THE CHEMICAL NEWS,

VOL. CXXXII. No. 3438.

MERTON HOUSE, SALISBURY SQUARE, LONDON, E.C.4.

TELEPHONES :

Administrative: Central 6521.

Printing Works: Hop 2404.

REACTION REGIONS. I.

By W. P. JORissen.

When we mix three substances, A, B, and C, which can react with each other (viz., all three, or two by two, or two of them), there is a possibility that the reaction, started locally, propagates itself through the whole mass. This will happen if the quantity of heat developed by the reaction is sufficient to bring the adjacent parts of the mixture to the temperature necessary for the starting of the reaction.

When we gradually change the composition of the mixture we shall get, at last, mixtures in which the locally-started reaction does not propagate itself.

The points where the boundary curves of the reaction region reach the sides of the triangle (p and q, r and s, u and v) will indicate the limits of the reacting mixtures of two of the substances.

In the case represented by Fig. II, the reaction can propagate itself in a mixture of the substances C and B between the limits p and q.

Fig. III represents the case where also the substances C and A can react with each other, on condition that the composition of the mixtures varies between r and s.

Fig. IV. represents the case where the three substances can react with each other and also two by two.

AA

We also see the possibility that the cases represented by the Figs II, III, and IV change into each other by changing the temperature. The shape of the curves will further depend on the nature of the substances A, B, and C, and other circumstances. If the substance B is endothermic the reaction region can extend itself to the point B (Fig. V, as may easily be constances A and C.

The case of two endothermic substances (A and B) is represented by Fig VI.

The reaction region of three endothermic substances will fill the whole triangle when the reaction is so started that always each of the substances comes into reaction.

A. Let us consider now three gases.

a. Two of the substances can react under the circumstances, but the third one does not react under those circumstances.

Example: The explosion of mixtures of hydrogen, oxygen and nitrogen. We may expect a region as shown in Fig VII.

The lower and upper explosion limits of hydrogen in mixtures with oxygen are represented by the points of intersection of the curve with the line O,H,; addition of nitrogen will bring those points nearer and nearer to each other till they finally coincide. If we take air instead of oxygen and we consider the air as one substance (a less active oxygen) we can add carbon dioxide and will get a similar region (compare also J. K. Clement, Technical Paper 43, Bureau of Mines, 1913, in which are treated the explosion limits of mixtures of methane or natural gas with oxygen, carbon dioxide and nitrogen). A second example is given by the system phosphorus vapour-oxygennitrogen (or carbon dioxide, etc.). these reaction regions, only a few points have been determined; see W. P. Jorissen, Rec. trav. chim., 1920 XXXIX., 717; 1921, XL., 539.

Of

b. Two of the substances are able to react with the third one, apart or together. See Fig. VIII, in which the experiments with mixtures of hydrogen, ethylene and oxygen are represented. The explosion region is ADCB. (See W. P. Jorissen and B. L. Ongkiehong, Rec. trav. chim., 1926, XLV., 162). The small triangles represent Davy's observations (Phil. Trans., 1817, I., 59). If we take air instead of oxygen we get a much smaller region (upper limit : EF, the lower limit nearly coincides with AB). This region follows from A. G. White's observations (J. Chem. Soc., 1925, CXXVII., 48).

How case a may be changed into case b is shown by Fig. IX, where the explosion regions of chloroform, methylene chloride.

A. P. Langen van der Valk. See also: W. P. Jorissen and Langen van der Valk, Rec. trav. chim., 1925, XLIV., 813, 1045. The explosion region of CCI,-CO-air mixtures is very small (it is not drawn here); that of CHCI,-CO-air mixtures is somewhat larger, AFBA; that of CH,Cl2-COair, AEBA, approaches already the left side of the triangle. The explosion region CH,C-CO-air ACDBA reaches that side of the triangle; the lower explosion limit of CH,Cl-air mixtures is represented by C, the upper explosion limit by D.

Fig. IX shows that methylene chloride at ordinary temperature has already a disposition to give explosion limits with air and we may be sure that those will be found at a higher temperature provided that the substance does not decompose before that temperature is reached.

How Fig. II can change into Fig. III, when we replace air by oxygen, is shown by experiments with hydrogen-ammonia -air hydrogen-ammonia-oxygen mixtures. (W. P. Jorissen (W. P. Jorissen and B. L. Ongkiehong, Rec. trav. chim., 1926, XLV., 224), see Figs X and XI.

and

From these figures we may conclude that the addition of nitrogen to the oxygen has only a small influence on explosion limits of hydrogen. On the other hand the explosion limits of ammonia will approach each other rapidly, chiefly by a movement of the line of the upper limits.

By determining the explosion limits of ammonia in mixtures of oxygen and nitrogen we found that those limits coincide when the oxygen content is about 32 per cent. Therefore, if we were to determine the explosion region in hydrogen--ammonia -oxygen (+ nitrogen) mixtures with 32 per cent. of oxygen we should see that it is enclosed by a curve, which touches the O2(+N2)-NH, side of triangle.

3

When we determine the explosion region at a lower oxygen content, the curve will approach more and more closely the curve of Fig. X. If the oxygen content becomes less than 21 per cent., the explosion region will become smaller and smaller and at last disappear. When we mix the ammonia with a substance which is influenced in about the same way by the addition of nitrogen to the oxygen as ammonia itself, then we should expect that the curve which encloses the explosion region ammoniasecond substance-oxygen (+ nitrogen) will leave at a certain nitrogen content the O2 (N)-ammonia side of the triangle and also the O, (+N)-second substance side. The explosion region will then contract to a closed curve within the triangle. (See Fig. 1).

2

Thus it will be possible under certain circumstances (of temperature, etc.), to make explosive mixtures of two gases or vapours with air while each gas or vapour separately cannot give explosive mixtures with air under the same circumstances. In the second place we can turn our attention to vapours, such as methylene chloride, which with gases, such as carbon monoxide, and air, give an explosion region of the form drawn in Fig. IX, AEBA. In that

case the vapour does not give explosive mixtures with air alone. But certain mixtures of two such vapours and air will certainly be explosive (under the same circumstances). In the third place we should expect that mixtures of two gases or

29

An investigation, the results of which may be put in practice, is that of the explosion regions in mixtures of methane, air and an incombustible vapour. A number of observations has been made with the vapours of CCI, CHCI,, C,H,C, CC, C,H,Cl, and C,HCl,. (W. P. Jorissen and J. Velísek, Rec. trav. chim., 1924, XLIII., 80; W. P. Jorissen and J. C. Meuwissen, Rec. trav. chim., 1924, XLIII., 591; 1925, XLIV., 132). An extensive systematical investigation is now started in our laboratory in order to find the vapours (or gases) which give the smallest explosion regions.

Among the cases in which one of the gases is endothermic several are to be