Dislocation Etch Pit Formation in Lithium Fluoride
J. J. Gilman, W. G. Johnston, and G. W. Sears
Citation: Journal of Applied Physics 29, 747 (1958); doi: 10.1063/1.1723277
View online: http://dx.doi.org/10.1063/1.1723277
View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/29/5?ver=pdfcov
Published by the AIP Publishing
[This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded
to ] IP: 128.151.161.43 On: Thu, 14 Nov 2013 18:17:35
Journal
of
Applied Physics
Volume 29, Number 5
May, 1958
Dislocation Etch Pit Formation in Lithium Fluoride
J. J. GILMAN, W. G. JOHNSTON, AND G. W. SEARS
General Electric Research Laboratory, Schenectady, New York
(Received January 27, 1958)
Dislocation etch pits can be formed on LiF by a dilute aqueous solution of FeF 3 • In this report the etch pit
formation is described in detail, and the mechanism for pit formation is discussed. The nature of the etch
pits depends on the character of the dislocation, and on the exact composition of the etchant. Edge dislocations and screw dislocations etch slightly differently; the former produce deeper pits. The etching is
inhibited by some segregated impurities at dislocations, therefore aged dislocations and fresh dislocations
etch much differently. Etch pit formation is probably due to the preferential nucleation of' unit pits one
molecule deep at a dislocation, and the movement of the monomolecular steps across the surface. The
relative rates of these two processes determine the shape of the etch pits. The nucleation rate for unit pits
depends upon the dislocation energy, hence upon the character of the dislocation and the impurity content
as suggested by Cabrera. The nucleation rate is faster at edge dislocations, because of their higher energy.
The nucleate rate is low at dislocations with segregated impurities, because the impurities lower the dislocation energy. The ferric ion is adsorbed on the surface and inhibits the motion of steps, so that steeper,
more visible pits are produced as the iron content is increased.
I. INTRODUCTION
ETALLOGRAPHY is dependent upon the painM
staking discovery, by trial and error, of etchants
which reveal microstructures. The effective reagents
are often complex in composition, and there exists little
possibility of determining the detailed mechanism of
etching. When a particularly simple etching reagent is
encountered it merits a detailed study, because the
knowledge gained in studying a simple case may aid in
the discovery and understanding of other etching
reagents.
In recent years a number of reagents have been
reported which reveal dislocations in crystals. l One of
these etchants is a very dilute aqueous solution of
ferric fluoride which has been used by Gilman and
Johnston2 to study dislocation behavior in lithium
fluoride. Consisting of only a solvent and a very small
concentration of one impurity, this reagent is probably
the simplest reliable etchant for dislocations that has
1 A brief review of this work is presented by W. G. Pfann,
Solid State Physics, edited by F. Seitz and D. Turnbull (Academic
Press, Inc., New York, 1957), Vol. 4, p. 424.
2 J. J. Gilman and W. G. Johnston, Dislocations and Mechanical
Properties of Crystals (a report of the 1956 Lake Placid Conference)
(John Wiley and Sons, Inc., New York, 1957), p. 116.
been reported. The purpose of this paper is to describe
the mechanism by which this etchant operates.
It was reported in the earlier paper that the dilute
aqueous solution of ferric fluoride produced sharply
defined pyramidal pits on the (100) cleavage face of
lithium fluoride at the points of emergence of all "fresh"
dislocations. The fresh dislocations are ones produced
by room temperature plastic deformation. In contrast,
the aged dislocations which were present in the as-grown
and slowly cooled crystal prior to the deformation were
marked by shallow indistinct pits by the ferric fluoride
solution. It was suggested at that time that the difference in etching between fresh dislocations and aged
dislocations may be due to the segregation of impurities
around the latter. *
II. EXPERIMENTAL
The present experiments have been carried out on
the (100) cleavage faces of optical grade LiF crystals
* ]. J. Gilman and W. G. Johnston [reference 2 and J. Appl.
Phys. 27, 1018 (1956)J have described another etchant (1 part
conc. HF, 1 part glacial acetic acid, 1% HF saturated with FeF 3)
which does not discriminate between fresh and aged dislocations.
It etches both alike.
