Enhanced aerosol test for assessing filter integrity

Abstract

A method of aerosol integrity testing of filters, capable of detecting single defects that are less than 20 m in diameter, and even as small as 2 m in diameter, in liquid sterilizing grade filters such as filter cartridges. The method challenges the filter in a dry state with a particle stream of aerosol particles of the appropriate size and in the appropriate concentration, such that at least one or more of the particles in the stream will penetrate a defective region or regions within the membrane but will not penetrate in the integral region of the membrane. Wetting of the filter is not required.

Claims

1. A method of integrity testing a filter in the dry state, comprising: providing a liquid sterilizing grade filter to be tested in a dry state, wherein said liquid sterilizing grade filter is a filter capable of totally retaining a B. diminuta challenge level of 10.sup.7 cfu/cm.sup.2 at a differential pressure of 30 psi; generating an aerosol particle stream at a pressure of at least 5 psig, wherein the particles in said stream have a suitable size and a suitable concentration to challenge said filter and penetrate any defective region in said filter but will not penetrate integral regions in said filter, wherein said concentration of said particles in said stream is at least 10.sup.3 particles/cm.sup.3; and detecting particles that penetrate said defective region.

2. The method of claim 1, wherein defects in said filter that are as small as 2 microns are detected by said step of detecting particles that penetrate said defective region.

3. The method of claim 1, wherein said filter is pleated.

4. The method of claim 3, wherein said aerosol particle stream is generated with a plurality of atomizers at a pressure of 50 psig.

5. The method of claim 1, wherein said filter is housed in a cartridge.

6. The method of claim 1, wherein said aerosol particle stream comprises NaCl.

7. The method of claim 1, wherein said filter is a PVDF filter.

8. The method of claim 1, wherein said filter is a PES filter.

9. The method of claim 1, wherein said particles in said aerosol particle stream have a size and concentration to penetrate a defect less than 20 m in diameter in said filter.

10. The method of claim 1, wherein the aerosol particle stream is generated at a pressure of 5-60 psig.

11. The method of claim 1, wherein said aerosol particle stream is generated by a plurality of atomizers.

12. The method of claim 1, wherein the particles in said stream have a concentration of 10.sup.6 particles/cm.sup.3.

13. The method of claim 1, wherein the particles in said stream have a concentration in the range of 10.sup.5-10.sup.7 particles/cm.sup.3.

14. The method of claim 1, wherein said filter is a 0.2 m rated filter.

15. A method of integrity testing a filter in the dry state, comprising: providing a liquid sterilizing grade filter to be tested in a dry state, wherein said liquid sterilizing grade filter is a filter capable of totally retaining a B. diminuta challenge level of 10.sup.7 cfu/cm.sup.2 at a differential pressure of 30 psi; generating an aerosol particle stream with at least one atomizer at a pressure of at least 5 psig, wherein the particles in said stream have a concentration of at least 10.sup.3 particles/cm.sup.3 and have a size effective to penetrate a defect less than 20 m in diameter if present in said filter but not effective to penetrate integral regions in said filter; challenging said filter with said aerosol particle stream; providing a particle detector downstream of said liquid sterilizing grade filter to detect any particles that penetrate said defect if present in said filter; and classifying said filter as non-integral if a particle is detected by said particle detector, and classifying said filter as integral if no particles are detected by said particle detector.

16. The method of claim 15, wherein the particles in said stream have a size effective to penetrate a defect less than 2 m in diameter if present in said filter.

17. The method of claim 15, wherein said filter is pleated.

18. The method of claim 17, wherein said aerosol particle stream is generated with a plurality of atomizers at a pressure of 50 psig.

19. The method of claim 15, wherein said filter is housed in a cartridge.

20. The method of claim 15, wherein said aerosol particle stream comprises NaCl.

21. The method of claim 15, wherein said filter is a PVDF filter.

22. The method of claim 15, wherein said filter is a PES filter.

23. The method of claim 15, wherein the aerosol particle stream is generated at a pressure of 5-60 psig.

24. The method of claim 15, wherein said aerosol particle stream is generated by a plurality of atomizers.

25. The method of claim 15, wherein the particles in said stream have a concentration of 10.sup.6 particles/cm.sup.3.

26. The method of claim 15, wherein the particles in said stream have a concentration in the range of 10.sup.3-10.sup.7 particles/cm.sup.3.

