SOME HOMONUCLEAL TRI-DERIVATIVES OF NAPHTHALENE. 765 LXVIIL-Some Homonudeal Tri-derivatives of Naphthalene. By RAPHAEL MELDOLA, F.R.S., and C. H. DESCH. WHEN parabromo-a-acetonaphthalide is nitrated in acetic acid solution, it readily forms a homonucleal mononitro-derivative (Lie- bermann and Scheiding (Aiinalen, 183, 258). The nitro-group is ortho- t o the acetamido-group in this compound, as is shown by its yielding an anhydro-base on reduction (Meldola, Trans., 1885, 47, 505).” The nitrobromo-a-acetonaphthalide is easily and quantita- tively hydrolysed into the nitrobromo-a-naphthylamine, NH, : NO2 : Br = I : 2 : 4, by means of sulphmic acid (ibid., p. 500). * At the time of the researches above referred to, I was led to the conclusion that when the mixed nitro-derivatives obtained by nitrating a-acetonaphthalide in scetic acid (Liebermann’s method) are brominated, it is the paranitro-compound only which is attacked.This conclusion was based on a supposed isomerism of the aitrobromo-derivative, which apparently differed in melting point from Lieberrnann’e. But the isomerism was more especially inferred from the observation that the dibromonaphtliylamine obtained from this compound by hydrolysis, replacement of NH2 by H, reduction, acetylation, bromination, and hydrolysis gave a bromophthalic acid by oxidation (loc. cit., p. 511). Liebermann and Scheicling’s compound by similar treatment gave a dibromonaphthylamine which gave doubtful results CR oxidation (Zoc. cit., p. 515). Both dibromonaplithylamines gave the same dibromo- naphthalene by the diaza-reaction.I had always intended to repeat these experi- ments in order to clear up the discrepancy, but a3 my colleague, Dr. Armstrong, was engaged in a detailed studg of the naphthalene derivatives with a view to settle the VOL. LXI. 3 H766 MELDOLA AND DESCH: Some years ago experiments were commenced by one of the authors of this note with the object of ascertaining whether in homo- nncleal tri-derivatives the hydrogen in the second &position could be substituted by NOz, so as to produce homonucleal tetra-derivatives from which the missing di-P-qninone might possibly be prepared. It was thought that if the quinone itself could not exist, its di-haloyd derivatives of the type CI,H,BrzOz [Br : 0 : 0 : Br = 1 : 2 : 3 : 41 might be more stable. The discovery of tetrabromo- and tetrachloro- ortho-benzoquinone by Zincke in 1887 (Ber., 20, 1776) gave encouragement to continue the work in this direction. As a pre- liminary step, it was attempted to replace the amido-group in the nitrobromonaphthylamine by bromine by decomposing the diazo- perbromide, CloH5Br(NO2) *Nz-Br3, in the usual way.It was found, however, that by this method the nitro-group was also replaced by bromine, and the product consisted only of the tribromonaphthalene (m. p. 113-114') previously described by one of us (Trans., 1883, 43, 4). About the same time, and quite independently, the same result was announced by Prager (Rer., 1885, 18, 2163). This author further stated in his paper that the amido-gronp in the nitrobromo- naphthylamine could not be replaced by bromine by Sandmeyer's modification of the diazo-reaction. Accepting Prager's statement as correct, the experiments were accordingly abandoned. orientation of the substituents in the di- and tri-chloro and di- and tribromo-deriva- tivee, this field was left to him.In a recent communication by Armstrong and Roseiter (Proc., No. 104, 7, 186), these authors have shown that the product of the action of bromine on the mixed nitro-a-acetonaphthalides is identical and not isomeric with Liebermann and Scheiding's compound, and that it is therefore the ortho- and not the para-compound which is attacked, as I had originally supposed. They have also shown that the two bromonaphthalenes are identical, being the 1 : 4- modification (m.p. 82-83'). These results, which I have since repeated with the assistance of Mr. E. M. Hawkins, and can now corroborate, are instructive as en- forcing the conclusion to which I was led seven years ago, that in the study of isomeric naphthalene derivatives, melting points are of little or no value (Zoc. cit., p. 516). It appears also that oxida- tion with nitric acid gives very uncertain information respecting the homo- o r heteronucleal character in many cases. These and other difficulties which the in- vestigator of naphthalene derivatives has to guard against are pointed out by Arm- strong and Rossiter in another note (Proc., No. 104, '7,182). The effect of the revision of the work by these authors is to remove my bromonaphthplamine (m. p. 6.2") and its derived compounds from the list of naphthalene derivatives.I am glad of the present opportunity of prercnting the perpetuation of these compounds in chemical liberature. I may alEo take occasion to remark that the 1 : l'-(peri)di- bromonaphthalene which formed the object of the investigation by Armstrong and Rossiter referred to above has recently been obtained by me in conjunction with Mr. F. W. Streatfeild from the 1 : 1'-nitronaphthylamine. The melting point of peri-dibromoiiaphthalene is 108'5-109". Details concerning this and other peri- di-derivatives will be given when our investigations are more complete.