[CONTRIBUTION FROM r n LABORATORY ~ OF ORUANIC CHEMISTRY OF OF WISCONSIN]
THE:
UNIVERSITY
CHRYSENE DERIVATIVES BY THE ROBINSON-MANNICH BASE SYNTHESIS A. L. WILDS
AXD
RICHARD G . WERTH’
Received April 9, 1962
In previous reports from this laboratory improved procedures have been described for carrying out the Robinson-Mannich base synthesis of cyclic unsaturated ketones, and these were employed to prepare 2- and 3-substituted derivatives of chrysene (1). The method has also been shown to be readily extended t o the synthesis of 1-substituted chrysene derivatives (2). The present paper reports further improvements and variations in this synthesis and its use to prepare 4-methylchrysene. In an accompanying communication the method has been extended to the synthesis of benzo[c]phenanthrene and derivatives (3).
z
Z
CHn
/
(yp\ 0
---+
I, z = COOCHd VI, Z = =CHOW
11, Z = C O 0 C H a ; R = H IV, Z = COOCHs; R =: CHO VII,Z = H ; R = H
VIII, R = H I X , R = C&
_3
111, R =
w
V,R = CH;
’->
CI-4C OCHI CHeN \-
X
I n previous work when 2-carbomethoxy-1-ketotetrahydrophenanthrene(I) was condensed with the methiodide of 1-diethylamino-3-butanone(VIII), it mas found that careful purification of this Mannich base was essential. With incompletely purified reagent the desired 2-y-ketobutyl derivative (11) was accompanied by varying amounts of the 2-@-methylene-y-ketobutyl) derivative resulting from 1 ,1-bis(diethylaminomethy1)acetone present as an impurity in VI11 (1). The piperidino Mannich base (X) has now been found t o be superior to the 1 Wisconsin Alumni Research Foundation Research Assistant, 1948-1950. Present Address: Department of Chemistry, Concordia College, Moorhead, Minnesota. 1149
1150
A. L. WlLDS AND R. G . KVEETH
diethylamino derivative' TT7ith piperidine hydrochloride in the 3iannich rcaction with acetone the tendency toward bis-condensation is negligible and the Uannich ba6e is obtained in superior yield. In addition the crystalliae methiodide of X is less hygroscopic and more stable than that of VIII, and may be isolated if desired. Excellent yields of uncontaminated products result from its use in the unsaturated ketone synthesis. In condensations involving these Mannich base methiodides with sodium enoIates such as that of I, it is presumed that methyl vinyl ketone is liberated in situ and undergoes a Michael condensation as it is formed. The early attempts by Robinson and coworkers to utilize methyl vinyl ketone itself were unsuccessful due to the ready polymerization of this reagent by strong bases (4). We have found, however, that it may be employed Kith excellent results to form 11 (92% yield) provided only a small amount of sodium methoxide is used as the catalyst. Moe and Warner (5) have described successful Michael condensations using acrolein under these mild conditions. Recently Henecka has also obtained good resu1t)swith methyl vinyl ketone in a similar condensation ( 6 ) . In another series of compounds it was found that the hydroxymethylene group possessed advantages over the earbomethoxy group in activating ketones for this synthesis (7). This variation has also been evaluated in the present series. 2-Hydroxymet~iyle~ie-1-ketotetrahydrophenaiithrene (VI), prepared in 96-98% yield from the ketone, reacted readily n-ith the methiodide of the piperidino Mannich base (X) l o give a mixture from which the diketone VIS could be isolated. Acid cyclization of the mixture, which probably contained some of the 2-formyl derivative of VII, gave the tetracyclic ketone I11 in as high as 78% over-all yield from VI although the coinpound was not as pure as that obtained in 84--88'% yield from the carbomethoxy derivative I. When one wishes to prepare the diketone lacking the activating group (e.g. 1711) the use of the hydroxymethylene ketone (e.g. VI) is indicated, since the formyl group is eliminated extremely easily from the initial reaction product. We have not yet been able to prepare such diketones from the keto ester derivatives ( e g . 11),the diketones instead undergoing cyclization to the unsaturated ketones (e.g. 111). In extending the method to the synt,he& of derivatives of 4-methylchrysene, the keto ester I was condensed with the methiodide of l-diethylamino-3-pentanone (IX) to afford in excellent yield the y-ketopentyl derivative IV. Initial experiments with dilute alkali indicated this compound to be less readily cyclized than the lower homolog 11. By extendiiig the time of reaction, however, good results were obtained. Reduction of the ketone V by the Huang-Miiilon modification of the Wolf-IGshner method afforded crystalline 4-methylhexahydrochrysene which was dehydrogenated smoothly with palladium-on-carbon catalyst to 4-methylchrysene. These results, together with those previously reported, indicate that the method should be applicable to the synthesis of a wide variety of chrysene derivatives from 1-ketotetrahydrophenanthrene (or its derivatives) and the appropriate Maniiioh bases or unsaturated ketones. 2 The morpholino derivative appeared to be less satisfaciory than X in a preliminary evaluation.
