Next Generation Pharmaceutical Impactor (A New Impactor for Pharmaceutical Inhaler Testing). Part I: Design ABSTRACT

Transcription

1 JOURNAL OF AEROSOL MEDICINE Volume 16, Number 3, 2003 Mary Ann Liebert, Inc. Pp Next Generation Pharmaceutical Impactor (A New Impactor for Pharmaceutical Inhaler Testing). Part I: Design VIRGIL A. MARPLE, Ph.D., 1 DARYL L. ROBERTS, Ph.D., 2 FRANCISCO J. ROMAY, Ph.D., 2 NICHOLAS C. MILLER, Ph.D., 3 KEITH G. TRUMAN, B.Sc.(hons), 4 MICHIEL VAN OORT, Ph.D., 4 BO OLSSON, Ph.D., 5 MICHAEL J. HOLROYD, M.A.(hons)Cantab., 6 JOLYON P. MITCHELL, Ph.D., 7 and DIETER HOCHRAINER, D.Phil. 8 ABSTRACT A new cascade impactor has been designed specifically for pharmaceutical inhaler testing. This impactor, called the Next Generation Pharmaceutical Impactor (NGI), has seven stages and is intended to operate at any inlet flow rate between 30 and 100 L/min. It spans a cut size (D 50 ) range from 0.54-mm to 11.7-mm aerodynamic diameter at 30 L/min and 0.24 mm to 6.12 mm at 100 L/min. The aerodynamics of the impactor follow established scientific principles, giving confident particle size fractionation behavior over the design flow range. The NGI has several features to enhance its utility for inhaler testing. One such feature is that particles are deposited on collection cups that are held in a tray. This tray is removed from the impactor as a single unit, facilitating quick sample turn-around times if multiple trays are used. For accomplishing drug recovery, the user can add up to approximately 40 ml of an appropriate solvent directly to the cups. Another unique feature is a micro-orifice collector (MOC) that captures in a collection cup extremely small particles normally collected on the final filter in other impactors. The particles captured in the MOC cup can be analyzed in the same manner as the particles collected in the other impactor stage cups. The user-friendly features and the aerodynamic design principles together provide an impactor well suited to the needs of the inhaler testing community. Key words: impactor, pharmaceutical inhaler, calibration, design INTRODUCTION THE PHARMACEUTICAL INDUSTRY uses pressurized metered-dose inhalers (MDIs) and drypowder inhalers (DPIs) as a means to aerosolize medicines for inhalation. The regional deposition in the lung is a strong function of the aerodynamic diameter of the particles. In effect, the lung 1 University of Minnesota, Mechanical Engineering Department, Minneapolis, Minnesota. 2 MSP Corporation, Shoreview, Minnesota. 3 Nephele Enterprises, White Bear Lake, Minnesota. 4 GlaxoSmithKline, Ware, Hertfordshire, United Kingdom. 5 AstraZeneca R&D Lund, Lund, Sweden. 6 Phoqus Pharmaceuticals Ltd., West Malling, Kent, United Kingdom. 7 Trudell Medical International, London, Ontario, Canada. 8 Boehringer Ingelheim Pharma AG, Ingelheim am Rhein, Germany. 283

2 284 is a type of aerosol classifier, with the larger particles collected in the upper airways, mainly at bifurcations in the flow path, and the smaller particles depositing deeper in the lung. Impactors are the instruments of choice for the in vitro assessment of delivery efficiency of inhalation products for three predominant reasons. First, by using a specific drug assay, the size distribution of drug particles that is obtained is drug specific, not confounded with any non-drug material that may be in the sample. Second, the size distribution is measured on all of the drug that is delivered, rather than on a sub-sample that may or may not be representative. Third, impactors classify particles according to aerodynamic diameter (Appendix). Thus, the cascade impactor is a natural choice for an instrument to evaluate the aerosol emitted from an MDI or DPI. However, it is important to recognize that the cascade impactor is not a lung simulator because of many features, including the geometry at the point of impact, collection surface hardness and coating, and operation at constant flow rate. In particular, collection stages in the impactor do not correspond to any specific deposition sites in the lung. The first MDIs were commercially produced in Shortly thereafter, a method was published for sizing aerosols using the light scatter decay technique 2 and was put into use in Riker Laboratories at about that time. 3 This technique specified that the test aerosol be fired into a dark cubic box, the scatter of a beam of light inside the box was monitored, and the mean size of aerosol could be calculated. The use of a cascade impactor for pharmaceutical aerosols was first described in 1969 by Polli et al. 4 from Merck Scharp & Dohme, using the impactor later known as the Delron. This impactor had been first described 10 years earlier. 5 The Andersen cascade impactor was developed in 1958 as a sixstage design to sample airborne viable particles onto agar gel held in a Petri dish. 6 In 1971, an eight-stage Andersen Mark I impactor was introduced in Riker Laboratories. 3 The Multi- Stage Liquid Impinger was originally developed in 1966 by May to sample viable particles into a liquid. 7 The use of this instrument for pharmaceutical applications was subsequently described by Bell et al., 8 and the impactor was later modified by adding one stage to obtain additional resolution for the smaller particle sizes. 9 A feature of this instrument is that particles are collected on a glass frit that is in contact with MARPLE ET AL. liquid, so the problem of particle bounce and reentrainment is thought to be eliminated. Two other devices to size particles by inertial impaction were developed as quality control tools within the laboratories of pharmaceutical companies. These single impaction stage instruments, termed twin impingers, have been used with pharmaceutical aerosols because they permit significant labor savings compared to the considerable time required to analyze samples from a multistage impactor. A glass twin impinger described by Hallworth and Westmoreland 10 alluded to an earlier version described by Hallworth et al. 11 This instrument, along with a metal twin impinger of different design, first appeared in the British Pharmacopoeia. 12 Comparisons were reported by Aiache et al. 13 The two devices, although convenient to use, had inherent limitations in the adequacy of size description, 14 and the measurements were not comparable between the instruments. Nevertheless, they remain in use in some countries because of their operational economy and their historical use in registering some products. Figure 1 shows the instruments recommended for use by the current standards of the U.S. Pharmacopeia 15 and European Pharmacopoeia. 16 The first cascade impactor designed specifically for pharmaceutical aerosol applications was the Marple-Miller Impactor 17 (MMI). Although the MMI did not find favor within industry, it demonstrated that a cascade impactor could be designed specifically for sampling MDI and DPI aerosols and that its unique external sampling cups could offer productivity improvements over existing impactors. Discussions among pharmaceutical industry scientists about the inadequacies of the various available impactors led to the establishment in the mid-1990s of the Next Generation Impactor (NGI) consortium. In time, the consortium recruited members representing all of the major participants of the worldwide inhalation product testing community. The specific purpose of this consortium was to define the requirements for a new impactor and to fund its development and testing. The consortium considered proposals from several parties to undertake this development and awarded the project to MSP Corporation (Minneapolis, MN). The first project meeting was held in December The design of the NGI was a cooperative effort between the NGI Consortium, spearheaded by its Executive Committee, working with the impactor

