Clemo and Cocker The Constitution of +Suntonin. Part I. 11. The Constitution of +Santonin. Part I. By G. R. CLEMOand WESLEYCOCKER. In addition to I-santonin and fl-santonin a further compound C,,H2,0, has been isolated from certain species of Artemesia. Its chemistry is now discussed but no final structure can yet be given to the new compound, which is named +-santonin. INthe course of an examination of a number of species of Artemesia collected from the N.W. frontier of India Messrs. T. and H. Smith Ltd. of Edinburgh discovered a further crystalline compound in addition to I-santonin (111; Clemo Haworth and Walton J. 1929 2368; 1930 1110; Clemo and Haworth J. 1930 2570) and (3-santonin (Clemo J. 1934 1343). (1.) (11.) (111.) (IV.) The new compound C,,H,,O, termed $-suntonin is Imo-rotatory .Whilst I-and p-santonin impart little colour to concentrated sulphuric acid +-santonin gives an immediate dark brown coloration. Further Z-and p-santonin and artemisin (IV; Tettweiler Engel and Wedekind Annalen 1932,492 105) are unaffected by alkaline sodium nitroprusside but +-santonin gives a deep red coloration when the test is performed according to Jacobs and Hoffmann (J. BioI. Chem. 1925 67 333). When however the test is modified according to Paist Blout Uhle and Elderfield (J. Org. Chem. 1941 6 273) the latter compound gives a pink coloration after the addition of the third drop of sodium hydroxide when the solution is still acid to Congo red and this colour shows little fading. When the fourth drop of alkali is added and the solution is faintly alkaline to phenolphthalein a deep red coloration is produced which fades after three or four minutes.Addition of further drops of nitroprusside again produce the deep red coloration which changes to blue-green on acidific-ation. Potassium ferrocyanide has no effect. These results strongly suggest the presence of a p~ unsaturated lactone. Tollens' reagent is unaffected by either I-or p-santonin but like the cardiac aglycones and other a-p unsaturated lactones a silver mirror is very slowly produced by @antonin. Immediate reduction of the reagent usually takes place with a @-y lactone (Jacobs Hoffmann and Gustus J. Bid. Chew&.,1926 70 1). 4-Santonin gives no colour reaction with tetranitromethane but the absence of ethylenic bonds cannot be concluded from this since protected or 61-p double bonds very often fail to give a reaction with this reagent (Haworth Ann.Reports 1937 34 328). Tested side by side tincture of digitalis and an aqueous-alcoholic solution of 4-santonin of comparable strength give similar cardiac effects. Depression of conduction and cardiac inhibition take place similarly in each case but neither I-nor p-santonin shows similar reactions. In the latter cases there is slight augmentation of the heart beat probably due to the alcohol used as solvent. These results point to the existence of an un- [1946] Clemo and Cocker The Constit&ioH of +Santonin. Part I. saturated lactone in +-santonin (compare Kraft Arch. Pharm. 1912,250 126 ; Windhaus Bohne and Schwieger Ber.1924 57 1388). 4-Santonin gives an oxime and a 2 :4-dinitrophenylhydrazone,but all attempts to prepare a piperonylidene or j5-nitrobenzylidine derivative have so far failed. Recently however it has been possible to obtain a nitro- genous compound C,,H,,O,N presumably a 2 3-substituted quinoline by condensation with o-aminobenz- aldehyde thereby showing the probable existence of the -CO-CH,-system. 4-Santonin has so far failed to yield an osazone. +-Santonin is not immediately soluble in sodium hydroxide but on shaking it slowly passes into solution from which it is reprecipitated unchanged even on careful neutralisation with dilute acetic acid. In addition several attempts to prepare a methyl ester from the sodium salt yielded a substance which gradually lost methyl alcohol and yielded 4-santonin.The evidence therefore points to the existence of a relatively stable y-lactone and together with the carbonyl group accounts for three of the four oxygen atoms present in the molecule. 4-Santonin yields a monoacetyl derivative with difficulty and in poor yields and this together with Zere- witinoff determinations points to the probable presence of one tertiary hydroxyl group. Incidentally when the above-mentioned monoacetate is hydrolysed with sodium hydroxide and the solution acidified a lactone C15H2004, isomeric but not identical with 4-santonin is obtained. The structure of this lactone has not been established but in view of the mobility of tertiary hydroxyl groups it would not be surprising if a molecular rearrangement were to take place during acetylation.No change in structure has yet been encountered when +-santonin is treated with potassium hydroxide under a variety of conditions (compare Paist Blout Uhle and Elderfield Zoc. cit,). A point of interest is that whilst I-and p-santonin are soluble in cold water to the extent of only one part in four hundred