Determination of the water droplet size distribution of fat spreads using confocal scanning laser microscopy

J Microsc. 2002 Nov;208(Pt 2):116-33. doi: 10.1046/j.1365-2818.2002.01075.x.

Abstract

Knowledge of the water droplet size distribution of fat spreads is necessary for the development and production of high quality microbiological safe products. Fat spreads are water-in-oil emulsions. The water droplet size distribution can be determined by confocal scanning laser microscopy (CSLM) after staining the fat with Nile Red. The profiles of the non-fluorescent water droplets in the 2D images are identified and measured using image analysis. The 'true' water droplet size distribution is calculated from the distribution of the measured profile diameters using a Wicksell transformation of log-normal distributions. The influence of the fluorescent staining and CSLM parameters on the information were studied. The CSLM method was tested on fat spreads with a fat content ranging from 40% to 80%. The results were compared with those obtained by nuclear magnetic resonance spectroscopy (NMR). The distribution parameters [volume weighed geometric mean diameter (D3,3) and the standard deviation (sigma) of the logarithm of the droplet diameter] calculated for 80% fat spreads are in good agreement with those obtained by NMR (within +/- 7% relative). Small differences were found for 65% fat spreads and large differences were identified for 40% fat spreads. The precision for the determination of the D3,3 value by CSLM is worse than that of NMR, even when three images were used to calculate this parameter [3-11% and 1-6% relative standard deviation (RSD), respectively]. The precision for the determination of exp(sigma) by CSLM is comparable or better than that of NMR (1-5% and 3-6% RSD, respectively). CSLM proved to be a reliable method for the determination of the water droplet size distribution of margarines (80% fat). The advantage of CSLM compared to NMR is that visual information is given about the water droplet size distribution in the sample.