747
Copyright
© 1958 by the American Institute of Physics
[This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded
to ] IP: 128.151.161.43 On: Thu, 14 Nov 2013 18:17:35
748
GILMA~,
JOHNSTON,
FIG. 1. Dislocation etch pits on (100) cleavage face of LiF. The
pyramidal pits in the horizontal row mark the fresh dislocations
introduced by room temperature plastic deformation. Aged dislocations which were present in the as-grown crystal are located
along the sub-boundary running from lower right to upper left.
Magnification: 1200X.
obtained from the Harshaw Chemical Company,
Cleveland, Ohio.
The best dislocation etch pits are obtained with
abou t 2 X 10- 6 mole fraction of F eF 3 in distilled water.
A crystal is etched by immersing it in the etchant and
stirring vigorously. After one minute the crystal is
removed and rinsed in alcohol and ether in turn. This
treatment produces pyramidal etch pits with a uniform
width of about ten microns at the points where fresh
dislocations emerge from the crystal. Fresh dislocations
can be etched as soon as they are formed, so it is clear
that the etching does not depend on diffusion of
impurities to the dislocations. The aged dislocations
which are present in the as-grown crystal are marked
by shallow indistinctly shaped pits. The striking difference between the pits formed at fresh dislocations and
at aged dislocations is shown in Fig. l.
The pyramidal pits which form at fresh dislocations
have a square base with the edges parallel to the (100)
directions. The pits are about 15% as deep as they are
wide, so the sides do not correspond to low index
crystallographic planes. In Fig. 4 it can be seen that
the sides of the pits are not planar for some of the FeF3
concentrations that have been used.
From the shape of a pit, the orientation of the
dislocation with respect to the surface can be inferred
as described by Amelinckx. 3 The line drawn from the
center of the base of the pyramid to the apex of the
pyramid will be tangent to the dislocation line, e.g.,
if the pit is a symmetric right pyramid, the dislocation
line must be perpendicular to the surface. The glide
plane in LiF is the {1l0} plane and the glide direction
is (fIO). Edge dislocations will emerge at glide bands
which lie in a (110) direction, and the screw dislocations
will emerge at glide bands which are parallel to a (100)
3
S. Amelinckx, Phil. Mag. 1, 269 (1956).
AND
SEARS
direction. This is represented schematically in Fig. 2.
From the location and the shape of the pit, the character
of a dislocation can be determined.
The depth of an etch pit corresponding to a pure
screw dislocation [Fig. 2 (b) ] can be determined by
measuring the asymmetry of the pit since the screw
dislocation makes an angle of 45° with the surface. It
was found that the depth of the pits is approximately
0.15 times the width. If all the pits are assumed to be
this deep, it can be shown from the extreme asymmetry
of some pits, that the minimum angle that a dislocation
can make with the surface and still produce a pit is
less than five degrees. This estimate of minimum angle
is undoubtedly high, and it seems that all dislocations
emergent from the surface can be etched.
If a dislocation is moved during the etching, the
original pit will get no deeper, but will continue to
grow laterally producing a pit with the shape of a
truncated pyramid. 2
Effect of Stirring
The size of the etch pits given above for a one
minute etch results only if the solution is vigorously
stirred during the etching. If the crystal is placed in an
unstirred solution, the etch pits are much smaller and
slightly rounded, as shown in Fig. 3.
y
I
I
x-~-~-w
OJ
I
5
(e)
X
I
i
<])$
e
y
i'
FIG. 2. (a) Drawing of a cleaved LiF crystal showing the (110)
glide plane, marked WXYZ, and the [1I0] Burgers vector, h.
In the glide plane are shown an edge dislocation, ee', a screw
dislocation, ss', and a dislocation of intermediate character, ii'.
(b) Two dislocation pits on glide band XW. The pit on the left
corresponds to the pure screw dislocation, and the pit on the
right corresponds to the intermediate dislocation. (c) Two dislocation etch pits on glide band XV. The pit on the left corresponds
to the pure edge dislocation, and the one on the right corresponds
to the intermediate dislocation.
[This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ]
IP: 128.151.161.43 On: Thu, 14 Nov 2013 18:17:35
ETCH
PIT FORMATION
IN
LiF
749
Effect of FeF 3 Concentration
The behavior of the etching reagent is very sensitive
to the concentration of FeF 3 • Figure 4 shows the effect
of varying the amount of FeF 3 in the etchant. Distilled
water produces very wide, indistinctly shaped pits.
