Physical characteristics of nanoparticles emitted during drilling of silica based polyamide 6 nanocomposites

During the past decade, polymer nanocomposites have emerged as a novel and rapidly developing class of materials and attracted considerable investment in research and development worldwide. However, there is currently a lack of information available in the literature on the emission rates of particles from these material. In this study, real-time characterization of the size distribution and number concentration of sub-micrometer-sized particles (5.6-512 nm) emitted from polyamide 6 nanocomposites during mechanical drilling was made. For the first time, four different silica based filler of commonly use were assessed. Further, the respective emission rates were determined based on the particle population and the time. The measurements showed that the particle emission rates ranged from 1.16E+07 (min−1) to 1.03E+09 (min−1) and that the peak diameters varied from 29.6 to 75.1 nm. Airborne particles in the nanometer range (11.1-46.8 nm), in the ultrafine range (51.3-101.1 nm) and in the accumulation mode range (111.9-521 nm) accounted for 34.1% to 76.6%, 8.3% to 47% and 4.1% to 24.2% of the total emission rates, respectively, depending on the type of filler. Additionally, deposited particles were sampled and characterized, to explore any possible correlation between deposited and airborne particles. The result clearly showed that with increasing airborne particle concentration the deposit particle concentration decreased and vice verse.


Introduction
Polymer nano-composites have proven huge potential in both mechanical and thermal application J. et al. [2010], Njuguna et al. [2011]. The global consumption of nanocomposites is estimated to exceed 214,081 metric tons and 1.38 billion USD, by 2014BCCResearch [2010]. Of which the consumption of clay-based polymer nanocomposites will increase to 181,094 metric International Conference on Structural Nano Composites (NANOSTRUC 2012) IOP Publishing IOP Conf. Series: Materials Science and Engineering 40 (2012) tons and 692.3 million USD by 2014. For example, one of the leading automotive manufacturers is using 300000 kg of nanoclay composites annually for various automotive exterior part/panel applications at presentBCCResearch [2010]. However, only little attention has been payed to the end user contact with nanocomposites by science and regulators. Despite the worst case scenario of a nanocomposite being the encounter with a do it yourself (DIY) worker C. et al. [2010]. The DIY-worker's personal protection in general consist of an low-cost filter mask, which might be inadequate for protection against nanoparticle particles inhalation. In the car tuning scene, a common way to reduce the weight of a car is the drilling of small holes. Further other car modifications e.g. integration of a new stereo, system or body kit are imaginable and will involve drilling into those nanocomposites Scott J.  D. et al. [2011]. In the current work, airborne particles emitted by silica filled polyamide 6 nanocomposites by mechanical drilling were investigated in a controlled environment. The particle number concentration and the size distribution were measured continuously in real time using a particle sizer. In addition, the particles emission rates were evaluated as this information could be useful in assessing the resulting human exposure.

Nanocomposite manufacturing
A twin-screw extruder (ZMK/116/10, ZAMAK -Cable Machinery Plant, Poland) was utilized to compound the PA6 matrix and 5 wt.% of montmorillonite (MMT), nanosilica (SiO 2 ) and foamglass-crystal-materials (FGC). Extruder screws diameter was 24 mm, had a length/diameter ratio of 32 and consisted of 6 barrel zones. A conventional screw configuration for material compounding was. The screws were equipped with two high-shearing zones in a second and fifth zone. Single-screw extruder (ZMK/117/10, ZAMAK Zaklad Maszyn, Poland) with a length/diameter ratio of 25 and a screw diameter of 25 mm was used for manufacturing of the PA6-glass fiber (PA6/GF) granulate. The single-screw extruder provided effective homogenization of classic fillers without changing the geometry of glass fibres. The panels were prepared by compression moulding technique. The mould temperature was kept at 250 • C over the compression time of 5 min. Due to the high dimensions (100x100x10 mm 3 ) of the panels, cooling of polymer melt was completed in the mould in order to avoid warping.Additional unreinforced PA6 panels have been manufactured in the same way as the reinforced ones, as reference material

Measurement of number concentration and size distribution
To characterize the physical properties of particles generated during drilling, composite panels were fixed in a chamber, with dimensions 68.58cm (width) x 33.02cm(depth) x 55.88cm(height) as illustrated in Fig. 1. The particle emissions were measured using a Condensation Particle Counter "CPC" 5.403 with Classifier "Vienna"-DMA 5.5-U (SMPS+C, Grimm Aerosol, Germany). SMPS+C measured sub-micrometer particles generated during drilling process over a particle size range of 5.6-1083 nm and a particle size resolution of 32 channels in total. An angle drill (Makita BDA351Z 18V LXT Angle Drill) was used for drilling with a maximum speed of 1800 min −1 , adapted with a conventional drill bit of 10mm diameter. Prior to measurements, the chamber was purged with laboratory air for about 20 min. Each sampling cycle comprised a 60 min background air monitoring in the chamber, 14 min. of active drilling, and a 60 min postdrilling period. The experiment was repeated 3 times for each material composition. A total of about 2000 to 3000 data sets were collected for each particular sequence.

Determination of emission rates
The emission rates of particles of different sizes were determined from the number concentration and size distribution data based on a single compartment population balance model. Liu et al. W. et al. [2003] used a mass balance model to estimate emission rates of particles based on mass concentrations. As reported by See et al. S.W. et al. [2007], the mass balance model calculation can also be used to estimate the particle emission rate based on particle population. Basic assumptions for this model were made as followed: a) particles are assumed to be perfect spheres with a constant density; b) background concentration is zero; b) particle concentrations are homogeneous within the chamber; and d) emission rate and decay rate of the particles remain constant throughout the entire period of generation. The respective removal rates k x (min −1 ) of a particle size x were evaluated using the following equation: where C x (cm −3 ) represents the number concentration of particles of size x at time t (min), while C max;x (cm −3 ) represents the maximum concentration of particles of size x which at the time T (min). The background concentration of particles, averaged over the 60 min pre-burning interval, was subtracted from the concentration readings to compute C x (cm −3 ) and C max;x .
After calculating the removal rates k x , the emission rate of particles of size x, P x (min −1 ) in the chamber volume V (1,26540x10 5 cm −3 ), was calculated from:

Sequential alteration of number concentration and size distribution
The sequential alteration of the total number concentration of particles for a typical sampling cycle is shown in Fig. 2. It consists of a 60-min background measurement (t ≥ T ), the 15min drilling period(0 ≤ t ≤ T ), and a 120-min post-drilling period (t ≥ T ). Figure  2, shows that the total number concentration of particles was essentially constant with an  et al. [2007] the influence of the machining engines as "background noises" on the results has been reported. As the drilling was conducting in a controlled environment, the influence of the machining engine could be reduced to a minimum, as shown in figure 2. Highest C max could be measured for the nanosilica filled composites. Wohlleben et al. W. et al. [2011] reported that, while comparing nanocomposites with their reference materials without nanofillers, the differences are insignificant in the number concentrations of aerosols during normal abrasion use. This statement can not be substantiated, as the results strongly suggest that the integration of nanofillers changes the material behavior and thus the amount of particle emitted during drilling. Diffusion, gravitational deposition, convection, impaction, and coagulation are some of the complex processes which influence formation and removal of particles in the chamber. Removal of small particles is primary dominated by diffusion while larger particles are mainly affected by gravitation. During diffusion, small particles collide with one another and form larger particles Grassian V. H. [2008]. This process is known as coagulation and strongly depended on particle size and concentration. While the coagulation rate of a simple monodisperse particle population can be calculated using equation 3 Hinds [1982], where N is particle number concentration and K is the coagulation coefficient. Based on this equation, the coagulation rate is directly proportional to the diffusion coefficient and particle size and therefore decreases with particle size. An indication of the coagulation rate can be identified by the half-life of an individual particle, which is introduced into an atmosphere embracing a defined concentration of such particles Grassian V. H. [2008]. However, according to Hinds Hinds [1982] coagulation can be neglected for laboratory experiments if the particle concentration is less the 1x10 12 m −3 . As the maximum concentration measured in this study did not exceed 1.7x10 11 m −3 , coagulation was neglected. The particle size distributions of drilling into different material systems is shown in Figure 3. The plots reflected particle size distributions at T, the time particle concentration reached it maximum and drilling was terminated. The plots presented the normalized distribution with ∆N/∆lnd p versus the particle diameter d p , where ∆N is the concentration of particles within a specified size interval and ∆lnd p is the difference in the natural logarithm of the largest and smallest particle sizes of that interval. From visual inspection of the graph, it could be noted that from different composite distinct modal diameters were obtained and that they appear to be dependent of the filler used. A major difference in peaks could be observed between the different fillers. The peaks varied from 29.6 nm for the nanosilica composite to 75.1nm for neat PA6. This showed that integration of nanofillers changes the physical properties of the emitted particles. This results contradict with the results obtained in some recent studies I. K et al. [2009], W. et al. [2011], M. et al. [2009]. For example, Wohlleben et al. W. et al. [2011] reported, while comparing nanocomposites with their reference materials without nanofillers, that both the differences in the number concentrations and in the actual size distribution of aerosols during normal abrasion use, were insignificant. However, these studies dealt with abrasion and sanding of surfaces and composites. The difference in mechanical processing could explain the alterations in the obtained results. To support the results obtained by SMPS+C particles were sampled via an electrostatic precipitator (ESP, Grimm Aerosols, Germany) and subjected to JFEI XL30 field emission scanning electron microscope. The scanning electron micrographs of the particles generated during drilling of different composites are represented in Figure 4. Micrographs of the particles generated from PA6/FGC (Fig. 4a)., PA6/SiO 2 (Fig. 4b)and the neat PA6 (Fig. 4e) composites showed very similar structures of coagulated particle. Smaller generated particles (20-50nm), coagulated to larger particles 150nm, which were then sampled by ESP over a time frame of 2h. According to the literature Grassian V. H. [2008], the half time of particles in the size range of 10-200nm in the measured concentration would be between 16-83 min, therefore coagulates of particles are natural for such a long sampling period. The SEM image of the generated particles obtained from drilling of the glass fiber panels (Fig. 4d), showed agglomeration of larger particles. Furthermore the particles seem to be glass fibers debris, as the structure is very similar to structure seen in the bulk panels. Visual investigation of the particles obtained from drilling of the nanoclay composites (Fig. 4c)were scattered over the sampling plate and no coagulation could be found. This could be explained by the low particle concentration that was measured for this composite.

