COLOUR AND CONSTLTUTION OF AZO-COMPOUNDS. PART V. 511LV.-The Coloui- and Constitution of Azo-compounds.Part V.By JOHN THEODORE HEWITT and FERDINAND BERNARD THOLE.DURING the progress of some work on the relationships existingbetween the constitution and selective absorption of certain polyazo-compounds, the authors of the present communication were muchstruck with the great colour change which takes place on diazotisin512 HEWITT AND THOLE: THE COLOUR ANDaminoazobenzene in hydrochloric acid solution. As is well known,azobenzene and aminoazobenzene both dissolve in alcohol, with ashade which is yellow in dilute solution, the head of the band ofslowest oscillation, but little persistence, lying a t about 2200oscillation frequency in the case of azobenzene (Hnrtley, Trans.,1887, 51, 152; Baly and Tuck, Trans., 1906, $9, 985; Hantzsch,Ber., 1909, 42, 2132), and apparently at about the same point withthe amino-derivative (compare Landauer, Ber., 1881, 14, 391, andC, Graebe, Zeitsch.plqsikul. Chem., 1892, 10, 689). The bestmarked and most persistent band given by alcoholic solutions ofazobenzene has its head at an oscillation frequency of about 3200.Similar observations have been made in the case of dimethyl-aminoazobenzene (2430, Hantzsch, Zoc. cit., p. 2133), p-a.mino-benzeneazophenol (2600), and pdimethylaminobenzeneazophenol(2400, Hewitat and Thomas, Trans., 1909, 95, 1297).I n all cases, however, where an amino-group occupies a para-position with respect to the azo-group, the formation of a monacidsalt by the addition of dilute hydrochloric acid in sufficient quantityconditions a change in colour to red, the oscillation frequencies ofthe heads of the persistent band of longest wave-length lying atabout the same point for the compounds enumerated; this is shownin the following table :p-Aminoazobenzene ........................2000 (see Fig. 1)p-Dimethylaminoazobenzene ........... 1900 (Han tzsch)p-Aminobenzeneazophenol ............... 1800 (Hewitt and Thomas)pDimethylaminobenzeneazopheno1 . , . 1800 ,, , , , ,This change, which corresponds with that observed on acidifyinga solution of methyl-orange, is now generally explained by supposingthe hydrochlorides to be derived from the equivalent quinonoidfGrm :C,H,*N :No C6H,*N H - C6H, *N H*N: c,H,: N H,CI.The possibility of a certain small amount of amino-aromatic saltbeing also present in solutions is not completely negatived, since onemay adduce in support of such a proposition the fact that Hantzschhas in some cases isolated yellow salts of aminoazo-compounds, andhas also obtained a tribromo-derivative of aminoazobenzene bythe direct action of the halogen in glacial acetic acid solution (Ber.,1908, 41, 1171, 1187, 2435); whilst, further, there is the well-knownaptitude for diazotisation of the aminoaza-compounds. In the lastcase one is inclined to formulate the reaction as taking placeaccording t o the scheme:C6H5-N:N*C6H,*NH3C1 + HO*N:O =2H,O + C6H,*N:N*C,H4*N( iN)ClCONSTITUTION OF AZO-COMPOUNDS.PART V.513although reaction in the sense :C6H5*NH*N:C,H4:NH2C1 + HO*NO =H,O + C,H5*NH*N:C6H4:N(:NOH)C1=2H,O + C,~,*N:N*C,H,*~(~~)C~does not seem so improbable when one considers the easy way inwhich the quinonoid hydrates of azophencls lose water, passingdirectly into the hydroxyazo-form :R*NHDN:C6H4:(OH), -+ R*N:N*C6H4*OH + H20.It is certainly highly improbable that the diazonium chloridespossess a similar structure to the salts of the parent aminoazo-FIG. 1.Oscillation frequencies.1600 18 2000 22 24 26 28 3000 32 34 36 38 4000 42 44 46 4800Benzeneazobenzenediazonium chloride.......................... Benzeneazophenyltrimethylammonium iodide.Aminoaxobenzene hydrochloride.compounds, for on adding nitrite to the bluish-red solutions of thelatter, the d o u r changes to an orange, which to the unassisted eyeappears rather more red in'shade than that of the free aminoazo-compounds or of azobenzene itself. In these circumstances itappeared very desirable to examine the absorption spectra, andfor this purpose we have compared benzeneazobenzenediazoniumchloride with the hydrochloride of aminoazobenzene on the onehand, and with benzeneazophenyltrimethylammonium iodide on theother.Comparison with other para-derivatives