As the iron concentration is increased the pits become
smaller and have steeper sides, attaining a plane-sided
pyramidal shape at the concentration of 2X 10- 6 • As
the FeF 3 concentration is increased beyond this amount
the pits become still smaller, and slightly rounded.
Ca)
Effect of LiF Concentration
The etching behavior described above is for a fresh
solution of FeF 3 in H 20, using 200 cc of solution to etch
a crystal with about 1 cm2 of surface area for one or
two minutes. If this procedure is repeated, using a
fresh solution after each one minute etch, the etch pit
size increases linearly with the total etching time. If
the same solution is used for repeated etching, the
etching rate will slow down.
LiF solutions were made up with various degrees of
under saturation, and it was found that the amount of
FeF 3 required to produce plane-sided pyramidal pits
decreases with increasing LiF concentration. Figure 5
shows a photomicrograph of the etch pits produced
(b)
(e)
(d)
FIG. 4. Effect of FeF 3 concentration on etch pit shape. All
crystals were etched for two minutes. (a) Distilled water. (b)
O.SX 10- 6 mole fraction of FeF 3 • (c) 2X 10- 6 mole fraction of FeF 3 •
(d) 8XlO- 6 mole fraction of FeF 3 • Magnification: 1000X.
Ca)
Cb)
FIG. 3. Effect of stirring on etching rate of dislocations. (a)
Crystal revolved in lO-cm diameter circle at 1 rev per second. (b)
Not stirred. Magnification: 800X.
with a solution 15% saturated with LiF and contained
0.75X 10- 6 mole fraction of FeF3. The etching rate was
only one-tenth as fast as with a solution of the same
FeF3 concentration, but with no LiF in solution. A LiF
solution of 25% saturation did not preferentially etch
dislocations.
Above 10% of LiF saturation the etchant distinguishes between the dislocations which are predominantly of screw character and those predominantly
of edge character. This is shown quite clearly in Fig. 5,
where the pits of the (110) glide band which contains
the edge dislocations, are much deeper than the pits
on the (100) glide band which contains the screw dislocations. Although the widths of all the pits are the
same, the edge dislocations produce deeper pits.
[This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to
] IP: 128.151.161.43 On: Thu, 14 Nov 2013 18:17:35
750
GIL l\!l
~'\
l\'",
J0
H N S TON,
~,\)i])
S E :\ R S
cooled crystals. This behavior will be discussed in a
later paper, and it suffices here to say that the mechanical behavior implies that a precipitation of impurities
is occurring below 200°e.
Anisotropy of Dissolution Rate
FIG. 5. Difference in etching at screw and edge dislocations.
Edge dislocations lie along the horizontal and vertical lines; while
the lines of screw dislocations are inclined at 45°. The widths of
all pits are the same, but the edge dislocations produce deeper
pits, as indicated by their higher contrast. 800X.
The effect of the FeF" which is present in the etching
solution on the dissolution rate of LiF was investigated
by comparing the rates of dissolution of LiF spheres
in a polishing solution and in the etching solution.
Spheres of one-half inch diameter were ground from
LiF single crystals. The diameters of the spheres were
measured, and then they were immersed in a polishing
solution or an etching solution. The polishing solution
consisted of H 20 plus 1.5% NH 4 0H.2 and the etching
solution contained 2X 10- 6 mole fraction of FeF3 in
H 20. In each case, the solution was circulated through
several jets in the bottom of the container in such a
Difference Between Fresh and Aged Dislocations
Two experiments were performed to determine the
effect of impurity segregation at dislocations on etch
pit formation.
(a) Fresh dislocations were deliberately aged by
heating to 300 0 e and cooling slowly to room temperature at the rate of 4°e/hr.
(b) An as-grown crystal containing aged dislocations
was heated to 300 0 e in order to "boil off" the impurities from the dislocations, and the crystal was withdrawn from the furnace into the room atmosphere to
cool it rapidly enough to prevent the segregation of
impurities.
The results of the two experiments are shown in
Figs. 6 and 7. Figure 6 compares the two halves of a
crystal which was deformed and cleaved. One half was
heated to 300 0 e and slowly cooled; the other half was
untreated. Both halves were then etched. The fresh
dislocations in the unheated crystal give pyramidal
pits. The same dislocations in the heated and slowly
cooled crystal yield very shallow, indistinct pits. It is
clear that the dislocations are still present in the glide
bands of the heated piece, but the etching is very much
inhibited.
