Patent Grant
8936160
The present invention relates to a nozzle device having a nozzle for atomisation of a fluid, the nozzle comprising a nozzle plate support body having a cavity extending from a first main surface to a second main surface thereof, and comprising a nozzle plate having at least one nozzle orifice in fluid communication with said cavity at said first main surface side of said nozzle plate support body. The invention further relates to a nozzles as used in such a nozzle device.
These devices are used for filtration purposes and or for atomisation of a fluid to produce small liquid droplets in air (spray) or into a liquid (emulsion) with a relatively narrow droplet size distribution and to make small air bubbles into a liquid (foam) and to methods of using the same. The device and especially the nozzle plate may be produced by micro-machining (Micro System Technology) which means that the subject nozzle part means are produced using lithography steps related to semiconductor fabrication methods. Alternatively spark erosion and laser drilling techniques may be used, but in general these tend to be less reproducible and less precise in comparison with micro-machining methods.
The performance of many atomisation devices can be improved if the atomising device provides very small droplets with a very narrow pore size distribution.
For example, small droplets between 2 and 3 micron in diameter improve the effectiveness of medical atomisers because of the high (80%) deposition intake deep into the lungs. Also the stability of an emulsion (o/w, w/o) is greatly improved if the emulsion droplets are all of equal size. Besides that, the structural and rheological properties of many foams in the dairy industry can be improved by the use of very small air bubbles with a narrow size distribution.
The disadvantage of many conventional atomising devices is that they break bulk liquid or gas into relatively large droplets through use of stirring or turbulence. By more input of energy the large droplets will be broken up in smaller droplets. As the droplets become smaller than 20-100 microns, they become harder to break and secondary atomisation typically ceases. The droplet size distribution is in most cases rather broad.
It is known from fuel injectors that nozzle structures may be used for obtaining a very fine spray for combustion improvement. Such small nozzle structures however are very sensitive for fouling and unwanted leakage due to blocked nozzle orifices. For a high throughput of equally sized droplets normally an array of identical nozzles is used. However if one or more nozzle orifices becomes blocked the size distribution will broaden. If a nozzle orifice becomes smaller through partial blockage the droplets of this orifice will also become smaller. Moreover if the blockage is very severe spraying(or jetting) will cease and liquid will flow through this orifice over the surface of the nozzle structure hence influencing or inhibiting spraying behaviour of the other orifices.
It is also known that very small nozzles suffer from a threshold pressure (Pascal pressure/capillary forces) before they start spraying. The threshold pressure is inversely proportional to the nozzle diameter. For a nozzle with a diameter of 1 micron this pressure is typical 1-3 bar. For an array of nozzles it is therefore very important that all nozzles have an equal geometry with narrow tolerances and that the threshold pressure is kept as low as possible.
A high flow rate can be achieved by choosing the flow resistance of each nozzle orifice as small as possible and/or by increasing the pressure difference over the orifice during jetting. Practically the jetting pressures are chosen to be fairly higher than typical 5-10 bar. Such pressures will exert high forces on the nozzle plate. The nozzle plate is therefore chosen fairly thick (>4-5 micron) in order to withstand such forces. However a thick nozzle plate implies a long orifice length and thus a high flow resistance and subsequently a reduced flow rate.
It is inter alia an object of the invention to provide a nozzle device and a nozzle of the type referred to in the opeing paragraph in which these drawbacks have been counteracted at least to an impressive exntend.
To this end a nozzle device as described in the opening paragraph is according to the invention characterized in that said support body is provided with filtration means which comprise a filtration plate which is in fluid communication with said cavity at said second main surface side of said nozzle plate support body.
A further object of the present invention is to produce a properly constructed nozzle plate for atomisation at operational pressures smaller than 10 bar.
Another object of the present invention is to provide nozzle plates that produce droplets typically with a mean diameter of 10 micron or smaller with a very narrow droplet distribution.
Yet another object of the present invention is to provide nozzle plates for small handheld atomising devices with a throughput nearly independent of the viscosity of the fluid (e.g. medicine) and means to reproducible facilitate atomisation.
