Supplemental Information

Development and evaluation of a nanoparticle generator for human inhalation studies with airborne zinc oxide

Christian Monsé, Christian Monz, Dirk Dahmann, Christof Asbach, Burkhard Stahlmecke, Norbert Lichtenstein, Karl-Ernst Buchwald, Rolf Merget, Jürgen Bünger, Thomas Brüning

Zinc oxide particle synthesis

Zinc oxide (ZnO) nanoparticles were produced by pyrolysis of atomized liquid precursors in a hydrogen-oxygen-flame. Four precursor solutions were prepared by dissolving 8.33, 16.67 and 33.33 g zinc acetate dihydrate (Zn(CH3CO)2 * 2 H2O, purity 99 %, Merck GmbH, Germany) with 4.0 mL acetic acid (HAc) which actedas a stabilizer, (purity 99 %, Merck GmbH, Germany), and brought to a total volume of 1000 mL with purified water to yield Zn concentrations of 0.038, 0.076 and 0.152 mol/L, respectively. The fourth solution was prepared by dissolving of 25.87 g zinc formate dihydrate (Zn(CHO2)2 * 2 H2O, purity 98%, Alfa Aesar GmbH, Germany), with 3.0 mL acetic acid as a stabilizer (purity 99 %, Merck GmbH, Germany), and brought to a final volume of 1000 mL with purified water to yield a Zn concentration of 0.135 mol/L. Solutions 1 to 3 were fedthrough the two-substance nozzle at 0.25to 1.5 mL/min by a syringe pump(Perfusor secura ST, BBraun GmbH, Germany) and atomized by co-flowing 4 to 6 L/min of nitrogen (nitrogen generator, model NGM 22, cmc instruments GmbH, Germany) or compressed air at 0.20 to 0.40 bar. The gas flow of hydrogen was varied between 5 and 20 L/min, the oxygen flow between 2.5 and 10 L/min and the argon flow between 0 and 4 L/min. Solution 4 was used to characterize the resulting ZnO material via elementary analysis and Fourier Transform Infra Red (FTIR) spectroscopy, and to perform homogeneity experiments in the ExpoLab.

Characterization of the generated ZnO nanoparticleat the nanoTest Center (IGF)

Figure S1 shows the complete experimental arrangement at the nanoTest Center (IGF).The nanoparticle generator 4 was positioned in front of a steel pipe 1 (diameter 0.5 m, length 20 m). The airflow from the ventilator 3 is adjustable from 353 to 5655 m3/h (the corresponding airflow velocity is 0.5 to 8.0 m/s). It sucks ambient air without prefiltration through the pipe 1. All outgoing nanoparticles from 4 are diluted, mixed and homogenized in the pipe to reach the measurement chamber 2, equipped with a Scanning Mobility Particle Sizer (SMPS, model 3936, TSI Inc.,USA, equipped with a long DMA and a butanol CPC TSI model 3010), operated with an aerosol and sheath flow rate of 0.6 and 6 L/min to measure the particle and number size distribution in a range between 9.8 and 414 nm, respectively, and the temporal stability of the particle number concentration. Typical particle number concentration of the ambient background air was in the range of 103 to 104 particles per cm3(#/cm³). These values are negligibly small in comparison to a magnitude of 106#/cm³ in our experimental settings.

Figure S1:

Experimental arrangement of generation and characterization of ZnO nanoparticles (1: 20 m steel pipe; 2: measurement chamber; 3: ventilator; 4: nanoparticle generator; 5: SMPS; 6: APS; 7: NAS).

An aerodynamic particle sizer (APS, model 3221, TSI Inc., USA) was used to measure the particle number size distribution of particles with aerodynamic diameters approximately >600 nm in order to characterize the efficiency of the pyrolysis process (6 in figure S1). Such large particles would only be detected in case of incomplete decomposition of the liquid precursor solution in the flame. The nominal measurement range was between 0.5 to 20 µm.

