Regional deposited dose in the human respiratory tract using different particulate metrics

The objective of the current study was to calculate the deposited dose rate in the human respiratory tract arising from particle number (PN) and particle mass (PM) measurements. A main objective was the investigation of deposition pattern and characteristics of the two metrics in the human respiratory tract. The dose rate was estimated for residents at a suburban background location (Chania, Greece). The total dose rate showed two peaks, one in the morning (1.6 × 10 9 particles/h at 7:00-8:00) and the other one at night (2.1 × 10 9 particles/h at 21:00-22:00), during the warm period, while the cold period showed two peaks, morning (2.0 × 10 9 particles/h at 9:00-10:00) and afternoon (3.6 × 10 9 particles/h at 18:00-19:00). The peaks during the warm period were associated with traffic emissions, whereas the peaks during the cold period were associated with both heating and traffic emissions. A higher dose rate of PN 10 was found in the alveolar region while for PM 10 it was found in the extrathoracic region. These findings are linked with increased contribution of ultrafine and coarse particles to PN 10 (65%-78% and 54%-62% for cold and warm periods, respectively) and PM 10 (63% and 55% for cold and warm periods, respectively) concentrations, respectively. The current study showed the importance to use both number and mass aerosol metrics for determining the human exposure and regional dose and their related health effects. The novelty of the current study is the simultaneous measurements of the two particles metrics and the full particle size distributions to make accurate estimates of regional deposited dose.


INTRODUCTION
Human exposure to particle number (PN) and particle mass (PM) concentrations has been studied extensively in the scientific literature [1][2][3][4][5][6][7][8] . Investigation of population exposure, especially in urban areas, is important because of the associated health hazard that comes to the residents. Inhalation of airborne particles and the corresponding deposition in the human respiratory tract are directly linked with their concentration. Ambient levels of airborne particles depend on several parameters such as the geographical position, climatological characteristics, urbanization, population density, and anthropogenic activities [7,[9][10][11][12] , as well as different characteristics that accompany the particle size distributions (number vs. mass).
Airborne particles can be classified into different size classes (ultrafine, fine, and coarse). Accordingly, fine particles are particles with a diameter less than 2.5 μm, while coarse particles are considered those having diameters ranging from 2.5 to 10 μm. Ultrafine particles are a subset of fine particles, defined as those particles with a diameter less than 100 nm. The main sources of ultrafine and -generally -fine particles are combustion processes such as traffic exhaust, biomass burning, industry, energy production, and aviation [3,13] , whereas coarse particles usually originate from soil resuspension, sea spray/salt, and Saharan dust [3,14] . In principle, ultrafine particles are the dominant contributor to the particle number size distribution, whereas the opposite is found when evaluating the particle mass size distribution [2,3,13,15] . The latter is dominated by larger particles due to the significant mass input of coarser particles in the size distribution.
Estimation of the deposited dose of nanoparticles is important due to their ability to penetrate and deposit in the alveolar region, causing significant human health consequences. The increased concern arises for ultrafine particles (< 100 nm) because they are toxic and can cause several adverse health effects on the respiratory and cardiovascular systems such as ischemic heart diseases, oxidative stress, cancer, systemic/pulmonary inflammation, and arrhythmia events [16][17][18] . Moreover, ultrafine particles are therefore retained longer in the lungs and cause more serious pulmonary inflammation compared to larger particles [19] . Particles in the size range 0.1-1.0 μm are exhaled and, hence, have the lowest total deposition in the human respiratory tract [20] . Coarse particles are responsible for the majority of inflammatory responses to PM 10 [21] .
