150 Separation of Illuminants in Mixed Coal and Water Gas. CHEMICAL NEWS, March 26, 1915 17. Baxter and Frevert, Am. Chem. Journ., 1905, THE SEPARATION OF THE ILLUMINANTS IN xxxiv., 109 18. Mixter and Du Bois, Journ. Am. Chem. Soc., 29. Balard, Comptes Rendus, 1844, xviii., 909. 30. Souchay and Lenssen, Lieb. Ann., 1858, cv,, 253. 31. Nickels, Ann. Chem. Phys., 1865, (4), v., 161, 169; 1867, x., 318; Comptes Rendus, 1867, lxv., 107. 32. Harcourt and Esson, Phil. Trans., 1866, clvi., 193. 33. Eaton and Fittig, Ann. d. Chem., 1868, cxlv., 157. 34. Fischer, Journ. Chem. Soc., 1878, xxxiii., 409. 35. Volhard, Ann. Chim. Pharm., 1879, cxcviii., 337. 36. Pickering, Journ. Chem. Soc., 1879, xxxv., 654. 37. Berthelot, Comptes Rendus, 1880, xci., 251 38. Bemmeln, Journ. Prakt. Chem., 1881, xxiii., 379. 39. Christensen, Ibid., 1883, (2). xxviii., 1; 1885, (1), xxxi., 163; 1886, (2), xxxiv., 41; 1887, (1), XXXV., 57, 161; Ber., 1883, xvi., 2495; Oversight. Köng. Selsk. Forh., 1896, 94; Zeit. Anorg. Chem., 1897, xiv., 141; 1901, xxvii., 321. 40. Franke, Journ. Prakt. Chem., 1887, (2), xxxvi., 31, 453. 41. Kehrmann, Ber., 1887, xx., 1595. 42. Vernon, CHEMICAL NEWS, 1890, lxi., 203: Phil. Mag., 1891, xxxi., 469. 43. Brauner, Monatshefte, 1891, xii., 34. 44. Schierning, Journ. Prakt. Chem., 1892, xlv., 515. 45. Neumann, Monatshefte, 1894, xv., 489. 46. Straus, Zeit. Anorg. Chem., 1895, ix., 6. 47. Hutchison and Pollard, Journ. Chem. Soc., 1896, Ixix., 212. 48. Werner, Zeit. Anorg. Chem., 1898, xxi., 201. 49. Wagner, Zeit. Phys. Chem., 1898, xxviii., 33. 50. Rice, Journ. Chem. Soc., 1898, lxxxiii., 258. 51. Reitzenstein, Zeit. Anorg. Chem., 1898, xviii., 290. 52. Gooch and Peters, Am. Journ. Sci., 1899, vii., 461. 53. Meyer and Best, Zeit. Anorg. Chem., 1899, xxii., 169. 54. Jorissen, Zeit. Angew. Chem., 1899, xii., 521. 55. Engler, Ber., 1900, xxxiii., 1097, 1109. 56. Georgievics and Springer, Monatshefte, 1900, xxi., XXXV., 59. Meyer, Ber., 1901, xxxiv., 3606; 1902, 3429; Zeit. Anorg. Chem., 1913, lxxxi., 385. 60. Manchot, Ann. Chem. Pharm., 1902, cccxxv., 105 Ann. Chem., 1902, cccxxv., 95. 61. Schilow, Ber., 1903, xxxvi., 2735. 62. Ehrenfeld, Zeit. Anorg. Chem., 1903, xxxiii., 117. 63. Hahnel, Zeit. Electrochem., 1909, xv., 841. 64. Muller and Koppe, Zeit. Anorg. Chem., 1910, lxviii., 160. 65. Dietz, Chem. Ztg., 1910, xxxiv., 237. See also the references under the title "Experimental." MIXED COAL AND WATER GAS.* By G. A. BURRELL and I. W. ROBERTSON. Introduction. of Mines that resulted in separating the illuminants in the In this paper are shown experiments made by the Bureau artificial gas of Pittsburgh. This gas is made by mixing gas. The separation was made by fractionally distilling one part of carburetted water gas with three parts of coal the gas in a vacuum at low temperatures, and follows the method detailed by the Bureau in separating natural Fractional Distillation at Low Temperatures," Journ. gases (G. A. Burrell and F. M. Seibert, "Gas Analysis by Amer. Chem. Soc., xxxvi., No. 7, July, 1914, 1538-1548). (a) The boiling-point of N-butylene could not be found in the literature. The N-butylene was separated from the iso-butylene. Neither was the N-butane separated from the isobutane. * Presented at the meeting of the American Gas Institute, October 22, 1914, New York City, with the permiss on of the Director of the U.S. Bureau of Mines. as Analysed by Ordinary Methods. (Analysis made September 1, 1914). Constituents. Carbon dioxide, CO2.. Per cent. 2.64 Oxygen, O2 0.81 Illuminants.. 8.67 Carbon monoxide, CO 13'34 Hydrogen, H2 37'04 Methane, CH4 30.96 1.82 4'72 100'00 In both the natural gas work and coal gas work advan- TABLE II.