Analyst, January 1994, Vol. I I9 19 Measurement of Fine Aerosols in Workplaces* A Review James H. Vincent Division of Environmental and Occupational Health, School of Public Health, University of Minnesota, Box 807 Mayo, 420 Delaware Street SE, Minneapolis, MN 55455, USA The fine aerosol fractions of interest in occupational health are those associated with the regional deposition of inhaled aerosols in the different parts of the human respiratory tract. The recent recommendations of the International Standards Organization, the American Conference of Governmental Industrial Hygienists and the Comit6 Europetin de Normalisation define the thoracic fraction (inhaled particles which penetrate into the lung) and the respirable fraction (inhaled particles which penetrate further into the alveolar region).Based on the results of inhalation experiments with humans, conventional curves for these fractions are proposed, which provide ‘yardsticks’ for health-related sampling devices. Samplers for the respirable fraction have been in existence since the early 195Os, and a fairly large range of acceptable instrumentation is available commercially. Recently, however, attention has been drawn to the effect of even small departures from the ‘target’ curve on the respirable mass sampled. This suggests the need for additional criteria for sampler acceptability, based on sampled mass in addition to particle size selectivity. Samplers for the thoracic fraction may be based on similar physical concepts to those for respirable aerosol. In addition, versatile devices capable of providing information about more than one fraction simultaneously are being proposed and are finding increasing use as investigative tools for industrial hygienists.Finally, measurement of fine fibrous aerosols is recognized as a special case, and so is treated separately. Keywords: Fine aerosols; workplace monitoring Introduction The preceding paper1 sets out the background describing the nature of aerosol exposure in workplaces and the ideal basis of standards aimed at protecting workers. It then goes on to describe the rationale and variety of technical options for the sampling and measurement of coarse aerosol-the inhalable fraction (previously referred to as ‘total’ aerosol). This paper proceeds to extend those ideas to the measurement of fine health-related aerosol fractions.The Human Respiratory Tract The human respiratory tract is a complex system, its primary objective being to supply the body with oxygen (during inhalation) and to eliminate carbon dioxide (during exhala- tion). For the present purposes, a rudimentary description is sufficient. As shown in Fig. 1, the inhaled air enters the respiratory tract through either the nose or the mouth. It first encounters the upper region of the respiratory tract, compis- * Presented at the Conference on Modern Principles of Workplace Air Monitoring: Pumped and Diffusive Sampling for Contaminants, Geilo, Norway, February 15-18, 1993. ing the airways of the nose and mouth which, together, form the nasopharynx. The over-all ‘head’ region is assumed in the present context to include both the nasopharynx and the larynx.The whole region below the larynx is the lung, referred to as the thoracic region. Air enters through the larynx into the trachea, which in turn leads to a branching system of conducting airways referred to as bronchi and, eventually, the conducting and terminal bronchioles. Beyond the terminal bronchioles is the alveolar region where the gas exchange takes place. This consists of the respiratory bronchioles, alveolar ducts and alveolar sacs. Aerosol Regional Deposition in the Respiratory Tract The first part of the process of aerosol exposure occurs during the entry of particles into the respiratory tract during inhalation. The second part involves the arrival of particles at relevant sites inside the respiratory tract as the result of deposition by a combination of physical mechanisms (includ- ing sedimentation, impaction, diffusion and electrostatic forces). This has been investigated by aerosol inhalation experiments involving human volunteer subjects.Such experi- ments have permitted the determination of the, probability of deposition as functions of partical aerodynamic diameter (d,,) in the extrathoracic (nasopharyngeal), tracheobronchial (the conducting airways) and the alveolar regions. The efficiency of thoracic deposition is defined as amount deposited in tracheobronchial + alveolar regions amount inhaled Cthor = (1) Based on typical data from the literature,2-4 this is shown as a function of d,, in Fig. 2. The efficiency of alveolar deposition is amount deposited in the alveolar region amount inhaled G I ” = (2) Nasopharynx Larynx Trachea Bronchi Bronchioles Alveoli Fig.1 Simplified schematic diagram of the human respiratory tract20 Analyst, January 1994, Vol. I 1 9 for which typical data2 are shown in Fig. 3. These show that Calv falls to zero for particles with d,, larger than about 10 pm, reflecting the non-availability of larger particles due to the filtration effect of the upper respiratory tract. As d,, decreases, Calv exhibits a maximum around d,, = 3 ym, then decreases again. Although not revealed by the data as plotted, Calv reaches a minimum at around d,, = 0.5 pm then rises again for still smaller particles (as a result of increasing particle deposition by diffusion).The dip in Calv for dae below about 3 pm represents the exhalation of undeposited particles. It should be noted that none of the above relates to fibrous aerosols. Because there have been no experiments with human subjects for such substances (because of the very significant health risk that is associated with the fibrous and, frequently, durable nature of such particles), there are no data which correspond to those for non-fibrous particles. Criteria for the Measurement of Finer Aerosol Fractions The experimental data in Fig. 2 provided the basis of the conventional curve for thoracic aerosol proposed in 1983 by 10 20 30 Particle aerodynamic diameter d,&m Fig. 2 Typical experimental data (for human subjects) for thoracic deposition; also shown is the conventional particle size-selection curve for the thoracic fraction6 1 - 8 2 8 5 .E 0.5 8 2 a, 0 5 Particle aerodynamic diameter d&m Fig.3 Typical experimental data (for human subjects) for alveolar deposition; also shown are various conventional curves for the respirable fraction: A, the BMRC curve;5 B, the 1985 ACGIH curve;6 and C, the proposed new curve7 the International Standards Organization (ISO)5 and in 1985 by the American Conference of Governmental Industrial Hygienists (ACGIH);6 it is the same for both. This curve takes the form of a cumulative log-normal function with its median at dae = 10 pm and having a geometric standard deviation (a& of 1.5. It is seen from Fig. 2 that the conventional curve lies above the majority of the data points, the rationale being that, if anything, the conventional curve should err on the side of the ‘worst-case’ situation.Of the thoracic aerosol that penetrates below the larynx, a further subdivision takes place between what is deposited in the tracheobronchial region and that which penetrates down to the alveolar region. The fine fraction penetrating down to the alveolar region is known as the respirable fraction. In Fig. 3 are shown, alongside the alveolar deposition data from experiments with human subjects, a number of the curves that have been adopted over the years as conventions for the respirable fraction. These include the historically important British Medical Research Council (BMRC) curve, which first emerged in the early 1950s, and the 1985 ACGIH curve. Also shown is the new version proposed as a basis for international harmonization on the matter.7 Again, it is noted that none of these conventional respirable aerosol curves truly reflect alveolar deposition at very small particle sizes.This does not pose a problem so long as the aerosol in question does not contain a significant proportion of its mass in particles of diameter below about 1 pm. However, caution should be exercised in applying such respirable aerosol conventions to mixed aerosols where there are important components falling almost entirely in the sub-micrometre particle size range (e.g., diesel fume in the presence of dust). The latest criteria for health-related aerosol standards, representing the results of harmonization between IS0 and ACGIH, and latterly the ComitC Europe& de Normalisation (CEN), include recognition of the various processes by which aerosols come into contact with the human respiratory system.These reflect (a) aerodynamic processes outside the body by which particles are inhaled through the nose andor mouth and (b) aerodynamic processes inside the body whereby those inhaled particles are deposited. Thus inhaled aerosol is seen to be a fraction of the true total workplace aerosol. In turn, thoracic and respirable aerosol are sub-fractions of the inhalable fraction. For the fine aerosol fractions, there is now general agreement (led by ISO, ACGIH and CEN) on the following quantitative definitions: For the thoracic fraction, described as a sub-fraction of inhalable aerosol, the median d,, value for the cumulative log-normal curve is 11.45 pm and the geometric standard deviation (og) is 1.5; this means that, as a fraction of true total aerosol, the median dae value would be 10 