Determination of Ignition Temperatures of Combustible Liquids and Gases Modification of the Drop Method Apparatus G . S. SCOTT, G. W. JONES, AYD F. E. SCOTT, Bureau of.Mines, Pittsburgh, Pu. A stainless steel block, heated electrically and containing a 125-cc. Pyrex flask in the center, is substituted for a liquid bath for the determination of the minimum ignition temperatures of liquids and gases by the drop method. The gases are liquefied by
I
N'CONNECTIOS with its work for greater safety in mining and industry, the Bureau of Mines determines the explosive properties of various combustible gases, vapors, and liquids as they become commercially important. One phase of the investigation is the determination of the ignition temperatures of liquids. For this purpose, the standard A.S.T.M. test (1) was originally employed. In this test, one or more drops of the liquid to be examined were dropped onto the bottom of a 125-cc. Pyrex Erlenmeyer flask suspended in a bath of molten solder or lead heated by a gas burner. In the Bureau of Mines tests, however, antimony and later a mixture of sodium and potassium nitrates were used instead of the solder. To facilitate temperature control, the iron pot was wound with Nichrome wire and the electric heating circuit was operated by means of a temperature controller (20). In accordance with the observations of Mason and Wheeler (26),it has been the practice in this laboratory to sweep out the test flask with nitrogen after each trial. REASONS FOR MODIFICATION OF APPARATUS
One source of annoyance in the use of the liquid bath was the necessity for frequent changes from nitrate bath to lead bath, or vice versa. Neither bath seemed to possess the exact characteristics required for the tests. The possibility of cracking the test flask containing a combustible liquid was a constant source of danger. The molten lead oxidized readily at the surface, and oxidation was augmented by stirring. Time was required for the heat to pass through the silica-protecting tube t,o the thermocouple junction, and for this reason continuity of temperature between one bath and the other seemed doubtful. Rather than engage in experiments on the baths themselves, it was decided to attempt a modification of the apparatus.
means of dry ice or liquid nitrogen before testing. Lower minimum ignition temperatures are obtained in this apparatus than in the liquid baths, by a closer approach to black body conditions. Selected ignition temperatures for 172 compounds are given.
.
.
couple line from the 110-volt winding. The other ends of the thermopouple were connected to a two-pole, two-throw switch, h, so that the thermocouple circuit could be momentarily disconnected from controller m and connected to the potentiometer, i , for a temperature reading. The cold junction of the thermocouple was buried in an ice bath, g, contained in a vacuum bottle. Mirror f was sometimes used to observe the results of a test, especially where the flames have little color. . The block was designed for a temperature of 750" C. Since Pyrex test flasks are not suitable a t this temperature range, quartz flasks were used. In carrying out a determination, one or more drops of the liquid to be tested are dropped onto the bottom of the flask, and a stopwatch is started a t the same time. If the liquid flashes, the time is noted. This is the "lag" or lag on ignition. The temperature of the bath is then lowered a few degrees and the test repeated. In this manner, a temperature can usually be found, above which the liquid will ignite and below which it will not ignite. After such a temperature is found, the number of drops is varied to determine whether a still lower temperature may be found. Temperature readings were taken inside the test flasks in the steel bath and in the nitrate bath, The results are shown in Table I. The steel-block apparatus requires about 2.5 hours to heat from room temperature to 400' C. The bomb can be brought to
DESCRIPTION OF MODIFIED APPARATUS
In the modified apparatus (Figure 1) a stainlesssteel block, e, with a stainless-steel lid, k , was substituted for the nitrate and lead baths. The steel block weighs about 50 pounds. I t was wound with 27 feet of 18-gage Nichrome wire, b, encased in a sheath of asbestos tubing. The block was placed on a disk of 85% magnesia insulation, q, supported above the table on an iron disk, T . Between the metal cylinder, c, and the block, e, was placed thermal insulation, p , consisting of infusorial earth. A 0.375inch Transite ring, I , was placed on top of the lid, IC, as shown. Over the top of the whole was placed a ring, s, consisting of 0.25-inch asbestos board. A smaller ring, concentric with this, was attached t o the glass tube, a, through which a slow stream of air or oxygen passed into the test flask before a trial test was made. Chromel-Alumel thermocouple j passed through a cavity in the block, so that the hot junction was directly under the center of the flask. The 0.25-inch iron pipe was used to protect the thermo-
Figure 1. Ignition Temperature Apparatus
238
'
V O L U M E 20, N O , 3, M A R C H 1 9 4 8
239 within a few degrees of any desired temperature a t the start of the day's work by setting an elect.ric t,imeswitch the night hefore. Gases may b'e liquofied and tested in this a p p ~ L a ~ u:..a the same manner as ordinary liquids, except that they must be kept cool enough to remain in the liquid state. Propane and the butanes were condensed out in
ANALYTICAL CHEMISTRY
240
used. From the, data of Table I average flask wall temperatures were calculated, and a straight-line relationship between bath temperature and average wall temperature was assumed to hold down to room temperature. T o obtain an average wall temperature of 345" C., a nitrate bath temperature of 393" C. is required; however, a steel block temperature of 361" C. pro-
duces the same average wall temperature. The wall temperature, rather than the type of bath, seems to be the controlling factor. Wall temperatures in the steel block are higher for a given bath temperature because of the closer approach to black body conditions. Longer lags in the steel bath indicate that more of the wall surface is involved in the ignition process.
Table 111. Selected RIinimum Ignition Temperatures of Gases and Liquids I n Air Substance
C.
Formula
Acetal Acetaldehyde Acetic acid Acetic anhydride Acetone Acetylene Acrolein Acrylonitrile Allyl alcohol Allyl chloride Ammonia Amyl acetate, n Amyl acetate, is0 Amyl alcohol, n 4mgI alcohol, is0 .%my1 benzene hniyl chloride Amylene Aniline Anthracene Benzaldehyde Benzene Benzoic acid Benzyl acetate Benzyl alcohol Benzyl chloride Bromobenzene 1.3-Butadiene Butane Butyl acetate Butyl alcohol Butyl alcohol, is0 Butyl alcohol. aec Butyl alcohol, tert Butyl bromide Butyl Carbitol Butyl Cellosolve Butyl chloride Butylene Butyl formate Butyl propionate Butyraldehyde Butyric acid Carbon disulfide Carbon monoxide Cellosolve Cellosolve acetate Cetane 2-Chloro-%methylchloride Creosote Cresol, o Cresol, m Crotonaldehyde Cyanogen Cyclohexane Cyclopropane Cymene Decalin Decane, n Diamvl ether. ijo 1,2-Dlchloro-n-butane Dichloroethylene Dichloroethyl ether Diethanol amine Diethylene glycol Diethyl peroxide DImet hylamine Dimethyl aniline 2.3-Dimethvlbutane Dimethyl ether 2,3-Dimethylhexane Dioxane Dipropyl e t h e r Dipropyl ether, is0 Dirinyl ether Dodecane Ethane Ether, diethyl Ethyl acetate Ethyl alcohol Ethylbenzene Ethyl bromide 2-Ethyl butane E t h y l butyrate Ethyl chloride Ethylene Ethylene chlorohydii n Ethylene dichloride ~~
.
F.
230 275 550 ' 392 561 305 278 481 378 487 651 399 379 427 343
446 527 1022 738 1042 581 532 898 712 909 1204 750 714 80 1 650
269 273 530 472 192 580 573 460 428 627 688 418 408 42 1 345 434 414 478 483 228 244 460 443 322 426 230 552 120 609 238 379 235 343
498 524 986 882 378 1076 1064 860 802 1161 1270 784 766 790 653 813 777 89 2 902 442 47 1 860 830 612 799 446 1026 248 1128 460 714 455 650
356 599 626 232 850
673 1110 1159 450 1562
498 466 262 250 428 276 441 369 662 413 189 402 371 420 350 438 266 189 443 360 534 472 193 484 392 477 511
928 87 P 504 482 SO2 529 826 696 1224 776 372 756
463 516 490 425 413
866 961 914 797 776
...
