Determination of Diethylene Glycol Dinitrate and Nitroglycerin by Infrared Spectroscopy SHRAGA PINCHAS, Weizmann Institute of Science, Rehotmth, Israel HE accepted methods for the determination of nitroglycerin T (15) in cordite are not specific for that particular ester of nitric acid; the reagents n-ill act in the same way on every organic
accuracy of 3%, so that it can easily be used for the quantitative determination of the dinitrate. The “nitrate” absorption band, too, follows Beer’s law, as has been shown separately both for nitroglycerin and diethylene glycol dinitrate. The extinction per nitro group is about 80% higher for the dinitrate than for nitroglycerin, possibly t m a w c of the close proximity of the absorbing groups in the latter.
nitrate-e.g., the dinitrate of diethylene glycol (11,12). The isolation of the free polyvalent alcohols by reductive hydrolysis with alkali sulfide (4, 6, IS) and their identification through crystalline derivatives (8, 9, 14) are tedious and would lead to undesirable complications mThen mixtures of the two nitrates are present. Infrared analysis provides a satisfactory solution for the problem. 9 comparison of the two formulas CH24XO2 CHz-OKO2
I I
I
CH-ON02
CH2
CH,--ONOz
0
I
I
CHz
I
CH2-ONOz shows that while either will exhibit the same absorption bands characteristic of the C O N 0 2 group, only in the latter is the ether group (3-0-C present. According to Barnes and coworkers (1) and Williams ( l e ) , the ether group has a specific absorption at about 1130 em.-’ (10). Analysis of the spectra has, indeed, proved that both substances have a common absorption band between 1270 and 1280 crn.-’ [nitroglycerin 1273 (7.68/~),diethyleneglycol dinitrate 1278 cm.-l ( 7 . 3 8 ~ ) ~ corresponding to the absorption a t 1283 cm.-l of nitrocellulose], and that only the dinitrate shows an additional peak a t 1140 em. -1 ( 8 . 7 7 ~ ) . This absorption band can, therefore, serve to indicate the presence of the diethylene glycol derivative in mixtures which contain no other substances of the ether type. In the range of concentration investigated, the band obeys Beer’s law with an 1099
Figure 2.
1140 CM. -1
1113
Ether Absorption of Diethylene Glycol Dinitrate
In acetone solution, 0.083 gram per ml.
1231
If the quantity of the dinitrate in the mixture is known from its specific absorption a t 1140 cm.-I, the quantity of nitroglycerin can be calculated from the extinction coefficient a t 1270 to 1280 cm.-’ by difference. The error does not exceed 10%. Aromatic nitro compounds do not interfere with the abovedescribed method of determination, as their characteristic absorption bands lie a t 1360, 1520, and 1610 cm.-’ (2). The influence of some of the stabilizers, commonly used in the manufacture of propellants, is being studied, and the results of this work will be published. I t has been found that both the “ether” and the “nitrate” bands of diethylene glycol dinitrate are only slightly affected by the presence of as-diphenylurea and ethyl diphenylcarbamate, whereas diphenylamine has a somewhat more disturbing influence on the nitrate band of both the dinitrate and nitroglycerin. Figure 1 shows the nitrate absorption band in acetonic solutions of nitroglycerin and of the dinitrate, respectively; Figure 2 reproduces the ether absorption band of the diethylene glycol derivative.
1213 1295 1278 CM.-’
1. Nitrate Absorption
Figure.
Nitroglpoerin in acetone, 0.023 gram per -
,MATERIALS
ml. - - Diethylene glycol dinitrate in acetone, 0.016 gram per ml.
Diethylene glycol dinitrate was prepared according to Rinkenbach (11) and Rinkenbach and Aaronson (19). To 40.3gram of a mixture of concentrated sulfuric and nitric acid (nitric acid 50%1 201
ANALYTICAL CHEMISTRY
202
sulfuric acid 45%, water 5%, all by weight), 11 grams of freshly distilled diethylene glycol (boiling point 184-187" a t 12 mm.) are added slowly with vigorous stirring and coolin so that the temperature remains within the range of 5' to 10" Care has to be taken that no diethylene glycol adheres to the wails of the reaction vessel. The suspension so obtained is poured into an excess of icecold distilled water, whereupon the nitration product separates as a heavy oil. The acid layer is removed bv decantation and the product is transferred with the help of several portions of cold distilled water into a separating funnel where it is washed repeatedly with water. In order to obtain a neutral (and, therefore, stable) product, the washed oil is dissolved in methanol which contains a small amount of ammonia, and precipitated again by addition of water. The final product is dried in a desiccator over Drierite. Yield is 84%. Analysis of the product according to the method of Silberrad (15)showed a purityof 98.8%. Nitroglycerin was prepared according to Naoum (7), using distilled glycerol and working in a temperature range of 15' to 25" c. Portions of 12 grams were nitrated. The neutralization of the product was carried out as described above, which proved to be a quicker and more efficient method than that described by Naoum. The product analyzed by the method of Silberrad (15) gave 95.5%; but using titanous chloride (3, 6), a figure of 99.2% was obtained.
