By Dr. Harriet Burge, EMLab P&K Chief Aerobiologist and Director of Scientific Advisory Board
Collection of interpretable data depends on many factors, not the least of which is the type of sampling equipment used. Choice of an air sampler depends on the hypothesis being tested, expected concentrations, and the nature of the analytical method to be used. Also to be considered is the expected particle size range of the aerosol and the acceptable amount of error. These and other factors related to the available sampling methodologies are the subject of this article.
Particle Collection Modalities
Inertia is the tendency of an object to resist any change in its motion. Inertia is directly proportional to mass. Thus, larger particles require greater force to be diverted from their original path than do smaller particles. Inertial impaction methods for spore collection generally increase particle inertia by accelerating an airstream closely subtended by a perpendicular collection surface.
Fungal aerosols consist of a wide range of particle types and sizes so a consideration of the size range of particles to be collected is essential. The size of importance is the aerodynamic size, which is the diameter of a sphere that would have the same mass as the spore in question. Mass is a function of density and the mass of a spore must be determined experimentally. Fungal spores are often considered to have unit density (1 gram/cc – the density of water) because water is a major component. The actual density of a spore depends on the amount of water present, as well as the density of other spore components. Regardless of these considerations, unit density is still used to estimate aerodynamic diameter.
Another problem is that many (or perhaps most) fungal spores are not spherical, and the effect of these nonspherical shapes is difficult to predict. For spores, the smallest diameter is considered the aerodynamic diameter. Also, many spores are not smooth, increasing drag caused by air molecules. The effects of surface ornamentation on aerodynamic diameter would have to be experimentally determined.
All of these factors contribute to the inevitable errors involved in aerosol sampling, especially when the aerosol is as complex as most fungal aerosols.
In absolutely still air, gravity is the dominant force that removes particles from the air. Particle settling depends on aerodynamic diameter, so that aerodynamically large particles fall from the air faster than smaller ones. Over a long period of time in a sealed environment, all spores will have fallen from the air. However, until all spores have fallen, a gravity sample in still air underestimates the concentration of small spores. The slightest air movement will overcome the effect of gravity on fungal spores, which will tend to follow the air stream. The smaller the spore, the larger this effect, and small spore concentrations are seriously underestimated in gravity samples.
One of the parameters of a sampler that indicates its relative usefulness for bioaerosol sampling is the d50. The d50 is the aerodynamic diameter of particles that are collected with 50% efficiency. D50 for a sampler is calculated using the following equation:
Stk = Stokes' number describes the behavior of a particle suspended in a fluid flow
W = width of particle
ŋ = air viscosity
p = particle density
V = average air velocity in nozzle
C = Cunningham slip correction factor
A good discussion of d50 calculations is presented in Trunov et al. (2001). Note that the published d50 for a sampler is only obtained when the sampler is operated at the required flow rate. In the best samplers, particles with diameters above the d50 are collected with nearly 100% efficiency. Likewise, efficiency for particles below the d50 is very low. Thus, the shape of the efficiency curve is important. The "best" samplers have steep efficiency curves. Shallow efficiency curves mean that the 100% efficiency diameter may be much larger than the d50, resulting in overall inefficiency for particles above the d50.
Inertial impaction is the most commonly used method for capturing fungal aerosols. The representativeness of the aerosol captured using inertial impaction depends on the use of a sampler with a d50 below the aerodynamic diameter of the smallest spore expected or of interest. Obviously, the sampler must be operated at the required flow rate with the appropriate collection medium in place.
The Rotorod sampler uses direct impaction to capture airborne particles. Narrow rods (< 2mm in diameter) with greased flat surfaces are whirled rapidly through the air capturing particles that get in the way. The efficiency of collection depends on the width of the sampling surface, with narrower surfaces collecting smaller spores more efficiently. The d50 for the Rotorod that is commercially available today is about 20µm. It is therefore only efficient for the collection of very large fungal spores, which account for only a small proportion of most fungal aerosols both indoors and out. It is very useful for pollen sampling (Solomon et al. 1980).
Capturing particles from a moving air stream involves two steps: 1) the particles have to be pulled into the sampling device; 2) the particles have to be removed from the moving air stream.
Pulling Particles into the Impactor (Isodirectional and Isokinetic Sampling)
Capturing complex fungal aerosol particles in a way that accurately represents the original population is a significant challenge. If you place a sampling orifice so that the opening is facing into the airstream (i.e., the opening is perpendicular to the air stream and the air flows directly into the opening) (isodirectional) and no suction is applied, the ambient airstream will tend to flow both into and around the orifice. The particles will follow the path of least resistance. The largest particles will tend to continue on their original path and enter the orifice. Smaller particles will tend to flow around the orifice and be lost. If an amount of suction is applied that matches the velocity of the air stream, both large and small particles should be collected with equal efficiency, since no change in inertia is necessary. This is called "isokinetic" sampling. As you increase suction velocity over that of the ambient air, streamlines will tend to bend into the orifice and oversampling of the smallest spores will tend to occur. Very high velocity sampling will tend to pull all particles into the sampler. This is usually not a problem outdoors, but such samplers essentially clean the indoor air, resulting in possibly serious underestimates of the aerosol concentration per cubic meter of air that would be present if the sampler were not operating. In other words, it is important that the sampler itself not change the environment. The degree of underestimation will depend on the volume of air sampled after the particles are gone. This is because as you continue sampling, the denominator of the equation for calculating concentration per cubic meter of air gradually becomes larger.