-Composition of the Artificial Gas of Pittsburgh tage was taken of the work on the subject by P. Lebeau and A. Damiens (Comptes Rendus, 1913, clvi., 325, 797), who separated mixtures of the paraffin hydrocarbons and coal gas by the same method. The Bureau, however, separated the paraffin hydrocarbons into single constitu ents, and found it necessary to refractionate distillates and residues in all cases to obtain pure gases. Lebeau and Damiens make no mention of this latter necessity, and separated the paraffins in pairs. Further there is shown in this paper a simple method for the determination of benzene in artificial gas. The principle of the proce dure rests on the fact that the gases at different stages in the analysis are subjected to temperatures at which certain constituents can be removed by a mercury pump from cer tain others which have low vapour tensions at the temperatures selected. It was found impossible to make a clean separation in any case at one fractionation, hence distillates and residues were refractionated until the separation was as complete as was desired. Table I. shows the constituents in artificial gas that can be separated at different temperatures. The first part shows the distillates that can be obtained at a particular temperature and the second part the residues-2.e., those gases that have appreciable vapour pressures at the temperatures given, Ethane, C2H6 Total In making the analysis (Table II.) the carbon dioxide was removed by the caustic-potash solution, the oxygen by alkaline pyrogallate solution, the illuminants by fuming sulphuric acid, the hydrogen by absorption in colloidalpalladium solution, the methane and ethane by slow combustion, and the nitrogen by difference (see G. A. Burrell and G. G. Oberfell, "The Absorption of Hydrogen by Colloidal-Palladium Solution," Journ. Ind. Eng. Chem., vi., No. 8, November, 1914. The above gas was next subjected to fractional distillation at various low temperatures in the apparatus shown at Fig. 1 (apparatus for the liquefaction and fractionation of gases). At a is shown a Dewar flask to hold the refrigerant used in cooling the gases; b is the bulb in which the gases were cooled; d is a gas-analysis burette, and c another gas container for measuring the gases prior to cooling; e is a pressure guage for registering pressures in the Töpler pump; ƒ is a drying tube containing phos phorus pentoxide for removing the water vapour from the gases; g, h, and i are containers for trapping the gases over mercury as they were removed from the pump; h and i are provided with 3-way stopcocks. The particular advantage of containers such as are shown at h and i lies in the fact that they could be filled with mercury by forcing down on them with the stopcock open to the air, finally filling the capillary tubes with mercury. The gas from the pump could then be introduced into them by means of the goose-neck tube attached to, leaving a mercury seal in the capillary tube all the time. is a pressure guage, which was of principal use in the benzene determination; o, n, and m are 3-way stopcocks. A counterpoise (p) attached to the mercury reservoir of the Töpler pump greatly facilitated the working of the pump. First Series of Fractionations. At Fig. 2 are shown the various steps in the main separation of the gas. The original volume of sample (2048 cc.) was first freed of the carbon dioxide by passing it through caustic-potash solution; 53 cc. of carbon dioxide was removed, leaving 1995 cc. The latter quantity of gas was then cooled in the bulb b (Fig. 1) (about 300 cc. at a time), at the temperature of liquid air. After the introduction of each 300 cc. pumping was started, and as much of the gas removed as possible. There resulted a distillate (a) and a residue (b). The distillate (a) was fractionated at the temperature of liquid air, resulting in two more fractions (c) and (d); (d) was added to the residue (b) and the total again fractionated at the temperature of liquid air. The distillate () thus obtained was added to the distillate (c) and the total again fractionated at the temperature of liquid air. The residue thus obtained (10 cc.) was added to the residue (4) and the total again fractionated. There resulted a distillate (g) of 3 cc., which was added to the distillate (i), making a total of 1782 cc. This first series of fractionations shows the general procedure. At the temperature adopted, in this case the temperature of liquid air, the gases were repeatedly fractionated until no more distillate of consequence could be obtained; 3 cc., the final distillate obtained, is only about o:15 per cent of the original quantity of gas (2048 cc.) taken for the experiment. The total distillate obtained at the temperature of liquid air was 1782 cc., and consisted of those gases that have an appreciable vapour tension at that temperature. This fraction consisted of 0.81 per cent of oxygen, 13.25 per cent of carbon monoxide, 37′33 per cent of hydrogen, 3113 per cent of methane, and 4.23 per cent of nitrogen. It will be observed that the quantities of these constituents check very well with the quantities of the same constituents found in the original ordinary analysis of the coal gas. The residue of the first series of fractionations (1916 cc.) should consist of those gases and vapours that do not have an appreciable vapour pressure at the temperature of liquid air: they are: -C2H6, C3H8, C4H10, C2H4, C3H6, C4H8, and C6H6. In other words, the so-called illuminants of coal gas, with the addition of ethane and probably propane. CHEMICAL NEWS, March 26, 1915 After the removal of the ethane and ethylene there remained 24°4 cc. of gas, which was fractionated at temperatures ranging from -130° C. to 120° C. The distillate, 211 cc., was analysed by burning it in oxygen, and measuring the resulting contraction and carbon method does not isolate the gases, and hence does not dioxide, and calculating to propane and propylene. This show that other gases than these two were not present In working on the separation of the paraffin hydrocarbons, however, the Bureau found that propane could be separated from ethane at temperatures between -130° C. and -120° C. Propylene has a boiling-point very close to propane, and the boiling-point of ethylene is not far different from that of ethane; hence, it is assumed that propylene and ethylene would respond to the same treatment as the propane and ethane. (To be continued). ON IRON CASTING. By SERGIUS KERN, M.E., Petrograd. In addition to my article in the CHEMICAL NEWS (cxi., 3), I want to state that the product obtained is in no case the so-called semi-steel, but close grained cast-iron, where the percentage of metallic iron is increased by the addition to the cupola charge of wrought iron or steel scrap up to no more than 15 per cent. By acting so the percentage of free crystalline graphite in the iron is reduced to nearly nought, giving close grained iron castings, as this graphite is absorbed by the free iron introduced, and transformed into amorphous graphite. CHEMICAL NEWS, Preparation and Hydrolysis of Ethyl Hydracrylate. March 26, 1915 153 I started always in the foundry I directed by test- | 175° under ordinary pressure. melting of new brands of iron in crucibles, by mixing them, firstly, with 5 per cent of iron or steel scrap, next with 75 per cent, and so on. When good mechanical tensile tests were here obtained in the cupola charge, the iron or steel scrap was increased by 2.5 to 5 per cent in comparison with the same charge for crucible tests. Once I had the following pig-iron from South Russia :— The method is obviously not suitable for the preparation of the ester in sufficiently pure condition to study its properties and rate of hydrolysis. It gave good results when melted alone with 5 per cent of steel scrap in the crucible; I added to the cupola charge an extra 2.5 per cent of steel scrap, and obtained very decent results:-10'5 to 11'5 tons per square inch in tensile strength, and 13 to 1'4 tons in the fracture tests. (c) Method of Blaise and Maire (Comptes Rendus, cxlii., 215).-Blaise and Maire passed into glacial acetic acid at 0° the formaldehyde vapour obtained by the depolymerisation of trioxymethylene. The saturated acetic acid solution was then allowed to warm up slowly, resulting in a polymer of formaldehyde less complex than trioxymethylene. This polymer was condensed with bromacetic ester in the presence of zinc in a mixture of equal parts of absolute alcohol and ethyl acetate. They obtained a 40 per cent yield of ethyl hydracrylate whose boiling-point was given as 81° at 13 mm. pressure. It is to be observed that this boiling point is about the same as that given for Curtius and Müller's admittedly impure ethyl hydra crylate. (d) Direct Esterification Method.