I recently presented a webinar titled “Basic Principles of Nanoparticle Tracking Analysis: Size, Concentration and Speciation with Single-Particle Resolution”. We got such a good turnout that we were not able to address the numerous questions asked during the event.

Thank you for the participation!


Since there was such strong interest, we have provided the following list of questions and answers to benefit all who are interested.

In DLS also we can get lognormal number distribution. How is the NTA number distribution different?

NTA yields a number weighted distribution because it is a particle by particle analysis. DLS naturally yields an intensity weighted distribution, which can be converted to volume weighted, then further to number weighted. In ASTM guides, there is specific guidance that the number based distribution derived from DLS is subject to large inaccuracies and is not recommended for routine use. NTA is by nature a number based technique, so the same sample on these two techniques may show somewhat different results.

Check our technical note on Comparison of Statistical Measures Reported by NTA and DLS Techniques.

We also have a white paper comparing NTA and DLS results over a range of sample types. This should give a better idea of how both techniques fit together.

Does NTA fit data to any mathematical equation?

NTA does not fit the data to any particular peak shape, adjust the size distributions after analysis, or bias different parts of the size range. The histograms are just a display of all the individual particles after their size has been analyzed. Since our basic data is a population of single particles, any statistical measures are straightforward population statistics rather than interpolations from a derived distribution.

Is dilution needed in sample preparation?

Dilution will most likely be required unless you have a sample that is very sparse to begin with. If you are familiar with DLS measurements, we are usually 10-1000x more dilute than that. Ideally in the range of 10^7 to 10^9 particles/ml.

Can this work for emulsion droplets?

Emulsions in the sub-micron size range are certainly suitable for this technique. As long as the samples are diluted with an appropriate liquid that maintains the stability of the droplets, then the measurement should be quite easy.

What can NTA do for small protein e.g. albumin (7 nm)?

Protein monomers are well below the lower detection limit of this technique, but monitoring protein aggregation is a strong application for these instruments because of their ability with polydisperse distributions and to provide a concentration of the aggregates that are measured. Our application note on protein aggregation should provide a little more information.

Can this method show changes due to coating of the core particle with surfactant or antibody? and how it is different from DLS?

The nature of the measurement yields the hydrodynamic diameter, meaning the size of the object moving in the liquid. When anything of significant size is added to the surface of the primary particle, it affects the particle’s motion in the liquid, therefore the size that is reported. Adding surfactants or antibodies to the surface of the particle should be enough to show a shift in the reported size by an amount equal to the size of the attached molecules. Size shifts as small as three nanometers have been reliably seen in some coated/uncoated experiments.

DLS and NTA/NanoSight will respond in the same way since the Brownian motion both techniques rely on will be affected an equal amount. NTA might be slightly more reliable in showing that shift since it can measure the primary particles accurately, independent of whatever else may be in the sample.

Another technique to consider for this application is Resonance Mass Measurement (RMM), as embodied in Malvern’s Archimedes product. If the primary particles are within the measurement range of the technique, the mass sensitivity of this technique means the shift due to the coating will be easily measured.

Is sample viscosity a required input for proper measurement?

The Stokes-Einstein equation requires temperature and viscosity variables to be input. For any aqueous based diluent, the viscosity can be automatically read in from the temperature-dependent table of values. Temperature is measured in the sample cell and automatically read into the software for most NanoSight models. Therefore, unless the diluent is significantly different from water, the user will not need to manually enter any values.

How does this method compare to tunable resistive pulse sensing?

Comparing two different techniques could take many pages but I will summarize the major differences. More importantly, you should consider what you are trying to learn about your sample and what technique might be most sensitive to that.

NTA and TRPS are often considered together because they have similar measurement ranges and parameters. NTA has a lower limit of detection, so is able to see the full-size distribution for samples that extend well below 100nm. NTA does not require a strong electrolyte solution, which may interfere with some samples’ stability. Samples may be measured in any aqueous based diluent or a wide range of organic solvents. The TRPS sensing pore must be calibrated frequently for accurate results and is susceptible to clogging. Because each particle is pulled through one at a time, it has the potential to be a higher resolution technique. NTA is generally more flexible, has a broader application range, is much faster and more reproducible to use in practice, has the ability to work under fluorescence mode, and is more broadly accepted in the scientific community due to these greater capabilities.

If you are considering these techniques, we would welcome the opportunity to test a few samples to show what the system is capable of. All techniques have strengths and weaknesses, as evidenced by the range of techniques in Malvern’s portfolio.

How would you recommend to best calibrate a machine to ensure the sizing is correct?

