The thermal decomposition of nitrous oxide.

Abstract

when the thermal decomposition nitrous oxide was first investigated reactions were usually thought of as belonging to simple integral orders. Over the range 100-700 mm. the nitrous oxide reaction proved to be more nearly of the second than of any other order, showing that the activation process was collisional (Hinshelwood and Burk 1924), and not one depending upon the absorption of radiation by isolated moleculess—a possibility at one time considered . For a first-order reaction the plot of the reciprocal of the half-reaction time against pressure is a line parallal to the pressure axis, while for a second-order reaction it is a line inclined to the axis and passing through the origin. In a series of investigations (Volmer and Kummerrow 1930; Vomer and Nagasako 1930; Musgrave and Hinshelwood 1932; Hunter 1934) it has come to light that the form of this curve for nitrous oxide is really rather complex, and may be divided into the following parts: ( a ) an initial steep increase, starting from the origin, which between 50 and 100 mm. becomes shallower and passes into ( b ) an almost linear curve containing up to several atmosphere when it gradually bends again passing into ( c ) an almost straight line of still smaller slope which continues up to 20-30 atm. when it bends round and becomes nearly parallel to the pressure axis, as for a reaction of the first order. Various interpretations have been suggested : (1) The curve results from the superposition of three "quasi-unimolecular" reactions, each of the second order at low pressures and of the first order at higher pressures (typical of reactions in which activations is by collision and is followed by transformation of isolated molecules). The three components represent different activation modes with different transformation proabilities of the activated molecules (Hinshelwood, Fletcher, Verhoek and Winkler 1934). (2) The curve is not really composite in form, but represents a single quasi-unimolecular reaction, the transformation probability of the molecules being a continuous function of the excesss energy they contain. This view was supported by Volmer who, however, was not aware of the existence of region ( c ) of the curve, and belived that the portion ( b ) became parallel to the axis above about 10 atm. He did not recognize the distinctness of portion ( a ) but plotted the reciprocal of the velocity constant against the reciprocal of the pressure so that points corresponding to low pressures were spread out into an indefinite sweep which masked the normal composite appearance of the curve. (3) The third interpretation has not been specifically formulated for the case of nitrous oxide, but is implied by Letort's treatment (Letort 1937) of the analogues example of acetaldehyde: it amounts to the viwe that the changes of slope shown by the curve are not so much due to changes from one integral reaction order to another as to the existence of fractional orders, such as the order 3/2 which arises in certain circumstances when the mechanism depends upon the intervention of atoms or radicles. Limited stretches of the nitrous oxide curve could be represented approximately by the equation of a reaction of the 3/2 order: and, although this would not apply over the whole range, nevertheless, if we assume that the curve is complex, we shoud not neglect the possibility that one of the components is of this type. The present paper contains new experimental data bearing upon these interpretations, and upon other problems presented by this interesting reaction. Experimental Details The reaction is 2N 2 O = 2N 2 + O 2 , with a very small proportion of 2N 2 O = 2NO + N 2 , each of which corresponds to a 50 %increase of pressure.