EXCHANGE A Study of the Relation of the Structure of J-Mercaptobenzothiazole and its De- rivatives to Their Value as Accelerators of Rubber Vulcanization DISSERTATION PRESENTED IN PARTIAL FULFILLMENT OF THE REQUIRE- MENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE GRADUATE SCHOOL OF THE OHIO STATE UNIVERSITY BY LORIN BERYL SEBRELL THE OHIO STATE UNIVERSITY 1922 A Study of the Relation of the Structure of J-Mercaptobenzothiazole and its De- rivatives to Their Value as Accelerators of Rubber Vulcanization DISSERTATION PRESENTED IN PARTIAL FULFILLMENT OF THE REQUIRE- MENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE GRADUATE SCHOOL OF THE OHIO STATE UNIVERSITY BY LORIN BERYL SEBRELL THE OHIO STATE UNIVERSITY 1922 TABLE OF CONTENTS Page Introduction 4 Part I Preparation and Properties Historical 6 Mechanism of the reaction 6 Preparation of materials 8 Apparatus .- 9 Procedure 9 Discussion of experimental results. . . 9 Derivatives of 1-mercaptobenzothiazole and their physical properties 12 Summary and conclusions Part II 1-Mercaptobenzothiazole and its Derivatives as Accelerators of Rubber Vulcanization Historical 15 Derivatives 15 Results of compounding tests 17 Mechanism of acceleration by benzothiazole derivatives 25 Summary and conclusions : 28 Acknowledgment 28 Autobiography 29 545101 INTRODUCTION By an accelerator of rubber vulcanization is meant any compound which when added to the rubber-sulfur mixture not only shortens the time required for vulcanization but also causes a decided change in the resulting physical properties. The first organic compound to be so used was aniline, in the year 1906. The use of this compound by Oenslager and Marks in Akron was the beginning of one of the most important periods in the history of the rubber industry. Since that time many hundreds of different or- ganic compounds have been tried with varying degrees of success. The exact mechanism by which these accelerators bring about the desired results has not been definitely determined in all cases, although con- siderable progress has been made. The present investigation consists of two parts. The first deals with the methods of preparation and the properties of a recently discovered series of accelerators. The second part is concerned with the use of these compounds in rubber, the relation of their structure to their curing Dower, and the mechanism by which they function as accelerators. PART I THE PREPARATION AND PROPERTIES OF 1-MERCAPTO- BENZOTHIAZOLE, ITS HOMOLOGS AND DERIVATIVES 1 2 The present paper comprises a study of the synthesis of 1-mercapto- benzothiazole and its derivatives which was made in connection with an investigation of the role played by these compounds when functioning as accelerators of vulcanization. Eight different mercapto-benzothiazoles were prepared and studied. Each of these substances excepting where the quantity of material was limited, were prepared by four separate methods. These methods were alike in that the reaction mixture was heated in an autoclave under pressure. The reaction mixtures were as follows: (1) the corresponding disubstituted thio-urea and sulfur; (2) the zinc salt of the corresponding aryl dithiocarbamic acid and sulfur; (3) the ammonium salt of the same acid and sulfur; (4) a mixture of the corresponding aryl amine, carbon disulfide and sulfur. The first method has been described by Romani. 3 The last three methods are new. 1 /. Amer. Chem. Soc. 45, 2390. 2 The system of nomenclature as outlined in C. A. Decennial Index, 1-10, 2345, namely, omitting the sulfur and beginning the numbering with the carbon atom of the thiazole ring has been followed throughout this article. 6 2 3 The four methods of preparation described in the following pages as applied to 1-mercapto-benzothiazole and its three monomethyl derivatives, together with their disulfides and metallic salts were reported before the Organic Division of the American Chemical Society at the Birmingham meeting, April 6th, 1922. [(a) See Science, 56, 55 (1922).] Shortly afterward, an article by Romani, [(b) Gazz. chim. ital, 52, 29 (1922)] became available, describing the preparation of the three methyl derivatives by one of the methods, namely, heating the corresponding disubstituted thio-urea with sulfur. To him, undoubtedly, belongs priority of publication of a description of these three derivatives by this one method. He did not describe the disulfides. Historical l-Mercapto-benzothiazole was first obtained by A. W. Hofmann 4 in an attempt to prepare the disulfhydryl derivative of thiocarbanilide by the action of carbon disulfide on o-aminophenol. He obtained the same substance by the action of sodium hydro- sulfide on chlorophenyl mustard oil (1-chloro-benzothiazole) and also when carbon di- sulfide was caused to react with 0-aminothiophenol disulfide. The product thus ob- tained, after recrystallization from alcohol, melted at 179 and was easily oxidized to a disulfide melting at 180. Jacobson and Frankenbacher 5 while studying the formation of benzothiazoles, heated azobenzene with carbon disulfide in a sealed tube at 250 for 5 hours. The prod- uct melted at 174 but was identical with Hofmann's 1-mercapto-benzothiazole. The disulfide obtained by the oxidation of this product with potassium dichromate in acetic acid solution after recrystallization from benzene melted at 186. In order to verify the assumption that phenyl mustard oil is an intermediate prod- uct in the formation of 1-mercapto-benzothiazole from azobenzene, these authors heated the former substance in a sealed tube with sulfur for 5 hours. The yield of mercapto- benzothiazole thus obtained was equal to 45% of the weight of mustard oil used. The constitution of the thiazole was further established by fusion with potassium hydroxide, thus regenerating o-aminothiophenol. Azobenzene when heated with phenyl mustard oil was found to yield 1-anilido- benzothiazole although the same substance could not be obtained by the direct action of aniline on mercapto-benzothiazole. Rassow, Dohle and Reim 6 have shown that 1-mercapto-benzothiazole is formed by the action of sulfur on dimethylaniline. , Bedford and Sebrell 7 as well as Bruni and Romani 8 have independently described the preparation of 1-mercapto-benzothiazole by the reaction of thiocarbanilide with sulfur when heated under pressure. More recently Romani 3b has extended this method of preparation to the three monomethyl derivatives of 1-mercapto-benzothiazole. He recommends the use of an excess of sulfur and zinc oxide as a catalyst. Romani pre- pared the metallic salts of these methylated mercapto-benzothiazoles but not the disul- fides. Mechanism of the Reaction Bruni and Romani, 8 apparently following the lead of Jacobson and Frankenbacher, have sought to explain the formation of 1-mercapto- benzothiazole from thiocarbanilide, monophenyl-thio-urea and methylene- aniline upon the assumption that these substances first decompose to give phenyl mustard oil. They mention an alkali-insoluble residue, which apparently was not further investigated. It is shown in the experimental part of this paper that this insoluble residue consists chiefly of 1-anilido- benzothiazole and that it is formed in largest amounts when thiocarban- ilide is used as the starting material. Indeed, it is almost entirely absent when ammonium phenyl-dithiocarbamate or the free aniline and carbon Hofmann, Ber., 20, 1788 (1887). 