Auger Electron Spectrometer (AES)

1. Theory

• Definition: An Auger electron spectrometer excites a surface with a focused electron beam and measures the kinetic energies of Auger electrons emitted when core holes relax non-radiatively, giving elemental and some chemical information from the top few nanometres (Diplas; Semitracks).
• Auger process: A core hole in level W is filled by an electron from X, and the released energy ejects an electron from Y, so $E_{WXY}\approx E_W-E_X-E_Y$ independent of beam energy (Diplas; Caltech lecture).
• Competition with X-rays: Auger emission and X-ray fluorescence share each core hole, and Auger dominates for light elements (background knowledge, Briggs and Seah).
• Surface sensitivity: Auger electrons below 2000 $\mathrm{eV}$ escape only from the top 5 $\mathrm{nm}$ or so (Semitracks).

2. Working Principle

• Excitation: A field-emission electron gun at about 2 to 30 $\mathrm{keV}$ forms a nanometre-scale probe in ultrahigh vacuum (Diplas; Semitracks).
• Analysis: A cylindrical-mirror or hemispherical analyser records counts or the derivative spectrum against energy (Diplas).
• Mapping: Scanning the beam gives Auger maps alongside secondary-electron images (Diplas).
• Profiling: Ar ion sputtering with sample rotation gives composition against depth (Diplas; Semitracks).

3. Applications

• Failure analysis: Surface contamination, small particles and pad corrosion (Semitracks).
• Thin films and interfaces: Depth profiles of metal stacks, silicides and coatings (Diplas; Semitracks).
• Grain boundaries and metallurgy: Segregation on fracture surfaces (Diplas).
• Semiconductor surfaces: Native oxide, nitride and residue identification (Diplas; background knowledge).

Auger Process

Name	Formula	Related Formulae	Variables/Constants
Auger electron energy	$E_{K{L}_{2,3}{L}_{2,3}}\approx E_K-E_{L_{2,3}}-E_{L_{2,3}}$	$E_{WXY}\approx E_W\left(Z\right)-E_X\left(Z\right)-E_Y\left(Z+\Delta \right)-\varphi$	$\begin{array}{r}{E}_{WXY}:kineticenergyoftheWXYAuger\\ electron\end{array}$ $E_W:initialcore-holelevel$ $E_X:levelofthefillingelectron$ $E_Y:leveloftheemittedelectron$ $\begin{array}{r}\Delta :correction0.5to1.5fortheextra\\ hole\end{array}$ $\begin{array}{r}\varphi :workfunction;Diplas:firstform;\\ Caltechlecture:Z+\Delta correctionwith\Delta \\ from0.5to1.5andaveragedform;Si\\ levelsK1839,L_{1}149.7and{L}_{2,3}99.8\mathrm{eV},P\\ L_{2,3}about136\mathrm{eV}(backgroundknowledge,\\ BriggsandSeah)\end{array}$ $\begin{array}{r}(computed:SiK{L}_{2,3}{L}_{2,3}is1639\mathrm{eV}\\ uncorrected,1603\mathrm{eV}with\Delta =1andabout\\ 1621\mathrm{eV}with\Delta =0.5)\end{array}$
Auger against fluorescence yield	${\omega }_A+{\omega }_X=1$	${\omega }_X\approx \frac{Z^4}{a+Z^4},a\approx 10^6$	${\omega }_A:Augeryield$ $\begin{array}{r}{\omega }_X:X-rayfluorescenceyield;\\ complementarity(backgroundknowledge,\\ BriggsandSeah);XRFpage:approximate\\ yieldformulawith{\omega }_K=0.037forSiand\\ 0.31forFe\end{array}$ $\begin{array}{r}(computed:{\omega }_{A}is0.96forSiKand0.69for\\ FeK,soAESisfavouredforlight\\ elements)\end{array}$
Chemical shift	$\Delta E_A\approx \Delta E_1-2\Delta E_2$	$\text{e.g. predicted 5 eV against measured 4.7 eV}$	$\Delta E_A:Augerchemicalshift$ $\begin{array}{r}\Delta E_1,\Delta E_2:core-levelshiftsofthe\\ initialandfinallevels;Caltechlecture:\\ relationandthethiosulphateexample;\\ Diplas:Si,Si{N}_{x}andSi{O}_{2}shiftcomparison\\ slide\end{array}$ $\begin{array}{r}(computed:thepredictedvalueis6per\\ centaboveexperiment)\end{array}$