Yet another object of the present invention is to produce a properly constructed nozzle plate (filtration membrane) for filtration of small and large amounts of liquid or gas and means to facilitate filtration with such a filtration membrane, which may be used in combination with atomisation applications.
Yet another object of the invention is to provide nozzle plates for large atomising devices capable of substantial throughput of atomised liquid or gas.
Yet another object of the invention is to provide nozzle plates with orifices with a reduced flow resistance that can withstand high operational pressures.
Yet another object of the invention is to provide atomising devices that are rather insensitive for microbiological fouling and unwanted leakage due to blocked nozzle orifices.
Yet another object of the invention is to provide atomising devices that are less sensitive for the Pascal threshold pressure.
These and additional objects and advantages of the invention will become apparent from the technical description which follows.
It is to be understood that both the foregoing summary and the following technical description are exemplary and explanatory and are not restrictive of the invention as claimed.
A first embodiment of a nozzle device 1 is shown in
Another embodiment of a nozzle device 1 is shown in
Nozzles for atomisation 2 can be made with known micro machining techniques. A mono crystalline silicon wafer 12 with thickness 400 micron is provided with a Low Pressure Chemical Vapour Deposition grown layer 10 of low stress silicon nitride with a thickness of 1 micron. With a suitable mask a photo lacquer pattern with 2 micron orifices at the front side of the wafer 12 and a similar pattern with 15 micron openings at the back side is being exposed and developed. With the aid of anisotropic reactive ion etching a nozzle orifice 11 with a diameter of 2 micron and a length of 1 micron is made in the silicon nitride layer and with use of dry and wet chemical KOH etching a cavity 13 with a diameter of 15 micron and a length of 400 micron is made in the silicon wafer 12.
The flow rate Φ of a medium or a liquid with viscosity η through an orifice (tube) with length L and diameter D for viscous flow at a pressure difference ΔP is given by the law of Poiseuille: ΦPoiseuille=πD4ΔP/128Lη. A parabolic velocity pattern with a low velocity along the wall and a high velocity in the middle of the tube will settle in case the length of the tube L is larger than typical six times the diameter D. The mean velocity v of the medium or liquid is always given by v=4Φ/πD2.
In case the length L is of the order of the diameter D the law of Poiseuille will change to the law of Stokes: Φstokes=D3ΔP/24η. The parabolic velocity pattern will not be valid in this regime. Dagan et al., Chem. Eng. Svi., 38 (1983) 583-596 have proposed an interpolation formula for both regimes: ΦDagan=D3ΔP/24η [1+16L/3πD]−1.
At large velocities v the viscous regime will not be valid any more because another force/pressure is necessary for a kinetic (inertial) contribution to accelerate the medium or fluid to a velocity v. This pressure difference is given by ΔPkin=0.5ρv2, with ρ the mass density of the fluid v (cf. Law of Bernoulli).
An important insight according to the invention is that the total needed pressure ΔPtot is the sum of the viscous and the kinetic contribution:
Ptot=ΔPvis+ΔPkin=6ηπ[D+16L/3π]−1 v+0.5 ρ v2.
Typical for a waterbased fluid and for a thin orifice this means:
ΔPtot≈18.000D−1 v+500 v2
(L<D, n=10−3 poise, ρ=1000 kg/m3, D in micron, v in m/s, ΔP in Pascal). At a ΔPkin of 4 bar (=4×105 Pascal) the maximum jet velocity will be 28 m/s. ΔPvis will be for this velocity 500.000 D−1. In case D>2 micron than ΔPvis<2.5 bar. ΔPtot is then 4+2.5=6.5 bar, less than the maximum of 10 bar. However in case L/D>2 then at D=2 micron the needed pressure will surely exceed 10 bar.