'
...
...
700
788 662 820 511 372 830 680 993 882 379 903 738 891 952
Ref.
I n Oxygen F. Ref.
C. 174 159 490 361 485 296
345 318 914 682 905 565
460 348 404
860 658 759
...
... ...
...
332
630
255
49 1
...
...
.. ..
566 556
10S1 1033
...
... ...
...
373
...
...
335 283
...
328 364 377 460
... ... ... ...
308
...
..
704
..
635 542
..
622 687 711 860
.. ..
.. 586
...
206
403
107 588
225 1090
...
... ... ...
..
318
604
...
.. ..
...
... ,.. ...
296 454
...
... ... ...
250
...
569 849
.. 482
..,
... ... 346
6i5
298 252
i68 486
... ...
...
... ... ...
, . , . .
... ...
i8i
360
...
...
...
...
...
468
ii4
273 351 468 485 400
524 664 874 905 752
...
...
I n Air Substance Ethylene glycol Ethylene oxide Ethyl formate Ethyl mercaptan Ethyl propionate Furfural Furfurvl alcohol (.~ i .i - 1'' 0ine regular) (;mAine - 7 3 octane) (;tisoline (92 octane, Gasoline (100 octane) Glycerol Heptane Hexane Hexane, is0 Hexyl alcohol Hydrocyanic acid Hydrogen Hydrogen sulfide Isododecane Isophorone Isoprene Kerosene Methane Methyl acetate Methyl alcohol Methyl bromide 2-hlethylbutane (isopentane) Methyl butyl ketone Methyl Cellosolve Methyl chloride hlethyl cyclohexane Methyl cyclopentane Methylene chloride Methyl ethyl ketone 2-Methyl-3-ethylpentane Methyl formate a-Methylnaphthalene 2-Methylpentane 2-Methylpropane (isobutane) hlethyl propyl ketone Methyl salicylate Monomethylamine Saphtha Naphthalene Kicotine Nitrobenzene Konane Octane Ozite A Ozite B Paraffin Paraldehyde Pentane Petroleum ether Phenol Pinene Propane Pronvl acetate Pro'pi.1 acetate, is0 Prop11 alcohol Propyl alcohol, is0 Propyl bromide Propyl chloride Propyl cyclopentane Propylene ProDvlene dichloride .. Propylene glycol Pyridine Solvasoi Stearic acid Styrene Tetrahydrofurfuryl alcohol Toluene Toluidine. o Toluidine, p Trichloroethylene Triethylene glycol Trimethylbenzene 2,2,4-Trimethylpen. tane Trioxane Turpentine Yinyl acetate Xylene, o
Formula
.....
..... .....
C.
F.
413 429 577 299 476 391 391 280 299 390 429 393 230 248
776 804 1071 570 889 736 736 536 570 734 804 740 446 478
O
... . .
.....
CHI CIH609 CHrO CHaBr CaHn CaHn0 CaHsOz CHiCl C7HI4 CeHii CHzClz ClHaO CaHis
538 572 292 500 462 440 254 632 502 470 537 420 533 288 632
....
ClOH8 CioHiaN? CsHsKOz CgHza CsHi8
....
.... .... ....
C5HiZ
CXO CloHls
I000 1060 558 932 864 824 489 1170 936 878 999 788 992 $50 1170
.. .. .. .. .. ..
_ I n Oxygen_ C.
F.
... ...
...
261 440
502 824
364
687
O
...
,
.
...
...
...
... ...
...
... ... ...
320 214
608 417
284 300
543 572
560 220
1040 428
322
612
556
1033
...
...
...
... ...
...
461
...
294
...
...
...
285 329 606
... ,
..
...
...
... ... 862 ... 561
... ... ...
545 624 1123
642 514 461
1188 957 862
236 566
457 I051
... ...
...
462
864
275 319
527 606
...
CiHioO CsHsOa CHsN
..
Ref.
