EXCHANGE The Catalytic Preparation of Mercaptans DISSERTATION SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BY RICHARD L. KRAMER BALTIMORE, MD. February, 1920. EASTON, PA. ESCHENBACH PRINTING COMPANY 1921 The Catalytic Preparation of Mercaptans DISSERTATION SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BY RICHARD L. KRAMER BALTIMORE, MD. February, 1920. ^ EASTON, PA. ESCHENBACH PRINTING COMPANY 1921 ' b EXCHANGE CONTENTS. 1 . Acknowledgment 4 2. Introduction 5 3 . Historical 5 4 . The Apparatus 5 5. The Method of Work 6 6 . The Product 6 7 . The Catalyst 7 8. Regereration of Catalyst 9 9 . A Study of the Reaction 9 I Effect of Temperature 9 II Rate of Flow 10 III Ratio of Reactants 11 10 . Comparison of Alcohols 11 11 . Purification of Product 11 12. Comparison with Sabatier's Results 14 13 . Summary 14 14 . Biography 16 544007 ACKNOWLEDGMENT. The author wishes to express his appreciation of the interest shown and assistance given by Professor E. Emmet Reid, under whose direc- tion this investigation was carried out. He also takes this opportunity to express his appreciation for the aid, instruction and encouragement received from Professors Frazer, Pat- rick, Lovelace and Ames. THE CATALYTIC PREPARATION OF MERCAPTANS. Introduction. The problem of the preparation of alkyl mercaptans, especially w-butyl mercaptan, was assigned to this laboratory in connection with inves- tigations carried out in the war work. A survey of the methods in the literature excited interest in a catalytic process which appeared to pos- sess exceptional possibilities for development. Accordingly an investiga- tion of it was made which resulted in the erection of a small plant for the production of butyl mercaptan which will be described in a separate article. 1 Later, after the war, the problem was more thoroughly inves- tigated to clear up a number of difficulties which arose in the operation of the process and for its general scientific interest. Historical. Sabatier 2 discovered the catalytic method in his work with metallic oxides as dehydrating catalysts in organic reactions. The method as de- scribed consists in passing a mixture of alcohol vapor and hydrogen sul- fide over thoria heated to 300-380, the resulting product being a mixture of mercaptan, unchanged alcohol and some sulfide. A portion of the alcohol is converted into the olefine hydrocarbon, which in the case of secondary alcohols is quite large. The mercaptan was separated and purified by fractional distillation. 3 In a later communication, 4 Sabatier compared the activities of a number of metallic oxides by means of iso- amyl alcohol at 370-384, and found thoria to be the best catalyst, credit- ing it with a 70% yield of fs0-amyl mercaptan. The description of the method does not include information as to the rate at which the mixture of alcohol and hydrogen sulfide was passed over the catalyst, the proportions of each in the mixture, or the amount of thoria used, though it may be inferred that the apparatus, quantities of materials and rates of passage were analogous to those used in his study of the dehydration of alcohols. 5 Apparatus. The catalyst was contained in a hard glass tube 20 X 640 mm., which was uniformly heated in a special horizontal electric-tube furnace auto- 1 In this investigation Dr. J. W. Kimball, Mr. George Holm, Mr. G. W. Livingston and Mr. R. W. Hale, Jr., also took part. Though their results are not included in the present article, credit is due them for assistance in laying the foundation for the present work. 2 Sabatier, Compt. rend., 150, 1217 (1910). 3 See section on distillation of mixtures of mercaptan and alcohol, 4 Sabatier, Compt. rend., 150, 1569 (1910). 5 Ibid., 146, 1377 (1908). , 6 matically regulated to about 1. The temperature was read on a ther- mometer placed between the catalyst tube and the wall of the furnace. Tests showed that this registered about 20 above the temperature inside the catalyst tube, hence the temperatures given in tables are 20 lower than those actually read. The hydrogen sulfide was generated in a Kipp apparatus and was washed with water, dried over calcium chloride and measured by a calibrated flowmeter. It averaged 95% pure. The alcohol was admitted from a dropping funnel with a calibrated tip to a tube reaching to the bottom of a distilling bulb through which the hydrogen sulfide was passed. This bulb was immersed in an oil-bath and served as a flash-boiler to vaporize the alcohol and as a mixing cham- ber for the vapor and the gas, the mixture being led directly into the furnace. The reaction products passed through