[CONI’RIBUTIONFROM THE SCHOOL OF CHEMISTRY OF THE UNIVERSITY OF MINNESOTA]
T H E CHEMISTRY OF VITAMIN E. VI. T H E ADDITION OF DIENES TO PHENOLS AND HYDROQUINONES’~* LEE IRVIN SMITH, HERBERT E. UNGNADE, HARVEY H. HOE”, STANLEY WAWZONEK
AND
Received March 91, 1939
According t o Claisen2 dienes can be condensed with phenols in the presence of acid catalysts t o give chromans. Thus when isoprene and phenol or p-cresol are dissolved in acetic acid and subjected to the action of dry hydrochloric acid or of zinc chloride, the chromans I and I1 respectively, are formed.
”>”
R\
//
CHz
+ c\ I CH, CH
//
CH2
---+
(01::
R\/ I., R II., R
= =
H CH3
Claisen proved the structure of the chroman I by synthesizing it from coumarin. We were interested in extending this reaction to polymethylated phenols and to hydroquinones and their mono ethers, having in view a possible synthesis of tocopherols from appropriate hydroquinones and phytadiene. It was found that, with certain modifications of Claisen’s conditions, the reaction was fairly generally applicable to methylated hydroquinones and their mono ethers, as well as to polymethylated phenols. The reaction, however, is quite difficult3, and the conditions must be very carefully controlled if good yields are to be obtained with hydroquinones and their derivatives. Using Claisen’s procedure and dry hydrochloric acid as the condensing agent, 2,3,5-trimethylphenol gave a mixture of three different products. These were the chroman 111, m.p. 40-41°, the phenol IVa or Paper V: J. ORG.CHEM.,4, 305 (1939). (in part) at the 96th meeting of the American Chemical Society, Milwaukee, Sept. 5-9, 1938. * CLAISEN,Ber., 64, 200 (1921). SPATH,Ber., 70, 2276 (1937), has also experienced considerable difficulty in applying the reaction to derivatives of dihydric phenols. Thus from 7-hydroxycoumarin and isoprene, he obtained only 1-2 per cent yield of condensation product. 311
* Presented
312
SMITH, UNGNADE, HOE",
AND WAWZONEE
IVb, m.p. 84-86', and the chroman V, a liquid. The structure of IV is not certain, although we favor IVb because of the ready solubility of the subCHa
I11
V
VI
VI1
stance in alkali. Substance V was obtained as an oil which, because of lack of material, could not be purified sufficiently for detailed examination. When the condensation between this phenol and isoprene was carried out in acetic acid using zinc chloride instead of hydrochloric acid, a good yield of the chroman I11 was obtained. Mercuric chloride was also found to be a good catalyst, although the yields of chroman when it was used were not as good as those produced by zinc chloride. In a similar manner, with the use of zinc chloride as the catalyst, 2,2,5,7-tetramethylchroman(VI) was produced from 3 ,5-dimethylphenol and isoprene. Hydroquinone itself was found to be inactive toward dienes under practically all of the conditions used by Claisen. The inactivity may be due partly to adverse solubility relationships, necessitating such high dilutions for solution that the reaction does not occur, but in any event the temperature necessary for a reaction to occur under any of the conditions tried by us is apparently above the decomposition points of the reactants. The mono methyl ether of hydroquinone, however, reacts readily with dimethylbutadiene in acetic and hydrochloric acids to give 2,2,3trimethyl-6-methoxychroman (VII), and with isoprene to give first the
313
CHEMISTRY O F VITAMIN E
halogen compound VIII, which can readily be cyclized to 2 ,2-dimethyl6-methoxychroman (IX) with alcoholic potassium acetate. OH
VI11
IX
The reluctance to react with dienes extends to the methylated hydroquinones, although to a considerably lesser degree than is the case with hydroquinone itself. Thus trimethylhydroquinone, the most reactive of the hydroquinones examined, reacts well with isoprene but only in boiling acetic acid containing zinc chloride and sulfuric acid, and the reaction between hydroquinones and dienes in general appears to require high concentrations of the components and an active catalyst. Under these circumstances, 2 2 , 5 ,7,8-pentamethyl-6-hydroxychroman (X) is produced in good yields. The structure of this substance was proved by synthesizing it from 5,7 ,8-trimethyl-6-hydroxy-3 ,4-dihydrocoumarin (XI)4 and methylmagnesium iodide.
CHaMgI + --
X
XI
Trimethylhydroquinone also was found to react with phytadiene to give dl-a-tocopherol, and although the product was not obtained absolutely pure, the analysis and melting point of the allophanate (155", poor yield) indicate that the substance was about as pure as specimens obtained by other met hods. The reaction between 2 ,5-dimethylhydroquinone and isoprene under 4
SMITH AND DENYES, J . Am. Chem. SOC.,68, 304 (1936).
314
SMITH, UNGNADE, HOE",
AND WAWZONEK
all the conditions tried gave only viscous oils having no definite boiling points and which could not be crystallized. Some hydroquinone was recovered in most of the experiments, but the products were mixtures which we have not, as yet, been able to separate. EXPERIMENTAL
Reaction between dJ8,5-trimethylphenol and isoprene.-The phenol (30 g.) and isoprene (13 9.) were dissolved in acetic acid (32g.), and the solution was saturated with dry hydrogen chloride a t 0". (3 hours), after which the mixture was allowed to stand in the ice box for 36 hours. More isoprene (5 9.) was then added, and dry hydrogen chloride was passed in for 12 hours longer. At this stage most of the phenol had dissolved. The mixture stood a t room temperature for 12 hours, and was then warmed on the steam bath for one hour. An excess of 40% potassium hydroxide solution was added and the solution was extracted thoroughly with ether. The ether layer was washed several times with water and then extracted, first with 40% potassium hydroxide and then with 2% sodium hydroxide. The ether solution was washed with water, and dried over calcium chloride. The solvent was removed, and the residue was fractionated through a Hickman still under 0.1 mm. pressure. The distillate, which came over a t 81-124', weighed 24.1 g. and gave a strong positive halogen test (Beilstein). The material was refluxed under the vacuum of the water pump for 5 hours, when i t was halogen-free. It was then fractionated under 0.1 mm. pressure. Fractions were taken as follows: A , b.p. 58", 5.13 g.; B , 68-108",6.2 g.; C , 108-113°, 9.1 g. A was 2,2,5,7,8-pentamethylchroman(111) (see below). B solidified completely, and C partially, on cooling. The solid in C was found by its melting point to be the same as B ; accordingly C was filtered, and the solid was combined with B. Recrystallization twice from petroleum ether gave white needles; m.p. 84-86'. The substance was soluble in aqueous potassium hydroxide, and Claisen's alkali. Anal.? Calc'd for CMHSOO (IVa or IVb): C, 82.24;H, 9.87. Found: C, 82.26;H, 9.87. The oil remaining after the solid was removed from fraction C was very small in amount. It could not be crystallized and there was not enough of i t to distill. 9 , 2 , 6 , 7,8-Pentamethylchroman (III).-The trimethylphenol (34 g.), fused zinc chloride (4 g.), and isoprene (17 g.) were shaken with acetic acid (30 cc.) for one hour, when the phenol all dissolved. The mixture was allowed to stand a t room temperature for 12 hours and then was refluxed for 7 hours. The mixture was diluted with twice its volume of water, the oily layer was removed and the aqueous layer was extracted three times with petroleum ether. The oil and the extracts were combined and extracted three times with Claisen's alkali to remove any unchanged phenol. After washing with water, the organic layer was dried over calcium chloride, the solvent M-as removed, and the residue was distilled under 2-3 mm. pressure. The chroman thus obtained boiled a t 104-110' and weighed 22 g. I t crystallized from methyl alcohol as a white solid melting a t 40-41". A n d . Calc'd for C l r H 2 ~ OC, : 82.24;H,9.87. Found: C, 82.14;H, 9.88. In a similar experiment, the phenol (5g.), isoprene (2.5g.), and mercuric chloride (0.379.) were heated under an efficient reflux condenser on the steam bath for 6 hours
t Micro analyses by J. W. OPIE AND C.0.Guss.
