THE CHEMISTRY OF THE DITERPENOIDS By D. H. R. BARTON PH.D. E.R.I.C. (IMPERIAL CHEMICAL INDUSTRIES RESEARCH FELLOW IMPERIAL COLLEGE OF SCIENCE AND TECHNOLOGY LONDON S.W.7) ALTHOUGH impure specimens of abietic acid and the primary resin acids were the subject of investigations more than a century ago a full understanding of the chemistry of the diterpenoids has been developed only in the last twenty years. This period has seen remarkably rapid progress and apart from certain minor points of stereochemistry the structures of all the majcr diterpenoids have now been elucidated with certainty. The importance of the method of dehydrogenation in establishing the nature of the carbon skeleton in sesquiterpenoid compounds is well known. This method is of even greater importance in the study of diterpenoids and the whole chemistry of the group depends upon basic experiments involving dehydrogenation to aromatic compounds.Indeed the first application of the dehydrogenation method was made in the diterpenoid field by A. Vesterberg who obtained retene (I) from abietic acid by heating it with sulphur. Retene was later isolated in a similar manner by the dehydrogenation of levopimaric acid. A different derivative of phen- anthrene called pimanthrene 1 7-dimethylphenanthrene (11) was first obtained by L. Ruzicka and Balas by the dehydrogenation of dextro- pimaric acid and this hydrocarbon results also usually together with other hydrocarbons from the dehydrogenation of a number of other diterpenoids. The most numerous group of diterpenoids gives either on direct dehydro- genation or dehydrogenation of suitable derivatives 1 7 $-trimethyl- phenanthrene which was first obtained in this way by L.Ruzicka and J. R. Hosking It is possible to make a clmsification of diterpenoids (omitting phytol and a few miscellaneous diterpenoids) into bicyclic and tricyclic groups but since the members of the bicyclic group after suitable cyclisation give the same dehydrogenation products as do those of the tricyclic group it would seem preferable to adopt a system based purely on dehydrogenation experiments as has been done in the case of sesquiterpene compound^.^ One must distinguish therefore three main classes of diterpenoid those giving retene those giving pimanthrene and those giving 1 7 $-trimethyl- phenanthrene recognising also that a number of diterpenoids give both pimanthrene and 1 7 8-trimethylphenanthrene.from a derivative of agathenedicarboxylic acid. Ber. 1903 36 4200. a L. Ruzicka Balas and Vilim Heh. Chim. Acta 1924 7 458. 3 1 b i d . 1923 6 677. ‘ I b i d . 1931 14 203. (1 J. L. Simonsen and D. H. R. Barton “ Tho Terpenes ” Vol. 111 Cambridge Univ. Press. 36 37 The structures of retene pimanthrene and 1 7 8-trimethylphen- BARTON THE CHEMISTRY OF THE DITERPENOIDS anthrene have been rigidly proved by synthesis.6 (6 All the diterpenoids whose constitutions have so far been elucidated obey the " isoprene rule " as illustrated by the formule for abietic acid (111) dextropimaric acid (IV) and agatbenedicarboxylic acid (V). Diterpenoid Resin Acids."-These acids constitute the major non-volatile portion of many oleoresins especially those obtained from conifers.Abietic acid the best known of the resin acids is prepared from colophony (rosin) by treatment with acidic reagents.' .It is a so-called " secondary " resin acid being formed from a precursor levopimaric acid by isomerisation. Dextropimaric acid is also present in colophony but is most easily isolated from French " galipot " * obtained from the cluster pine (Pinus pinaster ; P. maritima). Fossil rosins are formed when resin-exuding trees decay under anaerobic conditions. The copal and kauri copal of various tropical trees are obtained largely as fossilised material and yield the diterpenoid 6R. D. Haworth B. M. Letsky and C. R. Mavin J . 1932 1784; J. C. Bardhan and S. C. Senguipta ibid. p. 2520 ; R. D. Haworth and C. R. Mavin ibid. p. 2720. G.Dupont Bull. SOC. chirn. 1921 29 727 1924 35 879 ; G. Dupont and It. Uzac ibid. 1924 35 394 ; L. L. Steela J . Amer. Chem. SOC. 1922 44 1333 ; 8. Pulkin and T. H. Harris ibid. 1934 55 1935 ; G. C. Harris and T. F. Sanderson ibid. 1948 70 334. * Galipot is that portion oE the untreated oleoresin which crpstallises spontaneously on standing at room t'emperature. Cf. G. C. Harris J . Amer. Chem. Xoc. 1948 70 3671. 38 QUARTERLY REVIEWS resin acid agathenedicarboxylic acid. Many fossil rosins from decayed pine trees have been examined and in the majority of cases the presence of retene and fichtelite (p. 60) has been established Abietie acid which may be obtained in a state of purity by crystallisation of the quarter sodium salt or with greater efficiency of the salts it gives with various nmines sa especially diamylamiize,s is a doubly unsaturated tricyclic acid C20H3002.The two ethylenic linkages are in conjugation lo and although abietic acid gives the same philodiene adduots as levopimaric acid,l0 l1 they must be distributed in two different rings as shown by the ultraviolet absorption spectrum.12 As mentioned above abietic acid (VI ; R = H) gives retene in high yield on dehydrogenation. The carboxyl group which is eliminated in this reaction has been shown to be tertiary in character l3 and t o he attached a t the l-position in the retenoid skeleton. The evidence for this is briefly summarised as follows. Bouveault-Blanc reduction of methyl abietate (VI ; R = Me) afforded abietinol (VII) dehydrated by phosphorus penta- chloride to methylabietin CZoH30.Tlie latter gave homoretene on dehydro- genation with sulphur or ~e1enium.l~ Although homoretene was a t first considered to be a dimethylisopropylphennnthrene it was subsequentlly realised l5 that the formation of methylabietin (VIII) from abietinol involved a molecular rearrangement and that homoretene was 1 -ethyl-’l-isopropyl- phenanthrene (IX). This was confirmed by R. D. Haworth’s synthesis of homoretene. l6 An important clue to the position of the double bonds in abietic acid was furnished a t an early date by the observation that trimellitic acid (X) was produced by oxidation with a variety of reagents.l7 Having regard to * G. Dapont L. Dzsdbres and A. Bernette Bull. SOC. chim. 1926 39 48s ; C. C. Keslor A. Lowy and W. F. Faragher J . Amer. Chem. Soc. 1927 49 2898 ; S. Pellrin and T.K. Harri? loc. cit. ref. (7). F. galas &sopis Cesk. Ldkarnictua 1027 7 320 ; S. Palkin and T. H. Harris Eoc. cit. ; compare V. N. Krestinskii and I. I. Bardyshev J . Gen. Chem. U.S.S.R. 1940 10 1894 ; I. I. Bardyshev ibid. 1941 11 996 ; R. Lombard and J.-31. Frey Bull. SOC. chinz. 1948 15 1194. G. C. Harris and T. F. Sanderson loc. c i t . ref. (7). Abietic acid. lo L. F. Fiesor and W. P. Campbell J . Amer. Chem. Soc. 1935 60 159. 11 L. Euzicka P. J. Ankersmit and R. Frank Helv. Chirn. Acta 1932 15 1289 ; B. A. Arbusow J . Oen. Chem. U.S.S.R. 1932 2 806 ; compare €1. Wimhaus and W. Sandermann Ber. P936 69 2202 ; L. Ruziclra R. G. R. Bacon R. Lukes and J. D. Eoso Helv. Chim. Actu 1938 21 583. l 2 K. Kraft Annalen 1035 520 133 ; H. Wienhaus H. Ritter and W. Sttndermann Ber.1936 69 2108 ; L. Ruzicka and L. Sternbach Helv. Chim. Actn 1938 21 565 ; W. Snndarmann Ber. 1941 74 154. 13W. Fahrion 2. angew. Chem. 1901 14 1197; P. Levy 2. anory. Chem. 1913 81 147. 14 L. Ruzickn and J. Meyer Helu. Chim. ,4cta 1022 5 681 ; L. Ruzicka and H. Jacobs Rec. Trav. chirn. 