PHENOL TAUTOMERISM By R. H. THOMSON B.Sc. PH.I). F.R.I.C. (UNIVERSITY OF ABERDEEN) CARBONYL compounds exist almost exclusively in the keto-form unless structural features are present (e.g. hydrogen bonding or steric hindrance) which stabilise the enol form as in the P-diketone (I) or the complex enol (11). On the other hand simple aromatic enols i.e. phenols exist ex- clusively in the enolic form but ketonic properties become more pronounced in polyhydric phenols and especially in the hydroxyl derivatives of polycyclic aromatic hydrocarbons. This Review deals with the tautomeric behaviour of aromatic enols particularly those which can be isolated in the ketonic form. The hydroxyl derivatives of thiophen and furan are included but not heterocyclic compounds which show lactam-lactim tautomerism.Hydr0xybenzenes.-Summation of the bond energies for the system )CH-C=O and )C-C-OH shows that the keto-structure is more stable than the enol by ca. 18 kca,l./mole. In phenol (111) however the energy gained by rearrangement to the hypothetical keto-structures (IV) is offset by the simultaneous large decrease in resonance energy (ca. 35 kcal./mole). These figures although only approximate indicate that phenol must be 1 1 OH 0 0 (111) (IV 1 predoniinantly enolic. [Conant and Kistiakowsky calculated the free energy of enolisation of (IV) to be - 18.6 kcal./mole.] There is in fact no evidence for the existence of the keto-forms although the presence of a minute proportion of them in equilibrium with the " enol " (111) is not ruled out (cf. a ~ e t o n e ~ at the other extreme where the enol content is only 2.5 x Fuson Gorse and McKeever J .Anzer. Chem. SOC. 1940 62 3250. Branch and Calvin " The Theory of Organic Chemistry " Prentice-Hall Inc. Conant and Kistialiowsky Chem. Rev. 1937 20 181. Schwarzenbach and Wittwer Helv. Chim. Acta 1947 30 669 New York 1946. 27 28 QUARTERLY REVIEWS The iiitroductioii of additional enolic centres should assist the develop- iiient of ketonic character since the energy which accrues in the formation of multiple keto-groups would compensate for the loss of resonance stabilisa- tion. However a survey of the literature reveals no clear evidence that ketonic forms are present in solution even in the most favourable case of phloroglucinol. (The infrared spectra 5 establish that the solids are entirely enolic.) Hayashi 6 claimed that carbonyl bands in the Raman spectrum of resorcinol (in methanol solution) confirm the existence of the ketonic form but as the measurements were obtained by irradiating the resorcinol solution for periods of 17 hours upwards during which it became pale red this claim is open to doubt.The ultraviolet spectrum of resorcinol is that of a normal phenol.' Sonn and Winter 8 concluded from bromina- tion experiments that phloroglucinol existed in alcoholic solution in a tautomeric monoketo-form but evidence based solely on chemical reactions is inconclusive and supporting physical evidence is lacking. As judged by the infrared spectrum phloroglucinol is entirely enolic in dioxan sol~tion.~ The classical evidence for the tautomeric nature of phloroglucinol is the formation of a trioxinie,10 the structure of which has recently been con- firmed,g and if hydroxylamine is replaced by ammonia reaction under similar conditioiis (long storage in a closed vessel at room temperature) yields 3 5-dihydroxyaniline and 3 5-diaminophenol.ll As these experi- ments with phloroglucinol were done in a basic medium the presence of the mesomeric anion (V) may be inferred but the free existence of cyci'o- hexanetrione molecules is doubtful.The only true ketones obtained from phloroglucinol are polysubstituted derivatives in which one or more of the 5 Barnes Gore Liddel and Williams " Infrared Spectroscopy " Reinhold New York 1944 ; Randall Fowler Fuson and Dangl " Infrared Determination of Organic St,ructures " Van Nostrand New York 1949. 6 Hayashi Sci. Papers Inst. Phys.Chenz. Res. (Tokyo) 1933 21 69. 7 Friedel and Orchin " Ultraviolet Spectra of Aromatic Compounds " Wiley New * Sonii a i d Winter Ber. 2928 61 2303. 8 Farmer arid Thonison Chem. c u i d Irtd. 1956 86. lo Baeyer Ber. 1886 19 159. 11 Pollak Moncitah. 1893 14 419. York 1951. labile hydrogen atoms has been replaced. Thus methylation with methyl iodide and aqueous potassium hydroxide gives a mixture of O-Me and C-Me compounds leading ultimately to the hexamethyl ketone (VI). l2 Chlorina- tion of phloroglucinol also gives a hexachlorinated end-product (VI ; C1 in place of Me).13 .Similar gem-substitution occurs in resorcinol but in phenols which lack a vicinal carbon atom activated by two enolic centres alkylation is restricted to simple C-substitution by the Claisen procedure using reactive ally1 or benzyl halides.14 The mesomeric anion (VII) of resorcinol contains an q?