505 934 454 849 806 430 531 277 587 1089 244 471 482 900 545 285 218 424 435 815 451 844 245 473 242 468 554 290 624 329 715 1319
...
..
493 920 450 842 572 1062 439 822 456 893 490 914 920 968 285 545 458 1036 557 1035 42 1 790 482 900 276 529 395 743 914 490 540 282 552 1026 482 900 482 900 463 866 371 700
... ...
...
... , . .
...
...
400
257
560 235
1040 635
...
...
... .
.
I
208
...
... ... ... 258 ...
... ...
... . .
406
... ... ...
496
500 275 468 388 448 328
932 527 874 730 838 622
255
491
. .
...
...
...
...
... , . .
392
558
2.58
496
460 273
842 524
516
96 1
... . .
...
,..
...
... ...
786 471 948 542
...
... ... ...
434
813
...
419 144 509 283
424 252 427 496
795 486 801 925
... ... ...
...
Ref.
~
V O L U M E 20, NO. 3, M A R C H 1 9 4 8 Longer lags also allon- more time for some of the compounds to decompose. TABLE OF IGNITION TEMPERATURES
Table I11 gives the ignition temperatures employed by the Bureau of Mines for a number of compounds. Many have been determined by the bureau; the remainder were selected from the results of other investigators. A variety of methods vias, of course, used in the different determinations. LITERATURE CITED rZm. So?. Testing Materials, Standards, Pt. 111, pp. 90-1 (B.S.T.M D286-30), 1942. Baron, J., and Laffitte, P., Compt. rend., 206, 1386 (1938). Runte. K.. and Block, A,, Gas- und Wasserfach, 78, 325 (1935). em. SOC.,1934, 1382. ., Dunkle, C. G., and Hess, B. E., , Bur. Mines, Carnegie Inst. Technol., Mining Met. Advisory Boards, Bull. 30 (1926). Dixon, H. B., Rec. trav. chim., 44.305 (1925). Dixon, H. B., and Coward, H. F., J . Chcm. SOC..9 5 , 5 1 4 (1909). Egerton, A., and Gates, S. F., J. Inst. Petroleum Tech., 13, 244 (1927). Factory Mutusls, I n d . Eng. Chem., 32,880 (1940). Freitag, Current Met. Abstracts, Metals & AZZoys, 3, 85 (March 1932). Goldman, F ,Z.physik. Chem., 5,316 (1929). Grebel, A., Mem. et compt. rend. trav. SOC. ing. civils France, 83, 35 (1930). Holm, H., 2.angm. Chem., 26,273 (1913). Jones, G. W., and Beattie, B. B., I d . Eng. Chem., 26,557 (1934). Jones, G. W., Kennedy, R. E., and Miller, W. E.,'U. 8. Bur. Mines, Rept. Investigations 3648 (1942). Jones, G. W., Kennedy, R. E., and Scott, G. S.,Ibid., 3597 (1941). (17) Jones, G. W., et al., U. S . Bur. Mines, unpublished data.
241
(18) Jones, G . W., and Miller. W.E., U. S.Bur. Mines. Rept. Investigations 3567 (1941). (19) Jones, G . W., Miller, W. E.. and Seaman, H., I n d . Eng. Chem., 25. 694 (1933).
done;, G.
k..and Scott, G. S., U. S. Bur. Mines, Rept. Investi-
gations 3881 (1946).
Jones, G . W., Scott, G. S., and Miller, TI'. E., Ibid.. 3640 (1942). Jones, G. W.,Seaman, H., and Kennedy, R. E.. I n d . Eng. Chem., 25, 1283 (1933).
Jones, G. IT., Yant, W,P., Miller, W.E., and Kenned>-,R. E., U. S.Bur. Mines, Rept. Inoestigations 3284 (1935). Laffitte, P., and Baret. G., Bull. soc. chim., ( 4 ) 51, 581 (1932). Lewis, J. S., J . Chem. Soc., 1930,2241. Mason, W., and Wheeler, R. V., Ibid., ,121. 2079 (1922). Masson. H . J., and Hamilton, W. F.,Ind. Eng. Chem., 20, 813 (1928).
