Carbon monoxide oxidation can show oscillatory behaviour, but, although oscillations in light emission were reported to exist in 1951, they have proved notoriously elusive and irreproducible. The present study is concerned with their nature (glow or explosion) and with the conditions for their occurrence in a closed quartz vessel. Kinetic mechanisms are mentioned here but a full discussion is deferred until part III, where an open system is used to generate true stationary states and to sustain oscillations indefinitely. Emphasis is placed on systematic procedures with the use of (i) continuous mass spectrometry for monitoring extents of reaction, (ii) instrumental recording of light intensity, reactant temperature and pressure, and (iii) the range of ‘interm ediate’ compositions containing 200 x 10
-6_
2000 x 10
-6
of added H
2
. In dry systems, proper conditioning of the reaction vessel allows long trains of oscillations to be generated wherever steady glow is possible. The oscillations are oscillatory luminescence, not multiple ignitions; very little CO is consumed at each excursion. If the vessel is treated with chlorpicrin, the stability and reproducibility are further enhanced. In intermediate systems, conditions for oscillation separate regions of non-oscillatory luminescence from those of ignition. In wet systems in this region (near the upper limit), ignition invades the glow region displacing both non-oscillatory and oscillatory luminescence.
Gold and iron supported on Y-type zeolite for carbon monoxide oxidation
