UC Berkeley

As applications of nanotechnology expand, there is an increasing need to develop inexpensive, sensitive, and specific procedures to measure occupational exposures to engineered nanoparticles (ENPs). The use of hand-held direct reading instruments to screen for airborne ENPs is attractive due to the relatively low cost of such commercially available instruments and the immediate feedback provided. However, because ambient air typically contains thousands of non-engineered nanoparticles per cubic centimeter (pt/cm3), this background particle concentration must be accounted for in quantifying ENPs. Incidental nanoscale particles are present in many workplaces due to sources including internal combustion engines, grinding and welding, electric motors, office equipment, printers, and infiltration of contaminated outside air. Because ENP toxicity has not been investigated comprehensively, and because some studies suggest that certain ENPs may be highly toxic, the risk management philosophy of many organizations is that exposure to ENPs above background nanoparticle levels is not acceptable. Industrial hygienists have generally approached ENP sampling with the notion that if ENPs cannot be detected above background levels in the sampling environment, exposure to ENPs is not occurring. If there is a low signal-to-noise ratio, however, this approach underestimates ENP exposure intensity, and large fluctuations in background concentrations could make it difficult to measure actual releases of ENPs. Therefore, it is important to accurately discern between background non-engineered nanoparticle and ENP concentrations. The few ENP exposure studies that have been published tend to inadequately account for background nanoparticles.

Tools and techniques were developed to improve the sensitivity and specificity of ENP exposure measurements in research laboratories based on direct reading instruments and filter-based sampling. These research facilities were located at the Lawrence Berkeley National Laboratory (LBNL). A portable fume-hood antechamber enclosure and a portable, bottomless glovebox enclosure were constructed. These enclosures are supplied with air passed through a high efficiency particulate air (HEPA) filter air, and ENP-handling tasks are performed within the enclosures. HEPA filtration greatly reduces the nanoparticle concentration in the supply air and thereby increases the ENP signal-to-noise ratio. Based on measurements made with a commercially available CPC, a 100- to 300-fold increase in sensitivity was achieved with the clean air enclosures. Combined with high-resolution microscopy analysis of particles collected on filter samples of air, employee exposures to ENPs that were previously not detectable using traditional sampling techniques were quantified.

Three different monitoring studies are described in this dissertation. The first study characterized airborne metal-oxide LiNi0.45Mn0.45Co0.1-yAlyO2 ENP emissions associated with a nanomaterial synthesis comprised of three steps - heating a precursor solution, combustion, and harvesting - all conducted within a laboratory fume hood. With the use of the cleanroom enclosure, background particle concentration levels of 0-2 pt/cm3 were achieved. A CPC was located in the cleanroom enclosure at waist-level or near the personal breathing zone (PBZ) of the LBNL researcher, and a second CPC was located inside the laboratory fume hood. The maximum of 465,129 pt/cm3 (measured inside the fume hood) was the highest particle concentration recorded in this research. For all the combustion events, I observed plumes of black particles which corresponded in time with the highest particle measurements. To my knowledge, there are no published studies that have shown such high particle number concentrations from ENP synthesis. At the conclusion of the synthesis, ENM was found on fume hood surfaces and equipment, and on the CPC instrument located in the fume hood. ANOVA demonstrated that the mean particle number concentrations were not equal across the three steps (p <0.05). The mean particle number concentrations were greater during the combustion (115,602 pt/cm3) than during other steps: heating precursor solution (55,023 pt/cm3), and harvesting (36,708 pt/cm3). Filter analysis confirmed that ENPs became airborne. All filters located inside the fume hood were heavily loaded with agglomerated source material. The primary particles were rounded, often spherical and averaged 5-10 nm in diameter. The EDS spectrum of the nanomaterial indicates that the particles were composed of Ni and Mn.

The second study evaluated the release of airborne ENPs during ultrasonication of ENP suspensions contained in sealed vials immersed in a water bath. The water bath was located within a clean air glovebox enclosure. There were four experiments - two involved a graphene ENP suspension in isopropanol, one involved an AlZnO ENP suspension in hexane, and the fourth was a control that involved ultrasonication of the bath water with no vial of ENP suspension present. CPC measurements and filter samples were collected inside the glovebox. Unfortunately, the results were internally inconsistent, for unknown reasons. During the two graphene ENP trials, the air sample filters were overloaded with carbon particles for which the structure could not be determined by the microscopy laboratory. However, the mean CPC particle measurement during one trial was 4,820 pt/cm3 and only 36 pt/cm3 during the other trial. In addition, the screwcap of the graphene ENP suspension vial was found to be loose at the end of the trial for which the mean concentration was 36 pt/cm3, but remained tight at the end of the trial for which the mean concentration was 4,820 pt/cm3. Given the unknown filter particle structure and the inconsistency in the particle concentrations, the source of the particles on the filters cannot be determined conclusively. Next, except for the graphene ENP suspension trial, in which a mean particle concentration of 4,820 pt/cm3 was measured, the average particle concentration was highest for the control trial (mean = 84 pt/cm3), which did not involve an ENP suspension. Aerosolization of bath water may have caused the concentration increase above background during the control trial, but logically the same magnitude of increase due to bath water aerosolization should have been observed during the other two trials. Future studies need to investigate water aerosolization and NSP emission from ENP production equipment (for example, the ultrasonicator) in contributing to background NSP levels.

The third study investigated the aerosolization potential while handling MWCNTs and CNFs with the use of the glovebox. CPC measurements and filter samples were collected inside the glovebox. Background laboratory particle number concentrations as high as 1,671 pt/cm3 were reduced to 0-2 pt/cm3 inside the glovebox. The CPC was positioned alternatively one inch and six inches from the emission source. As expected, higher particle concentrations were found when the CPC was located closer to the source. Particle concentrations were higher than background during the introduction of the nanomaterial into the glovebox and when handling the nanomaterial. CNFs demonstrated greater particle release as compared to MWCNTs. The CNFs reached a maximum of 70 pt/cm3, whereas the maximum particle number concentration for the MWCNTs was 8 pt/cm3. The results from this study can serve as a comparative indicator of emissions for different ENPs.

In summary, this dissertation work has permitted measuring airborne ENP concentrations with increased sensitivity and specificity. In turn, this improvement has permitted more reliably establishing ENP exposure controls in LBNL research-and-development settings. These same tools and techniques may prove feasible in assessing engineering control efficacy in small-scale commercial manufacturing and/or handling of ENP materials. In addition, they may permit more accurate assessments of airborne ENP exposure which would benefit future epidemiological studies of worker cohorts handling nanomaterials.