3 NEXT GENERATION PHARMACEUTICAL IMPACTOR DESIGN 285 FIG. 1. Instruments recommended for use by the U.S. Pharmacopeia and European Pharmacopoeia. (A) Andersentype. (B) Andersen-type with pre-separator. (C) Multistage liquid impinger. (D) Marple-Miller impactor. (E) Twin impinger (metal impinger). (F) Twin impinger (glass impinger). design team. The Executive Committee fielded a continuing stream of design questions and issues throughout the project, giving regular input and guidance. At major points of design decisions, the design team would present sound options to the entire consortium, and the consortium members would select the option that they felt best served the end-user needs. MSP interviewed users inside consortium member companies to get first-hand input. Twice in the design process, prototype

4 286 NGIs were built and tested by the consortium members in their laboratories. The qualitative and quantitative information from the prototype testing resulted in important improvements to the design. This development process was intended to ensure that the NGI would be an instrument that would serve the pharmaceutical industry well for many years. This paper describes the NGI design and the reasons for its particular configuration. The calibration of the NGI, defining the particle collection efficiency curves, is described in a companion paper in this issue. 18 USER REQUIREMENTS The NGI consortium met several times to define items that it considered to be important features of the NGI. The final result was a list of user requirements that were divided in two groups: 23 must features (Table 1) and 12 want features (Table 2). The 35 items listed in Tables 1 and 2 became the guide for all design decisions for the development of the NGI. From a designer s viewpoint, the user requirements can be divided into two categories: (1) those that affect the design of the impactor stages and (2) those that affect the overall impactor layout. The first category relates to the aerodynamic design of the stage nozzles and the flow passages between stages. The particular geometric configuration in these regions determines the collection efficiency characteristics of the impactor. The second category relates more to mechanical features, such as materials, clamping mechanisms and relative location of the stages. These characteristics have negligible effect on the aerodynamic performance of the impactor but have a large effect on chemical compatibility, reliability, and operating convenience. Particular focus was given to ensuring that the ergonomics of the new impactor were optimized from a user perspective and also that the impactor could be automated relatively easily. STAGE DEFINITION MARPLE ET AL. The first step in designing a cascade impactor is to define the number of stages and the particle cut size of the stages at the desired flow rate or flow rate range. The requirements M-2, M-4, M-5, M-7, and W-2 (Tables 1 and 2) most closely TABLE 1. MUST NGI FEATURES M-1 Automatable, but suitable for manual operation M-2 Operates over a range of flow rates L/min M-3 Calibration data for the flow rates of L/min are available M-4 Capable of fully characterizing the less than 10 mm cloud size range M-5 Needs right number of stages at right cut-offs (independent of flow rate); minimum of 5 stages mm, one stage between 5 and 10 mm plus high capacity stage at 10 mm or higher (pre-separator) M-6 The stages will be followed by a micro-orifice collector that is 90% efficient for particles larger than or equal to 0.2-mm aerodynamic diameter (or smaller particles; see W-6 in Table 2) M-7 Efficiency curves for stages must be appropriate; GSD for all stages similar; overlap between stages is minimized M-8 Deposition profile unaffected by stage loadings (up to 10 mg) M-9 Operates using defined entry conditions M-10 The empty volume of the impactor must be no more than 1.5 L, measured from the inlet of the USP/EP inlet to the exit of the impactor body, including the pre-separator M-11 Low wall losses; not more than 5% on any stage and not more than 5% total on all stages M-12 Good drug recovery (mass balance) M-13 No bounce/re-entrainment M-14 Capable of being grounded; unaffected by static M-15 Physically robust M-16 Constructed using inert/robust materials (can use common solvents) M-17 Good accuracy and precision (65%) M-18 Operator independent (no statistical difference between results between independent laboratories) M-19 Acceptable to regulators/pharmacopoeia M-20 Designed and manufactured to ISO 9000 or equivalent M-21 Applicable to all single shot inhaled delivery systems (MDIs, DPIs, aqueous inhalers) M-22 Easy to qualify and validate in the laboratory M-23 Fast; cycle time of less than 30 min for manual determinations