8
APP 4RATUS
A Perkin-Elmer infrared spectrometer Model 12 C was used; the solutions of the nitrates in aretone were measured in a cell of 0,025-mm. thickness. Absorption Absorption of diethylene glycol dinitrate ( D E G N ) G./Ml. At 1140 cm.'l 0.250 0.083 0.0964 Absorpt,ion coefficient At 1278 ern.-' 0,0482 0.012 .4bsorption coefficient
Optical Density
K
1.01 0.335 0,383 4.0
0.85 0.22
K 18.0
Ab
sorption of nitroglycerin ( N G ) Optical Density
G./LMI.
A t 1273 cm.-1 0.0572 0,0229 Absorption coe5cient Absorption of mixtures At 1140 cxn-1 DEGN 0.087 0.096 At 1278 cm.-' 0.0174
0.70 0,275
K 12.1 Found
Calcd.
NG 0.04 0.148
0.34 0 39
0.348 0.40
0,0086
0.46
0.42
ACKNOWLEDGMENT
Thanks are due to Eliahu Bograchov and Chaim Eiger for the preparation of the substances required. The present investigation was carried out under the auspices of the Scientific Department of the Israeli Ministry of Defence and is published with its permission. LITERATURE CITED
(1) Barnea, R. B.,and co-workers, "Infrared Spectroscopy," pp. 65 ff ,, New York, Reinhold Publishing Corp., 1944. (2) Ihid., pp. 93-6. (3) Becker. W. W.,IND.ENG.CHEM.,ANAL.ED., 5, 152 (1933). (4) Bloxam, Jahresher. Fortschr. Chem., 1883, 858. (ti) Fischer, H., 2.angew. Chem., A60,334 (1948). ( 6 ) Genevois, L., C;)im. a n d . , 29, 101 (1947).
(7) Naoum, Ph., Kitroglycerine and Nitroglycerine Explosives," p. 27, Baltimore, Williams & TTilkins Co., 1928. (8) Nef, Ann., 335, 284 (1904). (9) Orchin, M., J . Assoc. Ofic.A g r . Chemists, 26, 99-101 (1943). (10) Park, J. D., Sharrah, M. L., and Lacher, J. R., J . Am. Chem. SOC.,71, 2337 (1949) (11) Rinkenbach, W. H., I d . Eng. Chem., 19, 925 (1927). (12) Rinkenbach, W. H.. and Aaronson, H. A.. Ihid., 23, 160 (1931) (13) Schmidt, O., Z. angew. Chem., 62, 22 (1950). (14) Shriner and Fuson, "Systematic Identification of Organic Compounds," 3rd ed., p. 227, Kea Tork, John TTiley & Sons, 1948. (15) Silberrad, O., Phillips, H. .I., and Mertiman, H. J., J . SOC.Chem. IT&., 25, 628 (1906). (16) Williams, 1'. Z., Rev. Sci. Instruments. 19, 135 (1948). RECEIYED September 9, 1949.
Modified Method for Potassium Removal of Fluoride and Sulfate Ions by Chromatographic Ion Exchange JOHN A. DEAN' University of Alabama, University, Ala.
OST insoluble silicates may be completely decomposed by treatment with dilute hydrofluoric acid and sulfuric acid. However, the evaporation must be repeated once or twice t o remove all fluoride, which is rather tenaciously retained. Furthermore, in the resulting solution all the metals are present as sulfates, and sulfate must be removed if the solution is to be used for the determination of potassium by the perchlorate method. Several methods ( 7 ) have been proposed for the removal of sulfate, but all are troubleson~eand there is a great tendency toward the loss of potassium by coprecipitation during any gravimetric operation. The J. Lawrence Smith method ( 5 ) is extremely time-consuming for routine work, although it yields results beyond reproach. Of the alternative ways of eliminating fluoride following the decomposition of a silicate by hydrofluoric acid, the only successful method has been the volatilization of fluoride as hydrofluosilicic acid (8). A11 the preceding methods are lengthy. Because the difficulties center about the removal of either sulfate ions or residual traces of fluoride, the latter usually present as a complex fluoPresent address, Department of Chemistry, University of Tennessee, Knoxville, Tenn. 1
aluminate ion, the use of an acid-alumina anion exchange column appeared feasible. This type of column has successfully been used for the removal of fluoride from natural waters (4). Schwab and Dattler (3), and later Kubli (a),extensively studied the retention of anions on acidified alumina and constructed a comprehensive order for adsorption. Fluoride is strongly adsorbed, while sulfate exhibits somewhat less affinity; however, both are readily adsorbed on a perchloric acid-washed column. The process, as pointed out by Feigl (1), is probably an exchange adsorption taking place between a fluoride or sulfate anion, which is more firmly bound by alumina, and the perchlorate ion. Because the fluoride ions are even more readily exchanged than the sulfate ions, it is immaterial whether hydrofluoric acid is completely removed following the decomposition treatment. In fact, perchloric acid may be substituted for the sulfuric acid in the initial treatment of the sample, thus avoiding the danger of spattering inherent in the sulfuric-hydrofluoric acid method. CHEMICALS AND APPARATUS
Reagents. All acids were analytical reagent rade materials. Ethyl acetate, Eimer and Amend absolute graie. As early re-