—In view of the insta bility of 3-oxyacids on heating in the presence of minera acids, and of the work of Bogojawlensky and Narbut (Ber., xxxviii., 3344) on the use of anhydrous copper sulphate to facilitate esterification in the presence of acid catalysers, and the more recent work of Clemenson and Heitman (Am. Chem. Journ., xl., 319) on the use of ON THE PREPARATION AND HYDROLYSIS OF anhydrous copper sulphate in the esterification of certain ETHYL HYDRACRYLATE.* By W. A. DRUSHEL. HYDRACRYLIC acid is a well-known substance, and has been prepared by several methods; its esters, however, have received but little attention. In Beilstein's "Handbuch" the ethyl ester alone is mentioned in con nection with Klimenko's attempt to prepare paracrylic acid. Only three references are given to the ethyl ester in Richter's "Lexikon," and of these but one deals with the direct preparation of the ester. Apparently in no former investigation has hydracrylic ester been prepared by the esterification of hydracrylic acid, nor was the ester obtained in pure condition by any of the methods described in the literature. It seemed desirable, therefore, to prepare if possible hydracrylic ester by the direct esterification of hydracrylic acid in order to obtain the ester in pure condition for the purpose of determining its properties and supplying additional hydrolysis data for the series of esters studied in this laboratory. Preparation of the Ester. (a) Method of Klimenko (Journ. Russ. Chem. Soc., xxvi., 412).—Klimenko obtained what he called paracrylic acid by repeatedly evaporating hydracrylic acid with concentrated hydrochloric acid. He then heated 3 grms. of this acid with an excess of absolute alcohol at 150°, and obtained an ester which boiled at 185° to 190°, and which he believed to be ethyl hydracrylate. Ethyl hydracrylate, however, decomposes below 185° when heated at ordinary pressure. In attempting to repeat Klimenko's work the author was unable to prepare the so-called paracrylic acid by the method described. Baeyer (Ber., xviii., 680) did not believe that Klimenko had succeeded in preparing paracrylic acid (C3H4O2)2, but rather that Klimenko's acid was an unsaturated dibasic acid of the formula C6H8O4. If Baeyer's observation was correct Klimenko's ester could not have been hydracrylic ester. (b) Method of Curtius and Muller (Ber., xix., 850; xxxvii., 1276).-These investigators obtained impure ethyl hydracrylate by converting B-chlorpropionic ester into B-aminopropionic ester hydrochloride and treating this with nitrous acid. The ester contained a little chlorpropionic ester as an impurity, from which it could not be freed by fractionation. The crude ester which they obtained boiled at 80° to 84° at 12 mm. pressure or at 170° to From the American Journal o Science, xxxix.. p. 113. oxyacids in the absence of any catalysing acid, it seemed reasonable to expect this procedure to be applicable for the preparation of pure ethyl hydracrylate. The work of preparing ethyl hydracrylate resolved itself into two parts-(1) the preparation of a sufficiently large amount of pure hydracrylic acid, and (2), the esterification of the acid. 1. The Preparation of Hydracrylic Acid.-Of the several methods described for the preparation of hydracrylic acid the method using glycerin as a starting out material, with certain modifications, was found to be the most expedient for preparing the pure acid in considerable quantity. The modifications of former methods, which will be pointed out, resulted in increasing the yield and purity of the hydracrylic acid. oxidised in 200 cc. portions by fuming nitric acid in tubes A kilo. of glycerin was mixed with a litre of water and water, according to the procedure outlined by Mulder about 20 mm. in internal diameter kept cold by running (Ber., ix., 1902). After the oxidation was complete the liquid was concentrated in 400 cc. portions on the steam bath to remove any dissolved nitrous acid. On cooling some oxalic acid crystallised out. The filtrates from the several 400 cc. portions were combined and treated with