The regional deposited dose in the human respiratory tract incorporating particle number size distribution data is completely different than that incorporating particle mass size distribution data as a direct consequence of the different inherent characteristics of the two size distributions. Recently, Manigrasso et al. found that the highest deposited dose was observed in the extrathoracic (ET) region for PM 10 while for PN 10 the highest deposited dose was found in the alveolar-interstitial region [5] . Additionally, Vu et al. calculated the dose of traffic-generated particles (13-560 nm) and found that the dose rate of outdoor particles at a traffic site (Bologna, Italy) was 10 10 particles/h with the majority of particles deposited in the AI region [8] . Regarding PM, Chalvatzaki et al. found that the deposited mass dose was higher in the ET region, and 75% of the deposited mass dose in the ET region was associated with coarse particles [22] . In addition, Manojkumar and Srimuruganandam found that the deposition of PM 10 at an outdoor traffic site was higher than that at an outdoor residential site [23] . Therefore, particle deposition in the human respiratory tract is associated with the type of site. A similar finding was observed by Chalvatzaki et al., where the deposited doses of PM 10 for six major Greek cities were estimated, and they were found to be higher at all urban sites (traffic and urban) compared to suburban background sites [24] . Higher health risks were also obtained for residents in the urban areas. In addition, Gini et al. calculated the dose during Saharan dust outbreaks in a suburban site in Athens (Greece) and found that the deposited mass dose was higher (80% for ET region and 27% for lungs) during Saharan dust days compared to non-Saharan dust days [25] . Saharan dust is an important natural source of coarse airborne particles that causes increased deposited dose in the human respiratory tract.
Estimation of the regional deposited dose rate of airborne particles in the human respiratory tract is an important step for understanding the health effects arising from the inhaled particles. In turn, health effects depend on particle deposition within the human respiratory tract. In the current study, the deposited dose rate was calculated for two particles metrics (mass and number). The simulations were performed on an adult male in a suburban background location. In addition, the study focused on estimating the effect of different characteristics (age, gender, and activity level) of the exposed subject on the deposited dose rate.

Study area and field measurements
Field measurements were performed at the Akrotiri monitoring station, which is located within the campus of the Technical University of Crete (Chania, Greece). The station is approximately 5 km from the city center of Chania with the surrounding area characterized by medium traffic and consists of a mixed area with agricultural land/farms and low-density residential areas [12,26] . Therefore, the study area is considered a suburban background site with a significant input from the marine environment due to its proximity to the coastline (2 km).
Α Scanning Mobility Particle Sizer (SMPS 3938, TSI) consisting of an Electrostatic Classifier 3082 (TSI), a long Differential Mobility Analyzer 3081 (TSI), and a Condensation Particle Counter 3775 (TSI), was used for measuring the particle number size distribution of fine particles. The instrument was operated with aerosol sample flow rate at 0.3 lpm and sheath flow rate at 3 lpm, with the midpoint particle mobility diameter ranging from 13.1 to 661.2 nm (110 size channels). In addition, an Optical Particle Sizer 3330 (OPS, TSI) was used for measuring the particle number size distribution from 0.3 to 10 µm in 16 discrete size channels.
Measurement of the particle mass concentration was performed through a portable aerosol particulate monitor (Dust Trak 8530 EP, TSI) using a PM 10 pre-head, while an Andersen Cascade Impactor (ACI) was used for obtaining the particle mass size distribution. ACI collected size-segregated particle mass fractions into nine size intervals: four size fractions for fine particles (PM 2.1 ) and five size fractions for coarse particles (PM 2.1-10 ). Two sampling campaigns were performed: During the first (Campaign I), the particle number size distribution of fine particles was measured throughout a warm period and a cold period. During the second (Campaign II), a full size distribution was obtained by simultaneous measurements of the particle number size distribution (PNSD) with the SMPS and the OPS [Supplementary Figure 1]. In addition to the PNSD measurements, the particle mass concentration and size distribution were measured again throughout a warm period and a cold period. A detailed description of the sampling periods of each campaign is shown in Table 1. In total, the sampling period corresponds to 57 days (32 days for the warm period and 24 days for the cold period) during Campaign I, while, during Campaign II, field measurements were performed for 27 days (13 for the cold period and 14 for the warm period). Field measurements of the particle mass size distribution were performed for six days (four samplings during the cold period and two samplings during the warm period, Supplementary Figure 2).

Dosimetry model
Exposure Dose Model (ExDoM2) is a dosimetry model developed to incorporate ambient particulate concentrations in order to determine the deposited dose rate of the inhaled particles [27] . Simulations can be conducted using freely the particle mass or number measurements depending on the available dataset. In the current study, simulations were performed considering a 24 h exposure in the outdoor environment by assuming a light exercise activity level for an adult male resident.