pm, as described in the 1983 I S 0 and 1985 ACGIH recommendations.For the respirable fraction, again described as a sub-fraction of inhalable aerosol, the median for the cumulative log- normal curve is d,, = 4.25 pm and the geometric standard (og) is again 1.5; this means that, as a fraction of true total aerosol, the median d,, value would be 4 pm. In the new recommendations, tolerance bands are defined as before. However, as already discussed,’ the practical application of such tolerance bands in assessing the accept- ability of a given sampling instrument is not sufficient alone. In practice, ‘acceptable performance’ must be defined by also requiring that an instrument’s performance allows the collec- tion of aerosol mass which, under expected practical condi- tions, falls sufficiently close to what would be expected using a ‘perfect’ sampler.How to achieve this within a framework for an overall performance standard is still under consideration .g Fine fibrous aerosols can, as already mentioned, pose a particularly serious risk to health, and so represent a specialAnalyst, January 1994, Vol. 119 21 problem for the occupational hygienist. Asbestos-containing dusts come into this category. Because of their unusual morphological characteristics, such fibres are specifically not included in the above conventions. Instead, there are different criteria for measurement, based on an appreciation of both the aerodynamic characteristics of particle motion which govern fibre deposition in the deep lung and the biological effects that can then ensue and govern the subsequent fate of the particles.With these equally important factors in mind, it has been a common convention, since the 1960s, to assess ‘respirable’ fibres in terms of the airborne number concentration of particles which, when examined by optical microscopy under phase contrast conditions, have an aspect ratio greater than 3, a length greater than 5 pm and a diameter less than 3 pm, as recommended, for example, by the Asbestos International Association (AIA) .9 Standards Based on Fine Aerosol Fractions One health-related class of aerosols is for particles that may lead to adverse health effects after deposition in the tracheo- bronchial region of the lung.In this category are substances that can provoke local responses leading to such effects as bronchoconstriction, chronic bronchitis and bronchial carci- noma. For the health-related measurement of such aerosols, it is appropriate to think in terms of sampling according to a tracheobronchial criterion. However, for practical purposes, the overall thoracic fraction is more convenient to define and measure, and is sufficient. Another class is for particles that deposit in the alveolar region of the lung. Here local effects include the pneumo- conioses (e.g., silicosis and asbestosis), emphysema, alveolitis and pulmonary carcinoma. In relation to these, the respirable aerosol fraction represents a convenient and appropriate sampling criterion. As discussed previously for the inhalable fraction,l any such criteria need to be applied within sampling strategies that best reflect individual worker exposures to health-related sub- fractions and reflect the dynamics of possible health effects as well as the variability of exposure.Sampling for Fine Aerosol Fractions The history of sampling fine aerosols in workplaces began with the respirable fraction, in particular with the emergence in the 1950s of the BMRC respirable aerosol criterion. A number of types of sampling device have since been developed. Most have in common the fact that they first aspirate a particle fraction which is assumed to be representative of the total workplace aerosol, from which the desired fine fraction is then aerodynamically separated inside the instrument using physical options (e.g., elutriation, cyclone) with particle size-dependent penetration characteristics matching the desired criterion.The fine fraction of interest is that which remains uncollected inside the selector and passes through to collect on to a filter or some other collecting medium. Static (or Area) Samplers for the Respirable Fraction According to the philosophy embodied in the latest sampling criteria, if an instrument is to be used for collecting a fine aerosol fraction corresponding to deposition in a particular region of the respiratory tract, it should (ideally) first aspirate the inhalable fraction. Otherwise, if a sampler has a poorly defined aspiration efficiency, or one which varies in an uncontrolled way with (say) wind speed or orientation, then bias can result in the determination of the fine fraction of interest.Therefore, in samplers intended primarily for deter- mination of