1 A PC . ' 14. ^S: c. a STATE OF ILLINOIS DWIGHT H. GREEN, Governor DEPARTMENT OF REGISTRATION AND EDUCATION FRANK F. THOMPSON, Director DIVISION OF THE STATE GEOLOGICAL SURVEY M. M. LEIGHTON, Chief CIRCULAR NO. 136 SOME TRISUBSTITUTED BENZOTRIFLUORIDES G. C. Finger and Martin Knell SOME FACTORS IN THE SYNTHESIS OF 3,5-DINITROBENZOTRIFLUORIDE O. F. Williams and G. C. Finger PRINTED BY AUTHORITY OF THE STATE OF ILLINOIS URHANA, ILLINOIS 19 4 7 Hi Illinois Academy of Science Transactions, Vol. 38, 1945 SOME TRISUBSTITUTED BENZOTRIFLUORIDES G. C. FINGER and MARTIN KNELL State Geological Survey, Vrbana, Illinois A number of highly fluorinated aliphatic compounds have been known for some time, especially, some of the chlorofluoro derivatives of methane and ethane. Due to their unique properties they have found commercial application as re- frigerants, as propellants in "bug bombs," as gaseous and liquid die- lectrics, and in the synthesis of some very unusual polymers. These dis- coveries and applications warrant research in other types of com- pounds. Almost immediately the field of aromatic compounds sug- gests itself. Very few polyfluorides of such simple molecules as benzene and toluene have been described in the literature. The synthesis of two difluorobenzotrifluorides by Finger and Reed 1 in the Survey labora- tories immediately brought requests for samples and for further informa- tion on these and other aromatic compounds. This paper represents a preliminary report on some of the intermediate compounds studied in trying to find an appropriate syn- thesis for a trifluorobenzotrifluoride. The synthesis of higher fluorin- ated aromatic compounds presents two problems ; ( 1 ) the usual fluorine substitution reactions tend to be in- operative as the degree of substitu- tion increases, and (2) it is neces- sary to know the exact structure of the final compound in order to eval- uate it properly. Ordinarily the second problem is solved if the struc- ture of the compound just preceding the introduction of a fluorine atom is definitely known. These facts are, more or less, illustrated in this pre- liminary report. As a first approach to a probable synthesis of a trifluorobenzotrifluor- ide, 2-acetylamino-5-fluorobenzotri- iluoride (I) was used as the starting compound. This compound was available in the Survey laboratories as a result of some previous work 1 . The nitration of (I) followed by hydrolysis of the acetyl group gave 2-amino-3-nitro-5-fluorobenzo- trifluoride (III) . This was the com- pound desired although on a theo- retical basis two other isomers could be formed, namely, the nitro group could have entered also into the 4 or G positions. These isomers were not isolated although they may have been present in small amounts. The structure of compound (III) was established by conversion to 2, 3- diphenyl-6-fiuoro-8-trifluoromethyl- quinoxaline (V). After having estab- lished the structure of 2-amino-3- nitro-5-fluorobenzotrifluoride ( III ) , it appeared that this nitroamine may be the basis for the synthesis of 2,3,5-trifluorobenzotrifluoride. Attempts to apply the Schiemann reaction, in the usual way, to 2-ami- n o-3-n i t r o-5-fluorobenzotrifluoride (III) so as to obtain 2,5-difluoro- 3-nitrobenzotrifluoride failed. The nitroamine appears to be only slight- ly basic although it forms an amine hydrochloride in concentrated hy- drochloric acid ; the hydrochloride can also be formed in absolute ether by passing in dry hydrogen chloride. Indications are that diazotization did take place; however, when Illinois Academy of Science Transactions CF 3 CF 3 CF 3 CF 3 CF 3 ONHAC y\ NHAC S*\ NH 2 S*\. NH 2 f^Y ^ C " C « H 5 "^ F I^J NO, F l^J NO, — * F j^J NH 2 ~^ F l^^C-C.H, sodium fluoborate was added a pre- cipitate of sodium chloride and sodium fluoborate was formed. Although these results