Figure 15.11: Real LIDT test data with the LIDT value shown by a red
vertical line and a best fit 2-parameter Weibull distribution, showing that
there is still some probability of damage below the LIDT value
www.edmundoptics.eu/LO 49
Topography Analysis
Laser damage topography analysis involves generating a height map
of laser-induced damage sites characterizing the size and depth of the
damage,13 This method is not as commonly used as other detection
methods as it tends to be more tedious and time consuming. However,
valuable information for understanding the underlying mechanisms that
caused the damage is provided through this method. Topography analysis
can be performed using several different techniques including optical
microscopy, atomic force microscopy (AFM), scanning electron microscopy
(SEM), step profilers, and white light interferometry (WLI).
Different techniques are suited to detecting different types of laser induced
damage. Step profilers and AFM are ideal for precisely measuring
shallow damage sites (approximately 200 μm in diameter or less with a
depth on the magnitude of one nanometer). Both of these techniques
involve scanning a sample with a mechanical probe and generating a
height map based on the deflection of the probe. AFM systems are able
to achieve resolution on the order of fractions of a nanometer, which is
over 1000 times smaller than the optical diffraction limit for visible light.
SEM is more effective than step profilers and AFM at measuring deeper
damage sites with an aspect ratio (ratio of width to depth) of around
1, including pits left behind from the deposition of multilayer coatings.
SEM produces images by scanning a sample’s surface with a focused
beam of electrons, which are able to penetrate much deeper than photons,
14 Step profilers and AFM are not suitable for measuring deep damage
sites because the sharp slopes make it difficult for contact probes
to reach the bottom of the defect and produce accurate measurements.
Extremely deep damage sites that penetrate into the bulk material, or
pinpoint structures, are much more difficult to measure because conventional
topography analysis techniques only investigate the surface of the
optic. In order to measure these damage sites, the bulk material must be
exposed through cleaving or etching to allow one of the previously mentioned
techniques to be used to make cross-sectional measurements at
different depths. These cross-sections can then be combined into a complete
3D topography map.
Section 15.6:
Interpreting LIDT Test Results
The specified LIDT of an optic is determined by linearly extrapolating
the test data to determine the laser fluence at which the probability of
damage is zero. However, this is a linear fit to data that is not truly linear.
This single value does not provide all necessary information and damage
could still occur at or below the LIDT. Weibull and Burr distributions
are continuous probability distributions that provide a much more accurate
fit to LIDT data (Figure 15.11). In Figure 15.11, at a fluence around
5 J/cm2, there is a non-zero probability of damage, even though this is
below the specified LIDT value. The vertical error bars in damage probability
are caused by the number of test sites and the horizontal error
bars in fluence are caused by shot-to-shot variation of the test laser.
Because no laser is perfect, there will always be some level of hotspots,
or intensity fluctuations. This necessitates adding a factor of safety by
choosing an optic with a LIDT higher than the laser’s use conditions. The
safety factor required is heavily dependent on the application and type
of laser, therefore no general safety factor works for all situations. Common
industry practice is to use a safety factor of two of three. However,
if the laser-induced damage is defect driven, there are statistical models
that can be used to evaluate the damage probability at different safety
factors (see Section 15.7).
x
x
x
x
x
x
x
x
x
x x
x
x
x
x
x
Figure 15.12: Small diameter beams are less likely to overlap with low
density defects on an optic undergoing testing, resulting in overly optimistic
LIDT values
/LO