No, you cannot expect to see by dynamic light scattering the whole spectrum of sizes present in you system.

In general, the autocorrelation curve (from which you get the size distribution) is obtained from the fluctuating scattering intensity. In the simplest approach, each particle contributes to the total scattering intensity with a factor proportional to its square weight. In your specific case, if the aggregates are much larger than the primary particles, you might not be able to detect the latter ones.

As a rule of thumb, the particle which you want to detect on a sensitive light scattering apparatus (I am not referring to a zeta-sizer) should contribute at least to 1% of the overall scattering intensity. Note also, that the  scattering contribute is angle dependent.

thank you Leonardo for your answer. I see the point for small size particles: the smaller, the lower the light scattering contribution, so they are hindered. On the other hand, very large particles (about 10 um diameter) are above the upper size limit range of the technique. But my point is: how slow should the aggregation process be, to be able to caputure time-dependent (evolution) drifts in the curve? Does the aggregation typically occur on the same time-scale of the DLS measurements (i.e. minutes) or it is much faster (e.g. seconds), such that, in case of aggregation, one can never shoot a picture of the primary (instable) particles?

PS: you mean that zeta-sizer is not sensitive? I couldn't understand this comment.

Sorry, then I misunderstood your question.

The only thing I can answer to your question now, is that to see an evolution of your system, of course the kinetics should be slower than the experimental time. Can you somehow trigger the aggregate formation?

If yes, you might follow the transmission measured on a UV-Vis at a certain wavelength (measurement time of 1second should be sufficient). There you should get an idea on the aggregation time.

Concerning the zeta-sizers, they are pretty fine for an analysis of "simple", well behaving systems, e.g., dilute monodisperse nanoparticle solutions. However, for the characterizations of more complex solutions the use of a light scattering apparatus (as sold by a number of companies, ALV, Brookhaven just to mention two of them), which allow to perform measurement at different angles. with fast correlators are necessary (in my opinion).

Dear Marco, if you observe dispersed nano-particles which are electrostatically (not sterically) stabilised, than zeta potential can indicate stability of your system. In that case, if absolute value of zeta potential is >30 mV, than the particles will bi stable, as you said. If particles are stable, than there should not be any observable aggregation.

From my experience, there are particles that aggregate for days and those which aggregate in few minutes. Usually if particles are aggregating during measurement, quality report will inform you that something is wrong. I can not imagine that particles are aggregating so fast that the aggregation is finished before size measurement have even started and that there was not visible precipitation, because (in my opinion) if particles are that prone to aggregation and they are electrostatically stabilised (but very unstable), than they will keep forming even larger aggregates and precipitate in short time. So my guess is: if aggregation had ended before measurement has strated than aggregation should be very obvious.

I study aggregation of humic acids. At high pH values humic acid is disaggregated (dissolved), solution appears clean and transparent. Zetasizer shows multiple peaks: from 1000 nm. One can say that there is some aggregation going on, or that there is (dynamic) equilibrium between primary particles (in this case: molecules) and aggregates. (Multiple peaks with larger particles can also mean that there are some impurities like dust particles in the system.) But by lowering pH the aggregation is becoming very obvious: there are visible particles in solution and particles are precipitating in a matter of minutes or hours. Size measurement shows less number of peaks, usually single peak around 1000 nm. So if there is multiple peaks it could point to aggregation, but if there is single peak it doesn’t mean that particles are not aggregating.

Dear Marco!

It is a problem of the polymodal analysis of DLS. Intensivity of signal for particles from one material is proportional to diameter in the sixth degree! Therefore DLS sees generally big particles. Look at diameter average by number (Att. file, fig 8). If the smaller mode doesn't appear, separation, for example, filtering is necessary.

In fact, DLS can measure the size of particles ranging from 1-10000 nm. However, there are two conditions that particles can escape from the detection.

The first one is that the larger particles (e.g., 1000-4000 nm) in the dispersion mask the effect of the presence of smaller ones (

Hello Marco,

if electrostatic stabilization is the main contribution then you are correct: >30mV is a good indication of stability. If, however, there is a steric stabilization component, then even samples with much smaller zeta potential can still be stable, so it is not always a requirement to have high absolute zeta potentials in order to have a stable colloidal suspension.

In DLS, multiple peaks will only appear if the regularization algorithm used to fit the data is able to resolve them. In order for that to happen, there should probably be at least 1% of the scattering signal from a particular size peak. In addition, if sizes are closer than about a factor 3, they will not be resolved. So for example a mix of 10nm and 20nm nanoparticles will not appear as two seperate peaks in a DLS size distribution but instead as one peak at the intensity-weighed average between the two (in order to obtain two peaks one would need a separation technique such as for example GPC or FFF in conjunction with DLS). If one of the peaks is from a component with large size (for example several hundred nm) and the other with quite small size (for example a few nm) then the situation may arise where the scattering from the small peak is so small that it no longer gets picked up by the regularization. The angular dependent light scattering of large particles makes it possible for a backscattering system to still be able to see the small particles where at other angles the small particles would already have disappeared in the noise, in other words a high scattering angle gives you the best chance of still observing the smaller sized population. In the Zetasizer the forward angle can actually be used for DLS as well, however in many cases the largest species / aggregation may completely dominate the result. For standard size applications typical correlators are adequate, faster correlators may be needed when studying faster phenomena (like multilple light scattering or diffusing wave spectroscopy).

If your sample is changing over time, then you can try to reduce the measurement duration in your Zetasizer. Instead of automatic mode (which tends to measure and average runs for about 2-3 minutes) set the duration to manual and perform many short runs (as short as 5 second might work for you). There will be a larger variation on the results, but this setting will also make you aware of any aggregation kinetics if on that time scale. The "safest" parameter to observe is the cumulant or z-average size versus time. Under ideal conditions, you may be able to see that the size starts out small, then gets larger, splits into two peaks, the smaller size being close to the primary size, the larger to some aggregates. However, this is by no means the only possible scenario. If you can trigger/control the aggregation then that might be interesting to observe.

thank you Ulf, thank you all. I'll keep thinking about all your comments. Bye.