and one part in six hundred respectively 4-santonin is soluble to the extent of one part in forty a property which may be due to the hydroxyl group in the last compound. Catalytic reduction of 4-santonin by hydrogen in the presence of palladised charcoal and at 1 atm. yields a dihydm-compound C ,H9204,which dissolves readily in sodium bicarbonate with evolution of carbon dioxide and is reprecipitated unchanged even on warming with mineral acid.This dihydro derivative possesses a carbonyl group and it is evident that its preparation involves the hydrogenolysis of the lactone ring. Treat-ment of the acid with acetic anhydride yields a %lactone. Further hydrogenation using Adams’ catalyst (PtO,,H,O) yields an acid hexahydro-+-suntonin C&@4 which is also produced by reduction of 4-santonin itself under the same conditions. The latter acid has no carbonyl group and yields a rnonoacetyl compound which is also a &lactone. It appears therefore that two of the four additional hydrogen atoms are used in the reduction of the carbonyl group to the corresponding alcohol. When the reduction of 4-santonin with Adams’ catalyst is carried out quantitatively it is found that the absorption of the first molecular equivalent of hydrogen is rapid and the second and third equivalents progressively slower.Dihydro-#-santonin is readily reduced with sodium amalgam to an acid tetruh~dro-+-santoni~ C,,H2404 which fails to give the reactions of carbonyl compounds but is readily converted to a monoacetate also a &lactone. The tetrahydro compound is slowly reduced to the hexahydro compound by Adams’ catalyst and there appears to be no doubt that the reduction of the tetrahydro to the hexahydro compound corresponds to the third (slow) stage in the reduction of +-santonin itself. It appears therefore that the first stage in the catalytic reduction is the hydrogenolysis of the lactone ring the second is the reduction of the carbonyl group and the third is probably the reduction of the unsaturated lactone.The hydrogenoly sis of unsaturated y-lactones has been widely investigated by Jacobs and his co-workers and more recently by Elderfield and his co-workers and it now appears certain that the cardiac aglycones and other @-substituted a-p butenolides undergo hydrogenation by uptake of one molecule of hydrogen to give saturated lactones (Paist Blout Uhle and IZlderfield Zoc. cit.). On the other hand f3-y unsaturated lactones yield varying amounts of deoxy acids on catalytic reduction (Jacobs and Scott J. Biol. Chem. 1930 87 601) with uptake of two molecules of hydrogen. +-Santonin however takes up one molecule of hydrogen to yield an unsaturated carboxylic acid the double bond of which is only hydrogenated with difficulty This evidence favours the existence of a fL-y double bond as in structure (I),in which compound there is a ‘‘ pro-tected ” p~ double bond which might be expected to be resistant to hydrogenation.In view of the ready hydrogenation of the double bonds in I-santonin using palladised charcoal (Wienhaus and Oettingen Annalen 1913 397 219; Asahina Ber. 1913 46 1775) the possibility that the third molecule of hydrogen absorbed in the reduction of $-santonin is taken up by a double bond similarly placed to either of those in I-santonin is remote but as a+ double bonds are frequently inert or sluggish towards catalytic reduction (compare Haworth Zoc. cit.) the evidence cited does not rule out the possibility that this grouping is present in +-santonin.Its absence is however indicated by spectroscopic studies. Compounds with such a grouping have an absorption band due to the carbonyl group in the neighbourhood of 3000-3200 A. and a much more intense band (log E = 3-8 to 4-2) between 2200 and 2500 A. Isolated carbonyl groups absorb near 2800 A. (Gillam and West .J. 1945 95). I-Santonin has absorption bands at 2360 A. (log E = 4-05) and 3250 A. (log E = 1-54) corresponding to a carbonyl group conjugated with an a-p double bond (compare Gillam and West Zoc. cit. who show that piperitone has bands at 2350 A. ; log E 2 4-25 and 3210 A. ; log E = 1-73) On the other hand +-santonin shows the band at 2900 A. (log E = 1.56) charac- Clemo and Cocker The Constitution of +Santoiain.Part I. teristic of the isolated carbonyl group but it has much stronger absorption (no maximum) at 2000-2200 A. (log E = 2.2-4.0) than would be expected in this case. This spectrum has some similarity to those of p-cyclo-hexyl-Aa-butenolide and vinyl acetate (see fig.) supporting the conclusion that $-santonin possesses an unsaturated lactone ring. When 4-santonin is treated with one molecule of bromine it yields a nzonobronzo compound C,,HI,O,Br which is readily reduced with zinc and alcohol to #-santonin. Three to four molecules of bromine yield a tribronzo compound C,,H,,O,Br, which is a lactone and which is reduced in the presence of palladised charcoal to the monobromo compound. Titration by the method of Winkler (Pharm. Zentr. 