5 NEXT GENERATION PHARMACEUTICAL IMPACTOR DESIGN 287 TABLE 2. WANT NGI FEATURES W-1 Cheap; manual version to cost less than UK 5,000 W-2 Flow rates able to be continuously varied (see M-2 in Table 1) W-3 Although the must specification has a lower flow rate limit of 30 L/min, it is desirable for the lower flow rate limit to be 5 L/min W-4 Capable of simplification for QC use, e.g., able to remove one or more stages W-5 Works on single actuation by minimizing wash volumes W-6 The stages will be followed by a micro-orifice collector that is 90% efficient for particles larger than or equal to 0.1-mm aerodynamic diameter (up to 0.2 mm is acceptable, see M-6 in Table 1) W-7 Capable of characterizing particle cloud greater than 10 mm; pre-separator has to be well characterized and have a sharp cut off W-8 For automated systems, in-built diagnostics (obvious when instrument is functioning correctly) W-9 Capable of being operated over a range of temperature/humidity (20 30 C; 25 75% RH) W-10 Suitable for nebulizers W-11 The impactor must be easy for technicians to use, including handling, training, and overall ergonomics W-12 The empty volume of the impactor should be no more than 1 L, measured from the inlet of the USP/EP inlet to the exit of the impactor body, including the pre-separator (see M-10 in Table 1) relate to the choices that must be made during this first design step. Requirement M-2, which specifies the flow rate range, is the most quantitatively explicit of these requirements. Requirement M-5 puts a lower bound on the number of stages (namely six). Requirement M-7 puts a qualitative upper bound on the number of stages ( overlap is minimized ). Within these constraints, the design team considered several cascade impactor configurations having six or seven stages. These considerations included cases wherein the user would physically substitute nozzle pieces at will, so that the cut sizes could be nearly unchanged regardless of changes in the flow rate (interchangeable nozzles). Consortium members selected seven stages with no interchangeable nozzles so as to best achieve five stages in the range of mm over the entire flow range (M- 5, Table 1) and to eliminate possibilities for user error that might arise if interchangeable nozzles were available. Further, the constraint of minimization of stage overlap was met by insisting on a logarithmic spacing of the particle cut sizes (meaning that the ratio of the cut sizes for any two neighboring stages is a constant throughout the impactor). The logarithmic spacing of cut sizes also aids in the intuitive understanding of mass distribution bar charts, a qualitative issue of value to users. The aerosol dose can be introduced to the NGI through an induction port having the same internal dimensions as the port described in the USP and EP, to meet requirement M-9 and M- 19. OVERALL IMPACTOR LAYOUT The next step in designing the impactor was to choose a physical arrangement of the stages and the associated means of flowing air from one stage to the next. The requirements most related to the overall impactor layout were M-1, M-9, M- 10, M-11, M-14, M-15, M-16, and M-23. Of these requirements, ability to automate, yet suitable for manual operation (M-1) was considered the most important. Most of the requirements were qualitative, and therefore the design team was free to pose a wide range of options for the impactor layout. At least nine separate arrangements of the impactor body were considered. Both for automation and manual use, designs that incorporated external cups of some type appeared to have an advantage over layouts where impactor stages must be disassembled to get to the internal impaction plates. [The Andersen eight-stage impactor (USP apparatus 1) and Marple-Miller (USP apparatus 2) impactor, are both shown in Figure 1 and are examples of internal and external impaction plate layouts, respectively.] Nevertheless, internal impaction plate layouts were considered. One of these was a stacked layout, much the same as the Andersen impactor. However, in this configuration, shown in Figure 2, each nozzle plate incorporated the impaction plate for the previous stage. On one stage, the nozzles would be in the central portion of the stage, and the impaction plate would be in the annular area around the nozzles. The nozzle plate before, and after, this plate would have the noz-

6 288 MARPLE ET AL. FIG. 2. Stacked plate impactor (first three stages). zles in the annular area and the impaction surface in the center area. This layout made for a very compact cascade impactor. Several layouts with external impaction plates were also considered. Some of these layouts were rejected for reasons of anticipated machining difficulties, and the surviving layouts evolved to one with all of the impaction plates in one plane. This configuration resulted in what amounts to a twopiece cascade impactor, the lower part comprising the base that retains the collection cups, the upper part consisting of the lid and seal body that contains the stage nozzle assembles and interstage passageways. A study was made to determine whether automation was easier with the stacked impactor or the all impaction plates on one plane (planar) impactor. The steps that would have to be taken to remove the impaction plates and wash the deposits from the impaction plates for each of the layouts were enumerated. From this analysis, it was determined that the least number of steps would be taken for one sampling cycle with the external impaction plate layout of the two-piece impactor. Mock-ups of the competing stack and planar designs were presented to the full consortium along with the analysis of the speed and ease of use. The consortium voted heavily in favor of the planar, two-piece impactor. This layout was then further refined and became the basic layout for the NGI. DETAILED AERODYNAMIC DESIGN With the number of stages and the design cut points decided upon in the early project discussions, the next step was to define the exact sizes and numbers of nozzles on each impactor stage. The selection of the nozzle diameters and the number of nozzles followed the guidelines developed for impactor design initially described by Marple, 19 Marple and Willeke, 20 and later refined by both Rader and Marple, 21 and by Fang et al. 22 These guidelines are as follows: 1. The Reynolds number (see equation 1) of the flow through the nozzles should be in the range of 500 to FIG. 3. Schematic diagram of a typical nozzle/impaction plate stage.

