The Siemens Ultramat 6E analyser has a sample and reference measurement cell, each with a volume of 88 mL. The difference between the infrared (IR) absorption of the sample and reference air is measured and converted to a CO2 mole fraction based on the regular calibration procedures carried out that establish the analyser’s response function. The IR absorption of sample and reference are measured in quite a novel manner in the Siemens analyser; two sealed detector cells containing pure CO2 are located behind the sample and reference cells, and are connected to each other by a capillary tube with a microflow sensor. The CO2 in the detector cells absorbs all of the IR radiation that passes through the sample and reference cells. The difference in the transmitted IR due to the difference in CO2 mole fraction in the sample and reference cells causes the two detector cells to increase in temperature differentially. The temperature difference between the two detector cells results in a pressure difference that in turn causes a very small flow between the two detector cells through the capillary tubing connecting them. This flow is measured by the microflow sensor, and is proportional to the difference in CO2 mole fraction between the sample and reference cells.

O2 measurements are performed using an Oxzilla II (Sable Systems Inc.) analyser that employs fuel cell technology to measure O2 concentration. The fuel cells contain a weakly acidic electrolyte solution isolated from the air stream by a semi-permeable membrane. Air diffuses across the membrane, and within the fuel cell the following electrochemical reaction takes place: O2 + 4H+ + 2Pb → 2H2O + 2Pb2+

The lead atoms are supplied by the anode, the protons from the electrolytic solution and a gold cathode completes the cell. The resultant change in potential difference across the cell is measured and this is proportional to the rate at which O2 diffuses across the membrane and therefore also to the partial pressure of O2 in the air stream. Samples are always measured as a difference from the reference air being measured concurrently in the other cell. The sample and reference air streams are swapped to the alternate cell, via a 4-way valve every 60 seconds, with the first 30 seconds of measurement after each valve switch being discarded. The switching eliminates most of the drift that the individual fuel cells’ responses may experience on timescales longer then the switching interval, 60 seconds. This valve switching also has the beneficial advantage of doubling the sensitivity of the measurement.

Sampled air is dried in two stages; firstly liquid water is removed by a peristaltic pump as the air passes through two glass fridge traps at 1-2 °C, then the air is passed through two stainless steel chiller traps immersed in an ethanol bath at around -90 °C freezing out the remaining water and achieving a very low dew point.

The flows of sample and reference air are matched by ensuring that the pressure drops in both lines are identical using a Baratron 223B differential pressure gauge from MKS instruments.