Scanning Electron Microscopy and X-Ray Microanalysis

514

29

Selection of Candidate Crystallographic

Phases

EBSD requires the possible phases that will be analyzed to be

selected before the analysis is started. Generally, EBSD is

conducted on samples that are already well characterized

with respect to the phases that are present. Modern systems

are capable of sorting through a large number of phases to

match with the experimental patterns, but the operator

should try to keep this list to a minimum to allow maximum

speed of acquisition. There are many databases that provide

crystallographic data and it is also possible for the operator to

input specific descriptions of unit cells.

Microscope Operating Conditions and Pattern

Optimization

It is difficult to recommend specific operating conditions for

EBSD of all samples, but there are starting conditions that

should allow the system to be set up efficiently. It is suggested

that 20  kV and a beam current of a few nanoamperes is a

good starting point for EBSD analysis. The quality of EBSD

patterns can be rapidly assessed under these conditions. For

faster acquisition higher beam current is always better, as

long as the resolution is consistent with the microstructural

length scales that are to be studied. Higher operating voltages

are also sometimes useful with coated samples, and lower

voltages may be used to provide an improved spatial resolu-

tion at the expense of acquisition speed. The operator should

strive for clear patterns. In most commercial systems, the

operator has a choice of the pattern resolution that is to be

collected. For many orientation studies, the largest number

of pixels is almost never needed and the EBSD detector is

binned to produce larger pixels. For example, a typical EBSD

detector may have a maximum pixel resolution of 1600 ×

1200, but one would not use the full resolution and would

select to bin the result; so, for example, 4 × 4 binning would

result in an EBSD pattern with 400 × 300  pixels. Binning

helps with pattern quality as larger pixels collect more signal

and thus increasing the S/N of the pattern. Binning of the

detector allows higher speed acquisitions to be achieved.

Additional increases in pattern quality can be achieved at the

expense of collection speed.

Once the detector settings have been determined it is also

necessary to select the background removal method. Modern

EBSD systems have many methods for background removal

while older systems will be limited. Correct background cor-

rection is important to maximize the signal content of the

patterns while suppressing the high background contribu-

tion that is always present. For polycrystalline samples, it is

easy to scan a large representative region of the sample which

collects the average background levels without the sharp dif-

fraction features. Other methods that utilize a software blur-

ring algorithm may also be utilized and can be better than the

collected back ground method.

Now that the sample and the detector position are set and

beam conditions that provide useful EBSD patterns are estab-

lished, it is now time to calibrate the system. Calibration on

modern systems is entirely automatic provided a suitable

match unit has been specified. It is important that once a cali-

bration has been established that the sample to detector

geometry not be altered or a new calibration will need to be

determined.

Selection of EBSD Acquisition Parameters

Successful orientation mapping will depend on careful selec-

tion of the mapping parameters and the most important of

these is specifying the step size or the spacing between indi-

vidual measurement points. A selection of a spacing that is

too large risks missing the important microstructural fea-

tures and a spacing that is too small will require longer acqui-

sition times with little gain in information. A good starting

point is to plan on between four and ten pixels or measure-

ments across the smallest features to be studied. This sam-

pling will provide quality images and data while not wasting

time acquiring redundant information.

At this time it is useful to acquire electron images of the

region of interest. Secondary electron imaging may show

some surface features but imaging of highly polished samples

may not provide useful information. The use of forescattered

detectors is recommended as there is a good signal level and

forescattered images often show surprising high grain

contrast.

Collect the Orientation Map

Now that the experimental conditions for the EBSD acquisi-

tion have been selected it is often best to collect a small map

to determine if the parameters selected are capable of pro-

ducing a quality result. One of the most common ways to

judge a quality result is to look at the number or the fraction

of pixels in the map that have been successfully indexed to a

certain level of confidence. Modern systems on well-polished

samples can be capable of indexing 95 % or more of the pix-

els. Of course second phase fractions and grain size can influ-

ence the number of indexed patterns. It is not always

necessary to have 95 % of the pixels indexed to obtain a useful

result. If a high fraction of the pixels are not indexed it is

important to understand the reasons. If the correct phase or

phases have been selected then it may be that the system was

not calibrated adequately. If the selected phases are correct

and the calibration is correct then it is possible that sample

preparation was not optimal or the sample is heavily

deformed, leading to the low fraction of indexed pixels. Once

a satisfactory indexing rate is achieved in the test map it is

now reasonable to select a larger area for orientation map-

ping and proceed with mapping.

References

Brewer L, Michael J (2010) Risks of ‘cleaning’ electron backscatter data.

Microsc Today 18:10

Britton T, Jiang J, Guo Y, Vilalta-Clemente A, Wallis D, Hansen L,

Winkelmann A, Wilkinson A (2016) Tutorial: crystal orientations and

EBSD – or which way is up. Mater Charact 117:113

Chapter 29 · Characterizing Crystalline Materials in the SEM