a little more than enough calcium carbonate to precipitate out the dissolved oxalic acid as calcium oxalate, which was removed by filtration. The filtrate was diluted with water to 3 litres and heated to about 80°. The hot solution was neutralised with calcium carbonate, making use of mechanical stirring to facilitate the solution of the carbonate. The solution was allowed to stand over night for the main portion of the calcium glycerate to crystallise out. This first crop of calcium glycerate after washing with a little cold water was sufficiently pure for further use. The mother liquor, after making faintly acid to prevent darkening on heating, was concentrated on the steam bath and set aside to crystallise. After re-crystallising this second crop of calcium glycerate it was sufficiently pure, and was added to the first crop. The calcium glycerate so prepared was dissolved in hot water and treated with the theoretical amount of dissolved oxalic acid, the ealcium oxalate was filtered off, and the filtrate concentrated on the steam bath to a specific gravity of 1.26, containing about 61 per cent. of glyceric acid. This is the most favourable concentration for converting glyceric acid into ẞ iodopropionic acid by the action of phosphorus iodide. For the preparation of -iodopropionic acid a modification was introduced into the method of Wislicenus (Ber., viii., 1207) and Erlenmeyer (Ann. Chem. Pharm., 154 Preparation and Hydrolysis of Ethyl Hydracrylate. cxci., 284), making use of solid yellow phosphorus instead, To prepare the sodium salt of hydracrylic acid a modification was introduced into the method of Wislicenus (Ann. Chem. Pharm., clxvi., 10) for the conversion of 8-iodopropionic acid into hydracrylic acid by the action of freshly prepared silver oxide. Sokolow (Ibid., cl., 167) observed that when a water solution of B-iodopropionic acid is boiled with freshly prepared silver oxide the iodine is quickly fixed as silver iodide and that in solution are dihydracrylic acid, C6H10O5, and an isomer of this acid besides the hydracrylic acid sought. Moldenhauer (Ibid.. clxvi., 10) also made the observation that considerable lactic acid is formed in this procedure. In order to avoid the formation of lactic acid and the acids identified by Sokolow, Wislicenus heated his solution below 100° while fixing the iodine by means of silver oxide. The silver which combined with the hydracrylic acid formed in the reaction he removed by means of hydrogen sulphide, neutralised the acids in solution with sodium carbonate, and evaporated to complete dryness. From the dry mixture of sodium salts Wislicenus extracted sodium hydracrylate with boiling 95 per cent alcohol, leaving in the residue the sodium salts of Sokolow's acids. crystallised sodium hydracrylate was treated with a little less than the theoretical amount of dilute (1 : 3) sulphuric acid in the cold. The excess of water was then evaporated off on the steam bath, the hydracrylic acid was ex tracted from the residue of sodium sulphate with absolute alcohol, and the alcoholic solution of hydracrylic acid, free from sulphuric acid, was used in the esterification experiments to be described. 2. The Esterification of Hydracrylic Acid.-In the preparation of one sample of hydracrylic acid a very slight excess of sulphuric acid was used in liberating the hydracrylic acid. The hydracrylic acid, without purification from the trace of sulphuric acid, was esterified by boiling with a large excess of absolute alcohol for twelve hours without the addition of any catalysing acid. About twothirds of the acid was esterified, but a relatively large proportion of the resulting ester proved to be acrylic ester. A second sample of hydracrylic acid free from sulphuric acid was esterified by boiling with a large excess of abso. lute alcohol in the presence of anhydrous copper sulphate prepared by the gentle ignition of crystallised copper sul. phate, but not freed from the traces of sulphuric acid formed in the process of dehydration. Here again the esterification proceeded rapidly until about 70 per cent of the acid was esterified, but as before the result