis, 2020. <b>26</b>(S2): p. 176-177.<br/>6. Simons, D.S., et al., Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena, 2007. <b>25</b>(4): p. 1365-1375.<br/>7. DeRocher, K., M. McLean, and F. Meisenkothen, Microscopy and Microanalysis, 2022. <b>28</b>(S1): p. 728-729.<br/>8. Certain commercial equipment, instruments, or materials are identified in this paper in<br/>order to specify the experimental procedure adequately. Such identification is not<br/>intended to imply recommendation or endorsement by the National Institute of Standards and Technology, nor is it intended to imply that the materials or equipment identified are necessarily the best available for the purpose.