54 RONALDS ON THE MOST VOLATILE CONSTITUENTS XI.-On the most Volatile Constituents of American Petroleum. By EDMUND Ph.D. RONALDS [From the Transactions of the Royal Society OP Edinburgh.] CRUDEAmerican petroleiim evolves at ordinary temperatures a quantity of gas which takes fire on contact with flame and when mixed in certain proportions with air produces an explosive mixture. It is in consequence of this property that it has been thought necessary to pass a very stringent law known as the Petroleum Bill with a view of preventing accidelits from thc incautious storing and handling of the oil. The more volatile liquid products obtained by distilling the crude oil are still more highly charged with combustible vapour which when these liquids are again distilled escapes cordensation even by the most powerful freezing mixtures.The liquid constituents of petroleum have now been carefully studied by Messrs. Pelouze and Cahours and some of them also by Mr. Schorlemmer. These eminent chemists have shown that the oil consists essentially of a mixture of the homologues of marsh gas having the general formula It was during the collection of the more volatile of this series of compounds with a view to their analysiu in which object I have now been forestalled that my attention was drawn to the large quantities of incondensable gas which escaped at each successive fractionation arid it appeared desirable to ascertain whether the gaseous ingredients of the oil belonged also to the same series or mere accompanied by other hydrocarbons.With this object in view and still waiting the arrival of some specimens of oil col- lected and secured in hermetically sealed vessels direct from the oil-wells I was euabled by the kind permission of Mr. Shand of Stirling to collect the gns which floated over the surface of the crude oil in the barrels in which it is imported into this country. 1 also obtained from the same manufacturer some of the very first proifucta of the stills employed in refining the petroleum on a ilianufact uri ng scale. The gas floating over the surface of Pennsylvanian oil was col- lccted at a temperature of -1" C. aud was observed to contain OF AMERICAN PETROLEUM. combustible ingredients. It took fire instantly on being brought into contact with flame burning with a very failit bluish light but without explosion.From Canadian petroleum which is of much thicker consistence no combustible gas was obtained at that temperature. The gas was collected over water by simply removing the ori-ginal wooden bung of the casks and inserting immediately a ccrk bung furnished with a tube for the delirery of the gas and a long shaiiked funnel tube through which liquid petroleum was poured. Thus obtained the gas was of course a mixture of air and liy-drocarbon; it was not affected by fuming oil of vitriol nor mils bromine-water discoloured by it It was hence inferred that no perceptible quantities of the olefiant series were present and the temperature of collection is suflicient guarantee fcr the absence of any known members of the benzole series.The qas was treated over mercnry with solid potash and pyro-gallate of potash successively when it yielded- 1.27 per cent. of carbonic acid and 6.58 , , oxygen. The residue analysed eudiometrically gave the following re-€111ts :-Gas collected from the surface of Pennsylvanian Petroleum at a temperature of -1" C.,jt.eed from Carbonic Acid by Potash and from Oxygen by Pyrogadlic Acid. m. pressure. I- I-1-1- GaS .. .. .. Do. + air.. * .. Do. + do. -I-oxygen .. Afterexplosion .. .. Aiter absorption of CU .. After addition of hydrogenAfter explosion .. . . . . .. .. .. . . .. 133.1 392.8 465 .6 421.3 383.4 474 3 346.5 0.3099 0-5666 0.6391 0.6940 0.5515 0 6395 0 ,5062 39.934 215 47 288.92 245 09 205.23 299 .I5 152*86 Deducting the nitrogen or 23.4 vols.= 54 per cent. of the original gas we have here a relation of hydrocarbon to condcn-sation and carbonic acid as- 16.534 43.83 39.86 or 100 265 241. 56 RONALDS ON THE MOST YOLATILE CONSTITUENTS The oxygen consumed amounts to 67.16 vols. or 4.06 times the volume of the hydrocarbon. The members of the olefiant and benzole series being absent it may fairly be inferred that the hydrocarbon resembles in constitution the liquids with which it is associated ; and if this be the case the gas must be a mixture of the hydrides of ethyl and propyl the former of which requires a relation of hydrocarbon to condensation and carbonic acid as-1 2.5 2 while the hydride of propyl requires a relation of 1 3 3.By calculation from the numbers above it can be shown that the gas analysed must have consisted of a mixture of these gases in nearly equal proportions or of- C,H6 hydride of ethyl 7-94 C,H, hydride of propyl 8.01 -the correctness of which is confirmed by the amount of oxygen consumed being about the mean of the quantities required for the combustion of these hydrides separately. Hydride of ethyl requires 3.5 times its volume of oxygen. Hydride of propyl requires 5 times its volume of oxygen. The ~RSfloating over the surface of the petroleum is tlierefore composed of-Carbonic acid .. .. l*27 Oxygen .. .. 