SYNTHESIS OF TEIE TERPENES. PART XV. 2147By BERNARD DUNSTAN WILKINSON LUFF (1851 Exhibition Scholar)and WILLIAM HENRY PERKIN, jun.IN a previous communication (Trans., 1905, 87, loss), it wasshown that 1 -met,hylcy clohexan-3-c arboxylic acid (hexahydro-m- toluicacid) is readily converted by bromination and subsequen2148 LUFF AND PERKIN: EXPERIMENTS ON THEelimination of hydrogen bromide into a mixture of 1-methyl-A3- and-A2-cy clohexene-3-carboxylic acids :-+ C H M e < ~ ~ ~ ~ ~ ~ ~ ~ > C H and CHMe<CH, CH:C(CO,H)\ CH,,4H,.A3- A2-The separation of acids of such similar constitution proved to bea matter of much difficulty, and, as a result, only small quantitiesof the pure isomerides were available for subsequent investigation.l-Methyl-A3-cycZohexene-3-carboxylic acid (ni.p. 58-60°) wasconverted into its ester, and this, by the action of magnesium methyliodide, into A3-m-menthenol(8), a substance which, when left iiicontact with a large excess of magnesium methyl iodide, loses waterwith formation of A3 :8(Q)-m-menthadiene :I n a similar manner, A2-mmenthenol(8) and Az:8(g)-m-menthadienewere prepared from ethyl l-methyl-A2-cycZohexene-3-carboxylate :CH:C(CMe,*OH)C'H M e < ~ ~ ~ c ( " o ~ ~ ~ > C H 2 -+ CHMe< CH,--- C*,>CH2It will be observed that A3 :8(9)- and AZ:8(9)-m-menthadiene containconjugated double linkings, and investigation showed that theseterpenes do, in fact, show the characteristic behaviour of substanceswhich contain such groupings; they are, for example, only able tocombine with two atoms of bromine or with one molecule of hydrogenchloride (compare Trans., 1905, 87, 641, 667, and 1068).A short time since (this vol., p.1428), Perkin and Wallach madea special study of another synthetical terpene which contains con-jugated double linkings, namely, A3 :8(9)-pmethadiene :and observed that this terpene had a higher boiling point, density,and refractive index than terpenes, such it5 limonene, which do notcontain conjugated double linkings (compare p. 2154), and it seemeddesirable to investigate the physical constants of other terpenes ofsimilar structure in order to determine whether this behaviour isa general one. The difficulty in preparing the m-menthadienesalready mentioned in amount sufficient for accurate physicalmeasurement is due to the difficulty in obtaining the l-methylSYNTHESIS OF THE TERPENES.PART SV. 2149cy clohexene-3-carboxylic acids in any considerable quantity by theprocess described at the beginning of this paper. An attempt wastherefore made to devise a new method for the preparation of theseacids, and this has been successful in the case of the A3-acid (m. p.Quite recently (this vol., p. 1760), it was pointed out that 1-methyl-cyclohexan-4-one reacts readily with sodamide and carbon dioxide,with the formation of l-methylcycZohexan-4-one-3-carboxylic acid :58-60').and w0 have found that, under the conditions described in thepresent communication (p. 2150), the yield directly obtained is 50per cent.of that theoretically possible.* When this acid isreduced with sodium amalgam, it yields 1-methylcycZohexan4-01-3-carboxylic acid, and this, on distillation, is almost quantitativelyconverted into l-methyl-A~-cycZohexene-3-carboxylic acid (m. p.58-60') :This method of preparation is still laborious, but it is much lessso than the original method, and it has now been found possiblet o prepare nearly 200 grams of the pure acid, of which parthas been used in the present research, and the remainder isbeing resolved into its active constituents with the object ofpreparing and carefully investigating the corresponding activementhenols and menthadienes. I n possession of sufficientmaterial, we have now very carefully determined the physicalproperties of h3-m-menthenol(8) and A3:8(9)-m-menthadiene (seepp.2153, 2154).We find that the values of A3-m-menthenol(8) show a remarkablesimilarity to those of A3-pinenthenol(S), and, when these numbersare compared with those of terpineol, it is seen that the proximity ofthe double linking to the *CMe,-OH group causes a fall of boilingpoint, density, and refractive index. Similarly, the values forA3 :8(9)-m-menthadiene are very like those of A3 :8(g)-p-menthadiene,but, when these values are compared with those of limonene, it isseen that the presence of conjugated linkings causes a considerablerise in boiling point, density, and refractive index.A similar study of A2-m-menthenol(8) and of A2 :*(g)-m-menthadieneis in progress, and the results will shortly be ready for publication.