IDS-Environment - White Paper

Review of Aerosol Sampling Methods and Introduction of a New Low Cost Aerosol Sampler


Mahmoud Abdel-Salam & John H Dennis


Assistant Lecturer (Salam), Senior Lecturer (Dennis)


University of Bradford

There is increasing interest in the measurement of aerosol concentrations in the atmosphere, to assess their impact on the environment and, in particular, to investigate any harmful health effects. In the workplace, the impact of aerosol on the health of the workers is of a major concern. The nature and magnitude of the health risk from aerosol hazards in any given situation depends on a complex combination of factors including: particle size distribution, airborne concentration, and chemical composition (Vincent, 1995). Therefore, it is essential to measure regularly aerosol concentrations in the size fractions known to affect the human health in order to examine whether they comply with the existing standards or not.

Aerosol measurement frequently requires that an aerosol sample be conveyed to a collecting or sensing medium. This is accomplished by withdrawing a sample from its environment and transporting it through sample lines to the collecting or sensing medium. It is desirable that the sample is representative of the aerosol in its original environment and is not affected by the sampling process. Such characteristics as particle mass and number concentration and size distribution should remain unchanged during the sampling process in order to get a representative sample. However, it is difficult to prevent changes in aerosol characteristics during the sampling process because of many potential factors (Brockmann, 1993).

During the early part of this century, impingers were primarily used to obtain suspended particulate samples for area measurements, followed by microscopic and chemical analysis techniques to determine size and composition of the samples. For “grab” samples, light microscopy techniques were generally used. During the last 20 years, membrane filters have been used extensively for personal sampling while impactors and light scattering instrumentation have been used more recently in investigative activities.

 Aerosol sampling requirements and methods
The general aerosol sampling requirements for a large variety of study objectives are: (1) well-defined size fractions; (2) filter media which are compatible with the intended analysis methods; (3) sampled air volumes which are stable and provide sufficient deposit for the desired analyses without overloading the filter; (4) sampling surfaces which do not react with the measured species; and (5) available, cost-effective, and practical sampling hardware and operating procedures.

An aerosol sampler always comprises a number of components that contribute to the overall accuracy with which a sample is taken. These components are the sampling inlet, the transmission section, the particle size selector (not always present), the holder for the collecting medium, the collecting or sensing medium, calibrated flow monitoring and/or control, and a battery operated or mains electricity pump to move air flow.

There are two basic approaches to the measurement of aerosol particles. The traditional approach is to sample the particles onto a collection surface (e.g., a filter) and then analyze the collected particles. A more recently developed approach is to sample the aerosol directly into a real-time, dynamic measuring instrument. The major advantages of the traditional sample collection methods using sample collection are the simplicity and low cost of the sampling devices, their suitability for use under severe environmental conditions, and that the collected material may be subjected to different analysis techniques. The major disadvantage of the traditional sample collection techniques is the long time delay between sampling and analysis. A further drawback is the potential for chemical instability of the collected particles prior to analysis. In contrast, real-time measurement methods provide a rapid data readout with almost instantaneous information on particle concentration and size distribution. However, most real-time instruments are expensive and cannot be used under severe environmental conditions. Furthermore, these instruments usually rely on indirect sensing techniques and more calibration efforts are usually required, and also do not provide any information on the chemical nature of the particles.

Sample collection methods
Aerosol measurements may be performed by collecting the particles onto a substrate followed by subsequent analysis. The collection surface may be a porous medium, e.g., a filter that passes the air or gas flow but retains all or a fraction of the particles suspended in the incident aerosol flow. Passage through such a porous medium generally has to be at a rather low flow rate and all of the aerosol flow must pass through this medium. However, it is desirable in many applications not to affect the flow but to extract the particles from the aerosol flow by an externally applied force. The external force may be due to the gravitational pull of the earth (such as with a horizontal elutriator used to differentiate particle size) or from inertia when the particles are subjected to centrifugal motion in the bend of a duct or the helical path of a cyclone. Extractive forces may also be obtained by the application of electric, thermal, or magnetic force fields (Lehtimäki and Willeke, 1993).

