THE BROMINE COMPOUNDS o$ PHICNANTHRBNE. PART I. 55 IV.-Ihe Bromine Compounds of Phenanthrene. By HERBERT HENSTOCK. WHEN it is coiisidered that phenanthrene is one of the principal aromatic hydrocarbons derived from coal-tar it seenis strange that up to now almost the only commercial use to which it has been put is the production of lamp black by its destruction. Several of the alkaloids notably morphine apomorphine and thebaine have the phenanthrene molecule as the chief basis of their structure and the hydrocarbon should also be a fruitful source of colouring matters. One of the chief reasons why its value has not been more appreciated is undoubtedly the very meagre knowledge which we have of its more common compounds. Its halogen derivatives offer a promising field for inquiry.The number of possible isoiiiexic bromophenan threnes is large, although few have been already described and still fewer thor-oughly investigated. Of the five possible monobromo-derivatives only 9-bromophen-anthrene (m. p. 63O) is known (Hayduck ,4?znaZen 1873 167 181). Of the twenty-five possible dibromo-derivatives only six are known, namely the 2 7 (m. p. 199-2OOO; Hayduck Zoc. cit. ; Schmidt and Mezger Ber. 1907 40 4562) the 9:lO (m. p. 181-182O; Hay-duck Zoc. cit. ; Schmidt and Ladner Ber. 1904 37 4404) the 3 9 (m. p. 146O; Zetter Ber. 1878 11 170; Schniidt and Ladner Ber., 1904 37 3571 ; Sandqvist Rer.. 1920,53 [B] 168) the 4 9 (or 10) (m. p. 113O Werner Aanalen 1902 321 331) the 10 (1 4 5 or 8) (ni. p. 123O Sandqvist B e r . 1915 48 1146) and a dibromo-deriv-ative melting a t 202O (Annnlen 1873 167 182).With the excep-tion of two o r three imperfectly described ones nothing is known of the sixty-one possible tribroino-derivatives. I n brominating phen-anthrene the 9 10-positions are those most readily attacked and this dibroniide is very unstable decomposing a t looo with the evolu-tion of hydrogen bromide and the formation of the 9-bromo-deriv-ative (Hapduck ZGC. cit.). I n the course of this work a 2( 2)- bronLoplzeiiarktitrene dibromide ( I ) was isolated which decomposed exactly as in the case of Hay-duck’s dibroniide (Zoc. c i t . ) leaving a 2( ?) 10-c7ib~omophenarthrene (11) 56 HENSTOCK THE BROMINB This will probably be found to be a general reaction for coin-pounds of this kind.m e preparation of the 9-bromo-derivative has been improved by Austin (T. 1908 94 1762) by using carbon tetrachloride as a diluting medium. A systematic examination of the reaction in this diluent has been made by the author by using varying amounts of bromine. With two atoms Austin’s results were confirmed. Three yielded 28 per cent. of the 9-bronio-derivative and 23 per cent. of 2( ?)-bromophenanthrene dibromide whilst four atoms gave 31 per cent. of the 9-bromo-derivative together with 30 per cent. of 2( 1)-bromophenanthrene dibromide which decomposed to 23 per cent. of the 2( 1 ) 10-dibromo-derivative melting a t 162O. The dibromo-derivative having the nearest melt,ing point ( 1 5 8 O ) to this is that of Zetter (Zoc. cit.) but his is readily soluble in alcohol and crystallises only in plates whereas the new one is soluble in alcohol only on boiling and crystalliscs most readily in needles but is dimorphous and can also be obtained in plates.The proof of the position of the bromine atom in position 10 is shown by the formation of a phenaiithrone on oxidation. Normally, phenanthrene or its compounds having the 9 10-positions unoccu-pied yield o-quinones but in this instance only one oxygen atom was found in positions was thus : the oxidised product indicating- that one- of these The reaction may be represented already occupied. (111.) The phenanthrone yielded a monoxinie but no dioxime as would have been the case had the 9 10-positions been occupied by oxygen. Additional evidence was f orthconiing on nitration where exactly analogous conditions hold ; a mononitrate