An in-house developed method is presented for the purity analysis of nitrogen (N2) built-in purifier (BIPTM)) gas for the trace contaminant gases carbon dioxide (CO2), oxygen (O2)) and carbon monoxide (CO), using gas chromatography with a pulsed discharge helium ionisation detector (GC-PDHID). Nitrogen BIPTM gas is used as a “matrix” gas or diluent gas for the gravimetric preparation of binary reference materials of CO, CO2), sulphur dioxide (SO2)) and nitric oxide (NO) at the CSIR NML gas metrology laboratory. Purity analysis of nitrogen BIPTM is required to decrease the measurement uncertainty of the calculated gravimetric concentrations of the gaseous reference materials produced. The aim of the research was to find a method where amounts <0.25 x 10-6 mol•mol-1 of CO2), O2) and CO could be simultaneously analysed in high purity nitrogen within a short time, with minimum cost and on a routine basis. Gas mixtures of trace amounts of CO2), O2) and CO in N2) were separated and quantified using a parallel dual capillary column configuration with temperature and pressure programming and a pulsed discharge helium ionisation detector (PDHID). The detection limits were 9 x 10-9 mol•mol-1 for CO2), 7 x 10-9 mol•mol-1 for O2) and 37 x 10-9 mol•mol-1 for CO with repeatability precision of 1% for carbon dioxide, 1% for oxygen and 10% for carbon monoxide for a 0.2 x 10-6 mol•mol-1 standard. The detection limits obtained were lower than those reported previously by other investigators for similar methods and the validation for the method as set out in this investigation seems to be the first for trace amounts of CO2), O2) and CO in nitrogen. The method was validated by comparison of the CO2) and CO results with results obtained using a flame ionisation detector and methanisation. The technique of sequence reversal was used to improve the peak shape of CO but there was no improvement on the results obtained with temperature and pressure programming. Although no helium purging was used to reduce atmospheric contamination, it was shown that the main source of contamination from the air was through the sampling system which was reduced to a level of ± 20 x 10-9 mol•mol-1 oxygen simply by using a higher sample flow rate. It was also found that even when large amounts of CO2) were adsorbed onto the molecular sieve column, this made no difference to the column performance at trace levels. The method has also been validated for the analysis of nitrogen in high purity oxygen and may also be used to analyse carbon dioxide and carbon monoxide in oxygen as well.
Dissertation (MSc (Chemistry))--University of Pretoria, 2008.