1924 65 385) gives a bromine uptake of four atoms which with the above evidence points to the substitution of one bromine atom and the addition of two others to a double bond.Bromination under the conditions of titration points to rather than a-p unsaturation in the lactone ring. Dihydro-$-santonin on titration by Winkler's method also shows a bromine uptake of four atoms and it appears that the double bond is preserved during the hydrogenation of $-santonin with palladised charcoal thus supporting the view that only hydrogenolysis of the lactone ring takes place under these conditions of hydrogenation. When #-santonin is treated with 550,/ sulphuric acid or with anhydrous formic acid at 50" a dark red solution is quickly produced from which a crystalline compound C,,H1809 is obtained as the main product.This compound now called desmotroPo-t,h-san2onin is dextro-rotatory and like the dextro-rotatory derivatives of I-santonin is phenolic. It fails to couple with diazo- tised p-nitroaniline and gives only a dull red-brown colour but it is readily benzoylated and methylated in alkaline solution. It does not give a satisfactory Legal reaction nor is it reduced with palladised charcoal and hydrogen. Attempts to obtain a fully aromatised compound from desmotropo+-santonin by heating with selenium were unsuccessful ; the compound sublimed unchanged. Aromatisation is however readily effected by fusion with potassium hydroxide at 320" (cf. Andreocci Gawetta 1895 25 545). The compound obtained has the formula C,,H,,O corresponding to a dimethyl-naphthol and by analogy with the naphthol obtained from I-santonin it was anticipated that it would be a dimethyl-p-naphthol with the methyl groups in the same ring as the phenolic group.It is not identical however with 1 4-dimethyl-2-naphthol (Cannizzaro and Carne- lutti Bey. 1879 12 1575) which is obtained from I-santonin or with 1 3-dimethyl-2-naphthol or with -1800 2200 2600 3000 3400 3800 3 4-dimethyl-2-naphtho1 both of which have now been A.A. synthesised (see Part 11,following paper). It was finally I l-santonin11,pseudo-~anlonin. 111 Pipevitone. shown to be identical with 2 4-dimethyl- 1-naphthol IV /3-cycloHexyZ-Aa-butenoZide.V Vilzyl acetate. (Cornforth Cornforth and Robinson J. 1943 168), and we are indebted to these workers for a sample of their compound for comparison.It is clear therefore that in +santonin there is a carbonyl group at position 4. When $-santonin is treated with cold concentrated hydrochloric acid a denfro-rotatory compound,C,,H&, quickly separates. This compound is readily soluble in sodium bicarbonate with evolution of carbon dioxide and is converted into desmotropo-$-santonin on treatment with sulphuric acid. Attempts to obtain an oximino derivative gave a high-melting compound which contained nitrogen and apparently also sodium. Further work on the new compound is necessary; the fact that it is dextro-rotatory is significant. When t,h-santonin or its dihydro derivative is distilled with selenium an intractable syrup is produced but after reduction with amalgamated zinc and hydrochloric acid the dihydro compound yields a gum which on distillation with selenium yields an oil whose picrate is identical with that obtained from 1-methyl-7-ethyl- naphthalene.This hydrocarbon is also obtained from I-santonin (Clemo Haworth and Walton Zoc. cit.) and artemisin (Tettweiler Engel and Wedekind Zoc. cit.). It appears therefore that @antonin possesses a similar carbon skeleton to the compounds named and that in all three compounds the a-propionyl group is at position 8. It should be noticed that in the production of 1-methyl-7-ethylnaphthalene a methyl group is lost which is characteristic of compounds containing angular methyl groups or derivatives of these. This evidence together with the fact that 2 4-dimethyl-1-naphthol may be obtained as indicated above shows that in #-santonin there is a methyl group at position 1and a methyl group or a derivative of this at either position 5 or 10 and that during the treatment with acid the latter methyl group migrates in some way to position 3.[1946] Clemo and Cocker The Constitution of $-Santonin. Part I. Evidence is presented below to show that the hydroxy group is at position 10 and therefore the angular methyl group is at 5. The migration of the methyl group from position 5 to position 3 during the treatment of $-santonin with sulphuric or formic acids is difficult to explain as there appears to be no analogy. Migration of methyl groups on treatment of 4-quinols with mineral acid is well known (cf.Clemo Haworth and Walton Zoc. cit. who give references to earlier work) but migration takes place to an adjacent position in all the cases mentioned by these workers. In the case of t,h-santonin the methyl group migrates to a position two carbon atoms away. This difficulty could be overcome by the postulation of a methylene bridge between carbon atoms 3 and 5 in which case simple