but no dinitrate was obtained.It may therefore be concluded that position 10 is occupied by one of the bromine atoms. In an attempt to prove the position of the second bromine atom the evidence was of a negative character. If we consider the posi-tims I 2 3 and 4 which it might occupy number 3 is excluded, because Sandqvist (Zoc. c i t .) describes a dibroniophenanthrene melt-ing a t 1 4 3 O which he demonstrates to be the 10 3(or 6)-derivative. Of the remaining positions 2 was chosen for att)ack. 2-Ethoxyphen-anthrene was subjected t o exactly the same conditions of experi-ment in which the dibromo-derivative melting a t 162O was formed. The sole product was the lO-bronio-2-ethoxyphenanthrene so that, when position 2 is occupied the bromine enters the bridge only; from this it might be inferred that the second bromine ato COMPOUNDS OF PHENANTHRENE.PART I. 67 occupies position 2 but this evidence cannot be taken as conclusive, and the matter is still under investigation. E X P E R I N E N TAL. 2( 1 ) -2lromophenanthrene Dibromide ( I ) , A solution of 32 grams of dry phenanthrene in 50 C.C. of dry carbon tetrachloride was treahed with 57.4 grams (4 atoms) of previously dried bromine in 50 C.C. of the same solvent. After forty-eight hours 11 grams of pale yellow crystals had formed, which were fcund to be 9 10-dibromophenanthrene. ?"he filtrate was evaporated to dryness in a current of dry warm air no heat being applied when a yellow crystalline solid remained.On extrac-tion with three successive portions of cold light petroleum part dissolved leaving 10 grams of R pale yellow substance which crys-Lallised from glacial acetic acid in long flat lemon-yellow needles melting and decomposing at 100-102 (Found Br = 57.64. C,,HgBr requires Br = 57.55 per cent.). The bromine evolved as hydrogefi bromide was estimated by heating to l l O o and passing the gas into silver nitrate solution (Found Br = 18.79. C,,HgBr,(less HBr) requires Br = 19-10 per cent.). The compound can be boiled in glacial acetic acid solution with-out change but it gradually decomposes into the dibromo-derivative and hydrogen bromide if exposed t o a warm atmosphere. 2( ?) lO-Dibronto/ihe?tanthren~ (11). After heating 10 grams of the preceding compound a t looo until all hydrogen bromide ceased to be evolved it yielded 8 grams of a white solid which crystallised from a concentrated solution in boil-ing alcohol in long slender colourless needles forming throughout the liquid.On allowing the filtrate to remain no needles appeared, but very small colonrless plates formed on the sides of the dish. These were again crystallised from a concentrated boiling alco-holic solution when they separated in needles like the first crop. The needles (first crop) on the other hand separated from a dilute solution in boiling alcohol in plates. The needles melted a t 162O, the plates a t 1 6 1 O (Found [needles] C=50.00; H=2*68; Br= 47.68 ; [plates] C = 49-83 ; H = 2.43. C,,R8Br requires C = 50.00 ; H = 2-35 ; Br = 47.61 per cent.).T'he plates were found to be triclinic and the needles rectangular. The substance is therefore dimorphous. When kept for some months the needles gradually break down into plates 58 HENSTOCK THE BROMINE 2( 2) 10-7~l)ibron~o2lhenanthrene is fairly readily soluble in glacial acetic acid ether acetone chloroform or benzene less so in light petroleum or alcohol and insoluble in water. The solubility of the plates in ether acetone or glacial acetic acid is slightly less than that of the needles whilst the latter dissolve in alcohol rather less readily than the plates; in other solvents their solubilities are alike. Both plates and needles yield the same phenanhhrone and the same nitro-compound. 