Manganese exposures in Toronto during use of the gasoline additive, methylcyclopentadienyl manganese tricarbonyl

A year-long population-weighted study of personal exposures to particulate matter (PM2.5) was conducted in Toronto while the manganese-containing additive, methylcyclopentadienyl manganese tricarbonyl (MMT), was present in gasoline at an average level of 11.9 mg Mn/l, which was higher than the maximum of 8.3 mg Mn/l allowed in the U.S. In this study, 925 three-day personal samples of PM2.5 lair concentration of aerosol with an aerodynamic diameter of less than 2.5 mu m) were collected, along with a record of participants' occupations, personal habits, surroundings, and activities during sampling. Stationary samples of PM2.5 were collected indoors and outdoors at a subset of participants' homes over the same 3-day periods. Three-day samples of PM2.5 were also collected at fixed locations. Personal exposures to PM2.5, were highly influenced by exposure to tobacco smoke, and were poorly correlated with outdoor levels (Kendall's tau=0.13). The mean concentration of PM2.5 in homes (21 mu g/m(3)) was significantly higher than the mean outdoor level (15 mu g/m(3)). By contrast, the mean PM2.5 Mn concentration (air concentration of Mn in PM2.5) was higher outdoors (9.7 ng/m(3)) than indoors (5.5 ng/m(3)). Other than from tobacco smoke, there were no indications of significant indoor sources of PM2.5 Mn in homes. The most important predictor of exposure to PM2.5 was time spent in the subway, and a high level (428 ng/m(3)) of PM2.5 Mn was measured in the subway. The source of this Mn was hypothesized to be friction erosion of subway rails. Small, but statistically significant correlations were present between personal exposures to PM2.5 Mn and several traffic-related variables (time spent in transit, in a motor vehicle, near a roadway with traffic, and in a parking garage). However, in a stepwise regression that adjusted for weather and personal activities, time in a motor vehicle was the only traffic-related variable significantly associated with PM2.5 Mn, and it was only the 10th most important personal activity variable in the final model. Concentrations of PM2.5 Mn were higher at two fixed locations than outside of participants' homes, which were likely further from high traffic areas than the fixed sites. Likewise, outdoor and fixed site samples collected during periods that included weekend days contained lower air concentrations of Mn than samples collected during weekdays when traffic was heavier. On the other hand, the monthly average concentration of Mn in gasoline was negatively correlated with both outdoor and personal PM2.5 Mn, which suggests that traffic-related sources of Mn other than MMT may be present. After omitting participants with exposure to Mn from certain identifiable non-MMT sources (subway riders, metal workers and persons exposed to tobacco smoke), the average (median) personal exposure of the remaining 325 participants to PM2.5 Mn was reduced from 14 ng/m(3) ( 8.5 ng/m(3)) to 8.3 ng/m(3) (7.0 ng/m(3)). Potential sources of this residual Mn exposure include, in addition to MMT, naturally occurring Mn in the earth's crust, other occupational exposure, airborne release of Mn from industrial operations, and friction erosion of Mn from steel-containing products. Taken together, these facts ( elimination of participants with Mn exposure from known non-MMT sources reduced average exposures by 40%, the existence of multiple non MMT sources of the remaining Mn exposure, and the negative correlation between MMT usage and PM2.5 Mn) suggest that the preponderance of personal Mn exposure was from non-MMT sources.