You are on page 1of 11

TEM characterization of dislocation image

Cite as: AIP Conference Proceedings 2262, 050009 (2020); https://doi.org/10.1063/5.0016430


Published Online: 17 September 2020

R. S. Duryat, and C.-U. Kim

AIP Conference Proceedings 2262, 050009 (2020); https://doi.org/10.1063/5.0016430 2262, 050009

© 2020 Author(s).
TEM Characterization of Dislocation Image
R.S. Duryat1, a) and C-U Kim2
1
Departemen Teknik Metalurgi dan Material Universitas Indonesia, Depok JABAR 16424 Indonesia
2
Dept. of Material Science and Engineering Department UT Arlington, Arlington TX 76019 USA
a)
Corresponding author: rahmat.saptono@ui.ac.id

Abstract. Dislocation is the most important defect of modern engineering materials involving structural and electronic
application. Proper operation and understanding of dislocation also serve the basis for other characterization work. This
paper deals with both underlying principles and practical aspects of dislocation characterization. Thin foil of electronic
grade Cu is loaded into a double tilt specimen holder of Analytical TEM, JEOL JEM 1200EX. [001] Low Index Zone
Axis was located with the aid of Kikuchi Lines. Two-Beam Condition was achieved for [ 200 ] [ 220 ] g vector. Bright
Field and Dark Field Image of the Dislocation and their related Diffraction Patterns were taken at the Sample Edge. Dark
field weak beam imaging has advantages in the accurate characterization of dislocations since the images produced by
this technique meet all basic requirements for dislocation study.

INTRODUCTION
Dislocation is the most important defect of modern engineering materials involving structural and electronic
application. Proper operation and understanding of dislocation also serve the basis for other characterization work.
This paper deals with both underlying principles and practical applications of dislocation characterization.

BACKGROUND THEORY

Geometry of Diffraction
Diffraction takes place when the reflecting sphere meets the reciprocal lattice. In term of electron diffraction,
reciprocal lattice is not a group of points but a group of small rods whose shape varies with the sample thickness.
FIGURE 1 represents the case where the reflecting sphere does not intersect but close to the reciprocal lattice. The
 
nearby reciprocal lattice is not in a diffraction condition due to g ≠ k . However, diffraction can actually occur
where a certain condition is met.
    
( )
ϕ g = ϕ 0 ∑ f j exp 2π ik ⋅ r ; k = g + s
atoms
(1)

   
ϕg = ϕ0 ∑ f exp (2π i ( g + s ) ⋅ (r + r ))
j j n (2)
atoms
# &
%       (
ϕ g = ϕ 0 ∑ % ∑ f j (exp 2π i ( g + s ) ⋅ rj exp 2π i ( g + s ) ⋅ rn ) ( (3)
all unit cell % all atoms (
$ /unit cell '

1st International Seminar on Advances in Metallurgy and Materials (i-SENAMM 2019)


AIP Conf. Proc. 2262, 050009-1–050009-10; https://doi.org/10.1063/5.0016430
Published by AIP Publishing. 978-0-7354-2022-9/$30.00

050009-1
FIGURE 1. Relationship between the reciprocal lattice and reflecting sphere when Bragg’s Law is not exactly satisfied

 
Since s ≈ 0 and ∑f j exp ( 2π ig ⋅ rj ) = Fg (structure factor )
j
    
ϕ g = ϕ 0 Fg ∑ (exp 2π i ( g + s ) ⋅ r ); n g ⋅ rn = integer (4)
allunit cell
         
ϕ g = ϕ 0 Fg ∑ exp 2π is ⋅ rj ; s = sx a* + sy b* + sz c*; rj = ua + vb + wc (5)
allunit cell

ϕ g Fg tx ty tz !
= ∫ ∫ ∫ 2π i ( sxu + sy v + sy w)#$dxdydz
ϕ 0 Vc 0 0 0 "
(6)

ϕ g Fg sin sx π t x sin sy π t y sin sz π tz


= (7)
ϕ 0 Vc π sx π sy π sz
As the sample is extremely thin in z direction, t z << t x , t y
ϕ g Fg sin sx π t x
≈ (8)
ϕ 0 Vc π sx
It can be seen from the final equation that the diffraction intensity varies with thickness and the deviation
parameter s. For a fixed s, the intensity is a regular function of the sample thickness. The maximum intensity
happens when st = 2n (n = integer) or t = 2n/s. This result also indicates that the diffraction can occur even though
the reciprocal lattice point is not exactly on the reflecting plane because the point becomes a rod-like extended along
a z-direction.

Electron Scattering in a Crystal


Dislocation, displacement of an atomic plane by amount of burgers vector, will cause the electrons wave to
experience different scattering and hence produce a characteristic contrast.
It has been previously discussed that the electron scattering in crystal is dictated by:
   
ϕg ~ ∫ {exp (2π i ( g + s ) ⋅ (r + r ))}dv
i n (9)
where:

ri represents location of unit cell in a crystal.