6 Jacobson and Frankenbacher, Ber., 24, 1400 (1891). 6 Rassow, Dohle and Reim, J. prakt. Chem., 93, 183 (1916). 7 Bedford and Sebrell, J. Ind. Eng. Chem., 13, 1034 (1921). 8 Bruni and Romani, Giorn. chim. ind. applicata, 3, 351 (1921). disulfide are used. The insoluble residue obtained when the zinc salt is used consists almost wholly of zinc sulfide. If the formation of 1-mercapto-benzothiazole from thiocarbanilide is assumed to take place through its decomposition products, such as aniline and carbon disulfide or phenyl mustard oil, by virtue of their reaction with sulfur, then little or no 1-anilido-benzothiazole should be expected since these latter reactions form only small amounts of the alkali -insoluble resi- due. These facts seem to indicate that direct sulfurization of the thio- carbanilide offers the most satisfactory explanation for the simultaneous formation of 1-mercapto-benzothiazole and 1-anilido-benzothiazole. The formation of these two products may be readily understood by use of the following mechanism. 1. Thiocarbanilide may be assumed to be in equilibrium with its tautomeric form. CeHfiNH C NHC 6 H 5 ^= C 6 H 6 N = C NHC 6 H 5 S SH 2. This tautomeric form should occur as two geometric isomers, A and B. CeHeNH CSH C 6 H 6 NH C SH A || B || N C 6 H 6 C 6 H 6 N 3. When A is sulfurized and hydrogen sulfide eliminated 1-anilido- benzothiazole is formed. I CSH C 6 HBNH C SH SR C 6 H 6 NH C S +S /^ H 2 S > ^~\_y > N 4. When B is sulfurized and aniline eliminated 1-mercapto-benzo- thiazole is formed. [CSH C 6 H 6 NH C SH S CSH +S HS II ^ 6 -n6iN-n 2 ; \ - -N In the experimental part of this paper it is shown that the combined yield of these two products will account for between 90 and 96% of the thiocarbanilide used. These facts are of especial significance since Jacob- son and Frankenbacher have shown that 1-anilido-benzothiazole is not formed by the action of aniline on 1-mercapto-benzothiazole. The formation of 1-mercapto-benzothiazole from phenyl-dithiocarbamic acid and its salts is not so easily traced. It may be explained in one of three ways. 1. The dithiocarbamic acid derivative may decompose with the forma- tion of phenyl mustard oil, which in turn reacts with the sulfur to form the mercapto-benzothiazole. This explanation would correspond to that offered by Jacobson and Frankenbacher 5 and by Bruni and Romani 8 to account for the formation of the mercapto-thiazoles in other cases. The former workers obtained a 45% yield of 1-mercapto-benzothiazole by heat- ing phenyl mustard oil with sulfur. In this Laboratory as high as 60% yields have been obtained by the same method. By reference to Table I it will be seen that phenyl-dithiocarbamic acid and its salts yield 75 to 80% of the same thiazole. It seems doubtful, therefore, whether phenyl mustard oil is an intermediate product in the formation of mercapto- benzothiazoles from phenyl-dithiocarbamic acid and its salts. 2. Phenyl-dithiocarbamic acid and its salts not only yield phenyl mustard oil but also undergo a second type of splitting with the formation of aniline and carbon disulfide. This is especially true of the ammonium salt, since thiocarbanilide is found among its decomposition products. Thiocarbanilide should be expected to react with sulfur yielding the mer- capto-thiazole according to the scheme given above. Such a reaction would produce a large amount of an alkali-insoluble residue which experi- ment shows is not the case. 3. There is no a priori reason why phenyl-dithiocarbamic acid and its salts should not undergo direct sulfurization as readily as its anilide. The present authors believe this to be a more logical explanation and that the mercapto-benzothiazole is produced by the loss of the appropriate hydrosulfide from the addition product thus formed. Experimental Part Preparation of Materials. The ammonium salts of the aryl dithiocarbamic acids were prepared by the method of Losanitch, 9 by the reaction of the arylamine and carbon disulfide in the presence of ammonium hydroxide or ammonium sulfide in alcoholic solu- tion. The rapid decomposition characteristic of these ammonium salts has caused many investigators to doubt their actual existence. Ammonium phenyl-dithiocarbamate prepared in the presence of strong ammonium hydroxide is quite stable and after careful drying shows little tendency to decompose. By recrystallization from strong ammonium hydroxide it is obtained as long hexagonal crystals, which when carefully washed and dried have been kept for more than a year without evidence of decomposition. In the presence of moisture however it is slowly converted into thiocarbanilide, carbon disulfide and ammonia, as reported by Losanitch. These ammonium salts may also be prepared by passing ammonia gas into a solution of aryl amine and carbon disulfide in benzene, but the method has no advantages over the one previously described. The zinc salts of the aryl dithiocarbamic acids were prepared by adding a solution of zinc acetate in water or ammonium hydroxide to an aqueous solution of the corre- sponding ammonium salt. They may also be prepared by the method of Krulla 10 who added the metallic oxide to a mixture of the aniline and carbon disulfide. The first method gives the purer product. Thiocarbanilide and its substituted derivatives were prepared by the action of the aryl amine with carbon disulfide in the usual manner. These products were purified un- 9 Losanitch, Ber., 24, 3022 (1891); Ann., 166, 142 (1873). 10 Krulla, Ber., 46, 2669 (1913). 