Exposure to Nanoparticles - Environmental and Human Health Impacts of Nanotechnology - page 335

tion of fi ne particles, suggesting that released particles tend to be larger then 1 m

or so in diameter. The fi ndings agreed with laboratory data indicating that
sub-micrometre particles are not readily released at low levels of agitation.
More energetic processes would be expected to release higher concentrations
of nanotube aerosol. There is some question over whether the use of a vacuum
cleaner during the clean up led to large releases of small nanotube clusters,
or whether the observed particles originated from the device itself. This is clearly
an area requiring further investigation but at this stage it would seem prudent
to use High Effi ciency Particulate Arrester (HEPA) fi ltered cleaners with correctly
fi tted fi lter units to prevent the potential release of large nanotube number
concentrations.
This study provided a fi rst indication of the propensity with which unprocessed
SWCNT forms an aerosol during handling. However, given the diffi culty of identi-
fying CNTs by SEM and the nature of the surrogate analysis (metals rather than
CNTs directly), much work still needs to be done.
A recent study (Han et al. 2008 ) is the fi rst published attempt to measure expo-
sure to CNT using methods similar to those used to measure asbestos. The study
measured exposures in the post-production recovery of MWCNT and in a blending
activity, which is part of a composite formulation process.
An SMPS, which consisted of an electrostatic classifi er (model 3080, TSI) equipped
with a long-differential mobility analyser (LDMA, model 3081, TSI) and ultrafi ne
CPC (UCPC, model 3025, TSI), was used to monitor the particle size distribution,
from 14 to 630 nm. An APS (model 3321, TSI) was used to observe the particle
size distribution, ranging from 0.5 to 20

µ

m in aerodynamic diameter. In addition,
a portable aethalometer (model AE42 - 7 - ER - MC, Magee Scientifi c) was used to
measure the mass concentration of carbon particles based on an optical absorption
analysis.
Air samples were taken by drawing air through mixed cellulose ester fi lters in
sampling cassettes (35 mm diameter, 0.8

µ

m nominal pore size and 2-inch cowl). The
samples were collected in the breathing zone using SKC-117 battery operated
sampling pumps at a fl ow rate of 1.5-2.0 l/min. After mounting on a suitable sub-
strate, the samples were counted using a TEM. All objects, identifi ed as MWCNT
with an aspect ratio greater than three were counted. The diameter and length were
measured and were 52- 56 nm and 1473 - 1760 nm, respectively. Tests were carried
out before and after control measures (essentially isolating the process) were put
in place.
Most of the laboratory MWCNT exposure levels (maximum 0.43 mg/m 3 ) were
lower than the current TLVs for carbon black (ACGIH 3 mg/m 3 and Particles Not
Otherwise Specifi ed (PNOS) (3.5 mg/m 3 ), although the number of tubular struc-
tures (maximum 194 tubes/cm 3 ) was over the current fi bre TLVs (asbestos 0.1/cm 3 ;
glass wool 1/cm 3 ; rock wool 1/cm 3 ; refractory ceramic fi bre 0.1/cm 3 ; etc.). The
MWCNT lengths that were shorter than 5

µ

m may differentiate MWCNTs from
asbestos and other fi bre structures. Health effects of durable shorter fi bres remain
controversial and the durability of MWCNTs is not clearly known. Other main
biological determinants known for fi bres, such as aspect ratio, dimension and depo-
sition, may differentiate MWCNTs from asbestos and other fi bre structures.

µ