[COXTR1HKJ'1.IO~FROM T H E ClWhlICAL KESEARCII LABORATORY UF ~ ' U L A K O I I I C O K P O K A T I O N AYn ' I t i E C O X V H R S l i x [ & M O R I A I . 1.ABORATORY O F HARVAIID UNIVERSI 1'Y
1
The Total Synthesis of Quihine R v 11. B. WOODWARD AND W. E.DOERIXG
The c$mination of the structural investigations on quinine in the proposal of the correct structure (I) in ,1908l may be considered the point a t which rational efforts toward total synthesis could be 4
3
CEI~--CI1--CH
- CI1 --CH
cri CI I
i
9
C U ( 0 H J--CI I
S-CI-1
had shown6 in 1853 that the 'ciiiclion:t :ilkal(~ids, on heating with tartaric or sull'uric aC-C-N-Q a
e C+ >CH-(!!-N=O
a e
(cf. note 41) plncw a high barrier in the way of proton release from
c. a.
II
CQCHi
OeN 0
-e le c I I1
(Nj
XLV alkali hydroxides, normal Hofmann elimination took place. With 60% sodium or potassium hydroxide, vigorous evolution of trimethylamine commenced a t 140°, and with simultaneous hydrolysis of the ester grou and the amide link, homomeroquinene (XLVIp's was formed.46 At* (44) Know, B I I . , 44, 184 (1889); WillstBtter, ibid., 48, 3288 (1895); Straw. Ann.. 401, 374 (1912).
(45) We wete able to p d with the synthesis d (XLIV), with ConBdence that the elimination would take place in the denired direction, with the introduction of a vfnyl m u p and formation d homomecoquinme. rather than with the formation of the ethylidene homer (XLVIII), on the basin of a large bcdy of evidence relating ta decomposition of quaternary hydroxides
+
-..
K& ]OH The Rojmomm R k k t ita original form &ed that in ruch dsompositbnw ethylene was eliminated, if poscllble. in prekencc to any OLhCr ol&. More gena&, it hor been found subsequeutly that the ole%% mat r d y dimhted b derived from ttut group containing the largest number of hydrogen etomr, on the carbon atom /3 to the nitrogen atom
PA-&-). I I
CHICHICOOH
H
XLVIII
"heme raulta indicate clearly that within
THETOTAL SYNTHRSIS OF QUININE
May, 1945
867
tempts to isolate the free acid directly from the distillation, and was obtained in 96% yield (overreaction mixture were not successful, but when all from [XLVII]) as a clear, colorless, viscous oil. Condensation of the N-benzoyl ester (XLIX) the reaction product was warmed for a short time in neutral aqueous solution with a slight with excess ethyl quininate (11, R = GHb) was excess of potassium cyanate, subsequent acidifica- efTected by heating the fused mixture of reactants tion resulted in the separation of pure N-uramido- with dry sodium ethoxide.l*J1 The crude interhomomeroquinene (XLYII), m. p. 165.2-165.8O mediate &keto ester (L),obtained as an alkali(dec.). The over-all yield in the conversion of CH-CH-CH-CH=CH*
I
CHrCHsCOOH
XLVI
CHa I
t
XLV~I
the crystalline oximino ester (XLI) to (XLVII) was 4270. N-Uramidohomomeroquinene (XLVII) was readily converted in quantitative yield into a mixture of homomeroquinene hydrochIoride and ammonium chloride on boiling with aqueous 0.1 N hydrochloric acid.“ From the former, by treatment with silver oxide, followed by hydrogen sulfide, pure dZ-cis-homomeroquinene (XLVI), m. p. 219-220°, was obtained. For synthetic purposes, the crystalline mixed chlorides, directly from the cleavage reaction, were treated with dilute ethanolic hydrogen chloride. Direct beqzoylation in chloroform solution of the resulting Cu-cis homomeroquinene ethyl ester hydrochloride with benzoyl chloride and potassium carbonate paste gave N-bemylhomomeroquinene ethyl ester (XLIX), which was purified by molecular CHtCH&OOEt
