Nitromethane Potential Hazards in Use D. S. JlcKITTRICK, Shell Deielopment Companj, Emeryville, Calif., R. J. IRVLUE, Coast Ilanufacturing arid Supplj Company, TreFarno, Livermore, Calif.. h h D I. BERGSTEINSSON, Shell Development Companj. Emeryvillc. Calif.
Detonation and Heat Tests
S VIEIT of the usefulness of nitronietliaiie in certain
laboratory operations, the authors have carried out anumber of tests to ascertain the conditions under which it might explode. I n view of the results, which indicate that precautions should be taken against subjecting nitromethane t o severe shock or to high temperatures and pressures, i t appeared advisable to call the matter to the attention of others who may be manufacturing or using this compound. The explosiveness of nitromethane has been suspected by various investigators, but no demonstration of it has been published heretofore. Although nitromethane cannot be exploded by impact (4), its sodium and ammonium salts are known to be very sensitive, exploding on slight jarring or moderate heating ( 2 ), and the polynitromethanes are listed in tests on explosives. On the other hand, nitroethane is reported not to detonate a t 96” C. with a No. 8 electric cap (4). Usually no special precautions are taken in the making and handling of the mononitroparaffins in the laboratory (3, 4) or commercially ( I O ) and their shipment comes under no special ban.
The tests applied were not those standard for t’he examiriation of explosives; rather, they were devised to show conditions under which nitromethane becomes explosive, in order to furnish some idea of its potential hazard in comparison with other materials of more or less similar chemical structure. The experiments, which are described below, show that nitromethane is only moderately sensitive to explosion by detonation or by heat and pressure. When it does explode, however, i t does so with a force bhat qualifies it t’o be called a very powerful explosive. The detonation tests on pure and diluted nitromethane and on ot’her materials for comparison were made at about 20” C. The sample, of about 7 ml. contained in a small vial, was placed, along with the detonator, in the cylindrical chamber (diameter 1.9 cm., 0.75 inch) of a steel block 12.5 cm. ( 5 inches) high by 6.9 cm. (2.75 inches) in diameter that rested on a steel plate and was covered with a 2.27-kg. (5-pound) weight. The height to which t’his weight was thrown by the explosion gave an approximate measure of the explosive power of the material tested. The results with a No. 6 fulminate cap are given in Table I. I t was also found that pure nitromethane could be exploded with the detonating fuses Cordeau, which is trinitrotoluene in a lead tube, and Primacord, which is pentaerythritol tetranitrate spun into textiles. Test,s on the response of nitromethane and other materials to heat and pressure were made on samples of about 15 mg., sealed in capillary tubes 5 cm. long and about 1.5 mm. in inside diameter with walls about 2.0 mm. thick, which were dropped into a hole drilled in a copper block heated to the required temperature nith a Bunsen burner. The results of these heat tests are assembled in Table 11.
Material The nitromethane was prepared by the method described in “Organic Syntheses” (1) and its properties are compared below with those given for nitromethane in the International Critical Tables: Material Used
Boiling point (760)
101.4-101.5° C . (uncorrected)
1 1354 2 1
1.38133 %.s 1.3821 2 ‘ 101. l o C. 101.90 C.
1 139 1; ~.
I. RESPONSE OF
Table I shows that, though nitromethane is much less sensitive to shock than 40 per cent gelatin dynamite, even a 10 per cent solution of it in methyl Cellosolve is more sensitive
A S D OTHER COMPOKTSDS T O S H O C K BY NO.
Distance from B u t t of C a p to Sample Inches
h Nitrobenzene, 100% Nitromethane, 100%
40% gelatin dynamite
0.5 0.625 5,125 5.25 0.a
Explodes Sone Explodes Explodes h-one
Nitromethane, 25y0 Nitromethane, 10% None
Methyl Cellosolve 75% Methyl Cellosolve: 90% Methyl Cellosolve, 100%
0 0 0
Explodes Explodes None
Lubricating oil extract, 20JOb
Coking distillate, 7 % C
0 0.195 0 0.125 0 0.125
Explodes None Explodes None Explodes None
h’itromethane Nitromethane: Nitromethane Nitromethane:
Coking distillate, 80% Lubricating oil extract 84% Lubricating oil extract: 90% Coking distillate, 90%
0 0 0 0
None Sone None &-one
20 16% 107 10%
Cap made by the California C a p Co. An aromatic furfural extract of a light lubricating oil. A highly aromatic cracked oil, 30% boiling in the gasoline range.
