domnick hunter - a division of Parker-Hannifin Corporation. World-Class Filtration, Purification and Separation Solutions to Industry

• View all products
• Customer Support
• Compressed Air Technical Centre
• Global Exhibitions and Events
• Careers
• Awards & Accreditations
• Legal

Introduction

Aerosol challenge integrity testing has been shown to establish the integrity of air sterilisation filters without any of the associated problems caused by liquid based integrity test methods. This article examines the correlation of an aerosol challenge to an airborne microbiological challenge in establishing the integrity of any cartridge used for gas sterilisation. Correlation tests are conducted between VALAIRDATA and aerosol bacterial challenges using Bacillus Subtilis, Brevundimonas diminuta and MS-2 Coliphage. Results show a direct correlation between aerosol and bacterial challenge. This correlation work and the supporting data are discussed in detail.

The Aerosol Challenge Tests

Aerosol challenge testing is a highly searching and sensitive test. During the test, filter cartridges are challenged with a minimum of 1012 particles of the most penetrating particle size, i.e. 0.2 - 0.3 micron. The aerosol challenge tester therefore measures the particle penetration under worst case conditions. Extensive studies have been undertaken both by domnick hunter and independently by the Centre for Applied Microbiology and Research (CAMR), to correlate aerosol challenge with bacterial challenge. In the independent tests, aerosols containing very large numbers of viable organisms per unit volume of air have been produced by a Collison spray (CAMR, Porton Down, UK)4 to provide a high challenge so that the testing time can be reduced. The bacterial tests used were:-

• washed suspensions of spores of Bacillus subtilis varniger in distilled water
• washed suspensions of Brevundimonas diminuta in distilled water.

In the system described here, the temperature and relative humidity of the air were 20°C (68°F) and 60% respectively. Under these conditions, any water associated with aerosols containing a single bacterial cell will rapidly evaporate and the filters are therefore challenged by monodispersed particles representing the size of the cells i.e.:-

• Bacillus subtilis 1.0 by 0.7 micron
• Brevundimonas diminuta 0.8 by 0.2 micron

If the bacteria were suspended in culture medium or a buffer solution then a range of sizes of particles containing viable bacteria would be produced after the water has evaporated from the original droplets formed by spraying. The final size of the particles would depend on the initial size of the droplets and the solute content of the spray solution.

Selection of Challenge Bacteria

Brevundimonas diminuta is often used as the preferred challenge bacteria to test the integrity of filters because of its exceptionally small size. One disadvantage of using Brevundimonas diminuta in this way is that its survival in the aerosol has not been fully characterised and it is expected that only about 20% of the aerosolised cells will survive the conditions used in the tests. The authors prefer to use B.subtilis spores to study the penetration of bacteria through filters. This is because it is well established that the spores are not de-activated after prolonged exposure in a wide variety of air conditions. B.subtilis offers another advantage in that the colonies formed can be counted within 18 hrs of incubation (compared with 40 hrs incubation for Brevundimonas diminuta) and can be easily recognised by their orange pigmentation. Nevertheless, Brevundimonas diminuta has been used as well as Bacillus subtilis spores to challenge the filters in these tests. The fluid used to recover the viable micro-organisms is widely used for this purpose.5 It is phosphate buffer manucol antifoam (PBMA), made up of:

• 4.5g KH₂PO₄
• 0.5g (NH₄)2SO₄
• 0.5g NH₄C₁
• 2.5g sodium alginate containing 1.0ml of a 10% emulsion of silicone

(Dow Corning Antifoam RD Emulsion, Merck Ltd, Lutterworth, UK). The fluid was adjusted to pH 7.6 with 0.1N NaOH and made up to 1L and autoclaved at 121°C (250°F) for 15 min. Filters were also challenged with airborne bacteriophage. Bacteriophage contamination of fermentation processes is a particular problem because rapid multiplication of bacteriophage could theoretically lyse a 1000-L culture of 1010cells/L in less than 1 1/2 hrs. To prevent this, any bacteriophage present in ambient air are usually removed by air sterilisation filter cartridges. A reliable method to test the integrity of these filters is therefore vital. A series of filters have been challenged with high concentrations of aerosolised MS-2 Coliphage and correlated with the penetration measured by the VALAIRDATA system. MS-2 is an unenveloped single stranded RNA coliphage, 23 nanometres in diameter with a molecular weight of 3.6 x 10⁶ Da. This coliphage is fairly resistant to inactivation during aerosolisation and partly because of this, and partly due to its more compact and smaller size, it is preferred to the T-Coliphage in filter penetration studies.