CHBYSENES BY THE ROBIKSON-MANNICN IL4BE SYLUTIIESIS
1151
EXPERIMENTAL^ I-N-Piperidino-8-butanone (X). The procedure of Mannich and Hof (8) was modified as follows. A mixture of 36.6 g. of piperidine hydrochloride,' 10.8 g. of paraformaldehyde, 110 ml. of acetone, and 16 ml. of methanol was refluxed for six hours, cooled, and treated with excess 45% potassium hydroxide. The water layer was saturated with potassium carbonate, extracted with ether, and the extract dried over sodium sulfate and finally over Drierite. After a small forerun, 32.9 g. (71%) of the bIannich base was obtained, b.p. 83-87' a t 6-8 mm., n: 1.4630; there was no evidence of any bis-condensation product in this material (1). Refractionation gave colorless material distilling a t 89-90' (7 mm.), n: 1.4628. A n a l . Calc'd for CQH~,NO: C, 69.6; IT, 11.0. Found: C, 70.0; H, 10.9. The hydrochloride, prepared in methanol with dry hydrogen chloride and obtained by removal of the solvent in VUCZLO, crystallized from acetone as colorless leaflets, m.p. 184.5185" [reported (8) 167'1. A n a l . Calc'd for CsHlsClNO: C1,18.5. Found: C1, 18.3; 18.3. The methiodide was prepared by adding in portions with cooling an equal weight of methyl iodide t o the amine and allowing t o stand a t 0" for one-half hour and then for several hours at room temperature while protected from moisture. The solid methiodide was mashed by decantation with dry ether; yield 79-83%, m.p. 85-90' (dec.) A n a l . Calc'd for CIOH~OIKO: I, 42.7. Found: I, 42.6; 42.6. Attempts to recrystallize the methiodide from mixtures of acetone, ether, and ethyl acetate led t o partial decomposition, as indicated by analysis for iodide (for example, Found: I, 45.6, 46.0, 45.9); on the assumption that methyl vinyl ketone and N-methylpiperidine hydriodide (Calc'd: I, 55.9) were formed, this corresponded to approximately 25% decomposition. Such samples were unsuitable for the subsequent reactions, as were the variable products obtained when this methiodide was prepared in ether solution. W - y - K e t o b u t y l - 2 - c a r b o m ~ ~ h o x ~ - I - ~ e t o - l , 2 , 3 , ~ - t e t r a h y d r o p ~ ~ e n a n(11). t h r e n e( a ) Using methyl vinyl ketone. T o a stirred solution of 10.0 g. of 2-carbomethoxy-l-keto-1,2,3,4-tetrahydrophenanthrene (9) and 0.050 g. of sodium methoxide (The Mathieson Chemical Corp.) in 40 ml. of dry methanol and 60 ml. of dry benzene there was added dropwise a solution of 5.5 g. of anhydrous methyl vinyl ketone6 in 10 ml. of methanol and 30 ml. of benzene over a ten-minute period (temperature rose from 23" t o 30"). After five hours a t room temperature water was added and the product was extracted thoroughly with benzene, followed by washing with dilute acid and water. A total of 11.7 g. (92%) was obtained in three crops all melting a t 143-145" [reported (1) for purest sample, m.p. 145-145.5'1. (b) Using I-A'-piperidino-8-butanone methiodide. This reaction was carried out as described earlier (1) except that the solid methiodide of l-N-piperidino-3-butanone was used and allowed t o react a t room temperature for twenty or more hours. The product, largely of m.p. 143.5145' with additional crops melting in the range 135-143', totalled 92-97%, with no evidence for the presence of the @-methylenederivative arising from a bis-methiodide (1). 