LabAdviser/314/Microscopy 314-307/SEM/Nova/Transmission Kikuchi diffraction: Difference between revisions
| Line 65: | Line 65: | ||
== Nanostructure evolution during heating == | == Nanostructure evolution during heating == | ||
Figures 3.12 and 3.13 show the IPFZ maps overlaid with pattern quality | |||
for temperatures varying from 20°C to 900°C. | |||
A preliminary investigation of the Au �lm at room temperature revealed a | A preliminary investigation of the Au �lm at room temperature revealed a | ||
| Line 70: | Line 73: | ||
range of 30 nm and large grains with size in the 150 nm range. The latter | range of 30 nm and large grains with size in the 150 nm range. The latter | ||
ones showed a strong [111] out-of-plane texture (Fig. 3.11) . | ones showed a strong [111] out-of-plane texture (Fig. 3.11) . | ||
During heating, the [111] grains tended to grow faster than the [110] | |||
and [100] ones. It is also possible to observe that grain growth started at a | |||
temperature below 150°C, while holes are visible at 170°C (highlighted with | |||
a red circle). The holes were formed in the vicinity of non-preferentially | |||
oriented (non-PO) [110] and [100] grains, which were also the site where the | |||
hole growth was continuing. The yellow rectangles follow instead the delayed | |||
hole growth due to the presence of preferentially oriented (PO) [111] grains: | |||
the hole is visible from 170°C and grows until it is completely surrounded | |||
by larger PO grains, subsequently its growth is retarded until 500°C, while | |||
other holes continue to grow. | |||
Since TKD measurements provided a large amount of data for each | |||
scanned point in the map (54000 total acquired points at the step size of 10 | |||
nm), statistically signi�cant quantitative data analysis could be performed | |||
from this series of measurements, showing again the big advantage of performing | |||
OIM inside the SEM rather than the TEM. The two classes of grains | |||
already observed in Fig. 3.11 were analyzed, de�ned as i) PO-grains with | |||
[111]//growth direction (using tolerance angle of 15°) and ii) non-PO grains | |||
with other orientations. The average grain size evolution of both classes of | |||
grains is shown in Fig. 3.14a for annealing temperatures up to 600°C. | |||
The graph con�rms the trend from the maps shown in Fig. 3.12, revealing | |||
that grain growth started already at 120°C. The PO grains were larger | |||
than the non-PO ones already from the starting nanostructure and grew | |||
considerably faster: by a temperature of 220°C, PO grains have almost triplicated | |||
their average size, while non-PO grains maintained their average size | |||
of 33 nm. Furthermore, up to 550°C practically only the PO grains grew. | |||
Fig. 3.14b shows the evolution of the number of indexed points for | |||
the two classes of grains, while Fig. 3.14c shows the number of grains of | |||
each class during annealing. The data shows that the fraction of indexed | |||
points and the number of non-PO grains started to decrease at the annealing | |||
temperature of 150°C. Considering that the holes were formed near non-PO | |||
grains as described above, the decrease of non-PO indexed points can be | |||
considered as a signal of hole formation in the �lm, even if holes are not | |||
visible in the image at that temperature. The detailed study of the hole | |||
formation and expansion mechanism is reported in Chapter 4. For PO | |||
grains, the fraction of indexed points kept increasing up to 350°C, while the | |||
number of grains already started to decrease at 180°C, indicating that such | |||
grains kept growing and coalescing before the dewetting process took place. | |||
== Determination of the temperature of formation of the | |||
�rst hole == | |||
As reported in Subsection 3.4.3, the TKD maps were initially recorded with | |||
a 10 nm step size up to a temperature of 500°C. The holes formed near nonpreferentially | |||
oriented (non-PO) [110] and [100] grains and it was initially | |||
supposed that the decrease of non-PO indexed points from a temperature | |||
of 150°C could be considered as a signal of hole formation in the �lm, even | |||
if the holes were not yet visible in the map. | |||
However, the indexed point number criterion alone is not enough to | |||
evaluate the starting point of the dewetting, because the total number of | |||
indexed points is a convolution of i) points that were initially indexed, but | |||
became not indexed with temperature due to the dewetting of the material | |||
and ii) points that were initially not indexed, due to the relatively large | |||
value of step sized used, but which started to get indexed with temperature | |||
when the growing grains became bigger than the step size. Therefore it has | |||
been necessary to �nd another reliable evaluation criterion to con�rm the | |||
exact starting temperature of formation of the holes in the �lm. | |||
The new criterion used consisted in the evaluation of the quality of the | |||
Kikuchi patterns on the non-indexed areas of the map. Fig. 4.2a shows | |||
the IPFZ map acquired at 210°C: in this map there are several non-mapped | |||
(and therefore black) areas. Fig. 4.2b shows the Kikuchi pattern recorded | |||
from a dewetted area of the sample, while Fig. 4.2c shows the pattern from | |||
an area with very �ne grains (in the range of 10-20 nm). The di�erence | |||
between those two patterns is evident. In c) the Kikuchi pattern is visible, | |||
but indexing was di�cult due to the chosen step size and to the fact that | |||
the grain size was close to the physical resolution of the TKD technique; | |||
thus many patterns originated from grain boundaries were di�cult to be | |||
indexed. In b) no pattern is instead visible, indicating lack of crystalline | |||
material, i.e. only the Si3N4 substrate was present at that position. | |||