-unsaturated carbonyl system the existence of which is neatly demonstrated by reduction with sodium amalgam to cycZo- hexane-1 3-dione (dihydroresorcinol).l5 Hydrogenation over nickel in alkaline solution is more efficient. l 6 In a search for further evidence on the tautomeric nature of polyhydric phenols Fuchs 17 studied the addition of sodium hydrogen sulphite to quinol resorcinol and phloroglucinol. After prolonged reaction adducts were obtained from quinol and resorcinol to which he assigned the struc- tures (VIII) and (IX) respectively.* (An adduct C6H60,,3NaHS0 from phloroglucinol was not obtained pure.) These structures imply that \/ HO/\SO,Na (VIII) (1x1 addition has occurred at the ethylenic double bond as well as a t the tauto- meric carbonyl groups which is plausible in the case of resorcinol but the structure (VIII) is doubtful.Apart from this work there are no other indications that quinol can react in a tautomeric form -f and the product obtained by Fuchs gave a transient blue ferric chloride colour suggesting the presence of a quinolsulplionic acid. Similar bisulphite compounds are of course intermediates in the Bucherer reaction which is of great technical importance in the naphthalene series.ls Prom a study of the reaction kinetics Cowdrey and Hinshelwood l9 suggest that bisulphite adds to il C=C and not to a C=O linkage. This seems to be a necessary postulate l2 Herzig and Erthal Monatsh. 1910 31 827. l3 Zincke and Kegel Ber. 1889 22 1467. l5 Merling ibid.1894 278 28. Claisen Annalen 1925 442 210. Fuchs and Elsner Ber. 1919 52 2281 ; 1920 53 886 ; Fuchs B e y . 1921 54 Reviewed by Drake '' Organic Reactions " Wiley New York 1942 Vol. I l6 Org. Xynth. Coll. Vol. 111 p. 278. 245. Chapt. 6. l8 Cowdrey and Hinshelwood J. 1946 1036 ; Cowdrey J. 1946 1041 1044. * The compounds were originally written with sulphite ester groups t A number of substituted quinols can be obtained in both onol and keto-form Tsutomerism is restricted to keto -+ enol conversion. > C(OH)*O*SO,Na. by independent methods. See p. 37. 30 QIJARTERLY REVIEWS in Bucherer reactions involving secondary amines and may well be true in other cases hut until we know more about the reaction of bisulphite with aromatic compounds in general addition products such as those of Fuchs are of limited value in the study of tautomerism.Bucherer and Schenkel 20 showed long ago that even pyridine forms an adduct with bisulphite of formula C5H5N,3NaHS0,,2H20 ; the nature of this com- pound is obscure but the pyridine molecule was profoundly modified as the adduct readily liberated ammonia on treatment with alkali. Many reactions of monohydric phenols involve the tautomeric form which becomes obvious if the labile hydrogen is replaced by an entering substituent thus Tautomerism is by no means confined t o the polyhydric phenols. enabling a ketonic product to be isolated. A well-known example is the Reimer-Tiemann reaction tjhe first step of which can be considered a special case of C-alkylation. The reaction is not impeded by small alkyl groups in the o- and p-positions so that e.g.o-cresol yields 21 the ketone (X) as well as the hydroxy-aldehyde (XI). The formation of a " blocked tautomer " in this way has been used by Woodward 22 as a method for the introduction of angular methyl groups. Thus the hydroxy-aldehyde formed in the reaction of 6-hydroxytetralin with chloroform in alkaline solution is accompanied by the dienone (XII) which may then be hydro- genated to it methyldecalol. Ar c N.0' Another old reaction in this category is the formation of indophenols by condensation of arylnitIroso-compoui-ids with pheiiols under basic con- dit,ions. This reaction is of practical importance in the preparat'ion of Zo Bucherer and Schenkel Ber. 1908 41 1346. 21 Auwers and Keil Ber. 1902 35 4201. 22 Woodward d . Awer. Chenz. SOC. 3940 62 1208.THOMSON PHENOL TA'ZJTOMERISM 31 oxazine dyes from p-nitroso-dialkylanilines ; with the latter basic condi- tions are obviously unsuitable and usually the phenol is condensed with the hydrochloride of the nitrosoamine in acetic acid. 23 More recently it has been shown that ketones may be formed by oxida- tion of monohydric phenols with lead tetra-acetate. 2 4 Here the essential intermediate is a mesomeric radical produced by dehydrogenation of a suitably substituted phenol e.g. (XIII). The final products depend upon the structure of the phenol and also on the solvent; dimerisation occurs in benzene solution but in acetic acid which assists the propagation of acetate radicals acetoxylation predominates. Phenol itself affords only a M e M e Me' h Me dimeric product (4 4'-dihydroxydiphenyl) but with increasing o p - substitution this reaction diminishes and with mesitol only acetoxylation occurs irrespective of solvent.Attack by acetate radicals a t substituted 0- and p-positions stabilises the dienone system 2 4-dimethylphenol yield- ing the ketones (XIV) and (XV). Similar products are obtained by oxidis- ing phenols with organic peroxides. 