WO2002059574A1

WO2002059574A1 - Method and apparatus for studying aerosol sources

Info

Publication number: WO2002059574A1
Application number: PCT/FI2002/000055
Authority: WO
Inventors: Esko Kauppinen; Juha Kurkela; David Brown; Jorma Jokiniemi; Terhi Mattila
Original assignee: Valtion Teknillinen Tutkimuskeskus
Priority date: 2001-01-24
Filing date: 2002-01-23
Publication date: 2002-08-01
Also published as: FI20010144A0

Abstract

The present invention relates to a method and an apparatus for studying the behaviour of unsteady aerosol sources (UAS) including aerosol delivery systems (ADS), especially that of a dry powder inhaler (DPI). The said method comprising a controllable system producing time-dependent pressure or velocity profiles at the inlet of the sampling apparatus, a conduit wherein the sampled flow is conditioned and transported to measuring/analysis instruments to characterise particles.

Description

METHOD AND APPARATUS FOR STUDYING AEROSOL SOURCES

Field of the invention

The present invention relates to a method and an apparatus for studying the behaviour of unsteady aerosol sources (UAS) including aerosol delivery systems (ADS), especially that of a dry powder inhaler (DPI).

Background of the invention

Unsteady aerosol sources (UAS) are common in industry. Often it is required in the design and testing of equipment that the aerosol particle size distribution, size dependent composition or morphology is accurately measured. Examples of such aerosol sources include engine emissions in unsteady combustion operations such as occur in smelters, power plants, recovery boilers, powder production, reciprocating engines and turbine engines, soot blowing operations, and aerosol delivery systems (ADS) such as, aerosol spray cans, dry powder inhalers (DPI) and metered dose inhalers (MDI). The described method uses the examples of a DPI and a wider aerosol source (WAS), i.e. the flue gas from a combustion process, a metallurgical process, a powder synthesis process, a reciprocating or jet engine, a spray coating or spray drying process, as the aerosol source, but the method is equally applicable to a wide range of additional unsteady and steady aerosol source measurements.

Aerosol delivery systems are used in a number of technologies related to, e.g. drug delivery and applications of coatings and sprays. Measuring the performance of these systems requires the accurate determination of the particle size distribution and mass loading of the aerosol generated in the device under unsteady flow conditions. Of particular importance is the use of ADSs for the delivery of therapeutic agents to the lung where such devices are commonly termed inhalers. Inhalation has become an important route of administration in the treatment of various diseases, e.g. asthma, diabetes and cancer. This is because, in addition to providing direct access to the lungs, medication delivered through the respiratory tract provides rapid and predictable onset of action and requires lower dosages compared to the oral route.

Pressurised metered dose inhalers (pMDI) are currently the most commonly used inhalation devices. Such devices comprise a canister containing a suspension of fine drug particles in a propellant gas. Upon actuation, the aerosol contents are expelled, through a metering valve, and a metered dose is propelled into the respiratory tract of the patient. However, many problems with pMDIs are causing them to be replaced, in large part, by dry powder inhalers.

Several types of dry powder inhalers have been developed, in which the inhalation air of the patient is used for dispersing the drug particles. DPIs are user-friendly, as they do not require co-ordination between actuation and inspiration, as do MDIs. The powdered medicament is arranged as unit dose containers, e.g. blister packs, cartridges or peelable strips, which are opened in an opening station of the device. Alternatively, the unit dose is measured from a powder reservoir by means of a metering member, e.g. a dosing cup.

To increase flowability and dosing accuracy of the powdered medicament, the fine drug particles of respirable size are typically mixed with coarser carrier particles to form an ordered mixture, wherein fine drug particles are attached to the larger carrier particles. This technique complicates the powder aerosolisation process and, in particular, necessitates the break-up of the drug/carrier agglomerates before they enter the patient's mouth and throat, where individual large particles and agglomerated large and small particles tend to deposit. Effective aerosolisation and deagglomeration of the powder requires that forces exerted on particles (be they on exposed surfaces of the device, between drug and carrier particles or between drug and drug particles) must be overcome under all expected inhalation profiles.

The aim of inhalers is to produce a controlled fine particle dose (FPD), i.e. the mass of drug per dose in the particles smaller than 5 μm in aerodynamic size. However, the ability of an inhaler to aerosolise and deagglomerate the drug particles into a respirable particle size range depends on the patient's inspiration technique for most DPIs currently available. An ideal DPI would provide uniform powder aerosolisation and deagglomeration over a wide range of inhalation profiles, so as to generate consistent doses of respirable particles in the final dispersion.

The FPD is a crucial property for determining the performance of aerosol drug delivery systems. In the inhalation cycle, particles smaller than 5 μm have the highest probability of being delivered to the deep lung region where the treatment for the disease, e.g. asthma, is required. However, other diseases may require deposition elsewhere in the respiratory tract, for instance in the bronchi or throat. Therefore, it is essential to measure the emitted particle size mass distribution, and thus also FPD, properly in order to predict where (or if) particle deposition will occur. Once the powder has exited the inhaler, the aerosolised powder must be sampled in such a way that no changes take place in the measured particle size distribution. In other words, the sampled aerosol must be transported from the outlet of the inhaler into the particle size classifier without further breaking up particle agglomerates and by avoiding the deposition of the fine particles, which are of primary interest. Due to the complicated characteristics of the inhalation process, e.g. rapid changes in flow and loose particle agglomerates, special attention needs to be paid to prevent such phenomena which distort particle size distribution from occurring after the particles exit the aerosol source and before they enter the particle characterisation device during the sampling.

In the prior art there are known sampling methods applied to aerosols delivered from inhalers. Inhaler evaluation methods currently used by industry and academia are based on the philosophy that a sampling system should simulate the inhalation process (breathing cycle) and biological geometry (mouth and trachea) of a patient. A schematic of the measurement system approved by the European and US Pharmacopoeias is depicted in Figure 1 (European Pharmacopoeia 2001, US Pharmacopeia 1999). Such a system is deficient in a number of key aspects. First, the 90 degree stainless steel bend has a very different flow characteristics than a human patient and particle deposition patterns and thus losses to the surfaces of the bend will tend to be significantly different than in the real case (Li et al. 1996). If particles impact with sufficient velocity, they can also break-up, thus releasing into the gas stream a number of smaller particles that will likely be reentrained in the flow. In the mouth and throat of a patient, particles which impact the surface will likely remain attached and, therefore will not reentrain. In addition, the metal surface can actively charge particles and change their deposition characteristics in the downstream measuring/analysis instruments. Thus the simulated geometry can actively change the particle size distribution being measured. Difficulties also arise due to the unsteadyness of the flow. The simulation of the breathing cycle with unsteady flow through the particle sizing instrument leads also to significant errors in the measured results due to the fact that the flow remains undeveloped through most of the measurement period. Many measuring/analysis instruments require a steady-state flow to function properly. The relaxation time to reach a steady state can be many times the characteristic time of the unsteady aerosol process under investigation. In addition, turbulence, which tends to alter the particle size distribution since particles can deagglomerate due to high local instantaneous shear forces, alters the distribution in a manner very dissimilar than in a human patient. The aforementioned methods have no means to reduce turbulence downstream of the aerosol source. Finally, the simulated flow pattern in the aforementioned methods does not equal (or even approximate) a human inhalation pattern and thus does not represent realistic unsteady flow characteristics within the aerosol source.

The inaccuracy in such methods is bora out by the continuous disagreement between in vivo and in vitro studies of the inhaled dose delivered from inhalers. The values for lung deposition from in vivo studies are typically only half of those predicted from in vitro studies. Details and drawbacks of these methods are given below.

The sampling method according to the European and US Pharmacopoeias comprises a right-angle bend metal tube with a 19 mm minimum inner diameter (subsequently called 'throat piece'), which connects the inhaler to the particle size classifier, e.g. an Andersen cascade impactor or multi-stage liquid impactor (European Pharmacopoeia 2001, US Pharmacopeia 1999). Downstream of the impactor there is a flow control valve, a two-way solenoid valve controlled with a timer, and a vacuum pump.

The system is operated such that, with the pump running, the solenoid valve between the impactor and the vacuum pump is opened for the required time to create the flow through the inhaler. Thus the flow through the impactor is not stabilised in the beginning of the sampling process, although that is necessary to ensure the correct performance of the impactor. The unstabilised flow within the impactor causes errors in the interpreted cut-off sizes of impactor stages during the sampling, i.e. sampling with an impactor has to be carried out under stabilised and known flow conditions. Moreover, most of the powder from DPI is typically released at very early stage of inhalation, i.e. when the flow is not yet stabilised (Burnell et al. 1998). Therefore, the sampling method tends to significantly overestimate FPD. Furthermore, the inhalation profile is not equivalent to that of a human, as the profile is induced by simply opening the two-way valve. Since the beginning of the inhalation is the most critical phase in respect of drug delivery from DPIs, the inaccuracies in the sampling method are amplified. In addition, the right-angle bend tube is used in the system to mimic a typical human's throat and pharynx. Research has demonstrated that it is incorrect to simulate the particle behaviour in the human oral-pharyngeal cavities with a simple right-angle bend (Li et al. 1996). In addition to the different flow characteristics in the bend, the use of a throat piece strongly distorts the measured particle size distribution since large particles carrying smaller drug particles attached to their surfaces hit the back wall of the throat piece at the bend with such a high velocity that additional particle break-up occurs.

Another important deficiency in the aforementioned method relates to turbulence generated in the throat piece. The most important parameter describing the onset of turbulence in a particular flow field is the Reynolds number. High Reynolds number indicates that instabilities in a fluid will tend to grow. Low values indicate that these instabilities will diminish. Reynolds number (Re) is defined as pUL

Re =

where velocity U, density p, viscosity μ, and characteristic length E, such as tube diameter, are relevant parameters. The turbulent characteristics of the flow in the throat piece easily leads to undesired particle break-up and fine particle deposition downstream of the DPI: there is likely a turbulent flow (Re > 2000) for the flows greater than 27 1/min in the throat piece. Due to the severe deflection of the flow, turbulent conditions are likely to occur at much lower flow rates. Thus, even if the impactor was operating at the steady state, the measured size distribution would not be representative of that exiting the DPI nor that which would occur in a human subject.

Burnell et al. introduced the Electronic Lung Device wherein, at the start of the inhalation, air is drawn through the DPI via a right-angled glass or metal tube into an 11 -litre sampling chamber. This inhalation process is carried out by a motor-driven piston/cylinder arrangement and controlled by a feedback system to produce a predefined inhalation profile. The inhalation phase is followed by the particle sizing process, where the valves of the system are controlled so that the inhaler is isolated from the sampling chamber and that the aerosol cloud is sucked into a particle sizing instrument via a pump equipped with a solenoid valve downstream of the sizing instrument. After the sampling, the throat piece, the sampling chamber and the stages of the particle sizing instrument are washed in aqueous methanol for high- performance liquid chromatography (HPLC) analysis to determine the quantities of deposited drug in each part of the system. The system also includes a real-time measurement of dose emission. This is carried out with a laser light source and a right-angle detector which detects the scattered laser light signal. The signal is then converted to dose emission intensity.

As in the previous method the flow through the impactor does not reach a steady state at the beginning of the particle sizing process, as the air is drawn from the sampling chamber into the particle sizing instrument by opening the solenoid valve between the vacuum pump and particle sizing instrument. Therefore the measured particle size distribution is likely to be distorted. Another reason for the distortion of FPD is the use of a throat piece, where particle agglomerates break up into smaller particles as with the Pharmacopoeia method described above. In addition, high turbulence in the throat piece leads to undesired fine particle deposition and induces strong shear forces in the flow, which again may cause further particle break-up during the sampling. Finally, motor driven piston/cylinder systems are expensive to construct and operate and they limit the applicability of the method to unsteady phenomena with finite total volume.

Miller et al. introduced a measurement method for an aerosol drug delivery system wherein the sampled air is drawn into a cascade impactor by a vacuum source through a US Pharmacopoeia (USP) inlet and a mixing inlet. The USP inlet is a right-angled bend metal tube identical to the European Pharmacopoeia inlet. In the mixing inlet the sampled aerosol is mixed with variable supply air prior to the impactor, which enables a constant flow rate through the impactor during the sampling. The flows in the system are controlled via a computer by a breath profile simulator (BPS). This consists of a piston for inducing higher flow rate through the inhaler than that constant flow rate which is needed to run the impactor at optimal conditions.

As in the previously described methods, the sampled particles, which are mostly agglomerates, are exposed to the severe deagglomeration conditions due to the collision of the particles with the throat piece and due to the turbulence in the USP inlet. High turbulence conditions also contribute to high particle losses in the device. These phenomena lead to undesired distortions in the measured particle size distributions. As in the previously described method, the motor driven piston/cylinder system is expensive to construct and operate and limits the applicability of the method to unsteady phenomena with finite total volume. To conclude, the sampling methods of the prior art are encumbered by the following drawbacks: 1) flow through the impactor is not stabilised during the entire sampling process, although that is necessary to ensure the correct performance of the impactor, 2) inhalation profiles used in the methods are not equivalent to those of a human, 3) right-angle bend tube used in the methods strongly distorts the measured particle size distribution as large particles carrying smaller drug particles hit the throat piece and induce severe particle break-up, and 4) turbulent characteristics of the flow in the throat piece easily leads to undesired particle break-up and fine particle deposition downstream of the inhaler and further distorts the measured size distribution. Currently used sampling methods, therefore, neither measure particles exiting the inhalation device nor particles entering the sizing equipment accurately.

Thus, the proposed method aims to

1. Assure measured particles are the same as those that exit the inhalation device (avoid changing characteristics of the particle such as size distribution and charge before sampling takes place)

2. Assure that all particles that leave device and are within the size regime of the particle sizing equipment are measured (avoid deposition of measurable particles on exposed surfaces before sampling takes place)

3. Assure particles are generated under realistic operating conditions (operate the inhaler as if it were being used by a patient)

4. Focus on measuring details of the respirable (useful) particles (particles larger than respirable are removed from the fluid stream)

5. Assure measuring instruments give accurate results (flow in measuring/analysis instrument is always maintained at its appropriate steady-state operating conditions during the entire sampling period)

Summary of the invention

The object of the present invention is to overcome the drawbacks mentioned above and to provide a means of accurately measuring characteristics of aerosols produced in unsteady aerosol sources (UAS). Since the behaviour of aerosols downstream of the UAS is dependent on a number of variable external factors (such as a patient's physiology and inhalation technique in the case of inhalers), it is unfeasible to design a sampling system that can account for all possible variations. The current system is designed to take into account the effects of a variable pressure/velocity cycle only, thus eliminating the effects of variability outside the source. This is accomplished by measuring the particle size distribution of the aerosol from the UAS in a manner which minimises distortion of the measured size distribution, and by separately controlling the gas flow rate from the UAS and that through one or more measuring/analysis instruments, by eliminating from the gas stream particles above the measurable particle size, and by drawing the aerosol from the UAS via a time- dependent pressure or velocity profile. Though, the present invention overcomes problems associated with measurement of size distribution properties in unsteady systems, it is equally applicable to steady systems.

The above mentioned objective is achieved by a method, wherein the particle deposition, resuspension, agglomeration and particle break-up during sampling is impeded by feeding sheath gas into a flow adjustment section connected to a sampling line, wherein the relative concentration of the emitted particles is measured as a function of time with an optical detector, wherein the measuring/analysis instruments are continuously running at optimal conditions during the entire sampling process and drawing the sheath gas enabling a zero flow rate through the aerosol source prior to the initiation of the time-dependent pressure or velocity profile, wherein the sheath gas and total out flowing gas are controlled so as to draw the aerosol from the aerosol source via a time-dependent pressure or velocity profile.

The main advantages of the presented method when compared to the prior art are a flow adjustment section and a controllable valve. The flow adjustment section enables the feeding of sheath gas through the tube wall to protect fine particles from depositing, allows the settling out of large particles if required, allows dissipation of the incoming jet, all of which have the potential to distort the measured size distibution due to deagglomeration, turbulent or inertial impaction, reentrainment, etc., and additionally enables the controlling of flows in such a way that measuring/analysis instruments are continuously running at their appropriate operating conditions during the entire sampling process. The controllable valve allows the specification of predefined pressure or velocity profiles through the UAS or specification of pressure or velocity profiles generated via feedback from conditions in the WAS.

The use of sheath gas in the present innovation differs substantially from that in the method introduced by Miller et al. in as much as the function of the supply air in the latter method is merely to enable a constant flow rate through the impactor during the sampling. The supply air of the method introduced by Miller et al. is fed into the sampling line downstream of the USP inlet, whereupon particle deposition has already occurred within the USP inlet. In the prior art there are also known sampling methods where dilution is executed by various techniques (Willeke and Baron 1993). However, these methods can only be applied to steady aerosol sources. The method introduced by Miller et al. also uses a piston to generate the flow and thus is limited to the study of purely unsteady phenomena of a fixed total volume.

The example of the invention is presented for an application in which the aerosol source is a commercial dry powder inhaler, though the invention is applicable to any unsteady or steady aerosol source be it a device, such as an aerosol spray can, a metered dose inhaler or other aerosol delivery system (ADS) or wider aerosol source (WAS) such as the flue gas from a combustion process, a metallurgical process, a powder synthesis process, a reciprocating or jet engine, a spray coating or spray drying process.

Description of the drawings

Figure 1 illustrates the schematic of the measurement system allowed by the

European and US Pharmacopoeias.

Figure 2 illustrates the invented apparatus of claim 16, seen in cross-section in side view.

Figure 3 illustrates the schematic diagram of the invented method when sampling from an unsteady aerosol source with one or more measuring/analysis instruments. I

= unsteady aerosol source (UAS), II = inlet providing sheath gas, in = flow adjustment section, TV = sampling line, N = measuring/analysis instrument 1, VI = exhaust 1, Nπ = measuring/analysis instrument 2 and Vm = exhaust 2. Qι is the variable flow rate from the UAS into the flow adjustment section. β₂ is the variable flow rate of sheath gas provided into the flow adjustment section. Q₃ is the constant flow rate from the flow adjustment section into the sampling line. Q₄ is the constant flow rate from the sampling line into the measuring/analysis instrument 1. Q₅ is the constant flow rate from the sampling line into the exhaust line 1. Q₅ can be divided into additional exhaust and measuring/analysis instrument pairs until there is insufficient flow to operate an additional measuring/analysis instrument. Figure 4 illustrates the schematic diagram of the invented method when sampling from a wider aerosol source (WAS) with one or more measuring/analysis instruments. I = wider aerosol source (WAS), π = inlet providing sheath gas, III = flow adjustment section, IV = sampling line, N = measuring/analysis instrument 1, VI = exhaust 1, VII = measuring/analysis instrument 2 and VIII = exhaust 2. Qj is the variable flow rate drawn from the WAS into the flow adjustment section. Q_mai- Qi is the variable excess flow rate which is left from the total flow rate Q_totaι after the flow rate Qj is sampled from the total flow rate. β₂ is the variable flow rate of sheath gas provided into the flow adjustment section. Q₃ is the constant flow rate from the flow adjustment section into the sampling line. Q₄ is the constant flow rate from the sampling line into the measuring/analysis instrument 1. Q₅ is the constant flow rate from the sampling line into the exhaust line 1. Q₅ can be divided into additional exhaust and measuring/analysis instrument pairs until there is insufficient flow to operate an additional measuring/analysis instrument.

Figure 5 illustrates the flow chart within the control unit of the invented apparatus of claim 16. 18 = air intake from compressed-air line, 19 = pressure reducer, 20 = manual pressure control, 21 = buffer tank, 8 = high-speed electronical control valve, 14 = sheath air flow Q₂, 10 = computer, 11 = infrared light detector, 9 and 22 = differential pressure gauges, 13 = ADS outlet, 6 = laminar flow element and 39 = AD converter. Figure 6 illustrates the sampling line of the invented apparatus of claim 16 when two inertial aerosol sizing instruments, i.e. Berner-type low-pressure impactors (BLPI) are connected in a row into the sampling line. Figure 7 illustrates the isokinetic sampling from the centre of the cross-section of the sampling line of the invented apparatus of claim 16.

Figure 8 illustrates the total mass size distributions of an inhalation aerosol from a commercial multi-dose DPI measured with two BLPIs at the two sequential sampling locations. The solid and dashed curves correspond to the size distributions from the first and second test, respectively. The curves marked with plus signs and crosses correspond to the size distributions measured with the first and second BLPI, respectively. Each size distribution is derived from 40 doses. Figure 9 illustrates the drag mass size distributions of an inhalation aerosol from a commercial multi-dose DPI measured with two BLPIs at the two sequential sampling locations. The solid and dashed curves correspond to the size distributions from the first and second test, respectively. The curves marked with plus signs and crosses correspond to the size distributions measured with the first and second BLPI, respectively. Each size distribution is derived from 40 doses. Figure 10 illustrates averaged relative concentrations of the emitted particles from a commercial multi-dose DPI as a function of time measured with the infrared light source/detector couple. The solid and dashed curves without symbols correspond to the measured pressure drops over the DPI with a fast and slow profile, respectively. The flow rate Q through a DPI is proportional to the square root of the pressure drop Δp over the DPI, i.e. Q -*^• (Δpψ². Both averaged relative concentration curves, marked with triangles and squares for the fast and slow profiles, respectively, are the averages of relative concentrations measured from 20 doses.

Detailed description of the preferred embodiments

In the following a preferred embodiment of the invented method and apparatus and their additional advantages are described in more detail with reference to the appended drawings.

The invention includes:

1. Variable bypass flow to maintain constant flow through the sampling line in conjunction with variable flow through the ADS. 2. Variable mass flow rate flow adjustment zone to remove the largest particles from the flow before the entering the measuring/analysis instruments and to minimise particle reentrainment back into the main flow field.

3. Variable mass flow rate through the flow adjustment section to focus particles to create a particle free layer of gas near the tube walls to reduce near wall turbulence and particle deposition.

4. Low turbulence low curvature bend and flow development tube to stabilise the flow entering the downstream measuring/analysis instruments

The ADS 1 which is under investigation is placed in the ADS chamber 2, so that it forms a leak tight joint with the adapter 3, which connects the ADS to flow adjustment section 15-17 adjacent to the sampling line 23-28, which is closed to the atmosphere (Figure 2). The ADS chamber is covered by a chamber lid 4. Gas is drawn into the ADS via the intake 5 which is connected to the laminar flow element

6 for measuring the steady-state flow rate through the ADS. The chamber and laminar flow element are not necessary for the proper operation of the apparatus. The pressure drop over the laminar flow element is measured and converted into an equivalent flow rate by the control unit 7 and computer 10. In the preferred embodiment, the time-accurate relative concentration of emitted particles is measured by the infrared light source/detector couple 11-12. In the preferred embodiment, the measurement of particle concentration is based on the attenuation of the infrared light beam which intersects with the emitted aerosol cloud on a vertical plane immediately downstream of the ADS outlet 13. However, any other appropriate optical apparatus can be incorporated which does not appreciably disturb the flow. In the preferred embodiment, the flow rate through the ADS is obtained by measuring the pressure drop between the ADS chamber 2 and the ADS outlet 13. Any appropriate flow rate measuring apparatus can be incorporated which does not induce particle losses, does not appreciably disturb the flow and for which the resistance within the sampling line is negligible compared to that within DPIs. To control the flow rate through the ADS, to moderate the turbulence experienced by the aerosol after exiting the ADS and to impede fine particle deposition during sampling, sheath gas 14 is fed into the flow adjustment section 15-17. This forces small particles from the surface and creates a film of clean, particle free gas along the pipe while allowing larger particles to settle. In the preferred embodiment, sheath gas is fed through a round porous tube 16 which is mantled with a metal tube 17, though any feed mechanism which provides a source of low turbulence particle free gas near the exposed surface of the tube is applicable such as a coaxial jet coflow, a perforated wall, a perpendicular slit injection, or other applicable sheath gas introduction method. The total flow within the sampling line is the sum of the flow through the ADS and sheath gas flow:

Qtotal — QADS + Qsheath

To minimise the generation of turbulence outside the device, the flow adjustment zone 15 and sampling line 23-28 are of sufficient cross-sectional area to slow the flow down to velocities at which turbulence is modulated in order to reduce particle deposition, deagglomeration and reentrainment and to allow a sufficient residence time for settling to pre-select particles of interest before turning the flow from a horizontal direction to the vertical for measurement. For a given Qtotal, the Reynolds number in the sample line is preferably less than 3500 and more preferably less than 2500 to maintain a low level of turbulence.

The flow chart and the main components of the control unit for the preferred embodiment are presented in Figure 5. In the preferred embodiment, the intake gas 18 is supplied to the control unit 7 via a compressed gas line. Within the control unit, the pressure of input gas is first reduced by a pressure reducer 19. The pressure of input gas that is fed into the flow adjustment section 15-17 via a buffer tank 21 and a high-speed electronic control valve 8 is adjusted by a manual pressure control 20 before each test series. A predefined time-dependent pressure or velocity profile is executed via the control computer 10. Consequently, the control unit triggers the high-speed valve setting to regulate the flow rate of the sheath gas fed into the flow adjustment section. A differential pressure gauge 9 is included in the control unit to measure the pressure drop over the ADS as a function of time. The volumetric gas flow rate through the ADS as a function of time is determined from pressure drop data.

From the flow adjustment section the aerosol is led into a modular sampling line 23- 28, which consists of a variable number of replaceable modules. The aerosol is sampled via a sampling probe inlet 30 connected to a sampling probe bend 27, i.e. a module attached to the sampling line. Sampling probe bends 27 can be equipped with a probe inlet 30 of a suitable size to enable an isokinetical sampling for each of the available measuring/analysis instruments. In the preferred embodiment, the sample is drawn isokinetically from the centre of the cross-section of the sampling line 29 into an upright-positioned impactor 32 by using a vacuum pump 35 which is continuously running at its appropriate operating conditions to ensure a stabilised flow within the impactor and accurate measurement of the sampled aerosol. The measuring instrument typically used in the tests is a Berner-type low-pressure impactor (BLPI) or an electronic low-pressure impactor (ELPI), though any appropriate measuring instrument can be substituted. The analysis instrument may include a TEM grid sampler, a filter, etc. The downstream pressure of the impactor is controlled by a valve 34 and pressure meter 33. The by-pass flow rate is controlled by another valve 37 and pressure meter 36. The by-pass flow is sucked through an absolute filter 31 by a vacuum pump 38. Alternatively the aerosol size distributions can be measured simultaneously with two or more measuring instruments so that there are two or more sequential sampling locations in the sampling line. In the beginning of the test, the incoming sheath gas flow rate < ₂ is adjusted to compensate the total flow rate Q₃, i.e. all the gas drawn into the measuring/analysis instrument and exhaust line (Figure 3). Thus zero flow rate through the ADS adapter orifice is obtained. That can be confirmed by a flow meter and/or a pressure drop meter or other similar device. The sheath gas is drawn through the measuring instrument so that the flow through the instrument has stabilised. Properties of the sheath gas, such as gas composition and relative humidity are controlled upstream of its introduction into the flow adjustment section. The ADS is inserted into the adapter orifice and covered by the chamber lid. To release the powder from the ADS, a time-dependent pressure or velocity profile is run from the control computer. This reduces the sheath gas flow rate >₂ and correspondingly increases the gas flow rate Qi through the ADS so that the total flow rate Q₃ remains constant. Properties of the ADS chamber gas, such as gas composition and relative humidity, are controlled upstream of its introduction into the ADS chamber. Accordingly, the method can be used to control the particle environment, such as humidity, gas composition, charge state, etc. Additionally the residence time experienced by particle can be controlled by varying the geometry of the sampling line to study unsteady particle dynamic effects such as surface reactions and condensation. Simultaneously with the running of the pressure/velocity profile, the measuring of relative particle concentration with the light detection device is automatically launched by the control unit. When the maximum of the flow rate through the ADS is reached, sheath gas is still fed to some degree into the flow adjustment section to impede fine particle deposition and reduce wall-induced turbulence. In order to reduce the distortions to the particle size distribution of the range of interest due to e.g. overloading of the measuring/analysis instruments and further particle deagglomeration, length and surface area of the flow adjustment zone can be selected so that the particles larger than the size of interest are removed from the sample flow by gravitational settling. Thus, this design slows the flow down to reduce the turbulence, pre-selects particles of interest before the bend and turns the flow from a horizontal direction to the vertical for measurement. Example 1

Two identical test runs were carried out to characterise the behaviour of a DPI with the invented method and apparatus. The tested inhaler was a typical commercial multi-dose DPI which contains micron-size drug as the active ingredient and coarse lactose as the excipient. In each test, 40 doses were delivered from the DPI. The total flow rate in the sampling line was 80 1/min. To release the powder from the DPI, a predefined time-dependent velocity profile with a fast ramp was executed from the control computer. The sheath gas was filtered and dried air. The relative concentration of emitted particles was measured with the infrared light device. The total and drag mass size distributions were determined with two simultaneous BLPIs and HPLC analysis of the BLPI samples. Homogeneity of the sampled aerosol was validated by connecting the two BLPIs in a row into a sampling line (Figure 6). The samples into the impactors were drawn from the centre of the cross-section of the sampling line (Figure 7).

The measured total mass size distributions are shown in Figure 8. The mass size distributions at the two sequential sampling locations are in a very good agreement, especially in the respirable size fraction (particles smaller than 5 μm). The results also imply a good reproducibility between the tests. For the particles greater than 5 μm, which are of secondary importance for drag delivery to the lungs, gravitational settling causes lower size distributions at the second sampling location compared to those at the first sampling location. Coarse particles of non-respirable size fraction (mainly larger than 50 μm) are mostly deposited by gravitational settling within the flow adjustment section and sampling line. The drug size distributions at the two sampling locations are in a good agreement as well (Figure 9).

Example 2

Two test runs were carried out to study the emission of the aerosol particles delivered from a DPI with the invented method and apparatus. The tested inhaler was a similar DPI as in Example 1. In each test, 20 doses were delivered from the DPI. The total flow rate in the sampling line was 100 1/min. A predefined time-dependent velocity profile with a fast ramp was executed in the first test ran and a predefined time- dependent velocity profile with a slow ramp was executed in the second test ran. The sheath gas was filtered and dried air. The relative concentration of emitted particles was measured with the infrared light device. The relative concentrations of the emitted particles are shown in Figure 10. It was observed that an overwhelming part of the powder is released at very early stage of the inhalation process.

Example 3

In addition to the study of an aerosol delivery system (as a DPI in Examples 1 and 2) the invented method and apparatus are equally applicable to the study of unsteady or steady behaviour of a wider aerosol source (WAS), e.g. engine emissions in combustion operations such as occur in smelters, power plants, recovery boilers, powder production, reciprocating engines and turbine engines and soot blowing operations. In such instances, the flow rate <2- is a portion of said aerosol source and flow rate Qι is introduced into the flow adjustment section according to a feedback mechanism obtained from conditions in said aerosol source which are based on e.g. measured time-dependent free-stream gas velocity for isokinetic sampling or gas concentration for maintaining constant dilution ratio in the sampling line (Figure 4).

While preferred embodiments of the invention have been shown and described, modifications thereof can be made by one skilled in the art without departing from the spirit of the invention. For instance, as a replacement for the use of BLPI and

ELPI as presented above, the sampling method can be coupled with any applicable measuring/analysis instrument, e.g. Andersen cascade impactor and Next Generation

Impactor (NGI) for pharmaceutical applications. Particles can also be collected onto any applicable filter by connecting the filter to the line of outgoing gas or to a sampling probe. Also instead of an infrared light device, as a particle concentration detector, any other applicable method of measurement can be used, e.g. laser light device. A replacement for the porous tube in the flow adjustment section could be, for instance, a coaxial jet coflow, a perforated wall, a perpendicular slit injection, or other applicable sheath gas introduction method. Although, the current embodiment shows a constant diameter mantle and porous tube with constant wall thickness and porosity, other embodiments can utilise a variable cross-sectional area mantles and porous tubes with variable wall thicknesses and porosities. Though, the preferred embodiment of the device uses centered, isoaxial, isokinetic sampling, other embodiments of the device may employ noncentered, anisoaxial, anisokinetic sampling.

References cited:

European Pharmacopoeia supplement (2001), Preparations for inhalation: fine particle dose and particle size distribution, 2.9.18. US Pharmacopoeia (1999). 24^th rev. Aerosols, metered-dose inhalers, and dry powder inhalers <601>, 1895-1912.

Burnell, P. K. P., Malton, A., Reavill, K. and Ball, M. H. E. (1998). Design, validation and initial testing of the electronic lung device. J. of Aerosol Sci. 29 (8),

1011-1025. Miller, N. C, Maniaci, M. J., et al. (2000). Aerodynamic sizing with simulated inhalation profiles: total dose capture and measurement. Proc. Respiratory Drug

Delivery VII, Serentec Press, Raleigh, NC, USA, Vol. 1, 191-196.

Li, W., Perzl, M., et al (1996). Aerodynamics and aerosol particle deaggregation phenomena in model oral-pharyngeal cavities. J. of Aerosol Sci. 27 (8), 1269-1286. Willeke, K. and Baron, P. A. (1993). Aerosol Measurement. Principles, Techniques and Applications. Van Nostrand Reinhold, New York.

Claims

1. A method for studying the behaviour of an unsteady aerosol source (UAS) characterized in that

- the particle size distribution of the emitted particles which are exiting from the UAS is measured in a manner which minimises distortion of the measured size distribution and ensures proper interpretation of that size distribution by

- separately controlling the gas flow rate from the UAS and that through one or more measuring/analysis instruments

- introducing sheath gas into a flow adjustment section to control particle deposition, particle break-up, particle reentrainment, particle reagglomeration, condensation and/or nucleation or other particle dynamic mechanisms which may alter the measured size distribution

- operating the measuring/analysis instruments at their appropriate operating conditions during the entire unsteady sampling process - controlling the sheath gas and total out flowing gas so as to draw the aerosol from the UAS via a time-dependent pressure or velocity profile

2. The method as defined in claim 1 characterised in that

- a sample gas flow is drawn from the UAS (I; Figure 3) into a flow adjustment section (HI), consisting of an inlet providing sheath gas (H) connected to a sampling line (IV)

- flow rate Qi into the flow adjustment section is variable

- flow rate β into the flow adjustment section is controlled so as to provide an essentially constant flow rate Q₃ into the sampling line

- a portion Q₄ of flow rate Q₃ is drawn into measuring/analysis instrament 1 (V) at its appropriate operating conditions

- flow rate Q₅ is exhausted (VI)

3. The method as defined in claim 2 characterised in that

- the exhausted flow rate Q₅ can be divided into additional exhaust and measuring/analysis instrament pairs until there is insufficient flow to operate an additional measuring/analysis instrument

4. The method as defined in any one of claims 1-3 characterised in that - the UAS is an aerosol delivery system (ADS) wherein the flow rate Qi is introduced into the flow adjustment section according to a predefined time- dependent pressure or velocity profile and Qj is essentially the entire flow rate through the ADS 5. The method as defined in any one of claims 1-3 characterised in that

- the UAS is a fraction of a wider aerosol source (WAS) wherein the flow rate Qi is a portion of said aerosol source and flow rate Qi is introduced into the flow adjustment section according to a feedback mechanism obtained from conditions in said aerosol source which are based on e.g. measured time-dependent free-stream gas velocity for isokinetic sampling or gas concentration for maintaining constant dilution ratio in the sampling line (Figure 4) 6. The method as defined in any one of claims 1-4 characterised in that

- the ADS is a dry powder inhaler (DPI), a metered-dose inhaler (MDI) or other aerosol drag delivery system 7. The method as defined in any one of claims 1-4 characterised in that

- the predefined time-dependent pressure or velocity profile is produced by a control unit.

8. The method as defined in any one of claims 1-7 characterised in that

- the sampled aerosol is conditioned by controlling the properties of sheath air, e.g. humidity, charge, gas composition and temperature.

9. The method as defined in any one of claims 1-8 characterised in that

- the residence time of the sampled aerosol in the sampling line is controlled by varying the cross-sectional area and/or length of the sampling line

10. The method as defined in any one of claims 1-9 characterised in that - relative concentration of the emitted particles is measured as a function of time by an optical detector.

11. An apparatus for carrying out the method of claim 1 characterised in that it comprises

- a flow adjustment section whereinto sheath gas is introduced and whereinto a sample is drawn from an aerosol source

- an inlet connecting an aerosol source to said flow adjustment section

- an inlet providing sheath gas into said flow adjustment section - a sampling line, which connects the flow adjustment section to one or more measuring/analysis instruments

- one or more measuring/analysis instruments for particle characterisation of the sampled aerosol

12. The apparatus as defined in claim 11 characterised in that

- said inlet providing sheath gas is a porous tube, a perforated wall, a perpendicular slit injection, a coaxial jet coflow, or other applicable sheath gas introduction method.

13. The apparatus as defined in claim 11 or 12 characterised in that - said sampling line and the flow adjustment zone of said flow adjustment section are of sufficient cross-sectional area to slow the total flow down to velocities at which turbulence is modulated in order to reduce particle deposition, deagglomeration and reentrainment and to allow a sufficient residence time for settling to pre-select particles of interest before turning the flow from a horizontal direction to the vertical for measurement, the Reynolds number in the sample line being preferably less than 3500 and more preferably less than 2500 to maintain a low level of turbulence.

14. The apparatus as defined in any one of claims 11-13 characterised in that

- the sampled aerosol is conditioned by controlling the properties of sheath air, e.g. humidity, charge, gas composition and temperature

15. The apparatus as defined in any one of claims 11-14 characterised in that

16. An apparatus for carrying out the method of claim 1 characterised in that it comprises

- an aerosol delivery system (ADS) chamber (2; Figure 2 and Figure 5) connected to a gas intake passage (5) and whereinto the ADS (1) under investigation is placed, a laminar flow element (6) connected to the gas intake passage, an adapter (3) connected to the ADS chamber and whereinto the ADS is inserted, - a flow adjustment section (15-17), consisting of a porous inner tube (16) and a gas tight outer tube (17) connected to the adapter (3) and whereinto the sample flow is drawn from the ADS, - a sheath gas feeding (14) into the flow adjustment zone (15) through the porous tube for preventing fine particle deposition of the sample within the flow adjustment section (15-17),

- a control unit (7) and a high-speed valve (8), to control the sheath gas fed into the flow adjustment zone (15) through the porous tube

- a modular sampling line (23-28) which is connected to the flow adjustment section (15-17),

- a variable number of measuring/analysis instruments which are connected to the sampling line (23-28) via corresponding number of sampling probe inlets (30) connected to the sampling probe bends (27),

- an optical detector (11,12) connected to the flow adjustment section (15-17) near the ADS (1) outlet for measuring the relative concentration of the emitted particles a control unit (7) combined with a computer (10) for producing desired time- dependent pressure or velocity profiles through the ADS and

- a differential pressure gauge (9) for measuring the pressure drop over the ADS 17. The apparatus as defined in claim 16 characterised in that

- the aerosol drag delivery system (ADS) is a dry powder inhaler (DPI)