Filtration in Particle Size Analysis

Abstract

Filtration is defined as the separation of solid particles from a liquid or gas by passing the suspension through a porous, fibrous or granular substance. In science and industry, filtration is commonly used for the concentrating and/or removal of particulate matter or for the separation of particles into size classes for subsequent physical, chemical or biological analysis. Separation of particles into multiple size classes is often required for the assessment of particle size. There are many different types of filters available (e.g. fibrous, membrane, polycarbonate) for particle characterization through filtration. The type of filter to be used for particle size characterization is dependent on a number of considerations such as sample (particle) characteristics, collection efficiency, pore size and structure and analytical compatibility. Filtration has the advantage of being one of the few techniques which allow for the physical separation of particles into different size classes for distribution or chemical–biological analysis. If appropriate precautions are taken to minimize sample perturbations (e.g. select appropriate filter for the application, filter in the field, avoid overloading a filter, ensure a low flow rate and stir if possible), a reasonably accurate estimate of particle size can be obtained. Generally, errors associated with filtration arise from filter clogging and when attempting to characterize flocculated/aggregated particles. The basic nature of a suspended aggregate (i.e. fragile, plastic, high water content) means that it is likely to be broken up or created owing to the often tortuous nature of filtration, or be deformed on the surface or within the filter owing to pressure effects. This has implications if optical image analysis of the filter is to be used for particle size analysis as the true particle morphology and size will have been altered. In such an instance, if only a grain size distribution is required, it is suggested that optical image analysis be employed in conjunction with a plankton chamber settling technique rather than with filtration.

Particle size distribution and optimal capture of aqueous macrobial eDNA

“…Filtration can bias particle size analysis by retaining particles smaller than the filter pore size or passing particles larger than the pore size (Droppo 2006). However, our analysis required chemical assay of the SPM to quantify its subcomponents, total eDNA and Carp eDNA.…”

“…For these reasons we used sequential filtration size fractionation (Figure 1) to generate the PSDs for SPM, total eDNA, and Carp eDNA. We followed the recommendations of Droppo (2006) to minimize filtration artifacts: low vacuum pressure, small water volume, and filters with precise pores. We used nylon net filters (180, 100, and 60 !m mesh size; Millipore) and track-etched polycarbonate (PCTE) filters (20, 10, 1, 0.2 !m pore size; GE Osmonics) because these provide the sharpest separation between particle sizes and the best agreement between nominal and effective pore size (Sheldon 1972;Droppo 2006).…”

“…We followed the recommendations of Droppo (2006) to minimize filtration artifacts: low vacuum pressure, small water volume, and filters with precise pores. We used nylon net filters (180, 100, and 60 !m mesh size; Millipore) and track-etched polycarbonate (PCTE) filters (20, 10, 1, 0.2 !m pore size; GE Osmonics) because these provide the sharpest separation between particle sizes and the best agreement between nominal and effective pore size (Sheldon 1972;Droppo 2006). By contrast, 'tortuous path' type filters such as glass-fiber and cellulose membrane actually have a wide array of effective pore sizes in every filter, rather than well-defined pores (see Droppo 2006 for a detailed discussion).…”

“…We used nylon net filters (180, 100, and 60 !m mesh size; Millipore) and track-etched polycarbonate (PCTE) filters (20, 10, 1, 0.2 !m pore size; GE Osmonics) because these provide the sharpest separation between particle sizes and the best agreement between nominal and effective pore size (Sheldon 1972;Droppo 2006). By contrast, 'tortuous path' type filters such as glass-fiber and cellulose membrane actually have a wide array of effective pore sizes in every filter, rather than well-defined pores (see Droppo 2006 for a detailed discussion). Thus our PSDs, while accurate, may not translate directly to the nominal pore sizes of tortuous path filters, and we recommend the use of PCTE filters for future particle size analysis of aqueous macrobial eDNA.…”

A multicomponent model of phytoplankton size structure

“…The two in situ methods (HPLC and SFF) have advantages and disadvantages. SFF data explicitly partition the size classes, but difficulties arise from inaccuracies in pore sizes, filter clogging, and phytoplankton cell breakage [Droppo, 2000]. Whereas the method of Welschmeyer [1994] was used in the determination of chlorophyll a fluorometrically, which minimizes the influence of chlorophyll b, chlorophyll c 2 , and pheopigments on chlorophyll a estimates, it is possible that this influence varies depending on the size fraction of interest, especially when considering chlorophyll b is likely present in higher concentrations in the smaller size classes [Vidussi et al, 2001;Uitz et al, 2006] and at the deep chlorophyll maximum .…”