Dry sampling of gas-phase isocyanates and isocyanate aerosols from thermal degradation of polyurethane

摘要

The performance of a dry sampler, with an impregnated denuder in series with a glass fibre filter, using di-n-butylamine (DBA) for airborne isocyanates (200ml min-1) is investigated and compared with an impinger flask with a glass fibre filter in series (1 l min-1). An exposure chamber containing 1,6-hexamethylene diisocyanate (HDI), isophorone diisocyanate (IPDI), and 2,4- and 2,6-toluene diisocyanate (TDI) in the concentration range of 5-205 μg m-3 0.7-33 p.p.b.; relative humidity (RH) 50, generated by gas- and liquid-phase permeation, was used for the investigation. The precision for the dry sampling for five series with eight samplers were in the range of 2.0-6.1 with an average of 3.8. During 120-min sampling (n = 4), no breakthrough was observed when analysing samplers in series. Sixty-four exposed samplers were analysed after storage for 0, 7, 14, and 21 days. No breakdown of isocyanate derivatives was observed. Twenty-eight samplers in groups of eight were collecting isocyanates during 0.5-32h. Virtually linear relationships were obtained with regard to sampling time and collected isocyanates with correlation coefficients in the range of 0.998-0.999 with the intercept close to the origin. Pre- or post-exposure to ambient air did not affect the result. Dry sampling (n = 48) with impinger-filter sampling (n = 48) of thermal decomposition product of polyurethane polymers, at RH 20, 40, 60, and 90, was compared for 11 isocyanate compounds. The ratio between the different isocyanates collected with dry samplers and impinger-filter samplers was in the range of 0.80-1.14 for RH = 20, 0.8-1.25 for RH = 40, 0.76-1.4 for RH = 60, and 0.72-3.7 for RH = 90. Taking into account experimental errors, it seems clear that isocyanic acid DBA derivatives are found at higher levels in the dry samples compared with impinger-filter samplers at elevated humidity. The dry sampling using DBA as the reagent enables easy and robust sampling without the need of field extraction.