Quality of Particle Sample Measurement
Quality in a particle characterization measurement can be defined using the term “accuracy”, which is composed of both bias and precision.
Bias is a measure of how the measured value compares with the “true” value or commonly accepted value. This true value may be established through other (certified) reference measurement methods or simply unknown. The best way to avoid bias, including bias in instrumentation, sampling, and operation, is to verify the whole measurement process or to calibrate the instrument using a reference material.
Reference materials can be a user’s own product reference, a national standard reference material, or a secondary reference material traceable to the national standard. Bias is also often associated with the inherent resolution of each technology. Figure 1 shows a comparison of the particle size distribution of a polystyrene latex sample measured using the electrical sensing zone method and the laser diffraction method. Although the mean values from the two methods are very close, the difference in the size distributions is due to the difference in resolution of these two methods.
Precision is the degree of agreement from one measurement to another. According to the condition of comparison, there are three terms related to precision:
- Standard deviation is defined as the degree of agreement in the results from repeated measurements on the same instrument by the same operator;
- Reproducibility is defined as the degree of agreement in the results from repeated measurements on the same instrument but by different operators following the same operational procedure;
- Repeatability is defined as the degree of agreement in the results from measurements with the same conditions but by different operators using different instruments that may or may not be at the same location.
In measuring particle size distributions, besides errors introduced due to imperfection in instrument design and malfunctions of instrument (not in proper working condition, inappropriate setup, abuse of the instrument, or the wrong choice of data interpretation model, etc.), there are several other error sources that should be mentioned. Improper sampling and sample dispersion commonly introduce errors. The former may be non-representative sampling from the batch sample or non-uniform sampling during measurement; the latter may be inadequate or wrong dispersion procedure, causing agglomeration, attrition, swelling, or air bubbles. Generally, the possibility of sampling error increases with increasing particle size; conversely the possibility of dispersion error increases with decreasing particle size.Posted on