The software further allows the user to select any image of the correct size to be used as background = [aera © 
and border for the topography images, e.g. to include a company logo. The header text and scale bar ~~ yr: # 
marking is automatically put into the image using any desired font type and color, including {he movem 
“engraving” as shown in the figures for the Depth, Time and Angle bars. hat the ten 
In normal operation, the white light interferometer makes a topography image in typically 10 s with gin 
a lateral resolution of about 1 um and a depth resolution of 3 nm (or 0.3 nm in the co-called “PSI the fact Iha 
mode”). As long as vibrations are avoided, the long-term stability of the instrument is therefore not causes a lin 
so crucial. However, for experiments where the sample is repeatedly measured at time intervals the con“ TU 
from minutes to hours, it is important that the microscope unit and the XY stage do not drift relative +" 
to each other, neither vertically nor laterally. The first tests indicated, however, significant image op 
drift both in depth (vertically) and in the XY plane (laterally). The vertical drift could be several um “mE 
over as little as 20 minutes (Fig. 2). The lateral drift (Fig. 4) was largest in the Y-direction often as - 
much as 10 ym per hour. as“ Ad 
An extensive study was therefore initiated to investigate the image drift. The suspected reasons for ” 
the drifts were temperature changes and mechanical creep in bearings or other movable parts. The size 
NT-2000 has a coarse focus and a fine focus that both mechanically move the microscope unit in eres: 
the vertical direction. A series of experiment were carried out in which the last movements of the temperafu 
coarse and fine focus were varied systematically. The results in terms of the vertical drifts are is lincar ar 
shown in Fig. 2. It seems clear from the figure that an upward movement of the fine focus gives a 
much more predictable and reproducible drift than a downward movement. The data in Fig. 2 have 
not been corrected for changes in temperature. A corresponding plot where thermal expansion of 
the microscope column is taken into account shows even more clearly that an upward fine focus 
gives the lowest vertical drift with time. Further studies are needed to determine if vertical drift can 
always be eliminated by using this focusing scheme. 
J 
Fra Last direction - 
, wt of focus: 
— 
= ; Coarse - Fine 
SS / Wa —— Down - Down 
ur 3 —o—Down - Down (2) 
a 2 dr-tt=d > Down-Up 
a al - —&— Down - Up (2 
A ER Up -D Po 
© 14] AEE oT oe p - Down 
® Jr dm 
= y ZO ES ee ——Up-Up 
Ops 0 — —o—Up- Up (2) 
: — 
1 — - 
0 100 150 200 250 300 350 Fig.d:2 
Stage, Th 
Time [min] Ay 
Fig. 2: Plot of vertical drift vs. time during four-hour drift experiments with different Thine. 
conditions for the last movement of the coarse and fine focus. The measured surface oe 
was the top of the XY stage in order to eliminate other possible sources of drift. boxe i 
These data are not corrected for changes in temperature during the experiments. . - 
50 = Co 
126