hydrolysis and dehydration would lead to the required naphthol according to the following scheme. OH Ye Me El With such a methylene bridge it would be necessary to postulate a saturated lactone ring against which there is considerable evidence. The suggested tertiary hydroxyl group may be at positions 1 5 or 10. It cannot . be at 8 as in artemisin since the lactone is unsaturated nor can it be at 9 (only possible in the a-p unsaturated lactone structure) because of the ready re-formation of the +-santonin after hydrolysis.Furthermore dihydro- +-santonin does not undergo acetylation but readily gives a theoretical yield of a lactone C15H2003, on being heated with acetic anhydride and is regenerated on warming with either dilute sodium hydroxide or dilute hydrochloric acid. The new lactone is much less stable than +antonin. It gives a monoxhne on treatment with one equivalent of hydroxylamine but with an excess of this reagent a compound C1,).I&4N, is produced which combines the properties of an oxime and a hydroxamic acid. It is also obtained by further treatment of the monoxime with excess of hydroxylamine and it appears that the lactone ring is opened during this reaction.Observations on the effect of hydroxylamine on a limited number of y and 6 lactones indicated that the %lactones were converted to hydroxamic acids whilst the y-lactones were unaffected. Thus N-$-toluene- sulphonyl-6 6-dimethyl-2-morpholone (Cocker J. 1943 373) readily gave N-fi-toluenesulphonyl-N-P-hydro-xyisobutylacethydroxamic acid. Coumarin gave P-2-hydroxy-phenylacrylhydroxamic acid and anhydro-2-hydroxycycZohexyl-~-propionic acid gave a small yield of a nitrogenous hygroscopic substance which gave the usual hydroxamic acid reactions. All attempts to obtain a compound containing nitrogen from anhydro-2-hydroxycycZohexy1-a-propionic acid were unsuccessful and of course #-santonin itself gave no hydroxamic acid.The above evidence supports the view that the lactone C15H2003 mentioned above is of the &type which could only be the case if the tertiary hydroxy group were at 10. Therefore the angular methyl group is at 5. It should be mentioned that the lactone C,,H,,,O, is unaffected by catalytic hydrogenation using palladised charcoal. The evidence so far presented gives a good picture of the carbon skeleton of $-santonin with methyl groups at positions 1 and 5 the keto group at 4 the a-propionyl group at 8 and the hydroxy group at position 10. The position of the double bond in the lactone ring is however uncertain since there is evidence in favour of each possibility nor is it certain whether fusion of the lactone ring takes place at 9 as in (I)and (11)or in the alternative position 7.On biogenetic grounds and analogy with +antonin etc. the former is favoured. Further work in this field is in hand. EXPERIMENTAL. $-Suntonin crystallised from alcohol in colourless rhombic prisms m. p. 183-184" (Found C 68.0; H 7.7. C,,H,,O requires C 68.2; H 7.5%); [CZ]~,~O" -169"(c = 2-49in chloroform) ; cell dimensions a = 8.40 A. b = 10.1 A. c = 15.9 A. ; d 1.245. The available atomic models did not permit the setting up of accurate scale models of the struc- tures (I)and (II),but measurements made on rough models showed that these structures agreed fairly well with the crystal measurements. $-Santonin dissolved slowly in aqueous sodium hydroxide and was recovered on neutralisation with dilute acetic acid at 0"(Found equiv.wt. (by titration) 251 253,245; M (Rast),275; active hydrogen 1-2-1.3 atoms per molecule. C,,H,,O requires equiv. wt. 264). The mime crystallised from boiling water in long pointed prisms m. p. 203-204' (Found C 64.8; 64.9; H 7-5 7.8; N 5-0. Cl,H,104N requires C 64.5; H 7.5; N 5.0y0). The 2 4-dinitvo-phenylhydrazone crystallised from a large volume of alcohol as bright yellow prisms m. p. 257" (Found C 57.3 ; H, 5.4. C,,H,,O,N requires C 56.8; H 54y0). Condensation with o-AminobenzaZdehy&.-$-Santonin (0.2 g.) o-aminobenzaldehyde (0.1 g.) alcohol (1 c.c.) and sodium hydroxide (0.4 g. ; 20%) were shaken together until a clear solution was obtained and then set aside for 45 hours at room temperature. The liquid was diluted with water extracted thrice with ether and the aqueous portion carefully neutralised with dilute acetic acid.The precipitate was collected washed with water dried and recrystallised from dilute alcohol giving colourless needles m. p. 210" (Found C 75.4; H 6-4. C,,H,,O,N requires C 75-6; H,. 