did not achieve the original objective, the nitration data is significant in pre- dicting the results which may be obtained in nitrating 2,5-difluoro- benzotrifluoride. The orientation factors are, more or less, the same thus lending credence to the forma- tion, of 2,5-difluoro-3-nitro-benzotri- fluoride as the chief product. The nitration of 2,5-difluorobenzotrifluo- iide gave a nitro compound and its composition was established through the acetyl derivative of the amine resulting from reduction. There- fore, it is possible to assign a struct- ure with some supporting evidence even though a more positive proof is necessary. The nitro compound is being studied further as to its structure and properties. The authors are indebted to Dr. F. H. Reed, chief chemist, for sug- gestions, and to Mr. H. S. Clark, microanalyst, for the analyses in this investigation. EXPERIMENTAL 2-Amino-3-nitro-5-fluorol)enzotrifluor- ide (III). — Twenty grams of 2-acetyla- mino-5-fluorobenzotrifluoridei (I) was dissolved in a mixture of 150 g. of concentrated sulfuric acid and 30 g. of glacial acetic acid at room temperature. After cooling to 10°, a solution of 12.6 g. of fuming nitric acid (sp. gr. 1.49- 1.5) in 25 g. of concentrated sulfuric acid was added slowly with stirring, and the temperature maintained at 10- 20° for ninety minutes. The reaction mixture was poured slowly into an ice- water mixture giving a white precipi- tate. After filtering and drying, the yield of crude product was 17 g. This crude product was found to be chiefly 2-acetylamino-3-nitro-5-fluorobenzotriflu- oride (II) upon recrystallizing a sam- ple from 50 percent aqueous ethanol. The crude material was heated on a water bath for one hour with 50 cc. of concentrated sulfuric acid, and poured into an ice-water mixture giving the crude nitroamine as a yellow precipi- tate. After filtering and drying, recrys- tallization from 60 percent aqueous ethanol gave 11.5 g. (57 percent) of 2-amino-3-nitro-5-fluorobenzotrifluoride (III) as yellow needles, m.p. 68-69°. Anal. Calculated for CILF^O,: C, 37.51; H, 1.8; N, 12.50. Found: C, 37.63; H, 1.85; N, 12.56. The acetyl compound (II) upon re- crystallization from 50 percent aqueous ethanol gave white needles, m.p. 179.5- 180°. Anal. Calculated for CH«F 4 N 2 3 : C, 40.61; H, 2.27; N, 10.53. Found: C, 40.70; H, 2.26; N, 10.45. 2,3-Diphenyl-fi-fluoro-8-trifluoromethyl- quinoxaline (V). — To a refluxing solu- tion of 2 g. of 2-amino-3-nitro-5-fluoro- benzotrifluoride (III) in 100 cc. of 50 percent aqueous acetic acid, 6 g. of powdered zinc was added in small por- tions. After the mixture was light yel- low in color, the unreacted zinc was re- moved by filtration. The 2,3-diamino- 5-fluorobenzotrifluoride (IV) was not isolated. To the filtrate was added slowly, a warm solution containing 30 cc. of 40 percent aqueous sodium bisulfite, 3 g. of benzil, and 40 cc. of ethanol. The yellow flocculent precipitate which formed was removed by filtration, washed, and dried. Recrystallization from methanol gave 2,3-diphenyl-6-fluoro-8-trifluorome- thylquinoxaline (V) as white needles, m.p. 133-133.5°. The formation of the quinoxaline proved the adjacency of the nitro and amino groups in III. Anal. Calculated for C 21 H )2 F 4 N 2 : C, 68.48; H, 3.28; N, 7.61. Found: C, 68.46 H, 3.28; N, 7.74. ome Trisubstituied Benzotrifluorides Diazotization of 2-Amino - 3-nitro - o- fluorobenzotrifluoride ( III ) . — Attempts were made to diazotize 2-amino-3-nitro- 5-fluorobenzotrifluoride (III) under a variety of conditions. The amine hydro- chloride may be formed in concentrated hydrochloric acid or by dissolving the nitroamine in absolute ether and pass- ing in anhydrous hydrogen chloride. In a concentrated hydrochloric acid solu- tion, the amine hydrochloride appears to diazotize slowly with sodium nitrite to give a clear solution. Upon the addition of a strong sodium fluoborate solution, a white precipitate of inorganic salt ( probably sodium chloride or sodium fluoborate) was formed. Diazotization with nitrosyl sulfuric acid in the pres- ence of phosphoric acid 2 or glacial acetic acid s or 60 percent fluoboric acid failed to give a diazonium fluoborate. 