Добавил:
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Joseph I. Goldstein, Dale E. Newbury [et al.]. Scanning Electron Microscopy and X-Ray Microanalysis. (2017). (ISBN 978-1-4939-6674-5). (ISBN 978-1-4939-6676-9). (DOI 10.1007978-1-4939-6676-9).pdf
Скачиваний:
19
Добавлен:
11.04.2023
Размер:
68.34 Mб
Скачать

References

a

Counts

489

 

28

 

 

 

b

4000

c

3500

3000

2500

2000

1500

1000

500

0

300

350

400

450

500

550

600

650

700

750

800

 

 

 

 

 

nm

 

 

 

 

 

. Fig. 28.12  Acetaminophen, E0 = 20 keV: a SEM SE image; b panchromatic CL image; c CL spectrum (Example courtesy of S. Wight, NIST)

28.3.3\ Organic Compounds

 

References

 

 

 

Despite the general vulnerability of organic compounds to radiation damage under electron bombardment, some organic compounds can be examined with CL spectrometry.

.Figure 28.12 shows an SEM SE image and a panchromatic CL image of acetaminophen (paracetamol, N-(4-hydroxyphenyl) ethanamide N-(4-hydroxyphenyl)acetamide) along with a CL spectrum showing broad CL bands.

Magee C, Teles G, Vicenzi E, Taylor W, Heaney P (2016) Uranium irradiation history of carbonado diamond; implications for Paleoarchean oxidation in the São Francisco craton (South America). Geology 44. doi: 10.1130/G37749.1

Zanetti M, Wittman A, Carpenter PK, Joliff BL, Vicenzi E, Nemchin A, Timms N (2015) Using EPMA, Raman LS, Hyperspetcral CL, SIMS, and EBSD to study impact meltinduced decomposition of zircon. Microsc Microanal 21:S3 1447–1448

491

 

29

 

 

 

Characterizing Crystalline

Materials in the SEM

29.1\ Imaging Crystalline Materials with Electron Channeling Contrast – 492

29.1.1\ Single Crystals – 492

29.1.2\ Polycrystalline Materials – 494

29.1.3\ Conditions for Detecting Electron Channeling Contrast – 496

29.2\ Electron Backscatter Diffraction in the Scanning Electron Microscope – 496

29.2.1\ Origin of EBSD Patterns – 498

29.2.2\ Cameras for EBSD Pattern Detection – 499 29.2.3\ EBSD Spatial Resolution – 499

29.2.4\ How Does a Modern EBSD System Index Patterns – 501 29.2.5\ Steps in Typical EBSD Measurements – 502

29.2.6\ Display of the Acquired Data – 505 29.2.7\ Other Map Components – 508

29.2.8\ Dangers and Practice of “Cleaning” EBSD Data – 508 29.2.9\ Transmission Kikuchi Diffraction in the SEM – 509 29.2.10\ Application Example – 510

29.2.11\ Summary – 513

29.2.12\ Electron Backscatter Diffraction Checklist – 513

References – 514

© Springer Science+Business Media LLC 2018

J. Goldstein et al., Scanning Electron Microscopy and X-Ray Microanalysis, https://doi.org/10.1007/978-1-4939-6676-9_29

\492 Chapter 29 · Characterizing Crystalline Materials in the SEM

While amorphous substances such as glass are encountered both in natural and artificial materials, most inorganic materials are found to be crystalline on some scale, ranging from sub-nanometer to centimeter or larger. A crystal consists of a regular arrangement of atoms, the so-called “unit cell,” which is repeated in a twoor three-dimensional pattern. In the previous discussion of electron beam–specimen interactions, the crystal structure of the target was not considered as a variable in the electron range equation or in the Monte Carlo electron trajectory simulation. To a first order, the crystal structure

29 does not have a strong effect on the electron–specimen interactions. However, through the phenomenon of channeling of charged particles through the crystal lattice, crystal orientation can cause small perturbations in the total electron backscattering coefficient that can be utilized to image crystallographic microstructure through the mechanism designated “electron channeling contrast,” also referred to as

“orientation contrast” (Newbury et al. 1986). The characteristics of a crystal (e.g., interplanar angles and spacings) and its relative orientation can be determined through diffraction of the high-energy backscattered electrons (BSE) to form “electron backscatter diffraction patterns (EBSD).

29.1\ Imaging Crystalline Materials

with Electron Channeling Contrast

29.1.1\ Single Crystals

The regular arrangement of atoms in crystalline solids can influence the backscattering of electrons because of the regular three-dimensional variations in atomic density in the crystal compared to those same atoms placed in the near random three-dimensional distribution of an amorphous solid. If a well-collimated (i.e., highly parallel) electron beam is directed at a crystal array of atoms along a series of different directions, the density of atoms that the beam encounters will vary with the crystal orientation, as shown in the simple schematic of .Fig. 29.1. A sense of this effect can be obtained by manual manipulation of a macroscopic, three-­dimensional ball-and-stick model of a crystal. For certain orientations of the model, the observer can see through the “atoms” along the open gaps between the planes in the model. In a real solid, the atoms are tightly packed, limited in their approach by the repulsive interaction of their atomic shells. The “channels” in reality are regions of the crystal where the atomic packing creates lower charge density with which the beam electrons will interact more weakly. When the beam is aligned with the channels, a small fraction of the beam electrons penetrate more deeply into the crystal before beginning to scatter. For beam electrons that start scattering deeper in the crystal, the probability that they will return to the surface as backscattered electrons is reduced compared to the amorphous target case, and so the measured backscatter coefficient is lowered compared to the average value from the amorphous target.