9 til the melting points varied not more than 3 or 4 from the highest recorded values in any case. Apparatus. The apparatus used in this work consisted of a steel autoclave mounted in an electric resistance oven. The autoclave, which had a capacity of 2 liters, was turned from tool steel and was capable of withstanding a pressure of 2000 pounds. The oven consisted of 3 heating elements specially wound from Chromel wire and mounted on a reinforced frame built up from Alundum cement. The heating unit was effectually insulated by packing in a mixture of asbestos and magnesite. Temperature control, secured by an external hot wire rheostat, was sufficiently exact to make it possible to reproduce any given set of temperature conditions. Temperatures up to 600 were easily obtained. Procedure. The substance used to prepare the thiazole was mixed intimately with one molar equivalent of sulfur and placed in the auto- clave. Practically all runs were made by bringing the oven to a temper- ature of 390-400, then lowering the autoclave to position and allow- ing the temperature of the reaction mixture to rise until the pressure reached a maximum. The pressure would rise gradually with the tem- perature to about 170-180 where the increase became much more rapid reaching a maximum at about 225250. The pressure generated varied with the substance used to produce the thiazole, being greatest with the ammonium salts and least with the disubstituted thio-ureas. Reproducing the heating conditions did not always give the same yield of product. Usually the autoclave was withdrawn from the oven immediately after the pressure reached the maximum to secure more rapid cooling, and when it was quite cold the accumulation of hydrogen sulfide was blown off and the autoclave opened. Method of Purification. The crude reaction product was removed from the autoclave by solution in warm dil. sodium hydroxide solution. This alkaline solution was submitted to steam distillation to remove any free aryl amine. The solution was then filtered from a residue of insoluble material and fractionally precipitated by the addition of small portions of hydrochloric acid. The first precipitates were very dark and carried down most of the coloring matter. The last fractions precipitated con- sisted of the 1-mercapto-benzothiazole in an almost pure form. The product was redissolved in sodium carbonate solution and reprecipitated further purification varied depending upon the nature of the product and will be described for each individual substance together with its physical and chemical characteristics. Discussion of the Experimental Results The data collected in the preparation of 1-mercapto-benzothiazole and 6 of its substituted derivatives are set forth in Table I. Not all of the runs made are listed, but those selected are typical. Failure to prepare all the thiazoles by each of the four methods was due to a lack of sufficient quantity of the aryl amines from which to prepare the necessary starting 10 materials. In such cases that method was selected which promised to yield the largest amount of easily purified product. TABLE I DATA ON THE PREPARATION OP I-MERCAPTO-BENZOTHIAZOLES Substances used Max. Max. NaOH Used Time temp, press, insol. Yield G. Hrs. C. Lbs. G. G. % Product 1 ThiocarbanilSde 456 4.00 265 300 105.0 245.0 74. 2 Ammonium phenyl-dithiocarba- mate 263 1 50 224 1000 7 3 174.5 74.0 1-Mercapto- 8 Zinc phenyl-dithiocarbamate. . . . 350 1.83 247 575 101 225.2 77.5 benzothiazole 4 Aniline and carbon disulfide 93 76 2 .00 271 571 9 128.0 76. 6 5 Di-o-tolyl-thio-urea 512 4 .00 293 520 194 n 177.0 49.0 6 Ammonium o-tolyl-dithiocarba- matc 170 1 .42 253 1050 18.5 103.0 67.0 1-Mercapto- 7 Zinc o-tolyl-dithiocarbamate 214 1 .25 266 625 137.5 65.7 37.0 3-methyl- 8 o-Toluidine and carbon disulfide. . 105 benzothiazole 76 1 .58 255 675 34 7 80.4 45. 5 ( 9 Ammonium m-tolyl-dithiocarba- 1 1-Mercapto- mate 186 1 .00 242 625 15 5 116.6 69 5 \ 4-methyl- 1 benzothiazole 10 Di-/>-tolyl-thio urea 384 1 .50 295 475 104. o 185.0 68.0 < 11 Ammonium -folyl-dithiocarba- 1-Mercapto- mate 308 1 .25 227 1050 22. 8 199.2 VI. 5 5-methvl- 12 />-Toluidine and carbon disulfide. 105 benzothiazole 76 1.50 240 738 44. 102 57. 9 < 13 2,4.2', 4' - Tetramethyl-diphenyl- thio-urea 260 1 .25 228 450 178. 6 30.7 17.2 14 Ammonium o,-xylyl-dithiocar- 1-Mercapto- bamate 198 1 .70 220 1000 37. 1 62.1 34.5 3,5-dimethyl- 15 Zinc o,/>-xylyl-dithiocarbamate.. . 239 1 .00 239 775 114. 5 66.2 32.4 benzothiazole 16 m-Xylidine and carbon disulfide. . 121 76 1 .58 247 725 84. 59.3 30.4 17 Di-/>-phenety!-thio ur*a 269 1 .83 244 325 109. o 104.5 58.4 ' 1-Ivtercapto- 18 Ammonium p - phenetyl - dithio- 5-ethoxy- carbamate 339 1 .10 206 1275 8. 5 227.5 73.0 Hpn r f nt h i a 7r1 f 19 Zinc p phenetyl-dithiocarbamate 288 1 58 246 550 125. 147.7 59.3 f 20 -Phenetidine and carbon disul- 137 fide 76 1 .30 249 455 49. o 117.4 55.6 21 0-Anisidine and carbon disulfide. 62 1-Mercapto- 38 1 .20 237 365 36. 5 61.5 62.0 5-methoxy- Runs 1 and 5 were made first, and without preheating the oven, thus accounting for the longer reaction time. In later runs the reaction time was shortened to avoid heat decomposition. The ammonium aryl-dithio- carbamates gave in general the best yields, but the products from the zinc salts contained less tarry material and were therefore more readily purified. ThQ more highly substituted the arylamine, the lower the yield of the thiazole so that only a 34.5% yield of the dimethyl derivative was obtained even from the ammonium salt of xylyl-dithiocarbamic acid. The disubstituted thio-ureas always yielded large amounts of an alkali- insoluble by-product. The yield of this substance is markedly lower when the ammonium salt of the aryl-dithiocarbamic acid or the aryl amine 11 and carbon disulfide were used. In the case of the zinc salts the alkali- insoluble product consisted almost wholly of zinc sulfide. Nature of the Alkali-Insoluble Material In the case, of thiocarbanilide the alkali-insoluble part of the reaction product was found to consist largely of 1-anilido-benzothiazole. The purification of this anilido derivative proved to be so tedious that the sep- aration of a similar derivative was not successfully completed in any of the other cases. 1-Anilidobenzothiazole. 4 The alkali-insoluble residue from the thiocarbanilide- sulfur reaction mixture was dissolved in alcohol and the solution poured into dil. hydro- chloric acid. A small amount of acid-insoluble substance was removed by filtration and the anilido derivative precipitated by the addition of sodium hydroxide. Fifty g. of the original residue gave 36 g. of the purified anilido derivative and 5 g. of the acid-insoluble substance, which proved to be mostly 1-mercapto-benzothiazole. After recrystallization from benzene the 1-anilido-benzothiazole was obtained as light yellow crystals melting at 154. A mixture of this product with 1-anilido-benzothiazole prepared by the action of phenyl mustard oil on azobenzene gave the same melting point. Analyses. Subs., 0.5470: 46.9 cc. 0.1 N H 2 SO 4 . Subs., 0.5102: BaSO 4 , 0.5293 Calc. for Ci 3 H 10 N 2 S: N, 12.38; S, 14.16. Found: N, 12.00; S, 14.25. Using the above data the calculated yield of pure 1-anilido-benzothiazole available from the 105 g. of alkali-insoluble residue from the thiocarbanilide in Table I is 75.6 g. or 16.5%. By combining this yield with the 73.3% of 1-mercapto-benzothiazole ob- tained, approximately 90.0% of the thiocarbanilide is accounted for. When one con- siders the losses incident to purification it is easy to conceive that the actual yield of the two derivatives is much higher, being about 96% on the basis of the crude products. Disulfides The disulfides of each of the mercapto-thiazoles described were prepared by the method of Hofmann. 