Deposit Maldistribution in the Andersen Non-Viable Mark-II Cascade Impactor (ACI): An Initial Investigation
Background: Flow uniformity is assumed in cascade impactors used in the particle size assessment of aerosols from orally inhaled products. We have observed severe radial maldistribution of the deposits on stages 0 and 1 of the Andersen 8-stage non-viable impactor (ACI) which we have studied by calibrating an ACI with monodisperse particles.
Materials and Methods: We introduced a 28.3-L/min steady flow of monodisperse uranine-containing particles into the ACI and varied the inlet particles to include sizes above, near, and below the 50% cut-point (D50) sizes of stages 0 and 1. We photographed the deposit patterns.
Discussion: There are four concentric rings of 24 nozzles each on stage 0 and 1 of this impactor. We found radial, but no angular, maldistribution indicating that each ring of each stage has its own efficiency for particle capture. The deposition pattern indicates that the inner ring has a D50 value smaller than that of the stage as a whole and that the two middle rings have D50 values larger than that of the stage as a whole.
Conclusion: These observations are consistent with studies reported by previous investigators of size distribution measurements made with transient, small total air volumes and of computer flow simulations at steady flow conditions. During unsteady, start-up flow, typical of dry-powder testing conditions, the actual D50 values of stage 0 and stage 1 are likely to be smaller than their steady-state values reported in the pharmacopeial compendia, since flow through the inner ring is clearly the path of least resistance.