LAPWORTH AND HA" : MENTHYL PORMYLPHENYLACETATE. l.kc)I CXLVII.-Opticakly Active Esters of ,&Ketonic and &Aldehydic Acids. Part I. Menthy1 Formyl- phe rLy lacetate. By ARTHUR LAPWORTH and A. C. OSBORN HANN. IT is well known that the desmotropic forms of such substances as P-diketones and nitroparaffins differ markedly in many physical properties, such as density, refractive, absorptive, and dispersive power, and magnetic rotation, and that by means of these properties it is frequently possible to determine to which of two possible forms a compound may be referred under given conditions, and, in many cases, to follow the change of one form into another. In these cases, however, it is not easy to follow the changes rapidly in dilute solution. When the compounds themselves are optically active, however, observations on dilute solutions may be very rapidly made, and, as Lowry has shown (Trans., 1899, 75, 2-24>, may somebimes afford an accurate method of estimating the relative amounts of the two forms present at any time.In spite of the value which LOWrfS method of investigation by the observation of mutarotation may certainly possess, as shown by his own experiments on nitro- and 7r-bromonitro-camphor, and by those of Forster on nitrocamphane (Trans., 1900, 77, 260) and benzoyl- camphor (Trans., 1901, 79, 987), the principle has not yet been widely applied, the reason being, no doubt, that the number of active labile compounds of the kind which may be prepared i n a perfectly straightforward way is limited. I n order that the method may be employed to the greatest advantage, it is, of course, necessary that both desmotropic forms should be isolable.Nevertheless, even when only one of the two forms can be isolated in 5 H 21492 LAPWORTH AND HANN: OPTICALLY ACTIVE ESTERS OF the pure state, but the other exists in solution and has a different rotatory pgwer, as in the case of nitrocamyhor, mutarotation may be observable and the effect of agents on the velocity of change studied with advantage, especially since it is clear that when the two forms exist in equilibrium in solution a trace of an agent which affects the velocity of the change in one direction should affect that of the reverse change in a similar manner. Some of the most interesting and most debated cases of tauto- merism exist among the esters of P-ketonic and P-aldehydic acids, such as acetoacetic, formylphenylacetic, and acetoneoxalic acids, and it occurred to us that among the esters formed from these and active alcohols of high molecular weight, one or both of the two forms might be crystalline, and therefore capable of complete purification ; moreover, it seemed that by observations of the mutarotation, the question of the nature of the agents which accelerate or retard the change might be easily investigated.It might appear, at first sight, very improbable that she two forms of such esters would differ sufficiently in rotatory power, as the position in the molecule at which the desmotropic change occurs is somewhat remote from the centre of asymmetry. Nevertheless, the well-known large divergencies which exist between the rotatory powers as calculated by Guye’s method and those actually observed, as well as the very marked effect which is produced on the rotatory power of many compounds by varying the concentration and solvent, made i t seem quite possible that the differences between the properties of the two forms might in some cases be sufficiently large for the purpose.The first optically active ester of the type which we succeeded in preparing was menthyl formylphenylacetate ; the same compound has been independently prepared by Cohen and Briggs (Proc., 1902, 18, 172), and for the same purpose, and the results of these workers serve in the main to confirm our own. The corresponding ethyl ester was prepared by Wislicenus (Be?., 1895, 28, 767; 1896, 28, 742; Amnden, 1896, 291, 147; 1900, 312, 34) by the action of metallic sodium on a mixture of ethyl phenylacetate and ethyl formate, and we have found that a similar process, in which the corresponding menthyl esters are used, affords a fairly good yield of menthyl formylphenyl- acetate.The ethyl ester appears to be capable of existing in two rnodifica- tions. One of these is oily at the ordinary temperature and generally exhibits the properties of a true hydroxymethylene