a condenser cooled with ice-water into a receiver packed in ice. The uncondensed portion was led through caustic alkali solution and the remaining gas collected over water in a large calibrated aspirator bottle. Method of Work. While the furnace was heating, the rates of flow of hydrogen sulfide and alcohol vapor were adjusted, the products, being run out through a 3-way cock placed between the condenser and receiver. When the desired conditions were attained one gram molecule of the alcohol was placed in the dropping funnel, the stopcock turned so as to connect with the receiver and the time taken. As the last of the alcohol ran in, the receiver was disconnected and the time taken again. The Product. The condensate usually consisted of 2 layers; the lower, or water layer, was discarded and the other weighed and entered as "product." The upper layer in the receiver, except in the case of the methyl compound, which demanded special treatment, consisted of mercaptan, unchanged alcohol, aldehyde, ether, water and condensation products saturated with hydrogen sulfide and unsaturated hydrocarbon. The liquid was freed from hydrogen sulfide by boiling under a well- cooled reflux condenser for 20 minutes, or until a test portion showed no hydrogen sulfide with alcoholic lead-acetate solution. The mercaptan present was then determined iodometrically. 1 The gaseous products col- lected consisted of the unsaturated hydrocarbon resulting from dehydra- tion of a part of the alcohol, hydrogen corresponding to the aldehyde pro- duced and hydrogen originally present as impurity in the hydrogen sul- fide, together with a small amount of air from the receiver. A measured sample was shaken with bromine water to determine the unsaturated 1 Kimball, Kramer and Reid, /. Am. Chem. Soc. 43, 1199 (1921). hydrocarbon. The analysis of the residue from an average run showed it to be nearly pure hydrogen. Methyl mercaptan boiling at 6 required special treatment. It was absorbed in the caustic soda wash-bottles, the resultant alkaline solution placed in a flask, the exit tube of which was connected with a coil con- denser surrounded with freezing mixture, and the mercaptan was dis- placed by passing hydrogen sulfide into the solution. No hydrogen sul- fide passed through till all of the mercaptan had been displaced. The mercaptan was collected in well-cooled tared tubes which were sealed and weighed. The Catalyst. Thoria has been used as a catalyst throughout the work; much time has been spent studying different methods of its preparation so as to obtain maximum activity. Comparing this oxide with alumina and others the activity of which is greatly influenced by mode of preparation, l Saba- tier says, 2 "On the contrary thoria does not present this inconvenience and its activity is not sensibly diminished when it is ignited at a red heat; it seems that so heavy a molecule can not undergo further important molecular condensations." Our experiments show that the mode of preparation of thoria greatly influences its activity in this reaction, some preparations being absolutely inactive. Commercial thoria and discs cut from Welsbach gas mantles were found to be inactive. When thorium nitrate is heated suddenly to a high temperature by being dropped into a red hot crucible a very light porous thoria is obtained, 7 g. of it occupying 200 cc. of space. On ac- count of its enormous surface we expected this to be a wonderfully active form but found it inactive. A thoria gel prepared according to Miiller* was found to be inactive alone but when the concentrated hydrosol was distributed on pumice, Catalyst E, before the final evaporation a fair catalyst was obtained. The precipitated and carefully washed hydroxide from 44 g. of pure thorium nitrate was suspended in 300 cc. of pure water, which was rapidly stirred at 90 while 4 g. of thorium nitrate in 20 cc. of water was added. The hydroxide dissolved to form an orange-yellow hydrosol. The volume was reduced to 20 cc. on the water-bath and the hydrosol evaporated in a vacuum desiccator to a hard glass-brittle light green solid. This was dehydrated at 400 in a current of air. On account of the high cost of thoria, its density, and tendency to pack, a number of experiments were made with different proportions of thoria on pumice as a carrier. 1 Sabatier, Compt. rend., 147, 106 (1909). 2 Sabatier, "La Catalyse," 2nd Ed., Paris, 1920, p. 26. 3 Mueller, Ber., 39, 2857 (1906). 