Radon Decay Product Aerosols in Ambient Air

The aerodynamic size distributions of radon decay product aerosols, i.e. Pb, Pb, and Pb were measured using low-pressure (LPI) as well as conventional low-volume 1-ACFM and high-volume (HVI) cascade impactors. The activity size distribution of Pb and Pb was largely associated with submicron aerosols in the accumulation mode (0.08 to 2.0 μm). The activity median aerodynamic diameter “AMAD” varied from 0.10 to 0.37 μm (average 0.16 μm) for Pb-aerosols and from 0.07 to 0.25 μm (average 0.12 μm) for Pb-aerosols. The geometric standard deviation, σg averaged 2.86 and 2.97, respectively. The AMAD of Pb-aerosols varied from 0.28 to 0.49 μm (average 0.37 μm) and the geometric standard deviation, σg varied from 1.6 to 2.1 (average 1.9). The activity size distribution of Pb-aerosols showed a small shift to larger particle sizes relative to Pb-aerosols. The larger median size of Pbaerosols was attributed to α-recoil depletion of smaller aerosol particles following the decay of the aerosol-associated Po. Subsequent Pb condensation on all aerosol particles effectively enriches larger-sized aerosols. Pb-212 does not undergo this recoil-driven redistribution. Even considering recoil following Po α-decay, the average Pb-labeled aerosol grows by a factor of two during its atmospheric lifetime. Early morning and afternoon measurements indicated that similar size associations of Pb occur, despite humidity differences and the potential for fresh particle production in the afternoon. In estimating lifetimes of radon decay product aerosols in ambient air, a mean residence time of about 8 days could be applied to aerosol particles in the lower atmosphere below precipitation cloud levels.


INTRODUCTION
The decay of radon, 222 Rn and thoron, 220 Rn in the atmosphere produces low vapor pressure decay products that coagulate with other nuclei or condense on existing aerosol particles.These decay products include 218 Po (3.0 min), 214 Pb (26.8 min) and 212 Pb (10.64 h).A long-lived decay product in the 222 Rn decay chain, 210 Pb (22.3 y) is produced about an hour after attachment of 218 Po onto aerosol particles.Very little data exist on the ambient aerosol associations of the decay products of both radon isotopes with respect to their aerodynamic size distributions relative to the formation and growth mechanisms of aerosols.
Fig. 1 presents the decay scheme for radon, 222 Rn and thoron, 220 Rn, illustrating the short-lived as well as the long-lived decay products, their mode of radioactive decay and their half-lives.
Diffusion batteries have been used to investigate aerosol size associations of gross -activity from the short-lived radon decay products (Soilleux, 1970;George and Breslin, 1980;George et al., 1984) in outdoor air situations, and high-volume cascade impactors for gross -and -activity from the long-lived radon and thoron decay products (Röbig et al., 1980).The residence time of atmospheric aerosol particles in the lower atmosphere assuming that the air in the troposphere is considered as a wellmixed reservoir (closed system) is a function of various removal processes, the most important being: (a) dry deposition by impaction, diffusion and sedimentation, and (b) wet deposition by rain drops (precipitation scavenging) as a result of processes occurring both within and below the rain cloud.The residence time of atmospheric aerosol particles can be estimated by means of radioactive nuclides as tracers, which become attached to aerosol particles and are removed with them as they are scavenged by precipitation or undergo dry fallout.Several methods have been used for estimating the mean residence time of atmospheric aerosol particles.These include measurements of the activities and ratios of radioactive decay products of radon, 222 Rn which emanate from continental surfaces into the atmosphere, such as 210 Pb (22.3 y), 210 Bi (5.01 d) and 210 Po (138.38 d) (Francis et al., 1970;Poet et al., 1972;Lambert et al., 1980;1983;Marley et al., 2000, Baskaran andShaw, 2001).However, there is disagreement between the derived values of the residence times due to various processes, including the fact that they refer to different portions of the atmosphere.
