On the performance of proton-transfer-reaction mass spectrometry for breath-relevant gas matrices

The accuracy of quantitative volatile organic compound (VOC) detection by proton-transfer-reaction mass spectrometry (PTR-MS) is substantially enhanced if the instrument is calibrated. Although quantification of a compound is in principle possible by mathematical methods based on kinetic theory, the underlying picture can become complicated depending on the gas matrix, leading to error. A simple, reliable method to overcome this is to calibrate the instrument using standard gas mixtures containing VOCs at known concentrations, which enables the compound-dependent sensitivity of the instrument to be determined. A dynamic gas calibration unit was developed to generate variable but known quantities of selected standard compounds in a carrier gas of variable relative humidity (RH; up to 100% at 37 degrees C) and CO2 content (<= 10%(v)) to reflect the changing conditions of a breath-gas sample matrix. Besides individual compound sensitivities, calibration also yields the limits of detection and quantification of the experimental method. Extensive calibrations of PTR-MS with several breath-relevant compounds were made at varying RH and CO2. Gas matrix effects of several compounds were negligible when appropriate mass-dependent transmission correction and normalization to the primary ions (m/z 21 and m/z 37) were applied. Two compounds are discussed in particular, namely acetaldehyde, which interferes with a CO2-related background, and formaldehyde, which shows a nonlinear dependence on sample gas humidity.