5 An alcoholic or alkaline solution of the mercapto-thiazole was oxidized by the gradual addition of an alcoholic solution of iodine. The resulting disulfide, insoluble in alcohol or alkalies, precipitated immediately, and after it was filtered, washed and dried was purified by recrystallization from the solvent indicated in each case. Zinc and Lead Salts The zinc salts of the mercapto-benzothiazoles may be prepared by either of two methods. A solution of the ammonium salt may be precipitated by the addition of a solution of ammonium zincate, or an alcoholic solution of the thiazole may be precipitated by adding an aqueous solution of any zinc salt. The first method gives the purer product. Both the normal and the basic lead salts may be prepared, depending upon the method used. The normal lead salts were obtained by precipi- tating an alcoholic solution of the free thiazole or an aqueous solution of its sodium salt by an aqueous solution of any soluble lead salt. The basic lead salts were obtained by precipitating an alkaline solution of the mer- 12 capto-benzothiazole by a solution of lead hydroxide in an excess of sodium hydroxide. These salts were thoroughly washed, dried and subjected to analysis without further purification. The zinc and lead salts of each of the several mercapto-benzothiazoles were prepared, but in only three cases were they actually analyzed and tested for their accelerating power as indicated in the following pages. 1-Mercapto-benzothiazole 5 is soluble in alcohol, benzene and acetic acid. By re- crystallization from dil. alcohol it was obtained as light yellow needles, melting at 177. Analyses. Subs., 0.7502: 44.8 cc. of 0.1 NH 2 SO 4 . Subs., 0.1016: BaSO 4 , 0.2840. Calc. forC 7 H 5 NS 2 : N, 8.38; S, 38.32. Found: N, 8.37; S, 38.38. THE BISULFIDE*; an amorphous slightly yellow powder, melting at 176 ; yield, 87%. THE ZINC SAI/T ; a white, amorphous powder. Analysis. Calc. for Ci4H 8 N 2 S 4 Zn: N, 7.05; Zn, 16.45. Found: N, 7.21 ; Zn, 16.05. THE NORMAL LEAD SALT; a bright yellow powder. Analysis. Calc. for Ci4H 8 N 2 S 4 Pb : Pb, 38.40. Found: 38.49. THE BASIC LEAD SALT; an amorphous white powder. Analysis. Calc. for C 7 H 5 NS 2 OPb: Pb, 53.05. Found: 52.35. l-Mercapto-3-methyl-benzothiazole 3b after repeated recrystallizations from dil. alcohol and finally from 50% acetic acid was obtained as white needles melting at 186. Analyses. Subs., 0.7205: 41.2 cc. of 0.1 NH 2 SO 4 . Subs., 0.1060, 0.1143: BaSO 4 , 0.2720; 0.2929. Calc. for C 8 H 7 NS 2 : N, 7.73; S, 35.36. Found: N, 8.00; S, 35.26, 35.21. THE BISULFIDE; white needles from chloroform; m. p., 162. Analysis. Calc. for Ci 6 H 12 N 2 S 4 : S, 35.55. Found: 35.63, 35.48. THE ZINC SALT; a white amorphous powder. Analysis. Calc. for Ci 6 Hi 2 N 2 S 4 Zn: Zn, 15.37. Found: 15.35. THE NORMAL LEAD SALT; an amorphous yellow powder. Analysis. Calc. for Ci 6 H 12 N 2 S 4 Pb : Pb, 36.50. Found: 36.78. THE BASIC LEAD SALT; an amorphous white powder. Analysis. Calc. for C 8 H 7 NS 2 OPb: Pb, 51.24. Found: 51.50. l-Mercapto-4-methyl-benzothiazole 3b was prepared only through the ammonium salt of w-tolyl-dithiocarbamic acid. After successive recrystallizations from 75% alcohol, 50% acetic acid and benzene, it was obtained as light yellow plates melting at 163. Analyses. Subs., 0.1585, 0.2270: BaSO 4 , 0.4075, 0.5878. Calc. for CsHzNSa: 3,35.36. Found: 35.30,35.52. THE BISULFIDE; white plates from benzene; m. p., 195. Analysis. Calc. for Ci 6 H 12 N 2 S 4 : S, 35.55. Found: 35.57; 35.65. l-Mercapto-5-methyl-benzothiazole 3b was recrystallized twice from 50% acetic acid and finally from benzene. It forms fine, very light yellow crystals melting at 181 . Analyses. Subs., 0.7234: 38.37 cc. of 0.1 N H 2 SO 4 . Subs., 0.1860, 0.1985: Ba- SO 4 , 0.4784, 0.5109. Calc. for C 8 H 7 NS 2 : N, 7.73; S, 35.36. Found: N, 7.44; S, 35.39, 35.34. THE BISULFIDE; This was insoluble in alcohol but soluble in benzene and chloro- form, white needles from chloroform melting at 201-202. 13 Analysis. Calc. for C :6 Hi2N 2 S4: S, 35.55. Found: 35.59. THE ZINC SAI/T; a white amorphous powder. Analysis. Calc. for Ci6Hi 2 N 2 S 4 Zn : Zn, 15.37. Found: 15.40. l-Mercapto-3,5-dimethyl-benzothiazole was recrystallized several times from al- cohol from which it separates in light yellow crystals melting at 250.5. It is only very slightly soluble in acetic acid or benzene. Analyses. Subs., 0.2050, 0.1888: BaSO 4 , 0.4904, 0.4523. Calc. for C 9 H 9 NS 2 : 8,32.82. Found: 32.85, 32.90. THE DisuiyFiDE; fine white needles, m. p. 193, obtained by repeated precipitation from chloroform with the gradual addition of alcohol. Analyses. Calc. for Ci 8 H 16 N 2 S4: S, 32.99. Found: 33.02, 33.13. l-Mercapto-5-ethoxy-benzothiazole was prepared by each of the four methods but only the ammonium ^-phenetyl-dithiocarbamate yielded an easily purified product. Recrystallized from 75% alcohol and twice from benzene it was obtained as well-formed, long, cream-colored needles melting at 198. It is only slightly soluble in benzene and other organic solvents. Analyses. Subs., 0.1950, 0.1852: BaSO 4 , 0.4303, 0.4044. Calc. for C 9 H 9 NOS 2 : 8,30.33. Found: 30.31,29.99. Attempts to prepare the disulfide were unsuccessful. l-Mercapto-5-methoxy-benzothiazole was prepared by each of the four methods but the product in all cases proved very difficult to purify. It was found that by rapidly heating the reaction mixture and rapidly cooling the autoclave a more easily purified product was obtained. Recrystallized from 70% alcohol and twice from benzene it forms light yellow needles melting at 201 . Analysis. Subs., 0.1505: BaSO 4 , 0.3546. Calc. for C 8 H 7 NOS 2 : 8,32.48. Found: 32.37. Attempts to prepare substituted mercapto-benzothiazoles from 0-anisidine, the o- and p-chloro- and bromo-anilines and ^-aminophenol were unsuccessful. Benzidine and carbon disulfide yielded an alkali-soluble product melting above 250 , which was not further investigated. Summary and Conclusions 1. The chemistry of the zinc and ammonium salts of phenyl-dithio- carbamic acid has been extended and the work of Losanitch verified. 2. Three new methods for the preparation of 1-mercapto-benzothiazole and its substituted derivatives have been described. 3. A fourth method previously described by one of us and independ- ently announced about the same time by Bruni of Italy, has been extended to these derivatives. 