6.58 Nitrogen .. {:::68)Hydrocarbon .. 54. 38.15 In this condition the gas is not explosive and would only be-come so on being mixed with a large volume of air.The most volatile liquid obtained by collecting the very first runnings from the stills employed in the process of refining pe-troleum has a specific gravity of 0.666. It is not seiisibly affected by nitric acid by oil of vitriol or by bromine. When distilled it begins to give off bubbles of gas in abundance at about 25" C. but after a few minutes all appearance of boiling ceases although large quantities of gas and condensable liquid continue to pass over up to 65' or 70' C. and the whole liquid is evapo- rated below 100" cent. OF AMERICAN PETROLEUM. This liquid resembles very closely the '' kerosolene" or "keroso-form" which an American physician of New Y ork has introduced as an anzesthetic agent ; and I am indebted to Dr.Simpson for the opportunity of comparing it with a specimen of the latter. The specimen lent me by Dr. Simpson was quite indifferent to the above reagents It had a specific gravity of -6336. It began to boil at 28O C. and was almost completely volatilised at 70" C. so that it must have been composed almost exclusively of a mixture of the Eydrides of amyl and hexyl while the crude volatile product from the manufactory contained in addition to these hydrides some incondensable gaseous products and a con-siderable quantity of the hydride of heptyl. The incondensable gases dissolved in this most volatile liquid were expelled by gently warming a large quantity (about two gallons) of-liquid and passing the gases before collecting them over water through a long metallic worm surrounded by a freez-ing mixture composed of ice and salt the whole apparatus having been filled previously with carbonic acid to expel air.The first two portions which were collected showed after sepa- rating carbonic acid and oxygen little difference in compositicn from that already analysed and which had been collected from the surface of the crude oil. I omit the details of the analyses of these two and submit only the results which correspond in both cases with a mixture of the hydrides of ethyl and propyl. Gas. Condensation. Carbonic Acid. 23.947I.{y;; . 2-77 Oxygen consumed .. 19*04.5 242 32.338 11.{7.y;; * 20.70 280 17.586 240 Oxygen consumed ..31-07 The gas coming over a little later from the same liquid was found to approach nearer in composition to pure hydride of propyI :1s is shown by the following analysis. This portion was treated with potash before being introduced into the eudiometer but the oxygen which it contained was not separated before combustion but was estimated in a separate experiment and found to amount to 2.44 per cent. of the gas burned. RONALDS ON THE i\lOS'r VOLATILE CONS'I'ITUENTS Corrected 1-01. Observed at 0" and 1 m. Volume. pressure. Gas .. .. .. .. 39 -723 0.2817 15 *1 10.604 After addition of oxygen.. .. 160 -0 *3939 16 -59.548 After addition of air .. .. 260 -125 0 4917 14 -5 121 -46 After explosion .... .. 236 386 0.4660 16 -5 104 .33 After absorption .. .. .. 204.386 0 451 75. 87 -276 After admission of hydrogen .. 357.161 0.603 14-204 *53 After explosion .. .. .. 231.225 0 *4613 13*6 102 +29 Deducting the nitrogen and the 2.44 per cent. of oxygen con-tained in the gas we have here the ratio of hydrocarbon to con-densation and carbonic acid as 5.984 17-13 16.954 -100 286 283 Hydride of propyl C,H = 2 vols. requires a ratio of 1 3 3. The quantity of oxygen consumed by the hydrocarbon is 4.67 times its volume while pure hydride of propyl would require 5 times its volume. The gas collected at a still later period from the same liquid was free from carbonic acid oxygen and nitrogen gases arid agreed in composition with a mixture of the hydrides of propyl and butyvl.Observed Pressure. Tern p. at 0" and 1 m. Volume. Cent. pressure. -I--Gtas .. .. .. 43 *034 0 -2821 19 *5 11.335 After addition of oxygen.. 151 -465 0*3837 19.9 54 .454 After addition of air .. 417 * 0 .6439 20 -6 249 -70 After explosion .. . . 372 644 0 .6038 16 -7 212.05 After absorption . . .. 321 * 0.566 15 .2 172.11 After addition-of hydrogen 405 * 0.649 17 -247 -45 After explosion .. .. 353.032 0 *5846 15 *2 195 52 The relation here of hydrocarbon to condensation and carbonic acid is as-11.335 37.65 39.94 100 332 352 The oxygen consumed is 5-88 times the volume of gas burned OF AMERICAN PE'I'ROLEUM. while hydride of butyl alone requires 6.5 times its volume of oxygen for combustion.The gas evolved on warming the light spirit of petroleum as it is prepared for sale after having been kept however for some months in a vessel not hermetically sealed mas found to be a mix- ture of nitrogen and oxygen with nearly pure hydride of butyl. After separating by potash the carbonic acid which had been allowed to occupy the space above the liquid the gas was analysed ; the oxygen which it contained was estimated by pyrogallate of potash in a separate experiment arid amounted to 15-37 per cent. Observed Volume. Presaure. pressure. I__--- Gas . . .. .. .. 