or enolic com- pound, both in its chemical and in its physical characters; thus, its solutions are rapidly coloured violet by ferric chloride, it dissolves readily in alkalis, and gives a copper derivative at once when its alcoholic solution is treated with copper acetate solution ; its specific,@-KETONIC AND @-ALDEEYDIC ACIDS.PART I. 1493 refractive power (Briihl, AnnuZen, 1896, 291, 217) and its absorptive power with regard to rapidly oscillating electrical vibrations (Wis- licenus, AmnaZen, 1900, 312, 32), are also i n agreement with the formula OH*C:CPh*CO,Et. The second modification has the chemical and physical characters of an aldehydic acid of the formula CHO*CHPh*CO,Et, The two forms me produced when solutions of the sodium derivative are acidified ; precipitation by carbon dioxide usually leads t o the production of the enolic form, whilst strong mineral acids in excess produce the aldehydic modification. The two forms are found t o be interconvertible, both at high and at the ordinary temperatures, and also in solution.It was found, moreover, that there are two stereoisomeric forms of the enolic modification, and isomeric copper and benzoyl derivatives were prepared. I n the case of the menthyl ester we have not been able to detect any tendency to persist in the ketonic form. Under all conditions in which Wislicenus found that the oily ethyl ester is produced, me have invariably obtained one and the same compound, and this we believe t o be the aldehydic form of the ester for the reasons stated on p. 1495. On the other hand, we have indications that equilibrium between the two forms is attained extremely readily in alcohol, and the importance of the non-production of a coloration with ferric chloride in benzene or chloroform solution may perhaps be over-estimated, as ethyl aceto- acetate or phenol gives only a faint coloration under these conditions.Further, we have found that, unlike the aldehydic ethyl eeter, the substance reacts rapidly with cold phenylcarbimide, interaction being complete in less than three days. From the description given by Wislicenus of the properties of the two ethyl esters, we are led t o think that the rapidity of change of the aldehydic into the enolic form and vice uersd mmt be far greater in the case of the menthyl ester, a difference which is very difficult to understand. We have also failed t o observe any indication of stereoisomerism in the metallic derivatives of the menthyl ester, Satisfactory evidence that we had been dealing with the menthyl ester of formylphenylacetic acid was afforded by the fact that when the substance is warmed with solutions of phenylhydrazine i t afforded diphenylpyrazolone, which is aIso obtained when Wislicenus' ethyl ester is treated in a similar way.With regard to the point which is held to be of the greatest importance in this investigation, namely, that of the mutarotation, the results were very disappointing. I n the first place, the amount of the mutarotation is very small, and this is the case in chloroform. I n the cases of oxygenated solvents such as alcohol, acetone, or ethyl acetate,1494 LAPWOR'I'H AND HA" : OPTICALLY ACTIVE ESTERS OF little or no mutarotation could be observed, and w0 believe this indicates that equilibrium between the two forms is attained too rapidly for observation, although it is impossible to decide this point satisfactorily.Moreover, the mutarotation in chloroform was so small that we do not feel justified in attaching any importance to the results of the experiments which were made on the influence of agents on the change, and i t may merely be stated that acceleration appeared to be caused both by acids and by bases. EX PER I M E N TA L. heparation of MenthyE Fo~rnplphenylacetate, CH 0 CH (C,H,)*CO,*C, -The rnenthyl phenylacetate required for these experiments was made by heating a mixture of 50 grams each of phenylacetic acid and menthol and 5 C.C. of sulphuric acid on the water-bath for six hours, dissolving the product in ether, washing repeatedly with sodium carbonate, drying with potassium carbonate, and distilling in a vacuum.The fraction employed boiled at 210-215Ounder 34 mm, pressure, had a sp. gr. 0.99, and a rotation [.ID - 68.15'at 16*5', and was found to be pure enough for the purpose for which it was required. I n the first attelhpt to convert it into the