A crystal that had been slowly cooled after growth
was cleaved in half; one half was heated to 300 0 e and
air cooled, and the other half was not heated. Both
halves were then etched. It is seen in Fig. 7 that disloca tions in the original crystal produce shallow, indistinct pits that are characteristic of aged dislocations,
while in the rapidly cooled crystal the same dislocations
have produced distinct pyramidal pits.
A study of the mechanical properties of the LiF
crystals shows that the strength is very sensitive to the
rate of cooling from 200 0 e down to lOO°e. The slowly
cooled crystals are several times stronger than the air
(a)
(b)
FIG. 6. Effect of aging on etching behavior of dislocations. A
crystal was deformed at room temperature and cleaved in two.
(a) One piece, as deformed and etched. (b) Matching crystal face
of second piece, which was heated to 300°C and cooled slowly
before etching. Photograph reversed in printing. Magnification:
400X.
[This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to
] IP: 128.151.161.43 On: Thu, 14 Nov 2013 18:17:35
ETCH
PIT
FORMATION
manner that the sphere was supported by the fluid
jets and tumbled rapidly in a random fashion. The
solutions were replaced from time to time to keep
them fresh, and the spheres were measured at intervals
to determine the radial rate of dissolution.
The results shown in Table I indicate that the rate
of solution was much higher and more uniform in the
polishing solution than in the etching solution. The
effect of the FeF 3 is to reduce the dissolution rate by
a factor of ten, and to make the dissolution rate
anisotropic. The dissolution rate is highest in the (111)
direction and lowest in the (100) direction.
It was observed that square etch pits formed in the
(100) directions and triangular etch pits formed in the
(111) directions, but no etch pits formed in the (110)
direction on the spheres when they were immersed in
the etching solution.
Chemistry of the Etchant
The small amount of FeF;l in the etching solution
controls the etching behavior. The chemical role of the
impurity was investigated by studying: (a) the effect
(a)
TABLE
IN
I. :\Iacroscopic dissolution rates of LiF spheres.
Solve-nt
Direction
Dissolution fatt>
(A/>ec)
Polishing solution
(10 hours)
100
110
111
910
960
964
Etching solution
(30 hours)
100
110
111
91
109
116
TABLE
II. Effect of anions on etching of LiF.
Salt
Etching
Optimum Fe3')
( concentration
FeF 3
FeCL
Fe(N0 3)"
FeBr"
Excellent
Excellent
Excellent
Excellent
1.2 X 10- 6
4.7X 10- 6
9.8XlO- 6
21.0XlO- 6
of changing the anion of the salt ; (b) the effect of
adding to the etching solution anions which form
complexes with the FeH ion; (c) the effect of varying
the cation. From the results of (a) and (b) it can be
inferred that the cation, Fe H , in the etching solution
is of major importance, the role of the anion is minor.
In (c) data was obtained that might give some insight
into the requirements that the cation must satisfy to
produce a good etchant.
Solutions were prepared of each salt, with concentrations ranging from about 10- 7 to 3X 10- 5 mole fraction
in 7 or 8 steps. A crystal was immersed in each reagent
for one minute, rinsed, and examined microscopically.
In order to test the importance of the anion, four
ferric salts were used to make etching solutions. Table
II shows that all four salts produced good etchants,
but that the optimum concentration of ferric ion varied
among the salts. Apparently the role of the anion is a
minor one, although it is not completely negligible
An etching solution containing the optimum amount
of FeF3 was modified by adding anions which form
stable complexes with the ferric ion. The results of this
experiment are summarized in Table III. It is seen
that the anions that form ferric complexes more stable
than the fluoride complex disrupt the etching action.
It was shown above that CI- and Be have little effect
(bl
FIG. 7. Effect of heal treatment on the etching behavior of
grown-in dislocations. An as-grown crystal was cleaved into two
halves. Ca) As-grown crystal, etched. Photograph reversed in
printing. (b) Matching crystal face of the second piece which was
heated to 300°C and air-cooled prior to etching. Magnification:
400X.
751
LiF
TABLE
a
Salt
.\nioll
NH,OH
LiOH
NaCN
NaC 2O,
NaP
FeCI"
FeBr-
OHOHCN!C 20,--
J.
F-
Cl-
Er-
III.