1038 57 509. * 5 F. Vocke Annalen 1932 497 247 ; I,. Ruzicka G. I3. R. de Graaff and H. J. Muller Helv. Chim. Acta 1932 15 1300. l7 J. Schreder Annalen 1874 172 93 ; 0. Emmefling Rer. 1879 12 1441 ; L. Ruzicka H. Schinz and J. Meyer Helv. Chirn. Actn 1923 6 1077 ; L. Ruzicka and M. Pfeiffer ibid. 1925 8 632. l6 J. 1932 2717. BARTON THE CHEMISTRY OF THE DITERPENOIDS 39 the formula of retene this was taken to imply that at least one of the ethylenic linkages must be in ring B. Support for this view was the fact that isobutpic acid could be obtained on energet,ic oxidation with potassium permanganate.la Rigid proof of the position of the conjugated ethylenic (VIT.) linkages as shown in (VI) has been provided f (VTII.) VX.1 only comparatively recently by the elegant experiments of L.Ruziclra and L. Sternbach.lg When abietic acid is oxidised under mild conditions the fist product of the reaction is dihydroxyabietic acid (XI),2o which then appears to be further attacked with formation of oxidodihydroxyabietic acid (XII). This oxide is unstable in aqueous media and is rapidly hydrated to y-tetrahydroxyabietic acid.* By treatment with dilute hydrochloric acid the latter was converted almost quantitatively into the stable chlorotrihydroxyabietic acid (XIII) whilst with dilute sulphuric acid it afforded the a-tetrahydroxyabietic acid (XIV) which had been obtained previously by P.Levy. 21 y-Tefrahydroxyabietic acid underwent a slow mutsrotation in neutral aqueous solution to afford @-tetrahydroxyabietic acid converted like the y-isomer into the a-acid with dilute sulphuric acid. It appears therefore that the usual product isolated from the potassium permanganate oxidation of abietic acid under mild conditions is a mixture of dihydroxyabietic acid and y-tetrahydrosy- abietic acid and that Levy’s a-tetrahydroxyabietic acid is an artefact formed l 8 P. Levy Ber. 1909 42 4305 ; L. Ruzicka J. Meyer and M. Pfeiffor Hel?]. lS Helv. Chim. Acta 1935 21 565 ; 1940 23 333 341 355 ; 1811 24 492 ; 1942 2O L. Ruzicka and J. &Toyer ibid. 1923 6 1097 ; comparo 0.Aschan Bes.. 1921 21 Ber. 1909 42 4305 ; 1926 59 1302 ; 1935 61 616 ; 1929 62 2497 ; compare * Strictly this name is incorrect and should be replacod by y-tetrahyclroxytetre- Similarly many of the other names uscci in this article are logically Chim. Acta 1925 8 637 ; L. F. Fieser and W. P. Campbell Zoc. cit. ref. (10). 25 1036; L. IZuzickn L. Sternbach and 0. Jogor ibid. 1941 24 504. 54 867 ; L. Ruzicka J. Moyer and M. Pfeiffcr Zoc. cit. rof. (18). 0. Aschan and P. Levy ibid. 1927 60 1923. hydroabietic acid. incorrect but are retained because of the familiar usago in the literatiire. 40 QUARTERLY REVIEWS during working up. Both the p- and the y-tetrahydroxyabietic acid mus% be regarded merely as stereoisomers of the a-acid. By dehydrogenation (XI) (XII) (XIII) and (XIV) all gave 7-hydroxy-1-methylphenanthrene (XV).These experiments place with certainty one of the hydroxyl groups in all these compounds at C arid thus prove the relative position of the isopropyl group and one of the ethylenic linkages in abietic acid. (XI.) (XII.) (XIII.) OT Pd/C I y-Tetrahydroxyabietic acid 130°C I I v\- I (XV.) (XIV.) When y-tetrahydroxyabietic acid was treated with hydrobromic or hydriodic acid it gave bromo- (XVI) or iodo-trihydroxyabietic acid (XVII) respectively. All three halogenot'rihydroxyabietic acids afforded the cor- responding halogenodiketo-acids (XVIII) (XIX) and (XX) on oxidation with two molecular proportions of lead tetra-acetate. With hydriodic acid (XX) was reduced to the diketo-acid (XXI) which by treatment with ammonia gave presumably via the dihydropyridine (XXII) followed by disproportionation 8-azadehydroabietic acid (XXIII) .Selenium dehydro- genation of the latter furnished 8-azaretene (XXIV) the identity of which was confirmed by its synthesis. Both (XIX) and (XX) underwent an unusual reaction when treated with ammonia giving 9-keto-8-azadehydro- abietic acid (XXV) from which (XXIII) was obtained on reduckion by the Wolff-Kishner method. This last series of reactions proves in a particularly elegant manner the correctness of the double bond positions in the abietic acid formula (VI ; R = H). The experiment's described above constitute an unambiguous proof of the main structural features of the abietic acid molecule and leave undecided only the position of the quaternary methyl group. The following is the more important evidencf with regard to this feature.By energetic oxidation abietic acid affords two homologous tricarboxylic acids (XXVI) BARTON THE CHEMISTRY OF THE DITERPENOIDS 41 HO,C HO,C (XILI) Pb(OAc) .1 HO,C X A 7 LIyO CO-CHMe v (xvm.) b\ I N I II \A 1 (XVI.) Pb( OA& or CrO 1 yyTr KO,C v'\Qo.IHnm (XIX.) NH3 (XXV.) Wolff-Kishner 1 HO,C Se I (XVII. ) I Pb(Obc) ( X X . ) HO,C i.' \/ PPI \/)A* ( ,CO*CHTUe (XXI.) HO,C \/\- I I ' (XXIV. ) (XXIII . ) (XXII.) and (XXVII). 22 Dehydrogenation of these with selenium gave respectively m-xylene (XXVIII) and hernimellitene (XXIX). This proves the 1 3- relationship of the two methyl groups in (XXVI) and (XXVII) and hence the position of the quaternary methyl group in (VI) as shown? A further proof of this relationship was provided by L.Ruzicka and zz L. Ruzicka J. Meyer and M. Pfeiffer Eoc. cit. ref. (18) ; P. Levy Ber. 1829 62 2497 ; L. Ruzicka 3%. W. Goldberg H. W. Huyser and C. F. Seidel Helw. Chirn. Acta 1931 14 545 ; compare. J. Schreder Zoc. cit. ; 0. Emmerling Zoc. cit. ; P. Levy loc. cit. ; 0. Aschan and P. Levy loc. cit. 23 L. Ruzicka M. W. Goldberg H. W. Hiiyser a.nd C. F. Seidel loc. cit. 42 QUARTERLY REVIEWS H. Waldmann,24 who showed that vigorous oxidation of abietic acid also afforded 1 3-dimethylcycZohexan-2-one (XXX). HO,C HO,C \/ CH,.CO,R >-<,,co2H /\/ + &I\ ise (171 ; R = €1) -+ I A 1 CO,H ‘+I\ CO,R IiNnO4 (XXV1.) (xx1711.) I be f) I 0.‘ \ \ (XXIX. ) I I,,*” v\ (XXVIII. ) I -4 v\ (XXX.) Additional evidence on this point was given by F. Vocke z5 who treated (XXVI) with red phosphorus and broiiiine and isolated amongst other product’s two bromo-anhydrides (XXXI) and (XXXII) both of which furnished an unsaturated acid (XXXIII) with sodium hydroxide.Although at the time these experiments were carried out they were open to a different interpretation,2G H. N. Rydon’s synthesis 27 of (XXXIII) has confirmed Vocke’s formulation. H02C ErOC H0,C \/ Br \,,/ Br I CO,H \/ <‘+I (p/ ( y > O \/\ C0,H ‘\/I\() / ~ c o \ o + p c o v’- / \/\- I (XXXI . ) (XXXII.) (XXXIII.) (XXXIV.) Levopimaric acid.* This resin acid was first isolated in a state of purity by A. Vesterberg,28 but little progress was made with it,s chemistry until GI-. Dupont 29 had described a reliable method for its preparatioii from French galipot. The most convenient inefhod of isolation is by crystallisa- tion of the butanolamine salt.g S.Palkin and T. H. Harris 30 isolated pure levopimaric acid from the primary resin acids of Pinus paZ,iistl-is and of P. caribhea and it is now certain that the acid is a primary constituent of all resins from pine and fir trees. It has also been shown that the so-called Helv. Chim. Acta 1933 16 842. 25 L O C . cit. 26 L. Ruzicka H. Waldmann P. J. Meier and H. Hosli Helu. Chim. Acta 1933 2 7 J. 1937 257. Ber. 1887 20 3248 29 Comnpt. rend. 1921 172 923 1184 ; Bull. SOC. chim. 1921 29 718 ; compare L. Ruzicka and Balm Zoc. cit. ; L. Ruzicka Balzs and Vilim loc. cit. 30 J . Amer. Chem. Xoc. 1935 55 3677. * The names levo- and dextro-pimaric acid do not of course denote optical anti . 16 169. podes. here because t o do so at this stage might prove confusing.They were originally introduced by A. Vesterberg and have not been altered BARTON THE CHEMISTRY OF THE DITERPENOIDS 43 sapinic acids formerly thought to be primary constituent's of the oleoresin all contain levopimaric acid. 31 * Levopimaric acid C,,H,,O (XXXIV) is readily isornerised by heat or by acids to abietic acid 32 and like abietic acid it gives retene on dehy- dr~genation.