hydroxymethylene deriva- tive, sodium and amyl formate in absolute ethereal solution were used. The product appeared t o consist of a mixture of the amyl and menthyl esters and did not crystallise. Menthyl formate was therefore pre- pared ; i t boiled at 215--217' under atmospheric pressure. The following process was found to afford the best yield of the desired product. Sodium (2.3 grams) in thin shavings was placed with 200 C.C. of dry ether in a dry flask cooled by R stream of water. A mixture of 30 grams of menthyl phenylacetate aad 20 grams of menthyl formate was then added in small quantities at a time ; the sodium dissolved fairly readily and the liquid acquired an orange-yellow colour.After the operation was completed, the whole was allowed to remain for several hours, when it was poured into about 500 C.C. of cold water and well shaken with it. The light yellow, aqueous layer was then run off, extracted twice with pure ether, and finally freed from the ether which remained dissolved by means of a stream of air. On adding a few drops of dilute acetic acid, an immediate precipitation of an oil occurred ; -on stirring, this rapidly became converted into a white, pulverulent solid, and the addition of acid was then proceeded with. The product was col- lected on a filter, allowed to dry in the air, and crystallised from hot light petroleum, from which i t separated at once in a nearly pure form.On analysis:@-KETONIC AND &ALDEHYDIC ACIDS. PART I. 1495 0.2773 gave 0.7680 CO, and 0.21 16 H,O. C,,H,,O, requires C = 75.7 ; H = 8.6 per cent. Menthyl formylphenylacetate is readily soluble in almost all the usual organic media ; it dissolves readily also in hot light petroleum and separates on cooling in the form of magnificent, transparent pyramids or prisms, which are sometimes nearly a centimetre in thickness. During the crystallisation, flashes of light are usually observable, even in daylight ; this is the result of the property of tribo- luminescence, which the compound possesses in a remarkable degree, and appears to be caused by the cracking of the crystals owing to the change of temperature; the phenomenon is rendered more obvious by warming the vessel to which the crystals are adhering.The crystals lose some of their transparency when kept, sometimes slowly, at other times rapidly, and it is interesting to note that the power to exhibit triboluminescence diminishes with the transparency. The crystals are tetragonal prisms and shorn the forms (110) and (101). Tbey are frequently elongated in the direction of the c-axis. The double refraction is strong and positive in sign. The compound melts quite sharply at 82--83* and exhibits very little tendency to persist in the liquid condition, even after it has been kept in the fused state for more than an hour ; after solidification, it melts a t the same temperature as a t first.It seems likely, therefore, either that the tendency for the compound to become converted into the tautomeric form is slight, or that the change occurs with great rapidity; the latter view appears to us to be the more probable. When the solid is dissolved in benzene, chloroform, or light petroleum and a drop of an ethereal solution of ferric chloride is added, no coloration is developed at first. On standing, however, the liquid slowly becomes a faint reddish-violet, a change which is brought about rapidly by heating the solution to its boiling point. Again, if the solution in benzene is boiled for a few minutes and then cooled before the ferric chloride is added, a faint coloration is observable as soon as the reagent is added.A much more striking effect is produced if a trace of a base such as pyridine is introduced before or after the addition of the ferric chloride, when a deep violet coloration is at once developed. The development of the colour in alcoholic solution is much more rapid. If some of the powdered substance is placed in a test-tube, dissolved in alcohol, and at once treated with ferric chloride, a reddish-violet coloration is produced, and the intensity of this increases in a few seconds to a maximum. If an alcoholic solution which has been kept for some minutes or has been previously boiled and cooled is used, the maximum coloration appears to be developed instantaneously. These facts, as well as the circumstance that the solid and more