050009-2

rn represents position of atom within a unit cell.
 
g and s are dictated by TEM condition.
 
Effect of ri is explored in the characterization of thickness fringe. Effect of rn is now going to be employed to
produce contrast in dislocation characterization.
 
In perfect crystal, rn determines structural factor, Fg . When dislocation exists rn is locally disturbed.

Visibility of Dislocation

In perfect crystal, lattice planes, which are represented by g = [ hkl ] , do not provide contrast. When the atomic
arrangement of atom in a crystal is locally disturbed this is not the case. In general, this situation can be divided into
 
two extreme conditions. First is the case when the burger vector b is parallel to g (perpendicular to the plane) and
 
the second case is when the burger vector b is perpendicular to g (parallel to the plane). In both situations, the
lattice structure is perfect but for the core of dislocation. At the core of dislocation an atom is displaced by the
 
burger vector b . Thus, rn can be considered affected by the vector. The equation (9), therefore, can be modified.
   
ϕg ~ ∫ {exp (2π i ( g + s ) ⋅ (r + b ))}dv
i (10)
   
Since s ⋅ ri and s ⋅ b are both too small to have a significant effect, the equation can be simplified as:
  
{ (
ϕ g ~ ∫ exp 2π i ( g ⋅ ri ) exp 2π i ⋅ b dv )} (11)

  
As g ⋅ ri = integer , the intensity of diffraction is mostly affected by the presence of b .

FIGURE 2 Burger Vector in Edge Dislocation Relative to g vector

 
When g ⋅ b = 0 there will be no intensity variation caused by the dislocation core because the condition is the
 
same as that in perfect crystal. When g ⋅ b ≠ 0 , the intensity variation will occur.
It can be learned so far that the visibility and invisibility of dislocation is dependent upon the choice of the
diffraction condition. Applying this condition, burger vector of any unknown dislocation can be theoretically
specified, and hence the type of dislocation.

050009-3
In general, two types of dislocations exist, screw dislocation and edge dislocation. Since in screw dislocation
 
b / / in edge dislocation b ⊥  , photographic analysis can be employed to determine the type of dislocation
without any difficulty.

EXPERIMENT
Thin foil of electronic grade Cu is loaded into a double tilt specimen holder of Analytical TEM, JEOL JEM
1200EX. [001] Low Index Zone Axis was located with the aid of Kikuchi Lines. Two-Beam Condition was

achieved for !" 200#$!" 220#$ g vector. Bright Field and Dark Field Image of the Dislocation and their related
diffraction patterns were taken at the sample edge.

RESULTS AND DISCUSSION

Dislocation Image at Visible and Invisible Condition


A low zone index axis was located with the aid of Kikuchi Lines and indexed properly to specify two g vectors
that were used to achieve two-beam condition. [001] zone index axis is characterized by square symmetrical
diffractions pattern with 1:1 aspect ratio[1]. Indexing was consistently done referring to the standard indexed
Diffraction Patterns[2].
Two-beam condition was achieved, following the procedure provided by William and Carter[2]. In this
experiment, deviation vector s was not set equal to zero (exact Bragg condition) but “Small Positive”. The condition
was selected as it was revealed in William and Carter[2] that defects are best revealed in small positive two beam
conditions. The condition was achieved by finely tilting the specimen with the aid of Kikuchi Lines. Crystal
geometry, Intersection of Kikuchi Lines with Diffraction Spots, and Relationship between the deviation vector for
exact Bragg condition and “Small Positive s” rotated crystal are illustrated in FIGURE 5[1].

200 220

001

FIGURE 3 SADPs of [001] Zone Axis with 2 g vectors and corresponding Standard Indexed DPs of [001] FCC

050009-4
FIGURE 4 Variation in the diffraction contrast when s is varied from (A) zero to (B) small and positive and (C) larger and
positive. After[2]

FIGURE 5 (a) Geometry of crystal before and after rotation by angle φ, and position of Kikuchi lines with respect to diffraction
spots. (b) The Kikuchi line intersects the g-diffraction spot when the specimen is at the exact Bragg orientation (left), but is
displaced by x when off the exact orientation (right). (c) Relationship between the deviation vector, s, for rotated crystal (right),
and for exact Bragg orientation (left)[1]

050009-5
g = [ 2 00 ]

FIGURE 6 SADPs of Two Beam Condition for g = [ 200 ] with “small positive s”

g = [ 2 00 ] g = [ 2 00 ]

FIGURE 7 Bright Field and Dark Field Image at g = [ 200 ] Two-Beam Conditions

050009-6
g = [ 2 20 ]

FIGURE 8 SADPs of Two Beam Condition for g =[ 220 ] with ‘small positive s”

g = [ 2 20 ] g = [ 2 20 ]