Stability
of iron
complexa
9.3
9.3
~7
~5.5
4.2
1.5
0.6
Concentration
to stop f'teh
pit formation
4.7XlO- 7
1.1 X 10- 6
1.1 X 10- 5
2.3XlO- I1
1.3X 10-"
Bjerrurn. Chern. Revs. 46, 381 (1950).
to show that the effect of the Na in the two previous
b NaF was inducted
~alts is n('gligible.
[This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to
] IP: 128.151.161.43 On: Thu, 14 Nov 2013 18:17:35
752
GILMAi\, JOHNSTON, AND SEARS
TABLE IV. Various cations as dissolution inhibitors in etching reagent.
Cation
Fe3+
AP+
Mn2+
\'2+
Mg2+
Zr2+
In3+
Ti3+
Pt3+
TI2+
Ca2+
SrH
BaH
Y3+
NiH
Cr3+
La3+
Sb3+(?)
PbH
Na+
ZnH
Cd H
C02+
HgH
Cu2+
Bi3+(?)
Ga3+
UREA
EDTA
H+
Salt
added
Fluoride
Fluoride
Chloride
Chloride
Chloride
Chloride
Chloride
Chloride
Chloride
Chloride
Fluoride
Chloride
Fluoride
Chloride
Chloride
Nitrate
Nitrate
Fluoride
Fluoride
Fluoride
Chloride
Chloride
Chloride
Chloride
Chloride
Chloride
Chloride
Fluoride
Etching
behavior
Very good
Very good
Very good
Good
Fair
Fair
Fair
Fair
Poor
Poor
Poor
Poor
Poor
Poor
Poor
Poor
Poor
No good
No good
No good
No good
No good
No good
No good
No good
No good
No good
No good
No good
No good
Optimum
concentration
1.2 X 10-6
~1.6XIQ-6
1.4X1(r~e
5.6XIQ-6
2.7XIO-3
~2.3XIQ-5
Fluoride
heat of
formation a
(kcal/mole)
Fluoride sol.
product
-243
-311
-260
~2.7XIQ-7
-263
-350
6.4XIQ-~
1.9XlO-5
~1.5XlO-5
-287
-290
-286
4.0XlO-u
2.8XIQ-9
1.7XlO-s
...:. i71
-265
-216
-156
-136
-188
-174
-159
-127
(A)
0.60
0.50
0.80
0.66
0.65
Complex stability
Constant
Bailarb
4.2
5.1
0.81
0.69
-315
-77
Ionic
radius
3.7XIQ-s
1.44
0.99
1.13
1.35
0.93
0.70
0.64
1.15
0.76
1.21
0.95
0.74
0.97
0.72
1.10
0.72
0.93
0.62
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
No
No
No
No
No
Yes
",2
1.5
<4
5.3
Yes
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
, Compare with value for LiF of -145 kcal/mole.
b Opinion expressed in J. C. Sailar, Chemical Coordinate Compounds (Reinhold Publishing Corporation, New York, 1956), P. 1.
e This relatively high concentration suggl'sts that Fe impurity in the Mn salt may be the true inhibitor.
on etch pit formation; therefore, the anions which
form complexes less stable than the fluoride do not
affect the etch pit formation.
Since the impurity cation seems to be the dominant
ingredient in the etching solution, various cations
were tested to determine what requirements a cation
must satisfy in order to produce good etch pits. The
following factors were considered: (a) valence of cation;
(b) size of cation; (c) stability of its fluoride salt; (d)
solubility of its fluoride salt; (e) stability of its fluoride
complex.
Table IV lists the various salts that were tried, the
effectiveness of the salt in producing pits, and the
properties of the salt. It is seen that FeH is not unique;
AlH forms just as good an etchant.
Although it seems clear that valence does not control
the effectiveness of a cation, one cannot draw strong
conclusions regarding the other factors from Table IV
alone. It appears that the cations which produce the
better etch pits are within 25% of the size of the Li+
ion, they have a stable fluoride salt with low solubility,
and they form a stable fluoride complex. The last point
combined with the data of Table III indicates that
fluoride complexes are important in the etching process.
m.
DISCUSSION
The growth and dissolution of crystals are thought
to occur by the advance and retreat of monomolecular
steps across the surfaces of the crystal. 4 The active
sites are taken to the places along the steps where
single molecular rows end. These places, called "kinks,"
are where individual molecules may be deposited or
removed.