7 NEXT GENERATION PHARMACEUTICAL IMPACTOR DESIGN The nozzle-to-plate distance should be at least one nozzle diameter and no larger than 10 for round-nozzle impactors. 3. The cross-flow parameter (see equation 2) should be less than 1.2. If an impactor stage is constructed according to these guidelines, experience has shown that it will have a steep collection efficiency curve and that its cut point, defined as the particle size that is collected with 50% efficiency (d 50 ), will be very close to the calculated value. Figure 3 is a schematic diagram of a typical nozzle plate and impaction plate with the relevant dimensions labeled. Conventional theory for single-nozzle impactors has shown that there are two dimensionless parameters that are important in defining the flow field through such a nozzle, 19 and it is the flow field that is important in defining the sharpness of cut of the stage. 20 These parameters are the ratio of the nozzle-to-plate distance to the nozzle diameter, S/W, and the Reynolds number, Re, defined as: rwv 4rQ Re 5 } o 5 } (1) m npmw For a specific flow rate, n and W can be selected to keep Re in the desired range. The above statements hold true for singlenozzle impactors, and in most cases, for multiple nozzle impactors. However, in multiple-nozzle impactor stages, there can be a cross-flow problem caused by spent air from the nozzles near the center of the nozzle plate flowing outward past the air jets located near the edge of the nozzle cluster. In some cases this cross-flow can prevent the air jets near the edge of the cluster from reaching the impaction plate. This phenomenon was studied by Fang et al., 22 who found that a cross-flow parameter, X c, could be defined as: nw X c 5 } (2) 4Dc If the value of this parameter is less than 1.2, the jets from the outer nozzles are not affected. Thus, the three parameters Re, S/W, and X c are important in defining the correct aerodynamics in an impactor stage. Another dimensionless parameter important in determining the cut size of the impactor stage is the Stokes number, St: 2 2 4r p QC p d p 4r ae QC ae d ae St 5 }} 5 }} (3) 9npmW 3 9npmW 3 FIG. 4. NGI seal body (top view).

8 290 To calculate C p or C ae, one uses the following expression for C: C Kn exp (4) with the Knudsen number, Kn, calculated from equation 5, using d p or d ae as appropriate, depending on whether C p or C ae is to be calculated from equation 4: 2l 2l Kn 5 or Kn 5 (5) } dp When the value of d ae in equation 3 equals d 50, the Stokes number is denoted as St 50, and the governing (implicit) equation for the cut size of an impactor stage becomes: ÏC 50 d 50 5! 9pmnW }} 3 ÏSt 50 (6) 4rae Q } Kn } dae The value of C 50 is computed from equation 4 using the diameter d 50. Consequently, for each selected impactor stage design and flow rate (m, n, W, r ae, Q), equations 4 and 6 must be solved simultaneously, either numerically or iteratively. However, for particles greater than 1 mm the Cunningham slip correction is equal to 1.0 for practical purposes, permitting a direct solution of d 50 from equation 6. Ideally, the value of ÏSt 50 should be near to if Re is , and S/W is at least 1.0 but not greater than about 10. However, if factors other than inertia are significant in the collection mechanism, such as gravitational effects for large MARPLE ET AL. or very dense particles, or significant deviation in C from unity associated with particles having d p, 0.1 mm, then ÏSt 50 will have a value other than Thus, when designing a stage with a specific value of d 50, n and W are selected so that equation 6 is satisfied, 20 while keeping Re in the range of By using this method, the values of n and W in Figure 4 were determined so that the theoretical cut size values at 60 L/min of stages 1 to 7 of the NGI ranged between 0.33 mm and 7.8 mm aerodynamic diameter, with equal spacing between the values on a logarithmic scale. Figure 5 depicts the resulting expected performance based on the above theoretical considerations for the NGI within the entire flow range for which the impactor was designed. Note that the requirement that there should be five stages with cuts between 0.5 and 5 mm is not strictly met. There are, however, five cut sizes between 0.5 and 6.5 mm at all flow rates. This result was judged to be acceptable, since five stages strictly in the 0.5 to 5 mm size range at all flow rates between 30 and 100 L/min would have required eight stages, leading to overlapping of the stage efficiency curves. The critical design parameters for the NGI stages including the MOC, including their associated dimensionless constants Re, S/W, and X c are listed in Table 3 for volumetric flow rates of 30, 60 and 100 L/min. The Re values for the seven impactor stages are in the generally desirable range of , with several stages below the range at low flow rates and small cut sizes, FIG. 5. Theoretical NGI stage cut size versus flow rate.

