Electron Backscatter Diffraction

Electron Backscatter Diffraction (EBSD) Geometry

The beam-sample-detector geometry is an important consideration when setting up an Electron Backscatter Diffraction (EBSD) analysis. Optimising the geometry will ensure the best results for a wide range of applications, although it may be necessary to compromise on some aspect of the set-up to accommodate unusual samples or special applications.

The parameters to consider when setting up the EBSD experiment are as follows:

Sample Tilt Angle
Working Distance (WD): this is the distance from the final lens (or pole-piece) of the scanning electron microscope (SEM) column to the focal plane of the beam (i.e. the sample surface)
Detector Distance (DD): this is the shortest distance from the EBSD detector’s phosphor screen to the beam-sample interaction distance. It can be measured in mm, but is often expressed as a fraction of the phosphor screen dimensions.
Pattern Centre (PC): this is the point on the phosphor screen that is closest to the beam-sample interaction point and will vary according to the previous 2 parameters.

These parameters can be adjusted either on the SEM (e.g. by adjusting the stage position) or by moving the EBSD detector (insertion/retraction or, on detectors such as the Symmetry S3, by adjusting the elevation to move the phosphor screen up or down). The image to the right summarises these variables for a typical EBSD set up.

In the tabs below you can read more about how best to adjust these parameters, plus some more details on the special case of EBSD in transmission: transmission Kikuchi diffraction (TKD).

Schematic illustration showing the typical geometry for EBSD and marking the main variables

Schematic illustration showing the main variables associated with the geometry set-up for EBSD experiments

Historically, a tilt of approximately 70.5° was used in early EBSD experiments because at that tilt angle, the <114> direction for a silicon (100) crystal would be positioned at the pattern centre, making system calibration much easier.

Modern EBSD systems perform the geometry calibration automatically, but we still use a tilt angle of ~70° from horizontal for 2 reasons:

The high tilt angle gives a much higher diffracted signal, thus improving the signal to noise ratio in the EBSD pattern and making analyses significantly faster using a constant beam current
Although the resolution down the tilted surface is significantly worse than parallel to the tilt axis (typically ~2.5 times worse), this value is acceptable for most experiments and would be significantly poorer at higher tilt angles.

The series of images below show the effect of tilting the sample to lower tilt values. Note the progressive loss of signal to noise, plus the inversion of Kikuchi band contrast in the lower parts of the EBSP.

EBSD pattern collected from a Silicon sample at 70° sample tilt

EBSD pattern collected from a Silicon sample at 60° sample tilt showing some inverted contrast

EBSD pattern collected from a Silicon sample at 50° sample tilt, showing inverted contrast and loss of signal

EBSD patterns collected from Si at different sample tilt angles. Left – 70°, Centre – 60°, Right – 50°.

The tilt angle can be set by tilting the SEM stage to the desired value, or by using a pre-tilted holder. The choice will depend very much on the SEM in question and the set-up of the EBSD detector relative to the stage tilt axis. The advantage of tilting the stage is that, for most SEMs, the sample can then be moved in the plane of the sample surface (i.e. the tilted X-Y plane) and the WD and beam-sample-EBSD detector geometry will remain constant.

There has been some recent exploration of the potential of performing EBSD at very low sample tilt angles or in a horizontal geometry. Early studies showed significant problems with contrast reversal, but the use of direct electron detection technology, coupled with electron energy thresholding, can circumvent this problem. However, the signal to noise ratio is significantly reduced and hence extremely high electron doses are required for successful analyses using this low-tilt geometry.

The working distance is an important consideration when setting up an EBSD experiment, primarily because it will affect the position of the pattern centre on the EBSD detector’s phosphor screen (and thus the strength of the signal).

For most standard experiments, the optimum position of the pattern centre is about ^¾up the phosphor screen, but this will be affected by the beam energy and the atomic number of the material to be analysed.