8 The best catalyst obtained was prepared as follows. Pumice sized between 6- and 12-mesh sieves was placed in a dish on a water-bath and a cone, water solution of analyzed thorium nitrate poured over it, the quantities being so taken that the ratio of pumice to thoria should be 3: 1. Other ratios were tried but gave poorer results. The mass was contin- ually turned during the evaporation of the water. The material may be further dried in an oven at 120. This was done with Catalyst F while G and H were not so dried. Catalyst F was easily duplicated and proved to be reliable and efficient; it was used in all of our subsequent work. The thorium-nitrate pumice was placed in a tube in a current of air and heated to 270, the decomposition temperature of the nitrate, till decomposition was nearly complete, after which the temperature was gradually raised to 400, and air passed until the issuing gas would no longer redden moist litmus. A snow-like coating of thoria covered the pumice. C and D were preliminary catalysts prepared in this way except that the proportion of thoria and the decomposition temperatures varied. Catalyst A was prepared by igniting precipitated thorium hydroxide which had been carefully washed to remove electrolytes. It was mixed with glass wool to keep it suspended. The table below gives the results with various catalysts, 74 g. of n- butyl alcohol with an equivalent amount of hydrogen sulfide being passed over the catalyst at 380 in 6 hours except with Catalysts E and H , where it was 4 hours. TABLE I. COMPARISON OP CATALYSTS. Thoria. Pumice. Temp, of prep. Yield. Expt. Catalyst. G. G. C. %. 1 A 20 none Below red 35 . 8 2 B 19 37.5 Below red 43.5 3 C 6 44 400 34.4 4 C 6 44 450 26.6 5 C 6 44 550 21.1 6 D 12.5 37.5 400 44.4 7 D 12.5 37.5 500 42.7 8 E 25 X ... 38.8 9 F 12.5 37.5 270-400 50.2 10 F 12.5 37.5 270-400 52.1 11 F 12.5 37.5 270400 52.7 12 G 8 42 270-400 44.0 13 G 8 42 270-400 45.7 14 H 12.5 37.5 270-400 49.9 15 H 12.5 37.5 270-400 50.1 From these figures it appears that a given weight of thoria is considerably more effective if distributed on a carrier. From Expts. 3, 4, 5, and 6-7, it appears that the activity of the thoria is considerably diminished by heating much above 400. Catalyst H is the most efficient. 9 Regeneration of Catalyst. The catalyst becomes coated slowly at 380, and more rapidly at higher temperatures, with carbonaceous material and its activity diminishes. A fouled catalyst may be cleaned by passing steam through it at 380 until all volatile material is removed and following this with nitrogen peroxide at the same temperature as long as there is any action. The oxides of nitrogen are then completely displaced with steam. The regenerated catalyst is snow-white and shows its original activity. Effect of Temperature. One gram mol of n-butyl alcohol was passed with an equivalent amount of hydrogen sulfide over Catalyst F in 6 hours, the product weighed, and the mercaptan determined. For the higher temperatures the butylene was determined and the aldehyde estimated from the amount of hydrogen remaining, allowing for the hydrogen present in the hydrogen sulfide, the alcohol being taken by difference. Some butyl sulfide was doubtless formed but it was not taken into account as the amount is relatively small and no convenient method of estimating it is known. The results are given in Table II below and in Fig. 1 the yields are plotted against temperature. In order to show the effect of the catalyst on the alcohol alone the last three runs were made without the hydrogen sulfide, under the same conditions. TABLE II. EFFECT OP TEMPERATURE. Product. Alcohol converted. . . . . Alcohol Temp. Weight. Analysis. BuSH. C 4 H 8 . PrCHO. by dif. Expt. C. G. %. %. %. %. %. 1 260 75 9.3 7.7 2 280 78 14.3 12.4 3 300 82 21.7 19.7 4 320 84 29.3 27.4 5 340 80.5 38.2 34.2 0.3 5.1 60.4 6 360 82.5 47.3 43.4 1.2 10.0 45.4 7 360 78.5 48.5 42.3 1.7 11.3 44.7 8 380 77 59.6 50.2 9 380 76.5 61.3 52.1 1.8 15.1 31.0 10 380 73.5 64.6 52.7 2.1 17.4 27.8 Alcohol alone. 11 340 0.5 7.7 91.8 12 360 .... .... .... 2.1 17.7 80.2 13 380 .... .... .... 2.7 32.7 64.6 From these results it appears that the yield of mercaptan increase regu- larly with the temperature but on account of the increasing prominence of side reaction it is not advisable to go above 380. Fouling of the catalyst interferes above this temperature. The rate of formation of butylene is surprisingly low, and is not serious even at 380; that is, this catalyst appears to be more active in the de- 10 32 S ^ / 1 Z 3 4 S 6 7 8 9 10 II 12 Expt. 1 2 3 4 5 6 7 8 9 10 11 Time. Hours. 1 2 3 3 4 5 6 6.3 7.1 8.5 12 Alcohol converted. 72.5 79.0 78.0 78.5 77.5 76.0 77.0 76.5 73.5 76.0 68.0 29.9 35.1 43.1 43.0 46.9 54.0 59.6 61.3 64.6 61.3 71.4 BuSH. C