I did a size measurement of cellulose nanocrystals in a Zetasizer ZS90, but I do not know if I must take the z-average or the peak value. Thank you very much for your help.
If you look at the distribution analysis charts, how many peaks are present?
Z Average is a recommended result to report from DLS as it comes from a direct fit of the measured correlation function, however it is only a reliable parameter when the sample of interest in monomodal, as it represents an average particle size.
If your distribution results only show 1 peak, then the ZAve and Peak size should be comparable to within some error, but if the distribution is multimodal than you should use the peak size.
@ Adolfo. Please show a plot of your results as I am unsure why you have more than one peak present (agglomerates?) and the sizes of these separate peak(s). If your material in truly 'nano' it will be < 100 nm according to definitions in the international standards.
I uploaded here the size distribution plot of my results. There are a peak at 94 nm (24.6%) and other in 504 nm (75.4%). The Z-average is equal to 255.6 nm and the PDI is 0.354.
As mentioned, Z Average is only meaningful as a measure of size if your sample in monodisperse, so here, peak sizes are appropriate.
Is this distribution intensity, number or volume weighted? Based on the shape, I would expect the former, so whilst the peak at 504 nm is larger than the peak at 95 nm, it is important to note that larger particles scatter more light (I ~r^6), so the particles at around 95 nm will be more abundant that this plot implies.
Alexander Malm Is the Z average size not only supposed to be used for spherical particles? I thought I read that in a manual however I may be mistaken? Also which of the measurement options would you use to quantify the size (intensity/volume/etc)?
Both the Z Ave and distribution results report the diameter of a sphere that diffuses at the same rate as the sample.
If your material has shape anisotropy, then it may be possible to see two different diffusion rates- one associated with translational diffusion and an addition compomemt for rotational diffusion. In a traditional DLS measurement these can not be differentiated and will both be reported as an apparent size component in the distribution result.
An extension to DLS called Depolarised Dynamic Light Scattering used polarising optics to differentiate rotational and translational diffusion.
In in terms of which to use to quantify, Z Ave is the most robust and is therefore the ISO recommended measurement. This is because distribution analysis results can vary depending on the inputs of the model used in the fit, I.e number and spacing of size classes, regularisation etc. If the sample is multimodal, then the Z Ave is no longer representative of the particle size, and I would refer to the intensity distribution.
The intensity distribution is biased to larger components which scatter more strongly, so you may therefore wish to refer to volume or number weighted results, however this is a further abstraction from the raw data and requires knowledge of material optical properties which either the intensity distribution or Z Ave don’t.