Another important insight is that with a very thin orifice (L≈D≈micron) both in the viscous and in the kinetic regime all the fluid will leave the orifice as a jet with constant velocity v (no parabolic velocity distribution). Especially the kinetic energy of the jet will make that the jet will prolong its track before it breaks up in small droplets, which is particularly useful for Rayleigh break-up of the jet in droplets in air. Rayleigh droplets have a typical droplet size 1.6 times the diameter D of the out coming jet. The fabrication tolerance in the diameter D of the nozzle orifice is an essential factor in determining the amount of liquid (ΔV=4π(1.6D/2)3/3) in a Rayleigh droplet. The United States FDA imposes a repeatability of 20% for 90% of the droplets and 25% for the remaining 10%. Only micro machining methods are capable of producing orifices with a tolerance less than 3% (=variation in ΔV<10%). Also because micro machining is done in a sterile and particle free Clean Room environment also the effect of fouling of the nozzles due to particles and/or micro organisms is avoided.
Another important insight according to the invention is that for a very thin orifice(L≈D≈micron) the flow rate at relatively low pressures (3-10 bar) is mainly determined by the kinetic contribution, which means that viscosity of the fluid (medicine) has a minor role as long as L≈D and η<10−2 poise. Jetting (e.g. Rayleigh break-up) with a nozzle plate with a thickness less than 2 micron and orifices with a diameter between 0.4 and 10 micron at a pressure in which the contribution of the kinetic regime (0.5ρv2.) is larger than the contribution of the viscous regime (6ηπ[D+16L/3π]−1.v) is therefore a very good method to deliver and dose medicines nearly independent of the viscosity of the medicine.
Another important insight according to the invention is that medicine (e.g. proteins and peptides) degradation is strongly diminished if such thin orifices are used at relatively low jetting pressures (<10 bar) with a minimum of shear in strength, time and length of the medicine in passing such an orifice.
Using the law of Stokes and Poiseuille (or Dagan) it is easily to calculate that the flow resistance of the 2 micron orifice 11 is still 5-10 times higher than the flow resistance of the cavity 13 with diameter 15 micron and length 400 micron. This means that the pressure/flow characteristics of this structure is still mainly determined by the 2 micron orifice.
In preference the thickness of the nozzle plate 10 for atomisation is less than six times the diameter of the nozzle orifice 11 and in preference less than one to two times the diameter in order to prevent the built up of a parabolic velocity distribution. The flow resistance may be further reduced through the manufacturing of tapering orifices although it is well known that the amount of tapering is very difficult to control precisely. In case the nozzle plate 10 has a thickness less than 2 micron it still has sufficient strength and it is not necessary to taper the orifices.
Nozzles can be used for as well atomisation and filtration. An embodiment of a nozzle with a nozzle orifice for atomisation or filtration 4 is shown in
The nozzle 4 comprises further at least one shallow flow channel 44 connected to the nozzle plate 40 with a mean depth of minimum 10 and of maximum 300 micron connected to the nozzle plate. This depth 43 is dependent on the size and number of the nozzle orifices 41 in the nozzle plate 40. The flow resistance of the flow channel 44 in the nozzle plate support should be at least one to ten times smaller than the flow resistance of the nozzle plate 40 itself. In case the total flow resistance of the nozzle plate support as defined by regions 44 and 45 is one to five times the flow resistance of the nozzle plate 40 a nice flow limitation has been constructed in case the nozzle plate 40 would disrupt. Alternatively two or more openings 46,47 can be provided in each nozzle plate to promote fluid flow and the removal of particles and air bubbles underneath the nozzle plate 40.
Cross-flow cleaning 90,91 on both sides of the nozzle plate is enhanced by the interconnection 81 in one or more directions of all nozzle plate support flow channels 44 (FIG. 7A,7B). Silicon bars 92 between the nozzle plates 40 may be provided for enhanced strength.