was a mixture of acrylic and hydracrylic esters. It seemed, therefore, necessary to avoid even very small amounts of mineral acids in the esterification of hydracrylic acid in order to prevent the formation of acrylic ester. Wislicenus (Ann. Chem. Pharm., clxiv., 181) suggested the use of the vapour from boiling absolute alcohol as an efficient dehydrating agent in esterification. The inefficiency of absolute alcohol alone, as well as that of absolute alcohol vapour passed through the reaction mixture, and the superiority of anhydrous copper sulphate as a dehydrating agent are shown in the following experiment:-32 grms. of hydracrylic acid free from sulphuric acid were first boiled, with a reflux condenser, with 200 cc. of absolute alcohol for three hours, giving 9'4 per cent of the theoretical yield of ester; at the end of six hours 17 per cent of the acid was esterified. Then the reaction flask was heated at 110° and absolute alcohol vapour passed through the reaction mass, giving at the end of two hours 40 per cent. of ester and after two hours more only 42 per cent of ester. The current of absolute alcohol vapour was then discontinued and about 30 grms. of anhydrous copper sulphate, free from sulphuric acid, were added and the reaction mixture gently boiled, giving after three hours 74 per cent. of ester, and after three hours more 83 per cent of the acid was esterified. The ester formed was found to be hydracrylic ester free from acrylic ester. In another experiment 26'4 grms. of hydracrylic acid were gently boiled with 200 cc. of absolute alcohol in the presence of about 50 grms. of anhydrous copper sulphate freed from sulphuric acid by means of absolute alcohol. At the end of three hours 66 per cent of the acid was esterified, at the end of six hours 71'4 per cent, at the end of nine hours 83 per cent of the theoretical amount of ester was formed, and the boiling was discontinued. To remove the unesterified hydracrylic acid in this and in the previous experiment a little less than the theoretical amount of anhydrous sodium carbonate was added to the esterification mixture and the excess of alcohol was distilled off on the water-bath. The hydracrylic ester was extracted from the residue with dry ether and the ether distilled off on a warm water-bath, finally heating the watercrude hydracrylic ester were obtained, which on fractionation under diminished pressure yielded 45 grms. of ethyl hydracrylyte boiling at 95.5° to 96° at a pressure of 20 mm. to 22 mm. The saponification equivalent of the ester so prepared was found to be 119, and the saponification equivalent calculated from the formula CH2OH.CH2.COOC2H5 is 118. When the ester was dissolved in an equal volume of water and was treated It was found preferable to modify this procedure by neutralising the B-iodopropionic acid with sodium carbonate, and then removing the iodine by acting upon a concentrated solution of sodium 8-iodopropioniate with a slight excess of freshly prepared silver oxide at room temperature, making use of mechanical stirring during the process of replacing the iodine by the OH group. To remove completely the iodine from the sodium salt of 400 grms. of B-iodopropionic acid by this procedure required about two hours, but the resulting sodium hydracrylate was apparently free from the salts of the acids formed by the older methods and identified by Sokolow and others. After the complete fixation of the iodine as indicated by no further change in the colour of silver oxide when added to the solution, the precipitated silver iodide was filtered off and the filtrate evaporated to complete dry-bath to boiling. From the two experiments 53 grms. of ness over the steam bath. It is important that the sodium hydracrylate should be evaporated to complete dryness in order to be able to purify the salt by crystallisation from 95 per cent alcohol. The dried residue of sodium hydracry. late was dissolved in boiling 95 per cent alcohol and allowed to crystallise out on cooling. The residue of sodium salt was completely dissolved by the hot alcohol, indicating the absence of the sodium salts of Sokolow's acids. The |