This paper summarizes results of an investigation designed to characterize the aerodynamic size distributions and the residence time of atmospheric aerosols in the context of 222 Rn-and 220 Rndecay products for better understanding of how radioactive aerosols behave in the atmosphere, and of the aerosol growth mechanism.The 1-ACFM design operated at 28 L/min (1 ft 3 /min).The eight stages had effective cutoff diameters, ECDs of 0.4, 0.7, 1.1, 2.1 3.3, 4.7, 7.0 and 11.0 m.The low-pressure modification, which alters the impactor's operation by increasing the resolution in the submicron region, involves a regulated flow rate of 3 L/min, five low-pressure (114 mm Hg) stages for the submicron region and eight atmospheric pressure stages for separating aerosol particles above 1.4 m.The ECDs of the low-pressure stages were 0.08, 0.11, 0.23, 0.52 and 0.90 m, whereas for the upper stages they were 1.4, 2.0, 3.3, 6.6, 10.5, 15.7, 21.7 and 35.0 m.A schematic diagram of a low-pressure cascade impactor, LPI complete system is shown in Fig. 2. The stainless steel plates supplied by the manufacturer were used for aerosol particle collection.Either polycarbonate or glass-fiber backup filters were used to collect all particles below the 0.08-m collection plate.The ECDs of the high-volume impactor stages were 0.41, 0.73, 1.4, 2.1, 4.2 and 10.2 m.The flow rate of this 20 cfm highvolume cascade impactor was about 34 m 3 /h.

Measurements
The length of each collection period varied from about 3 h for 214 Pb to 30 or 40 h for 212 Pb with low-pressure impactor measurements and from 1 to 24 h for 210 Pb and 210 Bi for highvolume impactor measurements.The samples were collected 13 m above the ground on the roof of the Environmental Sciences Division building at the Oak Ridge National Laboratory, ORNL, in Oak Ridge, Tennessee (35 o 58'N, 84 o 30'W).
The deposits on the stainless-steel collection plates of the lowpressure cascade impactor were leached with a solution of 1 M HNO 3 and the leachate rapidly evaporated on 5.08-cm stainless steel plates using a hot plate.The concentrations of 214 Pb and 212 Pb in the impactor plate leachates were measured using seven ZnS (Ag) alpha scintillation counters with backgrounds averaged 0.025 cpm.The 1 counting uncertainty was 5% or less for the stages corresponding to aerosols below 0.52 m and < 15% for the stages collecting to aerosols above 0.52 m.The calculated activities of 214 Pb and 212 Pb were used to determine the activity size distribution of each radionuclide and then the activity median aerodynamic diameter "AMAD" of the 214 Pb-and 212 Pbaerosols.The high-volume impactor plates were leached with 0.1 M HCl.Lead-210 was determined 30 days after collection stopped by separating and measuring 210 Bi (Poet et al., 1972).When 210 Pb was measured, the upper two high-volume impactor stages were coated with a thin layer of petroleum jelly to minimize soil particle bounce.While nuclepore polycarbonate membranes (0.4 m) were preferred as backup filters for lowpressure cascade impactors, glass-fiber filters had to be used in the high-volume impactor measurements.
The activity of 210 Pb-and 210 Bi-associated aerosol particles was measured by a low-background phoswitch scintillation detector system having a background of 2 cpm and efficiency higher than 40% for counting -radiation.This system consisted of a thin CaF 2 (Eu) primary crystal with a decay time of 0.23 s.The samples were counted for long enough to obtain a statistical accuracy better than 5%.It must be noted that 210 Pb and 210 Bi were chemically separated and measured for their activities.Detailed description of the analytical method and technique was established and presented elsewhere (Jaworowski, 1963).The activity of 210 Pb was also measured by a surface barrier Ge detector through its gamma-ray peak of 46.5 keV.The activity size distribution and the AMAD of the 210 Pb-aerosol particles were determined upon the measurement of the activities of the aerosol-associated radioactive nuclides.Fig. 3 shows a typical gamma-ray spectrum of a glass fiber air filter obtained by a Ge detector, in which the 46.5 keV -ray peak of 210 Pb is clearly shown.

