Quantitative electron probe microanalysis of Y-Ba-Cu-O superconducting materials

The high temperature superconductor YBa₂Cu₃O_7–x has been studied by quantitative electron probe microanalysis (EPMA). After investigation of appropriate standards and choice between three different matrix correction methods this technique provides very accurate analytical results including those for oxygen. The quantitative results even allow the determination of the stoichiometry with deviations less than 2% relative. The application of computercontrolled EPMA is demonstrated by the characterization of the microstructure of a ceramic Y-Ba-Cu-O sample. Finally the properties of copperL ₂ andL ₃ valence band spectra are presented for compounds with copper in different valence state.Dedicated to Professor Günther Tölg on the occasion of his 60th birthday     

Quantitative electron probe microanalysis: Fluorescence correction uncertainty

Summary In order to evaluate the fluorescence correction uncertainty in electron probe microanalysis, effects of two general types were investigated. The effects of uncertainty in 1. microprobe operational variables, including X-ray emergence angle and operating voltage, and 2. errors in model input parameters, such as fluorescence yield factors and mass attenuation coefficients, were evaluated. The major conclusion of this study is, that even in cases requiring large fluorescence corrections, microprobe operation at high (30°) X-ray emergences angles, and at the lowest voltage compatible with a reasonable X-ray output is entirely satisfactory. This conclusion is reached by substituting into existing fluorescence correction models reasonable input parameter errors, X-ray emergence angles, and operating voltages. Results show that absolute compositional errors vary quite slowly with X-ray emergence angle and operating voltage. The chief source of error among input parameters is the fluorescence yield factor. The role of model selection as a source of potential error is also discussed.

---

Zusammenfassung Für die Ermittlung der Unsicherheit der Fluoreszenzkorrektur bei der Elektronenstrahl-Mikroanalyse sind zwei prinzipielle Effekte zu berücksichtigen: 1. Unsicherheiten, bedingt durch den Austrittswinkel der Röntgenstrahlung und die Beschleunigungsspannung und 2. fehlerhafte Parameter, wie Fluoreszenzausbeute und Massenschwächungskoeffizienten. Das wichtigste Ergebnis der vorliegenden Arbeit ist, daß die Mikrosondenanalyse sogar in Fällen, die eine große Fluoreszenzkorrektur erwarten lassen, und selbst bei großem Austrittswinkel (30°) und bei niedrigster Spannung mit einer erträglichen Impulsausbeute völlig zufriedenstellend ist. Dieses Ergebnis wurde erhalten, nachdem annehmbare Parameter für Abnahmewinkel und Beschleunigungsspannung in bekannte Fluoreszenzkorrektur-Modelle eingesetzt worden waren. Die Resultate zeigen, daß der absolute Fehler für die Zusammensetzung nur wenig mit dem Abnahmewinkel variiert. Der Faktor für die Fluoreszenzausbeute ist die hauptsächliche Fehlerquelle. Die Auswahl der Korrekturverfahren als eine Quelle für einen möglichen Fehler wird ebenfalls diskutiert.

     

Quantitative electron probe microanalysis: Uncertainty in the atomic number correction

Summary The atomic number effect in quantitative electron probe microanalysis has been considered from the point of view of error propagation. The results indicate that the chief causes of analysis uncertainty are poorly known electron backscattering factors and mean ionization potential values. Operating variables such as working voltage or X-ray take-off angle cannot be manipulated so as to reduce the effect of the uncertainties. However, the overall analysis uncertainty remains nearly constant over the entire range of working voltages practical for microanalysis.

---

Zusammenfassung Der Atomnummerneffekt bei der quantitativen Elektronenstrahlmikroanalyse wurde im Hinblick auf seinen Beitrag zum Analysenfehler untersucht. Danach ist die Hauptursache für den Fehler die Ungenauigkeit der Rückstreufaktoren und der mittleren Ionisationspotentiale. Die Arbeitsbedingungen, wie Abnahmewinkel der Röntgenstrahlung oder Beschleunigungsspannung, können nicht so gewählt werden, daß diese Ungenauigkeit verringert wird. Der gesamte Analysenfehler bleibt jedoch fast über den ganzen für die Elektronenstrahlmikroanalyse üblichen Bereich der Anregungsspannung konstant.

     

Application of Monte Carlo method to the electron probe microanalysis of thin films

In the present paper the basic features of Monte Carlo method as applied to the electron probe microanalysis are outlined. In particular, applications to a large variety of experimental situations are reviewed. The problems examined are as follow: i) a binary film on a substrate of a third element; ii) a ternary film without substrate; iii) a ternary film on a substrate of an element present in the film; iv) a multi-layer elemental films; v) a multi-layer compound film.The output of the computer program consists of: a) spatial distribution of the penetrating electrons; b) depth distribution of the generated X-rays; c) spatial distribution of the deposited energy; d) spatial distribution of the electron-hole pairs created in semiconductors.By comparing X-rays intensities with experimental data, one is able to obtain both the unknown composition and the thickness of the film. In some parcular instances, additional informations, such as an independent determination of the thickness or measurements of the X-rays intensities at different electron energies, may be required.     

Development of X-ray electron probe microanalysis in Siberia

This paper describes the contribution of Siberian researchers to the development of X-ray electron probe microanalysis (EPMA) in 1969–2007. It is devoted to the 70th birthday of Valerii Petrovich Afonin (September 24, 1938–May 15, 1995).     

Quantitative separation of traces of copper by electrodeposition and determination by electron probe microanalysis

Summary A combination of ion-exchange and micro-electrodeposition techniques permits the rapid separation of sub-microgram amounts of copper in a form suitable for determination with the electron probe microanalyser. Recoveries of 98% and standard deviations < 5% have been checked by other methods at the 10-g level.

---

Zusammenfassung Eine Verbindung von Ionenaustausch und Elektrolyse ermöglicht die rasche Abscheidung von Submikrogrammengen Kupfer in einer für die Elektronenstrahlmikroanalyse geeigneten Form. Ausbeuten von 98% und Standardabweichungen von < 5% wurden mit anderen Verfahren im 10-g-Maßstab erwiesen.

---

---

Presented at the VI. International Symposium on Microtechniques September 1970, Graz, Austria.     

A computer program for use in quantitative electron probe microanalysis

Summary A computer program for use in the conversion of electron probe X-ray intensity ratios into chemical compositions has been written. The inclusion of several of the more successful correction procedures facilitates the comparison of these procedures and allows an investigator to select the correction procedure which will give the best results for the system being studied. Theoretical calibration calculations of the intensity ratio versus composition can be performed and these results can be plotted by machine. The program has been designed to be as versatile as possible and still retain simple experimental input requirements. Some results obtained using recently published correction procedure are discussed.

---

Zusammenfassung Ein Computerprogramm wurde geschrieben, um Intensitätsverhältnisse der Elektronenstrahlmikroanalyse in chemische Zusammensetzungen umzurechnen. Die Einfügung einiger besonders erfolgversprechender Korrekturverfahren erleichtert den Vergleich dieser Verfahren und ermöglicht dem Bearbeiter, jenes Korrekturverfahren auszuwählen, das für das eben untersuchte System die besten Ergebnisse liefert. Die theoretischen Berechnungen des Intensitätsverhältnisses in Abhängigkeit von der Zusammensetzung können als Eichung durchgeführt werden und diese Ergebnisse kann die Maschine ausdrucken. Das Programm wurde so vielseitig wie möglich erstellt, beinhaltet aber dennoch einfache experimentelle Eingabemöglichkeiten. Einige Ergebnisse, die durch den Gebrauch kürzlich veröffentlichter Korrekturverfahren erhalten wurden, werden diskutiert.

---

Résumé Un programme d'ordinateur servant à convertir les rapports d'intensité de rayons X de sonde électronique en compositions chimiques a été rédigé. L'inclusion de plusieurs procédures de correction parmi celles qui ont obtenu le plus de succès facilite la comparaison de ces procédures et permet à l'investigateur de choisir la procédure de correction qui, à son avis, donnera les meilleurs résultats pour lè système étudié. Des calculs théoriques de calibrage entre le rapport d'intensité et la composition peuvent être faits et ces résultats peuvent être rapportés sur graphique mécaniquement. De par sa conception, le programme jouit du maximum de souplesse tout en gardant le plus de simplicité possible aux données expérimentales nécessaires à l'entrée. Certains résultats obtenus à l'aide de procédures de correction récemment publiés, font l'objet de discussions.

     

Correction procedures in electron probe microanalysis of bulk samples

In recent years many of the advances in quantitative microprobe analysis have come from the improved ability to model the complex physics in the computer. This has enabled the accuracy of normal bulk analysis to be improved-right down to the very light elements as far as boron-and has extended the technique to the analysis of multiple thin films and layered samples. It has also allowed the user to shortcut the use of standards, though at the cost of somewhat reduced accuracy. Thus the analyst has come to rely more and more on correction procedures of increasing complexity, yet at the same time must understand their limitations. It is likely that in future the computer itself will be able to accumulate the best practice of expert users, advising the newcomer how these procedures can be best applied and where they are most likely to be in error.     

Accuracy of film thickness determination in electron probe microanalysis

The thickness of copper films (100–450 nm) on silicon substrates was determined by electron probe microanalysis (EPMA) applying (z) procedures of Pouchou and Pichoir. Film thickness was calculated from experimental k-ratios analyzed with electron energies between 6 and 30 keV using commercial software (LAYERF distributed by CAMECA). The influence of the incident electron energy and X-ray line chosen for analysis on the results was investigated. Accuracy of film thickness determination was evaluated by comparison with Rutherford backscattering spectroscopy (RBS) and secondary ion mass spectrometry (SIMS). The difference between layer thicknesses determined with EPMA and RBS is in general less than 2%, if EPMA measurements are performed with various electron energies. Layer thicknesses determined with Cu-L are mostly closer to values obtained by RBS than those derived from Cu-K radiation. Preliminary SIMS measurements yielded inconsistent results and, thus, cannot be used in this case to determine the layer thickness of Cu films on Si accurately.     

