Evaluation of the Airborne Particles Fraction Responsible for Adverse Health Effects

A new approach to evaluation of the fraction of airborne particles that is responsible for adverse health effects has been developed. In this approach a two-fraction concept for the adverse health effects is proposed. This concept considers aerosol particles in the air, the fraction of particles deposited in the respiratory system (captured by the system) and the fraction of particle accumulated in the body (captured by the body). The latter fraction of airborne particles actually causes the adverse health effects and associated with the health risk. The approach has been applied to workers in crystal glass industry exposed to lead. A size resolved sampling technique was employed to characterize nanoparticles at work places at several plants involved in lead processing. Lead aerosol size distributions were obtained across the entire aerosol particle size range from 1 nm to 30 μm (diameter). Size distributions were used to calculate the total lead intake due to particle deposition in the respiratory system of workers. A bio-kinetic model was employed to evaluate various size fraction contributions to the total blood lead level in workers. A comparison was made between calculated blood lead levels and actual measurements of blood lead levels of workers exposed to atmospheres containing lead. It was found that nanoparticles cause the major health risk due to high deposition efficiency and low clearance rate by the mucociliary escalator.

following exposure to nanoparticles. There is a growing concern that nanoparticles (less than 100 nm in 38 diameter) have an increased toxicity relative to larger particles composed of the same materials, e.g. of manufacturing processes such as incineration, smelting, welding and others. There is also increasing awareness of health risks that may occur with the production and commercialisation of novel nano-48 materials where specific properties are engineered into the material at the nano-scale. There is concern 49 that the enhanced properties engineered into the material will also impact biological activity and toxicity, 50 increasing the health risks posed by these new materials.

51
Currently exposure level is evaluated using total mass concentrations of material of concern in a size 52 range below a certain cut off diameter, for example 10 11m (PM 10 ) and 4 11m (the respirable fraction). The (1) 87 The Nd(D) represents the number distribution of particles that were captured by the respiratory system 88 and 0 is the particle diameter. The efficiency of deposition in the respiratory system was subject to 89 experimental research and modeling and widely available, e.g. Chamberlain (1985); Hinds (1999 111 Nh(O) is the differential form of the body burden representing the number of particles of concern 112 related to certain size range 0 to O+dO. The total body burden is the integral of expression (2

151
The aim of this paper was to develop a new approach for Health Risk evaluation based upon 152 quantification of the Accumulated Dose (HR-AD) using lead exposure data obtained at various factories 153 over the entire airborne particle size range. This approach was focussed on a study of workers exposed 154 to lead containing particles produced as a by-product of the manufacturing processes in the European industry. Data on aerosol particle size distributions and blood lead concentration were collected at working places at several European factories processing lead and lead compounds both mechanically and 176 chemically including high temperature stages. In addition, we investigate the hypothesis that significant 177 airborne lead exposure in the nano-size fraction is generated in crystal glass factories and we clarify the 178 role of nanoparticles in the total lead body burden. An area sampling approach was adopted, in part 179 because efficient personal samplers for nanoparticles were not (and still are not) available at the time of 180 the study.

181
The HR-AD method requires three sets of data: the aerosol particle size distributions 10(0), deposition 182 efficiency in the respiratory system £(0) and lead bio-kinetic model. A validated bio-kinetic model of 183 lead metabolism in humans (ICRP, 1975 andLeggett, 1993) was employed to predict blood-lead levels in 184 workers exposed to lead aerosols. The deposition efficiency of aerosol particles in the respiratory system 185 was taken from Hinds (1999) and ICRP (1994). Although there are other models available, e.g MPPD 186 model for aerosol deposition estimation in human lung, see Anjilvel and Asgharian (1995), the choice of

196
The initial stage of the process is the provision and mixing of the raw materials, which include

208
The main processes in a crystal glass production factory believed to generate the highest airborne lead  particles. There is the only one instrument on the market that uses both sizes and, therefore, emulates 315 the natural deposition in the respiratory system Nano ID Select.
This unique instrument (Nano-ID Select, Particle Measuring Systems) has been employed to collect 317 airborne particles in a wide range of aerosol particle diameters from 2.5 nm to 30 11m. Aerosol particles 318 were collected in 11 size fractions at a flow rate 20 L/min and a relative humidity of 80%. Sampling at a 319 controlled humidity is an important feature of the Nano-ID Select instrument that enables conditions in 320 the respiratory system to be emulated. Therefore, deposition within the instrument occurs under the 321 same effective humidity as in human respiratory system. This reduces effect of water uptake on the 322 particle size distribution measurements.  that represents the width of distributions <J g was assumed to be 1.6. The total mass of lead to have 378 entered the body was found from the integral of the size distribution over the size range of particles.

379
It clearly shows (Table 1) the link between the mean diameter of airborne particles and the blood level 380 of lead accumulated. The exposure to nanoparticles of 10 nm will cause greater blood level and 381 consequently more potential health risk than exposure to 100 nm or 1 11m particles. Even greater 382 deposition was obtained for 10 11m particles. Thus, neglecting the deposition efficiency can cause 383 considerable overestimation in the body burden and the health risk. However, it should be mentioned 384 that the health risk associated with these large particles is not necessarily very prominent due to other 385 factors discussed later.

Evaluation of the fraction of lead staying in the body compartments and causing the adverse health
388 effects (biological availability of lead particles) 389 Airborne particle size distributions along with the bio-kinetic model and the respiratory system 390 deposition efficiency enable the fraction of particles causing adverse health effects A(D) to be evaluated.

391
The blood lead content was calculated using aerosol particle size distributions in three stages.

398
Every A(D) model generated certain levels of lead in the blood of workers. Several simple models of 399 bio-availability were considered. The first model is a non-size selective model. It is described by a single 400 parameter that represents a fraction of bio-available lead (the same over the entire particle size range).

