1) From your explanation I understood that some kind of Joule heating of your particles has happened. Maybe you should try manually decrease the voltage.

2) What about conductivity and viscosity of your samples? As I know, for high conductive solutions you should change measurment from auto mode to high conductive mode. Also, measurment of zeta potential by M3-PALS (Malvern) allows to measure electrophoretic mobility of particles (and calculate zeta potential) and depends on viscosity: U=2*epsilon*Z*f(ka)/(3*nu) (nu is dynamic viscosity, epsilon - dielectric constant: http://www3.nd.edu/~rroeder/ame60647/slides/zeta.pdf) So, for each dispersant you sholud measure (or google) dielectric contant and viscosity and create a new dispersant (in measurement options).

Hello Daniel, there is a technical note with some tips on making measurements in high salt at http://www.malvern.com/en/support/resource-center/technical-notes/TN101104HighConductivitySamples.aspx which you could try. In your case the use of the diffusion barrier method ( http://www.atascientific.com.au/publications/wp-content/uploads/2013/03/MRK1651-02.-The-Diffusion-Barrier-Technique-Practical-Aspects-and-Data-interpretation.pdf ) or a dip cell may help to understand the details of your observation.

In high salt, the screening of the charges on your nanoparticles may lead to a very small effective zeta potential, which would explain the noisy phase plots you describe. You could also measure the DLS size before and after the zeta measurement, this would indicate whether the particles themselves are changing (maybe the high salt is destabilizing them, leading to aggregation). The color change of the whole sample could be an indication of some interaction between your sample and the electrode material. The DLS derived countrate of the sample before and after it turns may indicate if additional particles or larger particles were created in your sample in the process. Does the hydrodynamic size (z-average and polydispersity) change significantly after the voltage was applied in the zeta measurement?

First, let me say thank you for all of the answers, it is greatly appreciated.

Mr. Mikhailov, the appropriate viscosities and dielectric constants (at the temperature the measurements are being taken at) are being used for the various buffers in question. Interestingly, there appears to be a correlation between the discoloration of the sample and the conductivity (conductivity increases once the sample becomes brownish-black in color).

Dr. Schcharbin, I have tried titrating the concentration of the buffers, with limited success. At lower buffer concentrations the discoloration of the solution (from clear to a brownish-black hue) is more mild, and usually only occurs on the 3rd measurement (instead of the 2nd in the case of higher buffer concentrations). Also, when I perform a zeta measurement on the buffer solutions alone (without any nanoparticles), I run into the same issues of measurement variability and discoloration of the sample.

Mr. Nobbmann, the z-average size and count rates of nanoparticles suspended in the buffers do not change when the sample becomes discolored. Also, the buffers alone (without nanoparticles) produce the same discoloration on the 2nd-3rd measurement, and the count rate decreases substantially (1/10-1/100th of the count rate of the 1st measurement), and the conductivity of the sample increases when the sample becomes discolored.

All of this together (change in zeta potential, sample color, count rate and conductivity) when measuring buffers alone makes me suspect that the interaction between the electrodes and the buffer components is possibly what is causing the problem.  Does this seem likely?  

I guess your sample is a polyionic in nature. Try to reduce the concentration as much as possible and then run, u may run in manual mode with single run each time. have a look at mS value.

Dear Daniel,

I basically have the same problem: I have the sample in PBS 1X and I do not obtain a good response to the Zeta measurement (more random peaks and a "dirty" function plot). Finally, after few measurement, the electrodes of the cuvette became black (like burned). How did you manage finally? Thank you in advance.

I know this is an old question but here's the correct answer:

Just as electrical double layers form around charged particles, so they also form at the electrode surfaces when the voltage is applied. This creates polarization near the electrode that introduces capacitance. Increasing the salt concentration increases this polarization. This leads to a reduction in the voltage "seen" by the particles. Because instruments assume the voltage "seen" is the same as applied, the zeta potential results are wrong (underestimated by an unknown amount).

The extent of this problem depends on the material used to make the electrodes. Both gold and platinum (the two most common) cause too much polarization at the frequencies of the electric field used in commercial instruments.

The only solution to this is to use blackened platinum electrodes and higher electrode frequencies than achievable with commercial instruments.

IMPORTANT! Electrolysis occurs at high salt concentrations especially with gold electrodes in chloride solutions. If you almost certainly generate gas bubbles and corrode the gold. With time, this increases the polarization leading to unpredictable zeta potential errors not only at physiological salt concentrations even at 0.5mM. Some commercial electrodes can prevent contact of the particles with the electrodes but they do not solve the fundamental problems.

There's also significant amount of heat generated that can cause turbulence leading to false measurements. In PBS X1, a typical voltage used with a dip electrode will generate nearly one amp of current - enough to heat the sample by 1degC per *second*. Trying to generate such currents may even damage your instrument's electronics.

I have developed a next generation electrophoretic light scattering system (NG-ELS) that solves these problems and allows correct measurements as high as 4M.

I presented an overview of this to the ACS National Meeting in 2018:

https://tinyurl.com/ngels

I cannot emphasize enough...

1. No commercial instrument will give you meaningful results under physiological conditions, especially with gold electrodes. Sure, you will get nice looking results but they will be wrong.

2. YOU WILL DESTROY YOUR EXPENSIVE ELECTRODES if you continue to use them at high salt concentrations. Damage will occur before you can see it. It will invalidate your measurements unpredictably. This will happen after just a few hours of use. Subsequent measurements with the same electrodes at lower salt concentrations will eventually become erroneous. If you have already seen your electrodes turn black, the damage has begun.

Only platinum black electrodes can reduce the polarization sufficiently and also prevent irreversible damage. Such electrodes are not available commercially for zeta potential measurements.

These are bold assertions. My credentials are that I invented the phase analysis method used in instruments such as the Zetasizer Nano ZS, ZetaPALS, Mobius etc and have helped some companies develop their instruments.

To add to my previous answer, here's a photo of gold electrodes (Brookhaven AQ-927 dip electrodes for aqueous samples) that are extremely damaged.

They had been used previously in PBS X1 (150mM NaCl(aq)) a few times and then "cleaned" to restore the gold appearance.

This photo shows the build-up of the black material after approx. 1 minute with 5V applied in 10mM KCl(aq).

There were no particles present - just the electrolyte.

Contrary to common belief, the build-up of black material on gold electrodes isn't due to particles but simple electrochemistry.

I deliberately damaged these electrodes for my instrumentation development.

DO NOT try to make measurements in physiological-like liquids with your electrodes!

DO NOT try to disassemble the electrode assembly to clean the gold!

These are palladium "dip" electrodes - the type used in the leading zeta potential light scattering instruments.

They are immersed in PBS X1 (phosphate buffered 150mM saline).

A mere 8V was applied across the electrodes for a couple of minutes. This is what happens inside your sample when you try to measure under physiological conditions. This is why you see your lovely shiny electrodes turn black. And it happens at 10mM, too.

So now you know how to measure the zeta potential of some unknown brown palladium gunk instead of your sample.