73 *2 0 -2399 7. 17 *25 After addition of air .. .. 273 -3 0.4360 5 .2 117.11 After addition of oxygen..Alter explosion .. .. After absorption .. .. After admission of hydrogenAfter explosion ... .. .. .. .. .. 334. 288 *5 228 -5 330 317 * 0 4976 0 ,4523 0,3995 0,5022 0 -4799 5 .7 6.4 10.2 13 * 12 .2 162 82 127.5 158 .2 145 86 87 -398 Deducting the nitrogen and oxygen contained in the gas we have here a relation of hydrocarbon to condensation and carbonic acid as,-9.64 35.32 39.532,or as 100 366 409 closely corresponding to the relations in hydride of butyl which are,-] :3*5 4. The gas was therefore composed of-28.74nitrogen 15.37 oxygen 55-89hydride of butyl and it would appear from this experiment that the light volatile liquids absorb and retain oxygen in greater proportion than that element is coutairied in atmospheric air. The liquid condensed by the freezing mixture during the collec- tion of these gases and that obtained by subsequently heating the large body of liquid from which they were expelled to a higher temperature not exceeding however 30" C.or the boiling point 60 RONALDS ON THE R1os-r VOLATILE CONSTITUENTS of hydride of amyl mas redistilled. It began to boil at 0" C.; a considerable portion passing over betarem 0" and 4" was collected separately ;another fkaction between 6" and 8" was also collected apart ; the remainder nearly all distilled below 15" cent. The liquid distilling between 0" and 4" C. is nearly pure hydride of butyl which has not yet been described. It is a per-fectly clear colourless very mobile liquid having an agreeable sweet smell but eluding by its great volatility the sense of taste. It is insoluble in water but dissolves in alcohol and ether and alcohol of 98 per cent.absorbs between 11 and 12 times its volume of the vapour at a temperature of 21O.5 C. It burns with a yellow not very luminous flame. Mixed in the gaseous state with twice its volume of chlorine liquid chloride of butyl is formed and the original 3 volumes become condensed into 2 volumes of hydrochloric acid. The specific gravity of the liquid at 0" C. is 0.600. It is therefore the lightest liquid at present known. The vapour-density determined by Durn as' method the vapoiir being absorbed by alkohol gave he following results :-Temperature of air ...... 13'8 C. Temperature of sealing ........ 40" C. Barometer ............ T615 m. Capacity of globe .......... 185.6 ce.Empty globe .......... 30.577 grms. Air bubble ................ 7.8 cc. Globe and substance ,.,. 30.788 grms. Temperature of alcohol ........ 14" C. Hence vapour-density = 12-11. Hydride of butyl C4HI0,requires by calculation 2.006. The liquid analysed eudiometrically in the gaseous state gave the following numbers :-Analysis of Butyl Hydride. Temp. Corrected VO~. OybGr Pressure. pressure. Cent. at 0" and 1m. aas ........ 35-04 0.1944 M. 5°C 6 -691 After addition of oxygen.. .. 326 . 0.4810 4-8°C 154.11 After explosion ...... 294.8 0 .4507 4°C 130.95 After absorption of COZ .... 256. 9 0 *4215 9°C 104.33 Hence we have,- OF AMERICAN PETROLEUX. Gas. Condensation Carbonic Acid. 6.631 23.16 26.12 100 346 390 or Hydride of butyl requires- 100 350 400 The liquid collected between 6" and 8" C.is not very different from this last. It is however a mixture of hydride of amyl with liydride of butyl. Its sp. gr. at 0" C. was found to be *6004. The vapour-density was 2.178 and the composition in the gaseous state is shown by the following numbers :-Temp. Corrected vol. Pressure. Cent. at 0" and 1 m. pressure. -I-Gas .. .. . . .. 15.3 0.4392 19-3 9 -39 After oxygen .. .. .. 264.5 0.6912 19.3 185 *22 After explosion .. .. .. 223 * 0,6509 17.9 149 -12 After absorption .. .. .. 166'8 0.6154 19 *5 106 -78 Hence we have,-Qas. Condensation. Carbonic Acid. 9.39 36.10 42.34 100 384 450 or Hydride of butyl requires,- 100 350 400 It was not to be expected from the manner in which the gases were collected that any single portiou woiilcl correspond exactly in composition with any member of the series and some attempts which were made to separate the gases from each other by wash- ing with alcohol did not yield more conclusive results than those already obtained with the mixtures.From the foregoing experiments we may I thiuk safely con- clude that all the homologues of marsh gas excepting marsh gas itself are present in the liquid as it comes to this country and there appears to be little doubt that marsh gas and perhaps even iree hydrogen will be found among the gases which are evolved with the oil at the springs.