Milk is composed principally of casein micelles and droplets of fat together with some mineral salts . The particle size of the casein micelles is around 100 to 200 nm (diameter) with the distribution of fat droplet sizes extending to some microns. The most numerous particles by far are the casein micelles, which also dominate the volume size distribution. The purpose of this study is to investigate the dependence of the measured electrophoretic mobility (EPM) and hence derived zeta potential (ZP) on the sample concentration. The conductivity of a dispersion usually has a strong influence on the zeta potential since the presence of ions shields the surface charges, and generally increased ionic concentration produces a reduced magnitude of zeta potential. The conductivity of milk is around 5mS/cm and hence in this study the dilutions were made by adding approximately 50mM NaCl adjusted to the same conductivity as the original milk sample.
Samples of bovine milk, skimmed and semi-skimmed were obtained from normal commercial sources.
The milk samples were successively diluted by factors of 2, by volume, with an estimated tolerance of 5% which is considered sufficiently accurate for the purpose, since the dependence of measured zeta potential value on concentration is relatively weak. Around 1 mL of sample was injected through the high concentration zeta potential cell in each case to remove the previous sample and replace it with a fresh aliquot. The cell was held at a temperature 25°C (to a tolerance of 0.1 degrees C) by the Zetasizer Nano temperature controller. Two minutes equilibration time was allowed before running the measurement which consisted of 3 repeats using the standard measurement conditions for the ZEN1010 cell. A field of 40v was applied across the nominal electrode spacing of 16mm.
The 'raw' zeta potential value was converted from the measured electrophoretic mobility using the Smoluchowski equation [2-4], but assuming the viscosity value of water at 25°C. The 'corrected' value used an actual measured value for viscosity obtained using an SV-10 vibro-viscometer .
This technique measures a viscosity which must be divided by the liquid density to obtain the viscosity required by the Smoluchowski formula. The density of the milk samples was measured using a weighing bottle and balance and used to derive the required viscosity.
Tables 1 and 2 summaries the measurement data obtained for the various dilutions of skimmed and semi-skimmed milk respectively. The tables show the % concentrations, measured viscosities and raw and viscosity-corrected zeta potential values.
Figures 1 and 2 show the raw and viscosity-corrected zeta potential values plotted as a function of the % concentration for the skimmed and semi-skimmed milk samples respectively.
The results shown in figure 1 and summarized in table 1 indicate that the raw zeta potential data for skimmed milk shows a steady increase in magnitude attaining a steady-state around -19 mV when dilution has produced a concentration of less than 10% of the original value. The standard deviations of the zeta potential measurements are not shown but were all of the order of +/- 0.6 mV. Systematic variations from run to run of the order of 1 mV are to be expected in this type of measurement for a variety of reasons: not least that measurement of these samples of conductivity around 5 mS/cm involves some Joule heating due to the passage of the electric current.
|% Concentration||Viscosity (mPa.s)||Zeta Potential (mV)|
The clear trend seen in the zeta potential values (table 1 and figure 1) can be due to a number of factors. It could be real due to the change in sample equilibrium. However, by diluting using an indifferent or 'spectator' ion and preserving the overall conductivity, it is hoped that this is not the case. Another possibility is that obscuration of light transmission leads to the collection of signal from the edge, rather the center, of the cell so leading to the real electric field being lower than assumed and the measurement of more slowly moving particles near the edge of the cell. Again the high concentration of casein micelles and some remaining fat droplets not removed in the skimming process may lead to a higher net viscosity of the overall medium seen by the moving particles than has been assumed.
Measurement and use of the sample viscosity values certainly removes the bulk of the zeta potential trend for concentrations of 50% and lower for the skimmed milk experiment. However, somewhat paradoxically, the highest concentration now shows an apparently enhanced zeta value. It is not certain what the cause of this is, but the most likely reason is that the macroscopic measurement of viscosity does not entirely reflect the solution conditions experienced by the particles - the system in effect displays some non-Newtonian behavior.
The semi-skimmed milk data summarized in table 2 and illustrated in figure 2 shows a contrasting situation. Here the effect of viscosity reduces, but does not remove, the rising trend in apparent zeta potential until a concentration of around 10% is reached. This is likely to be due to the greater fat content in the form of micron sized droplets which contribute strongly to light attenuation and possibly short range particle interactions which are not sensed by the viscometer.
|% Concentration||Viscosity (mPa.s)||Zeta Potential (mV)|
The high concentration zeta potential ZEN1010 cell enables measurements of skimmed and semi-skimmed milk to be made from natural concentration down to very high levels of dilution. The lowest concentrations measured in this study are by no means limiting. The use of auxiliary viscosity and density measurements are useful in inferring true trends and enabling absolute measurements up to 50% in concentration. The data reinforces the need to dilute under conditions which do not change the chemistry of the dispersion, and the need to perform concentration studies to identify suitable measurement conditions. Semi-skimmed milk requires some level of dilution to obtain concentration independent results.