Issue link: https://www.e-digitaleditions.com/i/1518209
Inhalation April 2024 25 the influence of the filter support media. However, the influence of filter support media is an issue that likely requires the tester to validate the overall filter assembly, not simply the filter media. Conclusions We have undertaken a small-scale laboratory study to investigate the variability in air-flow resistance of a range of cartridge- and disc-based filter media using measurement arrangements comparable with the air flow configurations that apply when determin- ing DDU for inhalable drug products. We focused on the DUSA-based sampling arrangement used for DPI testing in the pharmacopeial methods firstly for experimental testing convenience and secondly because the overall flow resistance of the sampling system is known to determine the shape of the flow rate-rise time profile that influences how the powder in a DPI is aerosolized and subsequently transported the support screen for the filter contributes signifi- cantly to the total flow resistance for the filter assem- bly. We note that for the testing of filters using the rNGI configuration, the open area of the support screen was approximately 32%, whereas for testing filters in the DUSA, the open area of the hexagonal support mesh (Figure 2) was approximately 63%. So, the screen open area is double in the DUSA testing compared to the situation in the rNGI test- ing. And although it would be incorrect to sup- pose that filter media is inactive simply because it touches a support surface on the downstream side of the filter, this difference in open area is quite substantial. e fact that the non-linear factor "B" in the rNGI testing was approximately six times that observed for the testing in the DUSA is con- sistent with a greater influence of a factor outside of the filter media itself. Neither the current study nor that of Roberts [12] was aimed at investigating Pressure Drop Curve Fit Coefficients* 47 mm Diameter Disc in DPI DUSA 42.5 mm Diameter Disc in rNGI Configuration Filter Media A (Pa/(cm/s)) 1000*B (s/cm) A (Pa/(cm/s)) 1000*B (s/cm) Whatman Grade 934AH 61.6 1.05 83.9 5.97 Tremont Grade 934AH 56.1 0.981 69.5 6.65 Tremont Grade A-E 58.8 0.945 77.7 6.19 Advantec GA-55 64.7 1.07 74.1 6.73 *As in equation 3: ∆P = A*F v *(1 + B*F v ); pressure drop is dominated by the linear "A" coefficient Table 4 Comparison of Current Test Results with Results Obtained in a Reduced NGI