25 A related reaction is Bamberger's oxidation 26 of p-alkylated phenols to y-qixinols [e.g. (XVI) from mesitol] by using Caro's acid in the presence of magnesium carbonate where the substitution is presumably effected by *O*SO,- radical ions. Hydroxynaphthalenes (excluding 1 4-naphthaquinols) .-The tautomeric properties of the hydroxynaphthalenes are one of the many indications that naphthalene is less aromatic than benzene.The naphthols are in general much more reactive than phenol and share many reactions in common with resorcinol which is to be expected in view of the large resonance stabiliaation of the ketonic structures (XVII) as compared with (IV). There are however a few limitations due to bond fixat,ion as seen 2 3 Koechlin and Witt G. P. 15,915 ; Mohlau Ber. 1883 16 2843 ; 1892 25 1056. 2 4 Wessely et al. Monatd~ 1954 85 69 and previous papers ; Cavill Cole Gilham 25 Cosgrove and Waters J. 1951 388 ; Campbell and Coppinger J . Avzer. C'hem. 26 Bamberger Ber. 1903 36 2028. and McHugh J. 1954 2785. Soc. 1952 74 1469 ; Wessely and Schinzel Monatsh. 1953 $4 425 969. 32 QTTARTERLY REVTEWS in the failure of the naphthol (XVIII) to condense wit'h p-nitlrosodiniethyl- aniline.27 QJ-J* ~ &OH Me Numerous gem-substituted derivatives are known especially of /?- naphthol. A feature of these ketones stitnents arc eliminated to restore the is the ease with which blocking sub- more stable fully aromatic st,riicture. This tendency is demonstrated by the formation of 1 -nitro-2-naphthol by reaction of the ketone (XIX) in acetone with aqueous sodium hydroxide.25 In an extensive investigation of the halogenonaphthols Fries 29 and his associates showed that bromination of @-naphthol in acetic acid was a reversible reaction and for the gem-dibromo-ketone (XX) this results in rearrangement to 1 6-dibromo-2-naphthol when the ketone is treated with hydrogen bromide in acetic acid. Further debromination of the dibromo- naphthol (to 6-bromo-2-naphthol) is then possible.The selective removal of o-halogen substituents by acid reducing agents (commonly acid stannous chloride) is a useful reaction in the naphthol series. Nicolet 3O showed the reaction to be independent of the stannous chloride concentration and the first step is clearly the acid dehalogenation shown above. 27 Wedekind Ber. 1898 31 1675. 28 Fries Annalen 1912 389 316. 29 Fries and Schimmelschmidt ibid. 1930 484 245. 3u Nicolet J . Amer. Chem. SOC. 1927 49 1810. THOMSON PHENOL TAUTOMERISM 33 Ketones can frequently be obtained by reduction of naphthols under basic conditions the nearest parallel t o the sodium amalgam reduction of resorcinol being the formation of the diketone (XXI) by reduction of naphtharesorcinol with Raney nickel alloy and aqueous has not been adequately characterised.Reduction of 1 5- and 1 6-dihydroxynaphthalenes by the same method gives 5- and 6-hydroxytetralones respectively but a- and @-naphthol are unaffected. However reduction of /3-naphthol with (XXI) sodium and liquid ammonia in the presence of pentyl alcohol gives P-tetral~ne,~~ and the latter can also be obtained by hydrogenation of ,&naphthol over palladium-charcoal in the presence of a base.33 Hydroxythi0phens.-Since tautomerisation in phenol is restricted by resonance stabilisation it would be expected that enols which form part of less aromatic systems might in some cases exist in the ketonic form. This is not found in the naphthol series but is exemplified by the hydroxy- thiophens. 2-Hydroxythiophen 34 (XXII) is a weak acid it gives a red ferric chloride colour couples in alkaline solution with diazo-compounds FJOH CIA Or lT)=o )CH2-CH,*C0,H sodium hydroxide 31 although unfortunately the compound 0 o$.o HO--OH S S S S /-O (XXII) (XXIII) (XXIV) and forms an acetate benzoate and methyl ether; on the other hand in the tautomeric keto-form (XXIII) it yields a benzylidene derivative and readily undergoes hydrolytic fission (the keto-form is of course a thio- lactone).Its ultraviolet absorption (in aqueous solution) is in accord with the ketonic structure whilst the infrared spectrum of the liquid itself shows carbonyl and hydroxyl absorption and peaks corresponding to both ali- phatic and aromatic C-H bonds. These properties amply demonstrate the existence of a tautomeric equilibrium in marked contrast to phenol. It is interesting too that the enol derivatives (e.g.the acetate and the methyl ether) of 2-hydroxythiophen and of phenol show a close physical resemblance (b.p. odour) typical of thiophen and benzene analogues but the parent hydroxy-compounds themselves are physically dissimilar. 