I am studying the behavior of metallic nanoparticles in aqueous environments. I take a sample of water, from various locations, and introduce nanoparticles, type of NP varies, and at different time intervals I analyze the sample, one measurement being the particle size distribution (PSD). So I'm wondering how to compare the PSD at each time interval for these samples, which have been collected from different places and have different NPs in them.
Thank you for any help or advice with this!
Laura Wright What technique are you using for PSD? Usual comparison points are those that indicate something of the fine fraction (x10), central tendency (x50), and top end (x90) of the distribution. Take a look at the attached for some introduction to the math.
You can use boxplot to have a visual comparison between each PSD, with important representative points such as the mean, the median diameter (d50), other percentiles (d10, d90, ...) and the points at the extremities of the PSD. With that, you can also calculate the span=(d90-d10)/d50, which gives an indication of the PSD width.
If you want to go further, you can make some easy statistics to compare distributions. For instance, you can use the test of Mann & Whitney (ok for non normal distributions, and little samplings). This can help you to say if two distributions are significantly different or not.
Regards
Alan F Rawle you know, after reading about the math of psd I'm not sure using the mean diameter of a particle is the best option...
Im using DLS for the PSD analysis. And typically look at the intensity graphs to see the results, but i was going to try to do some statistical testing on the mean diameter size at each time step to see if there was significant differences.
Naïma Gaudel Thank you. I will look into how to do that.... stats is my weakest link. I appreciate your response.
Laura Wright The most robust paramenter in DLS according to the ISO and ASTM standards are the z-average and polydispersity index (PDI). You should use nothing else especially if the distribution is even a little broad.
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Electrodes