Collection of Airborne Microorganisms into Liquid by Bubbling through ...

244 downloads 78 Views 241KB Size Report
knowledge technical support provided by Drs. R. Gorny, A. Adhikari, ... impingers, such as the AGI-30 (Ace Glass Inc., Vineland, NJ), may lose a considerable ...
Aerosol Science and Technology 36: 502–509 (2002) c 2002 American Association for Aerosol Research ° Published by Taylor and Francis 0278-6826 =02=$12.00 C .00

Collection of Airborne Microorganisms into Liquid by Bubbling through Porous Medium Igor E. Agranovski,1 Victoria Agranovski,2 Sergey A. Grinshpun,3 Tiina Reponen,3 and Klaus Willeke3 Faculty of Environmental Sciences, GrifŽ th University, Brisbane, QLD, Australia Centre for Medical and Health Physics, Queensland University of Technology, Brisbane, QLD, Australia 3 Aerosol Research and Exposure Assessment Laboratory, Department of Environmental Health, University of Cincinnati, Cincinnati, Ohio 1 2

A new method for the removal of airborne particles by air bubbling through Ž brous Ž lters immersed into a liquid has recently been developed (Agranovski et al. 1999) and shown to be very efŽ cient for cleaning air environments with ultra-Ž ne aerosol particles. The principal objective of the present study was to evaluate the new bubbling technique for the collection of airborne bacteria into a liquid for subsequent physical and microbiological analysis. It was found that the technique is capable of achieving a physical collection efŽ ciency of 98.5% or higher for particles larger than 0.3 ¹m in aerodynamic diameter. The physical collection efŽ ciency of the prototype bubbler remained at that high level for 8 h of continuous operation with negligible variation of the pressure drop across the device. Evaporation of the collection  uid did not exceed 20% during 8 h, and the reaerosolization effect on the physical collection efŽ ciency of the bubbler prototype was 50% for particle sizes ¸0.3 ¹m, >80% for particle sizes ¸0.5 ¹m, and >95% for particle sizes ¸1.0 ¹m) (Lin et al. 2000). The BioSampler requires a pressure drop of 50 kPa for operation at an air ow rate of 12.5 L/min. A new method for highly efŽ cient removal of solid and liquid particles from gas carriers has recently been developed and evaluated for engineering control purposes (Agranovski 1998; Agranovski et al. 1998). In this method, a porous medium is submerged into a layer of liquid and a carrier gas contaminated with particles is blown through it. As a result, the gas carrier is split into a number of very small bubbles and particulates are effectively removed on their walls. The previous studies (Agranovski et al. 1999, 2000) have demonstrated that the method is capable of achieving a removal efŽ ciency of higher than 95% for particles larger than 0.01 ¹m. Since the control technology utilizes the wet collection method, low evaporative disinfectants can be added into the liquid to kill the collected pathogenic microorganisms, thus preventing their reentry with the ef uent air. In this study, the new method, which has been shown to be suitable for engineering control purposes, was extended to bioaerosol sampling. The objective was to evaluate the capability of the bubbling technique for the long-term collection of viable airborne microorganisms that requires maintaining their physical and microbiological properties in a sampler. A prototype device had been developed and tested over a range of experimental conditions with polystyrene latex (PSL) monodisperse spheres of typical bacterial and fungal particle sizes as well as with sensitive bacteria. The data have demonstrated that the new technique is suitable for the efŽ cient collection of airborne microorganisms into a liquid for total and viable enumeration.

503

MATERIALS AND METHODS

Prototype of the Bubbling Device A prototype device designed for this study is shown in Figure 1. It was built to operate at low sampling  ow rates of 2–4 L/min. The contaminated air stream enters the device through an air inlet pipe and, at the beginning of the operation, displaces the liquid from the bottom of the device to the area above the supporting sieve plate. The supporting sieve plate with an open area (ratio of the total area of holes to the total area of the plate) sufŽ cient to keep the liquid above the plate during operation is installed to prevent drainage of the liquid from the plate. To achieve this, the open area of the plate was calculated and experimentally adjusted to result in air velocity in the holes higher than 1.5 m/s for the liquid and materials used in this study. After passing the plate, the air stream goes through a Ž lter submerged into the liquid and protected against displacement by a restraining net. The puriŽ ed stream leaves the device through an outlet pipe. The prototype was made out of a 200 mm long clear acrylic tube with a 35 mm internal diameter and a 2 mm thick wall. A polypropylene Ž brous medium with a Ž ber diameter of 12 ¹m, packing density of 16%, and thickness of 6 mm was placed at a distance of 20 mm from the bottom of the device right above the supporting plate with an open area of 3%. A stainless steel restraining net with 5 mm cells was located above the porous medium to prevent its displacement during operation of the device. The air inlet and outlet pipes were also made out of acrylic with 10 mm internal diameter and 1.5 mm thickness of the wall.

Figure 1. Schematic diagram of the bubbler prototype.