* It is actually considerably more, because a good deal of ketone is recovered un-changed and may be employed in a subsequent preparation2150 LUFF AND PERKIN: EXPERIMENTS ON THEEXPERIMENTAL.Preparation of 1-Methylcyclohexan-4-one-3-carbozyZic Acid and1 -Met hyZcyclohexalz-4-ol-3-car~ oxylic A cid.I n order to obtain the large quantities of the above ketonic acidwhich we required for this research, we decided, as the result of along series of comparative experiments, to modify the method ofpreparation originally employed (this vol., p.1766) in the followingway.1-MethylcycZohexan-4-one (100 grams) is dissolved in dry ether(1 litre) in a three-necked flask fitted with a reflux condenser andmechanical stirrer, powdered sodamide (40 grams) is then addedin lots of 10 grams, and the stirrer set in motion, when a rapidevolution of ammonia takes place.After the reaction has subsided,the mixture is heated on the steam-bath for an hour, allowed tocool, and then a stream of dry carbon dioxide passed through a widetube so as to avoid stoppage, the whole being vigorously stirredduring the operation. This causes a rise of temperature, and thepassage of the gas is continued until the whole has cooled downto that of the atmosphere. The contents of the flask are thenwashed into a separating funnel with ice water, well agitated, theaqueous layer run off, and the ethereal solution kept in order torecover the unchanged ketone which it contains. The aqueoussolution is acidified, extracted with ether, the ethereal solutionshaken with sodium carbonate, and, after separating, the alkalineextract is cautiously treated with hydrochloric acid until the oilyimpurity has been completely precipitated and the crystalline acidcommences to separate.After filtering, the filtrate is acidifiedwith hydrochloric acid, and the colourless, crystalline precipitate ofnearly pure l-methylcycZohexan-4-one3-carboxylic acid collected,washed, and drained on porous porcelain. The yield is 45-50grams. The ethereal layer, containing the unchanged ketone, iswashed, evaporated, and the residue distilled in a current of steamin order to separate it from a quantity of the bicyclic condensationproduct described on p.2155. The ketone is extracted from thedistillate, and, after one fractionation, is sufficiently pure for afurther operation.I n preparing l-methylcycZohexan-4-ol-3-carboxylic acid, 1-methyl-c~cZohexan4-one3-carboxylic acid, in quantities of 10 grams, isdissolved in sodium carbonate, the solution made up to 600 C.C. withwater, and treated, in a bottle fitted with a mechanical stirrer,with freshly prepared sodium amalgam (500 grams), which is addedin three lots. Reduction takes place very slowly at the ordinarytemperature, and the most suitable temperature appears to bSYNTHESlS OF THE TERPENES. PART XV. 215150-60°, which is maintained during the whole operation by placingthe bottle in a trough of hot water ; it is also necessary to neutralisethe alkali produced by hydrochloric acid, which is dropped in insuch a way that the liquid is always faintly alkaline.The product is acidified with hydrochloric acid, saturated withsalt, and extracted at least five times with much ether; the etherealextract is dried over anhydrous sodium sulphate, and evaporated,when a syrup remains which soon becomes semi-solid, and thepurification and properties of which have already been described(this vol., p.1770).l-Methyl-A3-cyclohexene-3-carboxy~~c Acid.This acid has been prepared in large quantities in the followingmanner. Crude l-methylcycZohexan-4-ol-3-carboxylic