4. The mechanism of the reaction involved in the formation of 1- mercapto-benzothiazoles by the action of sulfur on disubstituted thio- ureas is fully discussed. Experimental evidence is offered in support of the view that this reaction takes place by virtue of direct sulfurization of the cis-mercapto form of thiocarbani^de and subsequent loss of the aryl amine to form the mercapto-benzothiazole. It is also pointed out that 14 sulfurization of the Zraws-mercapto form of thiocarbanilide and loss of hy- drogen sulfide explains the simultaneous formation of anilido-benzothiazole. 5. Direct sulfurization and subsequent elimination of the correspond- ing hydrosulfide is offered as the best explanation of the formation of mercapto-benzothiazoles by the action of sulfur on the aryl dithiocarbamic acids and their salts. 6. Six substituted mercapto-benzothiazoles are described together with four of the corresponding disulfides. 7. Methods are given for the preparation of the zinc, normal lead and basic lead salts of the 1 -mercapto-benzothiazoles and in three cases such salts are described. PART II 1 Mercaptobenzothiazole and Its Derivatives as Ac- celerators of Rubber Vulcanization 12 THE use of 1-mercaptobenzothiazole as an accelerator of vulcanization was first suggested by Bedford and Sebrell. 3 This announcement was closely followed by that of Bruni and Romani, * who set forth in detail a method for the preparation of 1-mercaptobenzothiazole by heating thiocarbanilide with sulfur under pressure. They also pro- posed a theory for the mechanism of the accelerating action as applied to thiocarbanilide and the thiazole derivatives. The second paper of Bedford and Sebrell 5 revealed that the method of preparation described by Bruni and Romani had previously been known to them, and pointed out that the mechanism used by the latter workers to explain the action of thiocarbanilide as an accelerator was untenable. The present investigation is a continuation of the work of Bedford and Sebrell. It has been carried out with the follow- ing purposes in mind: 1 To make a comparative study of the relative value of several derivatives of 1-mercaptobenzothiazole as accelerators of vulcanization. 2 By a process of substitution and elimination to determine what part of the mercaptobenzothiazole structure is responsible for the accelerating action of these compounds. PREPARATION OF MATERIALS I-MERCAPTOBENZOTHIAZOLE AND ITS DERIVATIVES The preparation and properties of 1-mercaptobenzothiazole and its 3-methyl, 4-methyl, 5-methyl, 3,5-dimethyl, 5-methoxy, and 5-ethoxy derivatives have been fully described in Part I. OTHER BENZOTHIAZOLE DERIVATIVES In order to deter- mine the particular grouping in the 1-mercaptobenzothiazole structure responsible for the accelerating action, it became necessary to prepare a series of related benzothiazole deriva- tives. In each case these compounds differed from the true 1-mercaptobenzothiazole by a single atom or grouping, the remaining part of the molecular structure being identical. All these substances have been previously described, but the details of their preparation are incomplete. 1 Presented before the Division of Rubber Chemistry at the 64th Meet- ing of the American Chemical Society, Pittsburgh, Pa., September 4 to 8, 1922. 2 J. Ind. Eng. Chem., 15, 1009. s Ibid., 13, 1034 (1921). < Giorn. chtm. ind. applicaia, 3, 196 (1921). Ind. Eng. Cbem., 14, 25 (1922). 16 1 -Hydroxybenzothiazole, f / s \ C 6 H/ ^COH, was prepared by the hydrolysis of 1-chlorobenzothiazole, according to the method described by Hofmann. 6 Long heating of the free 1-chlorobenzothiazole does not accomplish the hydrolysis. However, if the hydrogen chloride addition product of 1- chlorobenzothiazole is heated with alcohol for 40 hours, the hydrolysis is almost complete. The solution was rendered strongly acid and precipitated by dilution with water. After two recrystallizations from 75 per cent alcohol 1-hydroxy- benzothiazole was obtained as white needles melting at 136 C. This is the same melting point as given by Hofmann. s -^ > . - -^ 1 / ^ ~. B -^ 8 I ^ v, -~^- ^ ^ ~c f /I / / / ^ "^ ^ II / / ^*-- x< f *~* II / ^ 1 / f/G-Ufff NO 1 MAXIMUM TCNSIL fO'jT f/f rlCRCAP TO 3 ME THYi BCNZO THIAZOU -SMCTHYL J ' CTHOXY fflfTHOXY 7 Tine orcune IN MI The hydrolysis of the hydrochloride may be accomplished in a much shorter time by heating with alcohol under pressure. 1-Amidobenzothiazole, C 6 H 4 NH 2 , was first pre- pared by Hofmann 7 by the action of alcoholic ammonia on 1-chlorobenzothiazole at 160 C. In this laboratory the method, as described, always yields the product in the form of a noncrystallizable sirup. Excellent results were obtained by the following method : Fifteen grams of the 1-chlorobenzothiazole together with 45 cc. of saturated alcoholic ammonia and 15 cc. of aqueous ammonia were sealed in a tube and heated at 200 C. for 4 hours. On opening the tube the contents were mixed with 50 per cent alco- hol, strongly cooled in an ice bath and diluted very slowly with water to 1 liter. After standing in an ice box for 12 hours 1- amidobenzothiazole was deposited as fine white needles, with only a trace of the sirupy impurity. The yield was 10 grams. This product after recrystallization from benzene melted at 127 C., compared with 129 C. recorded by Hofmann. 6 Ber., 12, 1126 (1879); 13, 9 (1880). Ibid., 13, 11 (1880). 17 /Ov 1-Mercaptobenzoxazole, C 6 H 4 ( ^CSH, was prepared by refluxing an alcoholic solution of o-aminophenol with carbon disulfide according to the method of Diinner. 8 After the pre- liminary purification by precipitating from sodium carbonate solution, and recrystallization from water, the product melted at 193 C., the same as recorded by Diinner. CH 2 S\ H-MercaptotMazolen, I J^CSH, was prepared for the purpose of comparison with 1-mercaptobenzothiazole to determine the effect of the aromatic nucleus upon the ac- celerating power of the mercaptothiazoles. The compound was first described by Gabriel 9 who prepared it by the action of j8-bromoethylamine hydrochloride and carbon disulfide in alkaline solution. Some difficulty was experienced in the isolation of both the hydrobromide and its reaction product with carbon disulfide, but a sufficient quantity was finally obtained to determine its relative value as an accelerator. RESULTS OF COMPOUNDING TESTS All the compounds listed above were tested to ascertain their relative value as accelerators of vulcanization. The following experimental formula was used: 100 .00 parts of rubber (smoked sheet) 5 .00 parts of zinc oxide 3.50 parts of sulfur 1 .00 part of 1-mercaptobenzothiazole or a molecular equiva- lent of its derivative or analog With a few exceptions, each of these compounds produced very rapid curing in the formula given. Since many of these substances are rapid and powerful accelerators, slight variations in the milling and curing may produce differences of some magnitude in the tensile-time diagrams. For this reason the conditions of milling and cur- ing were standardized as follows : The stocks were all mixed in 1-kg. batches on a small experi- mental mill. After milling for 20 minutes to break down the rubber, the zinc oxide and accelerator were added and mixed thoroughly into the stock for 10 minutes. The stock was then cooled to the lowest workable temperature and the sulfur added. After 5 minutes further mixing it was removed from the mill. All samples were milled as nearly as possible as described above. The curing was conducted in a steam platen press in which the temperature was controlled as closely as possible. All cures were made in the same set of molds, which had been preheated before use by allowing them to remain for at least 30 minutes in the press at the required temperature. All the cures were made using the same decks of the same press. The vulcanized sheets after being removed from the molds were plunged into a tank of cold water to stop vulcanization. s Ber., 9, 465 (1876); 16, 1825 (1883). oibid., 21, 566 (1888); 22, 1137, 1152 (1889). 18 ly under slightly over od over fc -a-g, fc s s s % ,s.y > > > > > * !=) 000 > > > > Pf IS OS OS O T}< 00 CO CO O rH 00 00 00 Tj< OOOMOO'-H OS CO CO 00 iO OS fc '[fi T}OCO O H OJ i-HCOCOCOCOCN CO CO CO -^ CO * CO (NC^COCOCOCO C^INTOCOCO rH * O5 CO O> OS 1C iM o >-i coo T-I ON (M CO CO CO * CO CO oooioooo I rH O OS O rH OS 00 1C OS C 1C CO OINCOiOlO M OOOOOO OOOOOOO OOOOOO OiOOOiO a OrHOOOOOO OOOiOOOiOOOO COCOrHOOOOCN iCrHOOIN O OSOOOOOO OOOOOOOOO OOOOOOOOOO OOOOOOOOOO w N ^ CD CO CO 00 O O CO Tj< 00 x ^ r* 10 O O 10 rj< -* O 10 (N ^ rH (N iCCNOOOO't OOiO^eOT}* O QJ O U C<1COCOCOCO (N ^ ^ **< CO Tt* CO rHrHC<)lNCOCO rH CN CO CO C O O rH b- oo o t- r-i o .-H i>i>i^oa> COOOCOOO.-H a> CO eo-i t- I> MpSiOTflfOOO OOO5OJQ5 O5COt>-t^oO OOOCOOOO ^O5 CD^C^^ 00 00 OCN,-iiO O>-HCOCOTJ< i-HCO COCOt^OO ^ O i-l i-l iO i-l CON<-HOO t^ O CO CD 00 CO rf< (N-*iOicO'O rHCNCOCO 1 * rH M CO CO Tt* rl tNCNO . 00 . COiJ< . i-HCO . lOINCO * iHiOCO--.cO--.--Oit^---O^J-..(Mt>CO.-CO "T OOoO"3-*-*CO OS CO "3 ^ CO t- t^O5-*OCO O >O "3 T* (M COOOOCOOO 00 Ol~ i-l CO t^ CD i-H rH CN 1-1 OiO OOONiOO -<*COOlO -^CM U3COOOCO O * (NcDOOOOO OOOcMr^TttTX OOOOOO CcOOoO COTttt^cOOi COOO OOOO * IO CO CO CO CO OOOOOO OOOOO OOO OOOOOO OOOOOO OOOOO OOOOO OOOOO OO OOOO r-fC^ NCOCOOJIN -HO CO 22 2^2 2 a a a a a o, d at cs rt rt cB S ' 8 6 S 2 B . 9 6 w v s s s s s s 20 The physical data given in the following tables were ob- tained by two observers on a Scott testing machine using dumb-bell test pieces. The average of two or more closely agreeing strips is given for each cure. The results of the physical tests on the cures made with 1- mercaptobenzothiazole, its alkyl and alkyloxy derivatives, are recorded in Table I, and are represented graphically in Figs. 1 and 2. In comparing the relative activity of these compounds, the writers have used as a basis of comparison the time in which each compound was judged to give the best technical cure. This can best be determined by subjecting the stocks themselves to certain arbitrary tests, such as the effect of repeated flexing and resistance to tear. The nature of the grain as indicated by the appearance of the tear is also taken into consideration. The results of such tests are given for each cure in the tables in the column "State of Cure." In some cases the cures were not sufficiently close so that, for example, a 5-minute cure might be judged by means of these tests to be undercured, while a 10-minute cure would be con- siderably overcured. In such cases the writers have inter- polated for the time of best technical cure. The relative order of activity of these compounds as ac- celerators cannot well be determined from a comparison of the maximum tensiles obtained, since the test sheet which would give the highest tensile is in almost all cases overcured from a practical standpoint. Any attempt to classify the compounds on the basis of the so-called optimum cure, as given by either the maximum tensile product or the energy of resilience, or a combination of these factors, would be sub- ject to the same objection as given above for the classification in order of maximum tensile. 21 If it is desired to utilize the stress-strain data in determining the relative activity of these compounds, they can, for ex- ample, be ranked in the order of the highest tensile stress at 700 per cent elongation given by a 10-minute cure on each compound at 40 pounds steam pressure. In this way the er- rors inherent in a determination of the maximum tensile are avoided. If such a classification is made it will be found that the compounds arrange themselves in exactly the same order as determined by the best technical cure. The writers are therefore inclined to classify the various mercaptobenzothiazoles according to the time necessary to produce the best cure from a practical standpoint as shown by the above arbitrary tests, and to use the stress-strain data as a measure of the quality of the stocks. It is realized that the present method of determining the relative activity might not seem satisfactory to all, and for this reason the complete data for the physical tests, to- gether with the tensile product and the energy of resilience, have been included in the tables. Since only one formula was used in testing all these acceler- ators and because in some cases an insufficient number of cures have been obtained, it is obvious that the results herein presented cannot be considered as representing a complete or detailed compounding investigation. They are, however, considered satisfactory from the standpoint of determining the relative activity of the various mercaptobenzothiazoles and their derivatives as accelerators. Fig. 1 shows the tests made on cures at 40 pounds steam pressure, while Fig. 2 gives the same data for cures at 20 pounds steam pressure. The relative activity of the 1-mer- captobenzothiazoles and the time required to produce the best technical cure are given in Table II. TABLE II ORDER OF REACTIVITY OF THE I-MERCAPTOBENZOTHIAZOLES AS ACCELERATORS < 40 Pounds Pressure - / 20 Pounds Pressure - Time of Time of Best Tech- Order of Best Tech- Order of SUBSTANCE nical Cure Activity nical Cure Activity 1 - Mercapto - 3 - methyl- benzothiazole 5 min. 1 18 min. 1 1 - Mercapto - 3,5 - di- methylbenzothiazole. . . 8 min. 2 25 min. 2 1-Mercaptobenzothiazole. 10 min. 3 30 min. 3 1 - Mercapto - 5 - methyl- benzothiazole 18 min. 4 55 min. 4 1 - Mercapto - 5 - ethoxy- benzothiazole 20 min. 5 60 min. 5 1 - Mercapto - 5-methoxy- benzothiazole 1 hr. 6 2 hrs. 6 The data showing the relative curing power of 1-mercap- to-4-methylbenzothiazole are not given in Table II. A limited supply of this material prevented it from being in- cluded when the foregoing tests were made. It had pre- viously been tested in a different formula. Under these con- ditions its activity as an accelerator was found to lie between the 5-methyl and 3-methyl derivatives, being very close to that of 1-mercaptobenzothiazole. 