finer fractions, it is relevant to consider their aspiration efficiency in relation to the inhalability criterion. Static samplers commonly used in Europe for sampling the fine respirable fraction include, for example, the British 2.5 1 min-1 MRE Type 113A (see Fig. 4) with a pre-selector based on horizontal elutriation principles, and the German 50 1 min-1 TE3F50 (not shown) with a pre-selector based on the cyclone. Both were developed for applications in coal mines. Data for the aspiration efficiency of these instruments show that neither instrument matches the inhalability criterion particularly well for particles in the coarse size range exceeding about 10 pm.10 However, for finer particles in the respirable range, the agreement is much better.Here, reasonable consistency with the ideal inlet efficiency require- ment embodied in the latest sampling recommendations is maintained under most conditions. However, results for the performance of the MRE instrument in relation to the collection of respirable dust do show a strong dependence on wind speed, especially above 5 m s-1.11-13 These strongly suggest the onset of significant changes in aspiration efficiency at higher wind speeds. Caution is therefore recommended in interpreting respirable dust data, obtained using such instru- ments under these conditions.Regarding their performance for the finer respirable fraction, experimental results for the internal selection properties of both the above instruments are in good agreement with the BMRC curve. Personal Samplers for the Respirable Fraction Horizontal elutriators have been found to be very satisfactory for static respirable aerosol sampling, but they are inevitably rather bulky and not conducive to miniaturization. Therefore, horizontal elutriation is not promising for personal respirable (4 Parallel plate elutriator Glass-fibre membrane Filter clip reservoir - / - Terylene backing filter Adjustable crank \/ Valves Fig. 4 British MRE Type 113A static sampler for respirable aerosol (sampling flow rate 2.5 1 min-I), with pre-selector based on the horizontal elutriator principle22 Analyst, January 1994, Vol.119 aerosol samplers. On the other hand, cyclones are ideally suited for such purposes, and so have found wide application. Well known examples are the British 1.9 1 min-1 Casella cyclone (see Fig. 5) and the equivalent American 1.7-2.1 1 min-1 10 mm cyclone (not shown). Experimental data for their selection characteristics are in good agreement with the BMRC and ACGIH curves, respectively. Unfortunately, there are no aspiration efficiency data available for such samplers. However, based on the experience gained for other types of sampler, it is reasonable to assume that the aspiration efficiency will follow the inhalability curve fairly well for fine particles in the respirable size of primary interest in the use of these devices.Fig. 5 The 1.9 1 min-1 cyclone-based personal sampler of the type widely used for respirable aerosol in the UK There are two other devices that have emerged relatively recently as samplers primarily for finer aerosol fractions, having some interesting and unusual features and so deserving special mention. The first is the French CIPlO14 (see Fig. 6). It is particularly interesting because it incorpor- ates its own built-in pumping unit, consisting of a battery- driven, rapidly rotating polyester foam plug. Aerosol is aspirated through a downward-facing annular entry and is collected efficiently by filtration in two stationary, coarse- grade plastic foam plugs located inside the entry as well as on the finer-grade rotating plug.As a result of the low pressure drop characteristics of such foam filtration media, a flow rate of up to 10 1 min-1 can be achieved. By personal sampler standards, this is very large indeed. For this device, the fine-fraction pre-selector operates on the basis of foam filtration, where aerosol entering porous polyester foam media is collected by a combination of gravitational settling and inertial forces. Experimental data for the fine aerosol selection characteristics of the CIPlO lie approximately midway between the BMRC and ACGIH curves, and so come close to matching the new curve for the respirable fraction. One interesting feature of the CIPlO is the observed decrease in penetration for small particles with dae less than about 3 pm.This is the result of the penetration of fine particles through the final rotating collector foam (and so their escape from the instrument). This feature brings the performance of the CIPlO more closely into line with true alveolar deposition than any of the other respirable aerosol samplers described. The second interesting personal sampler is