Traditionally, FPM is collected on filters and analyzed by light microscopy—a manual task that is both time-consuming and subjective. Fully automated image analysis techniques are now accelerating the process, generating reproducible results and removing operator subjectivity. Accurate and reliable automated FPM characterization does, however, require the particle characterization system used to have a number of key attributes.
What are FPMs?
In automotive fluids, residual particles from manufacturing and assembly processes can increase wear rates with the potential for failure. Particulate contaminants require strict measurement and control, which the ISO 16232 series of standards describes. Hydraulic fluid systems, where particulates may cause serious problems that ultimately lead to failure and downtime, are similarly covered by ISO4406/ISO4407.
For the pharmaceutical industry, the source of contaminant particles is typically the active or excipient material or, from the administration device, through abrasion or shearing during operation. Glass, transparent fibers, stainless steel, rubber, aluminum and plastic particles are all encountered, and a variety of regulatory and guidance documents covers different devices and systems.
Keys to successful measurement
Analysis of FPM on filters has until recently been the preserve of manual light microscopy. Recently introduced particle characterization systems that combine light microscopy and image analysis bring new levels of speed, automation and objectivity while simultaneously delivering large amounts of information about the size and morphology of individual particles. This makes detailed investigation and effective troubleshooting much more possible, heralding better process and quality control. With many users of manual microscopy now looking to move to automated methods, some of the key factors for success should be considered. A system that combines advanced hardware with image analysis software dedicated to particle characterization is a given, but making measurements in filters demands some additional capabilities.
Figure 2. Sections of a field-of-view frame with (right) and without (left) z-stacking.
|
Filter integrity
Keeping the filter stretched flat is an absolute requirement for automated analysis and relies on a robust filter holder. A useful development is the glassless carrier (Figure 1), which eliminates any risk of counting particles on the glass surface or crushing particles beneath it.High resolution
In the latest systems, high-quality microscope objectives together with scientific-grade cameras enable the analysis of particles down to 2 µm. Even at such small sizes, images contain enough pixels to be confident that they are genuinely particles.
z-stacking
z-stacking allows images of each frame, taken at different focal points, to be combined before the particles are separated from the background. Large particles remain in focus while any small particles in the well of the filter paper are counted. Figure 2 shows a section of a frame containing a fibrous particle with and without z-stacking.
Particle stitching
If particles cross two or more frames, as is fairly typical for fibers, it is important that the software identifies the frames that contain the particle image, “stitches” the pieces of the particle together, and then extracts the whole edge-stitched particle from the background.
Detecting both reflective and dark particles
Whether particles are highly reflective metal or dark rubber, detection in a single system is important. Similarly, any particle containing both dark and light regions must be detected as a single particle rather than as fragments.
Figure 3. Composite image and X versus Y position scattergram showing the position of a particle of interest on the filter.
Click to enlarge.
|
Composite image generation
The ability to view a composite image of the entire scanned area serves a number of purposes, including verification that there were no faults in the filter paper itself. A combination of composite image and scatter plot can also visualize the position of particles of particular interest on a filter (Figure 3). Additionally, the composite image can help assess the quality of filtration, verifying that the particles are evenly distributed over the filter and not limited to the edge of the scanned wetted area.Looking forward
By removing the tedium and subjectivity from FPM analysis on filters--a task that is obligatory in many areas of industry, automated image analysis techniques open the door to easier, faster and more informative characterization. The amount and breadth of information these powerful techniques generate, their ability to store images of every particle, and the facility for extensive data analysis add a new dimension in understanding and controlling processes.
For more information, contact Ulf Willén, product manager, Analytical Imaging Systems, Malvern Instruments at ulf.willen@malvern.com or by phone at +44 1684 892 456.