If the sampling orifice is not directed into the airstream, then some particles will have to make a turn in order to enter the sampler. If the opening is parallel to the airstream, all particles will have to turn. With no suction, no spores will be captured. With suction, smaller particles will be able to make the turn and larger particles will not.
Increasing particle inertia
The principles of capturing particles by acceleration through small orifices are well illustrated by the Andersen 6-stage impactor. This sampler uses sieve plates with holes of descending size subtended either by greased metal plates or by agar plates to capture particles according to their aerodynamic size, which is more or less proportional to their mass and thus inertia. As the air stream accelerates through the largest holes, all particles gain inertia, and those particles that have gained sufficient inertia will impact on the collection surface. So, the top plate in the series captures the largest particles. The air is then accelerated through smaller holes and a set of particles with less mass is captured. As the hole size decreases, acceleration through the holes increases and smaller and smaller particles are captured.
Other kinds of samplers that use airflow restriction to accelerate spores are the spore traps and the impingers. The Air-O-Cell, the Allergenco and the Burkard spore traps are examples of commonly used spore traps. These samplers use a single rectangular orifice to accelerate particles that have entered the sampler. The amount of acceleration achieved is inversely proportional to the width of the rectangle or slit. Thus, as the slit narrows, the acceleration increases, as does the efficiency of collection for small particles. Common slit widths range from 1-2mm. Other spore traps use circular orifices (e.g., the Cyclex) and particle collection efficiency is proportional to the diameter of the orifice. All of these spore traps have collection surfaces that subtend accelerating orifices. An important engineering consideration is the distance of this collection surface from the orifice. It must be very close if small particle capture is to occur. The shape of the inside of the orifice is also important as it controls the shape of the airstream. Grinshpun et al. (2007) compared the Allergenco-D and the Air-O-Cell, which essentially differ only in the slit to surface distance and the shape of the accelerating orifice. The Allergenco-D is more efficient for the collection of small particles than the Air-O-Cell and the efficiency curve for the Allergenco is steeper. The d50s for the two samplers are about 2.5µm for the Air-O-Cell and about 1.7µm for the Allergenco.
Impingers use the same principles as the spore traps, but the impaction surface is under water (or some other liquid). The AGI (All Glass Impinger) is an example of this type of sampler. In the AGI, air is drawn through a critical orifice that ensures sonic velocity. The opening of the orifice is placed close to the bottom of the sampler, immersed in liquid. The particles impact on the sampler surface beneath the liquid and are then captured in the liquid. The AGI is not commonly used in routine fungal aerosol investigations, in part because of problems with foaming, and the fact that many fungal spores are hydrophobic and are lost in the bubbles (Lin & Li 2010).
The SKC biosampler is an impinger in which the air stream is directed tangentially against a wetted surface. Particles that impact are then washed to the bottom of the sampler.
The RCS centrifugal sampler draws air into the device using a fan with blades slanted such that the air and entrained particles are accelerated toward the collection surface (agar) using centrifugal force.
Cyclone samplers also use centrifugal force to capture particles, but the particles are generally captured in a liquid that collects in the bottom of the sampler.
Trapping the Particles
Once inside the sampler, the particles must be captured onto or into some kind of medium so that analysis is possible. Ideally, once a spore has achieved sufficient inertia to stop at the sampling surface, it will stay there until the sample is analyzed. However, unless the surface of the spore itself is sticky (and some are!), then contact with a hard surface will cause the spore to deform slightly, and the resulting elastic response will cause the spore to bounce away from the surface. The degree of bounce depends both on the nature of the collecting surface and the nature of the particle itself.
Agar is used as an impaction surface in culture plate impactors and the RCS centrifugal samplers. The assumption is that because the agar is soft, the particles with sufficient inertia for impaction will embed in the agar and not bounce. This phenomenon actually depends on the hardness of the agar surface and, as mentioned above, how many other particles have impacted at the same site. As the agar surface dries, bounce becomes an increasing concern.
On a hard surface some kind of adhesive is usually necessary to reduce bounce. The type of adhesive is an important consideration. Generally, silicone sealants have proven the best at preventing particle bounce but must be used soon after application since the sealant "cures" and becomes less sticky (Clauß et al. 2010). The widely used Air-O-Cell and Allergenco samplers use proprietary sticky surfaces.
Liquid can also be used to capture particles that bounce. The liquid impingers are designed so that particles bounce from the hard glass surfaces and are captured in the liquid.