acid, obtainedas described in the last section, is transferred to a distilling flaskwith a rather long neck, and heated in a metal-bath, wheneffervescence soon occurs, due to the decomposition of some unchangedketonic acid.In a short time, water commences to be eliminated,and as soon as this and the ketone have ceased to pass over, theresidue is distilled under diminished pressure, when almost thewhole passes over at 155-160°/25 mm., and solidifies on cooling.The mass is left in contact with porous porcelain until free fromtraces of oily impurity; it then melts at about 57-58O, and isalmost pure 1-methyl-A3-cycZohexene3-carboxylic acid. The yield isapproximately 10 grams from 50 grams of the crude hydroxy-acid.For analysis, the substance was crystallised from formic acid, fromwhich it separates in pearly plates, melting at 60°. (Found, C = 68.4 ;H=8.6.Careful comparison has shown that this acid is identicalwith the " Al-tetrahydro-m-toluic acid " obtained by Perkin andTattersall (Trans., 1905, 87, 1092) by quite a different process.Ethyl 1-methyl-A3-cycZohexene-3-carboxylate, which has already beendescribed (Zoc.cit., p. 1094), wm prepared in quantity by leavingthe pure acid (50 grams) in contact with alcohol (300 c.c.) andsulphuric acid (30 c.c.) at the ordinary temperature for two days,and then heating on the water-bath for two hours.After extracting in the usual way, it distilled at 146--148O/100 mm.It was noticed that the aS-unsaturated acid is esterified withsome difficulty at the ordinary temperature. Thus, for example,the mixture just mentioned, after being kept for twenty-four hours,contained a large amount of acid, and, even after forty-eight hours,a considerable quantity of unesterified acid was still present.Thisbehaviour has been utilised for separating the acid from 'theCalc., C=68*6; H=8*6 per cent.21 52 LUFF AND PERKIN: EXPERIMENTS ON THEisomeric By-acid by a process of fractional esterification (comparethis vol., p. 2146).ethulcvclohexarte-3-curb oxulic A cid.I n order to prepare this characteristic derivative, pure I-methyl-A3-cyclohexene-3-carboxylic acid, dissolved in a little dry chloroform,is cooled to -5O, and mixed with a few drops of dry bromine. Atthis low temperature, no action seems to take place, but, if allowedto rise to 15O, the colour of the bromine suddenly disappears, andthen the addition of the theoretical amount takes place rapidly at Oo.During the operation a good deal of the dibromo-acid crystallisesout, and the remainder is obtained by allowing the chloroform toevaporate.It is sparingly soluble in cold formic acid, and not veryreadily so on boiling, and separates in colourless, glistening leaflets :C8H,,0,Br2 requires Br = 53.3 per cent.0.3166 gave 0-3945 AgBr. Br=53-0.When rapidly heated, 3 : 4-dibromo-1-methylcyclohexane-3-carb-oxylic acid melts at 165O.It dissolves readily in sodium carbonate, and the solution, onboiling, does not cloud with separation of the bromohydrocarbon,aa sometimes happens in the case of afi-dibromo-acids of similarconstitution. After boiling for ten minutes with a large excess ofsodium carbonate, the solution deposited, on acidifying, a solidacid, which crystallised well from formic acid, melted at 148-150°,and appears to be a bromohydroxy-acid of the formulaC,H,,Br(OH)*C02H (Found, Br = 32.7.Calc., Br = 33*2), but itwas not further investigated.The aqueous filtrate from this acid contains a considerablequantity of a syrupy acid which may be extracted with ether, andis probably the corresponding dihydroxy-acid.4-Bromo-1 -m e t hylc yclohexane-3-curb ox ylic A cid,1-Methyl-A3-cyclohexene-3-carboxylic acid dissolves readily infuming hydrobromic acid (saturated at Oo), and, on keeping, butmore rapidly if slowly warmed to 70°, a syrup separates, whichgradually crystallises. After washing with water and draining onporous porcelain, the substance was dissolved in a little warm formicacid, in which it is very readily soluble, and from which it separatesin needles, melting at 107-109°SYNTHESIS OF THE