Activity Size Distribution of Radon Decay Product Aerosols
Fig. 4 shows representative plots from 46 low-pressure impactor measurements made over a ten-month period, illustrating the aerodynamic size distributions of 214 Pb-and 212 Pbaerosols.R is the radioactivity and Dp is the particle diameter, in m.About 46% of the measurements showed a radioactivity peak in the 0.11-to 0.23-m region (subplot a), while 39% showed a peak in the 0.23-to 0.52-m region (subplot b).The remaining 15% of the measurements resulted in activity size distributions similar to those in subplot c (8.7%) or subplot d (6.5%),where 214 Pb-and 212 Pb-activities were highest in different size ranges in the same spectrum.On the average, about 76% of the 214 Pbactivity and 67% of the 212 Pb-activity was found to be associated with aerosol particles in the 0.08-to 1.4-m size range.The activity-associated with aerosol particles smaller than 0.08-m can also be substantial, as indicated in Fig. 4. Fig. 4 illustrates that 214 Pb is generally enriched in the accumulation mode aerosol, particularly between 0.11 and 0.52 m, where most of the surface area and mass occurs, while the Aitken nuclei fraction (below 0.08 m) contained a higher percentage of 212 Pb-activity compared with 214 Pb-activity in 69.6% of the measurements.
The radon decay products become associated with atmospheric aerosol particles by condensation or coagulation processes that are surface-area related.The activity size distribution of 214 Pbaerosols showed a small shift to larger particle sizes relative to 212 Pb-aerosols.The shift of 214 Pb to a slightly higher size distribution compared to 212 Pb was also found using 1-ACFM and high-volume cascade impactors (Fig. 5).The higher flow rates of these impactors, as well as the ability to measure highvolume activity by gamma-ray spectrometry, made us confident that this shift was real and not a data analysis artifact.The larger median size of 214 Pb-aerosols was attributed to -recoil depletion of smaller aerosol particles following the decay of the aerosolassociated 218 Po.Subsequent 214 Pb condensation on all particles effectively enriches larger-sized aerosols.Pb-212 does not undergo this recoil-driven redistribution.Even considering recoil following 214 Po -decay, the average 210 Pb-labelled aerosol grows by a factor of two during its atmospheric lifetime.Early morning and afternoon measurements indicated that similar size associations of 214 Pb occur, despite humidity differences and the potential for fresh particle production in the afternoon.
The longer half-life of 212 Pb (10.64 h) compared to 214 Pb (26.8 min) might favor the presence of larger aerosol associations of 212 Pb if coagulation rates are fast relative to radioactive decay rates, although cascade impactor measurements may not be sensitive enough to record this effect.Instead, the measurements indicated that the shorter-lived chain ( 218 Po, 214 Pb, etc) was more often associated with larger aerosol particle sizes than the longerlived chain ( 216 Po, 212 Pb, etc).The high-volume impactor measurements reported by Röbig et al. (1980) also indicated a large particle shift of 214 Pb relative to 212 Pb (Fig. 6).They stated that this was due to shifts of the activity size distributions from night to day-time to smaller particle sizes by photochemical reactions and to shifts to larger particle sizes when the relative humidity increases during rainfall, and therefore depends on sampling time.They also observed that the activity size distributions of the long-lived radionuclides 210 Pb and 7 Be are shifted to larger particle sizes compared with the activity size distributions of the shorter-lived 214 Pb and 212 Pb.That shift was due to larger residence times for 210 Pb and 7 Be in the atmosphere and for 7 Be due to a different size distribution of stratospheric and tropospheric aerosols.
The 214 Pb shift might be explained by the fact that a significant fraction of the 3.05-min 218 Po atoms, precursor of 214 Pb atoms should attach to an existing aerosol particle or coagulate with other nuclei during their lifetime, as mean attachment half-lives are of the order of a minute less (Porstendörfer and Mercer, 1980).When this attached 218 Po -decays, the recoiling 214 Pb atom, decay product of 218 Po, can escape the aerosol particles.Complete recoil loss would occur if the particle diameter, D p of the aerosol were smaller than the range of the recoiling nucleus.In water, this recoil range is 0.13 m (Mercer, 1976) and should be somewhat less in the atmospheric aerosols that have a density closer to 1.5 g cm -3 (Friedlander, 2000).By contrast, very little of the 0.146 s-216 Po would attach before decaying to 212 Pb, because of its short life relative to attachment times.A considerable fraction of the 214 Pb should undergo recoil detachment, particularly from aerosol particles with diameters smaller than 0.1 m (diameter approximating the recoil range).The probability of loss would decrease with increasing particle radius (Mercer, 1976).If the recondensing 214 Pb behaves like the original 214 Po, the net effect would be a shift of 214 Pb to a larger particle size distribution.