Moore, H., J . Inst. Petroleum Tech., 6 , 186 (1919). Moore, H.. J . SOC.Chem. In,d., 36, 109 (1917). Naylor, C. A,, and Wheeler, R. V.,J . Chem. SOC.,1931,12456. Nuckolls, A. H., "Underwriters' Laboratories Method for Classi-
fication of Hazards of Liauids" (March 1929). Nuckolls, A . H., Underwriters' Laboratories, Report S-1528 (1919'1. Ibid., 2375 (1933). Ormandy, W. It., and Craven, E. C., J . Inst. Pptroleum Ttch., 12. 650 (1926). Sortman, C. W..Beatty. €1. A., and Heron, S. D., I n d . Eng. Chem., 33,357 (1941). (36) Tanaka, Y . , and Nagai, Y . , Proc. I m p . Acad. ( T o k y o ) , 2, 219 (1926). (37) Thompson. K.J.. I n d . Eng. Chem., 21, :34 (1929). (38) White, A. G., and Price, T. W., J . Chem. Soc.. 115, 1462 (1919). (39) Wollers, G., and Emcke, V., K r u p p . Monatsh., 2, 1 (1921). (40) Zerbe, C., and Eckert, F., A ngeiu. Chem., 45, 593 (1932). RECEIVED -4pril 28, 1947.
Presented before the Division of Gas and Fuel Chemistry a t the 111th Meeting of the AMERICANCHEMICAL SOCIETY, Atlantic City, ?;. J. Published by permission of the Director, Bureau of >lines, U. S. Department of the Interior.
Determination of the Gamma Isomer of Hexachlorocyclohexane J. B. LA CLAIR, California State Department of Agriculture, Bureau of Chemistry, Sacramento, Calif. A method is described for determination of the gamma isomer of hexachlorocyclohexane in the technical product and dust mixtures. The method is based on the dehydrochlorination of two 0.100-gram samples, dissolved in 50.0 ml. of 95% ethyl alcohol, one for 15 minutes and the other for 50 minutes, at 0" C. with 10.0 ml. of 1 N ethanolic potassium hydroxide. The per cent chloride difference multiplied by a factor yields the percentage of gamma isomer.
T
HE insecticidal value of hexachlorocyclohexane is believed t o be due t o the presence of the gamma isomer of hexachlorocyclohexane (?'), and in the enforcement of the Agricultural Code of the State of California, pertaining t o the labeling and sale of economic poisons, it is necessary to determine the gamma isomer in commercial dust mixtures offered for sale as insecticides. A search of the literature and written inquiry disclosed that the only workable methods developed at that time were based on infrared absorption (8) or bioassay using insects. The infrared absorption method, though accurate under certain conditions, demands the use of the relatively expensive infrared spectrophotometer, and the bioassay method requires the skill of a highly trained entomologist. EXPERIWENTAL
Hexachlorocyclohexane is dehydrochlorinated by alcoholic sodium or potassium hydroxide, losing three atoms each of hydrogen and chlorine and forming trichlorobenzene ( 2 , 9).
The individual isomers have different rates of reaction to dehydrochlorination (d), depending on time, temperature, and concentration of reactants. The reaction goes rapidly to completion a t elevated temperatures. Samples of highly refined alpha, beta, gamma, delta, and epsilon isomers of hexachlorocyclohexane were obtained and their melting points determined as shown in Table I. The reaction rates of the individual isomers mere determined a t various temperatures and concentrations. Reaction rate curves under the best conditions found were plotted (Figure 1). From these data it vias deduced that the alpha and delta isomers should be almost completely dehydrochlorinated before the gamma isomer was appreciably dehydrochlorinated, then using a longer period of time the alpha, delta, and gamma isomers should all be dehydrochlorinated. Since the beta isomer is inert under these conditions, and the epsilon isomer content in the technical product is low enough not t o interfere seriously, the difference in chloride produced in a definite short and long period of dehydrochlorination should be a function of the gamma isomer content.