. Fig. 29.1  Schematic illustration of the channeling effect: the atomic area density that the beam encounters depends on its orientation relative to the crystal

For other crystal orientations where denser atom packing is found, the beam electrons begin to scatter immediately at the surface, increasing the backscatter coefficient relative to the amorphous target case. As seen in the Monte Carlo simulation of an amorphous target, the elastic scattering of beam electrons rapidly randomizes their trajectories out of their initially well-collimated condition, reducing and eventually eliminating sensitivity to channeling in the crystal. For a bulk target, the modulation of the backscatter coefficient between the maximum and minimum channeling case is small, typically only about a 2–5 % difference. Nevertheless, this crystallographic or electron channeling contrast can be used to form SEM images that contain information about the important class of crystalline materials.

To determine the likelihood of electron channeling, the properties of the electron beam (i.e., its energy, E, or equivalently, wavelength, λ) are related to the critical crystal property, namely the spacing of the atomic planes, d, through the Bragg diffraction relation:

nλ =2d sinθB \

(29.1)

where n is the integer order of diffraction (n = 1, 2, 3, etc.). Equation 29.1 defines a special beam incidence angle relative to a particular set of the crystal planes with spacing d (referred to as the “Bragg angle,” θB) at which the channeling condition changes sharply from weak to strong as the beam incidence angle increases relative to that particular set of crystal planes. Since a real crystal contains many different possible sets of atom planes, the degree of channeling

29.1 · Imaging Crystalline Materials with Electron Channeling Contrast

encountered for an arbitrary orientation of the beam to the crystal is actually the sum of the contributions of many planes. Generally, these contributions tend to cancel giving the amorphous average backscattering except when the Bragg condition for a certain set of planes dominates the mix, giving rise to a sharp change in the channeling condition. What is the magnitude of Bragg angles for typical SEM beam and target conditions? Consider an incident beam of E0 = 15 keV, where the electron wavelength, λe is given by the de Broglie equation:

λe =h/p \

(29.2a)

where h is Planck’s constant and p is the momentum (m0v). In terms of beam energy, the wavelength is given by

λ (nm)=1.226 / E0.5

(1+ 0.979E 6 E)0.5

(29.2b)

e

 

 

\

 

 

 

where E is expressed in eV (Hirsch et al. 1965).

At 15 keV, λe = 0.00994 nm. If this 15 keV-beam is directed at a crystal of silicon, which has a diamond-cubic crystal lattice with a fundamental cube dimension of a = 0.543 nm, the series of possible “allowed” Bragg angles (expressed in the “Miller indices” [hkl] which designate the crystal planes) is given in .Table 29.1.

We can directly image the effects of crystallographic channeling on the total backscattered electron intensity for the special case of a large single crystal viewed with a large scanned area (i.e., low magnification). As shown schematically in .Fig. 29.2, the act of image scanning not only moves the beam laterally in x- and y-directions, but at each beam position the angle of the beam relative to the surface normal changes. For a 10-mm working distance from the scan rocking point, a scan excursion of 2 mm in width causes the beam incidence angle to change by ±12° across the field. As seen in

.Table 29.1, the allowed Bragg angles for a 15-keV electron beam striking Si have values of a fraction of a degree to a few degrees, so that a scan angle of ±12° will certainly cause the beam to pass through the Bragg condition for at least some of

. Table 29.1  Bragg angles for Si with E0 = 15 keV

Planes (hkl)

Spacing (nm)

θB (degrees)

111

0.314

0.908

 

 

 

220

0.192

1.48

311

0.164

1.74

400

0.136

2.10

422

0.111

2.57

511

0.105

2.72

493

 

29

 

 

 

qB

qB

. Fig. 29.2  Wide field (“low magnification”) image scanning produces sufficiently large changes in the scan incidence angle to pass through the Bragg conditions for the particular set of crystal planes. Note that the Bragg condition is satisfied in positive and negative going angles, giving rise to two sharp changes in contrast

the crystal planes. A large area image created at E0 = 15 keV by scanning a flat, topographically featureless silicon crys- tal—prepared with the surface nearly parallel to the (001) plane—is shown in .Fig. 29.3. The image consists of a pattern of bands, each running parallel to a particular crystal plane, creating a so-called electron channeling pattern (Coates 1967). The channeling effect appears as this series of prominent bands that span the crystal because the intersection of the crystal planes with the surface defines a linear trace, and the Bragg condition is satisfied along lines parallel to this trace where the scan angle relative to the planes equals the Bragg angle, ±θB. “Higher order” Bragg angles (n = 2, 3, 4, etc., with the specific integers that appear depending on “allowed reflections”) are satisfied as a series of progressively fainter lines parallel to the central bands, which can be seen in .Fig. 29.3 (b) for the family of {220} type bands. Note that it is appropriate to apply both a dimensional marker and an angular marker to this image. When the scanned area is decreased, that is, the magnification is increased, the range of the angular scan is reduced, and consequently, the bands on the display appear to widen and the angular extent of the electron channeling pattern decreases, as shown in

.Fig. 29.4a, b. If the magnification is made high enough and the angular range is sufficiently reduced, the beam will scan through such a small angle that the observer will see the center of the pattern swell to eventually fill the image with a single gray level. Effectively, the crystal presents a single atomic density that restricts the channeling effect to a single intensity value.