of azobenzene is alsopossible, and, as will be seen from the curve given in Fig. 1, benzene514 HEWITT AND THOLE: THE COLOUR ANDazobenzenediazonium chloride gives a curve which resembles some-what closely that given by benzeneazophenol (Tuck, Trans., 1907,91, 450).The head of the absorption band for benzeneazophenollies at an oscillation frequency of about 3000, whilst for thediazonium salt examined it is about 2950. Further, both of thesesubstances show an extension of the band towards the red end ofthe spectrum; a similar extension has also been observed withcertain aminoazophenols by Hewitt and Thomas (Zoc. cit.).On the same diagram will be found the curves for the absorptionspectra furnished by the hydrochloride of aminoazobenzene and bybenzeneazophenyltrimethylammonium iodide. The first of thesesubstances is almost without doubt of quinonoid structure, and itsabsorption, which closely resembles that of the hydrochlorides ofdimethylarriinoazobenzene and its hydroxy- and methoxy-derivatives,is absolutely different from that of the other two compounds, whichdo not, however, agree as closely between themselves as might havebeen expected for substances possessing similar structures. Ittherefore remains an open question whether the constitution of thediazonium salt is to be expressed by the formula:C,H,*N:N=C,H,*N( iN)CI,although it might.be urged that the difference in the groups*N( iN)Cl and -N( CH,),Clmight account for the comparatively minor differences in theabsorption spectra.Comparison of the absorption spectra of phenol (Baly andEwbank, Trans., 1905, 87, 1351) and benzenediazonium salts wouldhave been interesting, but since, according to Dobbie and Tinkler(Trans., 1905, 87, 273), the latter give absorption spectra similarto those of dilute solutions of the unstable diazotates, a referenceto the curve given in the latter case by these authors shows thatvery little aid can be expected.For the purposes of this work, aminoazobenzene was acetylated,the acetyl derivative recrystallised until of constant melting point,and hydrolysed.In examining its acid-alcoholic solution, hydro-chloric acid was added in such amount as not to produce anyfurther deepening of the red shade.The benzeneazophenyltrimethylammonium iodide was preparedby heating benzeneazodimethylaniline with methyl alcohol andmethyl iodide t o looo after keeping the mixture for eighteen hoursin the cold, and reaction was already practically complete. Theproduct was finally recrystallised from a large quantity of boilingwater.Benzeneazobenzenediazonium ChLloride.-The isolation of the soliddiazonium salt was effected by suspending 4.7 grams of aminoazoCONSTITUTION OF AZO-COMPOUNDS.PART V. 515benzene hydrochloride in 150 C.C. of alcohol (96 per cent.), andadding 2.4 grams of amyl nitrite diluted with 10 C.C. alcohol. Afterfifteen minutes’ stirring, any small solid residue was removed byfiltration and the diazonium chloride precipitated by ether, collected,washed with ether, and dried over sulphuric acid. The salmon-coloured powder obtained in this way proved, under the microscope,to consist of a mass of small but well-defined prisms:0.5506 gave 0.3211 AgCl. Cl=14*4.The salt dissolves in alcohol and water, showing considerablestability even in alcoholic solution.The absorption spectra wereobserved with freshly prepared alcoholic solutions, but even afterkeeping overnight a t the ordinary temperature a large amount ofthe diazonium salt was found to be still undecomposed. Onheating, the salt decomposes comparatively gently ; no detonationhas been observed, whilst the dry salt may be kept for monthswithout appreciable decomposition, still dissolving in water andcoupling with the usual azo-components. An aqueous solutionreacts, however, immediately with potassium iodide, giving p-iodo-azobenzene.PZatinichZoride.-This salt was prepared by precipitating anaqueous solution of the chloride with excess of chloroplatinic acid ;after washing and drying it formed a salmon-coloured powder.This salt also shows considerable stability, and decomposes in agentle manner when heated:C,,H,N,Cl requires Cl= 14.5 per cent.0.3228 gave 0.0768 Pt..Pt=23*8.