3-Hydroxy- dhiophen 34 is less stabie and has not been isolated in a pure state. 3 4- Dihydroxythiophen is known only as its dibenz0ate,~5 but the compound (XXIV) has been obtained and is as expected entirely en01ic.~~ Normally one carbonyl group of a cyclic o-diketone is enolised as in this form the repulsion of the cis-carbon-oxygen bonds is reduced 37 and in a thiophan 31 Papa Schwenk and Breiger J. Org. Chem. 1949 14 366. 32 Birch J. 1944 430. 33 Stork and Foreman J. Arner. Chem. Soc. 1946 68 2172. 3 4 Hurd and Kreutz ibid.1950 72 6543. 35Fager ibid. 1945 67 2217. 36 Karrer and Kehrer Helu. Chim. Acta 1944 27 142. 37 Dewar " The Electronic Theory of Organic Chemistry " Oxford 1949 p. 102. c 34 QUARTERLY REVIEWS ring greater stability is achieved by complete eiiolisation giving a fully conjugated system. Hydroxy€urans.-Much less is known about the hydroxyl compounds in this series. They appear to be predominantly ketonic in no way re- sembling phenols and with the exception of the 2-hydroxyfurans (XXV) [which exist as but-B-enolides e.g. (XXVI)] and the 2 5-dihydroxyfurans (i.e. succinic anhydrides) the majority are unstable. This agrees with the low resonance stabilisation of the furan ring. It is doubtful if the mono- hydroxyfurans theinselves have ever been obtained as such although both have been claimed.38 The substances were described as crystalline solids .~ ( )"O 0 (XXV) (XXVI) (XXVII) which darken and resinify spontaneously form dark solutions in aqueous alkali and do not reduce Tollens' reagent ; yet in spite of their instability both can be nitrated and then reduced to give seemingly stable amino- hydroxy-comp~unds.~~ 3-Hydroxyfuran forms an adduct with maleic anhydride but khere is no real evidence for the structure of the isomeric compound (prepared by fusion of 5-sulphofuran-2-carboxylic acid with sodium hydroxide in the presence of potassium chlorate) which appears t o be unlike its homologues. Cleavage of the acetate of 2-hydroxyfuran (XXV) (obtained by pyrolysis of 2 5-diacetoxy-2 5-dihydrofuran) yields crotono- lactone [double-bond isomer of (XXVI)].This occurs to some extent during the preparation 40 of 2-acetoxyfuran and again on its treatment with chlorine or bromine (at - 5" to - lo") halogenocrotonolactones result.41 The free existence of the enol (XXV) therefore seems very unlikely. Most known 3-hydroxyfurans are highly substituted and appear t o have the properties of aliphatic keto-enol systems but physical data are lacking. 3 4-Dihydroxyfuran has not been prepared but the compound (XXVII) is known 42 and exists essentially in the enol form. In addition to the normal enolisation of cyclic o-diketones this compound which can be regarded as two linked p-keto-ester systems is stabilised by chelation. Hydroxythionaphthens and Hydroxybenzo€urans.-As we have seen fusion of a second benzene ring on to phenol leads to increased ketonic character and the effect on the hydroxythiophens and hydroxyfurans is the same the benzo-analogues being best described as thionaphthenones and benzofuranones (coumaranones and isocoumaranones) .Both groups show the characteristic chemical properties of keto- and enol forms with the exception of the isocoumaranones which behave as lac tone^.^^^ 44 They 38 Hodgson and Davies J . 1939 806. 39 Idem J . 1939 1013. 40 Clauson-Kaas and Elming Acta Chem. Scand. 1952 8 560. 41 Elmicg and C lauson-Kaas ibid. p. 565. 4 2 Hoehn Iowa State Coll. J . Xci. 1936 11 66. 43 Hartough and Meisel " Compounds with Condensed Thiophen Rings " Inter- 4 4 Elderfield " Heterocyclic Compounds " Wiley New York 1951 Vol. 11. science New York 1954. THOMSON PHENOL TAUTOMERISM 35 probably all exist in the keto-form in the solid state ; the benzofuranones remain in the keto-form in solution but there is no reliable information concerning the thionaphthenones which may like the hydroxythiophens form an equilibrium in solution.Bromine estimations showed coumar- anone 45 to be almost entirely ketonic and thionaphthen-3-one 46 to contain 5% of the enol form in solution but again this is chemical evidence only. Marschalk *' apparently obtained thionaphthen-2-one in two modifications. Distillation of a sample of m.p. 44-45' gave a product of m.p. 33-34' and in one experiment he was able to convert this back into the higher- melting form by dissolution in aqueous sodium hydroxide and acidification. This suggesta that the compound m.p. 44-45" is the enol and the com- pound m.p.33-34' is the keto-form but apparently it is only the latter which gives a ferric chloride colour (blue). There is evidently some con- fusion here but it is just possible that Marschalk did isolate two tautomeric forms although he did not make this claim himself. If so this is the simplest aromatic enol known in both forms. isocoumaranone is known in two crystal modifications but these are not tautomeric forms.48 1 rl-Naphthaquinols.