stable form of the corresponding ethyl ester has the C = 75.5 ; H = 8.5.1496 LAPWORTH AND HANN: OPTICALLY ACTIVE ESTRRS OF aldehydic constitution, lead us t o believe that, in the present case, the solid is the aldehydic form of the ester.We have made many attempts to obtain a second form of the ester, but without success, By adding a solution oE the sodium derivative of the ethyl ester to excess of acid, an oily modification may be isolated, but under similar conditiohs we have always obtained the solid form of the menthyl ester; it is true that, for tl few moments after precipitation, the material forms an oil, but it invariably solidifies within twenty seconds, unless the sodium derivative has been pre- pared by using alcoholic sodium ethoxide ; i n this case, however, it is largely the presence of alcohol as an impurity which accounts for the smaller tendency to solidify, as may be shown by adding a little alcohol to water containing a small quantity of the finely-powdered solid in suspension, when the compound sometimes becomes oily, but solidifies once more if washed with water two or three times.There is, moreover, no reason t o suppose that the oily character which the material sometimes assumes under such conditions is connected with conversion into another form. In order to determine whether the compound exhibited mutarotation, which would serve to indicate a tendency t o undergo change into i t s desmotropic form, its rotatory power in various solvents was measured immediately after solution, and again after the solutions had been kept for various periods.In 2 per cent. solution in absolute alcohol, i t had [a], -64.9", and no alteration was observed after two days. To hasten any very slow change which might have been going on, a trace of piperidine was added, when the rotation fell t o - 63.9' immed ately and did not alter appreciably with time, so that the effect was probably not due to mutarotation at all. In chloroform solution (2 per cent.), the initial specific rotation mas [a], - '74.6' and fell slightly after the lapse of some days t o -71.3". Experiments were also made with solutions in benzene and petroleum, and in both cases a very slight fall of rotation with time occurred. I n no case, however, was the fall sufficiently great to render it likely that we should obtain any really useful results by studying the effect of agents on the change, and although several experiments were made Ghich appeared to indicate that both acids and bases accelerate the change, we do not feel justified in placing any reliance on them.The sodium derivative, ONa*CH:CPh*C0,*CloH19.-The ester is rapidly dissolved by alkali hydroxides with the formation of an opal- escent liquid, from which the ester is reprecipitated on addition of acetic acid. If the alkaline solution is too concentrated, much hydr- olysis occurs, and menthol is formed in considerable amount. How- ever, it is possible t o precipitate the sodium derivative of the ester from its aqueous solution by the addition of strong, aqueous, ice-cold&KETONIC AND P-ALDEHYDIC ACIDS.PART I. 1497 sodium hydroxide, when it appears in the form of thin plates with a pearly lustre, and may be crystallised from a mixture of benzene and absolute alcohol, It may also be obtained by adding sodium in the form of thin sheets or wire t o an ethereal solution of the ester. On analysis : 0.4168 gave 0.880 Na,SO,. C,,H,,O, Na requires Na = 7-1 per cent. The copper derivative, Cu(O*CH:CPh~CO,*C,oH,,),.-This is formed when a solution of copper acetate in water is added to a dilute alcoholic solution of the ester, and is precipitated in the form of a green oil which slowly solidifies, It is readily soluble in light petroleum, benz- ene, carbon disulphide, or ethyl acetate, separating from these on evaporation as a gum; i t is less readily soluble in methyl or ethyl alcohol, and crystallises from the former in green needles or prisms which melt and decompose at 92-95'.When the crystals are ex- amined in polarised light, they show varying angles of extinction; their double refraction is weak. Na= 6.8. On analysis : 1.9701 gave 0.2410 CuO. Cu= 9.7. (C,,H,,O,),Cu requires Cu = 9-4 per cent. The acetyl derivative, OAc*CH:CPh~CO,*C,,H,,, made by warming the ester with pure acetyl chloride on the water-bath for a short time, crystallised slowly from dry ether ;' these crystals were freed from adherent oily matter by being spread on porous earthenware and recrystallised from methyl alcohol. On analysis : 0.2122 gave 0.5687 CO, and 0,1591 H,O. C = 73.1 ; H = S.3. C,,H,,O, requires C = 73.2 ; H = 8.1 per cent. The compound is readily soluble in most of the usual organic media with the exception of light petroleum, in which it is rather more sparingly dissolved.It separates from methyl alcohol in fine needles melting at 51-52'; it solidifies very slowly on cooling to a mass of long needles. The crystals, in polarised light, show straight extinction ; the double refraction is strong, and the directions of greatest elasticity and length are coincident. The compound is insoluble in dilute alkalis and is slowly hydrolysed by these on boiling; when i t is heated with strong sulphuric acid, the odours of acetic acid and menthene are given o f f . The tenxoyl derivative, OBz*CH:CPh*C0,-ClOHl9, is precipitated as a n oil when the aqueous solution of the sodium derivative of the ester is shaken with benzoyl chloride.It showed no tendency to solidify, even when left for several weeks or when distilled in a vacuum, and then differs from the corresponding ethyl ester, which solidifies when1498 LAPWOltTH AND HA" : MENTHYL FORMYLPHENYLACETATE. subjected to this treatment. It gave the following results on analysis : 0.3132 gave 0.8794 CO, and 0.2161 H,O. The phenylcarbamate, NHPh*CO*O*CH:CPh*CO,~C,,H,,.-When phenylcxrbimide is added to the finely powdered ester, no noticeable development of heat occurs ; the solid slowly dissolves, and in a few hours a clear liquid is obtained if excess of the reagent is employed. If molecular proportions of the two materials are used, the whole finally sets to a semi-solid mass, which may be washed with a little light petroleum, spread on porous earthenware, and finally crystallised several times from ethyl acetate.C = 76.5 ; H = 7.6. C26H3004 requires C = 76.S ; H = 7.4 per cent. On analysis : 0.2825 gave 0.7658 CO, and 0.1872 H,O. C,,H,,O,N requires C = 74.1 ; H = 7.4 per cent. The compound is sparingly soluble in chloroform, carbon tetra- chloride, benzene, or light petroleum, and somewhat sparingly so in methyl alcohol; i t dissolves readily, however, in ethyl acetate or warm ethyl alcohol. It crystallises fairly well from ethyl acetate and melts somewhat indefinitely at 235-237O when slowly heated. The crystals are small needles and in polarised light show straight extinction; the directions of greatest elasticity and length are a t right angles. When melted on a glass slip beneath a cover glass, i t solidifies completely to masses of long needles making up broad patches with aggregate extinction.Action of Hydroxylamine and of Hydraxine on Menthyl _Formy@henyt- acetate.-Numerous attempts to prepare the pure oxime were made. When an alcoholic solution of the ester is mixed with a strong aqueous solution of bydroxylamine, or when aqueous solutions of the sodium derivative of the ester and hydroxylamine hydrochloride are mixed, a viscid, colourless oil separates a t once; this oil has the general properties of an oxime, being somewhat soluble both in dilute acid and dilute alkali, but all attempts to obtain it in a crystalline fqrm were unsuccessful. Efforts to prepare crystalline hydrazones of the ester were equally unsuccessful ; when solutions of the ester in either alcohol or acetic acid were mixed with the acetate of hgdrazine, phenylhydrazine, or p-nitrophenylhydrazine, or with t h e two latter in the free state, no precipitation occurred even when the mixtures mere allowed to remain for some weeks. If the solutions containing phenyl- hydrszine were heated, a crystalline material separated which was collected ; it crystallised from acetic acid in small plates and melted at 196-198'. On heating the compound with strong alkalis or acids, no menthol C = 73.9 ; H = 7.4.LAPWORTH AND HAKN : MENTEIYL ACETOACETATE. 1409 or menthene was liberated, and further investigation proved that the substance was identical with the diphecylpprazolone which had been previously obtained by -similar treatment of ethyl formyl- phen y lace ta te. Our thanks are due to the Research Fund Committee of the Chemical Society for a grant defraying most of the expense of this work. CHEMICAL DEPARTMENT, THE GOLDSMITHS’ INSTITUTE, NEW CKOSS, S.E.