Excitation, Depth and Resolution

Name	Formula	Related Formulae	Variables/Constants
Escape depth and surface sensitivity	$d_{info}\approx 3\lambda$	$\lambda \approx 1\text{to}5\mathrm{nm}(\text{50 to 2000 eV electrons})$	$\begin{array}{r}\lambda :inelasticmeanfreepathoftheAuger\\ electron;Semitracks:meanfreepath10to\\ 50\mathring{\mathrm{A}},electronsescapefromthetop50\mathring{\mathrm{A}},\\ Augerenergiesbelow2000\mathrm{eV};Diplas:\\ analysisdepthafewnm;3\lambda rule\\ (backgroundknowledge,BriggsandSeah)\end{array}$ $\begin{array}{r}(computed:\lambda =1to5\mathrm{nm}gives3to15\mathrm{nm}\\ for95percentofthesignal)\end{array}$
Ionization and overvoltage	$U=\frac{E_0}{E_W}$	$\sigma \left(U\right)\text{maximal near}U\approx 3\text{to}4$	$U:overvoltageratio$ $E_0:primarybeamenergy$ $E_W:corebindingenergy$ $\begin{array}{r}\sigma :ionizationcrosssection;Semitracks:5\\ to30\mathrm{keV}beams,vacuumbelow5\times 10^{-9}\mathrm{torr};\\ Diplas:2\mathrm{kV}forAugermapping;\\ overvoltagebehaviour(background\\ knowledge,Goldsteinetal.)\end{array}$ $\begin{array}{r}(computed:SiKat1.839\mathrm{keV}witha10\mathrm{keV}\\ beamgivesU=5.4;SiLat0.0998\mathrm{keV}\\ with2\mathrm{keV}givesU=20)\end{array}$
Lateral resolution	$d_{AES}\approx \sqrt{d_{beam}^2+d_{BSE}^2}$	$d_{AES}\ll d_{EDS}$	$d_{AES}:Augerlateralresolution$ $d_{beam}:probediameter$ $\begin{array}{r}{d}_{BSE}:spreadfrombackscatteredelectrons\\ generatingAugerelectronsnearthe\\ surface;Diplas:3\mathrm{nm}SEIresolution,8\mathrm{nm}\\ Augerprobe,lateralresolution0.1\mu \mathrm{m}at2\\ \mathrm{kV}against4\mu \mathrm{m}at20\mathrm{kV}forX-rays;\\ Semitracks:150\mathring{\mathrm{A}},uptotentimesbetter\\ thanEDS,field-emissionsubmicron;\\ Caltechlecture:about50\mathrm{nm}(seethe\\ Q/A);quadratureestimate(background\\ knowledge,Goldsteinetal.)\end{array}$ $(computed:4\mu \mathrm{m}over0.1\mu \mathrm{m}is40times)$

