|
Introduction
AFM has rapidly become an established technique for measuring the surface topography or morphology of individual, ensembles and clusters of nanoparticles1,2,3,4. One of the greatest advantages of AFM over traditional techniques such as optical and electron microscopes(SEM,TEM), is that the AFM directly produces three dimensional images.
Traditional microscopes measure only two dimensional images.
After a microscope image is measured it may be analyzed with advanced software packages to ascertain the parameters associated with the particles. Several parameters including particle volume, aspect ratio, and height are directly calculated from AFM images.
The steps required for characterizing nano-particle AFM images are:
1. Pre-Processing: In the preprocessing step, the background error associated with sample tilt and scanner bow is removed from the AFM image.
2. Outline: By outlining, the nano-particles are identified automatically with specialized software algorithms.
3. Count and/or Measure: After the nano-particles are outlined, they may be counted. Also, many parameters associated with a single particle or a cluster of particles may be calculated.
Software written for characterizing nano-particles with traditional microscopes such as the SEM is not appropriate for analyzing AFM images. This is because the SEM image characterization software does calculations in two dimensions. Thus, such software cannot give information about particle height and volume. AFM image characterization software utilizes all three dimensions for the calculations
|
|
|
|
Figure 1: Pre-processing steps illustrated
|
|
Pre-Processing
After scanning, all AFM images must be processed before they can be analyzed. The processing is intended to remove the background associated with sample tilt and scanner bow. It is important that the correct algorithms are used for this processing step. As illustrated in Figure 1, the first step is tilt removal. Next, depending on the relative amount and shape of the bow in the image, thresholding with line by line leveling should be used.
Care should be taken when processing the image to avoid introducing artificats into the image. A common artifact is shown in Figure 1B. The bands in the image are derived from line by line leveling without using the threshholding. Such bands can result in erroneous nanoparticle characterization.
Once the AFM image is processed, it is often useful to visualize the nanoparticle images. The particles may be visualized in a 2-D format, or a three dimensional format, Figure2. In either format, a color scale that indicates height can be applied to the image. Also, an image can be “shaded” so that it looks as thought a light is shining on the surface, Figure2C. The angle and direction of illumination is selectable.
When visualizing and comparing several nanoparticle images it is often helpful to show the images with the same Z scale. Such a capability is standard in nanoparticle characterization software. Figure 3 shows images of two sizes of nanoparticles: 261nm diameter and 100nm diameter displayed with the same Z scale (270nm) for comparison.
Particle characterization software includes algorithms for automatically locating the particles in the image. The methods are auto-detect and manual threshold setting. Sometimes it is helpful to test each of the methods to determine which is best for a particular image. Often both methods will give the same result.
Auto-Detect
This method has a software algorithm that searches for height transistions associated with particles in the AFM image. This technique is ideal if the nanoparticles in the AFM image are well defined. This technique is advantageous because it requires little or no expertise from the operator.
Threshold Method
In the threshold method, the particles are identified in an image by settings established by the person making the analysis. A “threshold” in the image is selected so that the nanoparticles are above the threshold. A color scale is used in the image to visually facilitate setting the threshold, Figure 4.
|
|
|
|
Figure 2: Particle Visualization Display
|
|
Count/Measure
Once the nanoparticles in the AFM image are identified they are automatically counted. The software typically does not count “partial” particles that are clipped at the edge of an image. However there is an option to enable counting partially measured particles. Each particle is assigned an individual number, Figure5.
|
|
Figure 3: Common Z Scaling: 200nm latex spheres on TEM grid and 100nm polymer spheres. Scan size 1.12 x 1.12 microns, Z scale the same for both images. |
|
|
Figure 4: Threshold Selection
|
z  |
Figure 5: Outline and count numbers
|
|
Parameters that are automatically measured for each particle and presented in a tabular format include (see Figures 6 and 7):
• Volume
• Height
• Area
• Aspect ratio
• Length
• Width
• Radius |
|
Figure 6: Table of Parametersn
|
z  |
Figure 7: Statistical information on measured particles and single
object information
|
|
Besides being able to analyze a singe particle, graphs may be created that show the distribution of particles as a function of any of the measured variables. Figure 8 shows the distribution of particles in an image versus the particle volume, volume vs height, volume vs aspect ratio.
|
|
Figure 8A: Volume Measurements
|
|
|
Figure 8B: Volume-vs-height particle distribution. Total number of particles
is 287. |
Figure 8C: Volume-vs-aspect ratio distribution. |
Figure 8: Distribution of particles
|
|
Surface roughness on the very top of the particle can be measured using a special option – selected area roughness measurement, Figure 9.
|
|
Figure 9: Surface Roughness Measurements on top of the particle.
|
|
Conclusion
To avoid errors caused by artifacts, correct pre-processing of AFM data is critical. Meaningful data interpretation has to take into account the finite size of the AFM tip as well as other contributing factors described in the literature4,5,6. Correct AFM functionality can be verified on pre-characterized spheres prior to imaging particles of irregular shape.
Nanoparticle characterization with an AFM requires specialized software that permits background removal, outlining the particles, and counting/measuring the particles. The advantage of AFM over traditional microscopes such as SEM is that AFM measures three dimensional images so that particle height and volume can be calculated..
|
print now