Performance and limitations of electron probe microanalysis applied to the characterization of coatings and layered structures

Electron probe microanalysis (EPMA) at low electron energies (typically 3–10 keV) has been used to study the composition of coatings with a film thickness of 0.1–1.3 m (bulk analysis). The accuracy of quantification, including the most critical situation of low atomic number elements (boron–oxygen), is about 5% relative by use of pure element or arbitrary compound standards. A special procedure of data processing (thin film analysis) enables the simultaneous determination of film thickness and composition, provided that the substrate is known. The film thickness may be in the range of 2 nm to 5 m. EPMA can be also applied in the mode of non-destructive in-depth analysis, which is based on the combined evaluation of experiments at different electron energies. The quantitative characterization of layered structures and implantation zones is demonstrated.     

Grazing-emission electron probe microanalysis of particles near the substrate edge

The distribution of X-ray radiation from a small particle, situated near the sharp edge of a flat substrate and excited by an electron beam, is analyzed theoretically taking into account the effects of the radiation reflection and its partial penetration through the substrate lateral surface. It has been shown that in some cases this dependence is strongly related to the distribution of primary X-ray sources above the substrate, which allows to use the signal — grazing-angle dependence for structure investigations of particles. The proposed technique was tested in experiments with CaCO₃ particles, situated on a pure Si substrate and exposed to a highly concentrated HNO₃ environment. Comparison of the observed angle dependencies for Ca and O characteristic X-ray radiation with calculated results revealed the intrinsic structure of the particles caused by their preliminary chemical treatment.     

Performance and limitations of electron probe microanalysis applied to the characterization of coatings and layered structures

Electron probe microanalysis (EPMA) at low electron energies (typically 3-10 keV) has been used to study the composition of coatings with a film thickness of 0.1-1.3 microm ("bulk" analysis). The accuracy of quantification, including the most critical situation of low atomic number elements (boron-oxygen), is about 5% relative by use of pure element or arbitrary compound standards. A special procedure of data processing ("thin film" analysis) enables the simultaneous determination of film thickness and composition, provided that the substrate is known. The film thickness may be in the range of 2 nm to 5 microm. EPMA can be also applied in the mode of "non-destructive in-depth analysis", which is based on the combined evaluation of experiments at different electron energies. The quantitative characterization of layered structures and implantation zones is demonstrated.     

Interference technique in grazing-emission electron probe microanalysis of submicrometer particles

The distribution of X-ray radiation from a small particle, situated on a flat support and excited by an electron beam, is analyzed theoretically. The interference pattern, appearing within the region of grazing take-off angles, is shown to contain valuable information about the particle structure and composition. It also provides the resources for extracting this information that are valid even in conditions of extremely low signals (e.g. for very small samples), when the effect of background suppression for angles of the total reflection region is insufficient. Characteristic elements of the interference pattern are found that are relatively independent on the noise caused by the support radiation. Algorithms for this noise correction and elimination are proposed. An approximate formalism for the particle structure retrieval, based on the Fourier-transformation properties, is described. Potential applications of the proposed technique are illustrated with numerical models of spherical and hemispherical particles, both homogeneous and of ‘core-and-shell’ structures, with total sizes between 30 and 200 nm. The additional possibilities of studying the shape, structure and elemental composition of small particles have been demonstrated. The general conditions of the experimental realization of proposed methods are discussed.     

Behavior of the continuous X-ray background in grazing-exit electron probe X-ray microanalysis

The energy distribution of the background radiation originating from two kinds of substrate materials has been studied using Electron Probe X-Ray Microanalysis (EPMA) at grazing-exit angles. The different behavior between tendencies of background reduction for the silicon and the gold substrates at the grazing exit angle is explained using a critical energy concept. It is also shown experimentally that a gold substrate results in a lower background intensity than a silicon substrate in the energy region 5–7 keV, while on the other hand, for elements with X-ray lines of 2–3 keV, it is advantageous to perform the analysis when such elements are deposited on the silicon substrate.     

Separate K-line contributions to fluorescence enhancement in electron probe microanalysis

The behaviour of the fluorescence enhancement correction factor in electron probe microanalysis, as a function of incident electron energy and take-off angle, is assessed for different binary samples in a wide range of compositions. Monte Carlo simulations are employed to validate Reed's correction algorithm [S.J.B. Reed, Characteristic fluorescence corrections in electron-probe microanalysis, Br. J. Appl. Phys. 16 (1965) 913-926], by means of estimating the primary excited radiation volume and the volume corresponding to secondary fluorescence generation. Then, Reed's expression for the fluorescence enhancement has been modified to account for Kα and Kβ line contributions separately. It is clearly shown that in certain cases the assignment of all fluorescent contribution to the Kα lines may be inadequate, particularly when trace element analysis imposes an accurate determination of elemental concentrations.