JULY, 1906. Vol. XXXI., No. 364. THE ANALYST. A NEW METHOD OF ESTIMATING MOISTURE, WITH SPECIAL APPLICATION TO MOISTURE IN CORDITE AND OTHER SUBSTANCES CONTAINING VOLATILE MATTERS OTHER THAN WATER. BY P. V. DUPRE. (Read at the Meeting, A p d 4, 1906.) IN a paper on ammonium oxalate, which I had the honour to read before this Society in June last, a method of estimating moisture by means of calcium carbide was mentioned, and it was stated that experiments were in progress to apply the method more or less generally. These experiments having been now completed, I beg to lay the results before this Society. With this object animonium oxalate was chosen, as affording the most convenient and A series of experiments was first undertaken to standardize the process,THE ANALYST. 214 accurate method of weighing out a definite quantity of water, since the proportion it contains is accurately known (12.67 per cent.), and there is no tendency to loss by evaporation during weighing.Various weights of the salt, from 0.25 to 0.56 gram, equivalent to 0.03 to 0.07 gram of water, were taken, the quantity being weighed out in a tube of about 1 em. diameter and about 12 ems. long; the salt was then covered with a thin layer of sand,* and calcium carbide filled in to a depth of about 5 cms. The tube was then connected with a nitrometer containing a saturated solution of common salt, and the apparatus adjusted to zero with the tube brought to a known temperature by immersion in water. The tube was then immersed in a water-bath at a temperature of 100" C, to a depth of 7 or 8 cms., and the heating continued until no further evolution of gas could be observed.The tube was then brought back to the original temperature and the volume of acetylene measured, the temperature and pressure being noted. As the mean of ten experiments, in the last three of which the water was weighed out direct, it was found that 1 C.C. of acetylene at N.T.P. was equivalent to 0.001725 gram of water. The theoretical number would be 0.00162, but at the temperature of boiling water the hydrate of lime formed retains a small pro- portion of water above the one molecule. As, however, the amount of hydrate of lime formed is always proportional to the amount of acetylene evolved, the relation between the amount of acetylene produced and the excess of water retained is always the same.The following table gives the results obtained; it will be seen that the variations are small, and that the process is therefore of adequate accuracy : Weight of Amount of Oxygen taken. 0.2430 gram 0.3003 ,, 0.3160 ), 0,3235 ,, 0.4890 ,, 0.5020 ,, 0,5565 ,, Weight of Water. 0*0308 0.0380 0.0401 0.0410 0.0620 0,0636 0.0705 0.0651 0,0298 0-0627 - Amount found. I Error. I 0.0311 0-0378 0.0400 0.0413 0.0619 0.0634 0.0703 0.0649 0.0301 0.0625 + 0*0003 - 0.0002 - 0.0001 + 0*0003 - 0-0002 - 0.0002 - 0*0002 - 0.0002 + 0-0003 + 0.0002 The moisture in cordite was then estimated in exactly the same way, cordite, ground as for analysis, being substituted for the oxalate, and from 1 to 2.5 grams being taken. The total volatile matter in each sample was estimated at the same time by the official method.cause loss of acetylene a t ordinary temperatures. See paper on ammonium oxalate. The following are the results : * The sand is used to prevent contact betneen the oxalate and the calcium carbide, which mightTHE ANALYST. - _- \-.- 215 I - Nature of Sample. Cordite M.D. ... ... Cordite M.D.. ... Cordite returned from India First sample after softening with acetone for grinding.. . Cordite hlk. 1, recent, size 20 Cordite Mk. 1, size 20, old ... Cordita M.D., recent ... Cordite M.D., old ... ... .. ... Total Volatile Matter. Per Cent. 0.41 0.72 0.36 5.39 0.27 0.40 1-04 090 Moisture. Per Cent. 0.29 0.24 0.23 0.60 0.11 0.10 0.34 0.44 Acetone. Per Cent. 0.12 0.48 0.13 4.79 0.16 0.30 0.70 0.36 In order to test the accuracy of these results the following experiments were made : ( a ) Two samples were examined, each in duplicate, in order to see whether the results were concordant.( b ) A sample was taken and the moisture estimated as before ; the experiment was then repeated with another portion, to which a drop of acetone had been added after weighing, to find out the effect, if any, of the presence of this body. (c) A portion, previously dried, was weighed, exposed to moist air and again weighed, and the moisture so absorbed estimated as before. The results were as follows : Sample I. Sample 11. ... ... 0.148 per cent. ... ... ... 0.146 ,, First experiment Second experiment 0.114 per cent. 0.118 ,, Sample 111. Direct ... ... ... 0.416 per cent.+ acetone ... ... 0.419 ,, Sample IV. Water absorbed Water found . . . . ... 0.344 , j ... 0.345 per cent. - The method has also been applied to other substances, and found very useful in the case of a, mixture containing a volatile constituent, such as naphthalene or camphor.216 THE ANALYST. For instance, about 0.25 gram of naphthalene was taken and 0.0255 gram of water added ; the moisture estimated by the above method came to 0.026 gram. Again, in an explosive containing such ingredients as ammonium oxalate or magnesium sulphate, the total moisture can be rapidly estimated by this method, whereas it takes a long time-several days in many cases-for these salts to lose all their water of crystallization in 'uacuo, especially in such explosives as gelignite.The water present as moisture can readily be calculated when the proportion of the salt has been estimated. Experiments are now in progress to find out whether the method can be applied to volatile liquids, 0 10 20 30 40 50 60 70 80 90 100 110 120 TIME IN NIIINUTES. Experiments made in yet another direction with the view of distinguishing, if possible, between water not in combination and combined water have yielded results which may prove to have