~~ The presence of two double bonds and therefore of three rings as implied by the above observations has been proved by hydrogena- tion 34 and by per-acid t i t r a t i ~ n . ~ ~ In addition the two double bonds must be in conjugation for levopimaric acid reacts quantitatrively with maleic anhydride at room temperature to afford the same adduct as is obtained from abietic acid under more vigorous condjtions.38 Since levopimaric acid shows an absorption maximum in the ultraviolet at 272.5 mp,37 it follows that these two conjugated ethylenic linkages must be contained in one ring as indeed would be expected from the ease of addition of maleic anhydride. Levopirnaric acid gives isobutyric acid on ozonolysis. 38 Therefore it is probable that the isopropyl grouping is attached directly to one of the ethylenic linkages. This has been rigidly proved by the elegant experiments of L. Ruzicka and S. Kaufinan~i,~~ the most important of which may be briefly summarised as follows. Ozonolysis of the trimethyl ester (XXXV) derived from the malsic anhydride adduct of levopimaric acid gave amongst other products a mono-unsaturated keto-tricarboxylic acid trimethyl ester (XXXVI) in which the double bond was in the 0r.P-posi- tion to the keto-group as shown by the ultraviolet absorption spectrum.On oxidation with hypobromite (XXXVI) was smoothly degraded (and hydrolysed) to an ap-unsaturated acid (XXXVII) the absorption spectrum of the tetramethyl ester of which confirmed the double bond position in (XSXV). Rigid chemical proof that the acetyl grouping of (XXXVI) was formed from the isopropyl group was obtained (a;) by Clemmensen reduction to (XXXVIII) followed by dehydrogenation to 1 -mcthyl-7-ethyl- phenantltirene (XXXIX) and ( b ) by reaction with ethylmagnesium iodide to give (XL),? dehydrogenation of which afforded l-rnethyl-7-sec.-butyl- 3 1 T. Hmselstrom and M. T. Bogert ibid. 1935 57 211s ; K. Kraft Annulen 1935 520 133 ; 1936 524 1 ; G. C. Harris and J.Sparks J . Amer. Chem. Soc. 1948 70 3674 ; compare F. Voclre Annnlen 1933 508 11 ; W. Sandcrmann Ber. 1938 71 2005. 3 = Inter id. G. Dupoilt Conapt. rend. 1921 172 923 1373 ; Bull. SOC. ckim 1921 29 718 727 ; L. Ruzicka Bslas and Vilim loc c i t . ; R. Lombcrd Bull. SOC. chirn 19-18 15 1186. 33 Idem. ibid. ; W. Sandermann Ber. 1941 74 164. 84 L. Ruzicka and R. G. R. Bacon Helv. Chinz. Acta 1937 20 1542 ; compare L. Ruzicka Ralns and Vilim Zoc. cit. 36 B. A. Arbusow Zoc. cit. ; L. Ruzicka P. J. An!rersmit and B. Frank Zoc. cit. ; L. Ruzicka aiid R. G. R. Bacon Zoc. cit. ; H. Wienhaus and W. Sandermann Ber 1936 69 2202. 35 I<. Kraft Annalen 1036 524 1. 37 K. &aft Anizalen 1935 520 138 ; G. C. Harris and T. F. Sanderson loc. cit s* L. Ruzicka R. G. R. Bacon R. Lukes and J.D. Rose Zoc. c i t . 3* Helv. Chim. Actn 1940 23 1346 ; 1941 24 939. * R. Lombard's recently isolated dextrosapinic acid may be an exception to thk; ?The carbomethoxyl groupings in (XXXVI) may also have reacted with the generalisation (p. 45). Grjgnard reagent but this does not affect the argnment. 44 QUUTERLY REVIEWS Me0,C \ / K A n (XXXVII.) r c o z M e - I C H C H . C O BMe" (XXXVIII.) I (XLII . ) (XXXIX. ) LVI . ) Zn/HCI MgEtI (XLI.) phenanthrene (XLI). The formation of (XXXIX) and (XLI) shows beyond doubt that the double bond of the adduct (XXXV) must be adjacent to the isopropyl grouping. Only the formulz (XXXIV) and (XLII) therefore are possible representations of levopimaric acid and of these the former is much to be preferred as it explains more readily the ease of rearrangement to abietic acid.The formula (XXXIV) is also supported by various theoretical arguments which have been advanced by W. Sandermann. 40 * This interesting resin acid is a primary constituent of the oleoresin of Pinus palusiris and has been isolated therefrom as the 40 Ber. 1941,74 154 ; W. Sanclermann and R. Hohn ibid. 1943 '76 1257 ; compare L. Ruzicka and S. Kaufmann Helv. Chim. Acta 1941 24 1425. 41 Chem. Zentr. 1942 ii 892 893. * Formula (XXXIV) for levopimaric acid is said to be provocl by certain experiments of B. A. ArbusomT,41 who has shown that the a-naphthaquinone adduct on dehydro- genation with air in alcoholic potash solution followed by pyrolysis and then oxidation with nitric acid affords anthraquinone- 1 3-dicarboxylic acid. The Reviewer has not ha? arcess t o Arbusow's original memoirs.neoAbietic acid. BARTON THE CHEMISTRY OF THE DITERPENOIDS 46 butanolamine salt by G. C. Harris and T. F. Sanderso~i,*~ as well as from the resin of this tree. It is also conveniently prepared by heating abietic acid at 300" in an inert atmosphere for short periods. The structure of neoabietic acid as (XLIII) has been established by G. C. Harris and T. F. Sanderson 43 in the following way. neoAbietic acid gave retene on dehydro- genation absorbed two molecular proportions of hydrogen on catalytic hydrogenation showed an intense band in the ultraviolet at 250mp and like levopimaric acid was almost quantitatively isomerised to abietic acid by the action of mineral acid. It must therefore be a simple double-bond isomer of abietic acid in which the double bonds are in conjugation with each other but not in the same ring of the carbon skeleton.On ozonolysis neoabietic acid afforded acetone and an ag-unsaturated ketone C1,H,,O (XLIV) thus showing the presence of an isopropylidene group. It was possible to distinguish between the alternative formula? (XLIII) and (XLV) both of which explain this observation since drastic ozonolysis of neoa,bietic I (XLIII. ) o* d (energetic) I CO*CO,H \/ (XLVa . ) I 1 (XLV.) (XLVb.) acid and dehydrogenation of the reaction product presumably (XLVa) gave 1- methyl- 5 - n-prop ylnapht halene (XLVb) . The relationship of neoabietic acid to t'he dextrosapinic acid recently isolated by R. Lombard 44 as a primary constituer;t of the galipot of Pinus halepensis is uncertain but like neoabietic acid dextrosapinic acid is said to be isomerised by acids to abietic acid.This acid was f i s t obtained in a state of purity by A. Vesterberg 45 by crystallisation of the sparingly soluble sodium salt but it is best isolated after removal of levopimaric acid as the maleic anhydride adduct by crystallisation of the butanolamine salt. 48 Dextro- pimaric acid is probably present to a greater or less extent in all resins 4 2 L O C . cit. 44 Compt. rend. 1944 219 587 ; 1944 219 253 ; 1946 222 237 ; Bull. Xoc. c l z i m 1945 12 395. 46 Ber. 1885 18 3331 ; 1886 19 2167 ; compare idem ibid. 1887 20 3248 ; 1905 38 4125 ; G. Dupont Bull. SOC. chim. 1921 29 718 ; L. Ruzicka and Balas Zoc. cit. ; L. Ruzicka Bdas and Vilim Zoc. cit. Dextropimaric acid. 4 9 J . Amer. Chem. Soc. 1948 'SO 339. 46G.C. Harris and T. F. Sanderson J . Amer. Chem. SOC. 1948 70. 2079. 46 QUARTERLY REVIEWS obtained from conifers but it is not alwa.ys possible to separate it from the accompanying isomeric resin acids. Thus it has been detected in the resins from Pinus caribbea P. tzda P. serotina and Picec excelsa,47 as well as in the oleoresins of P. palustris 4* and P. syl~estris.~~ Unlike most of the other primary resin acids dextropimaric acid is comparatively stable to heat and is not isomerised by treatment with mineral acids. As mentioned above (p. 36) dextropimaric acid (XLVI) gives piman- threne (XI) on dehydrogenation. The position of the carboxyl group has been proved by a similar series of experiments to those recorded (p. 38) for abietic acid. Thus Bouvcault-Blanc reduction of ethyl pimarate afforded dextropimarinol dehydrated by phosphorus pentachloride to methyl dextropimarin which on dehydrogenation gave an aroinatic hydrocarbon C,,H,,.Although the lat'ter was a t first considered to be a trimethyl- phenanthrene it was subsequently shown by 1;. Ruzicka G. B. R. de Graaff and H. J. &Iuller,51 to be 7-methyl-l-ethylphenanthrene (XLVII) and this identity has been confirmed by synthesis.52 From this it follows that niethyldextropimarin must be represented by (XLVIII) its formation involving a rearrangement similar to that observed in the dehydration of abietinol." A further indication of the position of the carboxyl group and a proof of the points of aktachment of the two quaternary methyl groups in ring A are that vigorous oxidation of dextropimaric acid gives the same two tricarboxylic acids (XXVI) and (XXVII) as are obtained in the same way from abietic acid (p.40).53 (XLVI.) (XLVII.) (XLVIII.) Dextropimaric acid is doubly unsaturated a>s shown by catalytic hydro- genation 54 and by per-acid experiment~.