Emission rates of different composite materials
The emission rates of size-distributed particles from different composites were determinate using Eq. 2, after taking into account the removal rates of particles of different sizes from Eq. 1. The stacked column charts of Figure 5 shows the number of particles that were emitted per minute  Figure 4: SEM micrographs of generated particles collected by ESP during drilling of the composite materials. The error bars represent the standard deviation associated with the total emission rate of 11.1-521nm particles resulting from the three replicate measurements. Table ?? represents the total emission rates for the different composite materials. Lowest emission rates could be calculated for the nanoclay filled composites, followed by the neat PA6 panel. Highest emission was calculated for the nanosilica filled material. However, while comparing the percentage of the individual size intervals on the total emission, it could be seen that for all materials (beside nanoclay) the majority of particles, 45-59%, were in the size range of 22.6-42.6 nm ( Figure 5). Further the un-filled material emitted the largest quantity (25%) of particles larger than 100nm. The emission rate increased by 56 times for the nanosilica filler and between 20-45 times for the glass fibre and foam-glass crystal filler. Interestingly, integration of nanoclay into the PA6 matrix reduced particle emission during drilling by 1.5 times. It is likely that the presence of nanoclay in some way retains the formation of high quantity of airborne particles. This could be caused due to modification of the nanoclay prior integration of the matrix, however other reason could be plausible e.g. shape of the nanoclay filler and integration mechanism into the matrix (intercalation/exfoliation) A. et al. [2007a,b].

Conclusion
Physical characterization of the number concentration and size distribution of sub-micron particles from 5.6 to 512nm was carried out, for the first time, for different silica based nanocomposites. Two kind of nanosilica fillers (nanoclay, nanocilica) and two kind of macro silica filler (glass fibres and foam-glass crystal) were successfully integrated into a PA6 matrix. Based on size-resolved particulate emission rates from various silica based composites during mechanical drilling process were evaluated. In general, nano and ultrafine airborne particles were emitted from all investigated materials, even the non reinforced polymer. However emission increased by 56 times for the nanosilica filler and between 20-45 times for the glass fiber and foam-glass crystal filler. Interestingly, integration of nanoclay into the PA6 matrix reduced particle emission during drilling by 1.5 times. However, the characterisation of deposited particles showed exactly the opposite particle behavior, as with decreasing airborne particle concentration the deposit particle concentration increased and vice verse. This study has certainly shown that the integration of  Figure 5: Emission rates of particles for different composites nanofillers into polymeric matrix, changes the behaviour of the composite material in terms of particle emissions. Therefore, future work will involve a series of comprehensive chemical characterization of size-distributed particles released during drilling of nanocomposites to make a better assessment of public health outcomes due to exposure.