some useful application. These experiments were carried out exactly as described in the paper on ammonium oxalate, except that the salt was taken in a moist condition, a known weight of water having been added, and the curve, representing the rate of loss of water at constant temperature, obtained. The results show that the proportion of uncombined water can be separated very approximately from the combined water by means of a sharp change in the steep- ness of the curve, which occurs directlyall the uncombined water has been driven off.I t might be thought a t first sight that a salt when so heated would lose added water only until all this had been driven off, and that it would only then begin to lose combined water. From the few experiments made, however, this does not appear to be the case ; but, on the contrary, combined water commences to be lost at once, theTHE ANALYST. 217 loss continuing during the whole period taken for the uncombined water to the driven off. Also, this loss appears to take place at an average rate, during this period, of one-half the rate of loss of crystalline water after all uncombined water has gone.The chart on p. 216 gives the results obtained. The black lines represent the observed volumes of acetylene, the observations being taken every five minutes and each marked by a circle. The dotted lines represent the heights at which the breaks the curves should, theoretically, have occurred, calculated from the amounts of uncombined water taken and allowing for temperature and pressure, and also making the allowance above stated for the loss of crystalline water. The rate of loss taken is one-half that indicated by the curve after the break. Weight of Uncombined Weight calculated Water taken. from Curves. 0.0080 ,, 11. 0.0825 ... ... ... ... ... 0.0323 ,, 111. 0.0275 ... ... ... ... ... 0.0274 ,, IV.0.0325 ... ..I ... ... ..* 0.0325 CURVE I. 04080 ... ... ... . * r ... DISCUSSION. THE PRESIDENT (Mr. Bevan) having invited discussion, Mr. GUTTMANN said that the differentiation of the various volatile matters in cordite was not so unimportant as might at first be thought. Having regard to the important r6le that any volatilizable matter in a powder must play as far as ballistic results were concerned, it would be easily realized what an enormous digerence there was between water and acetone. When it was remembered that 4 per cent. of moisture might make an explosive absolutely unexplodable, while with & per cent. it might still be explosive, and having regard to the narrow margin allowed by official specifications and shooting tests, the great influence of even a per cent.of moisture would be readily understood. He remembered a paper by Mr. A. Marshall in the Journal of the Society of Chemical Industry, in which special means were described for condensing the nitro-glycerin, which would also be evaporated in the process. Mr. Dupri! had not alluded to this, but presumably, at the temperature to which the cordite was heated, a good deal of nitro-glycerin would come off, and one wondered what happened to it. If so, these results would need bringing into order by some further work, and if Mr. DuprA could succeed in differentiating between the nitro-glycerin and the acetone-either by applying Mr. Marshall’s process or in some other way-he would make his method complete. Mr. A. MARSHALL said that the first attempt to find a satisfactory method of determining the total volatile matter in cordite was made by a cousin of Mr.Dupr6 at Waltham Abbey. The details of the method were not altogether satisfactory, but further experiments which he (Mr. Marshall) had carried out at Waltham Abbey had resulted in its improvement. The method brought forward by Mr. Dupr6 advanced the question a stage further, as it enabled one to distinguish the water from the acetone or other volatile solvent used in manufacturing. Mr. DuprA’s results brought out the interesting point that in cordite of uniform composition the amount of water was almost constant. Dr. Robertson, at Waltham Abbey, had by an entirely different method obtained similar results. The question arose whether that was really the Was it included with the acetone?218 THE ANALYST, proportion of water in the cordite as used for firing, or whether it might not depend upon loss or gain of moisture from the atmosphere during the necessary process of grinding the cordite to powder.He had found that ground dried cordite exposed to the air absorbed just about the same amount of moisture in a few minutes, and as the preparation of the sample took at least several minutes, he thought that the constancy of the moisture might perhaps be accounted for in that way. Mr. DUPRE, in reply, said that no doubt a certain amount of nitro-glycerin was volatilized; but he had heated some carefully-dried cordite in the apparatus for ten or fifteen minutes without the evolution of any measurable amount of gas, and he had therefore concluded that whatever nitro-glycerin might be driven off was not decomposed in its passage through the carbide, but was condensed again probably at some part of the tube between the cordite and the nitrometer, and did not affect the volume of gas in the latter. The acetone distilled off would also be condensed, and a small allowance was made for it in the vapour tension. He might say that the method he had used was not intended in the least to supersede the official method of estimating the volatile matter, but it had struck him that it might be useful if the moisture and acetone could be separated. I t was certainly possible that Mr. Marshall’s explanation with regard to moisture might be correct, but in that caae one would have t o assume that 0.26 per cent. was the most that the cordite would ever take up, which from some of the results quoted does not appear to be the case.