The deposition module of ExDoM2 is based on the International Commission on Radiological Protection (ICRP) model; therefore, it treats the human respiratory tract as a series of filters [28,29] . Accordingly, the human respiratory tract is divided into nine filters distributed as follows: two in the anterior nose region (ET1); two in the posterior nasal passages, pharynx, and larynx region (ET2); two in the bronchial (BB) region; two in the bronchiolar (bb) region; and one in the alveolar-interstitial (AI) region. Therefore, the deposition fraction can be sorted within five regions (ET1, ET2, BB, bb, and AI) and estimated as [28] : where DE 1 , DE 2 ,DE 3 , and DE 4 are the deposition fractions in the ET1, ET2, BB, and bb filters (inhalation), respectively; DE 5 is the deposition fraction in the AI filter (inhalation and exhalation); and DE 6 , DE 7 , DE 8 , and DE 9 are the deposition fractions in the bb, BB, ET2, and ET1 filters (exhalation), respectively. The deposition fraction in each filter depends on the fraction of tidal air that reaches the j filter (φ j ), the prefiltration efficiency (n 0 ), and the deposition efficiency of each filter (n j ).
The prefiltration efficiency (n 0 ) was estimated as [28] : where n I is the inhalability of particles.
The deposition efficiency (n j ) was calculated by [28] : where n ae is the aerodynamic deposition efficiency due to impaction and gravitational settling and n th is the thermodynamic deposition efficiency due to diffusion.
The φ j during inhalation was calculated by [28] : where V T is the tidal volume (mL), v jj is the volume of a preceding filter (mL), and N is the number of filters.
The φ j during exhalation was calculated by [28] : The mobility diameter was converted into the aerodynamic diameter using the following equation [5] : where d m is the mobility diameter, χ is the shape factor of particles, ρ is the density of particles, C(d m ) is the Cunningham slip correction for a particle of mobility diameter, and C(d ae ) is the Cunningham slip correction factor for a particle of aerodynamic diameter.
In the present work, the results were classified into three regions: the extrathoracic, tracheobronchial (TB), and alveolar-interstitial regions. The extrathoracic region includes ET1 and ET2, while the TB region includes BB and bb. Finally, a redistribution of 65% to ET1 and 35% to ET2 regions for individual depositions was incorporated, as proposed by ICRP [29] .

Clearance
According to ICRP [29] , the respiratory tract clearance model consists of 13 compartments. The retained dose of particles in each compartment j and the dose to the esophagus, lymph nodes, and blood was estimated by Equation (12), while Equation (13) was used for rapidly and slowly dissolving particles [22] : where m k,j is the mechanical movement rate of particles from compartment k to j; m j,k is the mechanical movement rate of particles from compartment j to k; f r is the fraction of particles dissolved rapidly; s r is the rapid dissolution rate; s s is the slow dissolution rate, H j is the deposited dose in the compartment j; I j and I k are the retained doses of particles dissolving relatively rapidly in the compartment j and k, respectively; and T j and T k are the retained doses of particles dissolving slowly in the compartment j and k, respectively. The absorption of particles in the blood was assumed to be moderate. The default values for f r , s r , and s s during moderate absorption were based on ICRP recommendations [29] . Therefore, f r , s r , and s s were set equal to 0.2, 3, and 0.005 d −1 , respectively [29] .

Diurnal and seasonal variation of deposited number dose rate
The diurnal variations of the deposited dose rate at the Akrotiri station during the warm and cold periods are presented in Figure 1A. The diurnal profile was characterized by two peaks, one in the morning (7:00-8:00) and one at night (21:00-22:00), during the warm period. Specifically, the dose rate was equal to 1.6 × 10 9 particles/h during the morning peak and 2.1 × 10 9 particles/h during the night peak. These peaks were associated with traffic and increased vehicular emissions during rush hours. Overall, higher values were obtained during traffic rush hours at night, which is linked with the closing time (21:00) of the commercial stores. It is well known in the scientific literature [8,30,31] that vehicular emissions increase the number concentration of ultrafine particles as well as the corresponding deposited dose. Regarding the cold period, the dose rate in the morning increased after 7:00 with a peak at 9:00-10:00 (2.0 × 10 9 particles/h), while the second peak was observed at 18:00-19:00 (3.6 × 10 9 particles/h). It was observed that the afternoon/night peaks were considerably higher than the morning peaks. These peaks were influenced by domestic heating emissions; therefore, increased dose rates were obtained. The opposite characteristic was observed by Vu et al., where the diurnal variation of the number dose rate in Bologna (Italy) showed two peaks with the afternoon peak (16 × 10 9 particles/h) being lower than the morning peak (22 × 10 9 particles/h) [8] . According to Aalto et al., the maximum particle number concentrations (between 7 nm and a few micrometers) in five European cities (Augsburg, Barcelona, Helsinki, Rome, and Stockholm) were detected in the morning hours during weekdays, while the opposite characteristic was observed on Sunday, indicating that traffic emissions influence the diurnal behavior [32] .