the Italian PERSPEC, a device aimed at collecting not only total aerosol but also the finer thoracic and respirable fractions (see Fig. Fig. 6 French CIPlO personal sampler for respirable aerosol.14 It contains its own air moving apparatus and operates at a sampling flow rate of 10 1 min-1 Fig. 7 Italian PERSPEC personal sampler,15 developed primarily as a sampler for the fine thoracic and respirable fractions, but also having potential for use for inhalable aerosol.Sampling flow rate is 2 I min-lAnalyst, January 1994, Vol. 11 9 23 7).15 Aerosol enters at 2 1 min-1 through a pair of crescent- shaped orifices and is separated by inertial forces into the finer sub-fractions of interest (which are deposited on different, well defined parts of the same filter). Sampling for ‘Respirable’ Fibres Because the definition of a ‘respirable’ fibre is based on purely geometric criteria, selection is best carried out not aerody- namically but visually under the microscope. This means that, in practical sampling, the main priority is to achieve deposition on a suitable surface (e.g., a membrane filter), which can then be ‘cleared’ and mounted for subsequent visual analysis by optical microscopy under phase-contrast conditions.It follows that actual physical sampling can be very simple, usually involving the collection of particles directly on an open filter (sometimes with the use of a cowl or some other baffle to protect the filter from large airborne material). Such sampling is carried out routinely in both the static and personal modes. In asbestos measurement, great emphasis is placed on the visual assessment of the sampled fibres. For routine assess- ment of workplace asbestos, this is usually carried out using an optical microscope under phase-contrast conditions at a magnification of ~ 4 5 0 . An appropriate graticule is used to provide ease of classification of fibres matching the criteria referred to earlier.Sets of ‘counting rules’ have been recommended to aid the microscopist in what, and what not, to count, guiding, for example, the assessment of fibrous aggregates, fibres in the presence of other, non-fibrous particles and fibres not fully contained within the microscope field of view. The technical methods for sample preparation and microscopy and the processes of selecting and counting fibres have been extensively researched and fully documented in the various reference methods that have been published.9 One important practical aspect is the setting of the sampling flow rate, as sufficient flow is required to achieve, over a sampling shift, a sample that is dense enough to provide good counting statistics and reliable visual counting,16 yet not so dense as to cause problems with fibre overlap.17 As far as the effects of sampling flow rate on aspiration efficiency for fibres are concerned, it has been demonstrated that, over a very wide range of flow rates, fibrous particles of asbestos are so fine that aspiration efficiency is nearly always close to unity.18 Therefore, for practical purposes, sampling bias due to aspiration effects can be neglected, providing considerable flexibility in the choice of flow rate in a given situation.Sampling for More Than One Fraction Simuhneously The important concept that thoracic aerosol is a sub-fraction of the inhalable fraction and that respirable aerosol is a further sub-fraction of the thoracic sub-fraction provides a framework by which all three aerosol fractions can be obtained simul- taneously.Such aerosol measurements could be important for assessing aerosol-related risk in certain situations, and appropriate practical sampling devices are just beginning to emerge. One interesting such personal sampler is the Italian PERSPEC already mentioned (see Fig. 7). A second instru- ment20-22 is derived directly from the Institute of Occupational Medicine (IOM) personal inhalable aerosol sampler (see Fig. 8)23 as described in the preceding paper.’ Here, aerosol is again aspirated through a 15 mm circular entry and, as before, the entry forms an integral part of an aerosol-collecting capsule which acts as a receptacle for the whole inhalable fraction. Now, however, the capsule is extended in length in order to house two porous polyester foam selectors, each using different grades of foam.The first is chosen ( i t ? . , grade of foam, dimensions) to provide penetration characteristics matching the thoracic fraction. The second selector, placed immediately behind the first, is chosen to provide penetration characteristics matching the respirable aerosol curve. In the practical use of this instrument, the whole capsule is weighed before and after sampling (to provide the inhalable mass fraction). Then the