22 In the absence of zinc oxide the 1-mercaptobenzothiazoles are without appreciable accelerating action. Compounded in the foregoing formula, except that the zinc oxide was omitted, 1-mercaptobenzothiazole cured at 40 pounds steam pressure for 2 hours gave a maximum tensile strength of 96 kg. per sq. cm. Curing the same mixture at 20 pounds steam pres- sure for the same time gave a maximum tensile strength of 50 kg. per sq. cm. These substances, therefore, have no prac- tical value as accelerators except in the presence of zinc or lead oxides. From these results it seems probable that in the process of vulcanization the mercaptobenzothiazoles are con- verted into the corresponding zinc salts, these salts acting as the true accelerating agents. To test this assumption the zinc and lead salts of 1-mercapto-, l-mercapto-3-methyl, and 1- mercapto-5-methylbenzothiazoles were tested in the foregoing c,,o riGURC NO 3 HA*.IMUM TENSIL ZO'ST ZINC SAL T Of - tt-inERc*PToaenzo THIAIOLC 0-2 &ritTHVI.BfN20TH C-J 5HCTHVL Tine or CURE formula. The results are shown in Tables Ilia and III6 and the tensile- time curves in Figs. 3 and 4. The zinc salt of the 5-methyl derivative is slightly faster in its curing action than the zinc salt of the free 1-mercapto- benzothiazole or its 3-methyl derivative. The last two are about equal in curing power. On the other hand, the lead salt of the 3-methyl derivative is the most active of the three lead salts, giving good cures in 15 minutes at 20 pounds steam pressure. The two remaining lead salts are about equal in activity but vary in the tensile strengths produced. In all cases the zinc and lead salts of these mercaptobenzo- thiazoles are much more powerful accelerators than the corre- sponding free compounds. The zinc salt is, generally, more powerful than the corresponding lead salt. These metallic salts must also be used with additional metallic oxide to obtain their maximum accelerating power. 23 o V 1 3 n tf 5* n fr 8 o_ T C ACCELERATOR ] [ercaptobenzothiazole f [ercapto-3-methylbenzothiazole i TABLE lllb PHYSI ACCELERATOR [ercaptobenzothiazole :ercapto-3-methylbenzothiazole ercapto-5-methylbenzothiazole H r M 1 OOt005tO t- OOOCn O OOtOOSCOH-" 00 tO CO 05 CO h-< $C 3 OOOOCn OOOOOCn Q "I ffi n > r 1 , , , . , . |5 i-J S r2 KJ enenenen CO '^ s. O2O5-4CnCn ^O5O5"^IO5"^ O J I OOtOO5COI-' OOtOOSCOl-" OOIOCOO5CO 2' ?| S 5 s ? t/3 H o C/3 CnO5O5O5CnCn O5 O5 O5 O5 O5 00 CO CO 00 00 H rrt oooco -en OtOt-*COCO O t- )-> to t- 1 00 O H $* 5 X$j o o to M " o o S OCOtOOCOCO tOtOtOH->O rf^ hfk rf*- rf^ CO xQ O CO 05Cn05^ -J OCO^-1^05 en ^1 00 00 O5 tO ^ " ? CO o ^ co to to to en en 05 O CO co to to to to to to co co to i- 1 OCOOO^OS Cn^cOCOCnOO ^r Is S- a a. d to to to i- 1 1 i-i )- H- to to i tototototo o OOOOOCOCO COOOh-'h-'en H-COCOtOOxO iK I! S. C3 oS 11 eococotototo cococococo cocococoto M ^ o 2 s o5 S o CncOi-'Cn O rf*OOOO5Cn O^O5O5tOO ^T 05 o O5O5Cni^co CnenOS^Jtn en O3 O3 O3 Cn o tootoo^i . to ^0 i! t-H-l-i > ^ ^ a -O O C5O05CntO CO i O en en 00 x^ ::::g:::::: 1 o n w o > H- tO tO I- 1 1- 1 H-i (-"tOtOtOt-" H-"tOtOtOtO C* 00 h-' O 00 rfi. CO COOh-'H-'CO OOOi-'tOO Cn H-I 00 CO to CO OOOOCOOO OO O5 CO H to vfj 00 to g a| D D s u> o o tOtOtOI-i H- to. to tooo i-i ooscno co tOtOtOtOtO ltOtOtO>-*l-' M CO rfx h^ en O5 OOOOOOU^ XZ ,00 1-00 Cn O COt005tOH-tO f' w H tototototo to tototototo tototototo W coo3^aO5O5en totococoto h-totocoH-" pi- OOOrf>-OOOCO O050000CO OOCO0500 " H w t g w > .*.CO*ken tOtOii-'tO a OCnOCncn OOOOO OOCnOOW o H u X z o ^ *$ Hj o OSCOh-'lO 00 OO'McOcOi-' ^enenrf'COO O. 3 tococoto.en*.^oo c t. o ^ i J> 00 CO^D tOtOO O5O5^O5O5 CnenOSOJCn O -> oo to > cooococoen a* c ID } i. S en r- n N Cn rt N ^ frl o *"*" o M H w w 5 Cn tO OO tO Oi OCO^OO >- oi-'cooen enco-jiH-co s:n > ' CO O O H OSCO^COCntO f^TQ O Cn ^ 00 O ^ CO 00 >^ O H- O *! Cn Cn ^ ~ o M CO Cn t^ Cn 00 ^ n -. *> en ikco bo'os oo 3 n S> o o OOOQ d K OOOOO OOOOOO OOOOcJ^ 1 OOOOO OOOOO co <<<i^Q.n?"inno. "% H n o c *"* G s. ? M 2 24 The disulfides of the several mercaptobenzothiazoles were also tested in the foregoing formula, and were found to be considerably less active than the free thiazoles. Two conclusions may be drawn from these facts: 1 The 1 -mercaptobenzothiazoles when used as accelerators first form the metallic salts by action with the metallic oxide present. These salts are the active accelerators. 2 The salts so formed tend to decompose during the process of vulcanization and an excess of the metallic oxide must be pres- ent to reform the salts and thus maintain the accelerating action TESTS ON COMPOUNDS SIMILAR TO I-MERCAPTOBENZO- THIAZOLE ju-Mercaptothiazoline (A), 1-mercaptobenzoxazole (B), 1-amidobenzothiazole (C), and 1-hydroxybenzothiazole (D) were also tested for their accelerating action Each of these |J MAXIMUM TENSIL 10'STPH A MCRCAPTO TKIAZOLIN B 1-MCKCAPTO BCNZ.OXAZOLC C l-Af1IOOBCNZOTHlAZOI.. a 1-HYOROXYOCNZOTHmZOLE. TIME or CURE compounds resembles 1-mercaptobenzothiazole in certain parts of its structure. The results of these tests are recorded in Table IV and shown graphically in Fig. 5. l-MERCAPTOBENZOTHIAZOLE CSH COH (D) The results show that the /x-mercaptothiazoline exerts a marked accelerating action, but much less than that of 1- 25 mercaptobenzothiazole. This would seem to indicate that the accelerating power is invested in the mercaptothiazole group. When the sulfur atom of the thiazole ring is replaced by oxygen as in 1-mercaptobenzoxazole, the accelerating action is still evident but much lower than that of either 1- mercaptobenzothiazole or ju-mercaptothiazoline. If tne mercapto group is replaced by an amino or an hy- droxyl group, as in 1-amidobenzothiazole or 1-hydroxybenzo- thiazole, respectively, the accelerating action is very greatly FIGURE NO TfNSIL A- 1 nCRCAPTO BEfiUOr B-l 3r}THYL C-l - SHfTHYt. TIMC OF CUftt diminished; indeed, in the latter substance it is almost entirely absent. The conclusions to be drawn from these results are as fol- lows: 1 The accelerating action of 1 -mercaptobenzothiazole and its derivatives is invested primarily in the mercaptothiazole group. The presence of the benzene nucleus adds markedly to the ac- celerating value. Whether this is due to the attendant increase in the molecular weight or more directly concerned with the chemical characteristics of the benzene nucleus, is not yet deter- mined. 2 The atomic grouping CSH, is directly responsible for the accelerating action. Any change in this grouping either greatly diminishes or completely destroys the accelerating value. 