Furthermore, the seasonal comparison showed that the deposited dose during the cold period preserved higher rates compared to the warm period. A similar observation was observed by Voliotis and Samara [33] with cold/warm ratios being equal to 1.2 and 1.6 at the urban traffic and urban background sites in Thessaloniki (Greece), respectively. Domestic heating emissions during the cold period increased the particle number concentrations and the corresponding deposited dose rates. Specifically, the total particle number concentration and dose rate reached 4047 particles/cm 3 and 2.1 × 10 9 particles/h (21:00-20:00) during the warm period, while the corresponding values during the cold period were 6553 particles/cm 3 and 3.6 × 10 9 particles/h (18:00-19:00), respectively. It was also characteristic that the deposited dose rate followed similar temporal variations with particle concentrations [Supplementary Figure 3]. In the study of Voliotis and Samara [33] , the peak dose rate during the cold period was observed at night (21:00-22:00) with a rate of 4 × 10 10 particles/h in the urban background site. Intensive wood burning for residential heating purposes was the reason for the high rates, whereas the urban traffic diurnal variations followed traffic flow (9:00, 12:00, and 22:00) with the deposited dose rate during morning peak reaching values close to 7 × 10 10 particles/h. Urban anthropogenic activities (traffic emissions and residential heating) are important sources which cause increased deposited dose rates.
The cumulative deposited dose for the warm and cold periods is shown in Figure 1B. A higher cumulative deposited dose was obtained during the cold period (4.6 × 10 10 particles/day) in comparison with the warm period (3.4 × 10 10 particles/day) as a direct consequence of increased concentration due to the heating emissions. In addition, it was observed that in the early morning hours (1:00-8:00, Figure 1B) the cumulative dose for the two periods was almost equal, as a direct consequence of the similar dose rate obtained for this part of the day [ Figure 1A]. Interestingly, the results of this campaign indicate that hourly fine particle number concentrations fluctuate at similar levels after midnight (cold vs. warm periods, Supplementary Figure 3), an observation that is influenced by several factors such as the intensity of local sources, temporal variations, and ambient conditions. The daily doses in the current study were close to the results reported by Hussein et al., where the daily deposited dose obtained for submicron particles for an adult male living in Helsinki was found equal to 5.7 × 10 10 and 4.0 × 10 10 particles/day on workdays and weekends, respectively [34] . In addition, they pointed out that the daily deposited dose would be higher in real-life condition, when indoor sources/activities are also considered.
The deposition fraction in the three regions of the human respiratory for both periods is given in Table 2. A higher deposition fraction was obtained for both cases in the AI region with the deposition fraction being slightly higher during the cold period in comparison with the warm period. Specifically, the deposition in the AI region was equal to 21.7% and 24.2% during the warm and cold periods, respectively. Besides the fact that the deposition fraction did not show important differences between the two periods, the deposited dose rate and the daily deposited dose during the cold period were higher than in the warm period as a direct consequence of higher PN concentrations. Therefore, seasonal variations showed only a slight influence on the deposition fraction but had a large effect on the dose rate and daily dose due to differences in PN concentrations. Regarding the other two regions (ET and TB), the deposition was considerably lower (5%-6%). These findings are in agreement with those of Vu et al. who found that the majority of particles (number) were deposited in the AI region, with the deposition fraction in the AI region being equal to 0.35 when walking in an outdoor traffic area in Italy [8] . The corresponding values for TB and AI were 0.09 and 0.07, respectively. Finally, a comparison between the two studies revealed that the deposition fraction in AI at the traffic site in Italy [8] was greater than that at this site (suburban background). Therefore, exposure to urban sites with increased traffic results in an increased deposition fraction in the AI region. The increased deposition fraction in AI is associated with health effects. Specifically, deposited particles in AI irritate and corrode the alveolar wall and, hence, cause impaired lung function [35] .