Subsequently the nozzle plate 50 may be chosen thicker than a few micron with corresponding tapering orifices 51 in order to reduce the flow resistance still further, shown in
With preference a number of nozzle orifices 61 are placed very close together (
Nozzle plates can be made substantially stronger (up to 250%) when the nearest distance 100 between all nozzle orifices and the nozzle plate support is at least six times the thickness of the nozzle plate
A next embodiment of a nozzle for atomisation is shown in
Jetting may be enhanced by using a piezoelectric actuator at a frequency between 100 kHz and 3 Mhz. Jetting may also be enhanced using the eigenresonance frequencies of the nozzle plate. This frequency should match the value of the initial jet velocity divided by two to two hundred times the diameter of the nozzle, typically a value between 50 kHz en 5 MHz. The eigenresonance frequency is mainly determined by the mass and a fortiori lateral dimensions of the free hanging nozzle plate (typical 1×10×10 micron to 4×250×2000 micron), the rigidity of and the tensile pre-stress in the nozzle plate (typical 106to 109 Pascal). A vibrating nozzle plate 150 is shown in
The nozzle plate may also be used for retaining and subsequent microscopic observation 192 of these particles, e.g. bacteria's, yeast cell's, blood cell's, etc. Fluorescent dyes may be used to simplify and identify specific species of the micro-organisms on the filter. Silicon nitride and other inorganic nozzle plate materials have the advantage in contrast to many organic polymeric materials that there is virtually no auto-fluorescence signal from the material itself. In some cases it is convenient to place the nozzle orifices further apart, in order to isolate the micro-organisms from each other for a more easy recognition and enumeration.
Very useful nozzle plates for this purpose are characterised in that the spacing 200 between the nozzle orifices is minimum three and maximum thirty times the diameter of the nozzle orifices.
Filter means or nozzles may be used for disposable filtration applications, with preference small nozzle plates 220 (e.g. 5×5 mm) are embedded in a ring shaped support 221 (e.g. ABS plastic discs) with outer dimension of e.g. 1.0, 2.5 and 5 cm in diameter and ready to use in standardised commercial filtration holders. With preference the nozzle plates are countersink 222 with a depth of 10 to 500 micron in the ring shaped support to prevent contamination, to facilitate packaging and mechanical rupture of the nozzle plate (
For reusable application an optic transparent cover slip 230 is placed over the nozzle plate in such a way that a cross-flow channel 231 with a depth of 50 to 500 micron exists between the nozzle plate and the cover plate (
With preference the nozzle plate support body has cavities 233 with at least the same size as the nozzle plate. It is then possible to use a microscope 192 with a light source that projects light 193 first through the nozzle support and next on the nozzle plate. Most microscopes with phase contrast mode work in this manner.
Large nozzle plates with an outer circular diameter of e.g. 2, 3, 4, 6 and 8 inches may be used for micro filtration applications like yeast cell filtration and clarification of beer and other beverages. Sterile filtration of milk and other dairy products is also possible with pore sizes between 5 and 0.22 micron. With a pore size of 0.8 micron it has been tested that a log reduction of 5 to 6 of micro-organisms in milk is well achievable in combination with back-pulse (pulsed permeate flow reversal) technology. Typical flow rates are 1000-2000 l/m2/hour at low trans-membrane pressures (0.03-0.1 bar) with a back-pulse rate of 0.01-5 Hz. The flow rate can be strongly increased (4000-20.000 l/m2/hour) using ultrasound in a broad frequency spectrum between 100 Hz-1 MHz. Preferably a frequency is used under 15 kHz or above 50 kHz in order to suppress the cavitation forces that might disrupt the nozzle plates between 15 kHz and 50 kHz. The ultrasound inhibits the forming of a dense cake layer just before the nozzle plate. Alternatively the performance for jetting, filtering, foaming and emulsification may be improved by moving the nozzle plate tangential and/or orthogonal to the fluid in contact with the nozzle plate with an actuator with an amplitude of 0.1 to 100 micron and a frequency of 10 Hz-10 MHz.
In a special embodiment the nozzle plates or nozzle plate support bodies are bonded to a glass plate in which flow channels 270,284 have been made with the use of grinding or powder blasting (
Nozzle plates made with a silicon support can be made chemically inert for caustic media by providing a thin LPCVD grown silicon nitride coating with a typical thickness between 0.01 and 1 micron. Other organic and inorganic coatings like e.g. Al2O3, TiO2, ZrO2, ZrO2/Si3N4 may be applied to alter the Zeta potential and/or the wetting properties of the nozzle plate to improve filtration characteristics. Other coatings may also be applied to promote anti-fouling like TiO2, PTFE, self assembling monolayers (SAM, e.g. based on nitryls, disulfides or thiols) or long polymer chains (e.g. polyethyleneglycol) coupled with an end- or side-group to the nozzleplate. Dense sol/gel coatings or gas permeation layers like Pd, PdAg may also be applied over and in the nozzle orifices to make ultrafiltration and gas filtration membranes.
