Issue link: https://www.e-digitaleditions.com/i/1518209
22 April 2024 Inhalation for testing DPIs (Figure 2). However, in principle, nearly any filter holder could be used to test any filter disc. Testing with the DUSA recommended for DPI testing [4, 5] as the filter holder, however, enabled manageable flow rates to be achieved at convenient pressure drop values and also provided for the testing of 47-mm-diameter filters that inhaler testers typi- cally use for this purpose. e DUSA also had a con- venient pressure measuring point (typically labeled as the P1 sampling port; "Pressure Tap" in Figure 2) that assisted with measuring the gas density approaching the filter disc and measuring the pressure drop across the filter-on-test. We checked the critical dimensions of the DUSA (34.85 mm diameter, 12 cm length; Table 1 of USP Chapter <601>). We noted that the filter support screen inside the exit cap of the DUSA is no longer made; rather the exit cap is now offered with an improved filter support (Copley Scientific product 8601; Nottingham, UK), whose surface is less likely to cause a filter-on-test to rupture under the stress of the measurement process. We experienced no such filter failures during our testing. In the standard DPI DUSA, with its inside diame- ter specified as 34.85 mm [4, 5], the front face of the filter disc through which the air flows has an area of 9.539 cm 2 . We used the face velocity (F V ), defined as the volumetric flow rate approaching the filter divided by the filter area, as the parameter for correlating with the observed pressure drop (∆P) across the filter. Initially, we set the flow rate at approximately 30 L/min (actual, at the inlet or front face of the filter) and increased it by increments of 10 L/min, up to 100 L/min when testing both the filter cartridges and filter discs. In the second part of the test pro- cedure, we decreased this flow rate slightly to 95 L/min, and then we continued to decrease the flow rate by increments of 10 L/min, down to 35 L/min. We then tested again (with the same filter cartridge but always with a new filter disc, when testing discs), by starting the flow rate at 100 L/min and decreasing it to 30 L/min and then, in the second part, return- ing from 35 L/min to 95 L/min. At every flow rate, we recorded the pressure drop across the cartridge or filter being tested. is test protocol guarded against unexpected experimental errors caused, for example, by flaws in any individual filter media disc and also enabled us to examine the possibility that the result would depend on the starting flow rate (hysteresis). e small number of replicates of our determinations did not support strong statistical analysis but was deemed sufficient to support the exploratory nature of this study. Results and discussion Filter cartridges e breathing filter made by the Pall Corporation had substantially smaller flow resistance than the Figure 3 Flow resistance of TSI HEPA 1602300 compared to that of Pall BB50T breathing filter Pall BB50T Pressure Drop (Pa) 2400 2200 2000 1800 1600 1400 1200 1000 800 600 400 200 0 0 50 100 150 Volumetric Flow Rate, Q (L/min) TSI HEPA 1602300 Curve Fit Parameters (Equation 3) Manufacturer and Grade A (Pa per cm/sec) 1000*B (sec/cm) Tremont 934AH 56.1 0.981 Tremont, Grade A-E 58.8 0.945 Whatman 934AH 61.6 1.05 Advantec GA-55 64.7 1.07 Tremont, Grade E 69.7 1.00 Whatman GF/A 70.8 1.12 Sartorius Glass Fiber Filter MGA 76.5 1.15 Pall 61631 A/E Glass Fiber Filter 82.1 1.13 Tremont, Grade A 98.0 1.16 Table 3 Flow Resistance Coefficients for Glass-Fiber Filter Media (listed in order of lowest to highest values of "A")