-In the naphthalene series when two hydroxyl groups are present in the same ring it becomes possible to isolate the tauto- meric diketo-form provided that both carbonyl groups are conjugated with the second benzene ring as in (XXIX). [The gain in bond energy as a result of ketonisation (m. 2 x 18 kcal./mole) is approximately equal to the loss of resonance energy of one benzene ring.] Both the dienol (XXVIII) and the diketone (XXIX) are stable compounds under normal 6 OH 0 03 0 (XXVIII) (XXIX) conditions and can be crystallised unchanged.A spectroscopic examina- tion 49 has shown that the diketone exists as such both in the solid state and in solution and there is no indication of the existence of an equilibrium a t ordinary temperatures. However in the estimation of naphthaquinol by ceric sulphate titration Braude et aL50 found that the titre gradually decreased to about 75% of its original value when the quinol was heated in phenetole at 131' tautomerisation would account for this but other factors may be concerned as the proportion of the keto-form seems rather high. The diketone * (XXIX) was first obtained by Madinaveitia and 45 Auwers and Auffenberg Ber.1919 52 92. 46 Auwers and Theis Ber. 1920 53 2285. 47 Marschalk J. prakt. Chem. 1913 88 227. 48 Stoermer Annalen 1900 313 79. 49Thomson J. 1950 1737. 5OBraude Jackman and Linstead J. 1954 3548. * Formulated as a monoketone by Olay Rev. Acad. Cienc. Madrid 1935 82 384. 36 QUARTERLY REVIEWS Olay 51 by fusion of the dienol (XXVIII) at ca. 210" in a vacuum followed by rapid cooling to " freeze the equilibrium " ; the diketone (2 3-dihydro- naphthaquinone) was then separated by chloroform extraction. In this way about 10% of the dienol can be isolated in the tautomeric form.49 All the simple 2 3-dihydronaphthaquinones have been obtained by Olay's procedure but the method is limited as some quinols decompose on being heated. A possible alternative is the catalytic method employed by Grob et u Z .~ ~ in the isomerisation of the enol (XXX) to the ketone (XXXI). OH 0 HN-CH In this case direct heating (at temperatures below the decomposition point) effected no rearrangement but by heating the phenol with palladium- charcoal in xylene-tetralin the ketone was obtained in 80% yield. Although keto-enol changes (e.g. carvone + carvacrol 53) and dienone -+ phenol rearrangements 54 have been brought about by heating with palladium- charcoal this enol + keto conversion appears to be unique. Rearrange- ment of the naphthalene system (in XXX) to the dihydroindole system (in XXXI) (AH ca. 4.5 kcal./mole) is possibly more favourable energetically than the conversion of the naphthalene system (in XXVIII) into the benzene system (in XXIX) but nevertheless this technique merits investigation.The reverse keto -+ enol change (XXIX -+ XXVIII) is readily effected by dissolution of the diketone in cold alkali and in practice enolisation occurs frequently under reaction conditions. Certain reactions of naphthaquinols proceed via the diketo-forms although this does not establish the existence of a tautomeric equilibrium. A number of substituents including halogen NHPh SR S02R SO,H and (in some cases) OH can be removed from positions 2 and 3 of 1 4-naphthaquinones by reduction with acid stannous chloride the first step being a rapid reduc- tion of the quinone to the quinol (XXXIII).55 A few o-substituents can also be removed from naphthols of the type (XXXII ; R = halogen SO,H and probably SR *) under the same conditions but the others are stable unless a second hydroxyl group is present as in (XXXIII).It is likely therefore that the reaction proceeds by tautomerisation to the 51 Madinaveitia and Olay Anal. Pis. Quim. 1933 31 134. 5 2 Grob and Voltz Helv. Chim. Acta 1950 33 1796 ; Grob and Hofer ibid. 1952 53 Linstead Michaelis and Thomas J. 1940 1139. 5 4 Homing J. Org. Chem. 1945,10 263 ; Leonard and Berry J . Amer. Chem. SOC. 5 5 Bruce and Thomson J. 1954 1428. * l-p-Tolylthio-2-naphthol can be reduced to j5-naphthol with stannous chloride. 35 2095. 1953 75 4989. THOMSON PHENOL TAUTOMERISM 37 diketone (XXXIV ; R = OH NHPh or S0,R) followed by acid-catalysed elimination of the substituent R as shown. .1 (XXXII) (XXXIII) (XXXIV) (XXVIII) Another reaction which involves the ketonic forms of naphthaquinols is the formation.of tetralins by Clemmensen reduction. Tetralin can also be obtained less readily by Clemmensen reduction of cc- and ,&naphthol and Madinaveitia 56 considered the rate of reduction to be proportional to the ease of tautomerisation. The diketone (XXIX) forms a bis-p-nitro- phenylhydrazone attempts to obtain this by reaction of the dienol (XXVIII) with p-nitrophenylhydrazine produce only 4-p-nitrophenylazo- l-naphthol the initial hydrazone formed by condensation at one enolic centre being oxidised in the tautomeric hydrazo-form. A number of naphthaquinols are known in both keto- and enol forms which are not interconvertible 2 3-dichloronaphthaquinol (XXXV) and 1 4-naphthaquinone dichloride (XXXVI) are typical. Enol -+ keto con- version cannot be brought about by fusion as the dienol decomposes a t high temperatures and the reverse change is restricted by the tendency of the diketone to form a quinone by elimination of hydrochloric acid.This is facilitated by both acids and bases so that in ketones of type (XXXV) (XXXVI) (XXXVII) (XXXVIII) (XXXIX) (XXXVI) enolisation and elimination reactions which proceed via similar ionic intermediates are in competition and the result depends upon the structure of the diketone and sometimes on the catalyst. I n naphthalene compounds elimination predominates but concurrent enolisation and elimination has been observed in a few instances whereas the corresponding benzenoid compounds show a greater tendency to rearrange to aromatic structures as shown by the natural product gliorosein (XXXVIII or XXXIX) which rapidly enolises in a basic medium.57 Benzoquinone dichloride slowly forms the corresponding dienol diacetate when warmed with acetic anhydride in the presence of sulphuric acid but if the catalyst is changed to toluene-p-sulphonic acid the elimination proceeds more 66 Madinaveitia Anal. PLs. Quim. 1934 32 1100. 57 Vischer J. 1953 815. 38 QUARTERLY REVIEWS rapidly and the product (after re-addition of the hydrochloric acid) is 2 5-dichloroquinol diacetate. Under the same conditions the dichloride (XXXVI) gives only 2-chloro-1 4-naphthaquinone. Again when benzo- quinone dichloride is suspended in acid stannous chloride at 0" both elimination and enolisation occur and a mixture of 2-chloro- and 2 3- dichloro-quinol is obtained. 58 Amongst related naphthalene compounds Russian workers 59 have shown that 2 3-epoxynaphthaquinone in hot aqueous solution gives via the glycol (XXXVII) a mixture of 2-hydroxy- 1 4-naphthaquinone (by elimination) and 2 3-dihydroxy-1 4-naphtha- quinone (by enolisation and aerial oxidation) and the latter is also obtained by aeration of the diacetate of (XXXVII) in alcoholic potassium hydroxide.6O Blocked tautomers are also found in this group but whereas they are obtained from monohydric phenols by substitution reactions here they are usually formed by quinone addition reactions e.g. addition of sodium hydrogen sulphite to 2-methyl-1 4-naphthaquinone gives both the quinol (XL) and the diketone (XLI).61 The compound (XLII) obtained by addition of hypochlorous acid to 3-hydroxy-2-methyl- 1 4-naphthaquinone (XL) (XW (XLII) (XLIII) is remarkable e 2 in that it does not revert to a quinone when heated in vacuo but loses water to form the triketone (XLIII).I n one instance a non- enolisable ketone has been obtained in the diphenyl series by a quinone addition reaction addition of hydrochloric acid to the diquinone (XLIV) 0 0 0 0 (XLIV) (XLV) gives the adduct (XLV) (and not the isomeric diquinol) which in spite of the extended conjugated system in the compound is very unstable and on attempted crystallisation dissociates into the original components.63 The introduction of hydroxyl groups at the peri-positions of 2 &&hydro- 58 Dimroth Eber and Wehr Annalen 1925 446 132. 59 Shchukina Khokhlov and Shemyaliin J . Gen. Chern. (U.S.S.R.) 1951 21 908. 6o Shchukina Vinogradova and Shemyakin ibid.p. 1661. Cormack Moore and Balis J. Arne?. Chem. SOC. 1950 72 844 ; Moore and Washbum ibid. 1955 7'7 6384. 6 2 Shvetsov and Shemyakin J . Cew. Chem. (U.S.R.R.) 1949 19 480. 63 Lindberg A d a Chem. Scand. 1951 5 885 ; Erdtrnan Proc. Roy. SOC. 1933 A 143 191. THOMSON PHENOL TAUTOMERISM 39 naphthaquinone as in F-hydrojuglone (XLVII ; R = H) and p-hydro- naphthazarin (XLVII ; R = OH) stabilises the diketo-system,* in part R OH HO OH 0 ry) )I Ho 0 (XLVI) (XLVII) by strong hydrogen bonding and lowers the activation energy for the enol-+ keto change so that these ketones can be obtained under much less vigorous conditions. Thus when a solution of a-hydrojuglone (XLVI ; R = H) in dilute hydrochloric acid (containing a little stannous chloride to prevent oxidation) is warmed it becomes yellow and the keto-form can be isolated by chloroform extraction.If the colourless solution is then set aside it slowly becomes yellow again as the equilibrium is restored.? The tetrahydroxynaphthalene (XLVI ; R = OH) behaves in the same way but as this compound is very susceptible to oxidation p-hydronaphthazarin is normally made by reduction of naphthazarin (5 8-dihydroxy-1 4- naphthaquinone) in hot acid stannous chloride solution from which it crystallises on cooling (90% yield 64). Elimination of substituents from naphthaquinols proceeds much more readily when a peri-hydroxyl group is present and a new feature arises in the case of substituted a-hydronaphth- azarins.65 66 The compound (XLIX) can tautomerise in two ways the reaction being controlled by the substituent.It can be seen that tauto- merisation at one enolic centre will be opposed by the + T effect of the group R and consequently rearrangement will occur preferentially in the other ring. Hence in the reduction of naphthazarins where R = OH or NHPh ketones of type (XLVIII) are formed but where R = SO,R the alternative (L) is produced and the substituent is then eliminated to give 6 4 Wheeler and Edwards J . Amer. Chem. SOC. 1916 38 387. 