When selecting the optimum WD, consideration needs to be made about the following factors:

The SEM spatial resolution may be better at shorter WD, especially for low accelerating voltages
The EBSD detector will be installed at a specific height, corresponding to an optimum WD. Although there will be some tolerance away from this position, the WD should be kept within ~5 mm of this recommended position
If the EBSD detector has elevation control (as for the Oxford Instruments Symmetry S3 detector), then the detector elevation can be adjusted to maintain an optimum geometry for any WD
An energy dispersive X-ray spectrometry (EDS) system will have a specific recommended analytical WD. If the sample is positioned away from that WD for simultaneous EDS and EBSD analyses, then the X-ray count rate will decrease or there may be variations in the count rates from the top to the bottom of the analysis area (especially at low magnifications). Retracting the EDS detector can help minimise this issue
If the WD is too long, then the EBSD detector may occlude the EDS detector preventing any simultaneous X-ray measurements. Adjusting the EBSD detector elevation or retracting the detector slightly may solve this problem
The sample size may dictate the minimum working distance that can be safely used

The series of images below show the changes in EBSP quality when moving from a 14.9 mm WD to a 22.5 mm WD. At the shorter WD, the EBSP is relatively noisy at the bottom of the pattern, whereas at the longer WD it is the upper part of the pattern that is noisy. EBSD indexing would still be possible for all of these geometries, but the loss of signal would become more significant the further the WD is extended beyond the optimum position.

EBSD pattern collected at a short working distance from a ferrite sample, with a high pattern centre position

EBSD pattern collected at a standard working distance from a ferrite sample, with an optimum pattern centre position

EBSD pattern collected at a long working distance from a ferrite sample, with a low pattern centre position

EBSD patterns from ferrite collected at different working distances. Left – 14.9 mm, Centre – 18.9 mm, Right – 22.5 mm. The position of the pattern centre is marked by the green cross.

When an EBSD detector is installed, typically the engineer will set a physical safe limit for the detector insertion that is recommended for that particular SEM chamber. This fully inserted position will normally have the detector’s phosphor screen close enough to the sample to subtend a large solid angle (>90°). Depending on the size of the phosphor screen, this will usually equate to a detector distance in the range of 15 – 30 mm.

However, it is not always beneficial to insert the detector to its fully inserted position and to work with the shortest possible detector distance. The following factors should influence the decision on what detector distance is best for a particular application:

Larger detector distances are safer: this might be worth considering if large area maps (involving automated stage movements) are planned, especially if the sample is very large or irregular
Moving to a longer detector distance will decrease the signal quite significantly. For a typical set-up, retracting the detector by 3 mm can reduce the signal by up to 50%, making any subsequent experiment much slower
As the detector distance increases, the solid angle decreases and fewer Kikuchi bands will be projected onto the phosphor screen. This could negatively impact the indexing process
As the detector distance increases, the Kikuchi bands will become broader. This can help in phase discrimination (especially if Kikuchi band widths are used) and for high angular precision analyses. The series of EBSPs to the right were collected from a ferritic steel sample, retracting the detector by ~40 mm from full insertion
Increasing the detector distance can also improve the quality of electron channelling contrast images collected using the lower forescatter detectors, as the signal from surface topography will be reduced. This is shown in the following 2 images from a broad ion-beam polished duplex stainless steel

Animated series of EBSD patterns collected from a ferrite grain at different detector distances

Forescatter electron image of an ion-polished steel collected at full detector insertion, with dominant topographic contrast

Forescatter electron image of an ion-polished steel collected at 10mm detector retraction, with dominant channelling contrast

Channelling contrast images of a duplex stainless steel collected using forescatter detectors below the EBSD detector phosphor screen. Left- full insertion, with the signal dominated by topography from the sample preparation. Right – detector retracted 10 mm, providing much greater crystallographic contrast.

To perform successful TKD analyses using a standard off-axis geometry, setting up in an optimised beam-sample-detector geometry is necessary. Unlike conventional EBSD, for TKD analyses the sample should be positioned in a horizontal geometry, or slightly tilted away from the EBSD detector.

The ideal geometry is shown in the annotated chamberscope image on the right.

There are several key considerations that are necessary for TKD:

As TKD analyses are primarily aimed at delivering the best spatial resolution, it can be beneficial to position the sample at a short WD (to improve the SEM resolution)
The resolution will be best with the sample in a horizontal position, as shown in the image. If a backtilt is unavoidable, it should be kept to a minimum (i.e. no greater than 20°).

$SEM chamberscope image showing the ideal geometry for off-axis transmission Kikuchi diffraction$

Chamberscope image showing the ideal geometry for off-axis TKD analyses

Many standard TKD sample holders will have a -20° pretilt (as shown in the image) – therefore it will be necessary to tilt the stage to 20° to position the sample in a horizontal geometry
In a horizontal position, the pattern centre will usually be off the top of the phosphor screen so as to avoid any shadowing problems. However, the pattern centre should ideally not be far off the top of the screen to minimise pattern distortion effects. This can be optimised by adjusting the stage height or the detector elevation.

Related products

Website by Miramar Communications Ltd