When a perfect crystal face is exposed to a solvent,
dissolution probably begins by the nucleation of "unit
pits" of one molecule depth.t These unit pits grow as
steps retreat across the crystal through the action of
the kinks, and this process is schematically represented
in Fig. Sea) and (b). On such a perfect crystal face the
nucleation of unit pits is random. On a real crystal
dislocations may be preferential sites for nucleation of
unit pits, and repeated nucleation at a dislocation leads
to the formation of an etch pit. This process is quite
different from the formation of a pit at a screw
dislocation by reversal of the growth process proposed
by Frank. 6
Cabrera 6 has recently considered the formation of
dislocation etch pits by evaporation. Cabrera empha4 See, for instance, Burton, Cabrera, and Frank, Phil. Trans.
Roy. Soc. London, A234, 299 (1951).
t For a large crystal the formation of steps at the edges can be
neglected for dissolution far from the edges.
• F. C. Frank, Discussions Faraday Soc. 5, 74 (1949).
6 N. Cabrera, Semiconductor Surface Physics, edited by R. H.
Kingston (University of Pennsylvania Press, Philadelphia, 1957),
p.327.
[This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded
to ] IP: 128.151.161.43 On: Thu, 14 Nov 2013 18:17:35
ETCH
PIT FORMATION IN LiF
sized the role of the dislocation energy in the nucleation
of pits; and he pointed out that the effectiveness of a
dislocation as a nucleation site depends on, among
other factors, the character of the dislocation and the
impurity content. The dislocation energy that contributes to nucleation of a unit pit is that energy which
is localized near the dislocation line. This "localized
energy" of a dislocation consists of the core energy
plus a small fraction of the total elastic strain energy.
The formation of visible dislocation etch pits depends
on the nucleation rate for unit pits at a dislocation,
and the rate of motion of the steps across the crystal
surface. These two quantities are reflected in the linear
solution rates normal to the surface at a dislocation,
and parallel to the surface; they are labeled Vn and v,,
respectively, in Fig. 8(c). If vn«v. very shallow pits
would be formed which could not be seen because they
would lack contrast when illuminated. It is found that
readily visible pits can be obtained when vn/v.~O.1.t
The problem of developing a suitable etchant to reveal
dislocations then can be solved by adjusting the ratio
of pit nucleation rate at a dislocation to the dissolution
velocity of steps. The ratio vn/v. can be increased by:
(a) increasing V n , as in the case of most etchants for
metals, which utilize a segregated impurity for this
purpose; (b) decreasing'll, by addition of an in~ibitor,
as is the case in the present work; (c) varymg the
temperature to take advantage of the difference in
activation energies of Vn and v•.
If a dislocation is moved during the etching process,
the solution rate normal to the surface, V n , at the
original position of the dislocation becomes negligible.
753
Since the velocity of the steps, v" remains the same,
the pit will assume a flat-bottomed shape as depicted
in Fig. 8(d).§
Etch Pit Formation in LiF
When a LiF crystal is immersed in pure water very
wide indistinct pits are formed as shown in Fig. 4(a),
because the steps retreat very rapidly from the point
of nucleation. In order to form more distinct pits it is
necessary to increase "lin or to decrease v•. By adding
ferric ions to the water, Vs can be reduced so as to
produce distinctly visible pits. The ferric ions probably
adsorb at newly formed steps and inhibit their motion.
For a given etching time, the etch pit width should
then decrease as the FeF 3 concentration is increased,
and this is experimentally observed (Fig. 4).
The present experiments indicate that the rate of
nucleation at a dislocation varies with the localized
energy of the dislocation. In Fig. 5, it is seen that an
edge dislocation etches faster than a screw dislocation.
The core energies and elastic strain energies of dislocations in NaCl have been calculated by Huntington,
Dickey, and Thomson. 7 who found that the core
energy and the elastic strain energy of the edge dislocation in the (110) glide plane were greater than the
respective energies of the screw dislocation. The results
of this theory are qualitatively the same for dislocations
in LiF, which would imply that the nucleation rate
should be greater at edge dislocations, in agreement
with experiment.
The results of the experiment summarized in Figs.
6 and 7 indicate that the presence of segregated impurities will lower the nucleation rate at a dislocation.