An important insight according to the invention is that the combination of nozzle plates, back-pulse technology and ultrasound has proven to be very powerful for the enhancement of flow rate and the prevention of irreversible fouling. Without ultrasound a typical clarification run for beer is 4-8 hours, with ultrasound dosed at intervals of 10 minutes for a few seconds the run can be extended to 4-8 days without the need of chemical cleaning procedures.
Backpulsing for a very short time 10-50 ms at regular intervals 0.01-5 Hz during cross-flow filtration at low trans-membrane pressure will lift the cake layer from the nozzle plate and will inject it higher in the cross flow channel where the fluid velocity is sufficient high to take it further away.
Backpulsers are also very suitable to use for up-concentration of samples for the detection and counting of food spoiling or pathogenic micro-organisms, e.g. lacto bacillus, E-coli and legionella. After the up-concentration all micro-organisms are present on the nozzle plate and can be processed for e.g. microscopic observation and PCR amplification. Small nozzle plates of e.g. 4×4 mm can be put easily with a clean and sterile pincer in a small PCR-cup. The nozzle plate can also be provided with an immuno binding (or elisa coupling) agent for the selective binding of certain species direct to the nozzle plate during filtration, especially when cross-flow techniques are used for up-concentration of the sample. Magnetic layers may also be deposited for the attraction of immuno magnetic beads. Metallic layers may also be provided on the nozzle plates for e.g. optic non-transparancy, non quenching or electrolysis applications, improvement of filtration under the applicance of a small voltage difference between the fluid and the nozzle plate, or the annihilation(electroporation) of microorganisms under the applicance of a high voltage pulse. Platina may be deposited in electrical resistor strips on the nozzle plate for heating purposes. Also a bacteria killing surface modification may be applied, for example a silver coating. Piezo materials may also be applied for direct vibration of the nozzle plates or for the detection of bending of the nozzle plates for pressure registration. The intensity and the frequency of the backpulsers may also be regulated by the registration of the nozzle plate trans membrane pressure. The trans membrane pressure will normally increase if there is a built up of a cake layer for the nozzle plate.
Nozzle plates can be made in various ways according to the invention.
A reinforced micromachined polymeric nozzle plate is made by
Another method of making a micromachined polymeric nozzle plate, comprises the following steps
A reinforced micromachined electroformed nozzle plate is made by
Another method of making a micromachined nozzle plate device comprises the following steps
Nozzle plates according to the invention may also be used for the extrusion of very viscous media like macromolecular solutions, gel-like solutions and protein-rich media, and for microstructuring of food and pharmaceutical products like e.g. synthetic meat (fibres).
Nozzle plates according to the invention may also used for micro-array and micro-titration applications, to make double emulsions and to apply them in bio-capsules because of the small diffusion length of the short nozzle orifice.
| Number | Date | Country | Kind |
|---|---|---|---|
| 1016030 | Aug 2000 | NL | national |
This application is a division of co-pending application Ser. No. 12/073,387, filed on Mar. 5, 2008, which is a division of co-pending application Ser. No. 11/101,391, filed on Apr. 8, 2005. Application Ser. No. 11/101,391 is a division of Ser. No. 10/362,762, filed on Feb. 26, 2003, which is is the national phase of PCT International Application No. PCT/NL01/00630 filed on Aug. 28, 2001 under 35 U.S.C. §371, which claims priority of Netherlands Application No. 1016030 filed Aug. 28, 2000. The entire contents of each of the above-identified applications are hereby incorporated by reference.