2( 1 ) 10-n;broniophena?Lthrone (111).Two grains of the above dibromo-derivative dissolved in 120 C.C. of glacial acetic acid were oxidised by heating for two hours with an equal weight of chromium trioxide. I f the solution becomes ccld yellow crystals appear but a better yield is obtained by cool-ing to about 30° and pouring into water. The precipitate (1.9 grams) crystallised from alcohol in slender lemon-yellow needles melting a t 2 G 3 O (Found C = 47-64 ; €I = 2-50. C,,H,OBr requires C =47*72 ; 13 = 2.27 per cent.). The compound is not so soluble in alcohol ether or glacial acetic acid as in most of the other organic solvents. Analyses of the substance prepared from both the plates and needles of the dibromo-derivative gave concordant results. It dissolves in warm concen-tiated sulphuric acid with a dark greenish-blue colour and is reprecipitated on dilution with water.The monoxinze is readily formed by boiling for three hours a 1 per cent. alcoholic solution of the substance milh two-fifths of its weight of liydroxylamine hydrochloride. After evaporating the alcohol the residize is boiled for some minutes with water and is then left as a greenish-yellow powder the yield being quantitative. It crystallises from benzene in bulky dull yellow hair-like crystals, which shrink considerably on drying and melt and decompose a t 239O (Found N = 4.05 ; Br = 43.44. C,,H,ONBr requires N = 3-81 ; Br = 43.59 per cent.). Although very readily soluble in ether acetone or alcohol it is insoluble in cold benzene. It does not form an anhydride when treated with alcoholic potassium hydroxide.A 1 per cent. solution of 2( 2) 10-dibromophenanthrene in glacial acetic acid is heated to 60-70° and then cold fuming nitric acid added until a slight permanent cloudiness remains the solution i COMPOUNDS OF PHENANTHRENE. PART I. 59 fipally boiled for five minutes and on cooling silky crystals (75 per cent. yield) are deposited which separate from alcohol in lemon-yellow feathery needles melting a t 1 8 8 O (Found N=3*39. C,,H70,NBr requires N= 3-61 per cent.). The nitro-compound is very readily soluble in chloroform or benzene but in acetone or alcohol only on boiling. On further addition of nitric acid or continued boiling of the mixture no dinitro-derivative was forined although several at tempts were made to obtain this; therefore the bromine atom in position 10 is not readily displaceable by the nitro-group.A similar type of compound namely 9-bromo-1 O-nitrophen-anthrene has been prepared by Schmidt and Ladner (Zoc. c i t . ) . $!(NH,)fBr C,H,-C,H,Br' 2 ( 1 ) lO-I).ibro~i.o-9-Q17i~iop~enanthre~e, The reduotion of the nitro-compound was effected by boiling for one hour with tin and hydrochloric acid. The amine was washed with hot waler dissolved in warm alcohol and the solution filtered. On pouring this into much water a brown flocculent precipitate was obtained which when dry was extracted with a little cold chloroform. The pale brown solid (35 per cent. yield) crystallised from dilute alcohol in colourless needles melting at 1 7 7 O (Found N =4.08.C,,H,NBr requires N = 3.98 per cent.). "he substance is very readily soluble in the usual solvents except light petroleum ether or chloroform although it does not crystal-lise readily from these. On prolonged exposure to light it gradually turns brown. It dissolves on boiling with dilute sulphuric acid, and on cooling flocculent crystals of the hydrogen sulphate appear. The acet91 derivative was prepared by heating the amine with twenty times its weight of acetic anhydride at 130-140° for six hours. The cold solution was poured into water and the resulting white granular solid after being washed with dilute sodium hydr-oxide solution and then with warm water crystallised from boiling alcohol in colourless flat tetragonal plates melting a t 2 0 2 O .The yield was SO per cent. of the theoretical (Found C =48*81; H = 3-01. C,,H,,ONBr requires C-=48-85; H=2.80 per cent.). It is very readily soluble in cold chloroform but dissolves in ether or alcohol only on boiling. Action of Alcoholic