504

I. E. AGRANOVSKI ET AL.

To provide access to the collection  uid for sampling during operation and for the proper removal of liquid for analysis on completion of the run, an 8 mm diameter access hole was drilled at the top of the device and hermetically sealed during operation. A mist eliminator was installed below the air exit tube to baf e relatively large droplets that may have been generated by the bubbling process.

Experimental Setup The experimental setup used in this study is schematically shown in Figure 2. A 3 jet Collison nebulizer (BGI Inc., Waltham, MA) was used to generate aerosols by nebulization of a liquid suspension at a  ow rate of 6 L/min of dry and Ž ltered compressed air. The aerosolized particles were then dried and diluted by Ž ltered air supplied at a  ow rate of 40 L/min. To avoid electrostatic removal of particles to the inner surfaces of pipelines, the air stream was passed through an electrostatic charge neutralizer (10 mCi 85 Kr, model 3012, TSI Inc., St. Paul, MN) before it entered the partially enclosed aerosol chamber (12 cm diameter, 15 cm high). The test particles were sampled into the prototype bubbling device located inside the chamber. The bubbler contained 30 mL of sterilized water and was operated at sampling  ow rates of 2, 3, and 4 L/min. The pressure drop across the device was monitored by a 25 Pa resolution Magnehelic differential pressure gauge (Dwyer Instruments Inc., Michigan City, IN). The concentration and size distribution of particles in the chamber was continuously monitored by sampling through the reference sampling line using an aerodynamic particle-size spectrometer (Aerosizer, API Mach II, Amherst Process Instruments, Inc.,

Figure 2. Experimental setup.

Hadley, MA). The Aerosizer measures up to 1,100 particles/cm3 over an aerodynamic size range of 0.3 to 200 ¹m at a  ow rate of 5.3 L/min (critical oriŽ ce controlled). The particle concentration downstream of the bubbler was also measured throughout the experimental run. Control valves 1 and 2 (Figure 2) were used to select the sampling line of interest. As the bubbler’s operational  ow rate is lower than that of the Aerosizer, a controlled amount of the HEPA Ž ltered air (equal to the difference of the above  ow rates) was added to the air stream before it entered the Aerosizer. This amount of additional air was monitored by a  ow meter and controlled by adjusting the valve in either the test or reference line, while the other line was closed. For example, for a  ow rate of 2 L/min in the test sampling line, valve 1 was closed and valve 2 was adjusted until the  ow rate in the Ž ltered air line was 5.3 ¡ 2 D 3.3 L/min. The sampling  ow rates were calibrated before the experiments by temporary installation of air  ow meters in the test and reference lines. The  ow rates through these sampling lines were the same in all experiments.

Test Particles Two types of test particles were used to evaluate the performance characteristics of the prototype bubbling device: PSL monodisperse spherical particles and vegetative cells of Pseudomonas  uorescens (P.  uorescens). The PSL particles were used to determine the physical collection efŽ ciency over a range of particle sizes (0.318–2.79 ¹m), sampling  ow rates (2– 4 L/min), and sampling periods (up to 8 h). The bacteria were utilized to evaluate both the physical and biological efŽ ciencies of the sampling technique. PSL particles (Bangs Laboratories, Fishers, IN) of 0.318 ¹m, 0.517 ¹m, 0.83 ¹m, 1.024 ¹m, and 2.79 ¹m diameter were used as these particle sizes represent bacteria and fungi. The rodshaped gram-negative P.  uorescens bacteria, ranging from 0.7 to 0.8 ¹m in diameter and from 1.5 to 3.8 ¹m in length (Palleroni 1984), were selected to represent microorganisms sensitive to aerosolization and collection stresses. The mean aerodynamic size of the aerosolized P.  uorescens bacteria is approximately 0.8 ¹m (Stewart et al. 1995). P.  uorescens is commonly found in ambient air (Nevalainen 1989). Thus the selection of test particles represents the worst-case scenario for performance evaluations of the prototype bubbling device. Preparation of the Microbial Suspension for Aerosolization A stock culture of P.  uorescens (ATCC 13532) was obtained from the American Type Culture Collection (Rockville, MD). The P.  uorescens routing cultures were inoculated from a loop of the stock culture and grown in 100 mL of Trypticase Soy Broth (Becton Dickinson Microbiology System, Cockeysville, MD) for 18 h at 30± C in a Gyrotory water bath shaker (Model G76, New Brunswick ScientiŽ c, Edison, NJ) at 150 rpm (2 cm circle). The stationary-phase organisms were harvested by centrifugation at 7,000 g for 7 min (Sorvall RC-5B refrigerated