Since the subsequent -decays of 212 Pb and 214 Pb + 214 Bi do not result in significant recoil (Mercer, 1976), the -measurement of  The AMADs of 214 Pb-aerosols determined with the lowpressure impactor data varied from 0.10 to 0.37 m (average 0.16 m) and for 212 Pb-aerosols from 0.07 to 0.25 m (average 0.12 m) (Table 1).For the 210 Pb-aerosols the AMADs varied from 0.28 to 0.49 m (average 0.37 m).Lead-210 is produced bydecay of 214 Po, the event used to quantify 214 Pb distributions on the low-pressure cascade impactors.While the relationship between the aerodynamic sizes of 214 Pb and 210 Pb is complicated because of the large differences in their atmospheric lifetimes, Fig. 6.The activity size distributions of the atmospheric aerosols in ambient air (Robig et al., 1980). 214Pb, as indicated by the differences in AMADs reported in Table 1.Table 1 shows data for the AMADs of the radon decay product aerosols.The results were obtained assuming lognormal distributions.The reported from the literature data for radon decay product aerosols are also presented in Table 1 and analyzed as follows: Grundel and Porstendörfer ( 2004) more recently carried out measurements in outdoor air at a northern latitude (51 o 32'N, 9 o 55'E) in Göttingen, Germany using an on-line alpha cascade impactor (OLACI) for a period of about one month and showed that the activity size distributions of the short-lived radon ( 222 Rn) and thoron ( 220 Rn) decay products are about 12-19 % in the nucleation mode (Aitken nuclei, 0.05-60 nm), 81-88% in the accumulation size range (60-1000 nm), and with no coarse mode (particle diameter > 1000 nm).The AMAD of the accumulation mode varied between 332 nm for 218 Po and 347 nm for 214 Po for the short-lived radon decay products and between 382 nm for 212 Po and 421 nm for 212 Pb for the thoron decay products.In comparison to the short-lived radon decay products, the AMADs of the thoron decay products were shifted significantly to larger values.For the long-lived radon ( 222 Rn) decay products 210 Pb and 210 Po they showed that these radionuclides are almost 93-96% adsorbed on aerosol particles in the accumulation size range and only 4-7% of their activities are attached on nuclei with diameters smaller than 60 nm.AMAD-values of 558 nm for 210 Pb and 545 nm for 210 Po were measured, i.e. significantly larger than those of the short-lived radon and thoron decay products.
In El-Minia, Egypt (28 o 04'N, 30 o 45'E), El Hussein and Ahmed (1995) (1999) by using a low-pressure Berner-type cascade impactor in El-Minia, Egypt reported that the AMAD for the very short-lived radon decay product 218 Po-aerosols varied from 280 to 386 nm (average 340 nm) and the geometric standard deviation, g, varied from 2.5 to 2.9 (average 2.7), while for its decay product 214 Pb-aerosols the AMAD varied from 296 to 360 nm (average 320 nm) and the g varied from 2.1 to 2.88 (average 2.7), which means that they are quite similar.For the relatively longer-lived thoron decay product 212 Pb-aerosols, the AMAD of the accumulation mode varied from 249 to 390 nm (average 360 nm) and the g varied from 2.1 to 3.2 (average 2.7) (Mohammed et al., 2000).Winkler et al. (1998) by using a nine-stage low-pressure Berner-type cascade impactor in Neuherberg, Germany (48 o 13'N, 11 o 36'E) reported that the activity size distribution of 210 Pb in ambient aerosols was unimodal (log-normal) and associated with submicron aerosols of about 0.5 to 0.6 m.On average, the AMAD of 210 Pb-aerosols (0.53 m) has been found to be significantly lower than the average mass median aerodynamic diameter "MMAD" (0.675 m), and higher than or at most equal to the respective surface median aerodynamic diameter "SMAD" (0.465 m) of the aerosols, i.e.SMAD < AMAD < MMAD.Variation of the atmospheric processes resulted in a variability of the AMAD, between 0.28 and 0.74 m for 210 Pb-aerosols.While in the winter period (October to April) the AMAD values of 210 Pb-aerosols averaged 0.595 m, in the summer period 210 Pb was associated with significantly smaller-sized aerosols (AMAD: 0.43 m).Suzuki et al. (1999) by using a 40-cfm high-volume cascade impactor with five stages reported that 77% of 210 Pb-and 70% of 210 Po-activities were measured in size-fractionated aerosols with a diameter smaller than 0.70 m from the west coast of the Japanese Islands (38 o 46'N, 139 o 44'E).
Apart of this, Papastefanou (2008) estimated the AMAD of another radionuclide in ambient aerosols, that is 7 Be (T ½ = 53.3days) of cosmogenic origin, varying between 0.76 and 1.18 m (average 0.90 m) in the accumulation mode of the activity size distribution of aerosol particles.The aerodynamic size  et al. (1999) distribution of 7 Be in atmospheric aerosols and then the AMAD was achieved by using Andersen 1-ACFM cascade impactors, in association with a gamma-ray spectrometer linked with a Ge coaxial-type detector for the 7 Be activity measurement of the collection plates and the back up filter of the impactor.

Residence Times of Radon Decay Product Aerosols
The method for estimating the residence time of atmospheric aerosol particles associated with the radon decay product radionuclides is based on the radioactivity of a pair of genetically related radioisotopes, such as 210 Pb, 210 Bi and or 210 Pb, 210 Po, according to the sequential disintegrations in the -decay scheme The residence time, R is described by the formula where Bi N Bi is the activity of 210 Bi, Pb N Pb is the activity of 210 Pb in air and Bi = 0.138 d -1 is the decay constant of 210 Bi.
The Eq. ( 1) was derived from the equation of the production and removal of radionuclides assuming asteady state equilibrium where R = 1/ R is the first-order rate constant for the removal of aerosol particles from the atmosphere by all processes, that is the inverse of residence time, R .
The ratio of the activities Bi N Bi / Pb N Pb in Eq. ( 1) varied from 0.40 to 0.68 (this work) or from 0.42 to 0.85 (Moore et al., 1972).
If the activity of 210 Po, Po N Po in air is considered and Po = 5.0 × 10 -3 /d is the decay constant of 210 Po, then the activity ratio of 210 Po, Po N Po and 210 Pb, Pb N Pb is given by the equation where The ratio of the activities Po N Po / Pb N Pb in Eq. (3) varied from 0.054 to 0.092 (Moore et al., 1972).
The residence time, R can also be determined through the ratio of the activities of radon, 222 Rn, Rn N Rn and 210 Pb, Pb N Pb in air according to the sequential disintegrations in the -and -decay scheme as where Pb = 8.5 × 10 -5 /d is the decay constant of 210 Pb and Rn N Rn / Pb N Pb >> 1.
The ratio of the activities Rn N Rn / Pb N Pb in Eq. 6 varied from 282 to 7700 (Moore et al., 1972).
Taking into account the 210 Bi/ 210 Pb ratios as determined in twenty one measurements of aerosol samplings carried out during an annual period at Oak Ridge, Tennessee at temperate latitude (35 o 58'N, 84 o 17'W) with high precipitation (wet climate), the estimated residence time of atmospheric aerosols in the boundary layer varied from 4.8 to 15.3 days (average 8.2 days).