3-Keto-I , 2 , 3 ,I f , I2,ISa-hezahydrochrysene(111).The procedure for cyclization employed by Wilds and Close (10) was found to give more uniform results than that described earlier; following the same procedure as used t o prepare A45 a'-2'-keto-l, 2-dihydro-3,4-cyclopen3 Melting points are corrected unless noted otherwise; microanalyses for carbon and hydrogen by Virginia Diekmann Miller, Richard Hunt, Edward Shiner, and Bennett Buell. Prepared from an ether solution of piperidine and dry hydrogen chloride, m.p. 240245" (dec.). 6 Prepared from the 85% azeotrope (du Pont Co., Organic Chemicals Dopartment) by treating with solid potassium carbonate and drying the organic layer over sodium sulfate and finally Drierite, b.p. 33-35" at 120 mm.; 80% recovery (modified procedure of Carl IT. Hoffman).
1152
A. L. \t'lLDS AKD R. G. WERTII
tenophenanthrene, the lretochrysene derivative I11 was obtained in a total yield of 91% (of which 58% had m.p. 187.5-188.5", 22% m.p. 180-184', and the remainder m.p. 173-180"). 2-7-Ketobutyl-I-keto-1 ,d,B,Q-lefrahydrophenanthrene(VII). To a solution of 0.14 g. of sodium in 10 nil. of dry methanol was added 1.37 g. of 2-hydroxymethylene-l-keto-l,2,3,4tetrahydrophenanthrene (11, 2) and then 3.62 g. of I-piperidino-3-butanone methiodide in 10 ml. of methanol. After 24 hours a t room temperature, the product was extracted thoroughly with benzene and washed r i t h dilute alkali, water, and dilute acid. After evaporation of the solvent the product nas crystallized from ethyl acetate and from methanol to afford a total yield of SO%, with most of the material melting at 72.5-73". Recrystallization from methanol gave a solid melting a t i7-78". Anal. Calc'd for Cl&IlsO~:C, 81.2; H, 6.8. Found: C, 81.6; H, 6.9. When the crude non-crystalline mixture mas cyclized with methanolic potassium hydroxide, or better m-ith hydrochloric and acetic acids (l),the crude ketone I11was obtained in as high as 78% yield (based on hydroxymethylene ketone), but less pure than from the crystalline carboinethoxy derivative I1 (25% with m.p. 182-184", the remainder melting as low as 16b173'). R-y-Ketopentyl-2-carbomethoxy-1-keto-1 ,R,B,d-tetrahydrophenanthrene(IV). The sodio derivative prepared from 5.37 g. of 2-carbomethoxy-l-ketotetrahydrophenanthrene in 50 ml. of methanol and 50 ml. of benzene containing sodium methoxide from 0.485 g. of sodium was allonred to react with the methiodide of I-diethylamino-3-pentanone (12),b as described earlier for the butanone derivative. After 18 hours a t room temperature the product was isolated and crystallized Prom benzene affording 5.87 g., m.p. 117-118', with additional material from the citrate (most melting a t 114-117") t o bring the total yield t o 96%. The analytical saniple crystallized from benzene as colorless prism8, m.p. 117.5-118.5". Anal. Calc'd for C21H2204: C, 745; H, 6.6. Found: C, 74.4; H, 6.6. 3-lCeto-~-methyl-l,2,9,11,12,iRo-hexahydroch1~ysene (Vi. Cyclization of 2 g. of the keto ester IV was carried out under nitrogen by the two-stage aqueous alkaline procedure described earlier (lo), except that with this example the second period of reflux with 3% potassium hydroxide was extended t o 18 hours. The product was crystallized from cyclohexane, 1.20 g., m.p. 121-124'; two additional crops were obtained from ethanol, 0.06 g., m.p. 115.5-136.5", and 0.08 g., m.p. 111-3 16". It is proaable that a still longer time of cyclization would be preferable. Recrystallization from absolute ethanol gave fine, colorless prisms, m.p. 126.5-727.5". The ultraviolet absorption spectrum in 95% ethanol showed maxima at 222 m p (log e = 4.47), 279 mg (4.50), 314 mg (4.377, and minima a t 240.5 nip. (3.95) and 291 n i p (4.15) (Cf. 13).
