Confocal Microscopy - Biomedical Optical Imaging Technologies: Design and Applications - page 214

Fig. 6.4 Images of an integrated circuit sample: ( a ) the measured surface height, ( b ) an autofo-
cused image where every point of the image is brought to its position of best focus, ( c ) an isometric
view of the sample, reconstructed in the computer
Fig. 6.5 The theoretical axial intensity recorded from a mirror for different numerical apertures
(NAs) of dry objective lens
of the optical section decreases strongly as the NA is increased. This stresses that,
in order to achieve good sectioning, the highest possible NA should be used, as
long as system aberrations are negligible. Observation of the axial image from a
mirror is in practice a useful approach to characterizing the imaging performance
and aberrations of the confocal system. It is seen that the images in Fig. 6.5 are
symmetrical about the focal plane: in the presence of aberrations, the axial response
becomes broader with stronger side-lobes.
Figure 6.6 shows axial images of a thin film of SiO 2 on a silicon substrate,
recorded using an NA of 0.8 and HeNe laser . D 633 nm/. Experimental results
(Fig. 6.6 a) and theoretical predictions (Fig. 6.6 b) are shown. There are reflections
from both the air/SiO 2 and SiO 2 /Si interfaces. This is a dramatic demonstration
of the optical sectioning property. Using some other techniques, such as digital
holographic microscopy, only the optical thickness of the sample can be measured,
so that the different interfaces cannot be observed. Also shown is the experimen-
tal image from a confocal interference microscope, recorded using a balanced