Force considerably greater t h a n 40% gelatin dynamite 3 out of 3 trials 3 out of 3 trials 3 out of 3 trial. 2 out of 3 trials 3 out of 3 trials
NOVEiMBER 15, 1938
As part A of Table I1 shows, nitromethane is not much more TABLE 11. RESPONSE OF NITROMETHASE AND OTHER COMPOCNDS sensitive to heat than nitrobenzene, a temperature of 410’ C. TO HEAT detonating both of them, the nitrobenzene after a short delay. ApproxiDilution of the two compounds with equal volumes of methyl mate TemueraCellosolve shows that the nitrobenzene mixture is less sensiture of Compound Explotive than the one containing nitromethane, requiring a temTested Diluent sion Remarks perature of 495’ C. instead of 450” C. A
Nitromethane, 100% Nitrobenzene, 100%
Immediate -4fter a few minutes 450 After a few minutes ZOO(@ Immediate
Sirrbmerhane. 7 3 % Sirromerhane, 50Cc Sirrobenzene, 5 0 5 Sirromethane. 205; Nitromethane, SO% Water, 100% Methyl alcohol lOOyo see-Butyl alcohbl, 100% Methyl cyanide, 100yo E t h y l cyanide, 100%
::: I ... ... ...
No explosion a t 500’ C. for 15 minutes
T o explosion a t 500’ C. for 15 minutes
than pure nitrobenzene. Methyl Cellosolve, though not explosive itself, forms explosive mixtures with even small proportions of nitromethane. Solutions of aromatic hydrocarbons in nitromethane are, as part C of Table I shows, explosive, though much less sensitive than pure nitromethane. Solutions of nitromethane in aromatic hydrocarbons, a t least up to 20 per cent concentration, appear to be safe (part D).
(1) Clarke, H. T., et al., editors, “Organic Syntheses,” Tol. 3, p. 83, New York, John Wiley &- Sons, 1923. (2) Colver, E. de W. S., “High Explosives,” p. 391, New York, D. Van Nostrand Co., 1918. (3) Ibid.. D. 392. (4j Hass’,k. B., Hodge, E. B., and Vanderbilt, €3. hl., IXD. ENQ. CHEM.,28, 339-44 (1936). (5) International Critical Tahles, Vol. I, p. 176, New York, McGraw-Hill Book Co., 1925. (6) Ibid., p. 276. ( 7 ) Ibid., 1’01. 111, p. 216. (8) Ibid., Vol. VII, p. 34. “Nitroglycerine and Nitro(9) Kaoum, P. (tr. by Symmes, E. M.), glycerine Explosives,” p. 139, Baltimore, Williams & Wilkins Co., 1928. (IO) Private communication. RECEIVED S o r e m b e r 27, 1937.
Iron Determination in Presence of Titanium Using Zinc Reduction EJIIL TRUOG
R. W.PEARSON, University of Fisconsin. Rladison, &is.
N T H E oxidimetric titration of iron, the use of amalgamated zinc for reduction of the iron is not only convenient, but, as indicated by Hillebrand and Lundell (a),leads t o more accurate results than some other reductants when such interfering substances as titanium, chromium, columbium, molybdenum, uraniuni, tungsten, vanadium, arsenic, nitrates, and organic matter are absent. The presence of considerable amounts of these interfering substances can usually be apprehended by the formation of a colored solution during reduction. Titanium, in particular, is easily recognized, since the presence of slightly more than 0.1 mg. gives a distinct violet color to the reduced solution. Vanadium, however, must be present t o the extent of 4 to 5 mg. before the color (lavender on leaving the reductor, owing t o diva.lent vanadium, but changing quickly to the green and blue tri- and tetravalent forms, respectively, on exposure t o the air) is easily observed. The usual smaller amounts (less than 1.0 mg.) are easily detected by applying the strychnine sulfate test to a fen- drops of the solution after permanganate titration. Vanadium is also reduced by the other common reductants including hydrogen sulfide, and when present in appreciable amounts necessitates a special procedure in the determination of iron. I n the presence of permanganate in acid solution, vanadium is oxidized to the pentavalent form, but aeration is not effective in promoting this complete transformation, or transformation to a definite stage so as to allow an accurate correction in the iron titration.
I n the analysis of highly siliceous rocks and clays, vanadium is seldom present in sufficient amounts to cause a serious error in the iron titration. Fortunately, of the substances interfering in the titration of iron after zinc reduction, only titanium is apt to be present in appreciable amounts in the ordinary run of soil and rock analysis; hence if the interference of titanium could be prevented, the desirable zinc reduction method could be used for the determination of iron in many additional cases. I n an attempt to accomplish this, Gooch and Sewton (1) developed a procedure in which bismuth oxide, cupric oxide, or cupric sulfate is added to reoxidize the titanium selectively. After filtration from the excess of oxidizing agent and reduced product formed, the iron solution is titrated with standard permanganate. Recently, Thornton and Roseman (S) attained the same result by simply bubbling air through the reduced solution for 10 t o 30 minutes, so as to reoxidize the titanium but leave the iron unaffected. The writers found t h a t this procedure gives good results when small amounts of titanium are present. but is somewhat slow with larger quantities.
Reoxidation of Titanium Prior to a knowledge of Thornton and Roseman’s work, the writers had found it possible t o reoxidize the titanium selectively by stirring or shaking the solution containing the iron and titanium in reduced form. After making a number