second (fine) foam plug and the backing filter are removed and weighed separately. The sum of the resultant two masses provides the thoracic mass fraction. The mass on the backing filter is the respirable mass sampled. This instrument is still at the prototype stage at the time of writing, so is not yet available commercially.Aerosol Spectrometers In principle, if we know the particle size distribution and the mass of the sampled aerosol, then we can determine the particle size distribution and mass contained in any sub- Sampling for the Thoracic Fraction Methodology for the sampling of thoracic aerosol in the occupational context was not widely considered prior to the emergence of the new IS0 and ACGIH criteria. For workplaces, the nearest we have come to a thoracic aerosol standard is in the US cotton industry, where a criterion was established in 1975 by the US National Institute of Occupa- tional Safety and Health (NIOSH), based on a selection curve which falls to 50% at 15 pm. This seems to imply recognition of the role of particle deposition in the large airways of the upper respiratory tract in cotton workers’ byssinosis.The recom- mended static sampling method employs the concept of vertical elutriation.19 Now, as the ISo and recommendations begin to be translated into new standards, more energetic consideration is Fig. 8 IOM 2 1 min-1 persona] sampler developed for the simu]- taneous sampling of inhalable, thoracic and respirable aerosol .20-22 It being given to the development of samplers for the thoracic fraction, and first attempts are being based On modification Of existing respirable aerosol samplers. incorporates an inhalable entry23 and two porous foam pre-selecton (located inside the nose-piece of the sampling head) operating on filtration principles24 Analyst, January 1994, Vol. 11 9 fraction.Aerosol spectrometers that can provide such infor- mation are more versatile than the dedicated samplers described above as they can provide data about any number of sub-fractions from just one sample. This can have important implications, in particular for epidemiological research. For example, in one recent study, the approach outlined above has been used to examine the effects on the actual dust uptakes (or deposited lung dose) in mineworkers of different breathing patterns, and hence different lung deposition characteristics, associated with different work rates.24 A wide range of physical possibilities exists upon which to base a family of aerosol spectrometer devices. The type that has achieved the greatest popularity since it first emerged in the 1940s is the cascade impactor.In this device, sampled aerosol passes through a succession of impactor stages, each taking the form of a jet directed on to a solid surface. Particle deposition takes place by impaction on to the surface, strongly dependent on particle aerodynamic size, jet width and jet air velocity. Decreasing jet width at each successive stage ensures that smaller and smaller particles are deposited as the aerosol penetrates from stage to stage. From the masses of aerosol collected at each stage, together with knowledge of the particle deposition (or ‘cut’) characteristics of the impactor stages, the cumulative and, in turn, the frequency size distribution of the sampled aerosol can be obtained. More detailed information of the principles, performances and types of cascade impactors and on data reduction appears widely elsewhere in the literature.Here just two specific instruments are mentioned. As they take the form of personal samplers, they are of particular potential value for applications in the investigation of aerosol- related occupational lung disease and, indeed, are finding increasing use as such. The first is the sampler proposed by Rubow et al.25 shown in Fig. 9 (the so-called ‘Marple’ device). It is an eight-stage device, with radial slot-shaped jets at each stage where aerosol is collected on polycarbonate membrane films. By weighing the films before and after sampling, the mass of aerosol collected on each is assessed gravimetrically. The second device is the IOM personal inhalable dust spectrometer (PIDS) shown in Fig.10.26 The general configu- ration is similar to that for Rubow et al.’s*5 device. exceDt that the slot jets are replaced with circular jets. The aerosol is collected directly on the back of each disc-shaped aluminium impactor surface, which also incorporates the jets for the next stage. All the collection surfaces are greased prior to sampling and the masses of collected aerosol are obtained by weighing each disc before and after. The key feature of this instrument that distinguishes it from the previous type is that it incorporates