Table 2 shows data for the residence times, R of atmospheric aerosol particles associated with radon, 222 Rn and radon decay products 210 Pb, 210 Bi and 210 Po.The reported from the literature R values are also shown in Table 2 and analyzed as follows: Poet et al. (1972) estimated tropospheric aerosol residence times ranging from 1.59 to 13 days (average 5.4 days) when based on the 210 Bi/ 210 Pb activity ratios and from 11 to 77 days (average 24 days) when based on the 210 Po/ 210 Pb activity ratios for twenty measurements of aerosol samplings carried out in surface air during a 4 ½-year period at Boulder, Colorado (40 o 01'N, 105 o 17'W).They concluded that, a mean tropospheric residence time of about 4 days could be applied for aerosol particles in the lower troposphere and about a week for aerosol particles in precipitation.They also found that the mean aerosol residence time increases with altitude within troposphere less than by a factor of 3 (Moore et al., 1973).Francis et al. (1970) estimated a mean atmospheric residence time for 210 Pb of 9.6 days ± 20% based on the 210 Po/ 210 Pb activity ratios from the dust of filtrate (> 0.22 m) collected during a 3month period (May-August) at Madison, Wiskonsin (43 o 05'N, 89 o 22'W).Marley et al. (2000) estimated residence times for seven aerosol samples collected at Argonne, Illinois (41.7 o N, 88.0 o W) during a 2-year period, three aerosol samples collected at Phoenix, Arizona, one sample collected at Socorro, New Mexico, and five samples collected at Mexico city, Mexico.Based on the 210 Bi/ 210 Pb activity ratios, they resulted in residence times of 6 to 67 days, while based on the 210 Po/ 210 Pb activity ratios, they suggested residence times of 33 to 66 days for the aerosols below 2 m in size.
Much earlier, Fry and Menon (1962) in twelve measurements carried out during a 7-month period (spring and summer) at Fayetteville, Arkansas (36 o 03'N, 78 o 54"W), indicated apparent aerosol residence times ranging from 2.4 to 25.6 days (average 8.5 days), based on the 210 Bi/ 210 Pb activity ratios in Arkansas rains.Gavini et al. (1974) also estimated residence times varied from 3 to 240 days (average 29 days) from the 210 Bi/ 210 Pb activity ratios in Arkansas rains also.Baskaran and Shaw (2001) based on the 210 Po/ 210 Pb activity ratios estimated residence times of arctic haze aerosols, from the upper atmosphere to the arctic atmosphere, varied between 12 and 32 days from the measurements carried out during an 1 1/2year period (winter) at Poker Flat (65.1 o N, 147.5 o W) and between 0 and 39 days from eight measurements carried out during a 2month period (winter) at Eagle (69.5 o N, 141.2 o W) in Alaska.
Very early, Lehman and Sittkus (1959) estimated aerosol residence times of 20 days from the 210 Po/ 210 Pb activity ratios in air at Freiburg, Germany (47 o 59'N, 7 o 51'E), while Peirson et al. (1966) estimated aerosol residence times of 40 days from the 210 Po/ 210 Pb activity ratios in air at Milford Haven, Wales (51 o 40'N, 5 o 02'E).Lambert et al. (1980)  residence times varied from 8.8 to 10.5 days based on the 210 Bi/ 210 Pb activity ratios in air, while later in another work they estimated aerosol residence times varied from 7 to 9 days (average 8.58 days) based on the 210 Bi/ 210 Pb activity ratios in air, in near Paris, France area during a 6-year period (Lambert et al., 1983).Gäggeler et al. (1995) during an one-year period carried out measurements on the activities of 214 Pb, 210 Pb and 212 Pb in air at high altitude in Jungfraujoch, Switzerland (46 o 32'N, 07 o 59'E) as high as 3450 m by using an epiphaniometer.They estimated aerosol residence times varied between 1 and 12 days (average 6 days) from the 214 Pb/ 210 Pb activity ratios.
Finally, Rangarajan (1992) based in the 210 Bi/ 210 Pb activity ratios in air estimated a mean residence time of the natural radioactive aerosols in the planetary boundary layer at Bombay (Mumbai), India (18 o 58'N, 72 o 50'E) around 8 days from forty three measurements carried out during the long dry period, from October to May.