27. The method of claim 15, wherein said filter is a 0.2 m rated filter.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) For a better understanding of the present disclosure, reference is made to the accompanying drawings, which are incorporated herein by reference and in which:

(2) FIG. 1 is a graph of particle count rate vs. nominal defect size for various membranes;

(3) FIG. 2 is a plot of particle count rate vs. nominal defect size;

(4) FIG. 3 is a graph of particle count rate vs. defect size for a PES membrane;

(5) FIG. 4 is a graph of particle count rate vs. defect size for a PVDF membrane;

(6) FIG. 5 is a graph of NaCl particle counts vs. number of atomizer jets;

(7) FIG. 6 is a plot of air flow rate vs. nominal defect size;

(8) FIG. 7 is another plot of air flow rate vs. nominal defect size; and

(9) FIG. 8 is a schematic diagram showing a test setup in accordance with certain embodiments.

DETAILED DESCRIPTION

(10) The sensitivity of an integrity test is constrained by its ability to differentiate the signal for a defect from any background noise that can compete or interfere with the signal. For example, for the conventional air diffusion integrity test that is commonly used to assess the integrity of sterilizing grade liquid filters, even perfectly integral filters will exhibit a significant diffusion flow rate through the liquid layer in the filter. This diffusive flow rate is sensitive to filter thickness, filter porosity, pore tortuosity, and operating condition variables such as temperature and test pressure. A defect in the filter will allow for a convective flow rate in excess of the diffusion flow rate but this excess flow rate must be high enough to be clearly distinguishable from the typical range of diffusion flow rates in integral devices. For small defects, the convective flow rate through a defect can be masked by the diffusional flow through the integral portion of the filter.

(11) Ideally, the background noise in an integrity test is as close to zero as possible. In the case of aerosol testing, the number of particles that are able to penetrate an integral filter should be zero, so that the detection of any particle that has penetrated a filter is an unambiguous signal for a defect.

(12) It has been found that sterilizing grade filters (often designated as 0.22 m rated filters) retain 100% of aerosol particles in the size range between about 10 nm and 800 nm. In certain embodiments, suitable particles include NaCl particles, KCl particles, as well as other materials that are commonly used to generate aerosols. Other common materials include di(2-ethylhexyl) phthalate (DOP), polystyrene (PSL) and polystyrene-divinylbenzene (PS-DVB) latex spheres, and powders and dusts such as silica, uranium-dioxide, coal, carbon black, pollens, and Arizona road dust (ARD).

(13) Table 1 below shows particle penetration as a function of particle size for PVDF membranes with nominal pore size ratings between 0.1 and 5 m:

(14) TABLE-US-00001 TABLE 1 PVDF Pore Rating Cumulative Particle Membrane (m) Penetration (%) Sample 1 0.1 0 Sample 2 0.2 0 Sample 3 0.45 0 Sample 4 0.65 0 Sample 5 1 0.0000089 Sample 5 5 1.2 Aerosol solution: 0.1 g/l NaCl Aerosol Inlet concentration: 6.5x 10.sup.6 p/cc
It can be seen that for membranes with nominal pore size rating less than about 1 m, the retention efficiency is 100%. If a defect exists however, then particles smaller than the defect size will have the potential to penetrate the filter.

(15) In accordance with certain embodiments, the method of integrity testing a filter includes providing a liquid sterilizing grade filter to be tested, wherein the filter is not pre-wetted (dry); generating an aerosol particle stream wherein the particles in the stream have a suitable size and a suitable concentration to challenge the filter and penetrate a defective region in the filter without penetrating integral regions in the filter; applying the aerosol stream to the filter for a predetermined period of time, and detecting particles that penetrate the defective region. For sterilizing grade filters, suitable particle concentrations may be in the range 10.sup.5 to 10.sup.7 particles/cm.sup.3, and particle sizes may be in the range 10-1000 nm in diameter. In certain embodiments, the method is able to detect single defects that are less than 20 m in diameter, and as small as 2 m in diameter. In certain embodiments, one or more of the solids concentration of the particle stream (typically 10.sup.5-10.sup.7 particles/cm.sup.3), the pressure at which the aerosol is created (typically 5-60 psig), and the number of atomizer generators (from 1 to 6, for example) is modified to ensure passage of particles through defects. In certain embodiments, the amount of time the aerosol particle stream is applied to the filter is modified to allow for sufficient resolution of small rates of particle passage. For example, if the rate of particles that penetrate the membrane is less than one particle per minute, then several minutes can be allowed to ensure that the particle passage rate is accurately determined. With respect to particle concentration, pressure, number of atomizers, and length of time the aerosol stream is applied, these parameters are determined for each type of filter, and can then be applied for all filters of that type.