6 5 Bruce and Thomson J. 1952 2759. 66 Idem J. 1955 1089. * The 7-methyl derivative of (XLVII ; R = H) occurs in Nature (Cooke Dowd and Webb Nature 1952 169 974; Cooke and Dowd Austral. J . Sci. Res. 1952 5 A 760 ; Austral. J . Chem. 1953 6 53). -f Olay (Rev. Acad. Cienc. Madrid 1935 32 384) found (by Meyer estimation) that 2-methyl-p-hydrojuglone was 43 yo ketonic a t equilibrium and 1 4-naphthaquinol 10% ketonic.Both figures are probably too high but they illustrate the difference between the two compounds. 40 QUARTERLY REVIEWS @-hydronaphthazarin as the final product. Weaker + 111 groups (SR and halogen) are also eliminated but this of course occurs in simple naphthols. Hydroxyanthracenes.-In the anthracene series the general pattern is very similar to that seen in the naphthalene compounds the hydroxyl derivatives showing a greater tendency to exist in the tautomeric keto- form. Of the monohydroxy-compounds a- and P-anthranol are very like a- and @-naphthol but in 9-anthranol (LI ; R = H) we have the simplest monohydroxy-aromatic compound (excluding 2-hydroxythionaphthen) which is stable in both tautomeric forms.The keto-form (LII) (cf. benzophenone) is the more stable which implies that the resonance energy associated with the central " benzene " ring is < 18 kcal./mole in accord with the absence & R OH QpJJ R 0 Ph 0-OH of aromatic properties. I n his classical work Meyer 67 showed that both forms tautomerise slowly in solution (in the absence of catalysts) the equilibrium attained being always predominantly ketonic. On slow cool- ing a melt crystallises as anthrone but after rapid cooling some anthranol is also present. The chemical properties of the two forms are quite distinct a t ordinary temperatures the keto-form being comparatively unreactive in the absence of enolising catalysts. It was observed by Julian et u Z . ~ ~ that 10-alkyl(and -aryl)-anthranols (LI) readily took up atmospheric oxygen to form peroxides considered to have the transannular structure (LIII) as they could be reduced catalytically to the corresponding 10-alkyl- & Ph 0 oxanthranols and on pyrolysis they yielded anthraquinone and the corre- sponding alcohol.The transannular structure was disputed by Dufraisse et and by a detailed study of the oxidation product of 9-phenylanthranol 68 Julian and Cole J . Amer. Chem. SOC. 1935 57 1607 ; Julian Cole and Diemer 69 Dufraisse Etienne and Rigaudy Bull. Soc. chim. (France) 1948 804. Meyer Annalen 1911 379 37. ibid. 1945 67 1721. THOMSON PHENOL TAUTOMERISM 41 they established the keto-hydroperoxide structure (LIV). These anthranols therefore fall into line with a number of other aryl-substituted enols which form hydro peroxide^.^^ Of particular interest is the formation of the keto- hydroperoxide (LIV) by aeration of an alkaline solution of the anthranol which involves oxidation of the mesomeric anion (LV) as shown.1 4-Dihydroxy- anthracene is converted into the diketo-form (LVI; R = H) by fusion (50% yield),65 and the diphenyl derivative (LVI ; R = Ph) is formed merely by treating a solution of 1 4-dihydroxy-9 10-diphenylanthracene with hydrochloric acid.71 As in the naphthalene series very stable diketones are formed when peri-hydroxyl groups are present. ZeucoQuinizarin (LVI ; R = OH) is easily obtained by reduction of quinizarin in hot acid solution 7 2 it enolises in alkaline solution but the tetrahydroxyanthracene obtained on acidification rearranges to the more stable diketone during crystallisation.The structure of the compound (LVI; R = OH) is established by its synthesis from naphthaquinol and succinic anhydride 73 and by its infrared specfrum,7* but the structures of the diacetate and dimethyl ether are uncertain. By treatment of Zeucoquinizarin with acetyl chloride in cold pyridine the diacetate can be formed without enolisation of the carbonyl Relatively little is known of the dihydroxyanthracenes. ( L W (LVII) groups (this is verified by acetylation of /I-hydronaphthazarin and /I-hydro- juglone in the same way). The identical diacetate can also be obtained by reduction of quinizarin diacetate in the cold with zinc and acetic acid followed by warming of the solution of 1 