505

COLLECTION OF MICROORGANISMS BY BUBBLING

superspeed centrifuge; Dupont Co., Boston, MA), the pellets were then washed 3 times with deionized and sterilized water that was also used to dilute the cells to obtain the Ž nal liquid suspension. A 50 mL aliquot out of 250 mL of the initial bacterial suspension was placed in the nebulizer to be aerosolized. The remaining suspension was kept at the room temperature of 23–25± C during the entire experiment. Before starting each experiment, an aliquot of the microbial suspension was analyzed for culturable bacteria by the plating technique that is described in the next section. Natural aging of the initial microbial suspension was investigated by analyzing the concentration of culturable cells in the suspension over 1 h intervals up to 8 h (the range of sampling time periods used in this study). It was found to be signiŽ cant for generation times exceeding 2 h. To minimize the effect of microbial stress caused by the nebulization process and to ensure an initial suspension of uninjured bacterial cells, the Collison nebulizer was recharged with the remaining (reserved) bacterial suspension every 2 h. The viability of bacterial cells in the suspension was almost not affected by the continuous operation of the nebulizer during a 2 h period (within about 10%), as was determined in the preliminary experiments by plating aliquots of the suspension over 30 min intervals.

Colony Enumeration The standard spread plate technique (Greenberg et al. 1992) was used to determine the bacterial viability by observing the ability of the cells to divide and form colonies. An aliquot (0.1 mL) of an appropriate tenfold dilution of the collecting  uid was spread on Trypticase soy agar (TSA) (Becton Dickson Microbiology Systems, Cockysville, MD). After incubation at 30± C for 48 h, colonies were counted with a Quebec colony counter (dark-Ž eld Reichert-Jung counter; Leica, Inc., DeerŽ eld, IL). Evaporation of the Collection Fluid during Operation of the Sampler The bubbling device was operated with sterilized water as the collection  uid. In contrast to the white mineral oil used in the BioSampler, water may evaporate during prolonged sampling periods. Therefore, the evaporation rate of the collection  uid was determined to correct the plate count results by the liquid reduction factor deŽ ned as the ratio of the current volume of water in the bubbler to the initial volume of water in the bubbler. The initial volume of water in the bubbler was equal to 30 mL for all experiments. The current volume was determined by weighing the bubbler on an analytical balance with 0.01 g resolution (Mettler, Columbus, OH) every hour during each 8 h experiment. Experimental Procedure First, the pressure drop across the bubbling device was measured as a function of the operating  ow rate and time. These tests were conducted when the bubbler was sampling PSL par-

ticles of 0.83 ¹m at an airborne particle concentration level of about 600 cm¡3 . The sampling  ow rates were 2, 3, and 4 L/min; the sampling times were 1 min (initial), 4 h, and 8 h. Second, the physical collection efŽ ciency of the bubbler prototype was determined with PSL particles of different sizes at  ow rates of 2, 3, and 4 L/min. The physical collection efŽ ciency, E, was calculated as E D1¡

Cout ; C in

[1]

where Cout and Cin are the particle concentrations downstream and upstream of the bubbler as measured by the Aerosizer. These tests were conducted 5 times for all 5 PSL particle sizes after 20 min of operating the device and were repeated at times t D 40 min, 1 h, 2 h, 4 h, and 8 h to determine the collection efŽ ciency as a function of sampling time for PSL particles of 1.024 ¹m. Third, the biological collection efŽ ciency of the bubbler prototype was determined as a function of sampling time. The bioefŽ ciency was deŽ ned as the bacterial recovery rate after collection by the bubbling device. The relative recovery rate, R, of P.  uorescens cells was obtained by measuring the number of colonies, N(t)CFU , cultivated from the collection  uid after the microorganisms had been collected for the time t; the total number of microorganisms, N(t)COLLECTED , collected by the sampler during this time and the natural viability decay factor, ° : RD

N (t )CFU N (t)COLLECTED ¤ °

[2]

N (t )CFU was determined by the standard plating technique. The amount of particles collected by the bubbler during the sampling period t was calculated as N (t)COLLECTED D C in ¤ E ¤ t ¤ Q;

[3]

where Q is the sampling  ow rate. The factor ° was determined through separate control experiments in which the microbial suspension was left in the bubbler for 8 h without air ow through it and the relative recovery rate was measured at t D 0 (initial) and at t D 0.5, 1, 2, 4, and 8 h: ° D

N (t )CFU(control) : N (t D 0)CFU(control)

[4]

Thus the relative recovery of microorganisms collected in the bubbler at the end of the sampling period t is RD

N (t)CFU N (t D 0)CFU(control) ¤ : C in ¤ E ¤ t ¤ Q N (t )CFU(control)

[5]

506

I. E. AGRANOVSKI ET AL.

As the data revealed that the physical collection efŽ ciency was sufŽ ciently high at all the tested sampling  ow rates, the bioefŽ ciency tests were only performed at Q D 2 L/min. All experiments were performed with at least 3 repeats. During all experiments, the air temperature and the relative humidity in the test chamber were monitored: T D 24–25± C, RH D 24–26%. RESULTS AND DISCUSSION The evaporation rate was found to be