Analyzing SEM Data

Last week the group obtained micro-scale pictures of the lab sample with the Scanning Electron Microscope (SEM). Because the specimen is nonconducting, the sample was first prepared with a thin film of conducting material, like platinum, in order to reduce charging in vacuum conditions. This was accomplished using the Cressington Sputter Coater 208 HR. With Marjorie's help, the prepared sample was then viewed under the SEM. The microscope used was the Zeiss Supra 50VP.

(Left: Cressington Sputter Coater 208HR used for preparing the sample with a conductive film. Right: Zeiss Supra 50VP used view the sample.)

With pictures of the sample in hand, the group has moved into a phase of analysis. The images offer much insight on both the effects of certain spinning parameters as well as resulting filtration properties. In regards to the spinning (experimentation) component, the fibers produced appear consistent, almost entirely continuous and appropriately sized compared to the intended result. Further analysis of the fiber diameter and density will allow the group to improve the testing parameters for continued optimization. The group will also address a few unique cases of beads and spindles that were identified within the sample.

(Image of the group's sample at a scale of 2 micrometers and 5,000X magnification.)

In the analysis of filtration efficiency, the group must use imaging software to measure porosity of the sample. Using the program ImageJ, the SEM data can be analyzed by taking many appropriately scaled measurements of fiber diameter and areas of pores. For each of the samples, the average fiber diameter and average pore size will be quantified. In the analysis of average pore size, some constraints exist for this method of porosity assessment. In the sample images, there exists a certain degree of depth to which the scaling is appropriate for measurement. Moreover, fibers in the foreground of the image would produce a much larger average pore size than those in the back when under the same scaling. The group will address issue by using color analysis techniques in ImageJ.