~~9 55 Since it has the formula C,,H,,O it must be tricyclic in agreement with the dehydrogenation evidence. The two double bonds are not in conjugation and differ greatly 47 Inter al. T. Hasselstrom and M . T. Bogert loc. cit. ; K. Kraft Annalen 1935 520 133 ; 1936 524 1 ; W. Sandermann Ber. 1938 71 2005 ; 1942 75 174. 48 S. Palkin and T. H. Harris loc. c i t . ; K . Kraft Zoc. cit. 49A. Vesterberg Ber. 1905 38 4125. 60 L. Ruzicka and Balas Helv. Chin&. Acta 1924 7 875. i51 L O C . cit. 62R. D. Haworth J . 1932 2717.6 3 L. Ruzicka G. B. R. de Graaff M. W. Goldberg and B. Frank Helv. Chini. Actci 6 4 L. Ruzicka H. W. Huyser and C . F. Seidel Rec. Trav. chim. 1928 47 363. 5 6 L. Ruziclre and B. Frank Helv. Chim. Acta 1932 15 1294 ; K. Kraft Annulem 1936 524 1 ; L. Ruzicka and L. Sternbach Helv. Ghim. Actcc 1940 23 124. * It should be pointed out that the 1 (2)-position of the ethylenic linkage in methyl- dextropimarin and in methylabietin has not been proved and it may well be at l(11) instead. 1932 15 915. BARTON THE CHEMISTRY OV THE DITERPENOIDS 47 in reactivity. Partial hydrogenation gives a very insoluble and characteristic dihydrodextropimaric acid (XLIX),54 56 which has been used by L. Ruzicka and L. Sternbach 57 in an elegant demonstration that the less reactive double bond is in the 7(8) or S(14)-position.Their evidence is briefly as follows. By treatment of the oxide of methyl dihydrodextropimarate (L) with methyl- magnesium iodide an alcohol (LI) was prepared which on dehydrogenation gave 1 7 8-trimethylphenanthrene (LII). MgMeI I' (XLIX. ) (LII.) The more reactive of the two ethylenic linkages is present as a tertiary vinyl grouping. This is proved by the facts that dext,ropimaric acid may be oxidised by potassium permanganate to a glycol (LIII) which on further oxidation by chromic acid gives a dicarboxyiic acid C,,H,,O (LIV) con- taining one carbon atom less,58 which is dehydrogenated by selenium to pimanthrene (11). 53 As would be expected ozonolysis of dextropimaric acid gives a high yield of f~rmaldehyde.~~ (LIII. ) (LIV.) (11.1 The experiments described above taken in conjunction with the " isoprene rule " indicate that dextropimaric acid must be represented by either (XLVI) or (LV).Conclusive evidence in favour of the former 6 6 L. Ruzicka and Balas ibid. 1923 6 681 ; L. Ruzicka and B. Frank Zoc. cit. ; S . Palkin and T. H. Harris loc. c i t . ; T. Hasselstrom and M. T. Bogert Zoc. cit. ; compare L. Tschugaeff and P. Teearu Ber. 1913 46 1773. 5 7 LOC. cit. 68 I,. Ruzicka and Balas Annakn 1928 460 202 ; compare P. Levy Ber. 1928 61 616. so L. Ruzicka and Balas Zoc. cit. ; compare L. Kuzicka C. E'. Soidel IT. Schinz and M. Pfeiffer Helv. Chim. Acta 1947 30 1807. 48 QUARTERLY REVIEWS of these has been reported recently by G. C. Harris and T. F. Sanderson.60 Dihydrodextropimaric acid (XLIX) was ozonised to give the keto-aldehyde (LVI) (negative iodoform test) which was reduced by the Wolff-Kishner method and then dehydrogenated to give a CIS disubstituted naphthalene (LVII).The keto-aldehyde from (LV) would have given a positive iodoform test and would have been converted into a C, trisubstituted naphthalene. Further proof for the correctness of (XLVI) was provided by partial dehydro- genation of dext,ropimaric acid which gave a C, trisubstituted naphthalene (LVIII). (XLVI.) (XLIX.) I I (LV.) I I a).. - I I/- \/\ (LVIII.) I 0 3 PdjC (LVIZ.) (LVI.) isoDextropimaric acid. This interesting resin acid has been isolated recently 46 as the butanolamine salt during the preparation of dextropimaric acid. G. C. Harris and T. F. Sanderson Go regard it as the C epimeride of dextropimaric acid on the basis of the following evidence.When isodextro- pimaric acid was subjected to the same series of degradations as was applied to dextropimaric acid the same hydrocarbons (LVII) and (LVIII) were obtained racemisation at C being presumed to occur during the formation of (LVIII). Also when dextropirnaric acid and isodextropimaric acid were ozonised and the products oxidised with hydrogen peroxide the same tricarboxylic acid (LIX) in which the asymmetry at C had been destroyed was isolated in both cases. Although Harris and Sanderson's interpretation of the experimental evidence is logical it should be pointed out that isodextropirnaric acid and its butanolamine salt were found to be optically inactive in all solvents and to This would be unexpected unless it were the racemate corre- sponding to dextropimaric acid.give an optically inactive dihydro-acid. 6 CO,H (LIX.) G. C. Harris and T. F. Sanderson60" have also 6o J . Amer. Chem. Soc. 1948 70 2081. 6ofJ Ibid. p. 3870. BARTON THE CHEMISTRY OF THE DITERPENOIDS 49 isolated the aldehyde corresponding to isodextropimaric acid from the neutral fract'ions of commercial gum and wood rosins. It is probably identical with cryptopinone obtained by N. A. Sorenseii and T. Bruun 60b from the twig roots and resinified trunks of pine trees. This bicyclic resin acid was first isolated in a state of purity by L. Ruzicka and J. R. Hosking 61 from Kauri copal and from the soft and hard grades of Manila copal. The structure (LX) which has been assigned t o this acid is due entirely to the experiments of L. Ruzicka and his collaborators,62 and is based on the following evidence.The carbon skeleton present in the acid is indicated by its dehydrogenation with sulphur or selenium to give 1 5 6-trimethylnaphthalene (LXI) and pimanthrene (11). The formation of the latter results presumably from the presence of an unsaturated side chain which appears partially in the naphthalene hydrocarbon as a methyl group. The acid contains two ethylenic linkages one of which must be in the ccp-position with respect to one of the carboxyls. This is shown by the absorption spectrum and by the ease wit'h which carbon dioxide is split out on pyrolysis to give noragathenemonocarboxylic acid (LXII) . Agathenedicafboxylic acid. I Ill v\ (LXIII.) (LXV.) (LXIV. ) MgMeI sob Acta chem. scand. 1947 1 112. G 2 L. Razjcka R. Steiger and H.Schinz Helv. Chim. Acta 1926,9 962 ; L. Ruzicka and J. R. Hosking ibid. 1930 18 1402 ; 1931 14 203 ; L. Ruzicka and H. Jacobs Zoc. cit. ; L. Ruzicka E. Bemold and A. Tallichet Helv. Chim. Acta 1941 24 223 ; L. Ruzicka and E. Bernold ibid. p. 931 1167 ; compare L. Ruzicka and E. Rey ibid. 1943 26 2136. 61 Annalen 1929 469 147. r) 50 QUARTERLY REVIEWS The position of the two olefinic linkages in agathenedicarboxylic acid was proved by a study of the ozonolysis products of the dimethyl ester. The most important of these was a 1 5-diketone (LXIII) which readily underwent intramolecular dehydration on treatment with alkali to give a tricyclic ap-unsaturated ketone (LXIV). By reaction with methyl- magnesium iodide," the latter afforded a conjugated diene (LXV) dehydro- genated by selenium to give pimanthrene.These experiments also indicate the position of the carboxyl group which is easily eliminated on pyrolysis. The position of the other carboxyl group was proved in the following manner. When agathenedicarboxylic acid was digested with formic acid it was isomerised to a tricyclic acid isoagathenedicarboxylic acid (LXVI). This like its progenitor possessed an orp-unsaturated carboxyl group readily eliminated by heat to give isonoragathenemonocarboxylic acid (LXVII). Reduction of the methyl ester of the latter t by a modified Bouveault-Blanc (LXVI.) (LXVII.) (LXVIII . ) (XLVII.) method furnished isonoragathenol (LXVIII) which was dehydrated and then dehydrogenated to give 7-methyl- l-ethylphenanthrene (XLVII) identical with the hydrocarbon prepared similarly from dextropimaric acid (p.46). Although podocarpic acid is not strictly a member of the diterpenoid resin acids its chemistry is closely related to that of the diterpenoids. It is convenient therefore to give a short account of it here. Podocarpic acid C,,H,,O (LXIX ; R == H) was first isolated by A. C. Oudemans 63 from the resin of Podocarpus cupressinum and was obtained later 64 from P. dacrydioides and from Dacrydium cupressinum. It is a phenolic carboxylic acid which gives l-methylphenanthrene and 6-hydroxy- Podocarpic acid. a 3 Ber. 1873 6 1122 1125 ; Annalen 1873 170 214 ; J. p r . Chem. 1874 9 385. 