3 The mercapto group is more essential to the acceleration than the sulfur of the thiazole ring, although both are necessary to the best results. MECHANISM OF ACCELERATION BY MERCAPTOBENZOTHIAZOLE DERIVATIVES Several theories have recently been advanced to explain the vulcanization of rubber by organic compounds containing 26 the mercapto group. Bruni and Romani 5 have proposed the following mechanism for the acceleration of vulcanization by mercaptobenzothiazoles : 2R SH + ZnO > (R S) 2 Zn + H 2 O x(R S) 2 Zn + Sx > xR S S R + xZnS R S S R > R S R + S (active) They have lately extended this theory to many other well- known accelerators, such as thiocarbanilide and aldehyde ammonia. 11 It is chiefly interesting as it applies to mercapto- benzothiazole derivatives. The disulfides are less active accelerators than the free mercaptobenzothiazoles, both with and without the presence of zinc oxide. In the absence of zinc oxide both types of derivatives are almost without ac- celerating power. These facts seem to show that the disulfide cannot be the active agent in this type of acceleration. The theory is therefore inadequate and cannot be accepted. Bedford and Sebrell 6 have proposed a theory to account for the action of mercapto compounds as accelerators, which may be stated as follows : 2R SH + ZnO > (R S) 2 Zn + H 2 O (R S) 2 Zn + S* > (R SS,) 2 Zn (R SS x ) 2 Zn > (R S) 2 Zn + S (active) This mechanism affords an explanation of the following facts: 1 The metallic salts of the mercaptobenzothiazoles are faster curing than the free compounds. 2 Both the metallic salts and the free compounds are faster curing than the disulfides, which must first undergo a reduction to the mercaptans before functioning as accelerators. 3 All mercaptobenzothiazoles require the presence of zinc or lead oxides for the development of full accelerating power. 4 The metallic salt of the mercaptan is therefore assumed to be the active agent. Excess metallic oxide must be present at all times to reform the salt if it should be decomposed by the action of heat or hydrogen sulfide. Evidence of such decomposi- tion is to be found in the low curing power of these salts in the ab- sence of metallic oxides, and their superior power when an excess of the oxide is present. While it cannot be shown that sulfur split off from poly- sulfides is an active form, it has been proven that the sulfur in trithioozone is particularly active, 6 and it is inferred that the sulfur yielded by the decomposition of the polysulfides is similarly active. The polysulfide theory, as given above, appears to offer the best explanation of the existing facts. Doubtless, both the mechanisms given above function in specific cases. It is also probable that no simple explanation of activation will be found which is equally applicable to all types of accelerators. Fur- ther work is being done to determine the validity of each of the reactions involved in the above mechanisms. 11 Romani, Caoutchouc gutta-percha, 19, 11626 (1922). 27 OOCntOCDOOtOCOOsenOO5tf^COCOOSrf.COtOi-'^i. > 3 I I OOOO OOOO OCnOcnO OiOCnOOO g'J* n < ^"2, I I- OH- OlOlOl-'toCOCOtOtOtO^-COCOtOCOi-'O ^ CnCn en Cn en Cn Cn en en en en .< KH rf\ Cn Cn W ^ iCOt*>-tOCOOoenrt.OCOOsCntfk.CnH-OcOCOOs S ft Cn en " OCn > * e? )^>^CO)f^OOOOsi(^tOOMCneoOsri^rfk.i^.OOo 2 w enen Cn Cn cj5 * ^ COMl-' CO >-< en ^ " % - 51 i-i K- eotoi-'*-' cocotoi-- oo 4^ en 4. co to o 2 OOOO- OO^IOOi*>. Cni-OiOOcO OsOOOCnenO O fi A i o COlOi-' CnCn^tO COOOOO*. COOOOCOOS OOOsOSCn*- ddd o^^^dd co O ES ..BOB <^'S.S.B B H g c re n g; " " n" H s : S 28 SUMMARY Mercaptobenzothiazole and its substituted derivatives show the follow- ing relative activity as accelerators of vulcanization: (1) 3-methyl, (2) 3,5-dimethyl, (3) unsubstituted, (4) 4-methyl, (5) 5-methyl, (6) 5-ethoxy, (7) 5-methoxy. Several compounds with a structure analogous to 1-mercaptobenzothiazole were prepared and their accelerating values de- termined. The mercaptobenzothiazoles have no value as accelerators except in the presence of zinc oxide. The zinc and normal lead salts of these thiazoles are faster curing than the corresponding free thiazoles. The zinc salt of the l-mercapto-5-methylbenzothiazole and the normal lead salt of l-mercapto-3-methylbenzothiazole gave the highest accelerat- ing values of any of the compounds tested. The disulfides have a lower curing power than the corresponding free thiazoles. The accelerating power of the mercaptobenzothiazoles is directly connected with the atomic grouping 5 C SH. Any alteration in this grouping removes almost entirely the power to accelerate rubber vulcanization. The mercapto group is more essential to the accelerating action than the sulfur atom of the thiazole ring. Both are necessary to develop the highest accelerating power. Aliphatic mer cap thiazoles show marked accelerating power but are inferior to the mercaptobenzothiazoles. The metallic salts of the mercaptobenzothiazoles are assumed to be the active agents in producing acceleration. The existing theories for the mechanism of acceleration by mercapto compounds have been given. The results obtained in the work are corre- lated with the polysulfide theory. ACKNOWLEDGMENT The writer wishes to acknowledge his indebtedness to Dr. C. E- Boord under whose direction this work has been carried out. His unfailing cour- tesy and optimistic spirit have been manifest at all times; and his helpful and kindly advice has helped to make the research a real pleasure. The writer also wishes to acknowledge the cooperation of the Goodyear Tire and Rubber Company, as expressed through Dr. B. B. Spear, for the loan of apparatus and the testing of samples. Further acknowledgment is due Mr. C. W. Bedford and Dr. W. J. Kelly for valued suggestions, and to Mr. C. M. Carson for assistance rendered. AUTOBIOGRAPHY I, Lorin Beryl Sebrell, was born at Alliance, Ohio, November 19, 1894. My elementary and secondary education was received in the public schools and high school of that city, and my undergraduate college training at Mt. Union College. I received the Bachelor of Science degree from the same institution in 1916, and in 1917, I was granted the degree of Master of Science by the Ohio State University. During the war I was connected with the Research Division of the Chemical Welfare Service. After the war I served one year as instructor in chemistry at Case School of Applied Science. In 1919 I entered the employ of the Goodyear Tire and Rubber Company as a research chemist, holding this position until February, 1921, when I left to resume graduate work at the University of Wisconsin. I returned to the Ohio State University in January, 1922, and completed the requirements for the degree of Doctor of Philosophy at the end of the summer quarter, 1922. 54 5101 UNIVERSITY OF CALIFORNIA LIBRARY