CHEMISTRY OF VITAMIN E
315
and then brought to a temperature of 240". The chroman (111), isolated and purified as above, weighed 2 g. and boiled a t 152-154' under 27 mm. pressure. d,d,3-TrimethyZ-6-methoxychroman (VII).-Hydroquinone mono methyl ether (12.4 g.) and dimethylbutadiene (8.2 9.) were dissolved in acetic acid (10 g.) and the solution was saturated with dry hydrogen chloride acid a t 0". After standing for 18 hours, the mixture was warmed on the steam bath. Water was added and the product was removed by ether extraction. The ether solution was washed thoroughly with 10% sodium hydroxide, then with water, and was dried over calcium chloride. The solvent was removed, and the product was distilled under high vacuum mm.) in a molecular still. The yield was about 4 g.; b.p. 50-55"; n: 1.5263. A n d Calc'd for CIJHUO~: C, 75.73; H, 8.73. Found: C, 75.97; H, 8.72. d,2-Dimethyl-6-methoxychroman ( I X ) and f-[o-hydroxy-m-methoxyphenyZ]-3-chlorobutane (IX).-Hydroquinone mono methyl ether (18 g.) and isoprene (11 g.) were dissolved in acetic acid (25 cc.), and the solution was saturated at 0" with dry hydrogen chloride. The mixture, after standing in the ice box for 12 hours, was warmed on the steam bath for 30 minutes. After cooling, an excess of 40% potassium hydroxide was added, and the solution was thoroughly extracted with ether. The ether layer .was washed with water several times and then dried over calcium chloride, After removal of the ether, the residue was fractionated in a Hickman still under 0.1 mm. pressure. Two fractions were collected: A , b.p. 74-80' (3.94 g.); B, = 1.5276, reacted with alcoholic silver b.p. 105-107', (3.38 g,). Fraction A had n nitrate only very slowly, and gave a slightly positive phenol test (Folin) and a light-orange color with cold concentrated sulfuric acid. It was undoubtedly the methoxychroman (IX), but i t was possible, with one distillation, to separate i t in a pure state. Fraction B was the chloro compound VIII. It had ng = 1.5310, gave a precipitate rapidly with cold alcoholic silver nitrate; the phenol test was strongly positive, and the color with cold concentrated sulfuric acid was deep-red. It was not analyzed, however, but was converted directly to the chroman I X by boiling for one hour with excess potassium acetate in methanol. The mixture was evaporated to dryness, and the reriidue was taken up in ether and water. The ether was removed, the solvent was evaporated, and residue was distilled under 0.1 mm. in a Hickman still. The distillate, which boiled at 83-90', weighed 2 g. and was the chroman IX. I t had ng = 1.5323, gave no precipitatewith alcoholic silver nitrate, and produced an orange color with cold concentrated sulfuric acid. A n d . Calc'd for ClaHleOz: C, 75.00; H, 8.33. Found: C, 74.48; H, 8.45. 2,~,6,7,8-PentamethyZ-6-hydroxychroman (X).-Trimethylhydroquinone (10 g.) and zinc chloride (1 g.) were dissolved in acetic acid (100 cc.). The solution was heated to 100" under an eflcient reflux condenser, and isoprene (10 9.) was slowly added. The mixture was allowed to stand for an hour, and then was refluxed for an hour. One drop of sulfuric acid was added, and the mixture refluxed for an hour longer The cooled mixture was poured into water, and the white solid was removed. It weighed 7 g. After three crystallizations from dilute ethanol, i t melted at 9494.5'. Anal. Calc'd for ClrHzoOz: C, 76.36; H, 9.09. Found: C, 76.57; H, 9.45.
