RESUMO
The quantitative analysis of gas mixtures with gas chromatography is based on the calibration with certified standards and the determination of a response relationship for each species by regression analysis. The conventional assumption of a constant amount-of-substance injected onto the column, the sample size, for all standards analyzed represents one of the largest sources of uncertainies in this analysis technique. For systems using time-based microinjectors, the sample size injected onto the column is determined by the opening time of a pneumatically actuated microvalve and therefore depends upon the gas velocity developed in the injector's microchannel. For this reason the sample size is not necessarily constant for a given opening time, but strongly correlates - according to fluid mechanic theory - with the gas mixture's dynamic viscosity. Neglecting these sample size variations leads to errors in the analysis up to 30%, depending on the diversity of the standards' viscosities, especially in processes with strongly changing hydrogen contents. A mathematical correction exploiting the inverse proportionality of the sample size and the sample's dynamic viscosity normalizes the injection amounts and minimizes the influence of the sample variations on the calibration regression. The data analysis based on the corrected normalized calibration is no longer viscosity dependent and can be applied to gas mixtures of any composition.