(C12H9N4)2PtC16 requires Pt = 23.6 per cent.No ferrichloride has been obtained, but solutions of the chlorideyield a yellow precipitate with potassium dichromate. The salt,presumably b enzeneazob enzenediazonium dichromute, was washedand dried, but on account of the character of the detonation whichoccurred on heating, an analysis was not carried out.. Meldola(Trans., 1905, 87, 4) has also prepared this salt, but did not analyseit owing to its explosive properties.The absorption spectra are not such that the authors feel justifiedin drawing any rigid conclusion as to the constitution of thebenzeneazobenzenediazonium chloride ; i t is, however, a veryremarkable fact that the most stable diazonium salts are alwaysthose possessing a para-substituent, especially when the latterhappens to be an unsaturated and negative group.Reference maybe made to the cases of pnitrobenzenediazonium chloride, theacetylaminobenzenediazonium chloride described by Meldola (Proc.,1899, 15, 196; Trans., 1905, 87, l), and to the benzenesulphonyl-VOL. XCVII. M 516 HEWIT" AND THOLE: THE COLOUR ANDaminoaryldiazonium salts of Morgan and his co-workers (Trans.,1905, 87, 73, 921, 1302; 1906, 89, 4 ; 1907, 91, 1311, 1505, 1512;1908, 93, 602).Morgan and Alcock (Trans., 1909, 95, 1319) adopt a modificationof Cain's formula (Trans., 1907, 91, 1040) for the constitution ofdiazonium salts, but apparently do not attribute the stabilityconferred by negative para-substituents to any essential difference inconstitution.I f one follows Cain in ascribing to diazonium salts ahemiquinonoid structure, the possibility certainly arises that saltsof the negatively substituted type may be fully quinonoid instructure, and their greater stabi1it.y be thus accounted for :Benzenediazonium chloride (Cain) : H\/='-NCI. /\=/-g/-\p-Nitrobenzenediazonium chloride : 0: N=-/ k N C I .I \=/ II I 0Acylaminobenzenediazonium chloride : Alk*C(OH)*N=/=N\= I/=\Benzeneazobenzenediazonium chloride : c' H *N*N=(=/\=NCI.5~ AThus, whilst Cain represents the simple diazo-salts as analoguesof the labile " chinols," the above representation of the negativelysubstituted compounds resembles the constitution attributed to thefar more stable quinones.Iodoazob enzene.-In connexion with the work described in thispaper, iodoazobenzene was examined with a view to seeing if itwas capable of iodonium salt formation.As amino- and hydroxy-azobenzene readily yield quinonoid salts, so it is not impossible thatthe iodo-derivative may give corresponding salts containing tervalentiodine :Since, however, azobenzene itself gives unstable salts with acids,it might be argued that any salts yielded by iodoazobenzene shouldbe referred to the same type. It happens that azobenzene gives adifferent absorption in acid solution from that given by the mono-hydrochlorides of its amino- and hydroxy-derivatives ; the absorptionspectra of the two latter salts being comparable, although the headof the band of slowest frequency lies always more towards the redend of the spectrum in the case of the nitrogen compound.When iodoazobenzene is dissolved in benzene and hydrogeCONSTITUlLON OF AZO-COMPOUNDS.PART V. 517chloride led in, the colour darkens considerably, and separation ofa certain amount of solid occurs. This method of preparing thesalt, which gives such good results for hydroxyazo-compounds, isnot very suitable for preparing salts of iodine derivatives, as theseparation of solid matter is but slight. Nevertheless, a comparisonof the absorption spectra of iodoazobenzene in alcohol alone and inalcoholic solution of hydrogen chloride (see Fig. 2) renders it quiteprobable that combinat ion with h j drogen chloride occurs with formationFIG. 2.Oscillation frcqzmacies.1600 18 2000 22 24 26 28 3000 32 34 36 38 4000 42 44 46 48001 1Iodonzobenzene in alcohol.lodoambenzene in alcoholic hydrogen chloride.of an iodonium salt. I n fact, when iodoazobenzene is moistened w i t hbenzene and exposed to gaseous hydrogen chloride, two moleculesof the latter are absorbed, and a beautiful, nearly black, crystallinehydrochloride is produced.The authors acknowledge with thanks the aid afforded them bythe Government Grant Committee of the Royal Society, by whichthe expenses of this research have been defrayed.EAAT LONDON COLLEGE