Deposition fraction size distribution characteristics
Simulations revealed that the deposition fraction was higher in the AI region for particles in the size range 13-322 nm, while for particles in the size range 334-661 nm the deposition fraction was higher in the ET region [ Figure 2]. This finding is linked with the deposition mechanism, especially diffusion, which is the main physical mechanism of particle deposition in AI and its characteristic to increase with decreasing particle size [29,36] . Previous studies have reported that the effect of diffusion is dominant for particles with diameters less than 300-500 nm.
The present results are in agreement with the results reported in previous studies. Particularly, Vu et al. found that while walking or cycling the deposition fraction was higher in the AI region for particles with a mobility diameter range of 10-300 nm, while the deposition fraction was higher in the ET region for particles in the 350-560 nm size range [8] . Chalvatzaki and Lazaridis [37] found that the deposition fraction of NaCl without considering the hygroscopicity effect was higher in the AI region for particles with aerodynamic diameters ranging from 15-550 nm, while taking into consideration hygroscopicity the deposition in the AI region was higher for particles in the size range 30-200 nm. This change occurs due to hygroscopic particle growth inside the human respiratory tract.
In addition, it was observed that the highest total deposition fraction occurred for nanoparticles in the lower part of the size distribution (13.1 nm). Patterson et al. asserted that the higher total deposition fraction in conjunction with the smaller size causes higher health risk due to the increased possibility to penetrate to the deep lung with more mutagens [38] .

Impact of activity level and exposed subject to the dose rate
The impact of the activity level (sitting, light exercise, and heavy exercise) of the exposed subject (adult male) on the deposited dose rate (particles/h) obtained from particle number size distribution data during the cold period is presented in Table 3. It was observed that the deposited dose rate increased with increasing the activity level with the median deposited dose rate increasing from 6.7 × 10 8 (sitting at rest) to 35.4 × 10 8 particles/h (heavy exercise). This finding is attributed to the higher inhalation rate [Supplementary Table 1]; the inhalation rates adopted from ICRP [28] for an adult male for heavy exercise, light exercise, and sitting were 3.5, 1.5, and 0.54 m 3 /h, respectively. The same finding was observed by Daigle et al. where the deposited dose of particles (number) in the human respiratory tract was 4.5 times higher  during exercise than at rest [39] .
To examine the effect of age and gender on the deposited dose rate, the same activity level (light exercise) for all exposed subjects was adopted. Table 3 shows that the median deposited dose rate in the human respiratory tract was higher in adult males (16.7 × 10 8 particles/h) than in adult females (14.1 × 10 8 particles/h). Likewise, the deposited dose rate in the human respiratory tract was higher in adult males than in children. These findings are attributed to the higher inhalation rate of an adult male in comparison with an adult female or children. Hussein et al. also showed that the deposited dose rate when walking (4 km/h) for sub-micron particles (PN 1 ) was higher for an adult male (77 × 10 8 particles/h) than an adult female (67 × 10 8 particles/h) [40] . In addition, the deposited dose rate was lower in a 5-10-year-old child than in an adult female [ Table 2]. However, it is misleading to conclude that adults have a higher health risk than children because children are more vulnerable than adults, and exposure to fine particles early in life may cause long-lasting effects [41] . In addition, a 15-year-old adolescent (male or female) has a higher inhalation rate than an adult female, and hence the median deposited dose rate was higher in the 15-year-old adolescent (14.9 × 10 8 particles/h for male and 14.3 × 10 8 particles/h for female) than in the adult female (14.1 × 10 8 particles/h). However, the deposited dose rates are more similar between adult females and 15year-old adolescents than between adult females and adult males. Therefore, the results in Table 3 show that the deposited dose rate varied depending on the age and gender of the exposed subject. However, the current study did not take into consideration different lifestyles (daily activity pattern) between adults and children and as well as between females and males. Specifically, females receive higher doses than males because they spend more time in indoor environments which corresponds to higher exposure levels [1] .