An electrostatic Nanometer Aerosol Sampler (NAS, model 3089, TSI Inc., USA; Dixkens and Fissan (1999) was used to sample particles onto 10 x 10 mm2 silicon substrates for consecutive SEM/ EDX analyses (7 in figure S1). The NAS was used with a custom-built unipolar corona charger upstream of the inlet area. The charger uses a sharp tungsten needle, around which a corona is formed when a DC voltage between +2.5 and +4.0 kV is applied. The corona generates ions that positively charge the particles to achieve nearly 100 % particle sampling efficiency in the electric field of the NAS (usually operated at – 10 kV). The sampling times were varied from 1 to 3 minutes with an internal airflow of 1.5 L/min.

The morphology of the collected particles on the silicon substrates were characterized by scanning electron microscopy (SEM, JSM-7500F, JEOL Ltd., Japan) with a nominal resolution of 2 nm.

The elemental composition of the collected particles as well as the silicon substrate was determined by energy-dispersive X-ray spectroscopy (EDX) (Apollo XLwith 30 mm2 detector, Ametec, EDAX Inc., USA).

Characterization of the generated ZnO nanoparticlesin the Expolab (IPA)

We integrated the flame generator into the air conditioning system of the exposure unit. Additionally, a cold water-cooling unit was installed near the generator to reduce the waste heat of about 3 kW and to prevent an influence of the climate parameters. Sound emissions of the burner head were completely eliminated with an integrated sound absorber.

Measurements of particle size distributions in the ExpoLab were performed with the same SMPS as at the nanoTest Center using the same measurement conditions. The system was installed on a mobile table. The number concentration and the size distributions were determined in the unit at a height of 1.20 m, which is representative of the height of a subject’s breathingzone whilesitting.

Trace gas analyses in the ExpoLab

All experiments were performed with a solution of zinc acetate in water, stabilized with 4.0 mL/L HAc, with a concentration of 33.33 g/L. The target exposure mass concentration was 2 mg/m3ZnO. The calculation was based on an airflow of 360 m3/h in the ExpoLab (air exchange rate is twelvefold per hour), regulated by the controller software of the air conditioner. Climatic parameters were set to 23.0°C and 45% relative humidity to ensure the subjects’ comfort. A ventilator was installed to enable a homogenous distribution of the particles and trace gases in theExpoLab.Initial trace gas analyses were performed to optimize the gas flow rates of hydrogen, oxygen and argon, and to point out the lab’s suitability for a human inhalation study with ZnO. First of all, selection of these trace gas analyses was limited to nitric oxides (NO, NO2) and acetic acid (HAc) via online chemical ionization mass spectroscopy (Airsense.net, MS4 GmbH, Germany). More technical details are available from Hornuss et al. (2007). After defining the suitable conditions of the nanoparticle generator, we also analyzed additional trace gases,besides NO, NO2 and HAc that could be formed by pyrolysis. We monitored NO and NO2 with a chemoluminescence measurement device (CLD 700 AL, Eco Physics GmbH, Germany), and carbon monoxide (CO) and carbon dioxide (CO2) with a NDIR (NonDisperse Infra Red) detector (NGA 2000/ MLT 3T IR, Fischer Rosemount GmbH, Germany). Ozone (O3) was measured with a UV-detector (ML 9810B, MS4 GmbH, Germany). Two impingers were installed to capture hydrogen peroxide (H2O2, IFA-workbook, index 8943). Sampling time was 2 h with a flow rate of 70 L/h and a limit of determination (LOD) of 0.050 mg/m3. Sampling of volatile organic compounds (VOC, IFA-workbook, index 8936) was performed with 5 thermo-desorption tubes (Tenax® TA, Perkin Elmer Inc., USA). Sampling time was 10, 20, 30, 40 and 60 min with a flow rate of 4 L/h, and a limit of determination of 0.005 mg/m3. The analytical determination was performed by thermo-desorption gas chromatography (Turbomatrix ATD, Clarus 500 GC, Clarus 500 MS, Perkin Elmer Inc., USA). One activated charcoal tube (Type B, Dräger GmbH, Germany) was in operation for 2 h with a flow rate of 20 L/h. The subsequent extraction was achieved with a mixture of methylene chloride, carbon disulfide and methanol (60:35:5). The analytical determination was performed by gas chromatography coupled with FID (GC 7890A, Agilent Inc., USA). The limit of determination was 1 mg/m3. One silica gel tube was installed with a sampling time of 40 min and a flow rate of 20 L/h. The subsequent extraction was achieved with sodium hydroxide solution. The analytical determination was performed by HPLC (Modell 1100, Agilent Inc., USA) with a limit of detection of 1.9 mg/m3(calculated for acetic acid).

