76 JONES CILYCERYL METHYL ETHER DINITRATE VII.-Glyoer~l Methyl Ether Dinitrate (a- Methylin Dinitrate .> By DAVID TREVOR JONES. DURING recent years considerable attention has been devoted to the study of the mono- and di-nitrates of glycerol and their chlorides and ethers. The interest in these substances has been stimulated by the technical possibilities which they appeared to offer as ingredients of non-freezing nitroglycerin blasting com-positions. Among the substances investigated have been the di-nitrate of monochlorohydrin (Kast Zeitsch. yes. Schiess- u. Sprenlystoffw. 1906 1 227) which has been more o r less extensively used in such explosives as gelatin astralit gelatin westfalit etc. The mono- and di-nitrates of glycerol have been very exhaustively studied by Will (Ber.1908 41 1107) who commenced the in-vestigation of these substances with the above-mentioned technical object in view. The dimethyl and diethyl ethers of glycerol mono-nitrate have been described by Paternb and Benelli (Gaazetta, 1909 39 ii 312) whilst Vender has described the dinitrates of monoacetin and monoformin (Zeitsch. yes. Schiess- u. S p e n g s t o f f u r . , 1907 2 21). Glyceryl methyl ether dinitrate which is here -described was prepared by the direct nitration of the a-monomethyl ether of Grun and Bockisch (Ber. 1908 41 3471). OMe.CH,*CH(OH)*CH,*OH + 2HN0 -+ OMe*C'H,~CH(NO,)*CH,*NO + 2H20. The product which was readily isolated was found t o solidify after being well supercooled and stirred. It is a powerful ex-plosive having about two-thirds the strength of nitroglycerin.It is however much less sensitive to shock although rather more readily exploded by heat. Its effect in lowering the freezing point of nitroglycerin is much the same as that of molecular concentra (a-METHYLIN DINITRATE .) 77 tions of monochlorohydrin dinitrate (Kast Zoc. c i t . ) and of ethyl nitrate (Nauckhoff Zeitsch. angew. Chem. 1905 18 21). All these values however are in complete disagreement with the value for the freezing-point constant of nitroglycerin as calculated from a carefully conducted determination of its latent heat of fusion (Hibbert and Fuller J . Amer. Chem. SOC. 1913 35 979). The subst,ance did not appear to exist in a second or labile form corresponding with the labile form of nitroglycerin.E x P E R I M E N T AL. Glyceryl a-monomethyl ether was prepared by Griin and Bockisch’s method (Zoc. cit.). The1 product distilled at l2Oo/ 18 mm. and the yield was 127 grams or from 200 grams of mane chlorohydrin 66 per cent. of the theoretical. The same yield was obtained on repeating the experiment. Glyceryl Methyl Ether Dinitrate. Sixty-three grams of glyceryl a-monomethyl ether were gradu-ally added to 480 grams of a mixture of nitric and sulphuric acids (HNO = 38.6 H,SO = 59.0 H,O = 2.4 per cent’.) which was cooled in icewater during the nitration. The initial temperature was 13O. During the operation the temperature was allowed to rise to 20° and was maintained a t that point until the end. The nitration proceeded quite smoothly and was easily controlled by regulating the addition of the glyceryl methyl ether.The time occupied by the nitration was from twenty to twenty-five minutes. The product was completely soluble in the mixed acid and the mixture; was slowly poured into 800 C.C. of water the temperature being allowed to rise to 40°. The bulk of the dinitrate separated, and after remaining for some little time the bulk of the aqueous layer was poured off and preserved for extraction. The residue containing the dinitrate was neutralised with a semi-saturated solution of sodium carbonate. The dinitrate was then run off from below the neutralised aqueous layer being added to the diluted acid which had been previously poured off. The dinitrate was then washed three times a t 50° with an equal bulk of 5 per cent.sodium carbonate solution then three times with water and was finally dried in a desiccator over calcium chloride. The yield