Mass and number dose rates
The hourly exposure concentrations (μg/m 3 or particles/cm 3 ) and the corresponding deposited dose rates (μ g/h or particles/h) for PM 10 and PN 10 during the cold and warm periods are presented in Figure 3. The results show that the regional deposited dose differed between the two metrics (mass and number). A higher deposited dose rate for PM 10 was obtained in the ET region (10-22 and 8-10 μg/h for cold and warm periods, respectively), while, for PN 10 the higher deposited dose rate corresponded to the AI region (6 × 10 8 -2 × 10 9 and 4 × 10 8 -7 × 10 8 particles/h for cold and warm periods, respectively). These findings are directly linked with the contribution of coarse and ultrafine particles to the particle mass and number size distributions, respectively. Table 4 indicates that coarse particles have negligible contribution to PN 10 but are the dominant contributor to PM 10 . The opposite characteristic was observed for ultrafine particles. Supplementary Figures 2 and 4 demonstrate the relative differences between number and mass size distributions. In more detail, the contribution of coarse particles to PM 10 concentration was equal to 63% for the cold period and 55% for the warm period, whereas the contribution of ultrafine particles (UFPs) to the PN 10 concentration was equal to 71% ± 3% during the cold period and 59% ± 2% during the warm period. The different regional deposition pattern in the human respiratory tract demonstrates different deposition mechanisms. Specifically, coarse particles are deposited mainly in the ET region due to the impaction mechanism, while ultrafine particles are deposited mainly in AI region due to the diffusion mechanism [28,42] . Additionally, the comparison of mass and number dose indicates differences in the sources between the two metrics. Particles from different sources differ in size as well as in the regional deposition and cause different effects on human health. Specifically, exposure to PN 10 indicates heating and traffic sources, which cause increased deposited dose in the AI region due to their small size, while exposure to PM 10 indicates natural sources such as windblown and resuspended dust, which cause increased deposited dose in the ET region due to their large size. This disproportional distribution of regional deposited dose in the human respiratory tract between the two metrics suggests different health implications and highlights the importance of investigating the personal deposited dose on both metrics. Therefore, model results depend strongly on the chosen metric.
It was observed that heating emissions during cold period increased the deposited dose rate for both metrics (PM 10 and PN 10 ). Specifically, the deposited dose increased by 131% and 59% during the cold period for PΝ 10 and PΜ 10 , respectively. In addition, the deposited dose rate for both cases (PM 10 and PN 10 ) in the human respiratory tract (total) was lower in comparison with previous studies [5,43,44] . Manigrasso et al. calculated the deposited dose of PM 10 at an urban background site in Rome and found that the total dose rate ranged from 8.4 to 72 μg/h in February while in the current study it ranged from 12-26 μg/h during the cold period and 10-13 μg/h during the warm period [5] . The urban background site in Rome is close to three roads, and hence higher values were obtained in comparison to the current study. Hussein et al. found that the deposited dose rate of PM 10 was in the range 49-439 μg/h for an adult male under different physical exertion levels (e.g., sitting, standing, driving a car, walking, and running) and environments (e.g., urban, rural, and main road) in an eastern Mediterranean city, highlighting the importance of the type of environment for dosimetry calculations [43] . Regarding PN 10 , Manigrasso et al. found that the dose ranged from 2.1 × 10 9 -2.4 × 10 10 particles/h in February, while in the current study it ranged from 8.3 × 10 8 -3.6 × 10 9 particles/h during the cold period and 6.5 × 10 8 -1.0 × 10 9 particles/h during the warm period [5] . In addition, Buoananno et al.  calculated the daily dose of sub-micrometer particles for the population in the northern and southern parts of Italy and found that the daily dose in the lungs (TB + AI) was equal to 1.5 × 10 11 particles, while in the current study the daily dose of PN 10 was equal to 4.0 × 10 10 and 1.7 × 10 10 particles [44] . The higher doses found by Manigrasso et al. and Buoananno et al. might be attributed to the sampling sites, as measurements were conducted within bigger cities [5,44] . In addition, Buoananno et al. considered a daily activity pattern with a high contribution (69%) from the indoor environment (and the associated sources), whereas, in the current study, only the contribution from the outdoor environment was considered [44] .