Potassium Hydroxide om 2( 12) 10-Dibromo-ph?enanthrene. When tha dibromo-compound (one-fifth of the weight of the potassium hydroxide) is gradually added to 10 per cent. alcoholi 60 THE BROMINE COMPOUNDS OF PRFNANTHRENE. PART I. potassium hydroxide and the solution boiled for four hours the liquid becomes reddish-purple and deposits a 90 per cent. yield of a blood-red precipitate which is insoluble in water and all the usual organic solvents with the exception of carbon disulphide and benzene in which it is sparingly soluble.It crystallises from the 1a.tter in small brilliant scarlet needles which do not melt below 350O. Its composition was not indicated by analysis (Found, C = 52.27; H = 3-03 per cent.). It leaves no residue on burning and is not acted on by strong acids or alkalis in the cold but is decomposed by the former on boiling. It is oxidised by chromium trioxide yielding a clear, yellow solution which on neutralisation with sodium hydroxide gives a yellow precipitat,e. The product is therefore not a quinone, neither is the red substance. Tbis is further shown by the fact that neither gives an oxime. The scarlet substance contains bromine. 9-Bromophenanthrene does not yield a red compound. Dibromo-hydrocarbon CI7HlGBr2. T%e three portions of light petrcleum extract from the bromina-tion of phenanthrene (p.57) were united and evaporated to dry-ness leaving a pale brown oil which solidified after twenty-four hours and crystallised from alcohol in pale yellow needles melting a t 216O (Found C = 53-64 ; H =4.04. C,,H,,Br requires C= 53-68 ; H=4*21 per cent.). The compound contains bromine and derivative of a hydrocarbon occurring phenanthrene. 2-Ethoxyphenanthrene (T. 1906 89, under exactly the same conditions as is evidently a dibromo-as an impurity in the yH=QEr cp C,H ox t 1527) was brominated when 2( P) 1Odibromo-phenanthrene was formed. The product when heated a t looo, evolved hydrogen bromide and left a yellowish-white mass which crystallised from glacial acetic acid in small colourless leaflets (81 per cent.yield) melting a t 148-149O (Found C=63*87; H=4"72; Br=26*34. Cl6HI30Br requires C=63*78; H=4.32; Br = 26-57 per cent.). It is very readily soluble in the usual solvents except methyl alcohol and glacial acetic acid. Since bromine attacks the 9 and 10 positions most readily it is reasonable to suppose that t,hese have been entered and the fact that hydrogen bromide is evolved on heating the crude produc BHATNAGAR STUDIES IN EMULSIONS. PART 11. 61 would substantiate this view a 9 10-dibromide being first formed and decomposing into the above substance and hydrogen bromide, just as in the case of Hayduck's dibromide (Zoc. cit.). Further evidence is afforded by the fact that 9 10dinitro-2-ethoxyphen-anthrene is not attacked by bromine under the same experimental conditions.Y(NO,):Y*NO, C,;H4-C6 H,*OEt' 9 lO-Dinitro-2-ethosyphenanthrene, ?"he dinitro-compound was prepared by treating a 3 per cent. glacial acetic acid solution of the ether with fuming nitric acid until a permanent cloudiness appeared and then adding a slight excess boiling for fifteen minutes and allowing to remain for twenty-f our hours. On pouring into water the dinitro-compound was precipitated and was dried a t 105O. After extracting with a little cold benzene the residue crystallised from the same boiling solvent in pale yellow shining rhombic prisms melting a t 247O (Found N t= 9'66. It is readily soluble in chloroform but not in carbon tetra-chloride. It. burns very easily with a small flash leaving no residue. C,,H,O,N requires N = 9.46 per cent.). The author wishes to express his thanks for grants from the Research Fund of the Chemical Society and from the Government Grant Committee of the Board of Trade towards the cost of the materials used in the work. CUEMIGAL RESEARCH LABORATORY, SCHOOL GARDENS, SHREWSBLJRY. [Received October 30th 1920.