Filter collection
The capture of aerosol particles by filtration is the most common method of aerosol sampling and is a widely used method for air cleaning. Filtration is a simple, versatile, and economical means for collecting samples of aerosol particles. Aerosol filtration is used in diverse applications, such as respiratory protection, air cleaning of smelter effluent, processing of nuclear and hazardous materials, and clean rooms. A wide spectrum of filters, with markedly different physical and chemical properties, is available commercially for aerosol measurement applications in a variety of filter materials, pore sizes, and collection characteristics, as well as a variety of shapes such as discs and sheets sized to fit commonly available filter holders. Filters widely used for aerosol sampling may be classified according to their structure as fibrous filters, porous-membrane filters, Nuclepore filters, and granular-bed filters.

Particle collection onto a fiber in a filter is due to the simultaneous action of several collection mechanisms (e.g., diffusion, interception, inertial impaction, gravitational settling, and electrostatic attraction) (Hinds, 1982). These five deposition mechanisms form the basis set of mechanisms for all types of aerosol particle deposition, including deposition in a lung, in a sampling tube, or in an air cleaner. The first four mechanisms are called mechanical collection mechanisms. The relative importance of these different mechanisms depends on the particle size, density, shape, and electrical charge, the filtration flow velocity and the mechanical and electrical properties of the filter.

Inertial and gravitational collectors
Inertial and gravitational collectors include impactors, cyclones, aerosol centrifuges, impingers, and elutriators. In contrast to filters that generally collect particles of all sizes, these instruments distinguish between particle sizes. They are used for size-selective sampling or size-segregated aerosol collection of airborne particles depending on the aerodynamic diameter of the particle (Hering, 1989). Cyclones, elutriators, and single-stage impactors are used as a precut for size-selective sampling and are commonly followed by a filter for aerosol collection.

Cascade impactors, cascade cyclones, aerosol centrifuges, and horizontal elutriators size-fractionate particles by aerodynamic diameter. Each stage of a cascade impactor can be analyzed to determine aerosol mass distributions or to assess chemical composition as a function of particle size. Cascade cyclones are useful under elevated temperature conditions such as those found in stack gas sampling. Aerosol centrifuges and some horizontal elutriators size-fractionate over a continuous size spectrum and are used to determine aerodynamic shape factors for irregularly shaped particles. The choice of sampler depends upon the application and the analyses to be performed.

1- Impactors
The term impactor encompasses a large class of aerosol collection instruments for which particle impaction in a non-rotating flow is the primary mechanism of particle capture. Air carrying particles passes through a nozzle and impaction occurs when streamlines bend as the air flow bypasses a solid object (i.e., a collection plate) perpendicular to the nozzle axis. Particles larger than the cut-size of the impactor will slip across the streamlines and impact upon the plate while smaller particles will follow the streamlines and penetrate the impactor stage to be either collected on a filter or passed into some other instrument for real-time mass or number concentration measurement.

A series of impactor stages (i.e., Cascade impactor) are commonly used, in which air passes from one stage to the next to remove particles in discrete size ranges (Lodge and Chan, 1986), and is exhausted through an after-filter. Each impactor stage consists of one or more jets followed by a collection plate and successive stages are designed to collect smaller particles. Those particles, which penetrate all the impaction stages are collected by a final absolute filter. The size collected by any individual stage depends upon the jet diameter and the air stream velocity in the jet as well as distance to the impaction surface; therefore, the collection of smaller particles in subsequent stages is achieved by using a combination of: smaller diameter jets, higher jet air velocities, or decreased distance to the impaction surface. Stokes number (Stk) is commonly used to describe the inertial behavior of aerosol particles. The higher the value of Stk, the more readily particles will diverge from the airflow streamlines and the more likely they are to deposit by impaction onto the obstacle. Generally, inertial impactors can be used over a wide range of conditions and have been designed for cut-sizes of 0.005 µm (Fernandez de la Mora et al., 1990) up to approximately 50 µm (Vanderpool et al., 1987) or even higher as in the rotary impactors (Hameed et al., 1983; Noll et al., 1985) and flow rates from a few cm3/min to several thousand m3/min. The impactor jets may be round or rectangular in cross section and many impactors have multiple jets per stage to permit larger sampling volumes.