iYcH=cH* XLIX a dngle poup, elidnation shontd take place in the direction pf -the moet heavily hydmgen-aubatitutai &carbon atom, and t h t comlusion haa been verified experimentally in a n u m w of casea. For a very i n t d n g and thorough. it not entirely convincing, dirugion of theme points from Ute theoretical point of view, cf. Hughes and Ingold. h 8 W . Par&9 Soc., W , 687 (1941). It not out of place to point out here that other elimination mctions involviqg a subrtltumt (#. #., Br.OH) om the 10-poaitionwould have giwn(XLVII1). (46) It u evideat that thin reactionis the point at which, with the introduction of the double bond, the asymmetry at C. 10 is destroyed, and that the two podble epimem of (XLIV) difiering in configuratioll at that -tar, give the -me rtaoochcmicOllyhomogeneour cishomoMIosulnmc (47) Qrpntltatfve studies d the cleavage of simple u r r ~[e. g.. cf. F a d t t , Z. Hysik. Chin., U , 610 (1902); J. Cbrr.Soc.: W , 494 (lO0s)l on w+b t h f r procedure WM baaed indicate that the reaction is notm direct hpirolpb of the d d e link, but rather, an irometirotion to a iubst3tuted ammonium cyanate, followed by hydrolyrk pf the cyanate ion ( b o - xih H+ -c C C ~ NIW m e eaae with which the uramido p u p is both i n t r o d d a d removed nuggesta that it should be of general value in the separation and character&ation df a & t o compound.. It haa hithato Imd limited w in wovk with 0-adwo midi [d. Dakin. Am. C h m . J.. M,4$(39lO)l in w+ ea& the dtaation t complicatd by the tendd the &cq.mido duinthes to pam readily into hydantolak
+
+
+
L
soluble oil, gave, on hydrolysis with boiling 6 N hydrochloric acid, crude dl-quinotoxine (IV, R = OCHJ as an orange oil in 50Y0 yield (from XLIX). Attempts to resdve the racemic alkaloid through its salts with d-tartaric acid were u11successfd” but the dibenmyl-d-tartrates were readily separable by crystallization from methanol, that of the dextrorotatory isomer being the less soluble. Pure d- uinotoxine dibenzoyl-dtartrate, m. p. 185.5-186 8,showedno depression in melting point on admixture with the salt of the natural alkaloid. On regeneration from the salt, pure synthetic d-quinotoxine was obtaiaed as a very light yellow viscous oil, [a]D +a’.For further characterization,. the .synthetic base was converted to the beautifully crystalline d-quinotoxine-d-tartrate hexahydrate, m. p. 55-63’, and from the latter, by crystallization from absolute ethanol, to the anhydrous dquinotoxine acid dtartrate, m. p. 150-153’. The melting points of these two salts were not depressed on admixture with samples of the corresponding salts from natural quinotoxine. Table I recapitulates the coni(48) This in spite of the fact that dLtutuiC add is -My red v a b l e by tlpturpl d-quinotoxine. It t a little known albeit h*toridly an important fact, that this, and the similar nrolutJon.by dnchotoxinc were the fint examplea [ M e w , Compt. rsad,, W, 181 (1853)1 of the now universally used method of resolution d a -ic compound by combination with an active material. followed separation of the resulting diqstweomw. Confurion . c a to ~ have resulted from the fact that quinotoxine and dnchotoxine (which wae discovered by Pasteur [ref. 61) were in 186a known 01qPtnidne and dnchonidne, and the rwemblaace to cinchonidine hpr led come to believe that the lattv alkaloid IU tbe e s t resolving agent [cf Lowrx. “Optkal Rotatory PDWC~,” L a m a p n a , Green pnd Q.. 1936, p. 841 w h i h othur have nmurneb W the well-tnpm EWquinine and cinchonine were wd. Further,Pasteur pnDO u p u i mental detrlk d his work; in fact the original note to the French Academy WPI lusely rmmrned .rDh other mattun, and the dbc o v q ia mentiomsl only in p d n g , there, a d I o t a , in the cotme .of a ser&a oflectures sucllmrtiring hi.mrk orrmokdnr wmmetry [“Reaeardwon the Molecular krmmctry of N.trrrol Ocgank RodWAS,” (1860); “Alembic Club Rednt.,” No. 14, p, 411, Further. the pabllrhed Wormation e k w h u e w&b rysrd to tke tutraw of qlrindoxlox(nc t fryrmcotnry at beat (d. ref. 7. Even Bdlrtcin emc trinr BO ralarma t o Putam‘s w S t t . A nof orn uperierica with the above Mutlolrand 4 t h the e b u r a a f v t b n o t the s u 4 o t o h e t a r h t a r ; n i l l bebtraclintho-N Putoft&
-pa.