64Compare I. R. Sherwood and W. F. Short J. 1938 1006. * The tertiary carbomethoxyl group in (LXIV) is strongly hindered storically and does not react. 7 More recent experiments have shown that (LXVII) was iiot homogeneous ; this does not affect the validity of the conclusions reached here because tho inhomogeneity depended only on isomerism in ring B.BARTON THE CHEMISTRY OF THE DITERPENOIDS 51 1 -methylphenanthrene on dehydr~genation.~~ Although these facts are explicable by other formulae besides (LXIX ; R = H) the correctness of the latter was proved by W. P. Campbell and D. Todd 66 in the following way. Podocarpic acid methyl ether (LXIX ; R = Me) was reduced by the Rosenmund method to podocarpinal methyl ether (LXX) which was further hydrogenated to podocarpinol methyl ether (LXXI). This was dehydrated and then dehydrogenated to give 6-rnethoxy-1 -ethylphenan- threne (LXXII). (LXIX.) (LXX.) (LXXI.) \/ OM0 OMe (LXXII.) Miropinic acid and isomiropinic acid.* These acids were isolated by C .W. Brandt and L. G. Neubauer from the resin exuded by the Miro tree (Podocarpus ferrugineus). Miropinic acid C20H3003 is tricyclic has two ethylenic linkages and gives pimanthrene on dehydrogenation. It is partly isomerised to isomiropinic acid by treatment with methanolic hydrogen chloride. Vouacapenic acid. The heartwood of Vouacapoua americana Aubl. contains the methyl ester of vouacapenic acid C,,H,,O,. This acid has two ethylenic linkages ; the third oxygen atom is probably ethereal.70 Cativic acid. The oleoresin from Prioria copaifera Griseb. (the cativa tree) contains the interesting resin acid cativic acid C20H3402 which occurs in the resin both in the free state and esterified with the corresponding primary alcohol cativyl alcohol.Although cativic acid is probably 6s H. Plimmer W. F. Short and P. Hill ibid. p. 694 ; I. R. Sherwood and W. F. 6 7 J . Pharm. SOC. Japan 1937 57 69. 68 J. R. Hosking and C. W. Brandt Ber. 1935 68 1313 ; J. R. Hosking New 70 D. B. Spoelstra Rec. Trav. china. 1930 49 226. * Mkopinic acid is possibly identical with cryptopimaric acid isolated by S. Keimatsu T. Ishiguro and G. Fukui 6 7 from Gryptomeriajaponica and with an acid obtained from Dacrydium. bqorme and D. birkii.68 Short Eoc. cit. 66 J . Amer. Chem. SOC. 1042 84 928. Zealand J . Sci. Tech. 1937 19 208. 69 J. 1940 683. 52 QUARTERLY REVIEWS diterpenoid it differs from abietic acid and related acids in that its carboxyl group is readily esterfiedO7l Diterpenoid Alcohols and Phenols.-These compounds have received compar at ivelj little at tent io n .Phytol. The mono-unsaturated aliphatic diterpenoid alcohol phytol C2J3400 was discovered by R. Willstatter 7 2 to be the alcoholic moiety of the chlorophyll molecule. The initial degradational experiments 73 did not lead to a successful elucidation of the structure of phytol and this was only effected later by the synthetic experiments of F. G . Fischer,74 who found that phytol (LXXIII) which had been already recognised as a primary alcohol gave a saturated ketone (LXXIV) and glycollic aldehyde on ozonolysis thus showing the double bond to be in the ccp-position to the alcoholic grouping. The structure of (LXXIV) as 2 6 10-trimethyl- H3C CH3 CH CK3 I I I ~~.[C€12]3.CH.[~H*]3.C~~.[CH,I,.C :CH*CII,-OH \ / (LXXIII.) H,C CH3 CH3 I cIr loa H3C I I CXI.[ CH,],*CK.[ CH,],.CH*[ C;H1] C O \ / I5,C (LXXIV.) " Ketuiiic 'I hydrolysis etc.CK3 I t CH3 I CH*[CH,I3.CH*[CH2I3*CH.[CH,I,.Br + CH,-CO*CHNa*CO,Et \ / H3C (LXXV.) pentadecan-14-one was proved by its synthesis as indicated * from hexahydrofarnesyl bromide (LXXV) . Natural phytol usually has a negligibly small optical rotation and this has been taken to imply that it is a r a ~ e m a t e . ~ ~ However Karrer et aE.76 H3C 71 N. L. Kalman J. Amer. Chew. Soc. 1938 60 1423. 7 2 R. Willstatter and F. Hochecler Annden 1907 354 305 ; R. Willstatter F. Hocheder and E. Hug ibid. 1909 371 1 ; R. Wjllstatter and 4. Opp6 ibid. 1911 3'98 1. 7 3 R. Willstatter E. W. Mayer and E. HLini ibid. p. 73 ; R. Willstiitter 0. Schuppli and E. W. Mayer ibid. 1918 418 121. 7 4 Ibid. 1928 464 69 ; F.G. Piselier and K. Lowenborg ibid. 1929 475 183. 7 5 F. G. Fischer and K. Lowenberg doc. cit. ; T. MTagner-Saurogg 2. physiol. Clzetre. '6 P. Kurrer A. Geiger H. Rentschler E. Zbinden and A. Kugler Helv. Chim. Acta * For later syntheses see L. I. Smith and J. A. Sprung J . Arrzer. Cliem. SOC. 1943 1933 222 21. 1943 26 1741. 85 1276; P. Karrer et ul. Helv. Chiin. Acta 1943 26 1741. BARTON THE CHEMISTRY OF THE DITERPENOIDS 53 have recently described the isolation of an optically active phytol [mi:"" + 0 - 2 0 4 4 1 O from nettles and have synthesised a lzevorotatory phytol tcD - 0.18". This synthetic phytol and the dextrorotatory phytol are not as might have been thought optical antipodes ; 77 it is concluded that phytol is not a racemate but a latently optically active compound.The ditertiary glycol sclareol C,,H,,O (LXXVI) was first isolated by Y. Volmar and A. Jermstad 78 from the leaves of Salvia sclarea I;. It was recognised as a diterpenoid by 31. M. J a n ~ t ~ ~ who showed by catalytic hydrogenation to the saturated dihydrosclareol (LXXVII) that it must be bicyclic and contain only one ethylenic linkage. The carbon skeleton of sclareol was partly characterised by L. Ruzicba and M. M. Janot's observa- tion 8O that sclareol gave 1 5 6-trimethylnaphthalene (LXI) on dehydro- genation with selenium. When dihydrosclareol was treated with potassium hydrogen sulphate it furnished amongst other products dihydrocyclo- sclarene (LXXVIII) which was dehydrogenated to a mixture of 1 7 8- trimethylphenanthrene (LII) and pimanthrene.8O The presence of a methylene grouping in sclareol was shown by the high yield of formaldehyde on ozonolysis and by the formation of a CIg dihydroxy-acid (LXXIX) amongst other products on oxidation with potassium permanganate.80.81 In view of these experimental facts particularly the isolation of (LII) on dehydrogenation of (LXXVIII) a carbon skeleton similar to that in agathenedicarboxylic acid (p.49) was suggested. The two uncharacterised oxygen atoms must be present as tertiary alcoholic groups and are placed as indicated in order best to account for the formation of (LXXVIII) on dehydration. Some recent work on manool (p. 55) discussed later con- stitutes an indirect proof of the correctness of the sclareol formula. Sclareol. (LXI.) OH OH (LXXVI.) (LXXVII. ) I OH (LXXIX. ) (LXXVIII.) 7 7 P. Karrer H.Simon and E. Zbinden ibid. 1944 27 313. 78 Compt. rend. 1928 186 517 783. 8oHelv. Chim. Ada 1931 14 645 ; M. M. Janot Ann. Chim. 1932 17 5. 81 L. Ruzicka C. F. Seidel and L. L. Engel HeEv. Chim. Aota 1942 25 621. 7 9 I b i d . 1930 191 847 ; 1931 192 845. 