2,5-Difluoro-3( *)-nitrobenzotrifluoride. — To a mixture of 40 cc. of fuming sul- furic acid (40 percent S0 3 ) and 19.3 g. of fuming nitric acid (sp. gr. 1.49-1.5), 18.3 g. of 2,5-difluorobenzotrifluoride was added at such a rate with stirring that the temperature was maintained at 50° for ninety minutes. The entire reaction mixture was poured into an ice-water mixture, and the crude nitro compound was isolated as a heavy yellow oil. The crude yield was 14.8 g. or 62 percent. A preliminary fractional distillation showed that 2,5-difluoro-3(?)-nitrobenzo- trifluoride boils about 191-192° or 91.5- 92.5°/19 mm. The sample was not pure indicating the possible presence of isomers and also a trace of a dinitro derivative. A small sample of the crude nitro compound was reduced with iron filings in an ammonium chloride (0.78 N) solu- tion* to the amine. The amine was isolated as an oil, and converted to the acetyl derivative by the usual method. Recrystallization from high-boiling pe- troleum ether gave 2,5-difluoro-3 ( ? ) - acetylaminobenzotrifluoride as white needles, m.p. 104.5-105.5°. Anal. Calculated for C„H„F 5 NO: C, 45.20; H, 2.53; N, 5.86. Found: C, 45.33; H, 2.43; N, 5.82. SUMMARY 2-Amino-3-nitro-5-fluorobenzotri- fluoride was synthesized from 2- acetylamino-5-fluorobenzotrifluoride. Attempts to convert 2-amino-3- nitro-5-fluorobenzotrifluoride to 2, 5-difluoro-3-nitrobenzotrifluoride by the Schiemann reaction failed. The nitration of 2,5-difluoroben- zotrifluoride gives a nitro compound which is probably 2,5-difluoro-3- nitrobenzotrifluoride. 2. Welsh, ibid., 63. 3278 (1941) BIBLIOGRAPHY J. Am. Chem. Soc. 3. Hodgson and Walker, J. Chem. Soc. 1933, 1620. 4. Lukasevich and Vorshilova, Compt. rend, acad. sci. U.R.S.S. >. 344 (1935). Illinois Academy of Science Transactions, Vol. 39, 1946 SOME FACTORS IN THE SYNTHESIS OF 3, 5-DINITROBENZOTRIFLUORIDE O. F. WILLIAMS and G. C. FINGER State Geological Survey, Urbana. Illinois The synthesis of certain symmetri- cal aromatic fluorine compounds under investigation in this labora- tory necessitated the development of a large-scale laboratory preparation of 3, 5-dinitrobenzotrifluoride. CF 3 CF 3 NO z H 2 SQ 4 2 N Adaptation of the original syn- thesis by Finger and Reed 1 of this dinitro compound to a large-scale laboratory preparation required further investigation to determine optimum yield conditions. The synthesis of 3, 5-dinitrobenzo- trifluoride involves the nitration of m-nitrobenzotrifluoride with a fum- ing nitric and sulfuric acid mix- ture. The nitro group entering the aromatic nucleus is directed to the meta position common to both the — CF 3 group and the original nitro group. The difficulty involved in nitrating m-dinitrobenzotrifluoride may be predicted when a comparison is made with m-nitrobenzene, as both compounds contain two meta directing groups in the 1,3 positions. The nitration of m-dinitrobenzene is extremely difficult; therefore, it would be expected that nitration of w-nitrobenzotrifluoride would also be difficult. This theoretical anal- ogy was verified experimentally as it was necessary to use a fuming nitric- sulfuric acid mixture at a tempera- ture of about 100° C. in order to obtain a satisfactory yield of 3, 5-di- nitrobenzotrifluoride. To develop a standard procedure for a large-scale laboratory prepara- tion of 3,5-dinitrobenzotrifluoride which would produce a maximum yield of product, consistent with eco- nomical use of both time and mate- rials, required a study of the effects on the