4-diacetoxyanthraquinol in an inert atmosphere. Zahn and Ochwat 73 proposed structure (LVII ; R = Ac) for this compound which implies that Zeucoquinizarin reacts in the tautomeric form (LVII ; R = H) in the first method of preparation and a rather improbable tautomerisation occurs in the second.The alterna- tive structure (LVI ; If this is correct migration of the acetyl groups to peri-positions must have occurred in the course of the synthesis from quinizarin diacetate; there are several pre- cedents for this similar migration of acyl groups has been observed under various conditions in glyceride^,'^ o-dihydroxyanthraq~inones~76 and 'O Criegee in Houben-Weyl " Methoden der Organischen Chemie " Georg Thiem Verlag Stuttgart 4th Edn. 1952 Vol. 8 p. 25. 71 Bichet Ann. Chim. (France) 1952 7 235. 7 2 Meyer and Sander Annulen 1920 420 113. 73 Zahn and Ochwat ibid.1928 462 72. 7 4 Flett J. 1948 1441. 75 Daubert and King J. Amer. Chem. Soc. 1938 60 3003. '13 Kubota and Perkin J. 1925,127,1889 ; Perkin and Storey J. 1928,229 ; Perkin R = OAc) is more plausible. and Storey J. 1929 1399. 42 QUARTERLY REVIEWS in one instance in a peri-dihydr~xynaphthalene.~~ ZeucoQuinizarin dimethyl ether presents a similar problem it is formulated.73 as (LVII ; R = Me) but it is not clear how this arises by reduction of quinizarin dimethyl ether. In general #I-substituents (C1 OH S03H NHPh) 7 2 7 78 are very readily eliminated on reduction of quinizarins. There is however one case (purpurin) in which reduction (in the absence of strong acids) leads to retention of a P-substituent the product (LVIII) on subsequent treat- ment with strong acid or alkali is smoothly converted into quinizarin and no doubt analogues of (LVIII) could be obtained by reduction in the same way.0 HO II NH HO 11 (LVIII) (LIW It is relevant here that certain aminoanthracenes have tautomeric properties not shown by the aminobenzenes and aminonaphthalenes. 9-Anthramine 79 is not unlike anthranol although it is known only in the amino-form,80 but imino-forms can be isolated if peri-hydroxyl groups are also present. The simplest example is Zeuco- 1 4-dianiinoanthraquinone (LIX). This can be obtained by hydrogenation of the diaminoquinone in the cold and then heating in an inert atm~sphere.~~ It is normally made on the large scale by heating Zeucoquinizarin with ammonia and the ready introduction of basic groups by direct condensation of primary amines with Zeucoquinizarin makes the latter a key intermediate in the manufacture (LX) (LX) of many aniinoanthraquinones.As the imino-groups are rather easily hydrolysed conversion of the Zeuco-compounds into the quinones by aeration of their alkaline solutions must be avoided and the usual procedure is to heat the Zeuco-compound in nitrobenzene preferably with an enolising catalyst. Again elimination of P-substituents can occur during the forma- tion of the Zeuco-amino-compounds and this extends also to heterocyclic derivatives. Reduction of the bromoanthrapyrimidine (LX) with acid 77 Hayes and Thomson J. 1955 904. 78 Marschalk Bull. SOC. chim. (France) 1927 41 943 ; G.P. 95,271. 79 Kauffler and Suchannek Ber. 1907 40 518 ; Meyer and Schlosser Ber. 1013 46 29. Craig and Short J. 1945 419. THOMSON PHENOL TAUTOMERISM 43 stannous chloride or sodium dithionite gives a bromine-free Zeuco-compound,*l presumably (LXI) and halogen is similarly lost in the reduction 82 of the phthaloylacridone (LXII) .In this case the stable Zeuco-compound formed must be (LXIII) or (LXIV). Hydro-1 Derivatives of More Complex Hydrocarbons.-In the higher polycyclic compounds keto-forms become increasingly stable. Compounds with hydroxyl groups located in a terminal ring are similar to cc- and p-anthranol but when an enolic centre occurs a t a meso-position in a linear hydrocarbon usually only the keto-form is known although tautomerism similar to the anthrone-anthranol system may occur in angular hydro- carbons. This distinction arises because the linear enol structures include more quinonoid rings and hence are less stable than the angular enol structures.* The simplest angular enol is 9-phenanthrol the keto-form of which is unknown (contrast anthranol) whereas the ketones (LXV) and (LXVI) are insoluble in boiling aqueous sodium hydroxide and their enols have not been isolated.On the other hand both tautomeric forms of the benzanthrone (LXVII) exist and can be crystallised unchanged from benzene although the enol isomerises fairly rapidly in acetone.83 0 (LXVII) Reports on Dyestuffs Intermediates etc. Microfilm P.B. 70,332 Reel lc Frames B.P. 587,006. 2 16-220. 83 Fieser and Hershberg J . Amer. Ghem. SOC. 1937 59 1028. * The para-localisation energies of a number of polycyclic hydrocarbons have been calculated by Brown ( J . 1950 691) and are in accord with the stability of the keto- forms.