Abstract
Nicotine is the main component of environmental tobacco smoke, and its presence in indoor air is widely used as a secondhand-smoke indicator. Environmental tobacco smoke is a major source of indoor air pollution, but sufficient investigation of the uncertainty of its measurement, which mirrors the reliability of nicotine measurement, has not been performed. We calculated the uncertainty of measurement of indoor air nicotine concentration at low, medium, and high concentrations of 11.3798, 10.1977, $98.3768{\mu}g/m^3$, respectively, and we employed the Guide to the Expression of Uncertainty in Measurements (GUM), proposed by the International Organization for Standardization (ISO). The factors considered in determining the uncertainty were uncertainty of the calibration curve (calibration curve and repeated measurements), desorption efficiency, extraction volume, and sampling airflow (accuracy and acceptable limits of flowmeter). The measurement uncertainty was highest at low concentrations; the expanded measurement uncertainty is $0.9435{\mu}g/m^3$ and is represented as a relative uncertainty of 63.38%. At medium and high (concentrations, the relative uncertainty was 13.1% and 9.1%, respectively. The uncertainty of the calibration curve was largest for low indoor nicotine concentrations. To increase reliability of measurement in assessing the effect of secondhand smoke, measures such as increasing the sample injection rate ($1{\mu}L$ or more), increasing sampling volume to increase collected nicotine, and using gas chromatography-mass spectrometry (GC/MS) or GC/MS/MS, which has a lower quantitation threshold, rather than gas chromatography with nitrogen phosphorous detector, should be considered.