TERPENES. PART XV, 21530.1544 gave 0.1314 AgBr.Br=36-2.C8H,,0,Br requires Br = 36.2 per cent.This bromo-acid is decomposed by boiling with sodium carbonate,and the solution, on acidifying, deposits a crystalline precipitateof 1-methyl-A3-cycZohexene-3-carboxylic acid.A3-m-Nenthenol(8) and A3 :8(9)-m-Menthadiene.The conversion of ethyl 1-methyl-A~-cycZohexen-3-carboxylate intoA3-mmenthenol(8) was carried out by adding the ester (13 grams)to an ethereal solution of magnesium methyl iodide, containing6.5 grams of magnesium. After fifty hours, the product was decom-posed by water (compare footnote, p. 2154), distilled in a current ofsteam, the distillate extracted with ether, and the ethereal solutiondried, evaporated, and the residue distilled under diminishedpressure.The whole quantity passed over a t 114--117O/35 mm.,and, after redistillation at 115O/35 mm., as a colourless, pleasantsmelling, rather viscid liquid, and the yield of this pure A3-m-men-thenol(8) was almost quantitative. The first time this substancewas prepared (Trans., 1905, 87, llOO), it was noticed that much ofthe ethyl ester remained unattacked; this was due to the fact thatthe amount of magnesium (2.9 grams to 10 grams of ester) wastoo small, and the length of contact (twenty-four hours) was also notsufficient. Under the conditions mentioned above, only a trace ofethyl ester remained unchanged, and this was removed, beforefractionation, by hydrolysis with a little methyl-alcoholic potassiumhydroxide in the usual manner.The phenylurethane of A3-m-menthenol(8) was prepared by leavingthe menthenol with an equal volume of phenylcarbimide for threedays.The mass was drained on porous porcelain, and thencrystallised from 80 per cent. methyl alcohol, from which itseparated in colourless needles, melting at 130° :0.1057 gave 5.0 C.C. N, at 21° and 758 mm.C&H,,O,,N requires N = 5.1 per cent.The physical properties of A3-m-menthenol(8) have been carefullydetermined, and found to be very similar to those of A3-p-men-thenol(8) (compare Perkin and Wallach, this vol., p. 1435), as thefollowing comparison shows :N=5.3.B p.............................. 102"/14 nim. 11 5"/35 iiim.d. 20"/20" ........................ 0.9268 0'9'210M ................................. 47.10 (calc. 47-16} 47'23M. p. of the phenylurethane 13U' 128"VOL. XCVII. 7 B?tu ................................. 1'4ii)S 1'4762154 LUFF AND PERKIN : EXPERIMENTS ON THEA3-m-Menthenol(8) has, so far, not crystallised, whereasA3-p-menthenol(8) is solid, and melts at 39O.A3:*(9)-m-Menthadiene has been prepared in two different ways,namely, (1) by the direct action of magnesium methyl iodide onethyl l-methyl-A3-cycZohexene-3-carboxylate, and (2) by the action ofaqueous oxalic acid on A3-mmenthenol(8).1. Ethyl l-methyl-A3-cycZohexene-3-carboxylate (25 grams) wasadded to an ethereal solution of magnesium methyl iodide (10 gramsof magnesium), and, after forty-eight hours, the product was decom-posed by dilute hydyochloric acid* and distilled in it current ofsteam.The ethereal extract of the distillate was dried, evaporated,and the product fractionated, when almost the whole passed over at178-183O/ 750 mm., the amount of As-m-menthenol remaining beingvery small.The A3:8(9)-nzrmenthadiene was then distilled three times oversodium, when it boiled constantly at 181-182°/760 mm.2. A3-m-Menthenol(8) was boiled for three hours with 6 percent. aqueous oxalic acid, and the product distilled in a current ofsteam. After extracting in the usual manner, it was found thatthe conversion into A3 : *@)-m-menthadiene had been almostquantitative, and that this terpene again distilled at 181-18Z0/760 mm.The physical properties of both the specimens of thisterpene were carefully determined, and found to be practicallyidentical, and it is interesting to tabulate these (I) with those ofthe two following terpenes :CH2*C(CMe: CH2)>cH,CH2 c=, I. A3 : s(g)-m-Menthadiene, CHMe<11. A3 's(g)-p-Mcnthadiene, C H 2 - C H > ~ e ~ ~ e : ~ ~ , .111. Limonene, CMeqCH2. CH-CH2>CH*CMe:CH2. CH,I. 11. ur.B. p. ............... 