Quantification and Profiling

Name	Formula	Related Formulae	Variables/Constants
Quantification with sensitivity factors	$N_A=\frac{I_A}{I_A+F_{AB}{I}_B+F_{AC}{I}_C+\dots }$	$F_{AB}=\frac{I_A/N_A}{I_B/N_B}$	$N_A:atomicfractionofA$ $I:peakintensity$ $\begin{array}{r}{F}_{AB}:relativesensitivityfrombinary\\ standards;Diplas:theserelations,highly\\ simplifiedbutreasonable;Semitracks:\\ relativesensitivityfactorsat3\mathrm{kV},\\ accuracyabout50percentwithoutand10\\ percentwithstandards\end{array}$ $\begin{array}{r}(assumed:I_B/I_A=0.8and{F}_{AB}=1.25;\\ computed:N_A=0.50)\end{array}$
Detection limit	$C_{min}\approx 0.1\text{to}0.5\text{at. per cent}$	$\text{Li to U detectable; H and He not}$	$\begin{array}{r}{C}_{min}:detectionlimit;Semitracks:about0.1\\ percentidentificationsensitivityand\\ below0.5atomicpercent,LithroughU;\\ Caltechlecture:0.1atomicpercent;\\ Diplas:notstated\end{array}$ $\begin{array}{r}(computed:0.1percentofa2\mathrm{nm}layerof\\ Si(5\times 10^{22}\mathrm{c}{\mathrm{m}}^{-3})is1\times 10^{13}atomspersquare\\ centimetre)\end{array}$
Sputter depth profiling	$z=\dot{z}{t}_{sp}$	$\text{Zalar rotation reduces differential sputtering}$	$z:depth$ $\dot{z}:sputterrate$ $\begin{array}{r}{t}_{sp}:sputtertime;Diplas:Ariongun,\\ automatedprofiling,Zalarrotationand\\ angle-resolvedanalysis;Semitracks:Ar\\ sputterprofiling,differentialsputtering\\ reducedbyZalarrotation,radiation\\ damageannealablebylow-temperaturebake,\\ Si3N4difficulttoanalyse\end{array}$ $\begin{array}{r}(assumed:5\mathrm{nm}perminute;computed:a100\\ \mathrm{nm}stacktakes20minutes)\end{array}$

Important Questions

Ques: How is the energy of an Auger electron determined?

By three atomic levels, not the beam energy. Diplas writes $E_{K{L}_{2,3}{L}_{2,3}}\approx E_K-E_{L_{2,3}}-E_{L_{2,3}}$. For silicon, with K at 1839 $\mathrm{eV}$ and $L_{2,3}$ at 99.8 $\mathrm{eV}$ (background knowledge, Briggs and Seah), this gives 1639 $\mathrm{eV}$ (computed). The Caltech lecture corrects the last level for the extra hole by evaluating it at $Z+\Delta$ with $\Delta$ between 0.5 and 1.5; using the phosphorus $L_{2,3}$ level near 136 $\mathrm{eV}$, $\Delta$ = 1 gives 1603 $\mathrm{eV}$ and $\Delta$ = 0.5 about 1621 $\mathrm{eV}$ (computed). Because the energy is fixed by the atom, AES can use any beam energy above the ionization threshold. Chemical shifts follow $\Delta E_1-2\Delta E_2$, predicting 5 $\mathrm{eV}$ against a measured 4.7 $\mathrm{eV}$ in the Caltech example. Reference: S. Diplas, ‘An Introduction to Auger Electron Spectroscopy’, MENA 3100 slides, SINTEF and University of Oslo (2017) (fetch-tool summary); D. Briggs and M. P. Seah, Practical Surface Analysis, Vol. 1, 2nd ed., Wiley 1990, and J. I. Goldstein et al., Scanning Electron Microscopy and X-Ray Microanalysis, 4th ed., Springer 2018 (background knowledge; not opened); ‘Lecture 20: Auger Electron Spectroscopy’, Electron Spectroscopies course, Caltech MMRC (2007) (fetch-tool summary; symbols garbled).

Ques: Why is AES better than XRF or EDS for light elements?

Because each core hole relaxes by either Auger emission or fluorescence, ${\omega }_A+{\omega }_X=1$ (background knowledge, Briggs and Seah). The XRF page estimates ${\omega }_X$ of 0.037 for Si K and 0.31 for Fe K, so Auger emission takes 96 per cent of Si K holes and 69 per cent of Fe K holes (computed). Light elements therefore give strong Auger signals but weak X-rays. Semitracks adds that AES covers lithium through uranium and has up to ten times better lateral resolution than EDS, because Auger electrons escape only from near the beam entry point, while X-rays come from the whole interaction volume: Diplas compares 0.1 $\mu \mathrm{m}$ for Auger at 2 $\mathrm{kV}$ with 4 $\mu \mathrm{m}$ at 20 $\mathrm{kV}$, a factor of 40 (computed). Reference: D. Briggs and M. P. Seah, Practical Surface Analysis, Vol. 1, 2nd ed., Wiley 1990, and J. I. Goldstein et al., Scanning Electron Microscopy and X-Ray Microanalysis, 4th ed., Springer 2018 (background knowledge; not opened); X-Ray Fluorescence Spectrometer (XRF and TXRF) page of this vault; Semitracks, ‘Auger Electron Spectroscopy’ failure-analysis reference page (fetch-tool summary; current equation shown as an image); S. Diplas, ‘An Introduction to Auger Electron Spectroscopy’, MENA 3100 slides, SINTEF and University of Oslo (2017) (fetch-tool summary).