54 QUARTERLY REVIEWS Manool manoyl oxide and ketomanoyl oxide. Because of a very close relationship in structure it is convenient to treat these three substances together. The diterpenoid alcohol manool C,,H,,O (LXXX) was isolated by J. R. Hosking and C. W. Brandt 82 from the wood oil of the yellow pine (Dacrydium biforme). The same authors also reported the isola.tion of manoyl oxide C2,H3,0 (LXXXI),83 and of ketomanoyl oxide (LXXXII)84 from the wood oil of the silver pine (Dacrydium coEensoi ; otherwise D. uiesthndicum).The accepted structure for manoyl oxide was established by Hosking and Brandt 8 5 7 86 in the following manner. Manoyl oxide contained only one ethylenic linkage and the oxygen atom which could not be characterised was assumed correctly to be part of an oxide ring. On dehydrogenation with selenium manoyl oxide afforded 1 5 6-trimethylnaphthalene (LXI) and 1 7 8-trimethylphenanthrene (LII). The presence of an exocyclic methylene group was proved by the formation of formaldehyde in high yield on ozonolysis and by oxidation with potassium permanganate which gave a C, acid (LXXXIII). This acid is important because it shows that (LXXXIV.) the oxide ring in (LXXXI) is correctly formulated. Oil treatment with hydrogen chloride both manoyl oxide and sclareol gave the same trichloro- compound (LXXXIV) thus confirming the structure assigned to the former.The formula (LXXX) for manool was deduced by Hosking and Brandt s2 86 from the following evidence. Since manool gave the same product (LXXXIV) by treatment with hydrogen chloride as was obtained from manoyl oxide they must possess the same carbon skeleton. Catalytic hydrogenation of manool furnished the saturated tetrahydromanool (LXXXV) and thus showed the presence of two ethylenic linkages. The action of hydrogen chloride on (LXXXV) afforded a chloride yielding on digestion with aniline a mixture of two hydrocarbons (LXXXVI) and (LXXXVII). On ozonolysis this mixture gave a C, ketone (LXXXVIII) 8 2 Ber. 1935 68 1311 ; N e w Zealand J. S c i . Tech. 1936 17 755. 83 Ber. 1934 67 1173 ; N e w Zealand J. S c i .Tech. 1936 17 7.50. 84 Ber. 1934 67 1173 ; 1935 68 286 ; N e w Zealand J. Sci. Tech. 1936 17 750. 85 Ber. 1935 68 37. su Ilbid. 1936 69 780. BARTON THE CHEMISTRY O F THE DITERPENOIDS 55 and a c16 acid (LXXXIX). The formation of these two degradation pro- ducts provides conclusive proof that the hydroxyl group in manool must be as in (LXXX) and also indicates indirectly the position of one of the ethylenic linkages as shown. The position I I (LXXXVI.) of the other ethylenic linkage '(XCI.) Vl\A -+ CO,H (h (LXXXIX.) 1 COMe (LXXXVII.) was proved by ozonolysis of manool itself which gave a C, diketone (XC). The keto-groups must have been in the 1 5-relationship in the latter because of the ease with which it was cyclised by alkali to the hydroxy- ketone (XCI). The formula (LXXX) for manool has recently received confirmation by the establishment of a direct relationship with abietic acid.87 The hydroxy-ketone (XCI) was reacted with isopropylmagnesium bromide to give the dihydric alcohol (XCII) which was dehydrated to the corresponding diene possibly (XCIII) and then dehydrogenated by bromosuccinimide to the hydrocarbon (XCIV).This hydrocarbon was identical with dehydro- abietane obtained from abietic acid since both gave the same 6 S-dinitro- dehydroabietane on nitration. The synthesis of dehydroabietane had been carried out earlier by W. P. Campbell and D. Todd,66 starting with dehydro- abietic acid (XCVI).* This was reduced by the Rosenmund method to \/ (LXXXVIII.) 8'0. Jeger 0. Durst and G. Buchi Helv. Chim. Acta 1947 30 1553. *Dehydroabietic acid was first obtained in a state of purity by L.F. Fieser and W. P. Campbell.lo It is a major constituent of the so-called pyroabietic acid formed by the action of heat on abietic acid.10 88 Although much interesting work of a preparative nature has been done in recent years on dehydroabietic z~cid,~g space restrictions do not justify a detailed account here. 56 QUARTERLY REVIEWS clehydroabietinal (XCVII) which on further reduction by the WolE-Kishner method gave dehydroabietane. ’I\/\ OH \ A/\ I (XCIII.) I (XCI‘L. ) 1 (XCIV.) I (XCVI.) (XCVII . ) (XCV.) The forinula (LXXXII) for ketoma,noyl oxide has been elegantly demonstrated by the experiments of Mosking and Brandt .s4 86 Reduction of this oxide by the Wolff-Kishner method afforded rnanoyl oxide (LXXXI) thus leaving only the position of the keto-group to be established.With methylmagnesium iodide (LXXXII) furnished via the carbinol (XCVIII) and catalytic hydrogenation (XCIX) the oxide ring of which was split by hydrogen chloride to give (C). Removal of hydrogen chloride from the latter by heating with aniline followed by dehydrogenation gave ,z mixture of 1 3 5 6-tetramethylnaphthalene (CI) and probably 1 3 7 %tetra- methylphenanthrene (CII) thus proving the original keto-group to have occupied the 3-position. This phenolic diterpenoid comprises the major part of the resin of the Miro tree and was isolated therefrom by C. W. Brandt and L. G. N e u b a ~ e r . ~ ~ Perruginol C20H300 gave 6-hydroxyretene (CIII ; R = a) on dehydrogenation with selenium,g0 91 from which result the formula (CIV) was deduced.The correctness of this formula was shown by W. P. Campbell and D. Todd 66 by partial synthesis of ferruginol (a) from dehydroabietic acid and (b) from podocarpic acid (p. 50). ( a ) 6-Meth- oxydehydroabietic acid (CV) was reduced by the Rosenmund method to Ferruginol. 88Inter al. E. E. Fleck and S. Palkin J . Amer. Chem. SOC. 1937 59 1593; 1938 60 921 ; 1939,61 247 ; L. Ruzicka R. G. R. Bacon L. Sternbmh and H. Waldmsnn Helv. Chim. Acta 1938 21 591 ; R. Lombard Compt. rend. 1939 208 1321 ; 1941 213 793 ; Bull. SOC chim. 1942,9 833 ; T. Hassslstrom E. A. Brennan,!and S. Hopkins J . Amer. Chenz. SOC. 1941 63 1759. a9 Inter nl. T. Hasselstrom E. A. Brennan and J. D. McPherson ibid. 1938 60 1267 ; T. Hasselstrom and J. D. McPherson ibid. p. 2340 ; L. F. Fioser and W. P. Campbell ibid.p. 2631 ; 1939 61 2528 ; W. P. Campbell and M. Morgana ibid. 1941 63 1838 ; T. Hasselstrom and S. Hopkins ibid. p. 421 ; L. Ruzicka and S. Kaufmann Helv. Chim. Acta 1940 23 288. J. 1939 1031. 9lW. P. Campbell and D. Todd J . Amer. Ghem. Soc. 1940 62 1287. BARTON THE CHEMISTRY OF THE DITERPENOIDS 57 (LXXXII. ) (XCVITT.) (XCIX . ) I I XXX I + the corresponding aldehyde which on further reduction by the Wolff- Kishner procedure gave ferruginol. (b) Podocarpic acid methyl ester methyl ether (CVI) underwent the Friedel-Crafts reaction with acetyl chloride to give the 7-acetyl derivative converted by reaction with methyl- magnesium chloride followed by dehydration of the resulting tertiary (CVI.) (CVII. ) (CVIII.) carbinol * into the 7-isopropenyl derivative which on hydrogenation furnished 7-isopropylpodocarpic acid methyl ester methyl ether (CVII).The cor- responding acid was then reduced as described above for 6-methoxyde- hydroabietic acid and gave ferruginol. These partial syntheses are not only a proof of the structure of ferrigunol but also show that the carboxyl group in abietic acid is epimeric to that in podocarpic acid. The diterpenoid phenolic alcohol hinokiol C20H3002 wits Hinokiol. * The carbomethoxyl group of methyl podocarpate is very hindered sterically and did not react wit,h the Grignard reagent. 