INTO
The localized energy of a dislocation is lowered by the
SOLUTION
segregated impurities, and when the crystal is heat
.J
treated so as to put the impurities back into solid
solution, the nucleation rate is increased again, and
deep etch pits are formed.
Since the ferric ions slow down the movement of the
1
J -•- - - - .•4
steps across the surface, the dissolution rate of a LiF
L ____
________
..t,
crystal should be slower in the etching solution than
(a)
(b)
in the polishing solution, and the experimental results
show that the dissolution rate is only one-tenth as
rapid in the etching solution. From the square shape
of the dislocation etch pits one may infer that the rate
of dissolution in the etching solution is slowest in the
(100) directions, which is observed.
Dislocation etch pits form to a depth of about 1.5
(d)
(c)
microns in one minute. This is also the rate at which
the (100) face of a large crystal is dissolved during
FIG. 8. (a) Unit pit one molecule in depth nucleated or: a
crystal face that has been exposed to a soIv~nt. (b) Enl:uged VIe~ chemical polishing. It therefore appears that the rate
of (a). (c) Etch pit formed by the successlve nucleatIOn. of umt of solution of a large crystal is limited by the rate of
pits at a dislocation. v.=velocity of steps; v.. =nuc~eatIon rate
x atomic spacing. (d) Shape assumed by an etch Pit after the nucleation of unit pits at dislocations.
dislocation has been moved away.
It has been established that the anion of the salt
FeF 3 has little effect on the behavior of the etching
t However, if Vn~v. the deep pits whifh are for~ed will ~nhibit
//
subsequent movement of the dislocatIOn, and In expenments
where the movement of dislocations is of interest it is advisable
work with the lowest value of vn/v, that will give visible pits.
§ See photograph of this effect in reference 2. Fig. 6.
7 Huntington, Dickey, and Thomson, Phys. Rev. 100, 1117
(1955).
[This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded
to ] IP: 128.151.161.43 On: Thu, 14 Nov 2013 18:17:35
754
GIL M .\ X,
J 0 H ~ S TON, .\ X n S E .\ R S
so the three remaining bonds would need to be satisfied
by three ions in the solution.
IV. CONCLUSIONS
FIG. 9. Cork ball model (to scale) of a kink on the surface of a
LiF crystal. Large gray balls are F- ions, small silver balls are
Tj+, and the small black ball is a Fe3+ ion adsorbed at the kink.
The Fe+ ion forms bonds with the three F- ions marked (X).
solution. The etching behavior is controlled by the
cation, and it appears that complexes like [FeF 6J3and [AIF 6Jl- are important in the process.
The specific effect of the inhibiting ions is to slow
down the movement of steps. It is suggested that this
comes about through adsorption of the cations at steps
and kinks. The most favorable sites for cation adsorption appear to be kinks. As the cork ball model of Fig.
9 shows, an ion with the size of Fe3+ could fit easily
against the three F- ions at a kink. The Fe3+ could then
form electrostatic bonds or coordination bonds with
the F- ions. If coordination bonds formed only three
of t he six bonds would be formed by t he surface ions,
The formation of dislocation etch pits in lithium
fluoride by a dilute aqueous solution of ferric fluoride
can be understood in terms of a preferential nucleation
of unit pits at dislocations, and the inhibition of the
motion of the edges of these pits by ferric ions. The
effect of ferric ions is to slow the motion of the steps
across the crystal, producing steep visible etch pits
instead of the wide indistinct pits that form in pure
water. The ferric ions appear to act by complexing
with fluoride ions of the crystal surface, protecting the
surface from dissolution.
Preferential nucleation of unit pits at a dislocation
is due to the energy (core energy and elastic strain
energy) that is localized there. The localized energy
depends on the character of the dislocation and the
impurity content; and, as suggested by Cabrera, these
factors affect the etching behavior. The pit nucleation
rate is slightly higher at an edge dislocation than at a
screw dislocation. Impurities which segregate at dislocations can reduce the nucleation rate.
The rate of dissolution of large crystals of LiF
appears to be limited by the rate of nucleation of unit
pits at dislocations.
ACKNOWLEDGMENTS
The chemical knowledge of Dr. A. E. Newkirk
contributed substantially to this investigation. We had
helpful conversations with Professor ~. Cabrera on the
theory of etch pit formation.
[This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded
to ] IP: 128.151.161.43 On: Thu, 14 Nov 2013 18:17:35