The shape of the orifice also affects particle bounce. For example, the Allergenco orifice has a straight section (a cylindrical tube with straight sides) that reduces turbulence and leads to a relatively compact deposit, while the Air-O-Cell lacks this straight section, resulting in spread of the spore deposit beyond the width of the slit (Grinshpun et al. 2007).
It should be noted that overloading impactor surfaces leads to increasing bounce as the surface becomes coated with non-sticky materials. In other words, spores may be bouncing off of other spores and particles. Concentration on impactor surfaces may also lead to analytical difficulties such as masking during microscopic identifications and loss of culturability with the culture plate impactors (Mainelis & Tabayoyong 2010).
Fungal spores are readily captured on filters. Capture is caused almost entirely by interception except for filters with very large pore sizes. Filters with pore sizes in the range of 1.5-2µm are very effective at capturing more than 99% of airborne fungal spores. Using the largest possible pore size will reduce resistance and allow for the use of small personal pumps at reasonably high flow rates. If filters are to be used for microscopic analysis, they need to be made of a material that can be cleared, such as cellulose acetate. Epifluorescence microscopy requires filters that have a relatively smooth surface. Black filters are often used to enhance the fluorescent image. Teflon or other smooth surface filters should be used if spores are to be washed from filters for analysis (Deacon et al. 2009).
Filtration is an attractive alternative for bioaerosol sampling, especially when long sampling periods are necessary, and when aerosols are expected to be extremely concentrated, such as in many farming environments. However, it should be noted that filtration seriously reduces culturability of some microorganisms, especially during long sampling times. In addition, increasing concentration of spores on the filter may clog the pores and increase backpressure on the pump, reducing flow rate. Also spores generally must be washed from filters for analysis. Electrostatic charges may cause spores to stick to the filter holder surfaces so that these surfaces must be washed as well. All of these factors introduce important sources of error.
Electrostatic samplers collect particles by applying a charge to collection plates and either taking advantage of the natural charge on airborne particles or by oppositely charging the particles, which are then collected on the plates. This type of sampling is rarely used for fungal aerosols although there is evidence that they would be quite efficient. Yao and Mainelis (2006) compared collection of fungal spores indoors and out for an electrostatic sampler and the BioSampler, which is a form of the N6 Andersen sampler. The electrostatic sampler, which took advantage of natural charges on the spores, collected many more culturable spores than the BioSampler. This points out the fact that, although Andersen sieve plate impactors are highly efficient at spore collection, culturability may be compromised by the force of impaction, or by competition of adjacent spores impacting at the same site.
Virtual impactors don't actually capture particles by impaction, but they do take advantage of particle inertia. Particles are drawn into the device at a rate that allows the largest particles to enter a reservoir, while the smaller ones pass by into a second chamber. Here, the airflow is faster and the particles that bypass the reservoir are smaller, while the remaining larger particles are captured. Successive reservoirs capture successively smaller particles. A good discussion of virtual impactors can be found at www.tsi.com/uploadedFiles/Product_Information/Literature/Application_Notes/ITI-051.pdf.
Comments and Conclusions
Sampling for bioaerosols is a significant challenge that requires consideration of many different variables. Understanding the principles of particle sample collection and the variables affecting the efficiency of sample collection is essential if realistic sampling protocols are to be designed. In addition, while using the same sampling device for all studies is attractive for a variety of reasons, keeping an open mind about new methods that might be more efficient, reliable, and cost effective is an important part of professional practice. On the other hand, within a study, the same sampler should be used throughout, and if comparisons to other studies are planned, then it is essential to use the same sampler under the same conditions.
1. Clauß M, Springorum AC, Hartung J. 2010. Effective collection of airborne micro-organisms by direct impaction on silicone sealants-comparison of different adherent surfaces. Aerosol Science and Technology 44(11): 993-1004.
2. Grinshpun SA, Willeke K, Ulevicius V, Juozaitis A, Terzieva S, Donnelly J, Stelma GN, Brenner KP. 1997. Effect of impaction, bounce and reaerosolization on the collection efficiency of impingers. Aerosol Science and Technology 26(4):326-342.
3. Lin X, Reponen TA, Willeke K, Grinshpun SA, Foarde KK, Ensor DS. 1999. Long-term sampling of airborne bacteria and fungi into a non-evaporating liquid. Atmospheric Environment 33: 4291-4298.
4. Mainelis G, Tabayoyong M. 2010. The effect of sampling time on the overall performance of portable microbial impactors. Aerosol Science and Technology 44(1):75-82.
5. Solomon WR, Burge HA, Boise JR. 1980. Performance of adhesives for rotating arm impactors. J. Allergy Clin Immunol 65(6):467-470.
6. Trunov M, Trakumas S, Willeke K, Grinshpun SA, Reponen T. 2001. Collection of bioaerosol particles by impaction: effect of fungal spore agglomeration and bounce. Aerosol Science and Technology 35: 617-624.
7. TSI Inc.: How A Virtual Impactor Works
8. Yao M, Mainelis G. 2006. Utilization of natural electrical charges on airborne microorganisms for their collection by electrostatic means. Aerosol Science 37: 513-527.
This article was originally published on July 2011.