Calibration method for carbon dioxide sensors to investigate direct methanol fuel cell efficiency

Abstract

Methanol crossover is a process in direct methanol fuel cells which causes significant reduction of cell efficiency. Methanol permeates through the membrane electrode assembly and reacts at the cathode with oxygen to form carbon dioxide. This process is undesirable because it does not generate electric energy, but rather only increases heat production.Different procedures have been used for the investigation of this crossover. One method uses the detection of carbon dioxide in the exhaust gas of the cathode by means of a carbon dioxide sensor. This technique is inexpensive and enables real-time measurements but its disadvantage is the low accuracy.This paper demonstrates a simple method to generate gas mixtures for the calibration of the sensor in order to increase the accuracy. The advantages of this technique consist in the fact that only the existing devices of a direct methanol fuel cell test rig are needed and that the operator can adjust the carbon dioxide concentration for the calibration process. This is important for dealing with nonlinearities of the sensor. A detailed error analysis accompanies the experiments. At the end it is shown that the accuracy of the determined Faraday efficiency can be improved by using the presented calibration technique.