K. 13. WOODWARD AND W. E. DOERING
868
parison of the natural and synthetic alkaloids and their derivatives. TAULE I icull)
d-Quinotouine
(1-Quinotoxine
1 rrnlurol)
(synthctrc)
+44""
$43"
35 - (53" d-Tartra tc kexuhydrtrh .ij - tisob Anhydroti5 d-tartratc 162 -155" 150 -153' 185 5-186" Dibeiiioyl d-tartrate 185 5-186" Hessc. ilriti , 178, 2(iO (1875) * CJ ref. 7
Crystallizatioli o f tlie tlibeiizoyl-l-tartrates" from the crude quiiiotoxinc regenerated from the inother liquors o f thc above resolution gave pure l-quinotoxiiie dibcnzoyl-l-tartrate, m. p. 185-186', from which 1m-e l-quinotoxine, [ a ] -Go, ~ was obtained. I n view of the cstablishctl conversion of quino-
toxine to quinine,I2 with the synthesis of quinotoxine the total synthesis of quinine was compkte.
Experimental Aminoacetal.285D-Into a aolution of 88.8 g. of chloroacetal ( b p. 151.4-156.2") in oiic liter of absolute methanol cooled t o 0' dry animotiia pasbcd until 283 g. had been absorbed. The reaction lure was then heated ten hours a t 140' in the autoclave. The colored solution was concentrated on the stcatn-bath t o 500 cc., 100 cc. of 5% aqueous potassium hydroxide was added, and concentration was continued until the vapors no longer burned. The solution was saturated with sodium chloride, treated with 100 cc. of 50% aqueous potassium hydroxide and extracted continuously with ether overnight. Concentration of the ether extract gave an oil from which on fractionation in nucuo 24.1 g. (72.5%) of aminoacetal (b. p 99-103' [lo0 mm.]) was obtained. Using twice the quantity of chloroacetal and the same quantities of methanol and ammonia, 40.0 g. of aminoacetal (60%) was obtained. By fractionation of the higher boiling material'from fourteen such double runs through a two-foot column packed with glass rings, 83 g. (10%) of diacetalylamine (b. p. 189.0-189.4' (100 rhm.); oxalate, m. p. 172.0-172.6' dec.) was obtained. m-Hydfoxybmzaldehyde.-To a solution of pure stannous chloride (450 g.) in 800 cc. of concentrated hydrochloric acid cooled t o 5', 100 g. of m-nitrobenzaldehyde was added in one portion. The temperature rose slowly at first, and then very rapidly. The reaction was quenched (by immersion in a n icesalt-bath) a t such time that the maximum temperature attained by the reaction-mixture was cu. 100' (the proper time for quenching depended on the size of the jUn, and the efiaency of stirring and cooling). During tfiis period vigorous mechanical stirring was necessary. After quenching, the mixture was stirred very slowly for one and one-half hours. The orange-red paste of m-aminobenzaldehyde stmnichioraride was collected on a au possible, and suspended sintered glassfunnel, sucked in 600 cc. of concentratr;d hydrochloric acid. While the reaction mixture was stirred continuously, a solution of 46 g. of sodium nitrite in 150 cc.of water was added gradually from a separatory funnel whose stem dipped below the surface of the solution; the rate of addition WEIS regulated to keep-the temperature of the reaction mixture between (49) M ~ g l - l - t n r t d ncid c was prepartd from l-tu+prie acid [Haskinsand Hudson, THISJOURNAL, 61,12fM(1989) 1 by the method d for the preparation of the $-isomer [Butler and Cretcher, ibid., 66, 2605 (leas), Zetzsche and Hubachrr, Hslv. Chim. Acto, S, 291 (1926)l. (50) W e mu indebted to Dr. L. P. Kyrides and the Monsnnto Chemical Co. who placed at our dimposol approximntely'two kilos of smiwacetd, prrnmably prepared by the method dcsrribed here, or an adaptation of it
Vol. A7
4 and 5". Stirring was continued further for one hour; the precipitated diazoniuiii 3aiinichloride was tlicii collected on a glass futinel, suckcd dry, and deconiposetl b y additioii in portions to 1700 cc. of boiling water (in a large vesscl). Lhriiig reaction, tvater lost by evaporation was replaced. The solutioti was then clarified with Norit (4 g.) and filtcred hot. Ovcrnight 48-52 g. (59-647c) of m-hydroxybenzaldehyde separated as orange or toti crystals, ni. p. 9!J-10l0. Farther purificatiori was cffected by crystallization after solution in ca. one liter of Imiliiig benzene, clarification with Norit, and coiiccntraLioii to 600 cc. (41-45 g. [51-Xi%], tu. p. 101-102°). fir-Hydroxybenzylideneaminoacetal(XIV).