a 15 mm circular entry similar to that for the IOM inhalable aerosol sampler, so it begins by aspirating the inhalable fraction. This entry is incorporated into a ‘cassette’, which, by also weighing before and after sampling, provides the mass of aerosol that is collected between the entry and the first impactor stage.Using this mass together with knowledge of the penetration characteristics of the entry stage, the particle size distribution obtained from the cascade impactor part of the instrument may be corrected to allow for Fig. 9 an actual workplace situation. Photograph courtesy of Andersen Fig. 10 IOM 2 1 min-1 personal inhalable dust spectrometer Instruments, Atlanta, GA. Marple 2 1 min-1 personal cascade impactor,25 shown in use in (PI D S)2625 Analyst, January 1994, Vol. 11 9 deposition (of both coarse and fine particles) in the entry, thus providing the particle size distribution of the whole inhalable fraction. Conclusion At present, limit values (e.g., in the ACGIH threshold limit value list) for fine aerosols are assigned only for the respirable aerosol fraction.For practical workplace measurement of this fraction, existing instrumentation appears adequate to meet most of the requirements of the latest criteria under most practical conditions. It is believed that, to match the compro- mise curve proposed by Soderholm,7 minor adjustments in sampling flow rate will achieve the desired purpose. Regard- ing assessment of sampler acceptability, however, there remain questions as to the best protocol.8 Few instruments are available specifically for the thoracic fraction. However, sufficient knowledge exists about the performance of pre-selectors such as those used for the respirable fraction to permit the design of adequate samplers for the thoracic fraction as soon as the demand is stimulated (i.e., when limit values for appropriate substances are expressed in terms of the thoracic fraction). Finally, the role of aerosol spectrometers as investigatory tools for industrial hygienists has been clearly identified. References Vincent, J. H., Analyst, 1994, 119, 13. Lippmann, M., in Handbook of Physiology; Section ZV, Environmental Physiology, eds. Lee, D. H. K., and Murphy, S., Williams and Wilkins, Philadelphia, 2nd edn., 1977, pp. Chan, T. L., and Lippmann, M., Am. Znd. Hyg. Assoc. J., 1980, 41, 399. Stahlhofen, W., Gebhart, J., and Heyder, J., Am. Znd. Hyg. Assoc. J . , 1980, 41, 385. International Standards Organization, Air Quality-Particle Size Fraction Definitions for Health-related Sampling, Technical Report ISO/TR/7708-1983 (E), ISO, Geneva, 1983.213-232. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 American Conference of Governmental Industrial Hygienists, Particle Size-selective Sampling in the Workplace, Report of the ACGIH Technical Committee on Air Sampling Procedures, ACGIH, Cincinnati, OH, 1985. Soderholm, S. C., Ann. Occup. Hyg., 1989,33, 301. Kenny, L. C., J. Aerosol Sci., 1992, 23,773. Asbestos International Association, Recommended Technical Method No. I : Reference Method for the Determination of Airborne Asbestos Fibre Concentrations at Workplaces by Light Microscopy (Membrane Filter Method) AIA, London, 1979. Vincent, J . H . , Aerosol Sampling: Science and Practice, Wiley , Chichester, 1989. Ford, V. H. W., Ph.D. Thesis, University of Newcastle-upon- Tyne, 1971. Ogden, T. L., Birkett, J. L., and Gibson, H., Improvements to Dust Measuring Techniques, IOM Report No. TM/77/11, Institute of Occupational Medicine, Edinburgh, 1977. Mark, D., Lyons, C. P., and Upton, S. L., Appl. Occup. Environ. Hyg., 1993, 8, 370. Courbon, P., Wrobel, R., and Fabries, J.-F., Ann. Occup. Hyg., 1988,32, 129. Prodi, V., Belosi, F., and Mularoni, A., J. Aerosol Sci., 1986, 17, 576. Cherrie, J. W., Jones, A. D., and Johnston, A. M., Am. Znd. Hyg. Assoc. J . , 1986,47, 465. Iles, P. J., and Johnston, A. M., Ann. Occup. Hyg., 1983, 27, 389. Johnston, A. M., Jones, A. D., and Vincent, J. H., Ann. Occup. Hyg., 1982,26,309. Walton, W. H., Br. J. Appl. Phys., 1954, 5 , Suppl., S29. Mark, D., Borzucki, G., Lynch, G., and Vincent, J. H., paper presented at the Annual Conference of the Aerosol Society, Bournemouth, 1988. Aitken, R. J., Vincent, J. H., and Mark, D., Appl. Occup. Environ. Hyg., 1993, 8, 363. Vincent, J. H., Mark, D., and Aitken, R. J., J. Aerosol Sci., 1993,24, 929. Mark, D., and Vincent, J. H., Ann. Occup. Hyg., 1986,30,89. Vincent, J. H., and Mark, D., Ann. Occup. Hyg., 1984,28,117. Rubow, K. L., Marple, V. A., Loin, J., and McCawley, M. A., Am. Znd. Hyg. Assoc. J . , 1987, 48, 532. Gibson, H., Mark, D., and Vincent, J. H., Ann. Occup. Hyg., 1987,31,463. Paper 3103201 I Received June 4, 1993 Accepted August 16, 1993