FIGURE 9 Bright Field and Dark Field Image at g = [ 220 ] Two-Beam Conditions

050009-7
Bright Field Image and Dark Field Images of Dislocations were taken for 2 g vectors (FIGURE 7 and FIGURE
9) in small positive s two-beam condition (FIGURE 6 and FIGURE 8) at the same location. The complimentary of
Bright Field and Dark Field were confirmed by taking image of dislocation at the edge of sample in which thickness
fringes were revealed.
Dot product of the Diffraction vector (g) and the Burger vector (b) of dislocation has been used as the criteria for
dislocation contrast visibility and invisibility. For perfect dislocations, it is generally accepted that if the burger
vector b of the dislocation is normal to diffraction vector g, the diffraction will not be visible because no diffraction
contrast is arose from the dislocations[1-3]. In practice, however, the dislocation can be invisible when the g.b is
1
small but not necessarily zero. The rule of thumb is that the dislocation will be invisible when g ⋅ b ≤ . In edge
3
dislocation, the displacement vectors δ rn will determine the conditions in which dislocations are completely out of
contrast. In screw dislocation the g.b=0 is a sufficient to condition for invisibility of dislocation as the displacement
vector is always parallel to the burger vectors. For partial dislocations, the conditions are more complicated due to
dynamical effect[3]. Thomas and Goringe[3], for example, revealed that for fcc structure, partial dislocations can be
2
visible or invisible when g ⋅ b ± depending upon deviation vector and sample thickness.
3
Schematic diffraction contrast of edge and screw dislocations have been revealed by Fultz and Howe[1] as
depicted in the figure. Effect of g vector on the visibility and invisibility of screw dislocations was well illustrated
by Goodhew et.al[4]. It was also learnt from our experiment at small positive s two beam condition that the
visibility of dislocations (the diffraction contrast of dislocation) were changed when the orientation diffracting
planes (g vector) were changed as illustrated in FIGURE 12.

FIGURE 10 Schematic diffraction contrast for Edge (Left) and Screw (Right) Dislocations (After [1])

FIGURE 11 Images of screw dislocation arrays in a Niobium stabilized steel (After [4])

050009-8
g = [ 2 00 ] g = [ 2 20 ]

FIGURE 12 Bright Field Image taken at the same location in small positive s two-beam condition showing different visibility of
dislocations.

Weak-Beam Dark Field Imaging and Their Advantages


Generally speaking Weak Beam Dark Field Imaging refers to the formation of diffraction contrast image when
the useful information is transferred by weakly excited beam[2]. The condition is illustrated in FIGURE 13, in
which the 3g is excited and the image is formed by the weak beam g. Detailed experimental procedures to obtain
weak beam condition are revealed in Thomas and Goringe[3] and Williams and Carter[2].

FIGURE 13 Ray Diagram and Corresponding Diffraction Patterns (After [3,4])

050009-9
Weak-Beam Dark Field Imaging provides higher resolution of diffraction contrast and hence very useful in
dislocation study. The improvement of the resolution can be explained by Dynamical Theory[4]. Referring to
Goodhew et.al.[3], the effective extinction distance for non- zero s is:
1 1
ξ eff = = −1
(12)
s' ') ! 1 $+) 2
2
( s + ## 2 &&,
)* " ξ g %)-

When s large, the width of dislocation image is approximately one-third of the effective extinction distance. At
5
deviation vector s = , which is readily obtained, the image width will be one-fifth of that at s = 0. The
ξg
significant improvement was illustrated in FIGURE 14.

FIGURE 14 A Bright Field (a) and Dark Field Weak Beam Image of Dislocation in Silicon (b) g.b = 2 (c) g.b = 1. Note the
decrease of dislocation width and the termination of thickness fringe at the end of dislocation [4].

According Thomas and Goringe[3], images for dislocation study must meet four requirements. Firstly, each
dislocation should produce a single peak that is very close to the position of the dislocation core. Secondly, the
criterion of standard invisibility should distinctly apply to the image peak of dislocation. Thirdly, the image half
width should approach the capabilities of instrument. Finally, the image should not be sensitive to minor changes in
foil thickness, dislocation depth, lattice orientation, etc. Assuming certain conditions of diffraction are fulfilled, it is
assumed that images produced by dark field weak beam imaging will meet all four requirements. Therefore, it is
obvious that dark field weak beam imaging has advantages in the accurate characterization of dislocations.

REFERENCES
1. B. Fultz and J. Howe, Transmission Electron Microscope and Diffractometry of Materials 3rd Ed., (Springer,
Berlin, 2007).
2. D. B. Williams and C. B. Carter, Transmission Electron Microscopy: A Textbook for Materials (Springer,
Berlin, 2009).
3. G. Thomas and M. J. Goringe, Transmission Electron Microscopy of Materials (John Wiley and Sons Inc.
(Reprinted by TechBooks), New Jersey, 1981).
4. P.J. Goodhew, J. Humphreys, and R. Beanland, Electron Microscopy and Analysis 3rd Ed. (Taylor and Francis
Book, London, 2001).

050009-10

You might also like