normal flora and with experimental aerosols of aid of a slight vacuum. The air-borne particles are

Transcription

1 APPLIED MICROBIOLOGY, June 1972, p American Society for Microbiology Vol. 23, No. 6 Printed in U.SA. Simple Liquid Scrubber for Large-Volume Air Sampling LEE M. BUCHANAN,' J. BRUCE HARSTAD,' JOHN C. PHILLIPS,3 EUGENE LAFFERTY,' CHARLES M. DAHLGREN,2 AND HERBERT M. DECKER2 U.S. Army Biological Defense Research Center, Ft. Detrick, Frederick, Maryland Received for publication 16 December 1971 A new large-volume air sampler called the "simple liquid scrubber" is described. It can recover a high percentage of microorganisms from large volumes of air, up to 950 liters/min, and concentrate them into a small volume of liquid at a ratio of about 400,000 to 1. The principle of operation of the scrubber is based on the production of a fine mist in a rapidly moving airstream with ultimate collection of the airborne particles by impingement into the film of liquid formed upon impaction of the mist droplets on the scrubber walls. The scrubber compared favorably with the all-glass impinger (AGI-30) and the slit sampler in tests with the normal flora and with experimental aerosols of Bacillus subtilis var. niger spores. The purpose of this paper is to describe and evaluate a new, large-volume air sampler called the "simple liquid scrubber." It can recover a high percentage of microorganisms from large volumes of air, up to 950 liters/min, and concentrate them into a small volume of liquid at a ratio of about 400,000 to 1. This sampler had its inception with the Cyclone- Scrubber, a large-volume air sampler, originally fabricated by Space General Corp. under a Government contract, and recently described by Decker et al. (1). The simple liquid scrubber is compared with the all-glass impinger (AGI-30), the cotton filter, and the slit sampler in tests with the normal flora and with experimental aerosols of Bacillus subtilis var. niger spores. MATERIALS AND METHODS The simple liquid scrubber is shown in Fig. 1. The collecting fluid is pumped through a needle into the throat of the inlet arm (1.9-cm diameter), where it is aspirated into a fine mist by the airstream. The air enters the right-angle arm tangentially and assumes a spiral path. The mist droplets are thrown out of the airstream onto the walls, forming a continuous film which is moved helically by the airstream to an outlet, where the fluid is collected in a flask with the ' Present address: Spearman, Tex Retired. 2Present address: Biological Defense Materiel Directorate, Edgewood Arsenal, Md 'Present address: Oregon State University, Corvallis, Ore 'Present address: 512 Valley Street, Frederick, Md aid of a slight vacuum. The air-borne particles are removed from the airstream primarily by impingement into the film of liquid. Removal of particles by interception with the mist droplets may be of little importance, because the droplet velocity is not greatly different from that of the airstream. A prototype made of acrylic resin (Lucite) was compared with the cotton filter sampler for the recovery of homogeneous aerosols of B. subtilis var. niger spores having particle diameters of 0.9, 1.8, and 3.6 im. The cotton filter was selected as the reference sampler because, in our experience, it is the most efficient sampler for the viable recovery of spore aerosols. However, it is applicable only for sampling spores or other highly resistant organisms and therefore is not widely used. The cotton filter sampler described by Wolf et al. (5) consists of a wad of absorbent cotton packed in a small drying tube to a resistance of 38 cm of water at an airflow rate of 10 liters/min. Spores are dislodged from the cotton by shaking in distilled water containing 0.1% Tween (Atlas Chemical Industries, Inc., Wilmington, Del.). The homogeneous spore aerosols were produced with a spinning-disc generator (Environmental Research Corp., St. Paul, Minn.) from aqueous spore suspensions and from suspensions containing 0.1 or 1% dextran 2000 (Pharmacia, Uppsala, Sweden) added to increase the aerosol particle size. The aerosols generated from spore suspensions without dextran were composed of single naked spores with an equivalent spherical diameter of 0.9 Am, whereas those originating from spore suspensions containing 0.1 and 1% dextran had particle diameters of 1.8 and 3.6 gnm, respectively. Procedures for producing and particle sizing homogeneous spore aerosols with the spinning disc generator are described in a recent paper by Harstad et al. (3). 1140

2 VOL. 23, 1971 SIMPLE LIQUID SCRUBBER 1141 FIG. 1. Prototype simple liquid scrubber. Funnel-like inlet is not shown. Fluid nozzle should have been shown at the center of the throat rather than at its periphery. The method used for evaluating the scrubber consisted of disseminating an aerosol continuously with the spinning disc generator into a 140-liter chamber at a rate of 2,300 liters/min. Aerosol was drawn from the chamber through a seamless duct (0.15-m diameter and 1.83-m long) and into the scrubber at a rate of 850 liters/min by a blower. The scrubber effluent passed through a particulate filter before being exhausted by the blower to the outside air. Another blower was used to remove the excess aerosol from the chamber. This aerosol was also filtered before passing through the blower to the outside air. After a 10-min aerosol equilibration period, the scrubber collecting fluid was recovered for 10-min periods, and aerosol samples were collected from the seamless duct with two cotton filter samplers by the use of isokinetic sampling probes centered in the duct about 0.6-m upstream from the scrubber's inlet funnel. Aerosol relative humidity and temperature were monitored with an electric hygrometer (Hygrodynamics, Inc., Silver Spring, Md.). The airflow resistance of the scrubber was measured with a U-tube water manometer. Sampler airflow rates were determined with rotometers (Fischer and Porter Co., Warminster, Pa.). Sampler fluids were assayed for spores on pour plates of Tryptose agar (Difco) after making the appropriate dilutions in distilled water. Colonies were counted after 24 hr of incubation at 37 C. RESULTS AND DISCUSSION The results of 41 tests comparing scrubber recoveries of spore aerosols with those obtained with cotton filters are summarized in Table 1. The collecting fluid used in the scrubber was TABLE 1. Effect of aerosol particle size on the collection efficiency of the prototype scrubbera for Bacillus subtilis var. niger spore aerosolsb Scrubber collection Aerosol No of efficiency particle No.on (%) 95% Confidence en interval (% diam (Mum) Mean to to to 102 ascrubber collecting fluid, 0.1% Tween ; fluid recovery, 2 to 4 ml/min; airflow, 850 liters/min; resistance to airflow, 1.25 m of water. Aerosol concentration, 60 to 4,000 spores/liter; temperature, 25 to 27 C; relative humidity, 19 to 43% Ṡcrubber collection efficiency = scrubber recovery/cotton filter recovery. 0.1% Tween. Fluid recovery rates ranged from 2 to 4 ml/min, but there appeared to be no advantage in using the higher flow rates. The data clearly indicate that the collection efficiency of the scrubber is a function of the aerosol particle size. The excellent results obtained with the prototype prompted the fabrication of a second scrubber, also made of Lucite, which was installed, along with the necessary auxiliary equipment (blower, peristaltic pump, magnehelic gauge, and collecting fluid containers), on a small laboratory cart to provide a portable air sampling unit (Fig. 2). This

3 1142 BUCHANAN ET AL. APPL. MICROBIOL. FIG. 2. Portable simple liquid scrubber. portable unit was compared with the all-glass impinger, AGI-30 (5), for the recovery of homogeneous spore aerosols. The AGI-30 was selected as the reference sampler because: (i) it is widely used for the viable assay of airborne microorganisms, (ii) the collecting fluid can vary from water for sampling spores to tissue culture medium for sampling viruses, and (iii) a learned committee recommended that any specialized sampler be compared with the AGI-30 operated at the maximal airflow rate of 12.5 liters/min (4). Three collecting fluids were used in the AGI-30: (i) distilled water, (ii) 0.06% Tween containing 0.1% Dow Corning antifoam A, and (iii) tissue culture medium containing 0.1% Tween and 0.2% antifoam. The tissue culture medium was Hanks balanced salt solution plus 5 x 10-3 g of lactalbumin hydrolysate per ml. The portable scrubber was operated at 950 liters/min, which resulted in a resistance of 1.25 m of water, whereas that for the prototype was 1.25 m of water at a lower airflow rate of 850 liters/min. This indicates that the two scrubbers were not identical, even though they were made from the same drawings. The collecting fluids evaluated in the portable scrubber were: (i) water, (ii) 0.06% Tween, and (iii) tissue culture medium. The results of 48 tests comparing the portable scrubber with the AGI-30 for the recovery of spore aerosols are summarized in Table 2. The data show that spore recoveries with both samplers were a function of the collecting fluid. Tween (0.06%) appeared to be the best scrubber fluid for the recovery of aerosols of naked spores (0.9-gm diameter), but tissue culture medium (TCM) was the best fluid for sampling the larger dextran-spore particles (1.8- and 4.0-gm diameter). TCM containing 0.1% Tween appeared to be the best AGI-30 fluid for the recovery of all three aerosol particle sizes. Therefore, the best estimates of scrubber performance compared to the AGI-30 are (i) 70% for 0.9-,um diameter aerosol particles and (ii) 103% for both the 1.8- and 4.0-,um diameter particles. The best estimates of scrubber performance with 0.06% Tween as the collecting fluid are: (i) 70% for 0.9-gm diameter particles, (ii) 89% for 1.8-gum diameter particles, and (iii) 94% for 4.0-,um diameter particles or comparable to those obtained with the prototype scrubber (Table 1) in which Tween was also the collecting fluid, but the cotton filter was the reference sampler. The portable scrubber was also evaluated for the viable recovery of microorganisms from outdoor air. In these tests, the scrubber was compared with the AGI-30 and with the slit sampler (2, 5) operated at an airflow rate of 28.3 liters/min. Slit sampler recoveries denote the concentration of viable or organismbearing particles, whereas clumps of organisms are believed to be disintegrated by the AGI-30 and, therefore, it will reflect the microbial