dinitro yield by the following conditions : ( 1 ) concentration and volume of sulfuric and nitric acid used in nitration; and (2) time and temperature of nitration. An investigation of the effects of the conditions given above on the yield of 3,5-dinitrobenzotrifluoride was conducted by performing a series of nitrations in which the con- ditions were varied from run to run. The yield of dinitrobenzotrifluoride as affected by various concentrations and volumes of acids used are shown in Tables 1 and 2. From the data obtained, indica- tions are that fuming sulfuric acid (30 percent free S0 3 ) and fuming nitric acid (sp. gr. 1.49-1.5), a nitra- tion temperature of 100°C., and a heating period of three hours pro- duced better yields of 3,5-dinitroben- zotrifluoride. Nitration temperatures higher than 100°C, and heating pe- riods longer than 3 hours were found to effect only negligible in- creases in the yield. The crude 3,5-dinitrobenzotrifluo- ride, washed free of acid with dilute alkali, is still contaminated with small amounts of isomers and unre- acted m-nitrobenzotrifluoride which make imperative an efficient purifi- cation process.- Fractional crystalli- zation of the crude product from methanol gave a pure product but Illinois Academy of Science Transactions Reagent Concentration Average Yield of 3, 5-Dinitrobenzotrifluoride Fuming H 2 S0 4 15% S0 3 25% S0 3 30% S0 3 60% S0 3 Sp. gr. 1.49-5 Fuming H 2 S0 4 Red fuming HN0 3 * Procedure wn- stamlnnlized mi t Table 2. — Yield oe 3,5-Dinitrobenzotrifluoride vs. Volume of Reagents Acid Volume Ratio Acid: C 6 H 4 (CF 8 ) (N0 2 ) 1, 3 Percent Yield of C fi H 3 (CF») (N0 2 ) 2 1, 3, 5 H 2 S0 4 (30% free S0 3 ) H2SO4 (30% free S0 3 ) HNO3 (Sp. gr. 1.99-1.5) HNOs (Sp.gr. 1.49-1.5) 5:1 3.6:1* 1.8:1 6.4:1* 41 45 41 51 low recovery, due to the additional .solubility effect of the unreacted m-nitrobenzotrifluoride. Purification by steam distillation was found to be impractical because the dinitro compound is slightly steam distil- lable. Fractional distillation under diminished pressure followed by re- crystallization of the dinitro from methanol proved to be the most effi- cient method for purifying the crude dinitro product. EXPERIMENTAL Nitrations were carried out in a 5- :liter three-necked (standard joints) round-bottomed flask submerged in a water bath containing a coil of copper tubing with connections so arranged as to permit circulation through the coil of either low or high pressure steam, tap water, or a cold brine solu- tion. The flask was equipped with a Claisen adapter to which was attached a dropping funnel and a reflux con- denser, a thermometer, and a mercury- sealed stirrer coupled to the shaft of a one-half h.p. motor. m-Nitrobenzotrifluoride was added to BIBLIOGRAPHY 1. Finger and Reed. J. Am. Chem. Soc. Gd, 1972-1974 (1944^ the stirred fuming nitric and sulfuric acid mixture maintaining the nitration temperature at 100° C. throughout the addition and the subsequent two-hour heating period. The efficient heat trans- fer system previously described makes possible precise temperature control at 100° C. of the nitration reaction which is exothermic. A modification of the above experi- mental procedure involves a nitration of benzotrifluoride with the spent acid to produce enough m-nitrobenzotrifluoride for a subsequent large-scale dinitro- benzotrifluoride preparation. This modi- fication is advantageous from the stand- point of more efficient utilization of ma- terials but has the disadvantage of in- creasing the dinitro production time. SUMMARY A study was made of the nitration of m-nitrobenzotrifluoride to deter- mine the conditions most favorable for obtaining good yields of 3,5-dini- trobenzotrifluoride. From the re- sults thus obtained, a large-scale laboratory method of preparation of 3, 5-dinitrobenzotrifluoride was de- veloped.