401
Two size-selective models were considered. For simplicity, these were based on the concept of the 402 critical particle size Dc. In the second model, it was assumed that particles of greater diameters than Dc 403 could be bio-available but smaller particles removed by unknown means. In the third model we assumed 405 bioavailable, but the larger particles (0 >Oc) could not be dissolved and were eventually removed by 406 cleaning mechanisms. In addition, small particles stay longer in the lung due to alveolar deposition and 407 consequently prolonged retention causes slower clearance.

408
In this way concentration of lead in the blood of workers C cal was found using airborne particles size Where C gt is the lead concentration associated with the consumption of lead through non-lung 447 pathways. Expression (9) was used to find both Dc and C gt •

448
Regression analysis revealed a poor correlation between measured and calculated blood lead content 449 for the second model. The R 2 was even less than for the non-size selective model.

450
In the third model mass was calculated using expression (10) that was similar to one used for the

466
Another cause of the origin of Dc is clearance by the mucociliary escalator of the respiratory system.

467
Larger particles are removed from the respiratory tract faster than nanoparticles because smaller and  (1994). Swallowing of particles from lung clearance potentially 470 may bring these particles into the blood stream. This mechanism is not very well understood/quantified 472 comparison to weeks and months for the respiratory tract and especially alveolar region, Leggett (1993).

473
Therefore, swallowed particles have higher probability to be excreted than to be dissolved into the 474 blood. The mechanism of the effect of particle sizes on the bio-availability is not well understood at the 475 moment. However, it is clear that nanoparticles and micro-size particles of lead have different bio-476 availability.

477
It is important to point out that toxicity of airborne lead is associated mainly with smaller particles 478 (D<200nm), see Figure 4. Thus, at working places and in general environment this fraction of airborne 479 lead is the subject of major concern and should be considered most dangerous for health. It is also 480 practical and may be cost effective to concentrate, at working places, on eliminating the sources of this  Table 2. 488 Table 2 reveals a poor correlation for the PM lO fraction (R 2 <0.36), for PM 2 . 5 (R 2 <0.68) and PM l (R 2 489 <0.72). These are considerably lower than R 2 value for Dc = 200 nm (R 2 =0.91). Thus, for the lead aerosols measured, the health risk would seem to be associated with the inhalation of the ultra-fine fraction of 491 the aerosol (Oc< 200 nm) rather than with the coarser fractions.

492
A good illustration of the importance of two-fraction concept for the HR-AD approach and hence for 493 the health risk evaluation can be seen comparing lead mass: in the air, captured by the respiratory the adverse health effects (black) is close to the fraction captured by the respiratory system for small 499 particles (range of particle sizes from 23 to 100 nm). However, for the larger particles (the size range of 500 from 2,000 to 4,000 nm) the black fraction (responsible for adverse health effects) is almost zero. Thus, 501 at each stage corresponding to the first and the second fraction the mass available to cause the adverse 502 health effects is smaller but the degree of bioavailability is not constant and is influenced by the particle 503 size.

504
In the past there were several publications suggesting the higher risk associated with nanoparticles. In 505 this paper, this is quantified and explained by a combination of effects of the size selective deposition in 506 the respiratory tract and removal from the body by cleaning mechanisms.

510
The total mass concentration of lead aerosols determined at working places ranged from 0.6 j.!gjm 3 to 511 50 j.!gjm 3 • The nanoparticle mass fraction of aerosols (sizes less than 0.1 j.!m) was found to vary from 512 10% to 60%. The fraction of lead remaining in the body (responsible for adverse health effects) has been 513 determined. It was found that lead particles with the diameter below 200 nm cause adverse health 514 effects but larger particles do not cause any measurable health effect. The evaluation of health risk 515 based upon PM 10 , PM 2 .5 or PM 1 fractions does not provide correct blood lead content. It is found that 516 PM 10 , PM 2 .5 or PM 1 will overestimate the health risk considerably. It was shown that the approach 517 proposed for health risk evaluation based on quantification of amount of particulate matter of concern 518 captured by body provides valuable information for health risk assessment and air quality control at 519 working places. Authors believe that in the future quantification of exposure to airborne nanomaterials 520 should be based upon size resolved chemical composition data measured over the wide size range 521 including nanoparticles as well as micron size particles, regional airway deposition efficiencies and  Windows platform using C++. It was developed to be run on a Pc. The set of constants is shown in Table   669 Ai in the Annex.

670
The program is based upon the direct integrating of the kinetic equations between compartments.

671
This method requires relatively large computing resources however direct integrating has one important 672 advantage: it enables calculations for a complex intake pattern of a toxic substance. Thus, the program 673 can be used to run the bio-kinetic model with either zero or lower intake, for example during holidays 674 and weekends.

675
The transfer rate from the small intestine to the large intestine (k sl ) was calculated as the ratio of the 676 product to the sum of the transfer rates from the small intestine to the upper large intestine (7) and 677 from the upper large intestine to the lower large intestine (1.8), see Leggett (1993

686
In addition, the code was verified using data on exposed volunteers to lead oxide aerosols obtained by 687 Griffin et 01. (1975). The difference between measured lead concentrations in blood after 100 days of 688 exposure and calculated with the code was less than 10%. Thus, the program provides good agreement 689 with existing published data.

690
The code has also been verified using data on lead in blood obtained by Rabinowitz et al. (1976). In 691 these experiments, lead was introduced to adult human subjects during a period of about 100 days.

692
Maximal value of lead blood content was found to be close to 35 flg/dl. Calculated curves obtained with 693 the code developed were in a good agreement with experimental data, see Figure 6. Lead is a bone-