58 QUARTERLY REVIEWS first isolated by Y. Yoshiki and T. Ishiguro 92 as one of the crystalline constituents of the resin extracted from the heartwood of Chamzcyparis obtusa Sieb. et Zucc. The chemistry of hinokiol has been extensively 93 On dehydrogenation with selenium hinokiol furnished amongst other products a substance which is undoubtedly 6-hydroxyretene (CIII ; R = H),90 91 and which accounts for the phenolic hydroxyl in hinokiol.The second hydroxyl is secondary as shown by oxidation of hinokiol to the corresponding ketone hinokione which is also a constituent of Chamzcyparis obtusa resin. It is probably attached a t the 3-position,94 as shown in the formula (CVITI). The only known phenolic ketone sugiol C,,H2s02 was isolated by S. Keimatsu T. Ishiguro and G. Fukui 95 from Cryptomeria japonica D. Don. Some progress has been made towards the elucidation of its structure.96 On reduction by the Clenimensen method and dehydrogenation of the product sugiol methyl ether afforded a substance which is undoubtedly 6-methoxyretene (CIII ; R = CH,).Sugiol probably differs therefore from hinokione only in the position of the ketonic oxygen function. The diterpenoid alcohol totarol C20H300 isolated from the wood of the totara tree (Podocarpzcs totara) has been investigated by W. F. Short and H. Str~rnberg.~~ Totarol is probably a secondary alcohol and it contains three ethylenic linkages. On dehydrogenation i t gives 7-hydroxy-l- methylphenanthrene which has been synthe~ised.~~ Diterpenoid Hydrocan;bons.-These constitute a numerous group of diterpenoids which apart from camphorene are of unknown structure. They are not a t present of acny economic importance. Camphorene was found by F. W. Semmler and I. Rosen- berg 99 to occur in the higher-boiling hydrocarbon fraction of camphor oil. Later,loO it was recognised as identical with the dimyrcene prepared by C.Harries lo1 by the action of heat on myrcene. It can also be prepared by similar pyrolytic methods from other substances. lo2 Camphorene is monocyclic and has four ethylenic linkages as shown by its catalytic hydrogenation to octahydrocarnphorene (CIX). The latter on oxidation with manganese dioxide and sulphuric acid furnished terephthalic acid. Slcgiol. TotaroE. Camphorene. 92 J . Pharm. SOC. Japan 1933 53 11 ; Chem. Zentr. 1933 i 3203. 93 S. Keimatsu and T. Ishiguro J . Phnrm. SOC. Japan 1935 55 45 ; Chew,. Zentr. 1935 ii 3664 ; G. Huzii and T. Tikamori J . Pharm. Xoc. Japan 1939 59 116. 94 L. F. Fieser and 11. Fieser " The Chemistry of Natural Products related to Phenanthrene " New Edition 1949. The Reviewer is much indebted to Professor L.F. Fieser for an opportunity to read the text of the new edition before pubIication. 95 J . Pharm. SOC. Japan 1937 57 92 ; Chern. Zentr. 1937 ii 596. 9 6 G. Huzii and T. Tikamori J . Phnrm. SOC. Japan 1939 59 124; Chem. Abs. 9SW. F. Short H. Stromberg and A. E. Wiles J . 1936 319. 99Ber. 1913 46 768. 1 o o F . W. Semmler and K. G. Jonas ibid. p. 1566; 19P4 47 2068. 101 Ibid. 1902 35 3264. 102 F. W Semmler and K. G. Jonas ibid. 1914 47 2068 ; 1939 33 4592. 9 7 J . 1937 516. I<. Iiafuku T. Oyamada and M. Nishi J . Chein. SOC. Japan 1933 54 364; L. A. Golclblatt and S. Palkin J . Amer. Chenz. Xoc. 1941 63 3517. BSRTON THE CHEMISTRY OF THE DITERPENOIDS 59 (CIX.) (CX.) (CXI.) (CXII.) This is in agreement with formula (CX) for camphorene lo3 derived from consideration of its method of formation by the polymerisation of myrcene.The diterpene phyllocladene C20H32 has been isolated from numerous essential l o 5 9 log and may occur together with isophyllocladene. lo6 Both phyllocladene and isophyllocladene are tetracyclic and contain one ethylenic linkage. They give the same hydrochloride and are related to each other respectively as are ,8- and cc-pinene. lo5 Phyllocladene gives amongst other products pimanthrene and retene on dehydrogenation and it has been suggested 107 that phyllocladene may be represented by (CXI) or (CXII) of which the former explains better the degradation to retene. cc-Dihydrophyllocladene produced along with the ,&isomer by the catalytic hydrogenation of phyllocladene is identical with iosene obtained from lignites.lo8 z'soPhyiiocIadene is pro babiy the optical' antipode of the di~erpene mirene isolated by J.R. Hosking and W. F. Short log from the leaf oil of Podocarpus ferrugineus (the &liro tree). This tetracyclic mono-unsaturated diterpene C20H32 has been isolated from various essential oils.ll0 It gives a monohydro- chloride which with potassium acetate affords 6-podocarprene. cc- and 6-Podocarprenes are probably related in the same way as phyllocladene and isophyllocladene. The diterpene kaurene isolated by J. R. Hosking ll1 from the leaf oil of the Kauri pine (Agathis australis) is probably an artefact formed by the action of heat or metallic sodium on cc-podocarprene which is the substance isolated from the oil under mild conditions.112 This tricyclic diterpene C20H32 isolated from the essential Phyllocladene and isophyllocladene.cc- Podocarprene. Rimuene. lo3 L. Ruzicka and M. Stoll Helw. Chirn. Acta 1924 '7 271. lo4B. H. Goudie J . Soc. Chem. I n d . 1923 42 3 5 7 ~ ; H. A. A. Aitken ibid. 1928 4'7 2 2 3 ~ ; W. J. Blackie ibid. 1929 48 3 5 7 ~ ; 1930 49 2 6 ~ ; L. H. Briggs J. 1937 79 ; compare C. W. Brandt New Zealand J . Sci. Tech. 1938 20 8. 1O5 K. Nishida and H. Uada J . Agric. Chem. Soc. Japan 1935 11 489; 1936 12 308; H. Uoda J . Dept. Aqric. Kyushu Imp. Univ. Japan 1937 5 117. loSL. H. Briggs and M. D. Sutherland J . Org. Chew&. 1948 13 4. lo' C. W. Brandt Zoc. cit. ref. (104). lo8 L. H. Briggs J . 1937 1035. log Rec. Trav. chim. 1928 47 834 ; J. R. Hosking ibid. 1930 49 1036 ; compare J. Kawamura Bull. Imp. Forestry Exp.Sta. Tokyo 1931 No. 31 93. 110 K. Nishida and H. Uoda J . Agric. Chem. SOC. Japan 1931,7 157 ; J. Kawamura Zoc cit. ; L. H. Briggs and R. W. Cawley J . 1948 1888. " ' R e c . Trav. chirn. 1928 47 578; 1930 49 1036. ll2 L. H. Briggs and R. W. Cuwley Eoc. cit. ref. (110). 60 QUARTERLY REVlEWS oil of the Rimn tree (Dacrydium cupressinum) 113 and as totarene from that of the Totara tree (Podocarpus totcar@,l14 gives pimanthrene on dehydro- genation and is isomerised to isophyllocladene on digestion with formic acid.lo7 It contains a methylene group and a further unidentified olefinic linkage. Miscellaneous Diterpenoids.-FichteEite. Although fichtelite is not strictly a diterpenoid it is very closely related to abietic acid and so may be considered here. It was first isolated by C.Bromeis 115 from the decayed wood of conifers in which it owes its presence to the decomposition of resin acids under anaerobic conditions. Although this saturated hydrocarbon was the subject of numerous investigations,ll6 it was not until L. Ruzicka F. Balas and H. Schinz 117 had shown that it afforded retene on dehydro- genation with sulphur that real progress could be made with its formulation. Subsequently L. Ruzicka and H. Waldmann 118 found by quantitative dehydrogenation with palladised charcoal that its (previously disputed) formula must be C19H34 and this was confirmed by X-ray analysis.119 In view of the relationship of fichtelite to abietic acid and to retene Ruzicka and Waldmann suggested that it is most probably represented by (CXIII). Marrubiin. The diterpenoid lactone marrubiin C,,H,804 which constitutes the bitter principle of the horehound (&?arrubium vuZgare L.) although known for 60 many years was not isolated in a state of purity until 1932.120 by various workers 121 and some progress has been made (~~111.) with the elucidation of its structure.It contains two ethylenic linkages and is bicyclic giving 1 5 &trimethyl- naphthalene (LXI) on dehydrogenation. Two of the oxygen atoms are present as a lactone ring one as a tertiary hydroxyl group and the fourth probably as an oxide ring. It is possibly related therefore to the 1 7 8-trimethylphenanthrene group of diterpenoids. Stereochemistry of the Diterpenoib.-Some progress has aheady been made in our knowledge of the stereochemistry of the diterpenoids and it is possible t o present a tentative scheme covering the more important centres in those compouncls of established structure.,),- I The chemistry of marrubiin has been investigated 1 113F. H. McDowall and H. J. Finlay J . SOC. Chem. Id. 1926 44 4 2 ~ ; M. S. 114 G. B. Beath ibid. 1933 52 3 3 8 ~ ; comparo H. A. A. Aitken ibid. 1929 48 115 Animlert 1841 37 304 ; compare J. B. Trommsdorff ibid. 1837 21 126. 