Test Organisms

Bacillus subtilis var niger -
Stock suspensions containing washed spores of Bacillus subtilis var -niger which had been washed and suspended in distilled water are available in CAMR. A spore suspension containing 1.5 x 10⁹ colony forming counts/mL was prepared and heated to 60°C (140°F) for 30 min before use (to inactivate vegetative organisms present which are more aerosol sensitive than the spores). When the suspension was assayed after heating and prior to aerosolisation it was found that no decrease in bacterial spore count occurred.

Brevundimonas diminuta -
A lyopholized vial of Brevundimonas diminuta obtained from The National Collection of Industrial and Marine Bacteria, Torry Research Station, Aberdeen (NCIMB), was reconstituted in 1 mL of nutrient broth and 0,1 mL of this was spread over the surface of a nutrient agar plate which was incubated overnight at 30°C ± 2°C (86°F ± 36°F). A loop was used to inoculate a fresh plate of nutrient agar from this plate. The inoculated plate was incubated at 30°C ± 2°C (86°F ± 36°F) for 40 hours. After 24 hrs incubation at this temperature, pin-point colonies were obtained and after 40 hours dull beige coloured colonies of about 2mm diameter were formed. The cells from a single colony were examined under the microscope. The microscopic examination showed that the bacterial cells were small gram- negative rods which correspond to the description of the Brevundimonas diminuta in Bergey’s Manual of Determinative Bacteriology. Another colony from the same plate was used to inoculate a 250mL flask containing 50mL of nutrient broth. The fluid was shaken for 40 hrs at 30°C ± 2°C (86°F ± 36°F). The cells were harvested by centrifugation at 8300g for 30 min and were washed by re-suspension in an equal volume of distilled water. The washed cells were recentrifuged again at 8300g for 30 minutes. The final suspension of Brevundimonas diminuta in water contained 2 x 10¹⁰ cfu/mL.

MS-2 Coliphage -
Suspensions of MS-2 Coliphage were obtained from the NCIMB. A stock suspension of coliphage was prepared by inoculating 0.1mL of the 1011 plaque forming units (pfu)/mL coliphage suspension into 500mL nutrient broth containing 1 x 10⁹mL Escherichia coli during the logarithmic growth phase. The suspension was mixed in a shaking incubator at 37°C (99°F). The bacterial cells lysed within 30 minutes to form a clear suspension. The suspension was centrifuged to remove the cell debris. The supernatant was transferred to a fresh flask and 10 drops of chloroform were added to kill any viable bacteria. The concentration of coliphage was determined by diluting the lysate in 1:1 mixture of nutrient broth and distilled water and adding 0.1mL of suitably diluted sample of 3mL of molten nutrient agar at 45°C (113°F) containing 10⁹ E.coli in the logarithmic phase of growth. After mixing, this was added as an overlay to phage typing agar plate. Each plate was incubated at 37°C (99°F) overnight and the clear plaques formed were counted.

Test Apparatus

The Test System - The system is routinely used in CAMR to test vent filters for the effluent treatment plants. It provides a high challenge of aerosolised micro-organisms to the filters within a short time to simulate the prolonged and much lower challenges that are likely to be encountered during the actual use of the filters. A schematic diagram of the apparatus used in the filter tests using Bacillus Subtilis and Brevundimonas diminuta is shown in Figure 5. The assembly consists of a Collison nebulizer for the generation of microbial aerosols in a chamber, connecting steel pipework, filter housing, compressed air supply, aerosol sampling devices, flow meters, and a powerful vacuum pump to draw the air containing the microbial aerosols through the test filters installed in the housing.

Figure 5: Apparatus for testing the efficiencies of domnick hunter filters against aerosol containing bacteria.

Sampling Systems -
There is a need to determine accurately the concentration of bacteria challenging the filter and the concentration of bacteria which has penetrated the filter. The former is done by taking a one-min sample upstream of the filter through a port linked to a Porton all-glass impinger (CAMR, Porton Down) operating at 11 L/min using 10 mL of PBMA as the collecting fluid. The design and function of this sampler has been described by May and Harper who have shown that nearly 100% of the airborne bacteria should be captured.7 The large capacity Casella slit sampler (Casella London Ltd, London, UK) fitted with a large volume sampling head was used to sample air downstream of the filters to capture any bacterial cells passing through the filter. Air was drawn through the sampler at around 300 L/min (measured by the rotameter) with two of the four slits sealed by the strips provided. With each filter in the housing the background count was determined by slit sampling on a rotating 150-mm Tryptone Soya Broth Agar (TSBA) plate for 3 min. Samples for both 2 and 1/2 min were taken when each filter was challenged by aerosolised bacteria. In these experiments excessive exposure to challenge aerosols was avoided to minimise gradual blockage of pores by bacterial cells. The slit sampler could not be used to determine the penetration of MS-2 Coliphage through the cartridge filter. An alternative system was therefore set up which uses a cyclone sampler⁸ (see Figure 6) to collect the challenge organisms penetrating the filters (i.e. with the cartridge filters installed in the housing). The cyclone is linked to a vacuum pump which draws air through the system at a rate of around 600 L/min (by measuring the linear air velocity into the inlet aperture by an Airflow anemometer). Collection fluid (PBMA) was fed into the cyclone inlet at a rate of 2 mL/min. The particles in the air stream are deposited by centrifugal force on the cyclone wall and are collected by the swirling liquid which was withdrawn by a syringe after 4 min. The exact volume of liquid collected was measured by weighing and the collected fluid was assayed for MS-2 Coliphage.