Ques: Is AES lateral resolution 8 nm, 15 nm or 50 nm?

The sources quote different quantities. Diplas gives an 8 $\mathrm{nm}$ probe diameter for Auger analysis with 3 $\mathrm{nm}$ secondary-electron resolution, Semitracks cites 150 $\mathring{\mathrm{A}}$, and the Caltech lecture about 50 $\mathrm{nm}$. The probe size is not the analytical resolution: backscattered electrons leaving the sample generate Auger electrons over a wider area, so $d_{AES}\approx \sqrt{d_{beam}^2+d_{BSE}^2}$ (background knowledge, Goldstein et al.). Values of 10 to 20 $\mathrm{nm}$ are therefore more likely for real elemental maps on a modern field-emission instrument, with 8 $\mathrm{nm}$ as the probe limit and 50 $\mathrm{nm}$ reflecting older or thermionic systems (my reading). Lower beam energies shrink the backscatter spread but reduce ionization for deep levels, a trade-off set by the overvoltage. Reference: S. Diplas, ‘An Introduction to Auger Electron Spectroscopy’, MENA 3100 slides, SINTEF and University of Oslo (2017) (fetch-tool summary); Semitracks, ‘Auger Electron Spectroscopy’ failure-analysis reference page (fetch-tool summary; current equation shown as an image); ‘Lecture 20: Auger Electron Spectroscopy’, Electron Spectroscopies course, Caltech MMRC (2007) (fetch-tool summary; symbols garbled); D. Briggs and M. P. Seah, Practical Surface Analysis, Vol. 1, 2nd ed., Wiley 1990, and J. I. Goldstein et al., Scanning Electron Microscopy and X-Ray Microanalysis, 4th ed., Springer 2018 (background knowledge; not opened).

Ques: How is AES quantified and how accurate is it?

With relative sensitivity factors, accurate to tens of per cent. Diplas gives $N_A=I_A/(I_A+F_{AB}{I}_B+\dots )$ with $F_{AB}=(I_A/N_A)/(I_B/N_B)$ determined from binary standards, calling it highly simplified but workable. With an assumed intensity ratio $I_B/I_A$ = 0.8 and $F_{AB}$ = 1.25, the atomic fraction of A is 0.50 (computed). Semitracks quotes accuracy of about 50 per cent without standards and 10 per cent with them, and sensitivity factors measured at 3 $\mathrm{kV}$. Matrix effects on backscattering and escape depth, and peak-shape changes with chemistry, limit accuracy (background knowledge, Briggs and Seah). Detection limits are about 0.1 atomic per cent (Caltech lecture; Semitracks), near $1\times 10^{13}$ atoms per square centimetre in a 2 $\mathrm{nm}$ layer (computed). Reference: S. Diplas, ‘An Introduction to Auger Electron Spectroscopy’, MENA 3100 slides, SINTEF and University of Oslo (2017) (fetch-tool summary); Semitracks, ‘Auger Electron Spectroscopy’ failure-analysis reference page (fetch-tool summary; current equation shown as an image); D. Briggs and M. P. Seah, Practical Surface Analysis, Vol. 1, 2nd ed., Wiley 1990, and J. I. Goldstein et al., Scanning Electron Microscopy and X-Ray Microanalysis, 4th ed., Springer 2018 (background knowledge; not opened); ‘Lecture 20: Auger Electron Spectroscopy’, Electron Spectroscopies course, Caltech MMRC (2007) (fetch-tool summary; symbols garbled).

Ques: How does AES depth profiling work, and what are its pitfalls?

