Routine analysis of molecular hydrogen (H2) and carbon monoxide (CO) is achieved using a modified commercial Reduction Gas Analyser (RGA3, Trace Analytical, Inc., California, USA), which includes gas chromatography followed by the reduction of mercuric oxide. Mercury vapour from this reaction is detected by UV absorption. Air is drawn from the top of the sampling tower through ¼” tubing (Synflex 1300) at a flow rate of 200 cm3 min-1 using an air pump (Rietschle Thomas, UK) downstreamof the sample loop. To prevent particulate matter from entering the analytical instrumentation the sample line is fitted with a 40-micron filter at the top of the tower and a 2-micron filter at the base of the tower (TF-series, Swagelok). The air is passed through a moisturetrap (silica beads) before entering a 1 cm3 sample loop. Theair pump is turned off 20 s prior to sample injection to allow pressure equilibration of the sample loop. Thepressure and the temperature of the sample loop aremonitored throughout analysis using a pressure sensor(WIKA A10) and temperature sensor (NOVUS TxRail), respectively.

When the sample loop has equilibrated the content is injected onto a pre-column consisting of Unibeads 1S (Grace, mesh 60/80; 1/8 in. OD x 76 cm length) at 95°C to separate H2 and CO from any remaining contaminant water vapour (H2O), carbon dioxide (CO2) and hydrocarbons. After 1 min the sample is redirected to an analytical column containing Molecular Sieve 5A (Supelco, mesh 60/80; 1/8 in. OD x 76 cm length) also at 95°C to further separate H2 and CO. At this time the pre-column is back-flushed. H2 and CO elute from the analytical column at 0.5 mins and 2.2 min, respectively, at which time they pass through the mercuric oxide bed (HgO). Any mercury vapour (Hg) liberated during the reduction of the HgO at 265°C enters the detector and is measured using UV absorption (Reaction 1).

             X + HgO (solid) → XO + Hg (vapour)       Reaction 1

where X is H2 or CO.

Synthetic air (BTCA air 178, BOC) is used as a carrier gas with a flow of 17 cm3 min-1. Prior to entering the system, any contaminant H2 and CO is removed using SOFNOCAT 514 (Molecular Products Ltd). In addition the carrier gas passes through a catalytic converter to convert any H2 and CO to H2O and CO2, respectively. This is then removed using a molecular sieve trap before entering the GC columns.

A run time of 6 min allows eight air samples to be analysed every hour with a workingstandard analysed after every fourth run. This results in eight fully calibrated measurements of H2 and CO each hour. Detector response, sample loop pressure and temperature are all recorded using a six-channel analogue to digital acquisition system (Model 302, SRI Instruments, California, USA) and subsequent peak analysis is performed using Peak Simple software (SRI Instruments). Normalised peak heights (peak height / average of bracketing working standards) measured from samples are referenced to the instruments non-linear response function determined using calibrated primary reference gases on the MPI-2009 scale for H2 (Jordan and Steinberg, 2011) and NOAA2004 scale for CO. The range of concentrations in the primary reference gases is 375 ppb to 1182 ppb for H2 (5 cylinders) and 62 ppb to 503 ppb for CO (6 cylinders). Instrument repeatability is assessed by calculating the deviation in concentration of the working standard from the bracketing two working standards as the concentration of this is known as it is determined at the same time as the non-linear response function. Using this approach the repeatability of the system is typically better than ± 5 ppb for H2 and ± 2 ppb for CO. The accuracy of the system is assessed through the analysis of a target gas every 6 hours. The target gas is a cylinder of gas with accurately assigned concentrations of the target species determined at a central calibration laboratory (in this instance MPI). This target gas is introduced to the system as a sample and following analysis the deviation of H2 and CO concentration from the assigned concentrations are calculated. Estimated accuracies using this approach are routinely better than ± 5 ppb and ± 2 ppb for H2 and CO, respectively.

Routine data work-up occurs on a monthly basis using bespoke procedures written for Igor Pro.

Jordan, A, and Steinberg, B. 2011. Calibration of atmospheric hydrogen measurements. Atmos. Meas. Technol. 4(3), 509521