Good Cascade Impactor Practice (GCIP) and considerations for "in-use" specifications

Abstract

The multi-stage cascade impactor (CI) is widely used to determine aerodynamic particle size distributions (APSDs) of orally inhaled products. Its size-fractionating capability depends primarily on the size of nozzles of each stage. Good Cascade Impactor Practice (GCIP) requires that these critical dimensions are linked to the accuracy of the APSD measurement based on the aerodynamic diameter size scale. Effective diameter (Deff) is the critical dimension describing any nozzle array, as it is directly related to stage cut-point size (d50). d50 can in turn be determined by calibration using particles of known aerodynamic diameter, providing traceability to the international length standard. Movements in Deff within manufacturer tolerances for compendial CIs result in the worst case in shifts in d50 of <±10%. Stage mensuration therefore provides satisfactory control of measurement accuracy. The accurate relationship of Deff to d50 requires the CI system to be leak-free, which can be checked by sealing the apparatus at the entry to the induction port and isolating it from the vacuum source and measuring the rate of pressure rise before each use. Mensuration takes place on an infrequent basis compared with the typical interval between individual APSD determinations. Measurement of stage flow resistance (pressure drop; ΔPstage) could enable the user to know that the CI stages are fit for use before every APSD measurement, by yielding an accurate measure of Deff. However, more data are needed to assess the effects of wear and blockage before this approach can be advocated as part of GCIP.

Fig. 1

The relationships between specifications for critical dimensions associated with cascade impactors

Fig. 2

Strategy for implementing CI System leak testing as part of “in-use” GCIP

Fig. 3

Cross-section through an ideal single-stage impactor showing the motion of particles of different sizes; W is the nozzle exit diameter (a critical dimension for verification) and S is the nozzle exit to collection surface distance (relatively insensitive to variability in a well-designed system)

Fig. 4

The relationship of effective diameter D _eff from annual mensuration (one data point per mensuration) to predicted D _eff from flow resistance measurements (stage pressure drop) for stage 4 (nominal D _eff = 1.207 mm) for six NGIs; the gray square represents the range for D _eff allowed by the manufacturer tolerance for this stage (±0.01 mm) in Table V

Fig. 5

Relationship between optically mensurated area-mean diameter and flow resistance-predicted area-mean diameter (D*) for stage 6 of an NGI, intentionally blocking two, four, six, and eight out of the 396-nozzle array for this stage; the fully open condition is also shown at far left of the figure