By alternating Auger measurement with Ar ion sputtering. Diplas describes an automated Ar gun with Zalar rotation, and Semitracks notes that rotation reduces differential sputtering, where different phases or grains erode at different rates and roughen the crater. At an assumed 5 $\mathrm{nm}$ per minute, a 100 $\mathrm{nm}$ metal stack takes 20 minutes (computed). The surface sensitivity that makes AES useful, with escape from roughly the top 50 $\mathring{\mathrm{A}}$ (Semitracks), means that ion mixing and preferential sputtering of one element directly bias the measured composition (background knowledge, Briggs and Seah). Insulators charge under the electron beam; Semitracks notes silicon nitride is particularly difficult, and Diplas suggests a low-energy ion flood for poorly conducting samples. Reference: S. Diplas, ‘An Introduction to Auger Electron Spectroscopy’, MENA 3100 slides, SINTEF and University of Oslo (2017) (fetch-tool summary); Semitracks, ‘Auger Electron Spectroscopy’ failure-analysis reference page (fetch-tool summary; current equation shown as an image); D. Briggs and M. P. Seah, Practical Surface Analysis, Vol. 1, 2nd ed., Wiley 1990, and J. I. Goldstein et al., Scanning Electron Microscopy and X-Ray Microanalysis, 4th ed., Springer 2018 (background knowledge; not opened).

Ques: How should the beam energy be chosen?

High enough to ionize the core level efficiently, low enough for resolution and minimal damage. The overvoltage $U=E_0/E_W$ governs ionization, with cross sections largest near U of 3 to 4 (background knowledge, Goldstein et al.). For Si K at 1.839 $\mathrm{keV}$ a 10 $\mathrm{keV}$ beam gives U = 5.4, while the Si L level at 99.8 $\mathrm{eV}$ is efficiently excited even at 2 $\mathrm{keV}$, U = 20 (computed). Semitracks lists beams of 5 to 30 $\mathrm{keV}$ and vacuum below $5\times 10^{-9}$ $\mathrm{torr}$, and Diplas maps at 2 $\mathrm{kV}$ for best lateral resolution. Electron-beam damage, including desorption and reduction of oxides, rises with dose; Semitracks notes damage is often annealable by a low-temperature bake. Reference: D. Briggs and M. P. Seah, Practical Surface Analysis, Vol. 1, 2nd ed., Wiley 1990, and J. I. Goldstein et al., Scanning Electron Microscopy and X-Ray Microanalysis, 4th ed., Springer 2018 (background knowledge; not opened); Semitracks, ‘Auger Electron Spectroscopy’ failure-analysis reference page (fetch-tool summary; current equation shown as an image); S. Diplas, ‘An Introduction to Auger Electron Spectroscopy’, MENA 3100 slides, SINTEF and University of Oslo (2017) (fetch-tool summary).

References

1. S. Diplas, “An Introduction to Auger Electron Spectroscopy,” MENA 3100 lecture slides, SINTEF and University of Oslo, 2017. Available: https://uio.no/studier/emner/matnat/fys/MENA3100/v17/mena-3100-aes.pdf
2. D. Briggs and M. P. Seah, Practical Surface Analysis, Vol. 1, 2nd ed., Wiley, 1990; J. I. Goldstein et al., Scanning Electron Microscopy and X-Ray Microanalysis, 4th ed., Springer, 2018 (cited from background knowledge).
3. “Lecture 20: Auger Electron Spectroscopy,” Electron Spectroscopies course, Caltech Molecular Materials Research Center, 2007. Available: https://mmrc.caltech.edu/XPS%20Info/Electron%20Spectroscopies%202007/L20%20Auger.pdf
4. X-Ray Fluorescence Spectrometer (XRF and TXRF) page of this vault.
5. Semitracks, “Auger Electron Spectroscopy,” failure and yield analysis reference material. Available: https://semitracks.com/reference-material/failure-and-yield-analysis/failure-analysis-materials-characterization/auger-electron-spectroscopy.php

Related pages in this vault: X-Ray Photoelectron Spectrometer (XPS), X-Ray Fluorescence Spectrometer (XRF and TXRF), Scanning Electron Microscope (SEM), Secondary Ion Mass Spectrometer (SIMS and ToF-SIMS), Rutherford Backscattering Spectrometer (RBS), EBIC and EBAC, Atom Probe Tomography (APT), Focused Ion Beam (FIB).