The major limitations to impactors when considering their use are particle bounce from the collection surface and re-entrainment because of the continuous blowing of the jet on the particle deposit; however, this is minimized by using a sticky surface on the impaction plate, e.g. numerous types of greases and oils. Furthermore, overloading of collected particle deposits on the impaction plates and inter-stage losses are other limitations.

Because traditional impactors do not provide much size resolution for sub micrometer particles, micro-orifice and low-pressure impactors have been developed to obtain smaller cut-point diameters as small as 0.05 mm. Low-pressure impactors resemble ordinary impactors but operate at reduced pressures of 5 to 40 kPa (0.05 to 0.4 atm), while Micro-orifice impactors operate closer to atmospheric pressure (0.8 to 0.9 atm) but employ very small orifices (40 to 200 µm in diameter) and the number of nozzles is large to obtain an adequate flow rate.

Virtual impactors and inertial spectrometers are two special types of impactors. Virtual impactor is very similar to a conventional inertial impactor, with the primary difference being that the impaction plate has been replaced by a collection probe below the impactor jet in order to avoid any errors due to particle bounce or re-entrainment. Inertial spectrometer, on the other hand, provides particle size distribution information on one stage rather than several stages as in the cascade impactor dependent on particle aerodynamic diameter and air velocity (Prodi et al., 1979; Belosi and Prodi, 1987). The inertial spectrometer provides aerodynamic separation for particles in the size range of 1-10 µm, where it uses filtration for particle collection and is not susceptible to particle bounce sampling errors.

 2- Impingers
Impingers utilize inertial properties of particles to effect collection and operate much like impactors, except that the sampled air stream jet is immersed in water at the bottom of a flask. The sampled air stream is accelerated in the impinger orifice to velocities of 60 m/s or greater. The air stream exits underneath the liquid surface immediately above an impaction plate or at a specified distance above the bottom of the collection flask. Particles impinge on the plate or flask bottom, stop, and are subsequently retained by the liquid. Impingers are effective for the collection of particles in the 1 µm to 20 µm size range. The lower size collected depends on jet velocity and diameter. The upper size collected by impingers is limited because large particles cannot follow the air stream into the impinger

3- Cyclone samplers
Cyclones use centrifugal forces for particle collection and utilize a vertical flow inside a cylindrical or conical chamber. Particles with sufficient inertia are unable to follow the air streamlines and impact onto the cyclone walls. The particles are either retained on the cyclone walls or migrate to the bottom of the cyclone cone, while finer particles remain entrained in the air during the upwards spiral.

Aerosol sampling cyclones both in the workplace and in ambient air are small, typically 1 cm to 5 cm in diameter, operating at flows as small as a few liters per minute. Respirable particle sampling is one of the most common applications of cyclones, wherein the cyclone is operated upstream of a filter. The cyclone is used to remove the larger, non-respirable particles such that the material collected on the filter is representative of that which penetrates into the non-ciliated deep lung spaces of humans. Cyclones are also used in ambient air sampling to separate the coarse mode aerosols (particle diameters > 2 mm) from the fine mode aerosols (particle diameters < 2 mm). In most cases, cyclones are used to provide a particle precut to another aerosol collector such as a filter or an impactor. The collection efficiency is dependent upon the particle aerodynamic diameter and the cut-size of the cyclone is governed by the flow rate, the size of the inlet and outlet tubes, the gas viscosity, and the size of the cylinder (Marple et al., 1993).