-To 170.0 g. of m-hydroxybenzaldehyde was added 80.0 g. of aniinoacetal. The reaction niisture became warm and homogeneous, and was warmed one-half hour on the steanibath, 300 cc. of benzene was added, and water was removed by filtration through solid anhydrous sodium sulfate. Sufficient ligroin (70-90") was added at. 35' t o cloud the solution, which on scratching and seeding, deposited long silky needles of. m-hydroxybenzylideneaminoacetal,in. p. 67.2-67.8". On long standing, the needles changed to a heavy prismatic form of the same melting point. From the concentrated mother liquor a second crop of pure material was obtained. The total yield was 128.4 g. (94'3 ). 7-Hydroxyisoquinoline (XV). (Method 1.)-m-Hydroxybenzylideneaminoacetal (50 g.) was dissolved as rapidly as possible by shaking and stirring in 118 cc. of 76.0% sulfuric acid previously cooled to 0'. The reaction mixture was allowed to stand twelve hours at 3', and then thirty-six hours a t room temperature. After pouring onto ice (700 g.) the brown solution was neutralized with ammonia and buffered with sodium carbmate. The precipitated crude 7 - h y d r o x y i s o q u i t e (30 g., d r y ) was sublimed in five batches at 2 mm. (bath temperature 160'). The sublimate (27 9.) on crystallization from 95% ethanol (36 g./lO g. phenol) gave 20 g. (64%) of nearly pure 7hydroxyisoquinoline as very faintly yellow plates, m. p. 227-229 '. Alternately the crude phenol could be purified through its sodium salt. In a typical experiment, 32 g. (dry)of the. crude material was dissolved in 160 cc. of water cbntaining 64 g. of sodium hydroxide. The yellow plates of the sodium salt which separated from the cooled solution were collected and recrystallized 'from three t o four times their (wet) weight of water containing one to two times'their (wet) weight of sodium hydroxide. On regeneration (24 g. obtained) and, recrystallization (Norit) from methanol (cu. 500 cc.), 21 g. (60%) of quite pure, very light tan 7-hydroxyisoquinoline, m. p. 229-231', was obtained in two successive crops. Method 2.-m-Hydroxybenaaldehyde (10 9.) and aminoacetal (11.4 g.) were heated on the steam-bath for half an hokr. The 'reaction mixture was dissolved in benzene, most of the water was removed (by anhydrous sodium sulfate, or by filtration through benzene-wetted paper), and the remainder driven off by evaporation twice with excess benzefie, the last time to a volume of 20-30 cc. T o the cooled benzene solution, 25 cc. of 80% sulfuric acid (previously cooled to Oo) was added with vigorous swirlin or shaking. The t a c t i o n mixture was allowed t o stanfi overnight in the c d d room (3-5") and then for twenty-four hours at room temperature (with occasional cooling if the temperature rose). The acid layer was then separated, combined with an aqueous washing of the benzene layer, and neutralized and buffered to precipitate the crude phenol, which w a s purified through the'sodium salt as above+$. method 1). The pure phenol (7.6g., 84%) was obtained as light yellow plates, m. p. 229.3-230.4". Under otherwise comparable conditions, variation in the strength of the acid used gave-the following results (10-g. runs) : Acid strength, % 72 76 78 80 82 84 Yield, q, 30 43 59 64 A4 31 With larger runs (100 g. of m-hydroxybenzaidehyde) better results were obtained' with 82% sulfuric acid. In nineteen runs (50-100 g.) using method 2 (82% acid) and
May, 1945
THETOTAL SYNTHEESIS OF QUININE
the sodium salt method of purification, pure 7-hydroxyisoquinoline was obtained in an average yield of 60%. The purest samples of 7-hydroxyisoquinoline melted at 829.5-230.5'. A small sample was converted t o the acetyl tlerivative by treating with excess acetic anhydride, blowing off excess reagent, and crystallizing the residue from ether. Pure 7-acetoxyiaoquiuoline has m. p. 73-75'. S-Hydroxyisoquinolie (-I).