1 8 1-1 82" 184-185" 175-1 76"d. 20"/20" ......... 0-8609 0'8580 0'8460no .................. 1 *4975 1 '4924 1.4746M. .................. 46'3 46'02 45'28(C1,,H1&*= 45'24.)* The curious observation (conipare p. 2163) has been made in this and in othercases, that the product of the action of maplnzsium methyl iodide on the ester of amethylcyclohexenecarboxylic acid, containing the double linking in the afi-position,yields the menthenol when it is treated with water, but the conjugated menthadienewhen i t is decomposed by dilute hydyochloric acid.In cases where the doubleliuking is in any other position, this difference has not been observed, and theproduct has always been the menthenol whether the decomposition has beon carriedout simply by water or by the addition of liyclrochloric acidSYNTHESIS OF THE TERPENES. PART XV. 2155Cases 1 and I1 prove again that the effect of conjugation is toraise the boiling point,, density, and refractive index above thoseof limonene.I n possession of considerable quantities of pureA3: *(g)-m-menthadiene, the authors have again investigated itsbehaviour with bromine and with hydrogen chloride (compareTrans., 1905, 87, 1101). The freshly distilled terpene (3.3588grams) was dissolved in twice its volume of chloroform, cooled to- 15*, and titrated with a solution of bromine in chloroform (1 in 3),when 12.6 C.C. were decolorised, the end-point being quite sharp,but a little hydrogen bromide was produced. This amount ofterpene had therefore decolorised 4.2 grams of bromine, whereasthe amount required for the formation of the tetrabromide,C10H16Br4, is 8.0, and for the dibromide, C10H16Br2, 4.0 grams. Itis clear therefore, as had previously been pointed out, that this andother conjugated terpenes are only capable of combining with twoatoms of bromine.I n investigating the action of hydrogen chloride, about 5 C.C.ofthe terpene were cooled to -15O, and a, current of dry hydrogenchloride passed for one hour; the almost colourless product was leftfor two days, then placed over potassium hydroxide in a, vacuumdesiccator for several hours, and analysed :0.1773 gave 0.1390 AgCl. C1= 19-4.CloH,,,HC1 requires C1= 20.6 per cent.It is therefore obvious that this conjugated menthadiene is onlycapable of combining with one molecule of hydrogen chloride.1 : 4~-Dim~ethyl-3-cyclohexy~~denecyclohexan-4-one,Considerable quantities of this ketone, as well as higher con-densation products of l-methyIcyc7ohexan-4-one, are produced duringt.he action of sodamide and carbon dioxide (p.2150). After extract-ing the sodium salt of l-met~hylcycZohexan-4-one-3-carboxylic acidwith water, the ethereal solution is evaporated, and the residuedistilled in a current of steam, when unchanged methylcyclo-hexanone passes over, and the condensation products remain in thedistilling flask. The dark brown oil is extracted with ether, theethereal solution dried and evaporated, and the residue fractionatedunder diminished pressure, when a quantity of a pale yellow oilis obtained, which distils at 173--174O/25 mm., and has a pronouncedfruity odour :0-1971 gave 0.5862 CO, and 0.1924 H,O. C =81.1; H = 10.8.Cl,H,O requires C = 81.5 ; H = 10.7 per cent.d 20°/200=0*9728; n=1'4986; M=62.1 (calc., 62.4)'7 ~ 2156 FORSTER AND ZJMMERLI !This condensation product is very similar to the (‘ bicyclic ketone,’’C,,HzzO, which Wallach (Ber., 1896, 29, 1595) obtained by thoaction of hydrogen chloride on 1-methylcyclohexan-3-one.When a drop of sulphuric acid is added to the solution of theketone in acetic anhydride, an intense crimson coloration is produced,which persists for a, long time. The oxime was obtained by addingan aqueous solution of hydroxylamine hydrochloride to an alcoholicsolution of the ketone, when, almost immediately, a, crystallineprecipitate began to form, which, after crystallisation from alcohol,was obtained in glistening needles, melting at 160O:0.463 gave 26.4 C.C. N, at 18O and 758 mm. N=6*7.C14Hz,0N requires N = 6.3 per cent.THE UNIVERSITY,MANOHESTER