Surface acoustic wave device for the nebulisation of therapeutic liquids

Abstract

A device is disclosed for the preparation of nebulised droplets, for inhalation. The device has: a surface acoustic wave (SAW) transmission surface; a SAW transducer adapted to generate and propagate SAWs along the SAW transmission surface; and an array of cavities opening at the SAW transmission surface for containing a liquid. In operation, SAWs propagating along the SAW transmission surface interact with the liquid in the cavities to produce nebulised droplets of the liquid. Operation of the device results in a nebulised plume of droplets of average diameter in the range 1-5 μm.

Claims

1. A device for the preparation of nebulised droplets, the device having: a surface acoustic wave (SAW) transmission surface; a SAW transducer that generates and propagates SAWs along the SAW transmission surface; and cavities provided in an array of cavities opening at the SAW transmission surface for containing a liquid, wherein a maximum dimension of the cavities in a direction perpendicular to a depth of the cavities is less than 500 μm, wherein the cavities are not tapered, wherein, in operation, the SAWs propagating along the SAW transmission surface interact with the liquid in the cavities to produce nebulised droplets of the liquid with a respirable fraction of at least 80%.

2. The device according to claim 1 wherein the SAW transmission surface is a surface of a superstrate coupled to the SAW transducer.

3. The device according to claim 1 wherein the cavities have substantially the same shape.

4. The device according to claim 1 wherein the cavities are closed at an end distal from the SAW transmission surface.

5. The device according to claim 1 wherein the cavities are open at an end distal from the SAW transmission surface.

6. The device according to claim 1 wherein the cavities have substantially the same dimensions.

7. The device according to claim 1 wherein the array of cavities is an ordered array.

8. The device according to claim 1 wherein each cavity has an interior surface, said interior surface of the cavities being chemically, physically or electrically modified in order to promote the containment of the liquid in the cavities.

9. The device according to claim 1 wherein the SAW transmission surface is chemically, physically or electrically modified in order to promote the containment of the liquid in the cavities.

10. The device according to claim 1 wherein the device includes a plurality of arrays of cavities, operable to contribute to a rate of nebulisation of liquid from the device.

11. A method for the preparation of nebulised droplets, the method including: providing a device having a surface acoustic wave (SAW) transmission surface, a SAW transducer that generates and propagates SAWs along the SAW transmission surface, and cavities provided in an array of cavities opening at the SAW transmission surface, wherein a maximum dimension of the cavities in a direction perpendicular to a depth of the cavities is less than 500 μm, wherein the cavities are not tapered; containing a liquid in the cavities; and causing the SAWs to propagate along the SAW transmission surface to interact with the liquid in the cavities to produce nebulised droplets of the liquid with a respirable fraction of at least 80%.

12. The method according to claim 11 wherein when the SAW transmission surface is facing upwards, the liquid is contained in the cavities such that a free surface of the liquid is below the level of the SAW transmission surface.

13. The method according to claim 11 wherein operation of the device results in a nebulised plume of droplets of average diameter in a range of 1-5 μm.

14. The method according to claim 11 wherein operation of the device results in a nebulised plume of droplets, the respirable fraction corresponding to the droplets having a diameter in a range of 1-5 μm.

15. The method according to claim 11 further including the step of supplying the liquid for nebulisation after the device is provided and before the step of containing the liquid in the cavities.

16. A method for the preparation of nebulised droplets and their delivery to a subject for therapeutic treatment, the method including: providing a device having a surface acoustic wave (SAW) transmission surface, a SAW transducer that generates and propagates SAWs along the SAW transmission surface, and cavities provided in an array of cavities opening at the SAW transmission surface, wherein a maximum dimension of the cavities in a direction perpendicular to a depth of the cavities is less than 500 μm, wherein the cavities are not tapered; containing a liquid in the cavities; causing the SAWs to propagate along the SAW transmission surface to interact with the liquid in the cavities to produce nebulised droplets of the liquid with a respirable fraction of at least 80%; and delivery of the nebulised droplets to the subject for therapeutic treatment by inhalation.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Embodiments of the invention will now be described by way of example with reference to the accompanying drawings in which:

(2) FIG. 1A shows a schematic plan view of a device according to an embodiment of the invention, in the form of an etched array of cavities in a superstrate on an interdigitated electrode transducer (IDT) surface.

(3) FIG. 1B shows a cross sectional view of the device of FIG. 1A.

(4) FIG. 2 shows the results of droplet size distribution (number-based) analysis of droplets generated from nebulisation (a) an embodiment of the present invention, with the liquid contained in cavities in a silicon superstrate coupled on a SAW device with excitation frequency of 8.6 MHz and input power of 1.5 W (b) Medix nebuliser (c) Medisana nebuliser and (d) directly on the SAW device of (a) with excitation frequency of 8.6 MHz and input power of 1.5 W.

(5) FIG. 3 shows the respirable fraction of nebulised droplet generated from the commercialised nebulisers (Medix (3) and Medisana (4)), directly on SAW devices (2) on silicon superstrate coupled on the SAW device (1) with excitation frequency of 8.6 MHz and input power of 1.5 W

(6) FIG. 4A shows a micrographic image captured from a video of nebulisation at 11.762 MHz and −4 dBm of DI water at 2 μl/min on a plain surface. Large individual drops are seen due to free capillary microjets at the surface of the drop.

(7) FIG. 4B shows a micrographic image captured from a video of nebulisation at 11.762 MHz and −4 dBm of DI water at 2 μl/min located in cavities arranged as a phononic lattice (900 μm diameter). No large individual drops are seen. As a guide to the scale of the images of FIGS. 4A and 4B, the syringe shown in the image is a 1 ml syringe, with a syringe body diameter of 5 mm.

(8) FIG. 5 shows a schematic cross sectional view of a single cavity.

(9) FIG. 6 shows a schematic cross sectional view of a single cavity which is a modification of the cavity shown in FIG. 5.

(10) FIG. 7 shows a schematic cross sectional view of a single cavity which is another modification of the cavity shown in FIG. 5.

(11) FIG. 8 shows a plan view of the cavity of FIG. 7.

(12) FIG. 9 shows a schematic cross sectional view of a single cavity which is a modification of the cavity shown in FIG. 8.

(13) FIG. 10 shows a plan view of the cavity of FIG. 9.

(14) FIG. 11 shows a graph of droplet size with cavity (pore) diameter, based on an assessment of largest droplet size viewed in video footage.

(15) FIGS. 12-14 show images from high frame rate video footage taken using a microscope when water is nebulised from cavities of diameter 80 μm (FIG. 12), 600 μm (FIG. 13) and 1500 μm (FIG. 14).

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS, AND FURTHER OPTIONAL FEATURES OF THE INVENTION

(16) Before discussing the features of the preferred embodiments of the present invention in detail, it is useful to consider the features and performance of known ultrasonic nebulisers.

(17) Ultrasonic nebulisers use the basic principle of applying a high frequency mechanical vibration to a surface. This leads to the excitation of deformations on the free liquid surface that result in microjets [Topp (1973)]. These nebulisers enable the atomisation of a wider range of liquids than other types of nebulisers (such as jet or compressed air). However the aerosol produced suffers from wide range of droplet size. Recently this principle has been extended to the use of SAWs [Reboud, Wilson et al (2012); Qi et al (2009)], which offer the advantages of lower powers and more versatility in integration of preparation functions. However these suffer from similar limitations in the control of the drop size, which generally leads to large mean diameters (above 10 μm) and multiple modes.

(18) To provide a tight droplet size distribution, meshes have been introduced as passive filters (MICROAIR® filter (Omron Healthcare, Inc., Kyoto, Japan), and MICROFLOW® filter (Pfeiffer Vacuum™ GmbH, Annecy, France)) situated after the nebulization process, to select the drops of the correct size. These systems require careful maintenance (to prevent clogging) and show limited efficiency.

(19) Vibrating meshes combine both approaches at the site of nebulisation [Maehara et al (1986)]. A mesh of apertures is vibrated at ultrasonic frequencies to generate the aerosol from a pinching off of the drops through the aperture, in a similar mechanism as the microjets mentioned previously for SAW nebulisation. A similar system is commercialised for droplet dispensing (Scienion AG).

(20) In the preferred embodiments of the present invention, the array of cavities is used to prevent the pinching off enabled by the vibrating meshes and thus provide the opportunity of a reduced size without requiring fine apertures (on the order of the size of the drop dispensed). This provides a cheaper manufacturing strategy. It is also not reliant on the surface properties of the mesh and thus can tolerate conditions that would lead to significant clogging, enabling the dispensing on difficult suspensions, such as those with high viscosity.

(21) Qi et al [2009] have shown nebulisation off a paper superstrate, using SAW. Although the paper superstrate could be viewed as a mesh, their work clearly show no capillary wave limiting effect on the selection of droplet sizes (see FIG. 6 of Qi et al [2009], clearly showing large (i.e. greater than 10 μm) droplets). This is due to the wide distribution of pore sizes compared to the embodiments presented here. Indeed, in their work, the paper superstrate is used as a matrix to feed the liquid, while the nebulisation happens in a bulk mode (as a drop—see FIG. 2c of Qi et al [2009]).

(22) A preferred embodiment of the present invention is illustrated schematically in FIGS. 1A and 1B. This is based on the inventors' previous SAW-based systems [see WO 2011023949, WO 2011060369, WO 2012114076, WO 2012156755, Reboud, Wilson et al (2012) and Reboud, Bourquin et al (2012)].

(23) The device includes a LiNbO.sub.3 actuator 10 (single crystal, self-supporting) with an interdigitated electrode 12 and a Si superstrate 14, with etched blind holes 16. The holes (i.e. cavities) are arranged in a square periodic lattice array. A liquid (the sample) 18 is positioned inside the cavities. Thus, the height of the liquid in the cavities is less than the depth of the cavities. This is ensured using highly hydrophilic wetting and a small sample volume.

(24) Upon actuation, the SAW propagates on the SAW transmission surface (the upper surface of the actuator 10) and is coupled onto the superstrate 14, via a coupling medium (not shown) such as gel or water or glue or a more permanent fixture (the array of cavities can be deposited on or etched into the piezoelectric layer). The material of the superstrate 14 is preferably acoustically non-dampening (e.g. Si or glass).

(25) The superstrate 14 holding the array of cavities can be fully in contact with the piezoelectric actuator 10 (as shown in FIGS. 1A and 1B) or coupled only using a small overlap.

(26) The SAW then interacts with the liquid contained in the cavities 16. This interaction creates a nebulised plume. Here the cavities are used to prevent the creation of microjets of sizes greater than about 10 μm that result in multimodal droplet distribution. The specific mechanism for this is still under investigation by the inventors. Without wishing to be limited by theory, the present inventors believe that the mechanism is linked to the damping, suppression or forbidding of capillary waves propagating at the free surface of the liquid in the cavities. This capillary mechanism has been reported as the primary mechanism for nebulisation using SAW [Qi et al (2008)], and leads to sizes outside the range of interest for drug delivery.

(27) In more detail, the SAW actuator 10 and the superstrate 14 are manufactured as follows. Positive photoresist, S1818 (Shipley) was used to lithographically define the electrode pattern on the 127.8° Y-cut LiNbO.sub.3 substrate. After the resist exposure and development, 10 nm of titanium and 100 nm of gold were deposited and lift-off was performed in acetone.

(28) The superstrate was fabricated using <100> silicon wafer and standard optical photolithography. The array of cavities was constructed using dry etch (STS ICP), down to half the wafer thickness (about 250 μm). Control experiments were carried out on unpatterned superstrates as well as on the LiNbO.sub.3 actuator.

(29) In order to control the volumes and shape of drops deposited on the surface as well as to create controlled spatial areas for nebulisation, the superstrate was patterned with a hydrophobic silane using standard optical lithography. The process involved developing the exposed S1818 photoresist (Shipley) and surface treatment in O.sub.2 plasma before silanisation in a solution of trichloro (1H, 1H, 2H, 2H perfluorooctyl) silane (Aldrich) in heptane (Aldrich). The superstrate was then rinsed in acetone to create hydrophilic (untreated) spots of varying sizes in the range of 1-15 mm on a hydrophobic surface.

(30) The frequency response of the SAW actuator was observed using a network analyser (E5071C ENA Series, Agilent Technologies). To perform nebulisation of the liquids on the substrate, a high frequency electrical signal was supplied to the electrodes using a MXG Analog Signal Generator (N5181A, Agilent Technologies) and amplifier (ZHL-5W-1, MiniCircuits).

(31) The silicon superstrate and the piezoelectric substrate were assembled with KY-jelly (Johnson & Johnson) between them to provide efficient coupling.

(32) Measurements of droplet size were performed at 8.64 MHz at the input power of 1.5 W. A sessile drop of 3 μL of deionised (DI) water was used for each nebulisation using the embodiment device of the invention. As comparisons, the nebulised droplet size by two commercialised nebulisers, Medix and Medisana were also measured. The Microneb Medix uses a titanium vibrator which oscillates at approximately 180 kHz with input power of 1.5 W to generate the droplet which is then passing through metal alloy mesh. The Medisana is an ultrasonic nebuliser that operates at 100 kHz with input power of 3 W.

(33) The distributions of nebulised droplet with different sizes were measured using a laser diffraction technique (SPRAYTEC® laser diffraction system, Malvern™ Instruments Ltd, Malvern, UK) and represented in the form of a frequency distribution curve.

(34) The diameters of the nebulised droplet has been reported by Kurosawa et al (1995) by using a number distribution. They obtained the linear mean diameter, D.sub.10 and surface mean diameter, D.sub.32 of 19.2 μm and 34.3 μm, respectively for tap water nebulised directly on a SAW device with excitation frequency of 9.5 MHz and input power of 2.5 W for 0.1 ml/min nebulisation rate. The droplet size distribution had two modes with peaks at 10 μm and 40 μm which were reported to be due to the capillary wavelength and the intermittent burst drive, respectively. Smaller droplets (D.sub.10=6.8 μm and D.sub.32=15.0 μm) were obtained using SAW device with higher excitation frequency of 48 MHz and lower input power of 2.3 W for 170 μl/min nebulisation rate [Kurosawa et al (1997)]. Alvarez et al (2007) successfully nebulised insulin with mean diameter of 4.5 μm using 19.3 MHz SAW device at 0.3 W input power. By using the same image processing technique as previous authors, Ju et al (2008) estimated the mean diameters of nebulised bovine serum albumin (BSA) to be 5.7, 4.4 and 2.7 μm using SAW devices with excitation frequencies of 50, 75 and 95 MHz, respectively. Smaller droplets with mean diameters of 0.36, 0.38 and 0.4 μm were obtained using 10 MHz SAW device with input power of 0.97, 1.00 and 1.03 W, respectively [Ju et al (2010)].

(35) FIG. 2 shows the distribution obtained for the different surfaces used. They are presented as frequency distributions. The results show that both commercial nebulisers, utilising an ultrasonic technology, provide drop sizes above the optimum size for lung penetration (modes above 5 μm). These distribution are also broad, leading to significant wastage of the targeted therapy.

(36) As shown in FIG. 2, SAW nebulisation from a plain surface is able to provide a smaller droplet size than the commercial nebulisers, which would fit the therapeutically-relevant range (between 1 and 5 μm). However this actuation leads to secondary peaks (large sizes above 10 μm), and a broad distribution. These features lead to inefficient nebulisation and wastage of the liquid.

(37) Using the array of cavities to contain the liquid for nebulisation enables the prevention of large secondary peaks, and sharpens the distribution of the peak (1-5 μm) of interest.

(38) The results can be presented using the concept of respirable fraction, which reports the proportion of the total size distribution that is enabled by the different systems (the ratio of integral below the curves between 1 and 5 μm, over the total integral), as shown in FIG. 3, while the data analysed is presented in Table 1.

(39) TABLE-US-00001 TABLE 1 Derived parameters of the nebulised droplets generated by the surface acoustic waves devices and commercialised nebulisers measured using the SPRAYTEC ® laser diffraction system (Malvern ™ Instruments Ltd, Malvern, UK). SAW + Si SAW MEDIX superstrate (transducer MICRONEB ® MEDISANA ® with cavities only) nebuliser nebuliser Linear mean 1.81 ± 0.13  1.32 ± 0.18 3.00 ± 0.27  5.73 ± 0.94 diameter, D.sub.v10 (μm) Linear mean 2.10 ± 0.16  5.21 ± 6.87 5.84 ± 0.49 13.19 ± 1.89 diameter, D.sub.v50 (μm) Linear mean 2.50 ± 0.30 52.28 ± 4.06 11.09 ± 1.06  27.13 ± 6.64 diameter, D.sub.v90 (μm) Surface mean 2.16 ± 0.17  3.38 ± 0.82 5.13 ± 0.64 10.04 ± 1.71 diameter, D.sub.32 (μm) Volume mean 3.45 ± 2.24 17.88 ± 3.47 6.54 ± 0.57 15.06 ± 2.93 diameter, D.sub.43 (μm) % Respirable 96.16 ± 4.93  36.44 ± 6.25 38.38 ± 6.14   6.92 ± 2.89 Fraction (1 μm < D < 5 μm)

(40) In order to increase the quantity nebulised and establish the steady-state properties of the system, a syringe pump (NE-1000 Multi-Phaser™, New Era Pump Systems Inc.) was employed to regulate the amount of liquid supplied to the superstrate continuously. The pump was set to a constant flow rate of 1.0 μl/min to maintain continuous production or rapid generation of nebulised droplets.

(41) FIGS. 4A and 4B show example frames extracted from movies on a plain superstrate (FIG. 4A) and on a superstrate having an array of cavities (FIG. 4B), illustrating the proposed hypothesis that the array of cavities prevents the generation of larger droplets.

(42) The cavities can be formed to extend through the superstrate. This allows the cavities to be replenished with additional liquid by filling from below under capillarity. This allows the delivery of the liquid to be more robust. In this embodiment the liquid can also be used as the coupling agent.

(43) Using a single spot and powers below 1 W, the illustrated embodiment enables flow rates around 20 ul/min. The flow rate can be increased significantly by increasing actuation power (up to 5-10 W) for a short period of time (<5 s).

(44) In order to further increase the flow rate, there may be provided a plurality of locations (each having an array of cavities) at which nebulisation is carried out substantially simultaneously. In this way, the flow rate can be increased to 5 ml/min, or higher.

(45) It is preferable to establish the nebulisation in specific areas, defined by wetting barriers. However, in the case of isotropic excitation on large arrays of cavities, the liquid may then be assembled in patches by irregularities and build up in volumes that are locally bulging above the surface of the SAW transmission surface of the superstrate. This behaviour would prevent the activity of the cavities and would result in multimodal droplet sizes, due to microdroplet ejection. Thus, it is instead preferred to provide wetting channels at the interface between the superstrate and the piezoelectric actuator. Such channels can also serve as acoustic waveguides to channel the SAWs to the nebulisation areas.

(46) In preferred embodiments, a suitable cavity diameter is 50-200 μm. The effect of the diameter on the ability of liquid in the cavities to support capillary waves can be considered based on the driving frequency f. As explained above, the diameter D of the cavities is preferably lower than a size that would allow the generation of unwanted large droplets, thought to be the result of unwanted capillary waves.

(47) The theoretical framework for the mechanism of large droplet suppression is not fully understood at the time of writing. The application of the progression of resonant responses from the fundamental mode upward, provided by the Lamb model, as set out in Blarney et al (2013), can be considered in which, for the sake of simplicity, the driving frequency f can be considered to be identical to f.sub.m for the fundamental mode. Whilst this is effective for low frequencies (in the kHz range or below) it is not effective for MHz range driving frequencies, for reasons which are not fully understood at the time of writing.

(48) It is therefore more suitable here to take an empirical approach to the design of the cavities. FIG. 2(d) shows that nebulisation on a flat surface gives rise to a bimodal distribution. However, for the intended use of the nebulised droplets, the second peak (larger drops) is not wanted. In a preferred embodiment of the invention, cavities are formed (cylindrical holes in a superstrate to be coupled to the SAW transducer) having dimensions that prevent the capillary wave instability, in order to suppress or avoid the formation of the larger droplets. In FIG. 2(d), the second peak is centred on about 40 μm. Therefore the diameter of the cavities can be controlled to be less than 40 μm. This indeed will show a performance in which the larger droplets are suppressed. The inventors also report that the drop formation requires deformations of the surface that are of a scale larger than the drop size. FIG. 4 for example shows that cavities of diameter 80 μm still prevent the bigger drops. Further results (not shown) have demonstrated that cavities up to 200 um in diameter can also prevent these secondary drops.

(49) In the schematic view shown in FIGS. 1A and 1B, the cavities 16 have a shape which is substantially hemispherical. This is intended to be illustrative. Hemispherical cavities can be used. However, more generally, cavities of other shapes can be used, e.g. cylindrical cavities, rectangular or square cylindrical cavities, circular cylindrical cavities, etc. Such cavities can have closed bottom ends. The bottoms of such cavities can be flat or rounded. In the case of etched cavities, some rounding of the bottoms is typical.

(50) In other embodiments, the cavities may have more complex internal structures. Examples are shown in FIGS. 5-10.

(51) FIG. 5 shows a schematic cross sectional view of a single cavity 40. The cavity has a closed bottom 42, internal walls 44, 46 and an upstanding pillar 48. Pillar 48 is supported on the closed bottom 42. In the cavity of FIG. 5, it is intended that the plan view shape of the cavity is a square, with the pillar formed at the centre. In alternative embodiments, the plan view shape of the cavity may be rectangular, other polygonal shape, round or circular. In those cases, it is possible for the pillar to be located at the geometrical centre of the shape, or located off-centre.

(52) FIG. 6 shows a schematic cross sectional view of a single cavity 60 which is a modification of the cavity shown in FIG. 5. Here, the internal walls 64, 66 and the pillar 68 have an array of projections 67. The projections are arranged based on a periodic arrangement with the intention of interacting phononically with SAWs and affecting the transmission, distribution or other properties of the SAWs in the cavity. In this way, a phononic structure 69 is formed.

(53) FIG. 7 shows a schematic cross sectional view of a single cavity 80 which is another modification of the cavity shown in FIG. 5. The cavity has an open bottom 82, internal walls 84, 86 and an upstanding pillar 88. Since there is no closed bottom to support pillar 88, it is supported by an arrangement of struts 90 extending from the internal walls 84, 86. As for the cavity of FIG. 5, it is intended that the plan view shape of the cavity is a square, with the pillar formed at the centre. In alternative embodiments, the plan view shape of the cavity may be rectangular, other polygonal shape, round or circular. In those cases, it is possible for the pillar to be located at the geometrical centre of the shape, or located off-centre.

(54) FIG. 8 shows a plan view of the cavity of FIG. 7.

(55) FIG. 9 shows a schematic cross sectional view of a single cavity 100 which is a modification of the cavity shown in FIG. 8. Here, the internal walls 104, 106 and the pillar 108 have an array of projections 107. The projections are arranged based on a periodic arrangement with the intention of interacting phononically with SAWs and affecting the transmission, distribution or other properties of the SAWs in the cavity. In this way, a phononic structure 109 is formed. Pillar 108 is supported by struts 110.

(56) FIG. 10 shows a plan view of the cavity of FIG. 9.

(57) The use of complex cavity structures allows the interaction of the SAWs with the liquid to be controlled further. This is achieved by consideration of interaction of the fluid with the additional structures and by consideration of the interaction of the SAWs with the additional structures.

(58) Additional investigation has been carried out to assess the effect of cavity size (also referred to herein as pore size) on aerosol droplet size. Cavities of different diameter were etched into silicon superstrates. The cavities were etched cylindrical pits with a closed bottom end, approximately 300 μm deep. Using blind cavities in this way did not allow a continuous feeding of the cavities with liquid. As a result, for each experiment, only a small volume could be nebulised at a time. For this reason, rather than carrying out an analysis of the particle size distribution based on light diffraction, as reported above, here the results are reported based on a visual observation of a small number of drops in the nebulised plumes (based on recorded microvideograph footage of the nebulised plumes).

(59) Initially, a drop of water was placed on top of each superstrate and SAW was applied until the top layer of water disappears (either evaporate or nebulised), leaving water only in the cavities without there being liquid communication between the cavities (no water present on the surface of the superstrate between the cavities). Nebulisation from the cavities was then monitored using a fast camera (>250 kfps) fixed to a microscope, enabling the recording of images.

(60) The largest droplets nebulised from the pores were visually assessed for their diameters.

(61) The results are shown in FIG. 11. FIGS. 12, 13 and 14 show images from the recorded footage, with FIG. 12 showing a superstrate with cavities of diameter 80 μm, FIG. 13 showing cavities of diameter 600 μm and FIG. 14 showing cavities of diameter 1500 μm. For each, the SAW was excited at 13.93 MHz and −5 dBM. The series of images was analysed and largest droplets were measured. Note that to estimate droplet size from the 80 μm cavities, the thickness of the plume were divided by the number of droplets (3 to 4 droplets) and for holes with 600 and 1500 μm diameter, single droplets were measured. For the 80 μm cavities, this was because the plume was so condensed that no single droplet could be measured.

(62) FIG. 11 shows the variation in droplet size with cavity diameter. This shows an increase in droplet size as the cavity diameter is increased. However, the change in droplet size is not as significant as expected, but this is likely to be due to the measurement approach. Measurement of the droplet diameter distribution using light scattering in a continuous plume would demonstrate a shift in droplet size distributions, in which for small cavities (80 μm diameter), only small drops are present (<10 μm), whereas for larger cavities (1500 μm diameter), small drops are still seen, but other modes also exist to provide droplets also of large diameter (about 20-50 μm mean size). The results reported here indicate an effect attributable to the cavity diameter and correlate with the theory outlined, which is that the pinned surface layer within pores is ruled by the meniscus curvature (i.e. contact angle with pore wall). This results in a surface rigidity, suppressing capillary waves.

(63) While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention.

(64) All references referred to above are hereby incorporated by reference.

LIST OF NON-PATENT LITERATURE REFERENCES

(65) P. P. H. L. Brun, A. H. de Boer, H. W. Frijlink, and H. G. M. Heijerman, Pharm. World Sci., 2000, 22, 75-81.

(66) J. Boe, J. H. Dennis, B. R. O'Driscoll, T. T. Bauer, M. Carone, B. Dautzenberg, P. Diot, K. Heslop, and L. Lannefors, Eur. Respir. J., 2001, 18, 228-242.

(67) M. B. Dolovich and R. Dhand, The Lancet, 2011, 377, 1032-1045.

(68) J. Reboud, R. Wilson, Y. Zhang, M. H. Ismail, Y. Bourquin, and J. M. Cooper, Lab. Chip, 2012, 12, 1268-1273.

(69) J. Reboud, Y. Bourquin, R. Wilson, G. S. Pall, M. Jiwaji, A. R. Pitt, A. Graham, A. P. Waters, and J. M. Cooper, Proc. Natl. Acad. Sci., 2012, 109, 15162-15167.

(70) A. Qi, L. Y. Yeo, and J. R. Friend, Phys. Fluids, 2008, 20, 074103.

(71) M. Kurosawa, T. Watanabe, A. Futami, and T. Higuchi, Sens. Actuators Phys., 1995, 50, 69-74.

(72) M. Kurosawa, A. Futami, and T. Higuchi, in Solid State Sensors and Actuators, 1997. TRANSDUCERS '97 Chicago., 1997 International Conference on, 1997, vol. 2, pp. 801-804.

(73) M. Alvarez, J. Friend, L. Yeo, and D. Arifin, in 16th Australasian Fluid Mechanics Conference, 2007, pp. 621-624.

(74) J. Ju, Y. Yamagata, H. Ohmori, and T. Higuchi, Sens. Actuators Phys., 2008, 145-146, 437-441.

(75) J. Ju, Y. Yamagata, K. Inoue, and T. Higuchi, in Service Robotics and Mechatronics, eds. K. Shirase and S. Aoyagi, Springer London, 2010, pp. 309-312.

(76) M. N. Topp, J. Aerosol Sci., 1973, 4, 17-25.

(77) A. Qi, J. R. Friend, L. Y. Yeo, D. A. V. Morton, M. P. McIntosh, and L. Spiccia, Lab. Chip, 2009, 9, 2184.

(78) N. Maehara, S. Ueha, and E. Mori, Rev. Sci. Instrum., 1986, 57, 2870-2876.

(79) Wu, T. & Chang, I., 2005. Actuating and detecting of microdroplet using slanted finger interdigital transducers. Journal of Applied Physics, 98(2), 024903-7

(80) H. Eixarch, E. Haltner-Ukomadu, C. Beisswenger and U. Bock, Drug Delivery to the Lung: Permeability and Physicochemical Characteristics of Drugs as the Basis for a Pulmonary Biopharmaceutical Classification System (pBCS), Journal of Epithelial Biology & Pharmacology, 2010, 3, 1-14.

(81) J. Blamey, L. Y. Yeo, and J. R. Friend, Microscale Capillary Wave Turbulence Excited by High Frequency Vibration, Langmuir 2013, 29, 3835-3845.