Hi folks,
This is a software-related question. I've seen many researchers asking how to export data using Malvern Zetasizer software. Most people simply answer the most obvious suggestion: click on the measurement, go to Edit -> Copy size values, then paste your data on excel, Origins or whatever.
In addition to this, you can also go file -> export and select exactly the kind of information (or the distribution) you want. Once you created your exporting template, you can automatize everything by enabling "export" while performing a measurement. All of these information are available in a 6 min video video made by Malvern Panalytical that automatically pops up if you google "Malvern how to export raw data".
But this is just a premise. The issue I'm facing it's the fact that I don't like how these distribution are plotted (it doesn't matter the kind of measurement). Let's say, for example, that I'm exporting a "Intensity vs size" distribution. My graph will have just 69 points. That means that, if I have a good gaussian distribution, only 20-30 points will be used for the distribution plot. As a consequence, my distribution will look like a segmented curve. I would love to find a way to improve my graph resolution. Until now, I haven't found any useful suggestion on Marvel website. Do you have any? I was thinking about exporting the correlation curve and using a Macro on IGOR for the cumulant analysis, but the correlation curve has a poor resolution, too.
Thanks you in advance,
Lorenzo
The number of points in the distribution is limited by the number of points in the correlation function. You won't gain anything by increasing the number of points in the distribution other than make it look nicer. You might as well use a smooth curve in something like Excel.
Also, if you are looking at a size distribution from the cumulants analysis, the graph is rather misleading since it isn't a real distribution - just a mean value with a standard deviation (from the polydispersity index). Given the numerical values of the mean and standard deviation, you could plot your own.
Dear J.F. Miller,
thank you for responding so quickly. You're totally right about the cumulant analysis. I will discard this option because you just remember me in my Master's I never actually used it to plot a distribution.
Still, I don't think that we should underestimate the aestetich issue: the distributions are poorly plotted. Do you know if there are any ways to increase the number points directly from the Malvern Software?
A couple of comments in reply to this discussion -
The Malvern Zetasizer software treats cumulants and distribution analysis separately, and all distributions originate from a non-negative least squares (NNLS) inversion method.
The software for the Zetasizer Nano does in fact allow you to edit the number of size points in the analysis, as well as the maximum and minimum size value so you could in theory increase the resolution.
This is however not recommended as it is quite easy to start over resolving the result and we may start to see artifact peaks in the analysis.
With regards to smoothing, the result is also smoothed via a regularization in the analysis.
If you would like more resolution from the analysis, you may wish to choose the multiple narrow modes analysis model which uses a different regularization value to give a higher resolution measurement.
In terms of wanting a smooth continuous distribution curve, this may be achieved by how you choose to "join the dots" if you plot the data in another package, but it is important to consider how the data is generated.
There are an infinite number of solutions to the distribution analysis, however these are reduced by firstly not considering negative contributions to the fit, and secondly by fitting a finite number of points to the analysis.
The distribution is an answer to the question of "what proportion of these sized particles would give a matching correlation function", and then we add a limit of only smooth solutions via the regularization.
You could therefore argue that the best representation for the distribution result is as a histogram...
Dear Alexander Malm , thank you very much for your illuminating response!
Hello Lorenzo Marsili - if your question is purely asthetics: there is an undocumented feature to allow you access to advanced graphing tools. To access these options, move the mouse into the graph and press these keys simultaneously Ctrl+Shift+Right-click on the graph . Any changes you make (ChartStyle, Labels, Markers, Axis, Titles, Legends,...) will be only temporary and revert to defaults when you exit the software.
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Enzymology
- Enzymes