Cyclones have several advantages in air sampling including their large capacity for loading and insensitivity to orientation. Furthermore, they have relatively low cost of construction, ease of operation, and are easily maintained. Unlike impactors, they are not subject to errors due to particle bounce and re-entrainment and do not require special collection surface coatings. However, the flow pattern inside cyclones is complex and not easily modeled; thus, it is not easy to predict cyclone performance without reference to empirical correlations.

4- Elutriators
Elutriators use gravitational settling in a laminar flow to separate particles by aerodynamic diameter. They provide segregation for particles greater than 3 µm. Two types of elutriators, vertical and horizontal, are available. The vertical elutriator consists of a vertical duct through which air flows slowly upward. Particles whose sedimentation velocity is greater than the duct velocity cannot follow the air flow and settle out. The ability of the device to distinguish particle sizes is reduced by the distribution of velocities in the duct; nonetheless, it is an effective method for removing large particles. One example of vertical elutriators is the cotton dust sampler designed for the sampling of particles below 15 µm using a flow rate of 7.4 l/min.

In a horizontal elutriator, aerosol is passed slowly along a horizontal channel. Particles settle onto the bottom of the flow channel at locations dependent upon the particle size, particle density, gas velocity, and channel height. The larger aerodynamic particles settle near the entrance while the smaller particles are deposited near the exit. Two devices that incorporate this technique for particle classification are the horizontal elutriator used for respirable dust sampling, e.g. Hexhlet horizontal elutriator and the Timbrell aerosol spectrometer (Timbrell, 1972). Their main advantage is the predictable performance based on gravitational settling of the particles during passage between horizontal collecting plates. Disadvantages include the restriction to a fixed orientation (wind-dependent), possible re-entrainment of particle deposits and the difficulty of miniaturization.

5- Aerosol centrifuges
Aerosol centrifuges refer to a class of aerosol samplers that spin at high velocity in order to subject particles to large centrifugal force which is used to deposit the particles on the outer edge of an aerosol chamber. Although several centrifuges have been developed, spiral centrifuge is the one that has seen the greatest application (Stöber and Flaschbart, 1969; Hoover et al., 1983). Generally, particles are collected on a foil which lines the outer wall of the channel and is removed for particle analysis after collection where the deposition distance along the foil is directly related to the aerodynamic diameter of the particle. The resolution of the centrifuge depends on the ratio of the aerosol sampler flow rate to the total flow rate and the overall particle classification may be achieved within the range 0.01 < dae < 5 μm at sampling flow rates up to about 1 l/min.

Conventional aerosol centrifuges can be easily calibrated and provide exceptionally high particle size resolution but generally operate at relatively low sample flow rates (usually less than 1 l/min). At such low sample flow rates, the amount of particulate material collected at any given part of the instrument is too small to be assessed gravimetrically and also substantial biases to the particle aerodynamic size distribution may be expected at such a low sampling flow rate. For particles greater than a few micrometers, most centrifuges are subject to inlet losses. However, in contrast to impactors, particle bounce is not a problem because centrifugal force holds the particles tightly to the collection surface.

Electrostatic and thermal precipitators
Electrostatic precipitators
Electrostatic precipitation differs from other particle collection mechanisms in that, electrical forces rather than inertial or thermal are used to separate the particles from the gas in which they are suspended. Electrical force is exerted directly on the particles instead of on the whole gas volume; thus, relatively little power is required to precipitate the particles or to move the gas steam through the collector. Electrostatic precipitator samplers have two significant advantages over filter samplers: 1) the sampling rate is not affected by mass loading; and 2) the sample is in a readily recoverable form.

The collection of a particle by electrostatic precipitation involves two separate and distinct operations. First, the particle must acquire electric charges, and second, the charged particle must be accelerated toward an electrode of opposite polarity by the electric field. Particles can acquire electric charges by several mechanisms; however, for speed, efficiency, and controllability of the charging process, the high voltage corona discharge is used.