Perhaps splitting hairs on semantics, but NTA is not a technique that requires calibration per se. Standards should routinely be run to verify correct operation and to check for drift in the longer-term. If those results are outside of specification or show drift, contact Malvern’s support desk. Access the detailed procedure for the polystyrene latex standards that are most commonly used to find out more.

Can NTA measure zeta potential?

Yes, the NanoSight NS500 model has the ability to measure zeta potential on a particle by particle basis. Our technical note provides some details on the process. If you wish to characterize individual components in a mixed material or if zeta potential is a way to characterize a change in the sample, this can be a valuable alternative to traditional electrophoretic light scattering zeta potential as used in Malvern’s Zetasizer models.

In flow mode, which flow (ml/min) can be used?

The simple answer is whatever setting will allow the particles to stay in the field of view for 5-10 seconds. This depends on which model instrument and top plate are being used. More details and recommended speed ranges are in our technical note on flow mode analyses.

What are your recommendation for sample preparation of dispersions of clay-sized minerals?

I have worked with various clay and environmental samples. For the many environmental water samples, samples are often measured as-received, usually with some level of dilution. If you are starting from any type of dry powder, there will usually be some effort required to disperse them into a liquid to get to primary particles (not agglomerates). Some surfactant would be required (e.g. sodium poly-metaphosphate) and mechanical energy like sonication. Depending on your goal (measuring agglomerates or primary particles) would determine what strategy you follow and how much energy you want to impart.

Can one evaporate a sample and watch precipitation of phases in real-time?

The NanoSight sample chamber is enclosed so the evaporation/precipitation would have to be done externally, then injected into the sample cell. Measurements can be as quick as a few seconds if you only need a quick estimation of the size. From a reaction vessel to the measurement chamber is just a matter of loading a syringe and inserting that to the instrument, so can be a quick response.

Can we see the shape of the particles? How does the particle shape influence the diameter measurement?

NTA is not able to see the shape of the particle. All particles will show up as a spot of light and the system only analyzes the motion of that spot of light. A high aspect ratio particle might generate a spot of light that isn’t perfectly round, but it also tumbles and fluctuates, so there isn’t any consistent information to work with. The output is an equivalent spherical diameter, so a sphere that has an equivalent volume of the particle being measured. If two-rod samples of different lengths are measured, the equivalent spherical diameter would show a difference relative to the difference in particle volume.

Can NTA measure differently shaped particles present simultaneously, for instance, spheres and rods?

Spheres and rods can be measured together in one sample, but the only way to distinguish them would be if the two equivalent spherical diameters are different enough from one another.

Is it possible to reprocess a video and change the viscosity value?

To change viscosity, open the files you which to change. They should be highlighted in blue as below. Click Change Settings>Viscosity, uncheck the box, then double-click where it says WATER in the picture now. You can then input the viscosity in cP. Click Update and it should reprocess and display the adjusted data. You will need to Export Results if you want updated spreadsheets or PDF reports.

If I wish to reproduce the result plot what is the data should I pick from the excel file?

From the summary spreadsheet files we export, just insert>graph>line graph, then select the ‘concentration’ column as the actual data. Bin center will be the x axis labels.

How do you decide what is the right detection threshold to analyze a sample?

The correct detection threshold will depend on the sample type and how the video was collected. If you adjust the camera level so all particles are visible, but the largest particles are not saturated (if that compromise is possible), it usually results in the analyses being run with the default detection threshold of 5.

The logic is to make the threshold low enough that all particles are marked with the red cross, but not so low that you are marking optical noise and background effects. The following image is from the manual and illustrates these two sides of the issue.

To get a good feel for the effect of changing detection threshold, take one video and analyze it over and over again with different detection thresholds. Until you get obviously too far in one direction, the results don’t change by large amounts with small differences in the setting. Being able to watch the particles being tracked on a screen should make it obvious when we are missing particles or picking up noise.

Regarding the Malvern’s Zetasizer Particle Size Analyzer. There is no sample preparation for the 20 nm – 10 um standards (Polystyrene beads). Where can I find it on your website?

We don’t have a documented procedure but I recommend this blog post.

DLS or NTA to measure diameter less than 80 nm?

DLS can measure most materials to sizes below 1nm, but NTA is also very capable in the <80nm range. Depending on the refractive index of the material, the lower limit might be ~10nm for metals, around 30nm for polymers, and ~40nm for liposomes. As long as we are able to see the particles, then the analysis of that image for size determination is easy and robust.

Can we confirm the internalization of gold nanoparticles into cells?

NTA will not be able to see particles inside of another object. The light scatter necessary to see the nanoparticles will scatter from the surface of the cell and that will be the only thing picked up by the camera. A direct microscopic technique would be required to see internalized particles.

Related content: