Method for coating microstructured components

Abstract

The present disclosure provides a method for the surface modification of microstructured components having a polar surface, in particular for high-pressure applications. According to the method, a microstructured component is contacted, in particular treated, with a modification reagent, wherein the surface properties of the component are modified by chemical and/or physical interaction of the component surface and of the modification reagent.

Claims

1. A discharge apparatus for fluids comprising a microstructured component comprising a surface modification obtained by contacting the surface of the microstructured component with a modification reagent, thereby modifying the properties of the surface by the chemical and/or physical interaction between the microstructured component surface and the modification reagent; wherein the modification reagent comprises at least one modifier, wherein the modifier is a silane of the general formula (III):
R.sub.4-nSiX.sub.n wherein R is C.sub.10-C.sub.16 alkyl, X is selected from C.sub.1 alkoxy and C.sub.2 alkoxy, and n is 3; and wherein the at least one modifier is at a concentration of from 0.001 to 2 mol/L based on the modification reagent.

2. The discharge apparatus according to claim 1, comprising at least one liquid medicinal product.

3. The discharge apparatus according to claim 2, wherein the liquid medicinal product is a dispersion or solution of at least one pharmaceutical active ingredient.

4. A microstructured component for use in a microfluidic system, comprising an outer surface, at least one inlet opening, at least one outlet opening, and inner surfaces formed by microstructures, wherein the inner surfaces are modified, at least in part, with a silane modifier of general formula (III):
R.sub.4-nSiX.sub.n wherein R is C.sub.10-C.sub.16 alkyl, X is selected from C.sub.1 alkoxy and C.sub.2 alkoxy, and n is 3.

5. The microstructured component according to claim 4, wherein the outer surface of the component is modified.

6. The microstructured component according to claim 4, wherein the inner surface of the component is modified to be rendered hydrophobic.

7. A microstructured component according to claim 4, wherein the silane is selected from the group consisting of C.sub.12 alkyltrialkoxysilanes, C.sub.14 alkyltrialkoxysilanes and C.sub.16 alkyltrialkoxysilanes, and mixtures thereof.

8. The microstructured component according to claim 4, wherein the inner surface of the microstructured component is activated before modification.

9. The microstructured component according to claim 4, wherein the microstructured component comprises two outlet openings and the microstructured component has channels which connect the outlet openings directly or indirectly with the at least one inlet opening.

10. The microstructured component according to claim 9, wherein the channels of the outlet openings are oriented towards one another at an angle of from 50° to 130°.

11. The microstructured component according to claim 4, wherein the microstructured component further comprises a filter region between the inlet opening and the outlet opening.

12. The microstructured component according to claim 4, wherein the component consists of two different materials.

13. The microstructured component according to claim 12, wherein the component consists of glass and silicon wherein the inner surfaces of the component are modified, at least in part, with a silane modifier of general formula (III).

14. The microstructured component according to claim 12, wherein at least one of the materials is microstructured on a side on which the material is bonded to a second material and thus there is a microstructure located within the component as a result of the bonding between the different materials.

Description

(1) Additional advantages, features, properties and aspects of the present invention are set out in the claims, the embodiments and the following description given on the basis of the drawings, in which:

(2) FIG. 1 is a schematic view of a microstructured component showing an illustration of the microstructure including the outlet openings and the channels leading to the outlet openings,

(3) FIG. 2 shows a cut-out of the microstructured component shown in FIG. 1, in which the region of the microstructured component around the outlet openings is shown in detail and the coating is visible,

(4) FIG. 3 is a schematic section through a discharge apparatus, in particular an atomiser, in the non-tensioned state,

(5) FIG. 4 is a schematic section through the discharge apparatus, in particular the atomiser, from FIG. 1 in the tensioned state, rotated through 90° compared with FIG. 3,

(6) FIG. 5 shows an example group I spray pattern (image in two different perspectives) for the tested inhalers comprising a DJI nozzle,

(7) FIG. 6 shows an example group II spray pattern (image in two different perspectives) for the tested inhalers comprising a DJI nozzle,

(8) FIG. 7 shows the change over time in group III spray patterns for inhalers comprising a DJI nozzle that has uncoated nozzle bodies, according to pH,

(9) FIG. 8 shows the 10-day average rate of group III spray patterns for inhalers comprising a DJI nozzle that has uncoated nozzle bodies, according to pH,

(10) FIG. 9 shows the change over time in group I spray patterns for inhalers comprising a DJI nozzle that has coated nozzle bodies, according to the modifier,

(11) FIG. 10 shows the change over time in group II spray patterns for inhalers comprising a DJI nozzle that has coated nozzle bodies, according to the modifier,

(12) FIG. 11 shows the change over time in group III spray patterns for inhalers comprising a DJI nozzle that has coated nozzle bodies, according to the modifier,

(13) FIG. 12 shows the 10-day average rate of group III spray patterns for inhalers comprising a DJI nozzle that has coated nozzle bodies, according to the modifier,

(14) FIG. 13 shows the change over time during long-duration tests in group I spray patterns for inhalers comprising a DJI nozzle that has nozzle bodies coated with 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane (F1308-OET) compared with inhalers having uncoated nozzle bodies,

(15) FIG. 14 shows the change over time during long-duration tests in group II spray patterns for inhalers having a DJI nozzle that has nozzle bodies coated with 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane (F1308-OET) compared with inhalers having uncoated nozzle bodies,

(16) FIG. 15 shows the change over time during long-duration tests in group III spray patterns for inhalers having a DJI nozzle that has nozzle bodies coated with 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane (F1308-OET) compared with inhalers having uncoated nozzle bodies,

(17) FIG. 16 shows the 10-day average during long-duration tests of group III spray patterns for inhalers having a DJI nozzle that has nozzle bodies coated with 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane (F1308-OET) compared with inhalers having uncoated nozzle bodies,

(18) FIG. 17 shows the change over time in group I spray patterns for inhalers having a DJI nozzle that has coated nozzle bodies, for other modifiers,

(19) FIG. 18 shows the change over time in group II spray patterns for inhalers having a DJI nozzle that has coated nozzle bodies, for other modifiers,

(20) FIG. 19 shows the change over time in group III spray patterns for inhalers having a DJI nozzle that has coated nozzle bodies, for other modifiers,

(21) FIG. 20 shows the 10-day average rate of group III spray patterns for inhalers having a DJI nozzle that has coated nozzle bodies, for other modifiers,

(22) FIG. 21 shows the yield of category I nozzle bodies depending on the selected modifier.

(23) In the drawings, the same reference numerals are used for like or similar parts, matching or similar properties and advantages being obtained even is the description is not repeated.

(24) FIG. 1 is a schematic view of a microstructured component 1 according to the invention, in particular a nozzle body. The illustration in FIG. 1 shows the microstructure of a nozzle body comprising a DJI-type nozzle, as preferably used in an SMI-type inhaler. The nozzle body is preferably a microfluidic sandwich system and consists of a square microstructured silicon chip that is bonded to a 0.5 mm thick, preferably very smooth glass plate (e.g. a glass produced using the “float” method, preferably based on borosilicate).

(25) The microstructured component 1 consists of two rigidly interconnected plate-like materials, preferably a silicon wafer and a glass wafer. The component 1 has inlet openings 2 and outlet openings 3 for receiving or discharging preferably pressurised fluids, preferably liquids. Channels 4 that are preferably oriented towards one another adjoin the outlet openings. The channels 4 and outlet openings 3 have either a round or non-round shape, in particular preferably an angular cross section of a diameter or depth of from 2 to 10 μm and a width of from 5 to 15 μm, in particular a depth of from 4.5 to 6.5 μm and a width of from 7 to 9 μm.

(26) The microstructures in the silicon chip are preferably produced using etching techniques. The etching depth on the silicon chip can vary depending on the solvent or dispersion medium or method used. Preferably, the depth is 5.6 μm for nozzle bodies intended for atomising aqueous formulations, or 7.0 μm for those intended for atomising ethanolic formulations, as determined by the different physico-chemical properties of the solutions or dispersions. The overall cross-sectional surface area of the outlet openings 3 is usually from 30 to 500 μm, a cross section range of from 30 to 200 μm being preferred. The outlet channels 4 preferably have a length of 40 μm and a width of 8 μm.

(27) In a component 1 having at least two outlet openings 3, the jet directions can be inclined relative to one another at an angle of from 50 to 130°; preferably, an angle of from 70 to 110°, from 85 to 95°, or most preferably of 90°, is obtained. The outlet openings 3 are generally spaced apart by 10 to 200 μm, in particular by 10 to 100 μm, preferably by 30 to 70 μm. Preferably, the spacing between the outlet openings 3 is 50 μm. The jet directions meet one another close to the nozzle openings (preferably at a distance of less than 1 mm, preferably of less than 100 μm, from the surface of the nozzle body) and the fluid is atomised by the liquid jets colliding.

(28) In the following, a preferred embodiment will be described in which the microstructures defined by the inlet openings 2, the outlet openings 3 and the channels 4 are made in the surface of a silicon wafer. The microstructures can be made in the silicon wafer by any suitable method; however, the microstructuring is preferably made by etching methods, as known for example from semiconductor technology. To create surfaces within the microstructured component 1 that are suitable for receiving and subsequently dispensing pressurised fluids, the microstructures are made in one of the materials, preferably the silicon wafer, of the component and connected to the second component, preferably a glass wafer. In this way, cavities in the form of microstructures can be obtained in the microstructured component.

(29) A fine filter 5 consisting of a multiplicity of filter channels can be arranged between the outlet openings 3 or channels 4. The passages or filter channels in the filter structure within the fine filter 5 are selected such that, as far as possible, even the smallest of impurities cannot enter the region of the outlet openings, i.e. the nozzle region, and clog them or change the geometry. The diameter of the filter channels or filter passages is usually from 0.5 to 20 μm, preferably from 2 to 5 μm. The fine filter 5 usually has a zigzag arrangement to increase the surface area. Particles that are larger than the cross sections of the filter channels can be retained in the fine filter 5 to prevent the outlet channels becoming clogged. The filter channels in the filters of the fine filter 5 are formed by protrusions that are preferably arranged in a zigzag shape so as to increase the filter surface area. For example, at least two rows of protrusions form a zigzag configuration. A plurality of rows of protrusions can also be formed, the rows preferably abutting one another at acute angles and forming the zigzag configuration. In embodiments such as this, the inlet and outlet can comprise an inflow region for unfiltered fluid and an outflow region for filtered fluid, respectively, the inflow region and outflow region being substantially exactly as wide as the filter 5 and substantially as high as the protrusions for the inlet and outlet sides of the filter 5. The zigzag configuration formed by at least two rows of protrusions preferably has a tilt angle of from 20 to 250°. Further details on this component design can be found in WO 94/07607 A1.

(30) The microstructured component 1 can comprise an outflow region or a plenary chamber 6. In particular, the plenary chamber 6 is arranged between the outlet openings 3 or channels 4 and the fine filter 5. The plenary chamber can comprise column structures 7 depending on the application. By means of the column structures 7 in the plenary chamber 6, a multiplicity of channels are produced and preferably run into the channels 4 of the outlet openings 3.

(31) The microstructured component 1 or nozzle body forms a rigid system designed to make two liquid jets impact against one another once they exit the outlet openings 3. When the impact is correct, an impact disc forms, at the boundary of which the fine aerosol is produced. The critical parameters for aerosol formation include, inter alia, the flow rate (around 100 m/s) and the angle of impact. Material deposited in the nozzle channels can thus noticeably affect the aerosol formation, for example by deflecting the jet, and cause “spray anomalies”, which could even prevent the spray cloud from appearing due to jet divergency.

(32) The surfaces of the microstructured component 1 have a coating 8 which determines the surface properties of the microstructured component. In this regard, both the inner and outer surfaces of the microstructured component 1 can be coated. Preferably, at least the inner surfaces of the microstructured component are coated in order to prevent particles adhering thereto and to thus prevent the nozzle becoming blocked or clogged.

(33) According to a preferred embodiment of the present invention, the outer surface of the component 1 is also modified or coated at least in the region of the outlet openings.

(34) FIGS. 3 and 4 are schematic views of a discharge apparatus according to the invention in the form of a manually operated medical device. The discharge apparatus according to FIGS. 3 and 4 is preferably a propellant-free atomiser 9 that discharges predetermined amounts of a liquid or a liquid medicinal formulation as a preferably respirable or inhalable aerosol 11 per actuation cycle. Aerosol droplets having an aerodynamic diameter of from 0.5 to 5 μm can be inhaled by a user. The average aerodynamic droplet size of the aerosol 11 is preferably within a diameter range of from 0.5 to 10 μm, in particular in the range of from 0.5 to 5 μm.

(35) For the atomisation, a suitable nozzle in the form of the microstructured component 1 according to the invention is used. When operating the atomiser 9, a distinction is drawn between the “non-tensioned state”, where the metering volume in the pressure chamber 12 is empty (FIG. 3), and the “tensioned state”, where the pressure chamber 12 is full (FIG. 4). When the atomiser 9 is tensioned, the upper housing part 13 is rotated by a fixed angle of rotation, e.g. 180°, relative to the inner housing part 14 and the lower housing part 15. By means of an internal screw thread mechanism, the relative rotation drives a plunger pump such that a predetermined, optionally adjustable amount of liquid 10 is conveyed out of the container 16 into the pressure chamber 12 while the mainspring 17 of the pressure generator 18 is tensioned at the same time (the final state of the tensioning process is shown in FIG. 4). When the atomiser 9 is triggered, i.e. when a locking ring 19 is actuated by means of a button 20, the energy from the pressure generator 18 stored in the mainspring 17 is released. The tubular piston 21 used previously to convey the liquid then pushes into the pressure chamber 12 while the return valve 22 is closed, such that the amount of liquid predetermined by the stroke movement of the tubular piston 21 is discharged from said chamber through the outlet opening 3. The apparatus is now back in the non-tensioned state, as shown in FIG. 3.

(36) In the embodiment shown, by means of the container 16, the tubular piston 21 is rigidly connected, e.g. integrally moulded, glued or snapped on, to a mount 23 belonging to the pressure generator 18. The container 16 is secured, in particular clamped or latched, in the atomiser 9 by means of the mount 23 such that the tubular piston 21 enters the fluid chamber of the container 16 and/or is in fluid communication with the liquid 10 in the container 16 and said liquid can be sucked up via the tubular piston. The container can optionally be replaceable. For this purpose, the device housing can be designed such that it can be opened or partly removed (e.g. in the form of a cap-like lower housing part as disclosed in WO 07/123381 A1).

(37) The container 16 used in the atomiser 9, which is equipped with a dose indicator or meter 24, is designed for removing a plurality of dosage units. For this purpose, the container has to be designed such that the internal pressure remains substantially the same, even during liquid removal, to ensure the same amount of liquid 10 is always removed during suction. For this purpose, it is possible in principle to use both a container 16 that comprise a rigid container wall, the internal pressure of which is kept constant by means of ventilation and which is in turn described for example in WO 06/136426 A1, and a container 16 having a flexible wall that is slid into the container interior at least in part when liquid is removed in such a way that the reduction in the internal volume keeps the internal pressure constant.

(38) Containers 16 in which the flexible wall is formed by a substantially deformable, collapsible and/or contractable pouch are preferred in this case. Various embodiments of containers of this kind are described in documents WO 00/49988 A2, WO 01/076849 A1, WO 99/43571 A1, WO 09/115200 A1 and WO 09/103510 A1. Particularly preferably, the container 16 consists of a flexible multi-layer film pouch that is closed at the bottom and is directly connected in its upper region to a supporting flange, preferably made of plastics material, a container cap welded thereon for connecting to the mount 23 of the atomiser 9, an outer protective sleeve and a top seal (for details see WO 99/43571 A1 and WO 09/115200 A1). The typical filling volume of a container 16 consists of from 3.0 to 3.6 ml of inhalation solution.

(39) A filter system 25 upstream of the microstructured component 1 can be located in the liquid outlet region of the pressure chamber 12. This filter system 25 preferably consists of a plurality of filter components that are arranged one behind the other and differ from one another in particular on account of the filter technology used. The filter thresholds of the individual filter components are of such a level that each filter lets through smaller particles than the one behind it in accordance with the largest exchange principle. By combining different filter techniques and arranging filters to have a gradually increasing degree of separation or gradually decreasing pore sizes, it is possible to achieve a high filter capacity, i.e. the precipitation of relatively large amounts of particles without the filter becoming blocked, and thorough filtering. In addition to collecting solid particles of a particular size, the filter can optionally collect additional material via adsorption. For this purpose, filters of different structures and different materials can be used, such that the adsorption properties are different from filter to filter. Accordingly, combining different filters makes it possible to catch even more particles and in particular particles that can deform under pressure owing to the various adsorption effects. For further details on preferred filter systems, reference is made in particular to WO 2012/007315.

(40) The entire system consisting of the pressure generator 18, having the mainspring 17, and the microstructured component 1 is preferably constructed such that, when the spray mist is produced, not only are respirable droplet sizes formed, but also the spray mist cloud itself remains there for enough time to allow the patient to adapt their inhalation thereto in a simple manner. Preferably, spray times are from 0.5 to 2 s, in particular from 1 to 2 s. The pressure at which the fluid, in particular the liquid drug, leaves the outlet openings is from 50 to 1,000 bar, in particular from 200 to 600 bar.

(41) Depending on the size of the components and the liquid formulation to be atomised, the aerosol 11 contains a distinctively high fine particle fraction (in this case: the fraction of the spray made up by particles having diameters of 5 μm) of for example ≥50%, preferably ≥65%, particularly preferably of 80%, in particular for ethanolic formulations, and the spray cloud produced is preferably slower than in other portable inhalers, e.g. MDI-type inhalers. This leads to significantly higher deposition in the lungs than in other conventional inhalers, such as pMDIs or DPIs. In addition, the atomiser 9 according to the invention is distinguished by a remarkably long duration of spray, which enables good patient coordination in terms of them triggering the atomiser 9 or inhaler.

(42) Depending on the required daily dosage and the intended period of application, the atomiser 9 can be designed to dispense from 10 to 200, in particular from 20 to 150, preferably from 60 to 130 sprays. A slide on the meter 24 indicates how many strokes have been consumed or how many are left. After the specific stroke number is reached, the discharge apparatus preferably locks itself automatically and is blocked from being used further. A “tail-off”, as can be noted in metering aerosols that use compressed air, is thus prevented.

(43) To prepare the atomiser 9 for application, a container 16, in particular in the form of a cartridge, must first be inserted. For this purpose, the lower housing part 15 has to be removed. After the container has been inserted, the removed lower housing part 15 is placed back on and the device is primed by being actuated a number of times (=discharging the air from the system). Only after this time is the atomiser 9 ready for operation and can guarantee constant dispensing of a dosage.

(44) The aim of the priming is to completely fill the metering chamber or pressure chamber 12. When the device is actuated while the mouthpiece is oriented vertically upwards, the lower housing part 15 is rotated towards the upper housing part 18 by 180° until the audible clicking and latching. In the process, the mainspring 17 is tensioned, the button 20 springs forwards when latching is complete and indicates that the atomiser 9 is in the tensioned state by sitting flush with the sides. Pressing the button 20 generates the aerosol; the position of the atomiser 9 can be freely selected.

Embodiments

(45) 1. Methods Used and Experiment Set-Up

(46) For the following tests, the surfaces of planar substrates made of silicon and glass, and a microstructured nozzle system were modified and their properties studied. Commercially available glass planar substrates produced from borosilicate glass using the “float” method were primarily used as planar substrates for the tests described below. Nozzle bodies having microstructures according to the drawing in FIG. 1 were used as the microstructured nozzle system.

(47) Si/glass planar substrates consist of the same materials as the nozzle bodies and are cut to size from the nozzle body starting materials, i.e. a glass wafer (borosilicate glass having a smooth surface according to the “float” method) and a silicon wafer (111). Their size is that of a conventional microscope slide (26 mm×75 mm) and they are used as a reference material since the nozzle body used is difficult to access for many characterisation tests.

(48) The nozzle body used for the following tests is a microfluidic sandwich system, consisting of a 2.05×2.55 mm.sup.2 microstructured silicon chip bonded to a 0.5 mm-thick glass plate (in this case a borosilicate glass produced using the “float” method). The internal microstructure of the nozzle body used consists of an inlet region having an inlet opening, a zigzagged fine filter, a columnar microstructure 7 and the front nozzle region. The etching depth on the silicon chip is 7.0 μm. The distance between the two outlet openings 3 is 50 μm.

(49) During the triggering, a liquid solution or dispersion flows through the inlet region into the microstructure under very high pressure. In the zigzagged fine filter, the fine filter structures having an opening width of 3 μm retain relatively large particles to prevent the outlet channels becoming clogged by such relatively large particles. The outlet channels have a length of 40 μm and a width of 8 μm. The inhalation solution leaves the nozzle body through the two front nozzle outlet channels. The aerosol is generated outside the nozzle body by the two liquid jets produced impacting against one another.

(50) The structure of the nozzle body forms a rigid system that ensures the two liquid jets impact against one another correctly. After having exited the nozzle channels, the two liquid jets collide at an angle of 90° to each other. When the impact is correct, an impact disc forms, at the boundary surface of which the fine aerosol is produced. The critical parameters for aerosol formation include the flow rate (around 100 m/s) and the angle of impact.

(51) As part of the tests described here, when fitted the functioning of the surface-modified nozzle bodies is checked on the basis of SMI atomisers as designed according to FIGS. 3 and 4.

(52) 1.1. Preparing Surface-Functionalised Nozzle Bodies and Si/Glass Planar Substrates by Means of Solution Coating

(53) The coating process begins by activating the silicon or glass surfaces. The activation cleans the surface of adherent dirt and increases the number of free reactive silanol groups on the surface by means of oxidation. The reactive silanol groups on the surface are capable of reacting with functional silanes. Tests were carried out on modification reagents based on various chloroalkylsilanes and alkylalkoxysilanes, which form a robust siloxane scaffold on the surface following the reaction.

(54) The quality of the surface coating on Si/glass planar substrates is determined by means of contact angle measurement and layer thickness measurement using spectroscopic ellipsometry. For the microstructure or nozzle body, the capillary action test can be carried out.

(55) 1.1.1 Preparing the Nozzle Body Sample

(56) The nozzle bodies are functionalised in PEEK trays. PEEK (polyether ether ketone) is a very inert material and is exceptionally suitable for this application. Nozzle bodies have to be functionalised in trays because the very small nozzle bodies would otherwise not be able to be handled in the process tanks. The tray has a reaction chamber that has space for around 50 to 75 nozzle bodies and can be closed by means of a lid. In addition, an agitator that ensures the chamber is thoroughly mixed can be positioned below the reaction chamber. Both the lid and the reaction chamber are provided with small pores, such that solution can flow through the chamber.

(57) Before coating, the nozzle bodies are transported into the reaction chamber by means of a vacuum cup. Since the nozzle bodies have already been cleaned, no special cleaning steps have to be carried out on the component, as is the case for silicon or glass planar substrates.

(58) 1.1.2 Preparing the Planar Substrate Samples

(59) The silicon and glass planar substrates do not undergo the cleaning steps like the nozzle bodies, so they have to be thoroughly cleaned again using water and isopropanol prior to being activated. The substrates are cut to the conventional slide dimensions and are also coated in an inert tray having space for around 20 substrates. The Si substrates must have been stripped as close to the functionalisation as possible, i.e. should undergo an HF treatment; this ensures a uniform SiO.sub.2 film on the surface. When the silicon substrates are stripped, the natural oxide layer on the surface is removed. After this removal, the oxide film regrows slowly and generates a uniform thin oxide film.

(60) For the stripping, a 6% hydrogen fluoride solution is used, in which the silicon substrates should remain for at least 20 minutes. Following the removal therefrom, the Si substrates are thoroughly rinsed with water and sorted into the substrate trays for coating.

(61) The glass substrates used have a fire-polished side and an extra-smooth side produced in the float method. To improve the reproducibility of the measurement results in the tests described here, the fire-polished side of the glass substrates was marked using a diamond scriber before the start of the process. This is the side on which the ellipsometric measurements will be taken later in the process.

(62) 1.1.3 Overview of the Surface Functionalisation Process Steps

(63) The process begins by cleaning the surfaces of adherent dirt. Next, the cleaned surface is oxidatively activated. After a rinsing step in which the activation solution is washed off the nozzle bodies or substrates, the actual coating process can begin. If the coating is carried is carried out using an alkyltrialkoxysilane, the coating reagent has to be transformed into the active silanol component by means of precursor splitting. After coating, the samples are dried and then dried in an oven at 120° C.

(64) 1.1.3.1 Cleaning and Activating Nozzle Bodies and Si/Glass Planar Substrates

(65) The coating process usually begins by cleaning and activating the silicon or glass surfaces. The samples are cleaned and activated in one operation. The following cleaning and activation solutions are used for nozzle bodies and Si/glass planar substrates in the following tests.

(66) 1.1.3.1.1 NaOH Solution (Highly Alkaline Activation)

(67) The samples are added to 25% caustic soda and stirred. Next, they are thoroughly rinsed with lots of water. In most cases, the method works very well, but relatively often leads to clouding on the glass or silicon surface since the surface has already been etched (glass corrosion). In the following, therefore, this method is only used for preliminary tests since clouding on the nozzle body is unacceptable due to the visual quality control.

(68) 1.1.3.1.2 RCA Solution (RCA=Radio Corporation of America)

(69) The next activation solution, referred to as RCA solution, is Standard Clean 1 solution (SC-1 solution), which was developed for the RCA method in order to clean silicon wafers. The samples are added to a solution of H.sub.2O:NH.sub.3 (aq. 25%):H.sub.2O.sub.2 (aq. 30%) in a volume-to-volume ratio of 5:1:1 at 70° C. and stirred. The Si/glass substrates are stirred constantly for 20 minutes, whereas nozzle bodies are treated for one hour alternating between treatment with ultrasound (20 minutes each time) and stirring (10 minutes each time). Next, the substrates or nozzle bodies are thoroughly rinsed with water. The substrates or nozzle bodies are stored in water until the start of the process.

(70) 1.1.3.1.3 Piranha Solution (Acidic Oxidative Activation)

(71) The samples are added to a solution of H.sub.2SO.sub.4 (conc.):H.sub.2O.sub.2 (aq. 30%) in a volume-to-volume ratio of 7:3 at 70° C. and stirred. The Si/glass substrates are stirred constantly for 20 minutes, whereas nozzle bodies are treated for one hour alternating between treatment with ultrasonic waves (20 minutes each time) and stirring (10 minutes each time). Next, the samples are thoroughly rinsed with water and can be stored in water until coating.

(72) 1.1.3.2 Surface Functionalisation of Silicon and Glass Surfaces Using Alkyltrialkoxysilanes

(73) The trialkoxysilanes are added to an alcoholic solution and, since they are a precursor compound, have to first be proteolytically split into the active silanol components. By way of example, this reaction can be taken from equation 1 below. The silanol compound is usually produced within five hours, as indicated in equation 1 for the example 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane (example substance used: Dynasylan® F8261 from Evonik) from the starting compound.

(74) ##STR00001##

(75) The active silanol components now associate with the activated surface and bond covalently following tempering.

(76) During the functionalisation, the solution is stirred constantly. Unlike the substrates, nozzle bodies are treated with ultrasound at fixed intervals during the functionalisation. Following functionalisation, the samples are removed from the solution and dried for one hour in air. After drying, the samples are tempered in an oven for 1 hour at 120° C. Following tempering, the substrates are carefully rinsed using a small amount of isopropanol. Nozzle bodies can also be used without any additional cleaning step.

(77) 1.1.3.2.1 Surface Coating Process Using Alkyltrialkoxysilanes in an Isopropanol-Water Mixture

(78) 1.1.3.2.1.1 Hydrolysis of Starting Compound

(79) A 0.1-1.0 vol. % solution of the silane is added to an acidic alcohol-water mixture (2-propanol/H.sub.2O/HCl.sub.conc 89.8:10:0.2) and stirred at room temperature for at least five hours.

(80) The compound can be used after five hours and must be used up within 24 hours.

(81) 1.1.3.2.1.1 Surface Functionalisation

(82) The samples are added to the functionalisation solution and constantly stirred during coating. Nozzle bodies are repeatedly treated with ultrasound at fixed intervals during the coating. After functionalisation, the samples are removed from the liquid, and the tray for the substrates and the tray for the nozzle bodies are passed to a cellulose cloth for drying. The samples are dried for one hour in air.

(83) 1.1.3.2.1.2 Condensation

(84) The samples are tempered in an oven at 120° C. Next, the substrates are carefully washed using isopropanol. Nozzle bodies can also be used after tempering without any additional cleaning step. The finished nozzle bodies and substrates can be stored under laboratory conditions.

(85) 1.1.3.3 Surface Functionalisation of Silicon and Glass Surfaces Using

(86) Chloroalkylsilanes Surface functionalisation by means of chloroalkylsilanes is based on the condensation reaction between the free silanol group on the surface and the chlorine function of the functional alkylsilane.

(87) Chlorosilanes are very sensitive to moisture and their reactions must be carried out in dry, non-polar solvents such as toluene, tetrachloromethane or alkanes such as hexane.

(88) 1.1.3.3.1 Schematic Surface Coating Process Using Alkylchlorosilanes in Toluene

(89) The surface of substrates and nozzle bodies are activated as described above.

(90) 1.1.3.3.1.1 Drying the Solvent

(91) The toluene is dried for at least 12 hours using a molecular sieve (type 4 A). The ratio is 10 g molecular sieve to 1 litre solvent.

(92) 1.1.3.3.1.2 Surface Functionalisation

(93) The samples are coated in 0.07 to 0.7 mol solutions of the chlorosilane while being stirred constantly. Nozzle bodies are repeatedly treated with ultrasound at fixed intervals during the coating. After functionalisation, the samples are removed from the liquid, and the tray for the substrates and the tray for the nozzle bodies are passed to a cellulose cloth for drying. The samples are dried for one hour in air.

(94) 1.1.3.3.1.3 Condensation

(95) After drying, the samples are tempered in an oven at 120° C. Next, the substrates are carefully washed using isopropanol. Nozzle bodies can also be used after tempering without any additional cleaning step. The finished nozzle bodies and substrates can be stored under laboratory conditions.

(96) 1.2 Categorising Nozzle Bodies after Coating

(97) The coated nozzle bodies are divided into the following nozzle categories I to IV.

(98) Category I

(99) The nozzle region, the zigzagged filter and the support structure of the nozzle body are as free of residues as possible.

(100) Category II

(101) The nozzle region is free of residues, yet there are some in the zigzagged filter and column structure.

(102) Category III

(103) One nozzle channel contains residues or is completely clogged.

(104) Category IV

(105) Two nozzle channels contain residues or are completely blocked.

(106) 1.3. Provocation of Nozzle Blockages in the Nozzle Body

(107) 1.3.1 Provocation Solutions

(108) Provocation solutions are intended to cause the phenomenon of nozzle blockage or the jet divergency effect as often as possible. Therefore, in terms of their parameters, they are selected such that they will cause a very high number of clogged nozzles during a test. This is necessary in order to cause a sufficient number of clogged nozzle using as few samples as possible to allow meaningful assessments to be made as to whether a modification influences the occurrence of clogged nozzles.

(109) 1.3.2 Provocation of Nozzle Blockages During in-Use Operation of an Atomiser Having a DJI Nozzle (=Provocation Tests)

(110) 1.3.2.1 the Phenomenon of Nozzle Blockages

(111) In an SMI inhaler having a DJI nozzle, the aerosol production is based on two microfluidic jets impacting against one another. These jets are generated in the nozzle body by two rectangular nozzle outlet channels that are oriented at a 90° angle to one another. For the spray performance, it is critical that the two liquid jets impact against each other correctly. Provoked particle deposits in one or both nozzle channels can disrupt the proper impact by deflecting the jets and can lead to changes in the spray pattern and deviations in the fine particle fraction. In this document, the term “jet divergency” should be understood to mean group III spray patterns, which have an altered fine particle fraction or in which at best only a small portion of the amount of liquid to be discharged by atomiser actuation is actually discharged as an aerosol. A more comprehensive definition of the term is as follows: The phenomenon of jet divergency (jet deflection due to nozzle blockage) is a reversible or permanent, partial or total blockage/clogging of one or both nozzle channels of the nozzle body that has been caused by particles in the inhalation solution and leads to sprays having a different fine particle fraction.

(112) 1.3.2.2 Allocating Spray Patterns in Provocation Tests

(113) Provocation tests allow various factors influencing the phenomenon of nozzle blockage to be assessed. The test is designed such as to block as many inhaler nozzles as possible over the duration of the test. For this purpose, the atomisers are triggered once a day and the spray pattern is determined visually. This type of test scenario is referred below as an “in-use test” since it reproduces the daily use of an atomiser by a user, at least in relation to frequency of use.

(114) The spray pattern (SP) of the atomiser is recorded over the entire duration of the test in accordance with the categorisation set out above and is then evaluated. For this, the individual spray patterns are summarised in the three following superordinate groups: Normal (“good”) sprays (group I), spray pattern anomalies (group II) and sprays having an altered fine particle fraction (group III=jet divergency). The effect of a test parameter on the incidence can be deduced from comparing it with a reference.

(115) FIG. 5 shows example group I spray patterns. Sprays of this kind have a dominant, undivided spray cloud (with at most just a small amount of side spray). Normally, the spray cloud is symmetrical.

(116) FIG. 6 shows example group II spray patterns. The spray patterns in this group show split spray clouds. FIG. 6 shows a spray cloud split symmetrically, although asymmetrically split spray clouds or sprays split multiple times also belong to this spray pattern group. It should be noted here that detecting the division of the spray cloud may depend on the perspective of the observer. In (automatic) spray cloud image detection, therefore, use is preferably made of two cameras that are arranged at a 90° angle to the aerosol cloud axis and take the images of the aerosol cloud from two different perspectives at the same time as the device is triggered.

(117) Group III spray images, for which the fine particle fraction is considerably different from or less than the group I and II sprays (due to the impaired aerosol formation), cannot be detected using simple image recording systems. However, when special illumination is provided, the liquid exiting the nozzle, e.g. in one or two very thin jets, can be detected visually. In the spray patterns for this group, the liquid jets do not impact against each other due to them being deflected (and so the spray cloud typical of DJI nozzles is not formed). Therefore, jet divergency has occurred.

(118) 1.3.2.3 Spray Pattern Assessment

(119) The spray patterns are assessed visually (in a series of tests carried out manually) or by means of camera technology (in series of tests carried out using a stroke robot) and are assessed upon each stroke of the device. The spray pattern detected on the test day is recorded together with the stroke number. In the evaluation, the recorded spray patterns are assigned to the individual groups and the progression of the spray pattern over time is observed.

(120) During a provocation test, the atomiser is mounted in the laboratory at constant temperature and atmospheric humidity. The spray patterns are assessed for all the devices within a test on the same day.

(121) Devices put into operation for the first time have to be primed before the start of the test.

(122) To subsequently assess the spray patterns, the primed device is lifted into a suction apparatus at an angle of 45°. The distance from the suction machine is around 20-30 cm to allow the aerosol cloud to be clearly visible. The contrast can be increased by using black cardboard as a base and a cold light source. The spray patterns are classified according to the aforementioned groups.

(123) The spray patterns can also be assessed in an automated manner using a stroke robot. For this purpose, the robot is fitted with the pre-primed atomiser and a robot arm then coordinates the tensioning and release of the inhaler. Two CCD cameras are oriented at a 90° angle to the aerosol cloud axis and take images of the aerosol cloud at the same time as when the apparatus is triggered. The spray patterns are assessed manually by an employee on the basis of the images recorded.

(124) 1.3.3 Test Set-Up for a Standard Provocation Test

(125) The standard test set-up for a provocation test is as follows:

(126) Number of inhalers: 30-150 (depending on the influencing parameters being tested)

(127) In-use mode: 1×1 (1 stroke/day) or 1×2 (2 strokes/day)

(128) Test point: Spray pattern (specification according to spray pattern catalogue)

(129) Duration of test: 28-120 days; similar to 1-month/4-month patient usage

(130) Formulation: Ethanol/water 90/10 (v/v), pH 2.0

(131) Number of references: 30-75

(132) 1.3.3.1 10-Day Average and Provocation Reference

(133) The individual spray pattern curves in a provocation test can be subject to large fluctuations over the duration of the test. The number of devices having clogged nozzles is time-dependent and increases linearly after some time, often then reaching a steady state around which the system spreads. Therefore, it is not always possible to determine a precise, final rate. For this reason, in addition to the spray pattern curves, the average rate of group III sprays is determined for each test branch over the last ten days of the test (i.e. the 10-day average rate is determined). The actual scale of the influence of one test parameter on the incidence in group III sprays can only be discovered by comparing it with an internal provocation reference. In this respect, a provocation reference is a group of inhalers that are missing the feature being studied, e.g. nozzle coating. They represent the original state of the device and thus make it possible to draw meaningful conclusions on any modification.

(134) 1.4. Instrumental Analytics

(135) 1.4.1 Contact Angle Measurement

(136) The quality of the coating on Si/glass planar substrates can be described on the basis of static contact angle measurement. The results give information on the presence and homogeneity of the coating. The contact angle measurements were taken using a commercially available measuring instrument (in this case a DSA 10 MKR from Krûss GmbH) having associated control technology and control software (evaluation according to the Young-Laplace model).

(137) 1.4.1.1 Bases

(138) To assess the extent to which a liquid can wet a surface, the contact angle can be used. The contact angle is formed and measured on the solid/liquid/gas phase contact line between the solid and liquid. In this regard, the extent to which a liquid can wet a surface depends on the surface tension ratios between the solid and liquid (γ.sub.f/fl), solid and air (σ.sub.f/g), and liquid and air (σ.sub.fl/g). The boundary surface tensions at the droplet are in an equilibrium that can be expressed by Young equation 2.
σ.sub.f/g=γ.sub.f/fl+σ.sub.fl/g×cos θ  (2)

(139) The smaller the contact angle, the better the wetting. A contact angle of 0° is referred to as total wetting; in turn, a contact angle of 180° denotes no wetting. The equilibrium described by the Young equation is adjusted according to time and temperature.

(140) 1.4.1.2 Static Vs. Dynamic Contact Angle Measurement

(141) Unlike dynamic contact angle measurement, in static contact angle measurement the contact surface between the solid and liquid is not changed during the measurement. On an ideal, chemically and topologically homogeneous solid surface, a pure liquid in a saturated vapour phase would have an identical dynamic and static contact angle. This state is described in the Young equation. However, the contact angle can vary according to time and location. This phenomenon is counteracted by always measuring the contact angle immediately after the placement of the droplet.

(142) 1.4.1.3 Description of Contact Angle Measurement

(143) Before each measurement, the coated silicon or glass substrates are freed of adherent dust by means of oil-free air and placed on the sample stage for the measurement. For this purpose, the silicon substrates are positioned with the matte side downwards. By means of an automated microlitre syringe, a droplet having a volume of 8 μl is placed on the substrate and measured immediately using the control software or control program. For all contact angle measurements, double deionised water and/or n-dodecane is used. All contact angle values stated in the document are average values produced generally from the measurement of five droplets. In the process, each droplet is fitted and measured ten times by the control program similarly to the Young-Laplace model. Generally, each coating is carried out using 3-5 substrates, each substrate also being used for contact angle measurement.

(144) 1.4.2 Testing Metering Behaviour Using Delivered Mass (DM) and Metered Mass (MM)

(145) The metered dispensing of formulation by an atomiser, i.e. the measured mass of formulation (hereinafter “metered mass” (MM)), is determined by the structural dimensions of the inhaler and the density of the formulation solution. This is the mass of the formulation solution released through the atomiser when it is actuated. The MM is determined gravimetrically.

(146) Generally, a small portion of the measured mass remains on the nozzle as a residue following triggering. This cannot be completely prevented since there is always a backscattering region in the atomisation principle used based on an impact disc formed by two liquid jets. Therefore, the mass actually emitted from the device as a spray cloud (hereinafter “delivered mass” (DM)) is usually smaller than the metered mass (MM). The DM is also determined gravimetrically.

(147) 1.4.3 Ellipsometry

(148) 1.4.3.1 Apparatus Set-Up

(149) Ellipsometry is a measurement method that was used as early as in the 19th century. Some of the essential components of an ellipsometer have been available for a very long time, whilst other components of modern ellipsometers have only been developed recently.

(150) In spectroscopic ellipsometry (SE), a white light source having a monochromator enables the ellipsometric measurements at different wavelengths. In the process, the monochromator can be positioned upstream of the polariser or downstream of the analyser. There are also spectroscopic ellipsometers having a rotating compensator (cf. Irene, E. A. and H. G. Tompkins, Handbook of Ellipsometry, 2005: William Andrew Pub).

(151) The ellipsometric measurements in this document were taken using a spectroscopic ellipsometer having a rotating compensator. More information on the devices used can be found in Table 1.

(152) TABLE-US-00001 TABLE 1 Device used in the ellipsometry. Ellipsometer Alpha-SE ® having a rotating compensator from J. A. Woollam Co., Inc. (USA) Spectral range 390-900 nm Angle of incidence 65°, 70° and 75°, and transmission at 90° Control software Complete EASE from J. A. Woollam Co., Inc. (USA)

(153) 1.4.3.2 Description of the Ellipsometric Measurements

(154) Before each measurement, the silicon or glass substrates are first freed of adherent dust using oil-free air. Next, the substrates are carefully placed on the sample stage of the ellipsometer using wafer tweezers. Care should be taken to ensure the silicon substrates are positioned with their matte side downwards. In transparent glass substrates (Borofloat® 33, Nexterion), Scotch tape is stuck to the “tin side” before the substrates are positioned on the sample stage. The aforementioned marking scratched into the glass before the coating process using a diamond scriber should be identified.

(155) The “tin side” is stuck down for two reasons. Firstly, rear-side reflections on the transparent glass substrate should be prevented, and secondly, the layer thickness should not be measured on the “float side” of the glass due to the tin residues and because it makes analysis more complicated. The glass used is a float glass, which is poured onto a tin bath during its production, resulting in high levels of tin contamination for a glass side.

(156) After positioning the planar substrates on the sample stage, the stage vacuum can be switched on (better orientation parallel to the plane) and the measurement carried out. Measurements are taken at the angles 65°, 70° and 75°. After the measurement, the raw data is analysed using the corresponding method for silicon or glass by means of the analysis function in the Complete EASE software.

(157) 1.4.3.3 Development of an Analysis Method for Layer Thickness Measurement on Silicon and Glass Planar Substrates

(158) 1.4.3.3.1 Analysis Model for Silicon Substrates

(159) For the elliptometric measurement of the layer thickness on silicon substrates, the following layer model is assumed: Silicon substrate/native oxide/self-assembled monolayer.

(160) On its surface, silicon forms a native oxide layer; this forms relatively quickly, even after the HF stripping. During the coating, the self-assembled monolayer of the reactive silane compound orients itself on this oxide layer.

(161) This layer model can be modelled using the software (Complete EASE) on the measurement instrument used, provided that the necessary optical constants and layer thicknesses are present. The optical constants for pure silicon and for the native oxide layer are known and are stored in the analysis software database. The optical constants for the SAMs to be generated (when present) are found in the literature or have to be determined by the laboratory itself. To determine the layer thickness of the self-assembled monolayer, the Cauchy dispersion equation in the analysis software is used. This equation describes the refractive index n as a function of wavelength A and is suitable for analysing ultra-thin, transparent, non-absorbent films. The layer thickness of the native oxide layer after RCA treatment was determined by experiment, the aim being a maximum of 1.5 nm.

(162) The test is set-up as follows:

(163) Substrate: two silicon planar substrates per test condition

(164) HF treatment: 20 minutes prior to RCA treatment

(165) RCA treatment [min]: 0, 5, 10, 20, 60, 90

(166) In the standard coating method, the RCA cleaning runs for 60-90 minutes. To measure the layer thickness of the SAMs on Si substrates, therefore, a silicon oxide layer of 1.5 nm has to be used.

(167) It is important to characterise the layer thickness of the native oxide layer in order to later determine the layer thickness of the self-assembled monolayer on Si substrates.

(168) 1.4.3.3.2 Analysis Model for Glass Substrates

(169) To measure the layer thickness on glass substrates, the following layer model is assumed: Glass substrate/self-assembled monolayer.

(170) The model does not have an oxide layer, as is the case for silicon substrates. It is assumed that the self-assembled monolayer bonds directly to the glass substrate. In addition to the specific optical constants for the glass used in this case, consideration was also given to how the glass behaves after RCA treatment and the material constants were adjusted accordingly. The glass was input into the software as a material along with its constants such that, when the SAM refraction index is known, the layer thickness can also be determined in this case in accordance with the Cauchy dispersion equation.

(171) 1.4.4 Capillary Action Test on Nozzle Bodies

(172) Since the interior of the microstructure is difficult to access for analytical methods, the evidence of whether a nozzle body is coated is gathered from a capillary action test. The test is based on a capillary being spontaneously filled with solvents of different polarities (generally water). The capillary action test makes it possible to check whether a coating is present in just a few seconds. As a polar liquid, water wets the nozzle body surface of an uncoated nozzle body and immediately penetrates the microstructure upon contact with the nozzle body as a result of positive capillary forces (capillary ascension effect). This effect can be tracked under a microscope and does not occur, for example, if a nozzle body is coated to be hydrophobic.

(173) If the hydrophobic coating is successful, the capillary ascension effect does not take place and “capillary depression” is noted. In the case of capillary depression, the cohesion forces between the molecules are greater than the forces of adhesion to the surface. The liquid becomes globular and the surface is not wetted. A liquid that does not wet the capillary surface is expelled from the capillaries when it enters them.

(174) 1.4.4.1 Carrying Out the Capillary Action Test

(175) Using PTFE tweezers, a coated nozzle body is oriented under the microscope such that the internal microstructure is clearly visible. Next, a cotton bud (Q-tip) soaked in test liquid is carefully moved close to the intake region of the nozzle body. When the Q-tip comes into contact with the intake structure, the microstructure is not filled with a polar liquid, e.g. water, if the nozzle body is hydrophobed. As a positive check in this case, the test can be repeated using a non-polar substance, e.g. n-dodecane. Here, the filling is successful due to good wetting. In this case, a nozzle body coated with a perfluorinated alkylsilane should in turn be filled with a perfluorinated test liquid, e.g. perfluorooctane.

(176) 2. Results

(177) 2.1 Parameters Influencing the Spray Performance of Uncoated Nozzle Bodies

(178) When determining a suitable provocation solution, the extent to which different external influences, in particular pH, affect the spray performance of the atomisers under consideration was investigated beforehand. In this respect, for example ethanolic formulations (90/10) of various acidity levels (pH 2.0, 2.4, 2.8, 3.2, 3.5, 5.0) were tested for their likelihood to cause nozzle blockages.

(179) As an example from these preliminary tests, FIG. 7 shows how the jet divergency rate is dependent on the pH of the provocation solution. As can be seen in FIG. 7, pHs 2.0 and 2.4 saw an early, sharp increase in group III sprays. At the end of the test, a value of almost 70% was achieved for pH 2.0 and a value of around 50% for pH 2.4. The two curves show the first group III sprays very early on from day 2. This is contrary to the spray pattern curves for pHs 3.5 and 5.0, in which the proportion is almost 0% at the end of the test. It is not until day 21 that a small number of individual group III sprays occur.

(180) It is very clear from the spray pattern curves that the frequency of a group III spray increases sharply as the pH decreases (from 5.0 to 2.0).

(181) FIG. 8 shows the 10-day average rate (in tests over 28 days) of group III sprays (jet divergency) at the end of the test. As the pH increases, the number of group III sprays decreases. Therefore, the pH of the inhalation solution has a significant proportion of blocked/clogged nozzles. The lower the pH of the inhalation solution in the range tested here, the more frequent the deviations from a group I spray (“good spray”).

(182) Overall, a suitable provocation solution was determined from the preliminary investigations: Ethanol/water solution in a volume-to-volume ratio of 90:10, made acidic using HCl at pH of ≤2.0 (no active ingredient added).

(183) 2.2 Functionalisation of Planar Substrates and Nozzle Bodies

(184) 2.2.1 Functionalisation of Silicon and Glass Substrates

(185) 2.2.1.1 Impact of Activation

(186) To determine a suitable coating process for simultaneously coating silicon and glass, various influencing factors have to be investigated. The actual activation of the silicon and glass surfaces plays a key role in this respect since it specifically ensures sufficient stability of the coating.

(187) In one test, the effect of different activations on the properties of the resultant coating was investigated. Silicon and glass planar substrates were coated simultaneously under the same conditions using the four different activation methods (piranha solution, RCA solution and NaOH solution). They were then functionalised using 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane (c=0.003 mol/l). Suitable 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane is sold, for example, by Evonik under the name Dynasylan® F8216. After coating, the substrates were washed using 2-propanol and the static water contact angle was then measured. The aim was to obtain a set of parameters for each activation that can be used to obtain water contact angles for each activation solution of more than 100° after coating. A water contact angle of more than 100° indicates a homogeneous, all-over coating and good hydrophobicity.

(188) It was found that all the activations tested could be used to successfully coat silicon/glass planar substrates and to obtain contact angles of more than 100°.

(189) However, it was also found that the NaOH activation method was only suitable for the coating process to a limited extent since it frequently causes clouding on the glass. This phenomenon is also known as glass corrosion. For this reason, the NaOH activation was intended for preliminary tests only, not for the subsequent tests on the nozzle body. For the quality check on the DJI nozzles produced, it is desirable to always be able to view the nozzles under a microscope. If the glass component of the nozzle body were cloudy, this would no longer be possible.

(190) 2.2.1.1 Effect of Coating Reagent Concentration

(191) The impact of the coating concentration on the water contact angle on silicon/glass planar substrates was analysed. For this purpose, the silicon/glass planar substrates were coated with a short-chain alkylalkoxysilane (methyltrimethoxysilane=MTMS). To analyse the effect of the coating concentration, the concentration was gradually increased. Then, the substrates were characterised using static water contact angle measurement.

(192) The silicon and glass planar substrates were coated in a similar manner to the description according to section 1.1. They were activated using an RCA activation for 20 minutes at 75° C. The substrates were coated for two hours without ultrasound at room temperature using the functionalisation reagent methyltrimethoxysilane (abcr GmbH, Karlsruhe, Germany) at various substance concentrations from 0.7 mmol/l to 70 mmol/l. Next, the substrates were dried for one hour at room temperature and lastly tempered in an oven for one hour at 120° C. The test results are shown in Tables 2 and 3 below.

(193) TABLE-US-00002 TABLE 2 Water contact angle (with standard deviation d) of glass planar substrates coated with MTMS according to MTMS concentration MTMS concentration Contact angle d [mmol/l] [°] [°] 0.7 67.11 12.45 3.5 87.45 6.03 7 78.32 4.47 14 83.60 8.13 35 97.39 8.46 70 103.22 2.68

(194) TABLE-US-00003 TABLE 3 Water contact angle (with standard deviation d) of silicon planar substrates coated with MTMS according to MTMS concentration MTMS concentration Contact angle d [mmol/l] [°] [°] 0.7 51.79 7.18 3.5 79.82 12.61 7 79.00 3.66 14 79.50 4.45 35 105.13 9.13 70 112.26 7.23

(195) It was found that the water contact angle increases on both glass and silicon as the MTMS concentration increases. At the same time, increasing the concentration at low concentrations has a somewhat greater effect than at higher concentrations. The system approaches a maximum in the range of from 110-120°. Therefore, the contact angle is affected less and less by further increasing the coating concentration. Overall, it can also be seen that, from a concentration of around 7 mmol/l, values showing a moderate spread can be detected. In general, this points to a more homogeneous surface covering with coating molecules.

(196) The data shows that it is possible to successfully coat both substrate materials in one coating process.

(197) As is clear, the contact angle increases as the concentration of the coating reagent increases, asymptotically approaching an apparent maximum at higher concentrations. This seems plausible since the surface is saturated after a certain point, i.e. the layer has formed all over the surface. This effect can be easily derived from the results for silicon and glass. In the low coating concentration ranges, a greater effect in terms of the contact angle is achieved when the concentration is increased slightly, whereas this effect becomes weaker and weaker in the higher concentration ranges. In this respect, doubling the concentration from 35 mmol/l to 70 mmol/l only increases the water contact angle by another 10°. There are no discernible general differences for silicon or glass (the related requirement for coating glass and silicon at the same time is thus met).

(198) On the basis of the results, an optimum minimum coating concentration for the process can also be derived. An optimum minimum coating concentration should be more than 30 mmol/l since it has been found that reliable and reproducible coatings are obtained above this concentration.

(199) 2.2.2 Functionalisation of Nozzle Bodies

(200) Nozzle bodies are generally functionalised according to the procedure under section 1.1. The external appearance of nozzle bodies after coating was tested.

(201) 2.2.2.1 Impact of Activation Solution

(202) In one test, nozzle bodies were activated using the various activation solutions and then coated with 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane. Next, the bodies were visually checked for noticeable problems under a light microscope. In general, RCA-activated and piranha-activated nozzle bodies have a similar appearance.

(203) Generally, coated nozzle bodies have a similar surface to uncoated nozzle bodies, even under the scanning electron microscope, and are thus considered optically identical. A monomolecular layer of an alkylsilane cannot be directly detected by scanning electron microscopes, so coated and uncoated nozzle bodies cannot be directly distinguished either. All coated nozzle bodies showed positive results in the capillary action test and were thus successfully coated. Morphological artefacts only occurred during the surface functionalisation in very few cases. These artefacts only occur in a very small number of cases, so the coating does not usually alter the surface morphology of the nozzles.

(204) 2.2.2.2 Capillary Action Test on Coated Nozzle Bodies

(205) Table 4 illustrates the results of the capillary action test on ten hydrophobed, perfluorinated nozzle bodies coated with 0.03 mol/l 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane. The method followed the description in section 1.1 and was carried out after each coating using randomly selected nozzle body samples.

(206) TABLE-US-00004 TABLE 4 Capillary action test on coated, perfluorinated nozzle bodies (N = 10) Coated Uncoated (perfluorinated) nozzle body nozzle body Solution/formula spontaneously spontaneously (property) filled filled Water/H.sub.2O Yes No (hydrophilic) Perfluorooctane/C8F18 No Yes (hydrophobic, oleophobic)

(207) Direct evidence of the coating in the coated nozzle bodies is difficult to obtain since it can only be deduced by means of complicated, time-consuming measurement methods such as TOF-SIMS and ESCA after splitting the sandwich system. This is due to the low sample volumes resulting from the thin layer thickness of just a few nanometres.

(208) The simplest way, therefore, is to use indirect evidence of the coating by making use of the altered capillary effect of the microstructure. The method is an identity test for the presence of the coating. This test is also suitable for large-scale quality control.

(209) 2.2.2.3 Impact on Spray Performance from Functionalisation Solution Residues in Nozzle Bodies

(210) In the laboratory situations used here, not all the samples emerged clean from the coating process. To be able to analyse the effect of coating residues on the spray pattern performance, a catalogue of fault images having images of characteristic nozzle faults was compiled. In the process, the nozzles were allocated to a relevant nozzle category depending on the location of the residue.

(211) 2.2.2.4 Effect of the Nozzle Body Category on Spray Performance

(212) To investigate the effect of the nozzle body category on the spray performance of the atomisers, nozzles coated for the long-duration provocation test (cf. 0) were analysed under a microscope and assigned to the appropriate nozzle body category. Table 5 shows the results of the microscopic analysis of three batches (A to C), which were produced for the long-duration provocation test sampling. Overall, 151 nozzle bodies were coated with 0.03 mol/l 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane in three different batches and then analysed under the microscope.

(213) TABLE-US-00005 TABLE 5 Results of the nozzle body categorising based on samples for long-duration provocation test from 0 Batch Category I Category II Category III Category IV Total A 26 (53%) 3 (6%) 14 (29%) 6 (12%) 49 B 28 (54%) 0 13 (25%) 11 (21%) 52 C 21 (42%) 1 (2%) 13 (26%) 15 (30%) 50 Average 49.64% 2.71% 26.52% 21.13% 151 SD  6.62% 3.12%  1.84%  8.88%

(214) For the long-duration provocation test, 60 atomisers having coated nozzles were provided. The original nozzles were removed from these devices and coated nozzles were fitted. The distribution of the coated nozzles used here in relation to the nozzle body categories can be seen in Table 6.

(215) TABLE-US-00006 TABLE 6 Distribution of sample nozzle body categorisation for long-duration provocation test Category I Category II Category III Category IV Total 30 0 15 15 60

(216) All the coated nozzles were fitted in provided atomisers in a traceable manner so that it was always known which nozzle category was fitted in which device. They were fitted in particle-free conditions, as can be produced under laboratory conditions. After the provocation test, the progression of the spray patterns of the samples was analysed in a large histogram analysis.

(217) After 120 test days in 1×2 in-use mode, i.e. two successive strokes per day, all the 14400 sprays dispensed in the test from the 60 coated nozzles were analysed in a comprehensive histogram analysis. It was found that initially individual group II spray patterns were formed. As the test continued, clear rows or bands emerged, which indicated that blocked devices preferably again displayed a group II spray in the following strokes. As the test time progressed, group III sprays developed from the group II sprays, and these would become permanent later in the test. In this test, the absolute number of group III sprays was relatively low, so the group II sprays were used for the bar chart analysis. Group II sprays are a precursor for group III sprays, meaning that they can be legitimately used for the further analysis.

(218) It was shown that category III and category IV nozzle bodies cause group II sprays more often than nozzle body categories I and II—15% for nozzle body categories I and II and between 25 and 30% for nozzle body categories III and IV.

(219) It was found that clean category I and II nozzle bodies have a lower number of spray pattern anomalies over the entire duration of the test than category III and IV nozzles. Residues in the nozzle region thus generally lead to more spray pattern anomalies. The number is even greater when both nozzle channels (category IV) are affected by deposits. The majority of the spray pattern anomalies in this test occur later on in the test. Category I and II nozzles always have better spray performance, even when broken down according to time interval. Both the short-term and long-term spray performance of clean category I and II nozzle bodies are considerably better than that of category III and IV nozzle bodies.

(220) Therefore, it can be concluded that the residues impair the coating or the formation of the coating on the surface and spray performance deteriorates.

(221) A RAMAN spectroscopic analysis revealed that the residues are polymer residues of the coating solution. During the drying phase after functionalisation, the coating solution presumably coalesced at the small cavities of the microstructure and polymerised out at these points during the tempering.

(222) The results reveal that the yield of category I nozzles should be as high as possible for large-scale processes. For this purpose, it is appropriate to use systems that expel the excess coating solution, in particular systems based on the use of rotary forces, e.g. SRD systems (spin rinse dryers).

(223) 2.2.3 Performance of Functional Alkylsilane Coatings in in-Use Methods (Provocation Tests)

(224) In the following, using a standard test method, the surface functionalisation of the nozzle bodies a coating method will be sufficiently analysed and evaluated as a measure for preventing the risk of nozzle blockage.

(225) 2.2.3.2 Provocation Test Using Functionalised Nozzles: Performance of the Surface Functionalisation in Relation to Plaque Deposits Using a provocation test (use of the predetermined provocation solution in atomisers), the behaviour of different layer functionalities when provoked plaque deposits occur was tested. In the process, nozzle bodies were coated with different coating reagents according to the method described in section 1.1 (with RCA activation).

(226) The provocation test is set up as follows:

(227) Coating Reagents: Dynasylan® F8261 (=1H,1H,2H,2H-tridecafluorooctyltriethoxysilane) (F1308-OET) n-octyltriethoxysilane (C8-OET) 1H,1H,2H,2H-perfluorooctyldimethylchlorosilane (F13C8-Cl) n-octyldimethylchlorosilane (C8-Cl)

(228) Concentration: 0.03 mol/l in each case

(229) Formulation: as in 1.3.3

(230) Number of inhalers: 50 for each coating, 30 for reference

(231) In-use mode: 1×1 stroke/day

(232) Duration of test: 28 days

(233) Test parameters: Spray pattern according to spray pattern catalogue

(234) The basic test set-up makes it possible to compare a large number of aspects within the experiment. Firstly, both fluorinated and non-fluorinated alkylsilanes are used here, and secondly, these are alkyltrialkoxysilane and alkyldimethylmonochlorosilane. This means that the basic coating chemistry in terms of the surface bonding can be tested, while so too can the actual layer performance determined by the alkyl side chains (fluorinated vs. non-fluorinated).

(235) In the following, the spray pattern curves for group I sprays (cf. FIG. 9), group II sprays (cf. FIG. 10) and group III sprays (cf. FIG. 11) are shown in graphs.

(236) 2.2.3.1.1 Group I Spray Pattern Curve

(237) FIG. 9 shows the group I (“good sprays”) spray curves for the coated nozzles and the uncoated reference. The number of group I sprays decreases for all test branches to a greater or lesser extent over time, although all the test branches in which coated nozzles are used display a better spray pattern curve than the uncoated reference. It is not until around the end of the test that the spray pattern curves for test branches C8-OET, C8-Cl and F12C8-Cl approach the reference. In this regard, the number of group I sprays drops below 20% for these test branches.

(238) This does not occur with coating reagent F13C8-OET, which shows excellent performance over the spray pattern curve. Up to the end of the test, a proportion of almost 80% remained in group I sprays (“good sprays”). This is an advantage of more than 60% over the other coating reagents. If this correlated with the coating reagent C8-OET, the advantage of the perfluorinated side chain of F1308-OET is clear.

(239) The coating reagent C8-OET obtains a similarly poor result towards the end of the test as the two chlorosilanes F1308-Cl and C8-Cl. It is clear, however, that it has a significant advantage over these two compounds in the preceding test days, taking an intermediate position between F1308-OET and the two chlorosilanes.

(240) Overall, the relatively poor spray performance of the two chlorosilanes is surprising. It is also surprising that, in relation to F1308-Cl, the benefit of a perfluorinated side chain is not detectable either compared with C8-Cl. The fact that C8-Cl provides similar results leads to the assumption that the surface bonding is not stable enough in the chlorosilanes. While they do have an advantage over the uncoated reference in the first few test days, this lasts only until day 17.

(241) 2.2.3.1.2 Group II Spray Pattern Curve

(242) FIG. 10 shows the group II spray curves for the coated nozzles and the uncoated reference. The proportion of group II sprays (“spray pattern anomalies”) increases over time to a greater or lesser extent for all test branches. The reference reaches its maximum as early as test day 4 at over 60%, whilst no specific maximum can be identified for the other test branches. There is a continuous increase of group II sprays over the test duration. What is also striking is the large advantage of the coating reagent F1308-OET. At the end of the test, the proportion of spray pattern anomalies is just 15%, and from day 5 onwards is always below the other test branches over the entire test period.

(243) 2.2.3.1.3 Group III Spray Pattern Curve

(244) FIG. 11 shows the group III (“jet divergency”) spray curves for the coated nozzles and the uncoated reference. The figure shows that the proportion of group III sprays also increases to a greater or lesser extent over the time period for all test branches. Towards the end of the test, the two chlorosilanes reach a similar rate of group III sprays (almost 60%) as the uncoated reference. Compared with the reference, therefore, the chlorosilanes show only a temporary benefit lasting a few test days since the chlorosilanes and the reference already have a similar number of group III sprays (around 40%) after day 16. In this view, the coating reagent C8-OET again assumes a position in the middle having a group III spray rate of almost 30% at the end of the test. The coating reagent F13C8-OET has by far the best spray performance: Less than 10% of the atomisers showed a group III spray over all test days.

(245) 2.2.3.1.4 Group III Spray Pattern Curve: 10-Day Average Rate Towards the End of the Test

(246) The advantage of the coating reagent F1308-OET is particularly clear from the 10-day average rate shown in FIG. 15. In the 10-day average, the reference reaches a rate of almost 50% group III sprays. This is similar to the two chlorosilane test branches, which exhibit a similarly high 10-day average rate of around 50%. The lowest rate of group III sprays is achieved by the test branch containing the coating reagent F1308-OET, which achieves a 10-day average rate of less than 5% towards the end of the test; this is remarkably better than all the other test branches. The coating reagent C8-OET also has a significantly better 10-day average rate of just around 25%. This coating reagent thus provides significantly better protection against plaque deposits than the alkylchlorosilanes.

(247) Overall, the advantage of alkyltrialkoxysilanes over alkylchlorosilanes is very unmistakeable.

(248) The result of the provocation test using functionalised nozzles shows that the likelihood of a clogged nozzle can be dramatically reduced when the correct coating reagent is used.

(249) For incomprehensible reasons, the alkyldimethylmonochlorosilanes performed relatively badly and barely brought any benefit to spray performance compared with the reference. This was unexpected since the alkyldimethylmonochlorosilanes also exhibit good binding properties in the literature (cf. Fadeev, A. Y. and T. J. McCarthy, Trialkylsilane monolayers covalently attached to silicon surfaces: wettability studies indicating that molecular topography contributes to contact angle hysteresis. Langmuir, 1999. 15(11): pp. 3759-3766). The spray pattern curve over the testing period is no better than the uncoated reference.

(250) Reasons why the alkyldimethylmonochlorosilanes bonded poorly to the surface can only be hypothesised. One possible critical disadvantage of alkyldimethylmonochlorosilanes is that this substance class only has one site for binding to the surface. By comparison, alkyltrialkoxysilanes have two more coordination sites, which appears to lead to increased bonding stability. If consideration is also given to the force conditions in the atomisers used, which can reach a pressure of more than 200 bar and flow rate of over 130 m/s in the nozzle channel, it may be that these forces are simply too great for just one coordination site, resulting in layer erosion or layer detachment.

(251) 2.2.4 Long-Term Performance of 1H,1H,2H,2H-Tridecafluorooctyltriethoxysilane in in-Use Tests

(252) In a provocation test, it was investigated how long nozzle blockages could be avoided by using coated nozzles. To do so, nozzles were coated with 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane and tested in a provocation test in in-use mode 1×2 over a period of 120 days. Overall, 60 coated nozzles of nozzle body categories I+II, III and IV were used.

(253) The provocation test is set up as follows:

(254) Coating Reagent: Dynasylan® F8261 (=1H,1H,2H,2H-tridecafluorooctyltriethoxysilane) (F1308-OET) Reference (uncoated)

(255) Concentration: 0.03 mol/l

(256) Formulation: as in 1.3.3

(257) Duration of test: Ref.=45 days, F13C8=120 days

(258) Number of inhalers: 60 coated with F13C8-OET, 30 reference (uncoated)

(259) In-use mode: 1×2 strokes/day

(260) Test parameters: Spray pattern according to spray pattern catalogue

(261) 2.2.4.1 Group I Spray Pattern Curve

(262) FIG. 13 shows the group I (“good sprays”) spray curves for the long-duration provocation test. FIG. 13 illustrates the great advantage the coating has over the uncoated reference. Whereas the reference had only around 20% group I sprays (“good sprays”) on test day 45, the coated nozzles had more than 80%. Even at the end of the test on test day 120, F13C8-OET still had more than 60% of the atomisers producing a group I spray (“good spray”). As can be seen in the graph, the reference was stopped on day 45 since the good-spray rate had reached a balanced state at around 20%.

(263) Within the first few test days, some of the coated atomisers actually began to deviate from a group I spray. This can be traced back to the presence of coating residues. Over the first few days, therefore, the reference tended to perform slightly better than the devices having coated nozzles. This effect lasted until around day 10, after which the number of good sprays in the reference began to drop considerably. In this case, the reason for this phenomenon at the beginning is the use of “good” and “bad” nozzle categories. In terms of coating residues, category I and II nozzles as well as category III and IV nozzles were used in this test.

(264) It is very clear from this test that coating solution residues can also cause group II and group III sprays.

(265) 2.2.4.2 Group II Spray Pattern Curve

(266) FIG. 14 shows the group II (“spray anomalies”) spray curves for the long-duration provocation test. For the reference in particular, the number of group II sprays increases very sharply over time. In this regard, more than 70% group II sprays were already observed on day 20. Looking at the spray pattern curve for the coated nozzles (F1308-OET), it can be seen that a moderate increase in group II sprays was also apparent in this case. However, the coated nozzles only had around 30-40% group II sprays even at the end of the test on day 120; this is considerably lower than the reference.

(267) It is also apparent that spray pattern anomalies (group II sprays) occurred in the devices having coated nozzles as early as on the first test day. This is down to the coating residues as mentioned above. Despite the coating, the F1308 devices also showed an increase in group II sprays after a certain time, which indicates the process of plaque deposit formation in this case too. Over the spray pattern curve, a larger increase can be seen from around day 40.

(268) In terms of the coating residues, the coated nozzle bodies having coating residues can be discarded before being fitted as part of a visual check on the coated nozzle bodies. Discarding these significantly reduces the spray pattern anomalies for devices having coated nozzles.

(269) 2.2.4.3 Group III Spray Pattern Curve

(270) FIG. 15 shows the group III (“jet divergency”) spray curves for the long-duration provocation test. As can be seen in the figure, very few group III sprays occur overall (batch effects also play a role in this respect). When comparing the coated nozzles with the reference, it can be seen that the reference already showed some group III sprays after test day 10, whereas with the coating this occurred for the first time as of day 70. This means that the coating was able to delay the occurrence of the first group III sprays for around 60 days.

(271) Under stress conditions, the atomiser batch tested here had a very low number of group III sprays from the outset, although a large number of group II sprays were formed very quickly in the reference. Generally, these group II sprays quickly react further to become group III sprays and can thus also be an indicator of nozzle blockages.

(272) 2.2.4.4 Group III Sprays: 10-Day Average Rate Towards the End of the Test

(273) FIG. 16 shows the 10-day average rate at the end of the test for the long-duration provocation test. FIG. 16 again shows a summary of the results of the spray pattern progression for group Ill sprays. On day 45, the reference had a rate of around 5% group III sprays. By comparison, the atomisers having coated nozzles had still not shown any group III sprays. These occurred sporadically from around day 70. Even at the end of the test on day 120, the atomiser devices having coated nozzles had a similar 10-day average rate of around 5%.

(274) 2.2.4.5 Results for Performance of 1H,1H,2H,2H-Tridecafluorooctyltriethoxysilane Nozzle Coatings

(275) The result of the long-duration provocation test using functionalised nozzles shows that the likelihood of a clogged nozzle can be dramatically reduced even over a very long test period. The duration of this test was 120 days altogether. The 120 days were derived from a potential 4-month use of the inhaler. For the nozzles coated in this case with 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane, a group III spray first occurred on test day 70. The reference, on the other hand, already showed the first group III sprays after day 10. This is an advantage for the coating of 60 days. The coating delayed the appearance of group III sprays for this long period of time, which is an excellent result.

(276) What is also striking in this test is that very few group III sprays occurred over the entire test duration. This is surprising even for the uncoated reference. The reason for this is the atomiser batch itself, which has also demonstrated very few group III sprays in other tests. The incidence of group II and III sprays is dependent on the device batch used. However, it is notable in this test that the number of group II sprays remained very high for a very long time. Generally, the number of group III sprays increases significantly when the maximum in the group II sprays is reached very quickly (i.e. devices having group II sprays become devices having group III sprays).

(277) The reference was stopped at day 45 since the advantage of the coating was already clear and the reference had reached a balanced state. At day 70, the inhalers having the coated nozzles also showed the first group III sprays. However, the rate of the increase did not match that of the uncoated reference; instead, it rose much more slowly.

(278) 2.3. Long-Term Stability of the Coating: Stability Study on Coated Si/Glass Planar Substrates

(279) A stability study is designed to investigate the effect of different stress parameters (pH, temperature, storage time) on the layer performance of 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane. For this purpose, silicon and glass substrates were coated to three concentrations in a plurality of coating approaches and stored over time in ethanolic placebo solution. The samples were each coated to identical coating parameters and differed only on account of the concentration of the coating solution. The stability of the samples was then assessed using layer thickness measurements and contact angle measurements.

(280) 2.3.1 Stability Study Set-Up

(281) Si/glass planar substrates were coated at the three following concentrations: 0.003 mol/l, 0.015 mol/l and 0.03 mol/l. The samples were stored in a sealed manner protected from light in polyethylene bottles in an ethanolic placebo solution at pHs of 2.0 and 4.5, and at room temperature in the laboratory (20° C.±3° C.) and 40° C. in the climatic test cabinet. For the high coating concentration, the storage time was 14, 30, 90 and 180 days. The two lower coating concentrations were stored for just 30 and 180 days. The study parameters are summarised in Table 7.

(282) TABLE-US-00007 TABLE 7 Set-up for Si/glass planar substrate stability study Sample Time Temper- concentration [days] ature pH [mol/l] Substrate 14 30 90 180 RT 40° C. 2.0 4.5 0.03 Si + glass x x x x x x x x 0.015 Si + glass x x x x x x x 0.003 Si + glass x x x x x x Reference x x x x x x x x Reference: x = carried out without silicon/glass substrates

(283) 2.3.2 Coating the Si/Glass Planar Substrates

(284) For the stability study, the Si/glass planar substrates were coated in several batches in accordance with section.

(285) 2.3.3 Storing the Substrates

(286) For the stability study, two ethanolic placebo solutions at pH 2.0 and 4.5 were used. On the day the samples were placed into storage, 170 g ethanolic placebo solution was placed in each polyethylene bottle. Next, two silicon and glass substrates were added to each bottle using wafer tweezers. In doing so, care should be taken to ensure the substrates are positioned in the bottle such that they do not adhere to one another. Once the bottles are closed, the weight of the entire PE bottle was determined and recorded in order to be able to determine the weight loss due to the storage when the substrates are removed from storage. In addition, reference samples without substrates were also put into storage for each storage duration, pH and temperature.

(287) 2.3.4 Checking the pH and Weight Upon Removal from Storage

(288) On the day the substrates are removed from storage, the weight of the PE bottle was determined again and the weight lost over the storage time thus ascertained. In addition to the weight loss, the pH of the solution was also checked. This makes it possible to check whether the PE bottle was sufficiently tight over the duration of the test and whether the pH has changed during storage.

(289) 2.3.5 Analysing the Stability Study

(290) The stability study was analysed using static contact angle measurement and layer thickness measurements by means of spectroscopic ellipsometry. The values of a sample upon removal from storage were related to starting values determined beforehand for the contact angle and layer thickness. The starting values were collected using a representative sample that had not been stored and contained samples from each coating process carried out beforehand.

(291) 2.3.5.1 Change in the Contact Angle Over the Storage Time

(292) Below, the results in term of the contact angle measurements on silicon and glass are presented.

(293) The tests show a clear concentration effect for the starting values on silicon. As the coating concentration was increased, a higher contact angle for water and n-dodecane was obtained and a smaller standard deviation was produced. In each case, 5 droplets of water and 5 droplets of n-dodecane were measured per substrate. With 10 measurements per droplet, this resulted in 50 angles.

(294) 2.3.5.1.1 Changes to the Contact Angle on Silicon when Stored at Room

(295) Temperature (25° C.) The change in the water contact angle and n-dodecane contact angle on silicon when stored at room temperature was tested. The test results are summarised in Tables 8 and 9.

(296) TABLE-US-00008 TABLE 8 Changes to the water contact angle on silicon when stored at room temperature (25° C.) MTMS Contact angle [°] concentration Starting Change [mmol/l] pH value 14 days 30 days 90 days 180 days 0.03 2.0 112.72 1.39 2.07 3.44 7.34 0.03 4.5 112.72 3.05 4.18 5.66 6.97 0.015 2.0 108.78 n.d..sup.1 4.73 3.25 3.96 0.015 4.5 108.78 n.d..sup.1 6.55 3.27 3.19 0.003 2.0 101.61 n.d..sup.1 1.76 n.d..sup.1 7.52 0.003 4.5 101.61 n.d..sup.1 0.01 n.d..sup.1 −1.41 .sup.1n.d. = not determined

(297) TABLE-US-00009 TABLE 9 Changes to the n-dodecane contact angle on silicon when stored at room temperature (25° C.) MTMS Contact angle [°] concentration Starting Change [mmol/l] pH value 14 days 30 days 90 days 180 days 0.03 2.0 68.82 0.51 3.11 4.40 4.82 0.03 4.5 68.82 0.33 3.20 3.51 5.42 0.015 2.0 66.40 n.d..sup.1 −1.71 0.82 1.55 0.015 4.5 66.40 n.d..sup.1 0.26 1.59 1.48 0.003 2.0 61.42 n.d..sup.1 6.68 n.d..sup.1 5.03 0.003 4.5 61.42 n.d..sup.1 2.20 n.d..sup.1 −1.72 .sup.1n.d. = not determined

(298) The tests show that the water contact angle and the n-dodecane contact angle only vary to a minor extent (x≤10°) over the storage period considered. However, a gradual fall in the contact angle over the storage period can be seen for the coating concentration 0.003 mol/l. This effect can be detected in both water and dodecane.

(299) It can be stated that the spread in the contact angle at a coating concentration of 0.003 mol/l is very high overall. In some cases, it is over 20° and thus conceals potential effects. By comparison, at the coating concentrations 0.015 mol/l and 0.03 mol/l, very precise values are achieved with very low spread. This indicates that a homogeneous coating is not always achieved at a concentration of 0.003 mol/l.

(300) Since all the average changes for the water and the dodecane contact angle remain below 10°, it can be concluded that, in this case, at most there has merely been a change to the layer alteration but no layer detachment. The water contact angle for an uncoated silicon or glass substrate would be at around 30-40°. An activated substrate would be even more hydrophilic. The water contact angles considered here are above 100° even after storage.

(301) Detachment of the layer would lead to changes in the water contact angle of well over 50°. This is clearly not the case here.

(302) There was also no general discernible difference between storage at pH 2.0 and pH 4.5 (i.e. the coating is acid-resistant in the tested pH region).

(303) 2.3.5.1.2 Changes to the Contact Angle on Silicon when Stored at 40° C.

(304) The change in the water contact angle and n-dodecane contact angle on silicon when stored in the climatic test cabinet at 40° C. was tested. The test results are summarised in Tables 10 and 11.

(305) TABLE-US-00010 TABLE 10 Changes to the water contact angle on silicon when stored at 40° C. MTMS Contact angle [°] concentration Starting Change [mmol/l] pH value 14 days 30 days 90 days 180 days 0.03 2.0 112.72 0.01 5.20 4.17 8.13 0.03 4.5 112.72 3.49 3.41 6.74 9.08 0.015 2.0 108.78 n.d..sup.1 −6.50 0.13 4.39 0.015 4.5 108.78 n.d..sup.1 −3.53 3.42 6.34 0.003 2.0 101.61 n.d..sup.1 0.90 n.d..sup.1 16.73 0.003 4.5 101.61 n.d..sup.1 3.83 n.d..sup.1 6.08 .sup.1n.d. = not determined

(306) TABLE-US-00011 TABLE 11 Changes to the dodecane contact angle on silicon when stored at 40° C. MTMS Contact angle [°] concentration Starting Change [mmol/l] pH value 14 days 30 days 90 days 180 days 0.03 2.0 68.82 0.88 2.17 1.73 7.19 0.03 4.5 68.82 2.29 2.42 5.36 5.93 0.015 2.0 66.40 n.d..sup.1 −1.71 −0.19 1.19 0.015 4.5 66.40 n.d..sup.1 0.26 2.11 3.21 0.003 2.0 61.42 n.d..sup.1 −4.68 n.d..sup.1 11.67 0.003 4.5 61.42 n.d..sup.1 1.67 n.d..sup.1 2.21 .sup.1n.d. = not determined

(307) The tests show that the contact angle only varies slightly in terms of the stress variables shown, as set out in Table 7 in section 2.3.1. The average change in the water contact angle remains below 10°, even when the temperature is increased. A gradual fall in the contact angle over time was also recorded here at the concentrations 0.015 mol/l and 0.03 mol/l, similarly to the values determined at room temperature.

(308) Overall, it can be stated that increasing the temperature does not reduce or alter the hydrophobicity of the substrates either. The data is absolutely comparable with the values at room temperature. Therefore, no temperature effects can be detected on the silicon substrates.

(309) 2.3.5.1.3 Changes to the Contact Angle on Glass when Stored at Room Temperature (25° C.)

(310) The change in the water contact angle and n-dodecane contact angle on glass when stored at room temperature was tested. The test results are summarised in Tables 12 and 13.

(311) TABLE-US-00012 TABLE 12 Changes to the water contact angle on glass when stored at room temperature (25° C.) MTMS Contact angle [°] concentration Starting Change [mmol/l] pH value 14 days 30 days 90 days 180 days 0.03 2.0 112.06 2.13 −2.26 −1.00 3.92 0.03 4.5 112.06 −4.31 0.75 −6.21 0.61 0.015 2.0 87.41 n.d..sup.1 −9.10 −14.92 −13.81 0.015 4.5 87.41 n.d..sup.1 −12.65 −12.05 −19.27 0.003 2.0 91.40 n.d..sup.1 −5.48 n.d..sup.1 22.08 0.003 4.5 91.40 n.d..sup.1 −3.65 n.d..sup.1 2.03 .sup.1n.d. = not determined

(312) TABLE-US-00013 TABLE 13 Changes to the dodecane contact angle on glass when stored at room temperature (25° C.) MTMS Contact angle [°] concentration Starting Change [mmol/l] pH value 14 days 30 days 90 days 180 days 0.03 2.0 68.91 3.74 1.27 2.61 2.79 0.03 4.5 68.91 2.64 5.84 −2.66 1.59 0.015 2.0 47.52 n.d..sup.1 −11.50 −12.20 −2.78 0.015 4.5 47.52 n.d..sup.1 −14.49 −9.92 −15.51 0.003 2.0 60.57 n.d..sup.1 16.11 n.d..sup.1 2.85 0.003 4.5 60.57 n.d..sup.1 9.65 n.d..sup.1 6.89 .sup.1n.d. = not determined

(313) It can be seen that the layer does not detach on glass either. In the substrates, the average change in the contact angle also remains below 10° at a coating concentration of 0.03 mol/l, as already seen with the silicon substrates. Looking at the data for the coating concentrations 0.003 mol/l and 0.015 mol/l, a very high spread in the values can be seen here too, in particular for the coating concentration 0.003 mol/l. This is similar to the spread on silicon for the same coating concentration and also conceals potential effects here.

(314) The results for the coating concentration 0.015 mol/l also display a noticeable feature. Here, the contact angle for water and n-dodecane is greater than the determined starting value. This trend was measured for all samples, though it is insignificant once standard deviation is taken into account.

(315) Overall, it can be concluded that no storage effects can be noted for glass substrates either. This also includes the two pHs tested.

(316) 2.3.5.1.4 Changes to the Contact Angle on Glass when Stored at 40° C.

(317) The change in the water contact angle and n-dodecane contact angle on glass when stored at 40° C. was tested. The test results are summarised in Tables 14 and 15.

(318) TABLE-US-00014 TABLE 14 Changes to the water contact angle on glass when stored at 40° C. MTMS Contact angle [°] concentration Starting Change [mmol/l] pH value 14 days 30 days 90 days 180 days 0.03 2.0 112.06 0.79 −2.19 17.55 −1.56 0.03 4.5 112.06 −1.89 2.74 −1.67 −1.08 0.015 2.0 87.41 n.d..sup.1 −17.06 −13.58 −14.47 0.015 4.5 87.41 n.d..sup.1 −17.54 −18.90 −23.69 0.003 2.0 91.40 n.d..sup.1 −16.71 n.d..sup.1 14.21 0.003 4.5 91.40 n.d..sup.1 −13.74 n.d..sup.1 −16.11 .sup.1n.d. = not determined

(319) TABLE-US-00015 TABLE 15 Changes to the dodecane contact angle on glass when stored at 40° C. MTMS Contact angle [°] concentration Starting Change [mmol/l] pH value 14 days 30 days 90 days 180 days 0.03 2.0 68.91 2.03 −0.67 13.38 −0.03 0.03 4.5 68.91 7.04 2.51 3.96 −1.25 0.015 2.0 47.52 n.d..sup.1 −17.96 −13.12 −13.24 0.015 4.5 47.52 n.d..sup.1 −17.34 −19.10 −21.52 0.003 2.0 60.57 n.d..sup.1 12.30 n.d..sup.1 0.80 0.003 4.5 60.57 n.d..sup.1 4.83 n.d..sup.1 −3.87 .sup.1n.d. = not determined

(320) In principle, the values determined for 40° C. are comparable with those at room temperature.

(321) Overall, it can be concluded here too that no specific storage effect can be detected for glass at a temperature of 40° C. This also includes the two pHs tested 2.0 and 4.5.

(322) 2.3.5.2 Effect of Storage on Layer Thickness

(323) Below, the results in term of the layer thickness measurements carried out on silicon and glass using ellipsometry are presented.

(324) Each substrate was measured three times at the angles 65°, 70° and 75° using spectroscopic ellipsometry.

(325) The sample having the coating concentration 0.003 mol/l stood out in terms of their standard deviation in the ellipsometric measurements too. In addition, it can also be seen that as the concentration was increased further, the layer thickness on the silicon substrates did not increase by a measurable amount. However, this effect can be noted as a trend on glass substrates.

(326) The model assumed for the analysis corresponds to the explanations under section 1.4.4. The refractive index required for the SAM layer thickness measurement was taken from the literature and is 1.256 (cf. Jung, J.-I., J. Y. Bae, and B.-S. Bae, Characterization and mesostructure control of mesoporous fluorinated organosilicate films. Journal of Materials Chemistry, 2004. 14(13): pp. 1988-1994).

(327) 2.3.5.2.1 Changes to the Layer Thickness on Silicon when Stored at Room Temperature and at 40° C.

(328) The change in the layer thickness on silicon when stored at room temperature (25° C.) and at 40° C. was tested. The test results are summarised in Tables 16 and 17.

(329) TABLE-US-00016 TABLE 16 Changes to the layer thickness of the MTMS coating on silicon when stored at room temperature (25° C.) MTMS Layer thickness [nm] concentration Starting Change [mmol/l] pH value 30 days 90 days 180 days 0.03 2.0 0.93 −0.10 −0.02 −0.06 0.03 4.5 0.93 −0.10 −0.14 −0.11 0.015 2.0 1.06 −0.05 0.11 0.16 0.015 4.5 1.06 −0.02 0.03 −0.01 0.003 2.0 1.05 0.23 n.d..sup.1 0.17 0.003 4.5 1.05 0.15 n.d..sup.1 −0.02 .sup.1n.d. = not determined

(330) TABLE-US-00017 TABLE 17 Changes to the layer thickness of the MTMS coating on silicon when stored at 40° C. MTMS Layer thickness [nm] concentration Starting Change [mmol/l] pH value 30 days 90 days 180 days 0.03 2.0 0.93 0.04 −0.11 −0.08 0.03 4.5 0.93 0.01 −0.17 −0.14 0.015 2.0 1.06 0.00 0.04 0.07 0.015 4.5 1.06 −0.08 −0.04 −0.03 0.003 2.0 1.05 0.04 n.d..sup.1 0.16 0.003 4.5 1.05 0.01 n.d..sup.1 −0.07 .sup.1n.d. = not determined

(331) It can be seen that the coating produced on the silicon is very thin. For the starting values, the layer thickness is around 0.7 to 0.9 nm and thus corresponds to the values in the literature (cf. Plueddemann, E. P., Silane Coupling Agents, 2 ed. 1991, New York: Plenum Publishing Corporation). If the spread in the starting values and in the values upon removal from storage are taken into account, there is no discernible storage effect on the layer thickness. This relates to both the temperature change from room temperature (25° C.) to 40° C. and the change in pH from 2.0 to 4.5

(332) As regards the impact of the coating concentration, the aforementioned observation can be included in the development of the standard deviation. The spread in terms of the layer thickness also becomes more moderate as the coating concentration increases, which indicates a more homogeneous layer formation at higher concentrations.

(333) 2.3.5.2.2 Changes to the Layer Thickness on Glass when Stored at Room Temperature and at 40° C.

(334) The change in the layer thickness on glass when stored at room temperature (25° C.) and at 40° C. was tested. The test results are summarised in Tables 18 and 19.

(335) TABLE-US-00018 TABLE 18 Changes to the layer thickness of the MTMS coating on glass when stored at room temperature (25° C.) MTMS Layer thickness [nm] concentration Starting Change [mmol/l] pH value 30 days 90 days 180 days 0.03 2.0 1.50 0.29 −0.96 −0.26 0.03 4.5 1.50 −1.03 −0.27 −0.27 0.015 2.0 1.43 −0.10 −0.12 −0.13 0.015 4.5 1.43 −0.26 −0.14 −0.18 0.003 2.0 1.34 −0.19 n.d..sup.1 −0.36 0.003 4.5 1.34 0.16 n.d..sup.1 −0.28 .sup.1n.d. = not determined

(336) TABLE-US-00019 TABLE 19 Changes to the layer thickness of the MTMS coating on glass when stored at 40° C. MTMS Layer thickness [nm] concentration Starting Change [mmol/l] pH value 30 days 90 days 180 days 0.03 2.0 1.50 −0.68 −0.49 −0.64 0.03 4.5 1.50 −0.17 −0.28 −0.19 0.015 2.0 1.43 −0.01 −0.24 −0.58 0.015 4.5 1.43 −0.35 −0.09 −0.30 0.003 2.0 1.34 0.06 n.d..sup.1 −0.71 0.003 4.5 1.34 0.03 n.d..sup.1 −0.38 .sup.1n.d. = not determined

(337) It can be seen that the layer thickness of the film generated on glass is significantly greater than on silicon. When the starting values are considered, the thickness is approximately 2 nm and is thus more than twice as thick as the film generated on silicon. In this case, this is down to a higher density of binding sites (in particular OH bonds) between the generated film and glass (compared with silicon) and/or a greater degree of surface roughness on the glass substrate used (compared with the silicon substrate used).

(338) As has already been demonstrated, there is a discernible concentration effect in terms of layer thickness for the coating on glass. As the concentration of the coating solution increases, the layer thickness tends to become greater. However, this effect is insignificant and is concealed by a considerable degree of spread. Even for the highest coating concentration of 0.03 mol/l, the spread in the starting values is still close to around 0.5 nm.

(339) It can therefore be concluded that the layer is detectable and there are no discernible significant changes to the layer thickness due to the pH or temperature. In addition, an effect of the coating concentration on the layer thickness can be noted. For silicon, the use of higher coating concentrations resulted in a moderate spread within the layer thickness measurements. In the case of glass, at high coating concentrations, increasing the concentration tends to lead to higher layer thicknesses being detected (this effect was within the standard deviation of the measurements determined).

(340) 2.3.5.3 Stability Study Results

(341) The results of the stability study show that, under the framework conditions tested here, the coating can be deemed stable. This is demonstrated by the values from both the static contact angle measurement and the ellipsometric layer thickness measurements.

(342) For the substrates used, complete layer detachment was not identified under any of the tested stress conditions; however, the static contact angle measurements on silicon indeed show that the layer becomes slightly more hydrophilic over the storage period. This correlates well with the data from the provocation tests carried out since group III sprays occur in coated nozzles after a certain point of the test in this case too, which can also be considered an indicator of a possible layer thickness change. The rate of the increase in group III sprays in coated nozzles does not match that of an uncoated reference. In this respect, the increase in group III sprays is always much slower in coated nozzles, indicating the layer effect is still present.

(343) The results of the elliptometric layer thickness measurements also indicate a stable layer in all samples. The layer is always detectable and essentially no decrease in the layer thickness is noted.

(344) The results again show that glass and silicon substrates can be coated simultaneously in one coating method.

(345) 2.4 Effect of Surface Functionalisation on Atomiser Performance

(346) The effect of coated nozzles on atomiser performance was tested. The nozzles tested were transformed using 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane as a coating agent in accordance with section 1.1.

(347) 2.4.1 Effect of the Coating on Priming Behaviour

(348) The effect of coated nozzles on the priming behaviour of the atomiser was tested.

(349) Priming refers to the initial first operation of the device. In this case, the first five strokes immediately after the container used and filled with provocation solution was inserted were compared in terms of delivered mass and metered mass (this was an investigation of the priming behaviour, it being possible to discern the number of strokes it takes for the discharged weight to reach its complete or target value).

(350) 2.4.1.1 Priming Behaviour Progression: Delivered Mass (DM) and Metered Mass (MM)

(351) The priming behaviour of atomisers comprising coated and uncoated nozzle bodies was tested.

(352) It was found that, in terms of the priming behaviour progression in relation to delivered mass and metered mass, there was no difference between the results from the atomiser test groups having coated nozzle bodies and uncoated nozzle bodies. Therefore, the coating has no impact on the priming behaviour.

(353) 2.4.1.2 Effect on the Metering Behaviour: Comparison Between Delivered Mass and Metered Mass in 120-Day in-Use Mode

(354) The effect of coated nozzles on the metering behaviour was tested over an in-use period (provocation mode) of 120 days. The progression of the delivered mass (DM) and metered mass (MM) for atomisers 9 having coated and uncoated nozzles was determined. The determination of the delivered mass and metered mass can be found in section 1.4.3.

(355) 2.4.1.2.1 Delivered Mass

(356) The progression of the delivered mass for atomisers 9 having coated and uncoated nozzle bodies was tested over an in-use period of 120 days.

(357) It was clearly shown that, initially, there was no difference between the progression of the delivered mass when using devices having coated and uncoated nozzle bodies. However, this changed after day 20, after which point the test group of devices not having a coating on the nozzle body saw an increase in the delivered mass. By day 45, the increase in the delivered mass reached almost 12 mg and thus remained significantly above the average delivered mass for devices having a coating on the nozzle bodies.

(358) This increase was caused by the appearance of spray pattern anomalies in the group of atomisers not having a coating on the nozzle body. The test was run in provocation mode and, as it progressed, exhibited more group II sprays (spray pattern anomalies) and group III sprays in the reference group. Plaque deposits can indeed have an effect on the delivered mass since they cause deviations in the impact angle (the spray anomalies thus affect the formation of the impact disc in the DJI nozzles and thus also influence the aerosol backscattering or formation of residue droplets on the nozzle).

(359) Fundamentally, however, it was found that the two test groups display totally comparable metering behaviour. This is only changed by the occurrence of undesirable spray pattern anomalies.

(360) 2.4.1.2.2 Metered Mass

(361) The progression of the metered mass from atomisers having coated and uncoated nozzle bodies was tested over an in-use period of 120 days.

(362) In this case too, the tests carried out did not show any difference in the progression of the metered mass between atomisers having coated and uncoated nozzle bodies, but rather the results were completely comparable with one another.

(363) Therefore, it can be concluded that the coating does not have any effect on the metering behaviour of the atomiser. This relates to both the priming behaviour and the delivered and metered mass.

(364) 2.4.2 Effect of Coated Nozzles on Particle Size Distribution

(365) The effect of coated nozzles on particle size distribution was tested. Experiments were carried out on the progression of particle size distribution for atomisers having coated and uncoated nozzles. The particle size distribution was determined by measurements taken on the Andersen cascade impactor (according to Ph. Eur.) and via laser refraction (in this case, a Helos BF measurement instrument from Sympatec).

(366) The tested atomisers having coated and uncoated nozzle bodies had identical particle size distributions within the accuracy limits of the relevant measurement method.

(367) Experiments were also carried out on the duration of spray of atomisers fitted with coated and uncoated nozzles. In each case, ten devices were tested, with five individual measurements being taken on each one.

(368) It was found that the duration of spray for atomisers having coated and uncoated nozzles did not differ significantly. A duration of spray of 0.99±0.03 seconds was determined for coated nozzles, and a duration of spray of 0.96±0.03 was determined for uncoated nozzles. Therefore, within the measurement accuracy limits, no difference can be discerned in the duration of spray for coated and uncoated nozzles.

(369) 2.4.3 Overall Results Regarding the Effect of the Coating on Spray Performance

(370) No significant difference can be discerned between atomisers having coated nozzles and those having uncoated nozzles in terms of the device parameters tested. In this respect, the nozzles can be deemed identical.

(371) The results of this performance analysis of coated and uncoated nozzle bodies showed that the coating has no impact on the device parameters tested in this case. This relates to the priming behaviour, the metering accuracy, the particle size distribution and the duration of spray.

(372) In light of this, the method tested here for coating nozzles and nozzle bodies fulfils a basic requirement for measures intended to counteract the phenomenon of clogged nozzles: The possibility of the coating affecting the characteristic functional parameters of the atomiser can be ruled out.

(373) 2.5 Testing Other Coating Reagents: Effect of Alkyl Side Chain Length

(374) The tests described above show that coatings based on fluorinated silanes, in particular fluoroalkylsilanes are exceptionally suitable for preventing clogged or blocked nozzles. In addition, some non-fluorinated silanes, in particular alkylsilanes, showed promising results.

(375) The tests below are aimed at identifying alternative effective coating molecules.

(376) The focus of the following test is analysing the effect of alkyl side chain length. For this purpose, tests were carried out on coating molecules that tend towards the homologous series of alkanes in terms of their alkyl side chain. The test again focuses on alkylalkoxysilanes and alkyldimethylchlorosilanes.

(377) The first tests in relation to producing a successful coating will be carried out on the basis of silicon/glass planar substrates. On the basis of these samples, the homogeneity and hydrophobicity of the coating will then be characterised by means of static contact angle measurements. Using this data, a selection of coating molecules will be determined to be used subsequently for a provocation test.

(378) 2.5.1 Coating Silicon and Glass Planar Substrates

(379) The substrates are coated according to the procedure known for alkylalkoxysilanes and alkylchlorosilanes from section 1.1. Table 20 provides an overview of the coating reagents tested.

(380) TABLE-US-00020 TABLE 20 List of the alternative coating reagents tested Coating reagent Abbreviation Description Methyltrimethoxysilane C1 Homologous series Ethyltrimethoxysilane C2 n-butyltrimethoxysilane C4 n-octyltriethoxysilane C8 n-decyltriethoxysilane C10 n-dodecyltriethoxysilane C12 Trimethylchlorosilane C1-Cl Homologous series Ethyldimethylchlorosilane C2-Cl n-butyldimethylchlorosilane C4-Cl n-octyldimethylchlorosilane C8-Cl 1H,1H,2H,2H- F13C8-Cl Perfluorinated perfluorodecyldimethylchlorosilane

(381) 2.5.1.1 Screening the Coating Reagents Using Static Contact Angle Measurement

(382) Table 21 shows the results of the static contact angle measurements for silicon and glass planar substrates.

(383) TABLE-US-00021 TABLE 21 Averages together with standard deviation for water contact angle for additional coating reagents Glass Silicon Standard Standard Average deviation Average deviation Coating reagent [°] [°] [°] [°] Methyltrimethoxysilane 85.14 1.97 85.85 1.98 Ethyltrimethoxysilane 88.38 2.08 88.29 1.80 n-butyltrimethoxysilane 88.05 2.13 87.83 0.99 n-octyltriethoxysilane 108.35 1.39 103.38 1.15 n-decyltriethoxysilane 108.15 2.17 108.96 2.09 n-dodecyltriethoxysilane 108.19 1.89 104.45 2.88 Trimethylchlorosilane 91.51 4.77 86.20 6.55 Ethyldimethylchlorosilane 88.45 2.05 76.67 1.95 n-butyldimethylchlorosilane 90.46 2.20 81.38 0.88 n-octyldimethylchlorosilane 100.99 1.76 88.87 6.30 1H,1H,2H,2H-perfluoro- 115.36 2.703 110.47 1.158 decyldimethylchlorosilane 1H,1H,2H,2H-tridecafluoro- 110.47 2.20 105.40 2.88 octyltriethoxysilane (reference)

(384) Whereas for alkyltrialkoxysilanes an increase in the contact angle is recorded as the length of the alkyl chain increases, this observation cannot be made for alkylmonochlorosilanes, in which a non-uniform development was observed in the water contact angles in relation to chain length.

(385) In almost all the reagents, the contact angle on glass was always slightly higher than on silicon, i.e. for glass surfaces, there was a higher density of binding sites on the surface than for silicon surfaces. The largest contact angle was reached by the perfluorinated coating reagent F13C8-Cl. All the silicon/glass planar substrates tested showed stable contact angles with moderate standard deviation. No substrate showed layer detachment.

(386) For the provocation test carried out afterwards using atomiser devices, the following selection was made on the basis of the coating reagents tested here: methyltrimethoxysilane (C1) n-octyltriethoxysilane (C8) n-decyltriethoxysilane (C10) n-dodecyltriethoxysilane (C12) trimethylchlorosilane (C1-Cl) n-octyldimethylchlorosilane (C8-Cl) 1H,1H,2H,2H-perfluorooctyldimethylchlorosilane (F13C8-Cl)

(387) The selection makes it possible study the effect of the alkyl side chain and the effect of the coating chemistry. In addition, the selection includes coating reagents having very high and very low contact angles. Furthermore, this selection makes it possible check whether the contact angle is actually a suitable parameter for assessing whether a reagent is suitable for coating the DJI nozzles in question in order to prevent the phenomena of spray anomalies or jet divergency.

(388) The results of the contact angle study showed that the coating was fundamentally detectable on all the substrates. All the coating reagents provided stable contact angles above 80°, and with a very moderate spread. However, the results also showed that there were significant differences in the resultant water contact angle.

(389) In general, tightly packed, methyl-terminated monolayers have water contact angles of more than 110°. The contact angle becomes smaller as the molecules are packed less densely in the monolayer. This effect is presumably demonstrated in this case too in the small-chain alkyl chains C1, C2 and C2, and is thus representative of both the alkyltriethoxysilanes and alkyldimethylmonochlorosilanes. Overall, the reagents displayed water contact angles of less than 100° on both silicon and glass substrates.

(390) It was found that the water contact angle increases as the chain length increases. This effect can be noted in both glass and silicon for the two reagent classes tested. In relation to the materials used here, C10 and C12 also showed the highest contact angles; this is most likely due to the higher packing density of the resultant layer. Sieval et al. disclosed that the maximum load of Si (111) is generally only around 0.5-0.55 of the molecular modelling simulation (cf. Sieval, A. B., et al., Molecular modeling of covalently attached alkyl monolayers on the hydrogen-terminated Si (111) surface. Langmuir, 2001. 17(7): pp. 2172-2181).

(391) 2.5.2 Performance of Additional Coating Reagents in a Provocation Test

(392) As already known, the basic method for a provocation test can be taken from section 1.3.

(393) The provocation test is set up as follows:

(394) Formulation: as in 1.3.3

(395) Coating:

(396) Alkyltrialkoxysilanes: methyltrimethoxysilane (C1), n-octyltriethoxysilane (C8), n-decyltriethoxysilane (C10), n-dodecyltriethoxysilane (C12) 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane (F13-C8)

(397) Alkyldimethylchlorosilanes: trimethylchlorosilane (C1-Cl), n-octyldimethylchlorosilane (C8-Cl), 1H,1H,2H,2H-perfluorooctyldimethylchlorosilane (F1308-Cl), 1H,1H,2H,2H-perfluorodecyldimethylchlorosilane (F17010-Cl)

(398) Concentration: 0.03 mol/l in each case

(399) Number of inhalers: 30 for each reagent and reference

(400) In-use mode: 1×1 stroke/day

(401) Test parameters: Spray pattern according to spray pattern catalogue

(402) Test Duration:

(403) Decyltriethoxysilane (C10) and n-dodecyltriethoxysilane (C12), 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane (F13-C8): 120 days, all other reagents 28 days

(404) 2.5.2.1 Group I Spray Pattern Curve

(405) FIG. 17 shows the group I (“good sprays”) spray curves for the coated nozzles and the uncoated reference.

(406) The figure shows that the long-chain alkylalkoxysilanes ensured a high number of group I sprays (“good sprays”) for a long period of time, whereas all the other coating reagents showed no advantage over the uncoated reference. Within the alkylalkoxysilanes tested here, C12 (n-dodecyltriethoxysilane) showed the best results. In this test, it even had a slight advantage over the perfluorinated F1308 (1H,1H,2H,2H-tridecafluorooctyltriethoxysilane). The alkyl monochlorosilanes also performed relatively poorly again in this test. The test result from section 2.2.3 is thus confirmed and can even be extended to cover fluorinated compounds. In this test too, there was no advantage over the uncoated reference.

(407) This therefore shows that the best spray performance is ensured by coatings having an alkyltrialkoxysilane (both fluorinated and non-fluorinated) having a long side chain (i.e. in this case by C.sub.10 and C.sub.12 chain lengths tested here).

(408) 2.5.2.2 Group II Spray Pattern Curve

(409) FIG. 18 shows the group II spray curves (spray pattern anomalies) for the coated nozzles and the uncoated reference.

(410) FIG. 18 very clearly shows that the long-chain alkylalkoxysilanes produced significantly fewer group II sprays than the chlorosilanes or short-chain alkylalkoxysilanes. It can also be seen here that both the short-chain alkoxysilanes and all the chlorosilanes barely brought any benefit compared with the uncoated reference. The long-chain alkylalkoxysilanes had a clear advantage over all the other reagents tested.

(411) 2.5.2.3 Group III Spray Pattern Curve

(412) FIG. 19 shows the group III spray (“jet divergency”) curves for the coated nozzles and the uncoated reference.

(413) Examining the group III spray curve also confirms the idea gained from the previous spray pattern curves. The long-chain alkylalkoxysilanes did not once pass the 20% mark over 120 days of in-use time, whereas all the other reagents tested already exceeded this level after around 20 days.

(414) 2.5.2.4 Group III Sprays: 10-Day Average Rate Towards the End of the Test

(415) The advantage that alkylalkoxysilanes have over the alkylmonochlorosilanes is particularly evident again when examining the 10-day average rate of the group III sprays, as shown in FIG. 20.

(416) FIG. 20 shows the 10-day average rate of group III sprays for the alternative coating reagents after day 28. In this case, the uncoated reference reached a rate of around 40% group III sprays. The coatings using C1 (methyltrimethoxysilane), F1308-Cl (1H,1H,2H,2H-perfluorooctyldimethylchlorosilane) and F17010-Cl (1H,1H,2H,2H-perfluorodecyldimethylchlorosilane) are similarly poor. They show no advantage in terms of their coating. The aforementioned coating reagents F1308 (1H,1H,2H,2H-tridecafluorooctyltriethoxysilane), C10 (n-decyltriethoxysilane) and C12 (n-dodecyltriethoxysilane) achieve very good performance. For the cartridge period shown here (28-day mode), F1308 (1H,1H,2H,2H-tridecafluorooctyltriethoxysilane) and C12 (n-dodecyltriethoxysilane) was able to largely prevent the occurrence of group III sprays.

(417) The results show very good performance for the alkyltrialkoxysilane group. The relatively poor performance of alkyl monochlorosilanes is surprising, all the more so since stable contact angles were achieved in the study carried out previously. No member of the group of tested alkylchlorosilanes could reach a group III spray rate of less than 10%. This is in contrast to the alkyltrialkoxysilanes, three members of which achieved a value of 5% for the 10-day average group III spray rate.

(418) 2.5.2.5 Results of the Provocation Test

(419) As the results of the provocation test show, the alkyldimethylmonochlorosilanes performed relatively poorly, similarly to in section 2.2.3. For the alkyldimethylmonochlorosilanes, the 10-day average group III spray rate was much higher than the longer-chain alkyltrialkoxysilanes. Surprisingly, in this test methyltrimethoxysilane also showed similar performance to the alkyldimethylmonochlorosilanes.

(420) The reasons for the poor performance of methyltrimethoxysilane are unclear. However, the spray pattern curve showed that the progression was very similar to the range of the uncoated reference over the entire duration of the test.

(421) One possible explanation for the poor result for methyltrimethoxysilane could be an absent or insufficiently stable coating, or inadequate packing density.

(422) The alkylmonochlorosilanes had a similar curve to the uncoated reference, although some members of the group did indeed show a slight layer effect. Looking at the 10-day average rate for group III sprays, a slight effect can be identified for trimethylchlorosilane (C1-Cl) or n-octyldimethylmonochlorosilane (C8-Cl).

(423) The alkyltrialkoxysilanes have two sites more than the alkylmonochlorosilanes for coordinating with the surface. Therefore, they are bound to the surface much more strongly than the alkyldimethylmonochlorosilanes. This would explain the systematic difference between the alkylmonochlorosilanes and the alkyltrialkoxysilanes. However, this possible explanation does not apply to the poor performance of methyltrialkoxysilane since this reagent also has three sites for coordinating with the surface. Layer erosion or the aforementioned effect of the alkyl chain length on the packing density are assumed here too.

(424) The provocation tests showed that n-decyltriethoxysilane (C10) and n-dodecyltriethoxysilane (C12) were well suited as coating reagents. They display similarly good results in their spray pattern curves as 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane (F1308). Examining the 10-day average rate, it can be seen that n-dodecyltriethoxysilane (C12) was even slightly better. These results show that non-fluorinated alkylsilanes are a serious alternative to fluorinated silanes. This is surprising but fluoroalkylsilanes generally do produce the better anti-adhesion effect (cf. Giessler, S., E. Just, and R. Stôrger, Easy-to-clean properties—Just a temporary appearance, Thin Solid Films, 2006. 502(1): pp. 252-256). However, the data from the provocation test shows that a long alkyl chain length in the range of 010 and C12 was sufficient.

(425) The data collected here also leads to the conclusion that a higher water contact angle alone is not necessarily an indicator of an effective coating reagent. In the previous study, the perfluorinated chlorosilane F17010-Cl had showed excellent contact angles of around 110° but performed very badly in the subsequent provocation test. Therefore, it is not possible to use the contact angle alone to deduce the spray performance in the provocation test, since the stability of the coating and its packing density also have to be considered. However, a high contact angle as defined by Ishizaki et al. can indicate a high packing density.

(426) 2.5.2.6 Yield of Category I Nozzles from Coating with Alternative Coating Reagents

(427) FIG. 21 shows the results of the microscopic analysis of the coated nozzle bodies in accordance with the nozzle body categorisation catalogue.

(428) It can be seen in FIG. 21 that the coating reagent F1308 (1H,1H,2H,2H-tridecafluorooctyltriethoxysilane) generates the cleanest nozzles under the laboratory conditions tested. More than 90% of the nozzles tested here were classed as category I. The yield of category I nozzles is only around 25% for C12 (n-dodecyltriethoxysilane), which actually performed better than Dynasylan® F8261 in the provocation test. C10 (n-decyltriethoxysilane) also achieves a very good category I nozzle yield of around 55%.

(429) 1H,1H,2H,2H-tridecafluorooctyltriethoxysilane generates the most clean category I nozzles by far. In this test, the yield was surprisingly high at 90% and was achieved by additionally separating the nozzles at the drying phase before tempering.

(430) In principle, the yield of clean nozzles could be increased further by using additional automated process steps. This could be done, for example, by using an aforementioned spin rinse dryer, which expels residual solution from the nozzle by means of rotation. Alternatively or additionally, rinsing processes, e.g. using alcoholic solvents, can be carried out. This should further increase the yield of clean nozzles since much less residual solution is present in the nozzle bodies when they are dried further.

(431) 2.5.2.7 Overall Results in Terms of Additional Coating Reagents

(432) In conclusion, it is evident that the best results are achieved using a long-chain alkylalkoxysilane. In this test, good alternatives to the fluorinated compounds were found, namely n-decyltriethoxysilane (C10) and n-dodecyltriethoxysilane (C12).

LIST OF REFERENCE NUMERALS

(433) TABLE-US-00022 1 microstructured component 2 inlet opening 3 outlet opening 4 channels 5 fine filter 6 plenary chamber 7 column structure 8 coating 9 atomiser 10 liquid 11 aerosol 12 pressure chamber 13 upper housing part 14 inner housing part 15 lower housing part 16 container 17 mainspring 18 pressure generator 19 locking ring 20 button 21 tubular piston 22 return valve 23 mount 24 meter 25 filter system