6.6%). This condensation product was readily soluble in dilute mineral acid and insoluble in .cold dilute sodium hydroxide. It dissolved in the latter on warming and was reprecipitated on neutralisation. The condensation product was undoubtedly a quinoline containing a lactone ring. Acetyl-#-suntonin. $-Santonin (0-5 g.) was refluxed with acetic anhydride (3 c.c.) and anhydrous sodium acetate D Clemo and Cocker The Constitution of $-Santonin. Part I. (1.0 g.) for 3 hours.Excess acetic anhydride was removed under reduced pressure and the residue extracted with water and crystallised twice from dilute acetic acid as colourless plates (0.15 g.) m. p. 187" [a]T"-69.2"(c 2-24in chloroform) (Found C 66.6; H 7.0. Cl,H220 requires C 66-7; H 7.2%). On hydrolysis with dilute sodium hydroxide and crystallisation from dilute alcohol colourless needles were deposited m. p. 205-207' (Found C 67.95; H 7-59. Cl,Hm04 requires C 68.2 ; H 7.504). This new compound was not identical with $-santonin. Oxime of acetate prepared 1n alcohol in presence of sodium acetate crystallised from dilute alcoHol as colourless prisms m. p. 196-197" (Found C 63.9; H 7.1. Cl,H2,0,N requires C 63.6; H 7.1%). Hydrogenation Expeyirnents.-(a) Using palladised chavcoal.4-Santonin (169.45 mg.) in glacial acetic acid absorbed 17.2C.C. hydrogen at 16"/755 mm. equivalent to 1.1 mols. hydrogen per mol. In a second experiment #+antonin (4.05 mg.) absorbed 0.50 C.C. hydrogen at 19"/766 mm. equivalent to 1.2 mols. hydrogen per mol. The absorption was com- plete in 1 hour in each case. (b) Using Adams' catalyst. #-Santonin (4.56 mg.) absorbed 1.40 C.C. hydrogen at 20"/760 mm. in 1 hour equivalent to 3.4 mols. hydrogen per mol. In this experiment a large excess of catalyst was employed. In a second experiment using a smaller quantity of catalyst #+antonin (327 mg.) absorbed 88.5 C.C. hydrogen at 16"/770 Wm. equivalent to 3 mols. hydrogen per mol. of #-santonin. The absorption of the first 28.5 C.C. (1 mol.) hydrogen ook 15 minutes the second mol.took 3+ hours and the third mol. took 36 hours. Dihydro-#-suntonin. A mixture of #-santonin (1 g.) palladised charcoal (0.1 g.) and glacial acetic acid (10 c.c.) was shaken for 20 hours in hydrogen at a pressure slightly greater than one atmosphere. After removal of catalyst and solvent the residue was crystallised from dilute alcohol as colourless needles (0.9 g.) m. p. 188-189" [alto' -239" (c 0.96 in glacial acetic acid) completely soluble with effervescence in dilute sodium carbonate (Found C 67.4; 68.1; H 7.8 8.0 ; M (Rast) 282. C15H2,0 requires C 67.7 ; H 8.3% ; M 266). The same dihydro compound was obtained by hydrogenation in alcohol at 100 lbs./sq. in. using palladised charcoal. The quantitative experiments described above using palladised charcoal gave the same compound.The dihydro compound was also obtained by refluxing a mixture of 4-santonin (0.3 g,),sodium hydroxide (8% 4 c.c.) ammonia (d0-88,l c.c.),water (2 c.c.) aad activated copper (2 g.) for 12 hours cooling and acidifying with dilute acetic acid when after removal of some brownish flocculent material the dihydro compound was deposited as well-formed prisms m. p. 182-184" raised to 189" by crystallisation from dilute alcohol (Found C 67.8; H 8.8%). The dry sodium salt of the above acid (2 g.) and methyl iodide (20 c.c.) were refluxed for 24 hours. Excess methyl iodide was removed and the residue extracted with ether and distilled. The fraction b. p. 175-180"/1 mm. was again distilled and collected (1 g.) at 181"/1 mm. On standing the methyl ester solidified and crystallised from ligroin in long transparent needles m.p. 77" (Found 68.6; H 8-1. C,,H,,O requires C 68-6; H 8.6%). The oxirne of dihydro-#-suntonin crystallised from alcohol in short stout prisms m. p. 229-230" (Found C 64-6; H 7.8; N 5.0. C15H,,04N requires C 64.1; H 8.2; N 5.0%). The same compound resulted from the catalytic reduction of 4-santoninoxime and it reverted to dihydro-#-santonin on treatment with hydrochloric acid. A nhydro-dihydro-#-suntonin. A mixture of dihydro-#-santonin (0.2 g.) acetic anhydride (3 c.c.) and anhydrous sodium acetate (0.5 g.) was heated for 2 hours excess acetic anhydride removed the residue extracted with water and crystallised from dilute alcohol forming colourless needles (0.1 g.) m. p. 158" (Found C 72.65; H 7.8.C,,H,,,O, requires C 72-6; H 8.1yo). It was insoluble in sodium carbonate but slowly dissolved in sodium hydroxide from which dihydro-#-santonin was recovered on acidification. The lactone was also obtained by heating dihydro-#-santonin for a few minutes at 190-200". Oximes of anhydro-dihydro-#-suntonin.A mixture of the above lactone (0.2 g.) hydroxylamine hydrochloride (0.2 g. ; 4 mols.) anhydrous sodium acetate (0.5 g.) and alcohol (3 c.c.) was shaken for 12 hours. The alcohol was removed under reduced pressure and the residue was extracted with water and crystallised from aqueous alcohol as colourless prisms m. p. 209-210" (decomp.). It gave an insoluble copper salt with ammoniacal copper sulphate and was insoluble in sodium bicarbonate and generally gave reactions of a hydroxamic acid (Found C 60.7; H 7.9; N 9.3.C15H2404hT2 requires C 60.8; H 8.1 ; N 9.5%). Using only one molecule of hydroxylamine hydrochloride the true oxime was obtained and this crystallised from benzene-ligroin as colourless needles m. p. 188" (Found C, 68.3; H 7.9. C1,H2,0,X requires C 684; H 8.0%). Hexahydro-~-santonin. A