116 Inter al. T. E. Clark ibid. 1857 103 236 ; C. Hell Ber. 1889 22 498 ; E. Bamberger ibid. p. 635 ; C. Liebermam and L. Spiegel ibid. p. 779 ; L. Spiegel ibid. p. 3369; E. Bamberger and L. Strasser ibid. p. 3361. Carrie ibid. 1832 51 3 6 7 ~ . 34411. 117Helv. Chim. Acta 1923 6 692. l1*1bid. 1935 18 611. 119 D. Crowfoot J . 1938 1241. lZoL. J. Mercier and F. Mercier Conzpt. rend. 1932 195 1102. lz1 H.M. Gordin J . Amer. Chem. SOC. 1908 30 265 ; A. Lawson and E. D. Eustice J. 1939 587 ; F. Hollis J. H. Richards and A. Robertson Nature 1939 143 604. BAaTON THE CHEMISTRY OF THE DITERPENOIDS 61 It was mentioned on pp 40 46 that vigorous oxidation of abietic (VI ; R = H) and dextropimaric acids gives a C1,H1606 tricarboxylic acid derived from ring A. This acid is optically inactive and must therefore possess a plane of symmetry i.e. the 1- and the 3-carboxyl group must be in the cis-relationship to each other. The carboxyl attached a t position 2 has recently been shown by D. H. R. Barton and G. A. Schmeidler,122 from a study of dissociation-constant data to be related in the trans-sense to the other two carboxyl groups so that the A/C ring fusion in abietic acid and related acids must also be trans.* The remaining asymmetric centre in abietic acid a t C13 (see VI) is probably in the trans-relationship to the C, methyl group.This is so because it would be anticipated that treatment with acidic reagents (as in the preparation of abietic A acid) would provide a mechanism for the (' A/ '-$ at this centre. There is support €or this \ ;B\ ,Q Ro& assumption of the more stable configuration A v \/ \, argument in the work of W. Sandermann,lz4 13 who prepared an isomer of abietic acid by lo * Ti/ pyrolysis of the levopimaric acid malcic \ A- W I (VI.) anhydride adduct and from nbietic acid dihydrobromide. He called this isomer iao- abietic acid and showed that it was possibly epimeric with abietic acid a t the C13 position. On treatment with hydrogen chloride it was rearranged to abietic acid.From the strong lzevorotation of abietic and levopimaric acids it is probable that they have the same configuration at C13. neoAbietic acid which is strongly dextrorotatory (p. 44) must by this reasoning have the opposite configuration to abietic acid a t C,,.94 Since manool has been related directly to dehydroabietic acid (p. 55) it must have the same configurations as abietic acid at C, and C12. This also establishes the configurations of manoyl oxide (p. 54) ketomanoyl oxide (p. 54) a,nd sclareol (p. 53) a t these centres. Very recently,f25 L. Ruzicka R. Zwicky and 0. Jeger have shown that agathenedicarboxylic acid has the same ring fusion as for rings A/C of abietic acid. When the agathenedicarboxylic acid derivative (LXV) (p. 49) was dehydrogenated by the bromosuccinimide method it gave the correspond- ing dehydro-derivative (CXIV) converted by standard methods into the J.1945 1197. l P 3 D . II. R. Barton Cheit~. und Id. 1948 638. lZ4Ber. 1943 76 1257 1261 135 HeEv. Chi???,. Actu 1948 31 2143 ; the author is indebted to Dr. 0. Jeger for a copy of this paper before its publication. * Treatment of the various diliyciroabietic acids with strongly acid rcagents gives an isomeric lactoae C20H3209 m.p. 130-131° [alD ca. - 2' (in alcohol). This substance is usually considered t o be lactonised at the CIS position. Its formation in this way has been taken to support a cis A/C ring fusion since it caiinot possibly be built up on models if the A/C fusion is trans. The accepted formula for this lactone must be in error and it has been suggested elsewhere lZ3 that its formation involves a migration of the C, methyl group compa,rable to tkat taking place in the genesis of Westphalen's diol .62 QUARTERLY REVIEWS partly aromatic hydrocarbon (CXV) which was obtained also from the manool derivative (XCI) (p. 55) by reaction with methylmagnesium iodide to give the diene (CXVI) and dehydrogenation of the latter with bromo- succinimide. These experiments also provide an unambiguous proof of the point of attachment of t,he C,,-methyl group in agathenedicarboxylic acid. Both the agathenedicarboxylic acid derivative (LXV) and the manool derivative (CXVI) have strong lzvorotations similar to that shown by abietic acid. For this reason t'hey probably have the same configuration a t C, as in abietic acid.Assuming that inversion at this centre has not occurred during their preparation this may be taken as evidence for the same configuration at C, in agathenedicarboxylic acid manool manoyl oxide lsetomanoyl oxide and sclareol as in abietic acid. These conclusions on stereochemistry are summarised in the table. (LXV.) (CXIV. ) Abietic acid . . . . . isoRbietic acid . . . . neoAbietic acid . . . . Dextropimnric acid. . . Levopimaric acid . . . Podocarpic acid. . . . Agathenedicarboxylic acid. Manool . . . . . . Manoyl oxide . . . . Sclareol . . . . . . Ketomanoyl oxide . . . Ferriiginol . . . . . (XCI.) ! (CXV.) (CXVI.) C02H at C arid H at C,,. cis cis C i S Relationship between H at C, and Me at CL2. trans trans trans trans truns tram trans trans tram trans trans trans Me at C, and 11 at C13.trans cis ? cis ? > trans trans ? trans ? trans. ? trans ? tram ? - - Relationships between Di- and Tri-terpenoids.-This is a field of the greatest importance for t,he find clarification of the structures of the BARTON THE CHEMISTRY OF THE DITERPENOIDS 63 triterpenoids. Important advances can be expected therein in the near future. The formulation of rings D and E of ambrein (CXVII) 126 has been elegantly confirmed by the experiments of L. Ruzicka 0. Diirst and 0. Jeger.127 When ambrein is oxidised it affords7l2* amongst other products a saturated lactone (CXVIII) which can be degraded through the cor- responding saturated acid (CXIX) t o an acid identical with that (LXXXIX) obtained previously from inanool. These cxperiments also establish that the ring fusion between rings D and E (CXVIII.) in aibrein is trans (see above).(CXIX .) (LXXXIX. ) 1 E I \/ /\ (CXX.) (CXXI.) /y -+ C0,H C0,Me C0,bIe (CXXIV. ) (CXXIII . ) (CXXII.) A similar important correlation has now been established between oleanolic acid probably (CXX) and manool through a common degradation product of the former and of a m b r e i ~ ~ . l ~ ~ Pyrolysis of the keto-lactone (CXXI) obtained from oleanolic acid aEordec1 amongst other products las E. Leclerer and D. Morcier Exprientia 1917 3 18s ; 0. Jeger 0. Diirst and 127 Ibid. 353. lZ8 L. Ruzicka and F. Lardon ibid. 1946 29 912 ; E. Ledercr F. Marx D. Mercior and G. Perot ibid. 1354. 129 L. Ruzicka H. Gutmann 0. Jcger and E. Lederer iGid. 1948 31 1746 ; t h o author is indebted to Dr. 0. Joger for a copy of this paper before its publication.L. Ruzicka Helv. C h i m Acta 1947 30 1859. 64 QUARTERLY REVIEWS a keto-ester (CXXII) reduced by the Wolff-Kishner method to the saturated acid (CXXIII). The latter was also prepared from ambrein in the following way. Oxidation of ambrein with potassium permanganate gave amongst other products the saturated hydroxy-acid (CXXIV) which by dehydration and then hydrogenation was converted into (CXXIII). These experiments show that the A/B ring fusion in oleanolic acid must be trans in agreement with the X-ray data of G. Giacornello.l30* The Reviewer wishes to express his indebtedness to Sir John Simonsen F.R.S. for many valuable discussions and t o Mr. C. J. W. Brooks for help with the manuscript. 13* Gazzetta 1935 68 363 ; compare L. Ruxicka and H. Gubser Helv.Chim. Acta 1945 28 1054. * An account of the chemistry of the minor diterpenoids cryptomerene and cupres- sene will be found e1sewhe1-e.~ Cafestol (cafesterol) and kahweol have not been considered in the present Review although they are possibly diterpenoid in character.g4