The collection efficiency of an electrostatic precipitator sampler is dependent on the operating parameters, e.g., current, voltage, and flow rate; the particle parameters, e.g., particle size, shape, dielectric properties, and mass loading; and the carrier gas parameters, e.g., humidity, ambient pressure, temperature and composition (Hering, 2001). Collection efficiency is increased by high charging currents, high voltage gradients, and low flow rates. For particles > 0.5 μm diameter, the charging, and hence the collection efficiency, is strongly dependent on the potential gradient in the charging field; whereas for smaller particles, the charging current, i.e., the number of air ion charge carriers, is more significant. Re-entrainment of particles from the collection surface, which occurs as a result of high voltage sparking and strongly dependent on the air velocity, is a disadvantage where large particles are more easily re-entrained than smaller ones.

Thermal precipitators
A thermal precipitator removes particles from an aerosol by passing it through a relatively narrow channel having a significant temperature gradient perpendicular to the direction of the flow. The movement of a particle in the direction of decreasing temperature, called its thermophoretic velocity, causes the particle to deposit on a collecting surface appropriate to the type of subsequent evaluation. Thermal precipitators collect virtually all particles from 5 μm down to 0.01 μm and probably smaller with high collection efficiency, provided that a sufficient thermal gradient is established in the sampling region.

The very high efficiency of collection of sub-micron particles and the little effect of the degree of charge on the particle on the collection efficiency are important advantages of thermal precipitators. However, the low sampling rate of thermal precipitators (ranging from 7 cm 3/min to 1.0 l/min) is unsuitable for some applications. Another disadvantage is that, many relatively volatile aerosols could not be collected in a thermal precipitator because of the possibility of evaporation in the vicinity of the heated surface. Finally, the standard thermal precipitator is bulky and has rather poor size selection characteristics for separating distinct aerosol size fractions.

Real-time measurement methods
In real-time or direct-reading instruments, sampling and analysis are carried out within the instrument and the property of interest can be obtained almost immediately. Different aerosol properties, such as number concentration, mass concentration, opacity, size distribution, etc., are measured by different direct-reading instruments. Direct-reading instruments use a sensing region; an aerosol is either passed through or collects upon this region. The presence of the particles gives rise to a change in some property of the zone that is detected, followed by establishing a simple relationship between the detected change and a property of the aerosol.

Direct-reading instruments are now available to measure aerosol number concentration up to 106 particles/cm3 , mass concentration up to 1000 mg/m3, and size distribution over a particle size range of 0.003 μm to over 100 μm (Pui and Swift, 1994). These instruments may be classified into three main categories: optical, resonance oscillation, and beta particle attenuation.
Optical instruments employ the interaction between airborne particles and visible light in a sensing region, where they are based on the principles of light extinction and light scattering. Generally their response is dependent upon the size distribution and refractive index of the particles (Mark, 1998). Optical particle counters, photometers and nephelometers are examples of optical direct-reading instruments.

In resonant oscillation monitors designed for real-time monitoring of particulate mass concentration, the frequency of mechanical oscillation of an element, such as a piezoelectric crystal (e.g., quartz) or a tapered glass tube, in the piezobalance and the Tapered Element Oscillating Microbalance (TEOM) respectively, is directly proportional to the mass of that element. Changes in effective mass of that element, such as that due to the deposition of particles on its surface, are reflected in the change in its mechanical resonant frequency.
Beta attenuation concept, on the other hand, is where the mass of particulate material deposited on a filter or some other surface is determined from the reduction in intensity of beta particles passing through the accumulated layer. The attenuation of beta particles is directly dependent on the mass, and in almost independent of aerosol type or particle size distribution (Macias and Husar, 1976).