mixture of 6-santonin (0.5 g.) glacial acetic acid (15 c.c.) and Adams' catalyst (0.1 g.) was shaken in hydrogen at 15 atm. for 24 hours and the product was crystallised from dilute acetic acid as long needles m. p. 191-192" (Found C 66.5; H 9.8. C15H2,04requires C 66.6; H 9.6%). It readily dissolved in sodium carbonate with effervescence and was regenerated unchanged on acidification. It did not yield an oxime.A cetyl-anhydro-hexahydro-#-santoninwas obtained by refluxing the hexahydro compound with an excess of acetic anhydride. It crystallised from alcohol as colourless needles m. p. 125" (Found C 69.9; H 8-75. C H2,0 requires C 69.4; H 8.8%). It was insoluble in sodium hydroxide and was hydrolysed to hexahydro-#-santox& with 13% methanolic potash. Tetrahydro-#-suntonin. To dihydro-4-santonin (0.3 g.) in alcohol (4 c.c.) alcohol-washed 4% sodium amalgam (1 1 g.) was added in portions. The mixture was refluxed for 5 hours and left overnight. Alcohol was removed and the residue dissolved in water and acidified. The precipitate was crystallised twice from water and obtained as colourless prisms m. p. 189-190" (Found C 67.3; H 8.8. C,&f,,o.4 requires C 67.2; H 9.0%).All attempts to obtain an oximino derivative failed. Acetyl-anhydro-tetrahydro-4-santonzn crystallised from dilute alcohol in colourless prisms m. p. 104" (Found C 70.4; H 8.3. C1&404 requires C 69.9; H 8.2%). It yielded the acetyl anhydro-hexahydro compound described above when submitted to hydrogenation in presence of Adams' catalyst. Tribrorno-~~-santonin.Bromine (1 g. 3 mols.) was added dropwise to a stirred solution of #-santonin (0.5 g.) in chloroform (10c.c.) at room temperature. Stirring was continued for half an hour and the chloroform and excess bromine were removed under .reduced pressure. The residue was crystallised from alcohol as long colourless needles (0-6 g.) m. p. 243" (Found C 36.0; H 3.6; Br 48.3. C,,H,,04Br requires C 35.8; H 3.8; Br 47.7%).Monobrorno-#-suntonin obtained in a similar manner using 1.2 mols. of bromine crystallised from ligroin (b. p. 80-1CO") as colour- less needles or prisms m. p. 198-199" (Found C 52-3; H 5.5. C,,H,,O,Br requires C 52.4; HI 5.5%). The same monobromo compound was obtained when a solution of the tribromo compound in glacial acetic acid was submitted at 35" to hydrogenation at one atmosphere using palladised charcoal. When these bromo compounds were refluxed with zinc and alcohol for 3 to 4 hours a product m. p. 184" was obtained in each case which gave no depression of m. p. when mixed with #-santonin. Monobrorno-$-santonin oxime crystallised from dilute alcohol as colourless plates m. p. 251" (decomp.) (Found C ,50.4; H 5-4. C15H,0,NBr requires C 50.3; H 5.6%).Attempts to obtain the oxime of the tribromo compound were unsuccessful. #-Santonin (0.1058 g.) titrated by Winkler's method (Zoc. cit.),required 15.84 C.C. 0.1004~-sodium bromate equivalent to 0.127 g. bromine (four atoms' of bromine = 0.127 g.). Similarly 0.0684 g. dihydro-$+antonin required 10.06 C.C. 0.1004~-sodium bromate equivalent to 0.081 g. bromine (four atoms of bromine = 0.082 g.). Action of Mineral Acids on 4-Santonin.-(a) Sulphuric acid (55%). Finely powdered #+antonin (0.25 g.) was introduced with stirring into a mixture of conc. sulphuric acid (2.5 c.c.) and water (3.7 c.c.) and the mixture was warmed [1946] Clemo and Cocker The Constitution of +Santonin. Part I. to 48-55" for 24 hours. After a short time the #-santonin passed into solution with the production of a bright red colour and shortly afterwards a crystalline solid was deposited.This was crystallised twice from 80% alcohol when desmotropo-$-suntonin was deposited as colourless needles (0.06 g.) m. p. 185-186" [CL];~~+ 67.9" (c 2-461 in chloroform) (Found 73.2 73.3; H 7.4 7.3. C1,H+,03 requires C 73.2; H 7.3%). It was soluble in sodium hydroxide insoluble in sodium carbonate gave a red coloration with neutral ferric chloride and a deep pink nitroso compound. It gave a dull red coloration with p-nitrobenzenediazonium chloride. The benzoate prepared in alkaline solution crystallised from dilute alcohol as long needles m. p. 164" (Found C 74.8; H 6.2. C,,H,,O requires C 75.4; H 6.3%). The acetate crystallised from dilute alcohol as colourless plates m.p. 233" (Found C 70.5 ; H 6.8. C,,H,O requires C 70.8; H 6.9%). This compound was not readily hydrolysed but hydrolysis was effected by the use of concentrated methanolic potash when desmotropo-4-santonin was regenerated. This resistance to hydrolysis is characteristic of phenyl esters. The methyl ether prepared by using methyl sulphate in alkaline solution crystallised from dilute alcohol in tufts of silvery needles m. p. 159-160" (Found C 73.3; H 7.4. C,,H,,O requires C 73.8; H 7.7%). Thionyl chloride had no action on desmotropo-$-santonin. Attempts to reduce the desmotropo compound by palladised charcoal were unsuccessful. From the mother liquors obtained during the