was 75 grams or 64 per cent. of the theoretical. The neutralisd aqueous washings were extracted with ether and the ethereal solution was washed with 5 per cent. sodium carbonate solution, dried with calcium chloride filtered evaporated under diminishe 78 * pressure and preserved over calcium chloride in a desiccator. I n this way a further yield of 13.9 grams was obtained the total yield being thus 88.9 grams or 77 per cent. of the theoretical. The dry liquid constituting the first and major portion of the yield was analysed by the combustion method but on- account of its highly explosive nature1 the weighed,out substance was first converted into a weak dynamite by mixing with excess of pre-viously ignited kieselguhr the dynamite in turn being mixed with roughly powde'red copper oxide and introduced into t3he combus-tion tube.The nitrogen was estimated by the nitrmeter method using sulphuric acid as in the analysis of guncotton: 0.1397 gave 0.1248 COP and 0.0514 H20. C=24*36; H=4.03. 0.5492 , 132.3 C.C. NO a t 16O and 755 mm. N=14*15. 0.638 in 20.45 benzene gave At= - 0.833O. JONES GLYCERYL METHYL ETHER DINITRATE The combustion proceeded normally. M.W. = 188. C,H,O,N requires C= 24-28 ; H =4*08 ; N = 14.29 per cent. M.W. = 196. The subst.ance was therefure undoubtedly a-met-hylin dinitrate. GZyceryZ methyl ether dinitrate crystallises in white monoclinic prisms melting a t 24O.As first obtained it was a clear colour-less liquid which became pale yellow on keeping. It crystallised with difficulty and remained liquid even with occasional shaking, for more than two years in a magazine maintained a t 15-21O. It distilled a t 124O/18 mm. that is a t approximately the same temperature as the glyceryl methyl ether from which i t was derived, and some 22O lower than glyceryl dinitrate the corresponding alcohol. It is therefore more volatile than nitroglycerin and when tested a t looo on a watch-glass it) was found to volatilise a t from seven to eight times as rapidly. The liquid has D: 1.374 and ng 1.4478. It is soluble in benzene toluene acetic acid methyl and ethyl alcohols chloroform ether or acetone and insoluble in carbon disulphide or light petroleum.It gelat.inises nitro-cotton rapidly a t the ordinary temperature, and after warming it yields a gelatin softer and more plastic than that obtained from nitroglycerin. The chief interest of this substance lies in its explosive proper-tiee as compared with those of nitroglycerin. It has about t w e thirds the power of nitroglycerin although i t is much less sensitive to shock. Its comparative insensitiveness was demonstrated by submitting to the fall-hammer test' unfrozen dynamites each con-Both Eiubstances were placed under a Ateel disk and subjesbd to tha . tlaining three parts of explosive to one part of kieselguhr @-METHYLIN DINITRATE.) 79 impact of a weight of 1 kilogram falling from a measured height. The results are set forth in the following table : Dinitrate.Nitroglycerin. Height of fall. cm. Detonations. Failures. 100 2 8 95 1 9 90 1 9 85 0 10 Height of fall. cm. Detonations. Fdluree. 30 10 0 20 10 0 16 9 1 10 0 10 The solid substance was very insensitive. It did not explode even when scratched with the sharp edge of a thin melting-point tube. On the other hand the dinitrate proved to be more easily exploded when heated than did nitroglycerin. When heated in a glass test-tube in a metal-bath the temperature being raised a t the rate of 5* per minute it was observed to explode a t 182O the %ri-nitrate exploding at' 192O. Comparative power tlests of nitroglycerin and methylin dinitrate dynamites were made with the Trauzl lead block and mortar tests.I n the lead block the dinitrate dynamite gave an expansion of 22.9 c.c. a similar charge of nitroglycerin dynamite giving 30.0 C.C. I n the mortar test the relative powers indicated by the ballistic pendulum were 93.76 kilogram-metres (678 foot-lb.) for the dinitrate dynamite as compared with 124.43 