Development of a new low cost aerosol sampler
A new design of aerosol sampler (Figure 1) has been patented and developedin the University of Bradford. A battery powered fan is used to pull aerosol laden air through an electrostatic collection medium. The concept of using electrostatic filters to sample aerosol in this way is a new one as is the application of a fan to provide the flow through the sampler.

Figure 1. Prototype aerosol sampler incorporating AA battery operated low power fan, size selecting foams and low resistance electrostatic filter.

Filtration air sampling has traditionally relied on the use of pumps to produce the relatively high pressure required to pull air through those filters that are normally used, e.g., membrane filters. With a focus on personal sampling, this method despite developments in technology, still relies on heavy expensive pumps and loose tubing. This has implications both for safety in hazardous workplace and for employee acceptability and comfort.

Electrostatic filters of the type investigated can be used for several applications where high efficiency, low-pressure drop and long lifetime are required. This includes vacuum cleaner exhaust filters, room air-purifier filters, gas anesthetic breathing filters, pulmonary function filters, hard disc drive re-circulation filters, air-conditioning filters and vehicle ventilation filters. The filtrete fibers are made of polypropylene as a raw material in several basis weights (ranging from 20 to 300 g/m2) and the fibers are electrostatically charged. Electrostatic charges are in the bulk of the material and thus not affected by environmental influences. The filtrete medium is uniquely constructed of permanently charged rectangular split fibers that are able to capture airborne particles. The fibers are charged bipolarly and the charge density is about 50 nC/cm2 . In the cross-section, the rectangular fibers have a thickness of about 10 μm and a width of about 60 μm and the solidity is about 7%. The performance level of the filtrete media shows no significant degradation within a temperature range of -40˚C to +90˚C and a humidity range of 5% to 95%.

The main advantage of using an electrostatic filter in an aerosol sampler is the low pressure required to pass air through the sampler, e.g., the nominal 2 liters per minute used in personal sampling. This means that a suitable flow can be achieved by the use of a low-power fan rather than a high-power pump. This decreased air resistance leads to significant reduction in power requirement, which in turn reduces size, weight, noise emission, and allows development of a self contained sampling unit incorporating battery, pump, and filtrations housing in a small portable device. An additional advantage over the conventional precision pumps is that there is likely to be a reduction in cost per unit when manufacturing these small aerosol sampling devices in large numbers at the production stage – though the University of Bradford aerosol sampler remains is in (an advanced) prototype design stages

Furthermore, the new sampler is able to separate the aspirated aerosol sample according to the new aerosol sampling conventions (inhalable, thoracic, and respirable). The standard IOM sampling inlet is planned to be incorporated within the new aerosol sampler that enables the aspiration of the inhalable fraction. On the other hand, both thoracic and respirable fractions can be separated by using two specific types of foams.

The new aerosol sampler being developed at the University of Bradford may be used in a variety of applications in assessing both environmental and occupational aerosol levels as an alternative to currently available aerosol samplers. However, further development and validation is required over the next 6-12 months.

There are various methods for aerosol sampling both in ambient environment and workplace. Sample collection and real-time measurement methods are two basic approaches to the measurement of aerosol particles. The former is a traditional approach in which aerosol particles are collected onto a collection surface for further analysis, while the latter is to sample the aerosol directly into a direct-reading instrument for almost instantaneous information.

The University of Bradford has patented and is developing a new design of aerosol sampler that integrates a low-resistance filter and consequently low-power fan and battery system within an integrated unit. The new sampling system has many potential advantages over traditional aerosol sampling technology including low cost and portability. Though a prototype of the new sampler has been developed, it is still under investigation to evaluate it as a realistic alternative to other aerosol sampling systems.


Belosi, F., and Prodi, V. (1987). Particle deposition within the inertial spectrometer. Journal of Aerosol Science, 18, 37-42.

Brockmann, J. E. (1993). Sampling and transport of aerosols. In: Aerosol Measurement – Principles, Techniques and Applications (eds. K. Willeke and P. A. Baron), Van Nostrand Reinhold, New York, pp. 77-111.