crystallisation of the desmotropo compound an amorphous solid was obtained from which a crystalline acetate (needles m.p. 239") was obtained in small quantity. This compound has not yet been identified (Found C 71.4; H 7.1. C1,H,o04 requires C 70.8; H 6.9%). (b) Formic acid. The desmotropo compound was also obtained by refluxing 4-santonin (0.5 g.) with formic acid (4 C.C. ; 98%) for 2 hours. Excess formic acid was removed under reduced pressure and the residue was crystallised from alcohol as colourless plates (0.2 g.) m. p. and mixed m. p. 186". (c) Concentrated hydrochloric acid. Powdered $-santonin (1g.) was stirred into conc. hydrochloric acid (5 g.). The solution quickly became yellow and in a short time a crystalline solid was deposited. After 1 hour this was collected washed with a little concentrated hydrochloric acid and then with water leaving a colourless crystalline solid (0.7 g.) m.p. 172-173". It was crystallised twice from dilute acetic acid as fine colourless needles m. p. 175-176" [alp + 25.2' (c 1.8 in chloroform) readily soluble in sodium carbonate with effervescence and recovered on acidification (Found C 63.9 H 7.5 ; 111,286. C,,H,,O requires C 63.8; H 7.8% ; &I 282). Oximation in alcohol using sodium acetate gave a nitrogenous compound which was readily soluble in water and contained sodium. By crystallisation from the minimum quantity of water it was obtained in clusters of needles which darkened at 200" and did not melt below 400" (Found N 4.3. C,,H,,O,NNa requires N 4.4%). Attempts to acetylate the compound m. p. 175-176" gave inconclusive results but on treatment with 55 yosulphuric acid it gave desmotropo-#-santonin.Dihydro-#+antonin and its anhydro compound were both unaltered by conc. hydrochloric acid and 55% sulphuric acid. Attempted dehydrogenation of the desmotropo compound by selenium failed since the compound sublimed too readily. Dehydrogenation was effected by heating gradually and with stirring a solution of the desmotropo compound (0.5 g.0) in potassium hydroxide (0.5 g.) and water (2 c.c.). Fusion commenced at 220" and gas evolution took place at 240 . The temperature was then raised gradually to 310-320" and kept there for about 10 minutes until gas evolution had ceased. The cooled melt was extracted with water and the violet solution shaken with a little charcoal and filtered.The filtrate was acidified and on steam-distillation a colourless crystalline solid (0.15 g.) m. p. 76-78' was obtained. It was crystallised thrice from ligroin as colourless needles m. p. 81-82'. It was found impossible to raise the m. p. above this value but the product did not depress the m. p. of authentic 2 :4-dimethyl-l-naphthol (Cornforth Cornforth and Robinson loc. cit. give m. p. 84-85') (Found C 83.4; H 7.1. Calc. for C,,H,,O C 83.7; H 7.0%). Its carbaniiate crystallised from ligroin (SO/lOO") as clusters of colourless needles m. p. 174-175'. The m. p. when mixed with the carbanilate prepared from 2 :4-dimethyl-l-naphthol obtained according to Cornforth Cornforth and Robinson showed no depression (Found C 78.0; H 6.3. C,,H,,O,N requires C 78.4; H 5.8%).Its picrate crystallised from aqueous alcohol as deep red needles giving m. p. and mixed m. p. with authentic matexial 143-144" (Found C 54.2; H 3-8. Calc. for Cl,Hla0,C,H,0,N3 C 53.9; H 3.7%). Clewmensen reduction of dzhydro-#-santonzn. The dihydro compound (0.4 g.) amalgamated zinc (4 g.) and concen- trated hydrochloric acid (7 c.c.) were refluxed for 18 hours ; then a further quantity of the acid (3 c.c.) was added and the boiling continued for 8 hours. after cooling the odoriferous oil was extracted with ether dried over sodium sulphate and the ether removed. As the resulting colourless gum (0.07 g. b. p. 140°/1 mm.) showed no signs of crystallisation after standing for a fortnight it was subjected to selenium dehydrogenation. The product from several such experiments (1-55g.) was heated with selenium (4 g.) at 280-340" for 48 hours and finally at 340-350" for a further 12 hours.The residue was extracted with ligroin (b. p. 60-80') and the solvent removed. The residue was distilled from sodium giving an oil (0-25 g. b. p. 80"/1 mm.). This residue'(0-15 g.) in alcohol was treated with picric acid (0.18 g.) in alcohol and the orange prisms collected. On crystallisation from alcohol the picrate was obtained as long orange needles m. p. 96". The m. p. was not depressed on admixture with the picrate of l-methyl-7-ethylnaphthalene(Harvey Heilbron and Wilkinson J. 1930 423). Ozonolysis experiments. A number of experiments were performed in which $-santonin (1 g.) dissolved in purified chloroform (20 c.c.) was submitted to the action of 6% ozonised oxygen for periods of 6 to 9 hours at 0-5".The solvent was removed and the syrupy residue was steam distilled into an aqueous alcoholic solution of dimedon," from which I' the formaldehyde condensation product (0.06 g.) m. p. 190" (cp. Vorlander 2.anal. Chew. 1929 74 241) was obtained The residue remaining after steam distillation was an intractable syrup. Efforts to convert desmotropo-$-santonin to a $-santonous acid were unsuccessful. Sodium amalgam hydriodic acid and red phosphorus and zinc and acetic acid were employed but the compound was either recovered unchanged or it gave an intractable syrup. All attempts to dehydrate #-santonin were fruitless. It distilled unchanged from anhydrous sodium sulphate copper sulphate and calcium oxide.Some Experiments on the Opening of Lactone Rings with Hydroxylamine.