Clearance of deposited PM 10
Clearance of deposited PM 10 particles from the human respiratory tract estimated from concentrations of the cold period is presented in Figure 4. Simulations showed that 224 μg were transferred to the esophagus (gastrointestinal tract), 9 μg were absorbed into the blood, 1.1 × 10 −4 μg were transferred to the lymph nodes, and 167 μg remained in the respiratory tract (of the total 437 μg deposited to the respiratory tract at the end of the exposure). Thus, the majority of the deposited particles in the human respiratory tract were transferred to the esophagus (51%), a finding that is associated with the high deposited dose in the ET region due to the high contribution of coarse particles to PM 10 . The clearance of PM 10 depends on the deposition sites: the deposited particles in the ET region are transferred mainly to the GI tract (esophagus) via mucociliary clearance, while the deposited particles in the TB and AI regions require more time to transfer to the ET2 region and are swallowed to the GI tract, and hence they remain in the respiratory tract and absorbed to the blood [22] . Tomasi et al. asserted that mucociliary clearance is a fast process for PM 10 , while for particles with a diameter less than 3 μm the clearance is less efficient [45] . Thompson et al. asserted that 40%-50% of the total mass is retained in the human respiratory tract, which is slightly higher than what was found in the current study (38%) [46] . The present model (ExDoM2) has been used in other studies [22,47] to assess the clearance of PM. Specifically, Chalvatzaki et al. applied the model to data from three European cities and found that the dose to the esophagus was equal to 44, 84, and 172 μg for Kuopio, Athens, and Lisbon, respectively [22] . Therefore, the dose to esophagus for all three cities was lower in comparison to the current study. Nevertheless, PM concentrations in Athens (20 μg/m 3 ) and Lisbon (37 μg/m 3 ) were higher than that in the current study (17 μg/m 3 ). The lower dose in the aforementioned cities is linked with the size distribution characteristics; for example, coarse particles corresponded to 63% of the total measured mass concentration in this study, whereas in Athens, Lisbon, and Kuopio they corresponded to 44%, 42%, and 40%, respectively. In addition, Vicente et al. applied ExDoM2 and found that the dose to esophagus was equal to 380 and 159 μg for an adult male exposed to wood smoke (indoor) from fireplace and woodstove, respectively, with the corresponding outdoor values being considerably lower (152 and 95 μg, respectively) [47] . This finding is associated with a high risk of developing GI cancer due to indoor biomass burning. Specifically, Sheikh et al. found that household biomass burning increased the risk for GI cancer and proposed replacing biomass with natural gas [48] . Therefore, the dose received by a resident during the cold period is significant due to the higher concentration of PM 10 and the associated increased health effects (increased risk for GI cancer due to transfer of deposited particles to esophagus).

CONCLUSIONS
The dosimetry model ExDoM2 was used to obtain the deposited dose rates of airborne particles in the human respiratory tract in a suburban background site. The novelty of the current study is the combination of concentration and full-size distribution (0.01-10 μm) data for two particle metrics (number and mass), giving a more complete view of human exposure and dose. In turn, the main limitation of the current study is that only the outdoor environment was considered for the calculation of deposited dose, and hence different microenvironments (e.g., home) were not considered during a daily pattern.
Traffic and heating emissions were recognized as significant urban-associated sources that caused an increased deposited dose rate in the human respiratory tract. The activity level, age, and gender of the exposed subjects played significant roles in the calculation of the deposited dose rate. Specifically, the deposited dose rate in the respiratory tract was 1.2 times higher in an adult male than in an adult female. Furthermore, the deposited dose rate was 1.1-2.6 times higher in an adult male than in children, while the deposited dose rate for a 15-year-old male was 1.1 times higher than an adult female. Finally, calculations for three activity levels (sitting, light exercise, and heavy exercise) showed that the deposited dose rate was higher during heavy exercise due to the higher inhalation rate.
Regarding the simultaneous measurements of PM 10 and PN 10 , the results show that PM 10 was deposited mainly in the ET region (cold period, 10-22 μg/h; warm period, 8-10 μg/h) while PN 10 was deposited mainly in the AI region (cold period, 5.7 × 10 8 -2.5 × 10 9 particles/h; warm period, 4.4 × 10 8 -6.9 × 10 8 particles/h). The high deposited dose of PM 10 in the ET region was associated with a high contribution from coarser particles, whereas the high deposited dose of PN 10 was associated with a high contribution from ultrafine particles. Overall, simulations demonstrated that seasonal variations and particle origin from variable sources with differences in the size distributions resulted in different regional deposited doses, thus are linked with different health impacts.

Authors' contributions
Conducted model calculations with ExDoM2, performed data analysis and prepared the manuscript: Chalvatzaki E Performed the field measurements and Review/edit of the manuscript: Chatoutsidou SE Made substantial contributions to conception, design of the study and review/edit of the manuscript: Lazaridis M All authors read and approved the final manuscript.