Current Approaches to APSD Measurements of OIPs Based on Inertial Impaction - Good Cascade Impactor Practices, AIM and EDA for Orally Inhaled Products - page 30

Table 2.8 Stage d 50 values ( μ m) for the Marple-Miller five-stage CIs

Model and design flow rate ( Q )
150P 150 160

Stage 4.9 (L/min) 12.0 (L/min) 30 (L/min) 60 (L/min) 90 (L/min)

1 10.0 10.0 10.0 10.0 8.1

2 7.2 4.7 5.0 5.0 4.0

3 4.7 3.1 2.5 2.5 2.0

4 3.1 2.0 1.25 1.25 1.0

5 0.77 0.44 0.63 0.63 0.5

Table 2.9 Design characteristics of the Marple-Miller CI variants at Q = 4.9, 12.0, 30, 60, and
90 L/min
Model Parameter Q (L/min) Stage 1 Stage 2 Stage 3 Stage 4 Stage 5

150/160 Re f 30, 60 3,160 1,240 1,240 1,260 1,260

90 4,740 1,860 1,860 1,890 1,890

S / W 30, 60, 90 1.0 1.0 1.0 1.2 2.5

X c 30, 60, 90 N/A 0.37 0.37 0.37 0.36

150P Re f 4.9 918 539 453 386 548

12.0 1,765 1,320 1,109 945 1,342

S / W 4.9 1.71 5.36 4.43 3.84 5.03

12.0 1.31

X c 4.9, 12.0 N/A 0.18 0.21 0.25 0.17

These CIs were first designed for OIP testing to make use of collection cups as
particle collectors, with the purpose of improving productivity. However, API
recovery can be more difficult than with the simpler geometry of collection plates
with recovery procedures requiring more than contact with solvent to dissolve the
collected particles. The model 160 (high flow) MMI is the standard configuration
intended for use at 60-90 L/min. The model 150 MMI has half the number of
nozzles per stage compared with the model 160 MMI, for use as an alternative to the
ACI for MDI characterization at 30-60 L/min.
Although these CIs have not been as widely adopted as the ACI, NGI, or MSLI,
internal losses for models 150 and 160 MMIs reported by Marple et al. were no
more than 5% of the incoming aerosol at worst case (4 m), decreasing

to <1% for finer particle sizes and <2% for larger particles [ 35 ]. These measure-
ments were based on calibration with monodisperse droplets.
The model 160 MMI has subsequently been reported as having internal losses at
60 L/min with at least one DPI (Bricanyl Turbuhaler ® , AstraZeneca, Sweden) that
were comparable with those indicated by Marple et al. [ 35 ], provided that precau-
tions were taken to eliminate particle bounce and re-entrainment by coating the
collection surfaces with a tacky surface (silicone oil) [ 37 ]. By comparison, losses
within the low-flow MMI have also been reported as being <5% of the material bal-
ance from two types of MDI-generated formulations [ 36 ].

μ m < d ae < 6 μ