SODEAU: THE DECOMPOSITION OF CHLORATES. PART 1V. 93'3 XCIX.-Thc Decomposition of Chlomtes. Pcwt I V. The Supposed Meclzanical FacilitatioiL of the De- composit'ion of Pot Q s s i z m CWomte. By WILLIAM H. SODEAU, B.Sc. VERY many substances are known to facilitate the decomposition of potassium chlorate, and it has from time to time been suggested that the whole or part of their action is analogous to that of sand, &c., in940 SODEAU: THE DECOMPOSITION OF CHLORATES. PART IV. promoting the boiling of water. The supposed analogy mould, how- ever, appear t o be false, it haviog been shown that the ease of decom- position is not appreciably altered by reducing the pressure to 1 mm. (Trans., 1900, 77, 144). It is dificult to understand why an irre- versible decomposition should be facilitated by the presence of solid particles if these are chemically inert.Were the action purely mechanical, it would presumably be common to all finely divided sub- stances, irrespective of their chemical nature, but several workers have stated that zinc oxide, magnesia,-&c., produce no facilitation. There would seem to be a possibility of chemical action in the case of all substances known to facilitate decomposition of the chlorate. For example, the oxides of manganese, iron, cobalt, nickel, and copper cause the oxygen to be very readily evolved, but in each case the exist- ence of an unstable higher oxide is known or indicated. Platinum black is said to have some facilitating action, n’ot, however, comparable with that of the oxides of manganese, &c., but at the temperatures at which this occurs its chemical activity towards oxygen would appear to be considerable.Mond, Ramsay, and Shields (Phi!. I\*cc?zs., 1895, [A], 186, 689; 1893, 199, 153) have shown that after heating a t 444” in an atmosphere of oxygen, platinum black contains about 66 times its volume of that gas, and that under similar conditions platinous oxide retains more than a quarter of its oxygen even after many hours’ heating. As reduction of pressure mas found to diminish these pro- portions, it would seem that the formation add decomposition of an oxide were probably proceeding simultaneously, hence platinum black cannot well be regarded as chemically inert. References to sand, &c., are discordant, possibly owing to variability of composition.Veley (Phil. Tv-am., 1888, [A], 179, 270) maint.ains that c L finely divided chemically inert particles ” produce acceleration of the de- composition of potassium chlorate, but publishes only one experiment t o establish this, and cites no evidence in support of his assumption that the added substance (barium sulphate) was chemically inert. The table referring to this experiment appears to indicate that the presence of 1 per cent. of barium sulphate increases the rate of de- composition to the average extent of 500 per cent,, but a statement in the text * shows that before systematic observa.tians were taken the average increase amounted only to about 100 per cent. Veley also shows that a similar proportion of manganese peroxide produces a far greater effect. In the series of experiments describcd in the present paper (see Table, p. 942), both range and rate of decomposition, and therefore the range of temperature, have been varied in either direction, as corn- * “ 104 c.c. of gas were given off from the chlorate containing the barium sulphate, but only 51.2 C.C. from that without the sulphate.”SODEAU: THE DECOMPOSITION OF CHLORATES. PART W. 941 pared with Veley's experiment, but the average increase of rate for the presence of 1 per cent, of barium sulphate amounts t o only 16 pel. cent. This is practically negligible, compared with the influence of many other substances, and seems fully explicable by the fact that double decomposition takes place (p. 943) with formation of a small amount of barium ohlorate, a substance less stable than potassium chlorate.With 1 per cent. of barium sulphate, the decomposition is very slow a t 450°, whilst with 1 per cent. of the peroxide of manganese which is precipitated from the acetate by bromine, the action becomes violent a t about 340'. The author is of opinion that the supposed ability of chemi- cally inert solid particles to facilitate the decomposition of potassium chlorate is unsupported by experimental evidence, and, if existing, is inadequate to explain even a small fraction of the great facilitation produced by the oxides of manganese, iron, cobalt, nickel, and copper, The action of the latter substances, now engaging his attention, would therefore appear to be entirely chemical. EXPERIMENTAL. Two specimens of barium sulphate mere prepared by precipitating solutions of recrystallised barium chloride with excess of " pure " sul- phuric acid, the products being very thoroughly washed by decanta- tion, and dried under reduced pressure, A temperature of 200" was employed in drying the first specimen, but that used in experiment 152 was not