What is a Good AFM?
When it comes time to purchase an AFM, there are many features
that may be important based on the final system application.
These features will create a long list of questions that
will be unique to the user when they proceed to evaluate
an AFM system. Questions that may be important to individual
users include: Is the system designed for industrial or
research applications? Is the system going to be installed
in a multi-user facility? Will the system support all of
the available options that I require? Is the system easy
to use for low-level users? Is the AFM flexible for high-end
users? These and many other questions are common when a
customer first begins to gather information about available
AFM systems.
Once these questions have been answered, and a list of
viable systems has been compiled, the user needs to make
a fundamental comparison of the systems that best meet their
overall requirements. With this goal in mind, Park Systems
has compiled what we believe to be a comprehensive overview
of “What is a Good AFM”
|
|
The Basics
Historically, while AFMs have proven to be a reliable
instrument for determining relative sample dimensions, they
have struggled to provide accurate absolute dimensions of surface
features. As the dimensions found in research and industrial
applications become smaller and smaller, it is now more important
than ever for AFMs to measure the absolute dimensions of surface
features with accuracy and repeatability.
For taking nanoscale measurements, the accuracy and reliability
of measurements and the flexibility of available modes and options
are just as important as resolution. When it comes time to purchase
or upgrade an AFM, regardless of your diverse application needs,
AFM performance will be influenced by the seven attributes listed
and described below:
 Noise
Floor
XY Scan
Flatness
XY Scan
Linearity
Tip Life
Thermal Drift
Available
SPM Modes
Option Compatibility
Noise Floor
|
Figure 1. The noise floor is
measured by taking a “zero scan,” in which the
tip is brought to the sample surface, and a 0
x 0 nm, 256 x 256
pixel scan with 0.5 gain is performed. The signal
for this scan of just one point corresponds to
the noise floor of the instrument .
|
Even the smallest environmental
vibrations can add noise to AFM results, making it a
challenge to image the smallest details and characterize
the flattest surfaces. To measure the baseline noise,
or noise floor, the user brings the cantilever to the
sample surface and obtains the system response to a
“zero scan” (See Figure 1). A good AFM should be isolated
from vibrations to achieve a noise floor below 0.5A
0 x 0 nm scan, staying in one point
0.5 gain, in contact mode
256 x 256 pixels
|
XY Scan Flatness
The AFM scanner is the most important component
of any AFM, as the scanner performance determines the
accuracy of the imaging results an AFM can provide.
Therefore, it is very important to evaluate the presence
and amount of any scanner artifacts, such as high-order
and non-linear background motions, to avoid distorting
the resulting AFM image.
|
Figure 2. The Z run-out, or out-of-plane
motion, is the peak to valley height of
scan profiles without any software correction
except for the sample tilt correction
(no higher than 1st order plane fitting).
The sample can either be an optical flat
(reference mirror) or a flat portion on
an AFM calibration standard.
|
|
Hence, a customer must look for a scanner design which
minimizes out-of-plane motion when imaging a flat surface.
A good AFM scanner must keep the Z out-of plane motion
within a few nanometers over the entire scan range,
independent of scan rate (see Figure 3), scan size,
and scanner offset (see Figure 4).
|
|
Figure 3. Typical
out of plane motion of a good AFM has less than
+/- 2nm over the entire 100μm scan area, having
repeatability < 0.5 nm at scan rates 0.5, 1, and
2 Hz. The repeatability is defined as the maximum
variation between the averaged profiles of two
or more scans. The tests can be done in both fast
scan directions of X and Y.
|
|
|
|
Figure 4. Typical out of plane
motion of a good AFM has less than +/- 2nm over 40um scan
range, having repeatability < 0.5 nm with XY offsets of
(0, 0), (25um, 25um), (25um, -25um), (-25um, 25um) and (-25um,
-25um). The repeatability is defined as the maximum variation
between the averaged profiles of two or more scans. The
tests can be done in both fast scan directions of X and
Y.
|
|
|
|
|
XY Scan Linearity
|
|
Figure 5. The X-Y linearity
is evaluated by matching two orthogonal scans after changing
the direction of the fast scan axis. A test can be performed
on a 40 x 40um size site at a random offset coordinate.
|
|
|
Figure 6. A good AFM should
show no visible mismatch in the two orthogonal scans,
demonstrating the excellent XY scan linearity.
|
The XY scan linearity is another important feature
of any AFM that determines the accuracy of the imaging results
an AFM can provide. If the X and Y scan movements are coupled,
as in a tube scanner, an extension in the X direction will directly
influence the scanner’s Y position. The XY scan linearity is measured
using two orthogonal scans and subsequent image matching as shown
in Figure 5. The amount of mismatch in two images taken by the
scanner will reveal how linear its X and Y scan movements are.
Hence, a customer must look for a scanner design which maximizes
the XY scan linearity. A good AFM scanner must keep the XY scan
linearity less than 0.05% at a random coordinate (see Figure 6),
independent of scan rate (see Figure 7) , scan size, and scanner
offset (see Figure 8).
 |
|
Figure 7. A good
AFM should show no visible mismatch in the two orthogonal
scans, demonstrating the excellent XY scan linearity, at
different scan rates of 1 and 2 Hz.
|
Figure 8. A good AFM should show no visible mismatch
in the two orthogonal scans, demonstrating the excellent
XY scan linearity with XY offsets of (0, 0), (25um, 25um),
(25um, -25um), (-25um, 25um) and (-25um, -25um). |
|
|
|
Tip Life |
Tip life is an important factor in getting
high-quality images with consistency and reliability. When
the tip touches the sample and becomes blunt, it limits
the resolution of the AFM and reduces the quality of the
image. For softer samples, tip-surface contact damages the
sample as well as the tip, compounding the inaccuracies
of sample height measurements. Consequently, preserving
tip integrity enables consistent high-resolution and accurate
data.
A good way to test the tip life of an AFM system is by
imaging CrN, which has sharp and pointy features on a very
hard surface. If a tip becomes blunt or damaged, it can
no longer reach the bottom of the features, and the images
become blurred..With a good AFM, the sharpness of the tip
can be preserved even after 100 image scans of the CrN sample
as shown in Figure 9, while maintaining the same sample
surface roughness (see Figure 10).
|
 |
|
Figure 9. One
can easily perform tip wearing experiment with CrN
sample, so called Tip Checker. With a good AFM, the
sharpness of the tip is preserved even after 100 scans.
If tip life is not long enough, one can see blurred
images without sharp triangular features which are
consequences of worn and damaged tips.
|
Figure 10. With a good AFM, the sharpness of the
tip is preserved, and the sample surface roughness
value remains the same. |
|
| |
|
Option Compatibility
|
|
|
|
AFM Buyer's Guide
How
to Evaluate an AFM