-The mother liquor after rcnioval of the sodium salt of 7-hydroxyisoquinoline from :i2 g. of the crude mixed pheiiols was neutralized and buffcred. When the very crude precipitate (5.9 g.)'was subliiiied a t 2 r i i t i i . (bath NO'), 1.6 g. of light yellow crystalline material was obtained. The solution of the latter iu 8 cc. of water containing 3.2 g. of sodium hydroxide, when seeded with the sodium salt of 5-hydroxyisoquinoliiie, deposited pure material, from which on neutralization 1.2U g. (ca. 5y0)of 5-hydroxyisoquinoline, m. p. 227-2%82", was obtained. This material did not depress the melting point of authentic%5-hydroxyisoquinoline; mixed with an equal weight of 7-hydroxyisoquinoline, it melted quite sharply a t 181-185°.*4 Further crystallizatioii from iiiethaiiol of the 5-hydroxy isomer raised the melting point to 23 1-23s0, and acetylation (as above) gave 5-acetoxpsoquinoline, ni. p. 59-60'. A mixture of the 5- and 7acetoxy compounds melted below room tctnperature. Neither of the acetates was stable on staiidiiig for long periods, and both were readily hydrolyzed on shaking with aqueous 2 N sodium hydroxide. 7-Hydroxy-8-piperidinomethylisoquinoline(XVIII).-To a solution of 6.0 K. of 7-h~droxvisoauinolineand 3.5 P. of piperidine in 30 cc. of 95% ethanol, 2.7 g. of 35% aqueous formaldehyde was added. The solution was heated for six hours on the steam-bath, and then evaporated t o "dryness." The ethereal solution of the dark red oil was filtered from a small amount of ether-insoluble material, and diluted with petroleum ether. From the seeded solution, yellow crystals separated which gave pure material (1.1g., 11%) on one recrystallization from hexane. The combined mother liquors were concentrated and treated with 40 cc. of aqueous 2 N sodium hydroxide. The warm solution was decolorized (Norit) and cooled; a moldlike mat of extremgly f l u e yellow needles of the sodium salt separated. The salt was dissolvCd in water, the resulting solution was neutralized, and the colorless oily product was extracted with ether. The ether was evaporated; crystallization of the residue from hexane gave a further 3.6 g. (36%) of the pure piperidinomethyl derivative. Further .crops were obtained by combining the residue from the hexane mother liquors with material regenerated from the sodium salt mother liqimrs, and putting the total, material through a fresh sodium salt-hexane purification. In this way a total of 61% of pure 7-hydroxy-E-piperidinomethylisoquinoline was obtained as beautiful stout glistening blocks, m. p. 81.5-82.5'. Anal. Calcd. for CI~HI~ONZ: C, 74.36; H, 7.49; N, 11.57. Found: C, 73.79, 75.18; H.7.65, 7.55; N, 11.86. The substance was reduced very slowly and incompletely in ethanol over Adams platinum catalyst. The combined material from several long-continued hydrogenations was dissolved in aqueous 2 N sodium hydroxide. Ether extraction gave an acid-soluble oil, insoluble in strong aqueous base, which gave a crystalline carbonate, recrystallized from alcohol-ether, m. p. 103-110' (de.). A w l . Found: C, 69.59; H, 8.37. The residual, alkaline solution on seeding deposited much sodium salt of 7-kydroxy-8-piperidinomethylisoquinoline, which was removed. Neutralization of the residual solution precipitated an oil, sparingly soluble in ether. On long standing the oil crystallized in part. The solid material (probably XIX),m. p. 155.5-156.5", on recrystallization from ethanol, separated in needles. m. p. 158.5-159.5'. Anal. Calcd. for G&,OSNI: C, 74.04; I€, 7.48. Found: C, 74.18; H.7.44. 7-H droxy-S-methylieoquinoline (XX): (a) By Hydrogen &xchange.-7-Hydroxy-€&piperidinornethylhquinoline (0.45 8.) in 10 cc. of methylnaphthalene was boiled for
869
twenty-two hours over 0.07 g. of 30% palladium+harcoal catalyst. Seventy-three cc. of hydrogen was evolved. The catalyst was removed, the solution was extracted with dilute acid, the acid extract was made basic and extracted with ether. The alkaline solution was neutralized, and extracted twice with ether. The extracts were evaporated, the reddue was dissolved in alcohol, and treated with formaldehyde (to convert any 7-hydroxyisoquinoline present into unsublima1,le material of high molecular weight). The alcohol was removed; sublimation (2 mm.) of the residue gave a small amouiit of crystalline material, which on recrystallizatioii from methanol separated in shiny platelets, m. p. 227-228', which did not depress the melting point of an authentic sample of 7-hydroxy-8methylisoquinoline (b or c, below); mixed with 7-hydroxyisoquirioline, in. p.