previously heated above 100'.Equal weights of carefully recrystallised potassium chlorate, with and without the addition of 1 per cent. of barium sulphate, mere de- composed in similar receptacles placed side by side in a bath of melted pewter, together with a thermometer. The evolved oxygen was collected over water in graduated tubes, readings being taken at intervals of 2 to 10 minutes, and the loss of weight was determined at the conclusion, in order to guard against the possibility of a leak.The table summarises the results of all such experiments, together with those of the experiment recorded by Veley (Zoc. cit.) ; the num- bers given in the latter case were calculated on the assumption that he measured the gas a t about 20° and 760 mm., but the main issue would not be affected even if the actual temperature and pressure were widely different from these. All numbers were calculated directly from the gas volumes, and the heating was always continuous, although the earlier and later stages are in most cases given separately. The weight of each portion of chlorate was 1 gram in experiment 143, 10 grams in experiment 144, and 5 grams in each of the subsequent experiments. Tubes closed a one end were used in942 SODEAU: THE DECOMPOSITION OF CIILORATES.PART IV 143 144 147 148 149 150 151 152 Vcley 's (ZOC. cit.) { 1<c'10, nlo11e. -____ Range of de. coniposition (total o= 100). 0-20.3 !0*3 -61.1 0- 3'44 3.44- 9.6 0- 1'39 1.39- 4'24 0- 1.07 1.07- 5.94 0- 3-66 0- 3-37 0- 0.35 0.35- 4.33 0- 0.34 0.34- 2-06 0- 0'32 0'32- 2-13 Percentage of total 0 per minute. 0.5s 0.66 0.041 0,034 0*0084 0 *030 0.0071 0.0325 0 '01 52 0.0087 { 0'0024 0'0306 0'0024 0-0107 0.0023 0-0113 0- 0'25 1 0.25- 0'41 i 0*0010 RC10, nlld 1 per cent. of IhSO,. Range of de- composition (total o= 100). 0-25.1 !5*1 -02 0 0- 4'08 4 '08-1 0'6 0- 5.56 5 -56-1 4 -5 0- 1-47 1.47- 8'21 0- 4'20 0- 3-27 0- 2.54 0- 0'54 0.54- 4'46 0- 0.40 0.40- 2.42 0- 0'45 0.45- 2-58 0- 0.51 0.51- 1.48 I Ratio of , rates.Percent age Mixture -__ of total 1 I(C10,' 0 prr minute. 0 72 0 '61 0.049 0.036 0'034 0.094 0 *0098 0.045 0'0175 0.0085 0,0066 0.0037 0'0301 0-0028 0-0126 0-0032 0.0133 '1 0,0116 1.23 0.92 1.18 1-05 (4'00) (3.19) 1-38 1.38 1.15 0-97 0 '75 1.5 0 -98 1-19 1-17 1-40 1-18 \ 2'03 ~ 6'06 experiment 143, but submerged bulbs were employed in the other experiments ; continual interchange must have eliminated from the average any possible difference due to individual bulb tubes, or to position in the bath.* Two bulbs containing the mixture were employed in experiment 150; both chlorate and mixture were duplicated in experiment 152. I n some experiments, potassium chlorate mixed with 5 per cent. of potassium sulphate was placed in an additional bulb tube; the potassium sulphate did not appear to have much influence upon the rate of decomposition, but this proportion dissolves in the fused chlorate.In experiment 147, the behaviour of the pure chlorate was normal, as was shown by the readings of the thermometer and of the volumes of gas evolved from a comparison mixture of chlorate and sulphate of potassium (5 per cent.), but the portion of chlorate mixed with 1 * Veley employed two retorts, each containing 70 grams of potassium chlornte, r t placed side by side in a small square air-bath" which was " packed with asbestos, to distribute the heat of the lamp as uiiiforinly as possible."STERN: THE NUTRITION OF YEAST. PART 111. 943 per cent. of barium sulphate gave results a t variance with every other experiment.It seems practically certain that a trace of some impurity must have been introduced, for the subsequent experiments would appear to have excluded every other explanation ; experiment 148 mas all but identical as regards known conditions, yet it exhibited no such abnormality. Omitting experiment 147, the average increase of rate for tlhe presence of 1 per cent. of barium sulphate amounted to 16 per cent. (ratio of rates = 1 :1*16), a result entirely a t variance with that obtained by Veley. Barium sulphate was found to completely dissolve in 500 times its weight of fused potassium chlorate contained in a test-tube placed in a bath of fused potassium nitrate a t about 480°, and the ‘‘ solution ” gave a thick precipitate on adding fused potassium chlorate containing potassium sulphate. It is evident that (as on adding barium chlor- ide, Trans., 1500, 77, 146) double decomposition takes place, and that in the present instance barium chlorate and potassium sulphate must be formed. The latter substance does not retard the decomposition of potassium chlorate, consequently the production of barium chlorate, a substance less stable than potassium chlorate, would seem to fully account for the trifling facilitation produced by barium sulphate. When 1 per cent. of barium sulphate is added, the amount rapidly transformed into chlorate must be about a fifth of the whole, EXPLOSIVES COMMITTEE’S LABORATORY, ROYAL ARSENAL, WOOLJVICH.