(16) In certain embodiments, the filters are pleated filters, such as PVDF pleated filters. Pleated filters are typically made with the filter media folded in an accordion-like fashion. The filters may be spiral pleated filters. The filter element may be a membrane. In certain embodiments, the filters, such as pleated filters, are housed in a cartridge.

Example 1

(17) In order to assess the capability of the aerosol test to identify defective filters, cylindrical holes of sizes between 2 m and 20 m were laser drilled into 142-mm membrane discs. Two types of membranes were evaluated: a 0.2 m rated sterilizing grade PVDF membrane and a 0.2 m rated sterilizing grade PES membrane. These membrane filters were challenged with a NaCl aerosol stream (0.12 g/1 NaCl, 310.sup.6p/cc aerosol inlet concentration) generated using a TSI model 3076 aerosol generator at test conditions recommended by the aerosol equipment supplier. The aerosol generator pressure was set at 30 psig and the particles were counted for one minute. The aerosol particles were counted using a TSI model 3772 condensation particle counter. A suitable test set up is shown in FIG. 8. FIG. 1 shows that while the integral membrane showed no passage of particles, the membranes with the laser hole defects showed very high passage of particles. This test demonstrates that particles could readily pass through the defects and be detected by the particle counter. The integral membranes did not exhibit any particle passage.

Example 2

(18) Sterilizing grade membranes in pleated cartridge format were tested under the same conditions, including aerosol and test pressure, as the 142 mm discs described in Example 1. As was done with the 142 mm discs, pleated 10 cartridges were constructed with membranes containing single laser hole defects between 2 m and 20 m. It can be seen from FIG. 2 that in pleated cartridge format, the defect signal was much reduced compared to flat discs. Defect sizes that were easily detected in 142 mm discs (containing about 127 cm.sup.2 of membrane area) format were not detectable in 10 pleated cartridge (containing about 5000 cm.sup.2 of membrane area) format. No particle penetration was detected in the PES membrane cartridge containing a 2 m hole, or, in a PVDF membrane cartridge containing a 5 m hole. This was in part due to the more tortuous pathway that the particles must travel in a pleated membrane, which also contains porous upstream and downstream support layers. The tortuous pathway hinders access to the membrane surface, and therefore increases the opportunity for particle interception either upstream or downstream of the defect and upstream of the particle detector. In addition, as membrane area increases, the flow through the defect becomes an increasingly smaller portion of the flow through the entire membrane and therefore there is a dilution effect on the measured downstream sample.

(19) To overcome the low passage of particles through small defects in 10 pleated membrane filters, the concentration and flow rate of particles challenging the filter were increased. Particle concentration can be increased by increasing the solids concentration in the atomizer solution, increasing the pressure at which the aerosol is created, and increasing the number of atomizer generators. In addition, the test was run for at least 5 minutes to allow for sufficient resolution of small rates of particle passage. This is in contrast to the typical practice of aerosol testing in which the test is often terminated in one minute or less. An enhanced combination of aerosol test conditions were developed and are summarized in Table 2:

(20) TABLE-US-00002 TABLE 2 NaCl Atomizer Number of Concentration Pressure Condition Atomizers (w/w %) (psig) Standard 1 0.012 30 Enhanced 5 1.2 50

(21) TABLE-US-00003 TABLE 3 Membrane Particle Defect Size Particle Penetration Rate Type Type (m) (p/l/min) PES NaCl None <1 5 148 KCl None <1 5 180 PVDF NaCl None <1 5 215 KCl None <1 5 273

Example 3

(22) The PES and PVDF membrane cartridges were also tested using a standard air diffusion integrity test. The cartridges were wetted and then the air diffusion flow rate was measured using a test pressure of 40 psig. FIGS. 6 and 7 show that while the air diffusion flow rate was slightly elevated for cartridges containing defects smaller than about 10-15 m compared to the integral controls, the increase in flow rate was not enough to differentiate integral from non-integral devices. Typical flow rate ranges of integral cartridges is indicated by the shaded areas in the plots. Defects less than 20 microns could not be detected.