9 NEXT GENERATION PHARMACEUTICAL IMPACTOR DESIGN 291 TABLE 3. VALUES OF RE, S/W, AND X C FOR THE NGI STAGES AT 30, 60, AND 100 L/MIN Reynolds number at each flow rate (L/min) Stage W, mm n S, mm S/W D c, mm X c na na na na MOC 70 mm and above the range at large cut sizes at high flow rates. This outcome is a consequence of the design choice that the NGI should not have interchangeable nozzles while operating over a large flow rate range. The high Re for stage one is unavoidable with a single nozzle design, which in turn was forced by the finding that only a single nozzle design could accept the airflow directly from the USP/EP induction port without unacceptable upstream losses. The values of S/W are all in the desired range of greater than 1.0 and less than 10. Also, the X c values are all less than 1.2. Values of D 50 calculated from the archival calibration in the companion publication 18 are summarized in Table 4 for inlet flow rates of 30, 60, and 100 L/min. The aerodynamic design of the flow passages was expected to result in very small interstage losses. Investigations of losses for a range of typical pharmaceutical aerosols, including 14 MDI formulations and 16 DPI formulations, were conducted on prototype NGIs. The sum of the deposition on all surfaces other than collection cups TABLE 4. STAGE CUT SIZES FOR THE NGI AT 30-, 60-, AND 100-L/MIN IMPACTOR INLET FLOW RATE CALCULATED FROM THE ARCHIVAL NGI CALIBRATION 18 D 50 (microns) at each flow rate (L/min) Stage was measured to be in the range 1 5% of the total delivered sample mass, depending on the nature of the formulation. 23,24 For dry-powder aerosols, of course, the stages in the NGI, as for all cascade impactors, must be coated with an adherent material unless specific tests show that the coating is not needed. Consequently, particles should need to be removed only from collection surfaces following a particle size determination, and the non-collection surfaces in the impactor should need to be washed only after multiple determinations. All non-collection surfaces are either on the lid or the seal body; these two parts disassemble easily and optionally can be placed in a dishwasher for cleaning. DETAILED LAYOUT DESIGN After the basic layout of a planar impactor was decided upon, there were several iterations made to establish the appearance of its components, how the impactor should open and close, whether the flow direction should be from left to right or right to left, and to determine the user-friendly features that could be incorporated. Much of the input specifying these features came from the user comments during the testing of the first prototypes. The final latching mechanism of the impactor was one of the more challenging design problems, involving careful setting of tolerances for the various components to achieve both a good seal (leak rate smaller than 100 Pa/sec) and the proper jet-to-plate distances. Figure 6 is a cross-section through a representative stage of the final NGI design. The nozzle assemblies are all held in one plate called the

10 292 MARPLE ET AL. FIG. 6. Cross section of one NGI stage. seal body. The impaction plates are tear-shaped cups located below the stage nozzles. The large end of the impaction cup is located directly below the nozzles and is where particle collection occurs. Air flows from the impaction region of the cup to its small end, where the flow is withdrawn upward into a cavity in the lid of the NGI. The cavity in the lid directs the air to the next stage. FIG. 7. The optional false lid for stage 1.

11 NEXT GENERATION PHARMACEUTICAL IMPACTOR DESIGN 293 The nozzles of the stages are staggered back-andforth in the seal body so that the tear shaped cups can be nested in a compact configuration. Although Figure 6 shows three pieces (lid, seal body, and bottom frame) instead of two, as outlined in the original concept, the lid and the seal body are secured as one assembly when the NGI is used for routine sampling. To change the impaction cups between sampling runs, only the cups and tray must be separated from the seal body/lid assembly. Two options were considered for the impaction cups. One option was to have all of the cups machined into a one-piece manifold. The other option was to have individual cups held in a rack that can be inserted into a frame below the seal body. In practice, both options can be used interchangeably, since the internal airflow path configuration would be the same. For manual operation the use of individual cups makes the washing of the deposits from the cups more manageable and is the option shown in this paper. However, for automation, the manifold option may be more satisfactory. The roughness of the surface of the collection cups can affect wetting of drug recovery solvents FIG. 8. NGI components. (A) Lid with internal air passages (bottom view). (B) Seal body with nozzles (bottom view). (C) cup tray with cups (top view). (D) Bottom frame with locking handle (top view).

12 294 but has no effect on the particle capture efficiency so long as the surface roughness measure is much smaller than the particle stop distance. 25 The stop distance is approximately 25% of the nozzle diameter for round jets. 19 Applying this criterion to the smallest nozzles (0.206 mm on stage 7), we conclude that approximately 2.5 microns is the upper limit of acceptable surface roughness (five percent of the stop distance on stage 7). Cups offered commercially by MSP Corporation have a surface roughness between 0.5 micron and 2 micron. Because native steel typically has a surface roughness of 0.3 micron to 0.5 micron, the choice of 0.5 microns to 2 microns for the surface roughness represents a slight roughening of the surface to improve wetting without affecting the aerodynamics. The cup for stage 1 is larger than for the other stages to minimize impaction of larger particles near the cup s vertical wall, identified in early prototype testing. This issue is not important for the other stages since these stages collect smaller particles that are less prone to this so-called secondary impaction. So, for stages 2 7, the dimension of the nozzle holder determines the cup size. For stage 1 only, the option is provided for a lid to retain particles within the cup of stage 1, if significant secondary impaction is encountered with a specific formulation. This lid rests on the edge of the stage-one cup (Fig. 7). Deposits on the inside surface of this lid can be rinsed easily into the stage-one cup. Without this false lid, these deposits would be on the underside of the seal body of the impactor and would therefore be cumbersome to recover. In prototype testing, as much as five percent of the mass of material on stage one was found on the false lid for a formulation that had a mass median diameter approximately equal to the d 50 value for stage one. In general, however, this false lid is not needed. FIG. 10. MARPLE ET AL. NGI in the open position. The cup size for the MOC is also larger than for the cups for stages 2 to 7, and is the same size as the cup for stage 1. The larger diameter was necessary to accommodate the 4032 nozzles in the MOC, and conveniently permits a standard 75- mm filter to be used in an internal filter configuration. Figure 8 shows the individual parts of the NGI. The lid contains the air passages between stages, the seal body contains the stage nozzle plates, and the bottom frame holds the collection cups in a cup tray. The bottom frame also contains hinge pins and a past-center clamping handle. The lid contains the receivers for the hinge pins and the clamping handle. Figures 9 and 10 shows the NGI in the closed and open positions, respectively. In Figure 9, the USP/EP induction port is shown on the inlet of the first stage. The NGI pivots open on two hinges located at the back of the NGI between the bottom frame and the lid. The seal body is held to the lid by two limited torque screws (the black knurled knobs on the upper surface of the lid). When closed, the lid is locked to the bottom frame by a past-center cam lock handle mechanism located at the front of the NGI. The cup tray and the seal body are clamped between the lid and bottom frame when the handle is pushed downward into a locked position. Tolerances on all parts from the hinges to the clamping mechanism are such that no adjustment screws or spring components are required to ensure a leak-tight seal when the NGI is clamped. FIG. 9. inlet. Closed NGI with USP/EP induction port on the Pre-separator design Design criterion M-5 states that the NGI must have a high-capacity pre-separator stage to prevent oversized particles from depositing at unwanted places. Because the need for a pre-sepa-

13 NEXT GENERATION PHARMACEUTICAL IMPACTOR DESIGN 295 FIG. 11. NGI with pre-separator. rator is formulation dependent, it was designed as an add-on unit, typically used for DPIs, that can be positioned on the NGI between the USP/EP induction port and the first stage, as shown in Figure 11. This pre-separator incorporates two collection surfaces working in tandem, as shown in Figure 12. The first (scalper) collection surface is a circular cup, containing solvent, beneath the inlet of the pre-separator and is intended to remove very large particles (e.g., lactose carrier particles). Typically, this cup is filled half-way with solvent before testing to reduce particle re-entrainment from the cup. The liquid reservoir gives the unit a high solids capacity along with a fairly low internal volume, but by itself does not provide a satisfactorily sharp cut. The volume of this first cup is approximately 30 ml, so most users load the cup with 15 ml of solvent. The pre-separator performance should be insensitive to the exact volume of solvent in the cup. The scalper is immediately followed by the second collection surface comprising a more conventional impaction stage. This stage has a much sharper cut than the scalper to eliminate most particles that are larger than the cut size of the first stage of the impactor itself, while not removing particles that are of a size to be captured on the second impactor stage. Sealing between the USP/EP induction port and the pre-separator, and between the pre-separator and the first stage is accomplished by 10 tapers at the interfaces, to reduce the number of O rings in the inlet of the NGI. If the pre-separator is not used, the tapered exit of the USP/EP induction port seals to the tapered inlet of the first stage. Whereas the outside dimensions of the induction port are customized for these tapers, the internal dimensions of the induction port are those described in the USP and EP. Micro-orifice collector (MOC) One unique feature of the NGI is that an impactor stage of 4032 nozzles, each nominally 70 mm in diameter, is used to replace the final filter used in other cascade impactors. There are two reasons for this feature. First, the collection cup beneath the micro-orifice nozzles can be handled, and particles analyzed, in the same manner as from any of the other collection cups. This aspect is especially important for automation. Second, particles are much more easily dissolved from a solid collecting surface than from the fiber network in a filter. The large number of nozzles in the MOC is necessary to keep the pressure drop within a range that can be managed by pumps normally found within aerosol laboratories. To FIG. 12. Exploded view of NGI pre-separator.

14 296 MARPLE ET AL. FIG. 13. External filter holder attached to exit of the NGI. keep the crossflow parameter below the critical limit, the diameter of the nozzle cluster was set to 75 mm (Table 3). The size of the collection cup for this stage is the same as for stage 1. Four dimples on the downstream face of the nozzle plate support the plate against the cup and maintain a minimum nozzle-to-plate distance. Some samples may contain particles that are so fine that they are not collected by the MOC. However, since the particle size distributions generated by most inhalers are larger than the cut size of the MOC, its efficiency is satisfactory for the majority of drug product formulations. Nevertheless, its effectiveness must be evaluated for any new formulation, or inhaler device, during method development by placing a filter downstream of the MOC and determining the magnitude of the portion of the dose (if any), that penetrates the MOC. Figure 13 shows an external filter designed especially for this purpose. This filter holder has a cassette into which the filter is placed. In this manner the filters are easy to handle and the cassette with filter can be transported with the cup tray for analysis. An internal filter holder has also been designed for the NGI, in the event that it is routinely found that a significant proportion of particles penetrates the MOC. The internal filter holder is contained in a special cup (Fig. 14) that replaces the cup normally used with the MOC. It is not necessary to remove the MOC nozzle plate, since any particles passing through the MOC nozzles impact directly on the filter upper surface. Although the MOC may appear to be an impaction stage, this is specifically not the intent for its incorporation into the design, and judgments about the size of material collected on the stage are not recommended. The MOC permits significant simplification of analytical procedures for the many samples for which it is shown, by the procedures above, to be applicable, but it will not A. B. C. A. B. C. FIG. 14. Internal filter holder (for insertion in place of the MOC cup).