-(i) Action of hydroxylamine on p-toluene- sulPhony1-6 6-dimethyL2-nzorpholone. This compound (0.5 g.) in alcohol (5 c.c.) was refluxed for 1 hour with hydroxyl- amine hydrochloride (0.5 g.) and anhydrous sodium acetate (1.8 g.). Alcohol was distilled away and the residue wil~ washed with water and crystallised from alcohol. Two crops of crystals were obtained. The first had m. p. 132' and was identical with the starting material the second was crystallised twice from dilute alcohol as long colourless prisms m. p. 105" (decornp.). It was readily soluble in sodium hydroxide and gave a blood-red coloration with neutral ferric chloride and a sparingly soluble copper salt with ammoniacal copper sulphate.Its properties generally conformed with those expected of N-p-toluenesulphonyl-N-~-hydroxyisobutylacethydroxa~icacid (Found C 48.9 ; H 6.4. C,,H,,O,N,S requires C 49.4; H 6.3%). (ii) ~ctionof hydroxylamine on coumarin. Coumarin (0.5g.) hydroxylamine hydrochloride (0.4 g.) sodium acetate (0.5 g.) and alcohol (5 c.c.) were refluxed for 2 hours. The product was isolated as in (i) and crystallised twice from alcohol when ~-2-hydroxyphenylacryZhydro%amicacid was obtained as colourless prisms (0.1 g.) m. p. 252O (decamp.) (Found C 60-4; H 5.2. C,H,O,N requires C 60.3; H 5.0%). Cocker The Constitution of +Santonin. Part II. (iii) Action of hydroxylamine on the lactone of cyclohexanoZ-2-/3-propionic acid (Cook and Lawrence J.1937 823). From this lactone (2g.) a very small quantity of a very hygroscopic nitrogenous material was obtained which on crystal- lisation from alcohol and ether was obtained as colourless plates m. p. 115-116". There was insufficient material for analysis. (iv) Lactone of cyclohexanol-2-U-propionicacid. Ethyl cyclohexanone-2-carboxylate(40g.) ethyl a-bromopropionate (42g.) and sodium ethoxide (from 5-4g. sodium and 120 C.C. alcohol) were refluxed overnight. The alcohol was removed and the residue extracted with ether and fractionated giving ethyl cyclohexanone-2-carboxylate-2-a-~ro~ionate (25 g.) as a colourless oil b. p. 158-163"/6-8 mm. (Found C,62.6; H 8.0. ClIHzz05requires C,62.2; H 8.1%). Attempts to hydrolyse this compound with hydrochloric acid (compare Robinson and Openshaw J.1937 944) gave poor results but hydrolysis with baryta gave somewhat higher yields than with aqueous methanolic potassium hydroxide (compare Haworth and Mavin J. 1933 1015). The ester (25 g.) was added to barium hydroxide (50 g. of monohydrate) in a mixture of methanol (250c.c.) and water (250 c.c.) and the whole heated with shaking for 2 hours. Methanol was dis- tilled away and the residue acidified and extracted with ether ; the extract was dried with sodium sulphate and frac- tionated. Two fractions were obtained (a) 2 g. b. p. 94-96"/0.25 mm. (b)6g. b. p. 153-156 (0-25mm.). Fraction (b) solidified on cooling and was crystallised from ligroin as rosettes of needles m. p. 135" (Found C 63.61 H 8.4. C9Hl,0 requires C 63-5; H 8.2%).This was the required cyclohexanone-2-a-propionicacid. Its 2 :4-dinitrophenyl-hydrazone crystallised from alcohol as orange needles m. p. 244" (Found C 51.7; H 5.2. C1&&,N4 requires C, 51.4; H 5.1%) and its semicarbazone from water as prisms m..p. 185-186" (decomp.) (Found C,51.1; H 7.5. (C,,H,,O,N,),H,O requires C,50.8; H 7.6%). Reduction of the above acid. The acid (1.7g.) in alcohol (50 c.c.) was refluxed with sodium (3g.) until all the sodium had dissolved. Alcohol was removed water was added and the mixture was acidified and extracted with ether. The extract was fractionated and a colourless liquid (1 g.) b. p. 106-108/0.5 mm. was collected. This was the required Eactone (Found C 69.7; H 9.0. C,H140 requires C 70.1; H 9.1%). All attempts to obtain a nitrogen containing compound from this lactone by treatment with hydroxylamine faded as did all attempts at hydrogenolysis using palladised charcoal.Fraction (a),b. p. 94-96/0.25 mm. dissolved slowly in sodium hydroxide readily decolorised bromine water and gave a 2 4-dinitrophenyZhydrazone which crystallised from alcohol as golden-yellow plates m. p. 154" insoluble in sodium carbonate (Found C 51.3 51.8; H 5.5 4.9. C,,H&,N4 requires C,51.4; H 5.1%). This compound was not investigated further. The authors desire to thank Messrs. T. and H. Smith Ltd. for gifts of +santonin. COLLEGE,UNIVERSITY DURHAM, KING'S OF NEWCASTLE-UPON-TYNE. [Received July 31st 1945.;