316
SMITH, UNGNADE, HOEHN, AND WAWZONEK
The substance formed an acetate which melted at 92.5-93.5". The preparation of X, m.p. 94-94.5' and its acetate, m.p. 92.5-93.5', from the coumarin and methylmagnesium iodide is described in the previous paper.' The allophanate of the hydroxychroman melted at 209-211.5"with decomposition. W.Johns who also prepared this allophanate, reported i t to melt at 230". Anal. Calc'd for C16H2204N2: C, 62.74;H, 7.19. Found: C, 62.78;H, 7.28. Phytadiene.-In a n attempt to purify phytyl bromide by distillation under high vacuum, i t was found that the distillate consisted largely of phytadiene.6 The distillate and residue were therefore combined (9 g.) and heated on the steam bath for 45 minutes with potassium hydroxide (10 g.) in methanol (30 cc.). After cooling, ether and water were added, the ether layer was separated, washed with water, and dried. The solvent was removed, and the residue was distilled. It boiled a t 186188" under 14 mm. pressure, and weighed 2.5 g. a-Tocopherol.-A mixture of trimethylhydroquinone (1.3g.), phytadiene (2.5 g.), absolute formic acid (5 g.), and acetic acid (2 9.) was refluxed for 3 hours. After cooling, ether was added, and the solution was washed thoroughly with water. The ether solution was dried over sodium sulfate, the solvent was removed, and the mm.). About 0.1 g. of-unchanged residue was distilled in a molecular still (5 x hydroquinone sublimed first. This was followed by the oily distillate, which was The first two fraccollected in five fractions, although all of i t boiled at 140-145'. tions were dark, due to some of the liquid in the still bumping over, but the last three fractions consisted of a pale-yellow oil with a slight blue fluorescence. The estimated yield was 1.5-2 g. Anal. Calc'd for C2~HaoOz:C, 80.86;H, 11.71. Found: C, 81.65, 81.62, 80.85; H, 12.03, 12.00, 11.95. The substance formed an allophanate which melted at 155", although the yield was poor, and the tocopherol was apparently contaminated with a high-boiling oil, insoluble in methanol.$ One of the later fractions was adsorbed on Brockman's alumina. The tocopherol layer could be distinguished in ultra-violet light. It was extracted with a mixture of methanol and ether. The residue after removal of the solvents reduced silver nitrate in methanol, and gave a n orange-yellow color with cold concentrated sulfuric acid. Attempts to prepare the dinitrobenzoate from a small amount of the material were unsuccessful. Reaction between 2,6-dimethylhydroquinone and isoprene.-The hydroquinone (13.8 9.) was dissolved in acetic acid (70 cc.) and ether (15 cc., absolute), and the solution mas cooled to 0'. Isoprene (6.8g.) was added, and the solution was saturated with dry hydrogen chloride. The mixture, after standing in a n ice box for one week and then a t room temperature for 24 hours, was poured into water and extracted with ether. Extraction of the ether with alkali removed some hydroquinone and left a dark, viscous oil which could not be crystallized nor distilled. An experiment analogous to that used for the preparation of VI11 above, involving 2,5-dimethylhydroquinone (9 g.), zinc chloride (1 g.), acetic acid (70 cc.), and
JOHN, Ber., 71, 2646 (1938).
* Paper IV, J . ORG.CHEM.,4, 298 (1939). $ The analysis of the tocopherol and preparation of the allophanate, were carried out by Messrs. Emerson and Haymrtn in the laboratories of Merck & Co., Inc., Rahway, N. J.
CHEMISTRY O F VITAMIN E
317
isoprene (10 g.) also gave only a dark oil which could neither be crystallized nor distil.led. SUMMARY
1. The condensation between dienes and phenols leading to chromans, has been extended to certain hydroquinones and to hydroquinone mono methyl ether. 2. Under proper conditions, these substances give good yields of chromans, but unless the conditions are carefully regulated, mixtures will result. 3. Hydroquinone itself does not react under any of the conditions tried, but its mono methyl ether reacts readily. Trimethylhydroquinone reacts well, but 2,5-dimethylhydroquinonedoes not. 4. a-Tocopherol has been synthesized in this way from trimethylhydroquin.one and phytadiene.