15 NEXT GENERATION PHARMACEUTICAL IMPACTOR DESIGN 297 be useful for all formulations, especially those containing a significant portion of very fine material. Special features of the NGI Several special features have been implemented that make the NGI easy to use: 1. The O rings are held in the seal body by dovetail O ring grooves. That is, the grooves are slightly narrower at the surface. Once the O ring is inserted in the groove, the taper will retain it in the groove until removed by the operator. 2. The seal body can quickly be separated from the lid by two thumbscrews. This feature allows convenient inspection of the air passages in the lid or washing of the stage nozzles, which is performed periodically. 3. The lid can be removed from the bottom frame hinge pins by opening the lid all the way and sliding the lid off of the pins. The lid can also be washed periodically. The bottom frame never has to be washed routinely as no components on it are exposed to any formulation. 4. The cups can all be changed in one action by simply removing the cup tray. This feature makes changing collection cups fast and keeps the cups in the correct location until analysis. 5. The sides of the cups are tapered so that they can be easily stacked during storage. 6. The lid support at the rear of the NGI also serves as a stand for storage by rotating the front of the NGI upward. In this position the NGI has a smaller footprint than when the impactor is positioned for normal use with the collection cups oriented horizontally. The design objectives related to stage performance were satisfied. The stages have been shown to have sharp cut off characteristics, with cut sizes at the desired places, and very low inter-stage losses so that the impactor body and nozzles do not have to be cleaned between every run. The collection efficiency curve for each stage has been determined at three flow rates and the results are reported in the companion paper. 18 A collaborative inter-company study has been conducted by members of the European Pharmaceutical Aerosol Group (EPAG) to establish interlaboratory variability, 26 and results of this study are expected to be submitted for publication in CONCLUSION The NGI is the result of a collaborative process, with the sponsoring pharmaceutical companies involved in all major decisions throughout development. Input was gathered from as broad a constituency as could be solicited. Design decisions were made in consideration of mechanical, physical, regulatory, and end-use viewpoints. The result is a seven-stage impactor with an optional pre-separator at the inlet for over-size particles, and a micro-orifice collector at the exit instead of a final filter to collect the finest particles. A unique external particle collection cup design was used so that all of the collection cups are in one plane and can be removed from the impactor as a single unit. This feature eases the ability to automate the NGI, which was a primary design requirement. The NGI is also unique in that it is not designed with stages having specific cut sizes at a specific flow rate. Instead, the NGI operates at any flow rate from 30 to 100 L/min with the cut sizes spanning a particle size range nominally from mm to 11-mm aerodynamic diameter. The NGI is designed so that the spacing of the stages varies between four or five cut points per decade of particle size, depending on the flow rate chosen for the measurement. Efforts are underway to incorporate the NGI into the European Pharmacopoeia and the U.S. Pharmacopeia. PERFORMANCE APPENDIX Aerodynamic diameter is a parameter defined as the diameter of a hypothetical spherical particle of unit density (i.e., r ae g/cm 3 ) that settles in air at the same falling velocity as the physical particle. In the most general case for non-spherical particles: r p C p d ae 5 d p! } xrae C ae (A-1) If the particles are spherical, x, the dynamic shape factor is unity and equation A-2 applies:

16 298 MARPLE ET AL. d ae 5 d p! The aerodynamic diameter, rather than physical diameter, is a useful concept for analysis of systems of aerosols because many fundamental equations describing particle motion can be expressed with aerodynamic diameter as a parameter. Aerodynamic diameter is a parameter that can be used, along with geometric description and air flow profile, to correlate with (A1) and predict (A2) regional deposition in the lung. Appendix references A1. Rudolf, G., R. Kobrich, and W. Stahlhofen Modeling and algebraic formulation of regional aerosol deposition in man. J. Aerosol Sci. 21:s403 s406. A2:. Martonen, T.B Mathematical model for the selective deposition of inhaled pharmaceuticals. J. Pharm. Sci. 82: C C 50 C ae C p D 50 d p d ae d 50 D c Kn n Q Re S St St 50 V o W X c l m ABBREVIATIONS r air density r p C p } (A-2) r p particle density rae C ae r ae unit density (i.e., 1 g/cm 3 ) x Cunningham slip correction factor Cunningham slip correction factor for a particle of size d 50 Cunningham slip correction factor for a particle of size d ae Cunningham slip correction factor for a particle of size d p calculated stage cut size from cut size-flow rate equation diameter of a spherical particle aerodynamic diameter of a particle aerodynamic diameter of particle collected with 50% efficiency diameter of the cluster of nozzles on a stage Knudsen number number of nozzles in a stage total volumetric flow rate through an impactor stage Reynolds number jet-to-plate distance Stokes number Stokes number at 50% collection efficiency average velocity of air in a nozzle nozzle diameter cross-flow parameter mean-free path of air air viscosity dynamic shape factor (x for spherical particles) ACKNOWLEDGMENTS Considerable technical input was obtained from representatives from all the sponsoring companies; their input is gratefully acknowledged. The representatives and their affiliated companies are Dave Warren (3M Healthcare Ltd.), Beatrix Fyrnys (Asta Medica now Sofotec GmbH), Bo Olsson and Lars Asking (AstraZeneca R&D Lund), Steve Nichols (Aventis Pharma), Dieter Hochrainer (Boehringer Ingelheim Pharma KG), Professor David Ganderton (Coordinated Drug Development Ltd), Bernie Greenspan (Dura Pharmaceuticals), Keith Truman and Mike Van Oort (GlaxoSmithKline), Mike Holroyd (Norton Healthcare now IVAX Pharmaceuticals), Dilraj Singh (Novartis Pharma AG), Terhi Mattila and Jari Kovalainen (Orion Pharma), Paul Miller (Pfizer), Tön Forch and Hans Keegstra (Pharmachemie), Bruce Wyka (Schering-Plough Research Institute), Jolyon Mitchell (Trudell Medical International), and Jeremy Clarke (Vectura Ltd.). REFERENCES 1. Thiel, C.G From Suzie s question to CFC free: an inventor s perspective on forty years of MDI development and regulation. In R.N. Dalby, P.R. Byron, and S.J. Farr, eds. Respiratory Drug Delivery V. Interpharm Press, Buffalo Grove, IL, Dimmick, R.L., M.T. Hatch, and J. Ng A particle-sizing method for aerosols and fine powders. A.M.A. Arch. Indust. Health. 18: Thiel, C.G Private communication, 3M Pharmaceuticals, St. Paul, MN. 4. Polli, G.P., W.M. Grim, F.A. Bacher, et al Influence of formulation on aerosol particle size. J. Pharm. Sci. 58: Mitchell, R. I., and J.M. Pilcher Improved cascade impactor for measuring aerosol particle sizes in air pollutants, commercial aerosols and cigarette smoke. Indust. Eng. Chem. 51: Andersen, A. A New sampler for the collection, sizing, and enumeration of viable airborne particles. J. Bacteriol. 76: May, K. R Multistage liquid impinger. Am. Soc. Microbiol. 30:

17 NEXT GENERATION PHARMACEUTICAL IMPACTOR DESIGN Bell, J. H., K. Brown, and J. Glasby Variation in delivery of isoprenaline from various pressurized inhalers. J. Pharm. Pharmacol. 25(suppl.): Asking, L, and B. Olsson Calibration at different flow rates of a multistage liquid impinger. Aerosol Sci. Technol. 27: Hallworth, G. W., and D. G. Westmoreland The twin impinger; a simple device for assessing the delivery of drugs form metered dose pressurized aerosol inhalers. J. Pharm. Pharmacol. 39: Hallworth, G.W., D. Clough, T. Newnham, et al A simple impinger device for rapid quality control of the particle size of inhalation aerosols delivered by pressurized aerosols and powder inhalers. J. Pharm. Pharmacol. 30(Suppl): British Pharmacopoeia. (1988). C. Pressurized inhalations: deposition of the emitted dose. In British Pharmacopoeia, Vol. II. Department of Health, London, A204 A Aiache, J. M., H. Bull, D. Ganderton, et al Collaborative study on the measurement of fine particle dose using inertial impactors. Pharmeuropa 5: Miller, N.C., V. A. Marple, R. K. Schultz, et al Assessment of the twin impinger for measurement of MDI sprays. Pharm. Res. 9: U.S. Pharmacopeia ,601. Aerosols, metered dose inhalers, and dry powder inhalers. In U.S. Pharmacopeia 25/National Formulary 20, U.S. Pharmacopeial Convention, Inc, Rockville, MD, European Pharmacopoeia Preparations for inhalation: aerodynamic assessment of fine particles. In European Pharmacopoeia, 3rd ed., Suppl. 2001, Council of Europe, Strasbourg, France, Marple, V.A., B.A. Olson, and N.C. Miller A low-loss cascade impactor with stage collection cups: calibration and pharmaceutical inhaler applications. Aerosol Sci. Tech. 22: Marple, V.A., B.A. Olson, K. Santhanakrishnan, et al Next generation pharmaceutical impactor (a new impactor for pharmaceutical inhaler testing). Part II: archival calibration. J. Aerosol Med. 16: Marple, V.A A fundamental study of inertial impactors [Ph.D. dissertation]. University of Minnesota, Minneapolis. 20. Marple, V. A., and K. Willeke Impactor design. Atmos. Environ. 10: Rader, D. J., and V. A. Marple Effect of Ultra- Stokesian drag and particle interception on impactor characteristics. Aerosol Sci. Technol. 4: Fang, C. P., V. A. Marple, and K. L. Rubow Influence of cross-flow on particle collection characteristics of multi-nozzle impactors. J. Aerosol Sci. 22: Mitchell, J.P The next generation impactor (NGI): results from the evaluation of prototype instruments with pressurized metered dose inhaler (pmdi) based formulations. In Drug Delivery to the Lungs XI. Aerosol Society (UK), London, Shrubb, I The next generation impactor (NGI): results from the evaluation of prototype instruments with dry powder inhaler (DPI) based formulations. In Drug Delivery to the Lungs XI. Aerosol Society (UK), London, Marple, V. A., Private communication, University of Minnesota. 26. Nichols, S Private communication, Aventis Pharma UK. Received on January 31, 2003 in final form, May 6, 2003 Reviewed by: Peter R. Byron, Ph.D. John E. Agnew, Ph.D. Address reprint requests to: Jolyon Mitchell, Ph.D. Trudell Medical International 725 Third Street London, Ontario, Canada N5V 5G4 jmitchell@trudellmed.com