kilogram-metres (900 foot-lb.) for a similar charge of nitroglycerin dynamite. Methylin dinitrate t-herefore would appear to have rather more than two-thirds the strength of nitroglycerin. I n order ta determine the lowering effect of the dinitrata on the freezing point of nitroglycerin a form of apparatus was adopted similar to that used by Kast (Zoc. cit.) for determining the melting points of the nitroglycerin isomerides and by Hibbert (Eie9hth Tnternutionnl Congress of Applied Chemistry 1912, About 5 C.C.of the mixture were inserted in a test-tube (15 x 1 an.) which was fitted into a slightly larger tube whereby the glyceryl nitrates were protected by an air-jacket from the too rapid action of the freezing mixture. The freezing agent con-sisted of ice where mixtures of higher melting point were con-cerned and of ice and salt for those of lower melting point. The thermometer was allowed to stand in the mixture direct contact between glass and glass being prevented by enclosing the lower portion of the thermometer bulb in a band of elastic. The stirrer consisted of a flexible piece of platinum wire which was attached t o a weighted string wound over a simple pulley and fastened a t IV 37) 80 JONES BLYOERYL METHYL ETHER DINITRATE.the further end to the outer edge of a wooden disk rotated by a motor. The liquid was first supercooled to the extent of aboutJ 4O or 5O. It was then inoculated with a small quantity of a frozen mixture of nitroglycerin wood-pulp and sodium nitrate and vigorously stirred. The maximum temperature was then carefully noted and taken as the freezing point. It will be seen that the values found for the molecular depression constant for nitroglycerin vary from 72.4 to 81.0 thus differing not very greatly from those obtained by Nauckhoff (Zoc. c i t . ) and by Kast (Zoc. cit.) with ethyl nitrate and monochlorohydrin dinitrate respectively These numbers are in fair agreement with the value 70.5 for the freez-ing-point constant for nitroglycerin obtained by Nauckhoff (Zoc.c i t . ) from a determination of its latent heat of fusion. Nauckhoff’s method was however admittedly defective and his r e d & differ very considerably from those of Hibbert and Fuller (Zoc. cit.) who found the latent heat of fusion ( h ) o’f nitroglycerin a t Oo to be 33.2 calories. Corrected to 13O the melting point of stable nitro-glycerin this would become 33*2+13*0 (c,-cz) where c1 and c2 art3 the specific heats of solid and liquid nitroglycerin respectively. Accepting Nauckhoff’s values of 0-356 and 0.315 for these the latent heat of fusion of nitroglycerin a t 13O would be 33.2+ 13.0 (0.356 - 0.315) = 34-33. Hence the freezing-point constant R P - 0,1991 x (2’73+ 13)2 = 48,5. lOOh 100 x 34.33 ~ -The results are set forth in the following table which includes Nauckhoff’s and East’s values derived from ethyl nitrate and chlorohydrin dinitrate : Molecular depression constant.Composition of - liquid. Calculated -’- from latent Nitro- Methylin Depression heat (Hib-glycerin. dinitrate. of freezing Calculated bert and Grams. Grams. point. (A) from A Fuller). 31-69 1.802 2-10 72.4 48-5 15-31 1.748 4.4 75.6 Y ? 8-22 1.729 8.7 81.0 9 9 Chloro-hydrin dinitrate. 21 2.1 4.4 88.4 $2 21 6.3 9.4 60.7 21 4.2 6.4-7.2 62.7-70’5 Ethyl nitrate. 74.1 76. THE INFLAMMATION OF MIXTURES OF ETHANE AND AIR. 81 Attempts to obtain a labile form of the substance analogous to that of nitroglycerin were made. The liquid was mixed with glass wool and supercooled to varying degrees with continual stirring with a glass rod. When some of the supercooled liquid which had not been previously frozen was inoculated with a trace of this product the solid obtained invariably crystallised a t 2 4 O . The author desires to express his indebtedness to Messrs. Nobel’s Explosive Co. and to Mr. W. Rintoul the manager of their Research Section for the facilities accorded to him for the carry-ing out and publication of this work. ARDEER. [Received November 7th 1918.