eTraining Introduction

Specimen Preparation

Leica Ultracut UCT



Hitachi S-4700 FE-SEM

Hitachi FB-2000A FIB


Veeco Dim 3000 AFM

Fluorescence Microscopes



What are the geometric effects on X-ray signal?

Quantitative microanalysis is based on the assumption that specimens are flat and polished. Any deviation from ideal conditions produces error in the results. Errors of 10 to 20% or more are possible when you analyze rough specimens. Therefore, data from rough surfaces must be characterized as qualitative or semi-quantitative in nature.

Normal/Ideal Situation

Matrix correction software, ZAF for instance, assumes that your specimen is flat, polished and perpendicular to the electron beam. In this case the matrix corrections are accurate. This leads to an ideal situation in X-ray analysis.

Normal or Ideal

Tilted Surface

Objective lenses were once square shaped. This required that the EDS detector enter the specimen chamber at 90 degrees. An analyst would have to tilt the specimen towards the detector to increase count rates.

Currently, objective lenses have a cone shape, which allows the detector to come in at a smaller angle, removing the necessity of tilting the specimen to achieve better results.

Tilting the specimen, however, has other benefits. The secondary electron signal may be enhanced by tilting the specimen towards the SE detector.

Tilting the specimen away obviously then decreases the signal.

Tilt Away From Detector
Tilt Toward Detector

Rough Surface

When the beam is placed in a low spot on the rough surface, it is possible that very few X-rays will reach the detector. The result will be very low dead times even with high beam currents.

Rough Surface

What would be your recommendations for analysis conditions?

Aperture Choice, Beam Current, Working Distance, Coating Thickness

The best settings for these values depend on the instrument used.

Accelerating Voltage

The analyst must consider what accelerating voltage will be required to efficiently excite X-rays for a particular element in the specimen. The appropriate voltage should be approximately 2x the beam energy of the emission line you want to excite.


Quantitative analysis is performed at a discrete point, not in scan mode at low magnification. There are two ways to obtain a discrete spot in the SEM. One is to use the “point mode” of the scan generator. The other is to zoom to the highest magnification. We prefer the second method.

What are the challenges to accurate detection and analysis?

Low ZLow Z

There are many obstacles to accurate analysis of a specimen containing low atomic number elements. For one, they are difficult to measure with an EDS system, as the sensitivity of that system is not great enough to accurately collect counts. Another issue lies in the detector itself. The window that the electrons pass through preferentially absorbs electrons from low Z elements. In the spectra, multi-phase specimens have L family lines occurring in the low Z, or low energy range, making it challenging to decipher which peaks are which. The last hurdle lies in the fact that the background of the spectra is nonlinear in the low Z (low energy) range, making peaks in that range appear smaller than they are. One way to tell whether an element has a higher or lower atomic number is to take note of the shade of grey that appears on the SEM viewing screen. Even in SE mode, low atomic number elements appear a darker shade of grey than do higher atomic number elements.

Low ConcentrationLow Concentration

Peak height (intensity) is governed by the mass concentration of an element in the specimen. The magnitude of the continuum background limits the ability to measure low elemental concentrations. This is described by the Peak to Background or P-B ratio. Elements in low concentrations will have a very low peak to background ratio and are therefore difficult to identify, much less quantify.

System Peaks

A system peak is derived from elements not present in the specimen; specifically, from the environment in which the detector lies. For example, the metals from the objective polepiece (lens), the stage, and the specimen holder, can contribute to the X-ray spectrum when bombarded with beam electrons or primary X-rays from the specimen. The most common system peaks include those of iron, copper, nickel and zinc. One way to avoid these peaks is to produce a flat specimen. System peaks are more of a problem when analyzing rough specimens. See the FAQ above on geometric effects.

Other Peak Artifacts

Escape peaks are the result of a photon escaping the detector. This artifact is much more likely to occur when a photon is produced somewhere close to the surface, and it often appears 1.74 KeV before a large peak in the spectra.

Sum peaks are the result of "pulse pileup", which occurs if a photon reaches the detector before the previous one has been processed, creating a signal exactly twice as large as it should be.

Peak overlaps occur when two elements have similar X-ray photon emission energies. The EDS detector senses them individually but does not have the capability to separate them on a graph, due to limited energy resolution. Their energies are therefore combined into one large peak instead of several smaller ones. This results in difficulties analyzing elements present in a specimen. In some cases it is almost impossible to do the elemental analysis with the EDS system. Peak overlaps seldom occur when using the WDS detector. Note that WDS analysis can only be performed with the JEOL JSM-6400 SEM. See WDS Quantitative Analysis Procedure under JEOL JSM-6400 SEM.

Standards for Microanalysis

Energy Calibration, Beam Current, Coating Thickness

The best settings for these values depend on the instrument used.

How do I choose Standards?

Always use standards that closely match the composition of your specimen. If you are uncertain of the composition of your specimen, perform a preliminary qualitative analysis. See lab personnel for assistance in locating standards.

How long should I let the program count X-rays? (EDS/WDS)

EDS - 60 seconds is adequate for qualitative work. Quantitative analysis requires a count time of at least 100 seconds for accurate results. When collecting spectra for standards, it is best to use 200 seconds. Remember that biological specimens are beam sensitive and can be distorted if left under the beam for even short periods of time.

WDS - Normally we use count times in the range of 5 -10 s for backgrounds and ~20 s for peak measurements. These general recommendations would increase for low concentration measurements. See lab personnel for special cases.

Operating conditions between Std and Spec (ED/WD same)

Rigorous quantitative analysis (using physical standards) requires that conditions used to measure specimens duplicate those used to measure standards. The following conditions are chosen and set by the operator. It is easy for novices to make errors in these settings.

Geometry (WD, flat/polished, tilt)

See the FAQ above on geometric effects.

Accelerating Voltage

Must be chosen to provide ~2 times the energy required for excitation of the heaviest element in your specimen. Commonly 15 or 20 kV are used.

Count Time

See the FAQ above on program counting.

How do I export Microanalysis data?

EDS - Spectra can be saved in a spreadsheet format or as a JPG image. EDS quantitative data can be saved as JPGs or exported as a spreadsheet.

WDS - Data are available only in a spreadsheet.

Top of Page