Data Analysis



After the group was able to successfully electrospun the first mat, pictures on a nanoscale were taken for data analysis of the distribution of the fibers, which created pores, hence showing filtration. With the use of ImageJ, an image processing program, the images were examined to inspect the mat. Using tool Threshold to show the different levels of the image, the length of fiber diameters were measured based on a scale system within the program. After converting 253 diameters, the diameters were sorted on Excel from smallest to largest, showing the wide spread. The mean value was calculated to be 180.2 nm, while the standard deviation was 30.5.

In order to show the distribution of the fiber diameter lengths, the difference between the smallest and largest number was used to calculate every ten percent increase within the spread. The table below shows what the diameter length was at the top and bottom of the list, the half point, and every ten percent in between.

Length of Fiber Diameter (nm)
Percentage Increase
Normal Distrubition
99.70
0%
0.0004
118.26
10%
0.0017
136.82
20%
0.0048
155.48
30%
0.0094
173.94
40%
0.0128
192.50
50%
0.0120
211.06
60%
0.0078
229.62
70%
0.0035
248.18
80%
0.0011
266.74
90%
0.0004
285.30
100%
0.0001


This table shows what was the top 20% fiber diameter length, which was 248.18 nm. It was also necessary to calculate the normal distribution for each of the 11 values using the function on Excel function called NORMDIST, which would return the normal distribution for the computed mean and standard deviation. The formula is as follows:

=NORMDIST(FiberDiameter,mean,standard-deviation,cumulative)

Using these new values, a graph can be obtained of the normal distribution of the fiber diameter lengths.

Figure 7 -Normal distribution of fiber diameter length


The graph has the lengths on the x-axis while the y-axis is the normal distribution, thus illustrating the distribution of all the fiber diameter lengths, with the highest point of the graph being the mean value, 180.2 nm. It is clear that there was a good overall diameter lengths of the fibers, showing that there were more numbers closer to the mean value, than there were numbers further from the value of the mean.

The next part of the images that was measured was the pore areas, which were created by the intertwining fibers. Again, with the use of Threshold, the pore areas visible on the images were measured on a scale system in ImageJ. This scale system was set up based on a measure scale on the image, which was then calculated into pixels/nm, with a conversion rate of 0.0445 pixels/nm. These measurements were then placed into Excel and converted into microns with the following for formula:

Area Pore (μm2) =( Area Pore (pixels/nm)* (Conversion Rate)2*(1/1000)

The actual value for each measurement was used with the use of Excel’s formulas. After the conversion of 193 areas, every value was sorted in an order of smallest to largest with the use of the Sort toolbar on Excel. The mean value was calculated to be 8.3 nm2, while the standard deviation was 5.75. The same method that was used to find the normal distribution for the diameter lengths was then used to calculate the normal distribution of the pore areas.

So the distribution of the pore areas was found by finding the difference between the smallest and largest number was used to calculate every ten percent increase within the spread. The table below shows what the pore area was at the top and bottom of the list, the half point, and every ten percent in between.


Pore Area (μm2)
Percentage Increase
Normal Distribution
0.2790
0%
0.0262
2.6969
10%
0.0431
5.1148
20%
0.0595
7.5327
30%
0.0688
9.9506
40%
0.0666
12.3685
50%
0.0541
14.7864
60%
0.0368
17.2043
70%
0.0209
19.6222
80%
0.0098
22.0401
90%
0.0039
24.4580
100%
0.0013


Like earlier, the normal distribution for each of the 11 values had to be computed using the function on Excel function called NORMDIST, which would return the normal distribution for the computed mean and standard deviation. The formula is as follows:

=NORMDIST(PoreArea,mean,standard-deviation,cumulative)

Using these new values, a graph can be obtained of the normal distribution of the pore areas.

Figure 8 - Normal distribution of pore area measurements

The graph shows the pore areas on the x-axis while the y-axis is the normal distribution, thus illustrating the distribution of all the areas of the different pores., with the highest point of the graph being the mean value, 8.3 μm2. The graph is not a perfect distribution, that the majority of the numbers were closer to the value of the mean, signaling that the higher the numbers were, the more random and uncommon the number was. Therefore, there were many values very close to the mean, demonstrating that most pore areas were very small, and only roughly about a few μm2.

No comments:

Post a Comment