Figure 6: Apparatus for testing the efficiencies of domnick hunter filters against aerosol containing MS-2 Coliphage.

Nebulisation of the MS-2 culture fluid by the Collison nebulizer produces a range of droplet sizes. After rapid evaporation of water from the droplets⁹ it was shown experimentally⁴ that a range of particles from the limits of optical microscopic analysis (less than 0.2 micron) to 30 micron were produced. The probability of MS-2 phage present in a particular size distribution depends on the size and number of the aerosol particle size. Analysis of the particle size distribution of the MS-2 nebulized in this way by the Andersen sampler¹⁰ showed that 55% of viable MS-2 were collected in the lowest stage (6th) of the sampler (corresponding to the particle size range 0.65 - 1.1 micron); 36% were collected in the 5th stage (corresponding to the size range 1.1 - 2.1 micron); 6% in the 4th stage (2.1 - 3.3 micron) and 3% in the top three stages (greater than 3.3 micron).

Microbial Assay Methods

Bacillus Subtilis var Niger -
The collection fluid from the impingers and cyclone were serially diluted (if required) in PBMA, and 0.1 mL of the diluent was spread on TSBA plates. TSBA plates containing Bacillus subtilis were incubated at 37°C (99°F) for more than 18 hrs. Aerosolisation had no effect on the size of the colonies formed. TSBA plates used in the slit sampler were incubated in the same way and the colonies were also counted.

Brevundimonas diminuta -
The assay of the impinger samples were carried out as described above except that the diluent was spread on nutrient agar plates which were subsequently incubated for 48 hrs at 30°C (86°F). The beige colonies formed were counted. All the Brevundimonas diminuta (formerley Pseudomonas diminuta) colonies counted were beige and their colony morphology was similar to the colonies produced by the original suspension. Nutrient agar plates used in the slit sampler were also incubated for 48 hrs at 30°C (86°F), and the colonies produced were counted.

MS-2 coliphage -
The number of viable MS-2 Coliphage particles collected by the impingers and the cyclone samplers were determined, as described earlier in this article (Test Organisms: MS-2 Coliphage), by adding 0.1 mL of the suitably-diluted sample in PBMA to molten agar at 45°C (113°F) containing 1011 E.coli in the logarithmic phase of growth.

Filters -
The filters used in this test programme were typical production HIGH FLOW TETPOR (domnick hunter limited, Birtley, UK) air sterilisation filter cartridges which utilise a single layer of expanded polytetrafluoroethylene (PTFE) filter media. Two sizes of cartridge filter were tested:

• ZCHT-AZ (5" Demi)
• ZCHT-BZ (2.5" Demi)

Flow rates were 300 L/min using the Casella system and 600 L/min using the cyclone system. Due to their inherent hydrophobicity and their high voids volume these filter cartridges offer exceptionally high flow rates at low differential pressures. A number of the cartridges were steam sterilised excessively until slight and gross failures were initiated. This provided a range of cartridge filters for testing to enable a correlation to be made between microbial penetration and penetration of aerosol by the VALAIRDATA. Using the MS-2 Coliphage suspension (10¹¹ pfu/mL) it was found that 5 x 10⁶ pfu/L actually challenge the filters at 600 L/min for a 4 min test period. The results of the penetration of these organisms through 5 inch HIGH FLOW TETPOR air sterilisation filter cartridges at this flow rate are expressed per hundred million of the challenge (in Table II and Figure 9).

Conclusions

Data comparing the penetration of aerosol test particles with aerosolised micro-organisms show that the VALAIRDATA is a valid system to determine the integrity of cartridge filters. The extremely high sensitivity of the system allows quite minute faults to be detected in filter assemblies where only few organisms were detected downstream of the filter when challenged by high concentrations of micro-organisms. The microbial integrity of seven filters “passed” by this system was confirmed because no micro-organisms penetrated the filters when challenged by extremely high numbers (10⁹) of micro-organisms in the aerosol phase.