-R. M. All subsequent experience has confirmed this.SOME HOMONUCLEAL TRI-DERIVATIVE8 OF NAPHTHALENE. 767 Having recently had occasion to prepare a quantity of the nitro- bromonaphthylamine, it appeared worth while to try whether a feebler acid like iodhydric acid would enable the NH2 group to be re- placed by iodine, without at the same time replacing the NO, group.The nitrobromonaphthylamine was dissoIved in strong sulphuric acid, the solution diluted with a little water, and when cold diazot- ised in the usual way. The solution of the diazo-salt was filtered into a solution containing tb e calculated quantity of potassium iodide. Decomposition occurs a t ordinary temperatures with the evolution of nitrogen, and the reaction is completed by gently heating for some time till no more gas is evolved. The resinous-looking product was washed with water, and then crystallised repeatedly from alcohol with the addition of animal charcoal.The purified compound con- sisted of small, ochreous needles, melting at 117-118". Analysis showed that the desired substitution had taken place :- 0.4819 gram gave 15.3 C.C. moist N at 15.6 and 764.9 mm. bar. 0.2982 ,, 0.3344 gram mixed AgBr and AgI. Calculated for CloH61Br-NOz. Found. N . . ............ 3-70 3.73 Br + I.. ........ 54.76 54-87 The constitution of this nitrobromoiodonaphthalen e is :- I The substance is very readily soluble in benzene, but slightly soluble in petroleum, and freely soluble in hot alcohol, acetone, and glacial acetic acid. By the action of zinc-dust and acetic acid, the compound can be reduced in alcoholic solution, without the removal of halogen, if the solution is kept cool. The base, bromoiodo-p- naphthylamine, is thrown out of the alcoholic solution on dilution with water, in the form of whitish needles, melting when crude a t about 89".The substance soon becomes reddish-brown by oxidation on exposure to the air, and has not therefore been isolated in a state of purity. Its basic character is practically destroyed by the halogen atoms present in the molecule. The acetyl derivative was prepared by heating the solution of the compound in glacial acetic acid with acetic anhydride. After several crystallisations from alcohol, the pure bromoiodo-p-acetonaphthalide was obtained in the form of opaque, white needles, melting at 235" :-768 MELDOLA AND DESCH: 0.2277 gram gave 7.1 C.C. moist N at 20" and 761 mm. bar. Calculated for The direct replacement having been so easily effected in the case of iodhydric acid, we were led to repeat Prrlger's work, using Sand- meyer's method.In the first place, the solution of the diazotised nitrobromonaphthylamine was filtered into an excess of cuproa s chloride solution, and the reaction completed by heating in the usual manner. The crude product was crystallised several times from alcohol, and finally obtained in the form of ochreous needles, melting at 117', and possessing all the properties of t.he preceding compound. A qualitative test showed the presence of nitrogen, Analysis gave the following results :- 0.2922 gram gave 13.7 C.C. moist N at 21" and 759.5 mm. bar. 0,1955 ,, 0.2257 gram mixed AgCl and AgBr. Calculated foil CI,B5C1Br-N0,. Found. N. .............. 4-89 5-53 Cl + Br .........40.32 40.22 The compound can be reduced in cold alcoholic solution with zinc- dust and acetic or even hydrochloric o r sulphuric acid, without loss of halogen. The amido-derivative, which is possessed of but very feeble basic properties, separates from thc alcoholic solution on dilu- tion with water, in the form of whitish needles. Although some- what more stable than the corresponding bromoiodo-P-naphthyl- amine, it also becomes coIoured by oxidation on exposure to the air, and has not been obtained as yet in a, pure stat?. A specimen crystallised two or three times from benzene was obtained in the form of Brownish, stumpy needles, melting at about 93". We do not give this melting point as final, as i t is proposed t o prepare a, lnrger quantity of the substance for future extension of the work.Analpis showed, however, that even at this stage the compound was f ;tirly pure :- 0.1756 gram gave 8.4 C.C. moist N a t 18.5" and 766.8 mm. bar. Calciilnt ed for CI(,H5C1!3r.NHn. Found. N . .............. 5.4fi 5.56 The acetyl derivative was prepared by acetylating in acetic acidSOME HOMONUCLEAL TRI-DERIVATIVES OF NAPHTHALENE. 769 with acetic anhydride. After several crystallisations from dilute alcohol^ it formed fine, white needles, melting a t 218" :- 0.1474 gram gave 5.9 C.C. moist N at 18" and 765.2 mm. bar. Cdculat,ed for CI,H,C1Br.NH.C2H80 Found. N , ................ 4.69 4% 5 By a similar application of Sandmeyer's method, the nitrobromo- nnphthylamine was converted into dibromonitronaphthalene, whicli, when pure, forms small, ochreous needles, melting at 117". The cou- Etitution of these last compounds is represented by the formula :- c1 Rr The results now submitted are snfflcient to show tha.t Pr;ger wa3 in error in stating tha% Sandmeyer's method is inapplicable to the mitrobromonaphtbylarine in question. It is proposed to extend the investigation of these compounds in the direction indicsted at the outset, and the preserft note is offered as a preliminary communica- tion, owing to urnadvoidable interruption of the work till the autumn. Finsbury Technical College.