Automated Static Image Analysis is a particle sizing technique for dry analysis of powders that makes use of digital imaging.
Dry analysis has always posed a difficult challenge for most particle sizing techniques because particles are difficult to disperse dry. Furthermore, the techniques that are most often used for dry analysis like sieving and laser diffraction provide only low resolution distributions. Automated Static Image Analysis characterizes individual particles to build up a distribution. Number based techniques like this always deliver higher resolution particle size information than ensemble techniques like laser diffraction. Furthermore, the size information from laser diffraction is expressed as a spherical equivalent diameter which by definition provides a particle size that is not derived from the actual particle.
The technique is implemented by dispersing a dry powder on a large glass slide or other optical surface. The slide is then translated in the X-Y plane underneath a digital camera mounted behind a magnifying lens. Images are captured that contain many particles and algorithms are used to identify the particle perimeters, then from that, to calculate a variety of size and shape parameters. After analyzing thousands of particles, distributions can be determined. Because of the power of modern desktop computers, the analysis of a dozen size and shape parameters for thousands of particles can be performed in minutes making the measurement much faster than sieving. Compact vacuum dispersers can be used to disperse the particles onto the glass slides in such a way that the dispersion is complete but the energy used is low enough to preserve delicate crystals like those from pharmaceutical active ingredients. An example of the power of image analysis can be seen below.
The images are from several carbon rod samples that were analyzed with the 500 Nano using it’s built-in vacuum-disperser. The length and width of 10,000 particles was determined for each sample. The distribution information is seen below. The first graph are the width distributions for each sample. It indicates that for multiple samples, the width of the rods was consistent at around 10 microns and narrow.
The same cannot be said for the lengths of the rods, the distributions of which are shown below. The various samples had different average lengths (from 60 to 80 microns) and all the length distributions were wide with one sample having rods as long as 400 microns. It is important to note that only Automated Static Image Analysis could have provided this detailed information.