A multi-component analyzer for CO, CO2, NO and NO2 with an internal suction pump (XAM 7000, Dräger GmbH, Germany) served as a backup system (LOD: CO: 2 ppm; CO2: 100 ppm; NO: 0.1 ppm; NO2: 0.1 ppm). Additional measurements were performed with online mass spectroscopy to exclude the formation of known byproducts acetone, propylene (Becker, 1948) and acetic anhydride formed from the pyrolysis of zinc acetate with a LOD of about 1 ppb.

Elementary analysis and FTIR spectroscopy of ZnO particles

ZnO particles were collected onto cellulose nitrate filters (47 mm diameter, pore size 8 µm, Type 11301-47-N, SatoriusStedin Biotech GmbH, Germany) with a volumetric flow rate of 46.5 L/min for about 3 h. The deposited material was mechanically separated from the filterswith a spatula (each about 10 mg), dried at 120 °C for 12 h and sentto an external laboratory (Mikroanalytisches Labor Pascher, Germany) to determine the carbon, hydrogen, nitrogen and carbonate contents. FTIR spectroscopy (Vector 22, Bruker GmbH, Germany) was performed with the collected material, ZnO (purity 99 %, Merck GmbH, Germany), zinc formatedihydrate (Zn(CHO2)2 * 2 H2O, purity 98 %, Alfa Aesar GmbH, Germany) and zinc nitrate hexahydrate (Zn(NO3)2 * 6 H2O, purity 98 %, Merck GmbH, Germany).

Measurements of mass concentration of airborne zinc oxide

To calculate the concentration of the liquid precursor with a target exposure mass concentration of 2 mg/m3 of airborne ZnO, the air exchange rate has to be exactly determined. To accomplish this, we monitored and analyzed a decay curve with the help of an inert gas (Lohmeyer 1992, Raatschen, 1995). Propylene (1.0 ppm) was continuously dosed into the ExpoLab until it reached a stable concentration. The propylene supply was closed while venting the unit with fresh air and the decay was recorded. A ventilator was used to homogenize the continuously diluting gas atmosphere. The software “GraphPad Prism 5.04” (GraphPad Software Inc., USA) was used to calculate the desired air exchange rate. The underlying mathematic model was the analysis of the 1 ppm plateau followed by one-phase decay with a least square fit.

Mass concentration measurements of airborne ZnO were taken continuously (recorded at 1-minute intervals) using a tapered elemental oscillating microbalance (TEOM, model 1400a, Rupprecht and Patashnik, USA).

Figure S2:

Average particle size distributions in the ExpoLab at 2.0 mg/m3. A: Without portable fan; B: with portable fan.

Table S1:

Mean particle concentrations and mean median diameters and their standard deviations with different liquid precursor flow rates.

Table S2:

Particle concentration and median diameter with different atomizing pressures.

Table S3:

Influence of different gas flows on particle concentrations and median diameters.

Table S4:

Results of the trace gas analyses compared to the respective German occupational exposure limits (OEL). LOD: Limit of determination

Table S5:

Median diameter with and without portable fan at different measurement point in the ExpoLab at 2.0 mg/m3.

Supplemental References

-Becker, F. (1948). Knowledge of the thermal decomposition of zinc acetate.

-Dixkens, J., Fissan, H.(1999). Development of an electrostatic precipitator for off-line particle analysis. Aerosol. Sci. Technol., 30: 438 – 453.

-Hornuss, C., Praun, S., Villinger, J., Dornauer, A., Moehnle, P., Dolch, M., Weninger, E., Chouker, A., Feil, C., Briegel, J., Thiel, M., Schelling, G. (2007). Real-time monitoring of propofol in expired air in humans undergoing total intravenous anesthesia. Anesthesiology, 106:665–674.

-Lohmeyer, G. (1992). Practical building physics.B. G. Teubner Verlag, 527.

-Raatschen, W.(1995).Measurements of tracer gases in the building technology: air exchange – measurement and simulation. gi – Gesundheitsingenieur, 116, 2, 78-87, 129 – 138.

1