Hi Aiala,

When you measured the particle with DLS and laser diffractometry, did you do in suspension?

The tiny particles (TEM result: apprx. 30nm) is not easy to be measured because in suspension they have high risks of agglomerate. The aggregates can be as big as several hundred nanometers or more even you keep zeta potential of them >30mV or

Hi,

Thanks for your answer. My particles are made of a mixture of lipids, but I don't know their density. Some of them contain chitosan in order to positively charge them, but they are not very different to their unmodified counterparts. I have used a Beckman Coulter for laser diffractometry and a zetasizer nanoseries for DLS. The samples are suspended in water containing 0.1% of Tween 20. The TEM images show a high polydispersion, but apparently, no aggregation.

The TEM results are the most trust worthy to me. Can you upload one example image here? It is alway easier to discuss on a real example... Thx

The principle of DLS is Rayleigh scattering,which falls off with the sixth power of particle size. So when the sample is polydisperse or easy aggregated. The large particles in the solution would be more influence to result, even a very small proportion of the sample.

Hi Aiala,

I've used the same DLS a few times and I think you may have gotten some information about the polydispersity of your sample, and the most abundant species in your sample according to their particle size, on the Malvern Zetaseizer DLS you were using. This should help you to determine the discrepancies in your results compared to the TEM. Also, you might want to try to dilute your samples and analyze again on the DLS in order to see if this is a concentration-dependent pattern (which it should). But I agree with Sebastian about the TEM results being the most trustworthy.

I'm far from being an expert but I hope it helps out.

Thank you very much to everyone! Here are some pictures from TEM (link). Yes, Cesar, I looked at the polydispersity in my sample (between 0.4-0.5), and to the most abundant specie, but still it does not correlate to the TEM.

I'll try to dilute my sample more and see what happens. Thank you!!!!  

https://www.dropbox.com/sh/3ca29e3j8ypbvg2/AABKm60yQb64dx7qiFDXNRZDa?dl=0

DLS size measurement is not reproducible due to high chance of agglomeration in the case of polymeric materials but low polydispersity index can be a considered as a good result. A sufficient diluted sample can give a good result. Please also check the SOP that must match to your material.  

Without being an expert on this from the TEM pictures I would say  the picture is quite complete. In your picture I see some particles much larger (up-to 30times?), i.e. largest particles may be indeed around 1 µm. By counting you may have got number based values. Your numbers by DLS and diffractometry are in intensity weighted or volume based? In both cases a few larger particles count much more than the smaller ones. This is not about right or wrong, but you should compare only number based with number based and volume based with volume based or consider what means intensity weighted for the 2 latter methods.

You have a bimodal distribution. The 2 latter methods use models to extract size information from the scattering data. If device software allows for, you have to give the input to use a model for bimodal distribution. Laser diffractometry, at least with common instruments, is not able to measure nanoparticles, i.e. smaller 100 nm, especially not if there is no narrow monomodal distribution. Therefore, you obtain the info on the distribution of the larger fraction only, maybe perfectly matching this part of reality. DLS should give a smaller size because partly it accounts also for smaller particles.

To measure bimodal distributions you better use separating methods like analytical centrifugation, however, you will need density data to obtain size values.

Hi Titus! Thanks for your answer. Yes, I compared the number based values from the main peak and it was still pretty different. The secondary peak was even of higher size, but I could not found in DLS the expected 30 nm peak seen on the TEM.

Hi， I met this same situation now.  I want to ask how you solve this problem.  Have you found the reason why there was pretty difference between DLS and TEM?  Waiting for your answer. Thanks very much!

@Josiah,

for the reason see discussion above. There is nothing to solve on this because these are the limitations of DLS, i.e. being much more sensitive to the larger particles. You have to live with this, use another method or first obtain fractions and measure them separately with DLS.

1) TEM measurement is performed in the dry state whereas differential light scattering (DLS) is measured in the solvated state. 2) TEM is a number based particle size measurement that show stronger emphasis on the smallest components in the size distribution whereas DLS is an intensity based measurement and emphasize on the larger particle size.

Rudra Pangeni -- thanks for the comment - however I do need to make you aware, what you say is only true, if you compare the number distribution of TEM and the volume distribution of DLS -- you can ofcourse transforme the number based distribution of the TEM to a volume based distribution and than will see the oposite phenomena than described by you.

@all: And ofcourse please consider the sample size of analysed particles. Usually due to the incredible work load of manual imagea analysis, the TEM data rely on a significant lower number of particles but - as they are a direct measurement including shape and morphology information the most reliable ones --- IF you have the time to count and measure enough particles.

It should be because of the aggregation impact in DLS experiments. In fact DLS does not give you the size of particles, but rather it provides the size of aggregated particles which may be different than the size of particles. The best method to correlate with TEM is Small-angle X-ray scattering which can get rid of the aggregation impact and is even more accurate than TEM