The data of Table 2 admit residence times of lower values as low as 1.59 days and higher values as high as 320 days.Mostly, the lower R -values were resulted from the 210 Bi/ 210 Pb activity ratios and higher R -values from the 210 Po/ 210 Pb activity ratios.Low R -values were also resulted from the 222 Rn/ 210 Pb activity ratios.Poet et al. (1972) concluded that longer apparent residence time values were based on the 210 Po/ 210 Pb activity ratios, than those based on the 210 Bi/ 210 Pb activity ratios.The difference may be explained by the presence of a mixture of aerosols of various apparent ages, of which older aerosols (those of greater ages) contribute most of 210 Po.On the other hand, the solid products of radon-decay, that is 210 Pb, 210 Bi and 210 Po might arise from sources other than radioactive decay within the atmosphere.Poet et al. (1972) showed that up to 85% of the 210 Po in the atmosphere is of terrestrial origin, and the vertical profile of 210 Po was found to differ appreciably from that expected from the decay of 222 Rn.Lambert et al. (1979) indicated that volcanic gases are very rich in long-lived radon decay products, especially in 210 Po relative to 210 Pb.Soil particles are the most likely contributors, since a part of the tropospheric aerosols originates at the Earth's surface.Coal burning and forests fires presumably are additional sources of radionuclides.
Apart of this, Papastefanou (2008) estimated the residence times of another radionuclide in ambient aerosols, that is 7 Be (T ½ = 53.3d) of cosmogenic origin varying between 7.4 and 8.9 days (average 8 days) at Thessaloniki, Greece (40 o 38'N, 22 o 58'E) in the category of the lower values of the residence times of atmospheric aerosols and leading to the same result for the residence times of radon decay product aerosols in ambient air varying between 4.8 and 15.3 days (average 8.2 days).The method used for the determination of the residence time of 7 Be aerosols was described elsewhere (Papastefanou, 2008).

CONCLUSIONS
The aerodynamic size distributions of radon decay product aerosols in ambient air were measured using low-pressure (LPI) and conventional (1-ACFM) and/or high volume (HVI) cascade impactors.Lead-214 and 212 Pb were largely associated with aerosol particles in the accumulation mode for smaller than 0.52 m diameter particles in size.Based on the 46 % low-pressure impactor measurements, the mean activity median aerodynamic diameter "AMAD" of 214 Pb-aerosols was found to be 0.16 m, while for the 212 Pb-aerosols, the mean AMAD was found to be 0.12 m.The slightly larger size of 214 Pb-aerosol particles confirmed with operationally different cascade impactors, was attributed to -recoil-driven redistribution of 214 Pb following the decay of the aerosol-associated 218 Po.Even considering recoil following 214 Po -decay, the average 210 Pb-labelled aerosols grow by a factor of two during its atmospheric lifetime.Estimated lifetimes of radon decay product aerosols in ambient air resulted in a mean residence time of about 8 days that could be applied to aerosol particles in the lower atmosphere below the boundary layer.
on aerodynamic sizes of atmospheric aerosols associated with radioactive nuclides, such as radon decay products 214 Pb, 212 Pb and 210 Pb were carried with Anderson type 1-ACFM ambient cascade impactors with or without the Andersen model 2000 low-pressure modification, as well as with Sierra model 236, six-stage high-volume cascade impactors, HVI.

Fig. 2 .
Fig. 2. A schematic diagram of a low-pressure cascade impactor complete system.

Fig. 3 .
Fig. 3. Plot of a -ray spectrum of an atmospheric aerosol sample (air filter) obtained by a Ge detector.

Fig. 4 .
Fig. 4. Aerodynamic size distributions of 214 Pb-and 212 Pb-aerosols obtained with low-pressure cascade impactors.R is the radioactivity, and Dp is the particle diameter.(a) Results occurred 46% of the time, (b) 39% of the time, (c) 8.7% of the time, and (d) 6.5% of the time.Lower Dp limits are arbitrary.

Fig. 5 .
Fig. 5. Aerodynamic size distributions of 214 Pb-and 212 Pb-aerosols obtained with 1-ACFM and high-volume cascade impactors, illustrating the large particle shift of 214 Pb.R is the radioactivity, and D p is the particle diameter.Lower D p limits are arbitrary.
by using a low-pressure Berner-type cascade impactor showed that the activity size distributions of the 214 Pband 214 Bi-attached aerosols were nearly identical and that most of the activities were associated with aerosol particles of the accumulation mode.The mean AMAD of 214 Pb-aerosols (range

Table 2 .
estimated aerosol Residence times, R of tropospheric aerosols (days).