4 VOL. 23, 1971 TABLE 2. Comparing the portable scrubber with the AGI-30 for the recovery of Bacillus subtilis var. niger spore aerosolsa Aero- Scrubber rela- Sol Collecting fluid tive recovery' ticle SIMPLE LIQUID SCRUBBER MaeRn diam Scrubber5 AGI-3Oc Mean' R 0.9 Water Water to 46 TCM' TCM' to % Tween TCM' to % Tween 0.06% Tween to Water Water 90 to 105 TCM TCM' to % Tween TCM' to % Tween 0.06% Tween to Water Water 84 to 90 TCM TCM' to % Tween TCM' to % Tween 0.06% Tween to to 3,000 spores/liter; a Aerosol concentration, temperature, 22 to 28 C, relative humidity, 31 to 47%. b Scrubber collecting fluid recovery, 2.22 to 2.78 ml/min; airflow, 950 liters/min; resistance to airflow, 1.25 m of water. c Duplicate samplers per test; collecting fluid volumes, 20 ml initially and a mean of 18.5 ml after sampling. d Scrubber relative recovery = scrubber recovery/ AGI-30 recovery. Based on four 5-min tests, all on a single day. 'TCM = Tissue culture medium. ' Contains 0.1% Tween. population rather than the particle population. The collecting fluid used in both the scrubber and the AGI-30 was 0.06% Tween, but antifoam was added to the AGI-30 fluid to reduce foaming. Tryptose agar plates, poured 1 to 2 hr before the tests, were used in the slit samplers. The use of freshly poured plates was indicated by preliminary tests. Scrubber and AGI-30 collecting fluids were assayed immediately after each test by use of an agar layer method. Duplicate 1- and 5-ml samples of each fluid were pipetted onto solid Tryptose agar plates, and a thin layer of melted agar was poured over the surface of each plate. The plates were rocked gently and allowed to harden. Colonies were counted after 24 hr of incubation at 37 C. The TABLE 3. Comparing the portable scrubberaslit sampler,' and AGI-3(Y for the viable recovery of microorganisms from outdoor air Organisms/ Particles/ Organisms/ Test' no. liter liter (slit liter (scrubber) sampler") (AGI-30t) Mean a Scrubber collecting fluid, 0.06% Tween ; fluid recovery, 2.67 to 2.85 ml/min; airflow, 950 liters/min; resistance to airflow, 1.25 m of water. ISlit sampler collecting medium, Tryptose agar plates poured 1 to 2 hr before test; airflow, 28.3 liters/min. CAGI-30 collecting fluid, 0.06% Tween containing 0.06% Dow Coming Antifoam A; fluid volumes, 20 ml initially and a mean of 16.7 ml after sampling. d Sampling period, 10 min/test. eduplicate samplers/test. results of a series of six 10-min tests (three tests/day for 2 days) are shown in Table 3. The scrubber compared favorably with the slit sampler and the AGI-30 for the viable recovery of microorganisms from outdoor air. However, the AGI-30 recoveries were considered too low to obtain good estimates of the populations of airborne microorganisms in that the average viable recovery was only 0.5 organism per ml of collecting fluid, whereas that for the scrubber fluid was 29 organisms per ml. The average slit sampler recovery was 19.6 colonies per agar plate, and there was good agreement between duplicate samplers. The results with the prototype and portable units show that the simple liquid scrubber has potential application in studies relating to the airborne transmission of disease and for monitoring air to detect viable microorganisms. Among its valuable features are its ease of fabrication and use, its high collection efficiency, and the ability to concentrate organisms from large volumes of air into a small volume of liquid at a ratio of about 400,000 to 1. ACKNOWLEDGMENT We thank George Bonnell and Ralph Stine for the engineering drawings and for valuable engineering support. LITERATURE CITED 1. Decker, H. M., L. M. Buchanan, D. E. Frisque, M. E.

5 1144 BUCHANAN ET AL. Filler, and C. M. Dahlgren Advances in largevolume air sampling. Contamination Control. 8: Decker, H. M., and M. E. Wilson A slit sampler for collecting air-borne microorganisms. Appl. Microbiol. 2: Harstad, J. B., M. E. Filler, W. T. Hushen, and H. M. Decker Homogeneous bacterial aerosols produced with a spinning-disc generator. Appl. Microbiol. 20:94- APPL. MICROBIOL Letter to the Editor. 12 June Standard sampler for assay of air-borne microorganisms. Science 144: Wolf, H. W., P. Skaliy, L. B. Hall, M. M. Harris, H. M. Decker, L. M. Buchanan, and C. M. Dahlgren Sampling microbiological aerosols. Public Health Monograph no. 60. U.S. Government Printing Office, Washington, D.C.