Fernandez de la Mora, J., Hering, S. V., Rao, N., and McMurry, P. H. (1990). Hypersonic impaction of ultrafine particles. Journal of Aerosol Science, 21, 169-187.

Hameed, R., McMurry, P. H., and Whitby, K. T. (1983). A new rotating coarse particle sampler. Aerosol Science and Technology, 2, 69-78.

Hering, S. V. (1989). Inertial and gravitational collectors. In: Air Sampling Instruments, 7th edn., (ed. S. V. Hering), American Conference of Governmental Industrial Hygienists, Cincinnati, OH, pp. 337-385.

Hering, S. V. (2001). Impactors, cyclones, and other particle collectors. In: Air Sampling Instruments, 9th edn., (eds. B. S. Cohen and C. S. McCammon, Jr.), American Conference of Governmental Industrial Hygienists, Cincinnati, OH, pp. 315-380.

Hinds, W. C. (1982). Aerosol Technology. John Wiley and Sons, New York.

Hoover, M. D., Morawietz, G., and Stöber, W. (1983). Optimizing resolution and sampling rate in spinning duct aerosol centrifuges. American Industrial Hygiene Association Journal, 44, 131-134.

Lehtimäki, M., and Willeke, K. (1993). Measurement methods. In: Aerosol Measurement – Principles, Techniques and Applications (eds. K. Willeke and P. A. Baron), Van Nostrand Reinhold, New York, pp. 112-129.

Lodge, J. P., and Chan, T. L. (1986). Cascade Impactor, American Industrial Hygiene Association, Akron, Ohio.

Macias, E. S., and Husar, R. B. (1976). Atmospheric particulate mass measurement with beta attenuation mass monitor. Environmental Science and Technology, 60, 904-914.

Mark, D. (1998). Transport and sampling of aerosols. In: Physical and Chemical Properties of Aerosols (ed. I. Colbeck), Blackie Academic & Professional, pp. 80-152.

Marple, V. A., Rubow, K. L., and Olson, B. A. (1993). Inertial, gravitational, centrifugal, thermal collection techniques. In: Aerosol Measurement – Principles, Techniques and Applications (eds. K. Willeke and P. A. Baron), Van Nostrand Reinhold, New York, pp. 206-232.

Noll, K. E., Pontius, A., Frey, R., and Gould, M. (1985). Comparison of atmospheric coarse particles at an urban and non-urban site. Atmospheric Environment, 19, 1931-1943.

Prodi, V., Melandri, C., Tarroni, G., De Zaiacomo, T., Formignani, M., and Hochrainer, D. (1979). An inertial spectrometer for aerosol particles. Journal of Aerosol Science, 10, 411-419.

Pui, D. Y. H., and Swift, D. L. (1994). Direct-reading instruments for airborne particles. In: Air Sampling Instruments, 8th edn., (eds. B. S. Cohen and S. V. Hering), American Conference of Governmental Industrial Hygienists, Cincinnati, OH, pp. 337-367.

Stöber, W., and Flaschbart, H. (1969). Size-separating precipitation of aerosols in a spinning spiral duct. Environmental Science and Technology, 3, 1280-1296.

Timbrell, V. (1972). An aerosol spectrometer and its applications. In: Assessment of Airborne Particles: Fundamentals, Applications and Implications to Inhalation Toxicology (eds. T. T. Mercer, P. E. Morrow and W. Stöber), Charles C. Thomas, Springfield, III., pp. 290-330.

Vanderpool, R. W., Lundgren, D. A., Marple, V. A., and Rubow, K. L. (1987). Cocalibration of four large particle impactors. Aerosol Science and Technology, 7, 177-185.

Vincent, J. H. (1989). Aerosol Sampling: Science and Practice. John Wiley and Sons, Chichester, UK.

Vincent, J. H. (1995). Aerosol Science for Industrial Hygienist. Elsevier Science.