| Number | Name | Date | Kind |
|---|---|---|---|
| 3586907 | Beam et al. | Jun 1971 | A |
| 4601777 | Hawkins et al. | Jul 1986 | A |
| 4628576 | Giachino et al. | Dec 1986 | A |
| 4768751 | Giachino et al. | Sep 1988 | A |
| 4789425 | Drake et al. | Dec 1988 | A |
| 4828184 | Gardner et al. | May 1989 | A |
| 4864329 | Kneezel et al. | Sep 1989 | A |
| 4871489 | Ketcham | Oct 1989 | A |
| 4875968 | O'Neill et al. | Oct 1989 | A |
| 5002230 | Norskav et al. | Mar 1991 | A |
| 5124717 | Campanelli et al. | Jun 1992 | A |
| 5201987 | Hawkins et al. | Apr 1993 | A |
| 5204690 | Lorenze et al. | Apr 1993 | A |
| 5320290 | Rohs et al. | Jun 1994 | A |
| 5543046 | Van Rijn | Aug 1996 | A |
| 5651900 | Keller et al. | Jul 1997 | A |
| 5753014 | Van Rijn | May 1998 | A |
| 5925205 | Zimmermann et al. | Jul 1999 | A |
| 6016969 | Tilton et al. | Jan 2000 | A |
| 6036105 | Sanada et al. | Mar 2000 | A |
| 6036832 | Knol et al. | Mar 2000 | A |
| 6044981 | Chu et al. | Apr 2000 | A |
| 6084618 | Baker | Jul 2000 | A |
| 6086195 | Bohorquez et al. | Jul 2000 | A |
| 6113976 | Chiou et al. | Sep 2000 | A |
| 6130688 | Agarwal et al. | Oct 2000 | A |
| 6189214 | Skeath et al. | Feb 2001 | B1 |
| 6189813 | Skeath et al. | Feb 2001 | B1 |
| 6207369 | Wohlstadter et al. | Mar 2001 | B1 |
| 6352209 | Skeath et al. | Mar 2002 | B1 |
| 6378788 | Skeath et al. | Apr 2002 | B1 |
| 6464347 | Kneezel et al. | Oct 2002 | B2 |
| 6513736 | Skeath et al. | Feb 2003 | B1 |
| 6652077 | Maeng et al. | Nov 2003 | B2 |
| 6679101 | Rohner | Jan 2004 | B1 |
| 6756019 | Dubrow et al. | Jun 2004 | B1 |
| 6769765 | Kneezel et al. | Aug 2004 | B2 |
| 6780340 | Conta | Aug 2004 | B2 |
| 6797945 | Berggren et al. | Sep 2004 | B2 |
| 6849459 | Gilbert et al. | Feb 2005 | B2 |
| 6878271 | Gilbert et al. | Apr 2005 | B2 |
| 7094345 | Gilbert et al. | Aug 2006 | B2 |
| 7531120 | Van Rijn et al. | May 2009 | B2 |
| 7963466 | Van Rijn et al. | Jun 2011 | B2 |
| 20010019029 | Tai et al. | Sep 2001 | A1 |
| 20030150791 | Cho et al. | Aug 2003 | A1 |
| 20030178507 | Maria Rijn Van | Sep 2003 | A1 |
| 20040028875 | Van Rijn et al. | Feb 2004 | A1 |
| 20050115889 | Schaevitz et al. | Jun 2005 | A1 |
| 20050178862 | Van Rijn | Aug 2005 | A1 |
| 20070227591 | Wissink et al. | Oct 2007 | A1 |
| 20080217262 | Van Rijn | Sep 2008 | A1 |
| 20080248182 | Jongsma et al. | Oct 2008 | A1 |
| 20120228238 | Van Rijn et al. | Sep 2012 | A1 |
| Number | Date | Country |
|---|---|---|
| 879635 | Nov 1998 | EP |
| 9323154 | Nov 1993 | WO |
| 9513860 | May 1995 | WO |
| 9740213 | Oct 1997 | WO |
| 9801228 | Jan 1998 | WO |
| 9801705 | Jan 1998 | WO |
| 2005105276 | Nov 2005 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20120228238 A1 | Sep 2012 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 12073387 | Mar 2008 | US |
| Child | 13448048 | US | |
| Parent | 11101391 | Apr 2005 | US |
| Child | 12073387 | US | |
| Parent | 10362761 | US | |
| Child | 11101391 | US |
