Oscillations, glow and ignition in carbon monoxide oxidation - I. Glow and ignition in a closed reaction vessel and the effect of added hydrogen

Abstract
Location of reproducible conditions for ignition, chemiluminescence (glow) and oscillatory behaviour in carbon monoxide oxidation is a classic unsolved problem. The present paper is primarily concerned with locating and distinguishing between the non-oscillatory instabilities (ignition and glow); subsequent parts will deal with oscillatory behaviour in closed and open systems and with the unifying mechanism. We use a clean, acid-washed, quartz vessel in the region of the second limit. Emphasis is placed on: (i) the behaviour of a range of compositions little studied hitherto and containing from 2000 to 200 x 10-6of added H2, intermediate between ‘wet’ mixtures (which are easy to study) and ‘dry’ mixtures (which are notoriously difficult to study); and (ii) continuous monitoring of reactions by mass spectrometry and continuous, instrumental records of light intensity, temperature and pressure. Ignitions are sudden and intense; reaction is complete and temperature rises are large. Chemiluminescence is more variable: it can be bright enough to be mistaken visually for ignition, but quantitative measurements show it to be less sudden and less intense; reaction is always incomplete and accompanying temperature rises are smaller, and may be imperceptible. Behaviour depends crucially upon the amount of added hydrogen. In ‘wet’ mixtures containing more than 2000 x 10-6H2, only slow reaction and ignition occur. Intermediate mixtures (2000 x 10-6to 200 x 10-6H2) display chemiluminescence also, but as the ‘dryness’ increases, the temperatures required for ignition rise so sharply that this mode becomes difficult to attain. In dry mixtures (200 x 10-6to 50 x 10-6or less), ignition is quite inaccessible. The boundaries established in classic work by Linnett and by Gordon relate to luminescence, not ignition. A novel, pulsed glow (distinguished from oscillatory chemiluminescence), induced by continued heating of the reaction vessel, is also reported here. Classical simple treatments of instabilities begin from chain-branching or thermal standpoints: they are often thought to be mathematically similar but physically different. Attention is drawn here to the sharp differences implicit in the treatments. Both are needed in the subsequent development.

This publication has 9 references indexed in Scilit: