| |
Pictures of Grain Boundaries
|
| |
Old
Pictures Taken be Me |
 |
Here are a few
transmission electron
microscope (TEM) pictures of grain boundaries in silicon (Si) that I took
around 1980. The HRTEM pictures were among the very first pictures ever taken
at high resolution. Looking at grain boundaries
edge-on you won't see
much, so the first examples are "top-down". |
|
 |
You are looking right on the grain
boundary. The silicon above and below the boundary is pretty much invisible.
It's like looking at an old-fashioned slide. The glass plates on top and bottom
of the film containing the picture are invisible.
These grain boundaries were artificially made (by me) by "welding" two single
crystals of silicon with a certain misalignment. This needs high temperatures
and pressure, and is not unlike the "hammer
welding" of two pieces of steel. In either case you produce a grain
boundary with some inclusions from the "dirt" still present on the
surface of the two pieces to be joined. In the case of silicon, the
"dirt" would be silicon dioxide (SiO2), in the
case of iron, we call it iron oxide, scale or slag. |
|
|
|
|
|
 |
| Grain boundary in Si made by welding. |
| It's low-angle twist boundary on a
{111} plane, if you
must know. |
|
|
|
|
|
|
There is a lot of structure on a rather small
scale. The lines forming a kind of six-fold pattern are "grain boundary dislocations"
with two kinds of embedded stacking
faults. You really don't want to know more about this.
The smooth blobs (one is marked "slag") are amorphous silicon dioxide
(SiO2) particles left over from the welding process—just like
real slag particles are always found in hammer welded blades. Amorphous silicon
dioxide, by the way, is just quartz glass; if it would be crystalline we call
it rock crystal. |
|
Here is a picture of
a very similar grain boundary: |
| |
|
|
|
 |
| Grain boundary in Si made by welding |
| Low angle twist boundary on
{100}, if you must
know. |
|
| |
|
|
|
|
Same thing once more. Grain boundary dislocations
forming a very regular pattern; hardly any "slag". |
|
Here are two pictures of naturally
occurring grain boundaries. They are all inclined to the viewing
direction. |
| |
|
|
|
 |
| Large-angle grain boundary in silicon. |
|
| |
|
|
| |
|
The "chicken wire" structure indicates
a network of very special grain boundary dislocations. The big black line is a
"real" dislocation, running through one of the grains and ending at
the grain boundary, interacting with the dislocations there. |
| |
|
|
| |
|
 |
| Grain boundary junction in Si |
|
| |
|
|
|
|
A junction of three large-angle grain boundaries.
The lower one has clearly visible grain boundary dislocations With the eye of
faith one also sees a fine-mashed network in the one branching off to the left.
The one going up appears to without a structure (the "zebra" fringes
have nothing to do with structures in the boundary) but that might simply be
due to the limitations of the electron microscope. |
|
Just for the hell of it, here are
high–resolution transmission electron microscope (HRTEM) pictures at
atomic resolution. Those pictures are among the very first ones taken with
atomic resolution around 1979, when electron microscopes became powerful enough
for that. |
|
|
|
|
|
 |
| HRTEM picture of the grain boundary
above. |
| Visible are 5 screw dislocations, causing the
typical shift of the lattice planes. |
|
|
|
|
|
|
What you see are "screw dislocations".
Look up the "dislocation
science" module if you feel you need to know what "screw
dislocations" are. Otherwise screw them. |
|
Here, just for the hell of it, is an
"edge-on" picture at atomic resolution of the ("small-angle
twist") grain boundary in the top most
picture. There is really not much to see. |
|
|
|
|
|
 |
| HRTEM picture of the grain boundary
above. |
|
| |
|
|
|
|
Things get a bit better with "edge-on"
pictures at atomic resolution of some slightly different kind of grain boundary
("small-angle tilt"). |
| |
|
|
|
|
 |
| HRTEM picture of a small-angle tilt grain boundary.
|
|
| |
|
|
|
|
The boundary runs from left to right in the
middle of the picture. The colored lines are only to guide the eye. The blue
lines indicate that the orientation of the crystal above or below the boundary
differs indeed by a "small angle".
The red lines indicate ending lattice planes, i.e. edge dislocations.
The picture shows directly (and for the first time) that this kind of boundary
consists indeed of a lot of dislocations in some special array. This was
predicted long before it could be imaged. Now the prediction has been
proved. |
|
The next picture shows some unusual
and unexpected behavior that could only be found with "edge-on
HRTEM". A simple (low-angle tilt) boundary is actually not so simple but
consists of three boundaries close together. |
| |
|
|
|
 |
| HRTEM picture of complex boundary consisting of three
boundaries.. |
|
| |
|
|
|
|
A low-angle tilt boundary composed of individual
dislocations as in the picture above is actually sandwiched between two
so-called "twin" boundaries, the effects of which cancel each other.
That is shown by the black lines that bend substantially at the twin
boundaries, but in opposite directions.
The misorientation between the upper part and the lower part of the crystal is
only determined by the low-angle tilt boundary. It changes position from
in-between the twin boundaries to being superimposed on one of the twins. |
|
|
This effect would be easily missed looking at the
grain boundary "top-down", so edge-on views do have some merits |
|
So what are you supposed to learn
form all this stuff? Not much, really. The messages are simple:
- The internal structure of grain boundaries is very complex. Trying to
understand in detail what one sees on those pictures will take a lot of time
and effort.
- Nevertheless, it is great fun (not to mention a lot of work) to take
pictures like the ones above.
|
| |
|
|
© H. Föll (Iron, Steel and Swords script)
With frame
9.1.1 Things are Complicated