Anal. Calc'd for C I O H ~ ~C, O :87.0; H, 6.9. Found: C,87.1; E,7.1. 4-Methyl-I,W,3,11,12,Ida-hexahydrochrysene.The Huang-hlinlon reduction was carried out under nitrogen using the special apparatus described previously (14). Reduction of 500 mg. of the above ketone as described for 14,15-dehydro-l6-equilenone gave an oil which crystallized from absolute ethanol, 342 mg. (72%), m.p. 61.567'. After evaporative distillation a t 100-130" (0.1 mm.) and further recrystallization from ethanol, the m.p. of the colorless needles was 83-85'. Anal. Calc'd for Cl~H20: C, 91.9; H, 8.1. Found: C, 91.9; €1, 8.2. /t-Methylehrysene. Dehydrogenation of 137 mg. ol the 4-methylhexah~-drochrysene,m.p. 67-69', with 70 mg. of 30% palladium-on-carbon catalyst (15) under nitrogen at 300-320" for one hour gave a total of 78% of 4-methylohrysene after crystallization from benzene and acetone, m.p. 151-161.5" [reported (16) m.p. 151-151.5"]. -
6 An improved procedure for preparing this aminoketone in 60-70% over-all yield from propiongl chloride, ethylene, and diethylamine mill be described by Wilds, McCaleb, Inglett, and Dessauer.
CHRTSENES BY THE ROBINSOK-MAXKICH BASE SYXTHESIS
1153
SUMW4RY
Some variations in thc Robinson-Mannich base procedure for synthesis of unsaturated cyclic ketones have been studied as applied to the preparation of 3-ketohexahydrochrysene (111). Methyl vinyl ketone and the methiodide of 1-piperidino-3-butanone have given superior results with carbomethoxy and hydroxymethylene ketones. The -1-methyl derivative of I11 has been prepared using 1-diethylamiizo-3-pentanone methiodide and this compound was converted to 4-methylchrysene. MADISOX 6, WISCONSIN
REFERENCES (1) WILDSLVD SHUNK, J . Am. Chem. SOC.,66,469 (1943). (2) WILDSAND DJERASSI,J . Am. Chem. SOC., 68, 1715 (1946). (3) WILDSASD WERTH,J . Org. Chern., 17, follom-ing paper (1952). (4) DU FEU, MCQUILLIN, .mD ROBINSOS,J . Chem. soc., 53 (1937). (5) MOE AND WARNER,J . Am. Chem. Soc., 70, 2763, 3470 (1948). (6) HENECHA, Ber., 82, 112 (1949). (7) SHUXK AND WILDS, J . Am. Chem. Soc., 71, 3946 (1949); 72, 2388 (1950). (8) &IA"ICH AND HOF,A&. Pharm., 266, 589 (1927). M D WILDS,J . Am. c'hem. Soc., 62, 2085 (1940). (9) B~CHM~YU'N (10) WILDSAND CLOSE,J . A m . Chem. SOC., 68, 84 (1946). (11) W. S. JOHNSON AND SHELBERG, J . Am. Ckem. Soc., 67, 1750 (1945). (12) ADAMSON, XCQUILLIN, ROBINSON, AXD SIXONEEN, J . Chem. SOC.,1578 (1937). AND SHUNK, J . Am. Chem. Soc., (13) WILDS, BECK,CLOSE,DJERASSI,JOHNSOS,JOI-~XSOS, 69, 1988 (1947). (14) WILDS,J. A. JOHNSON, ATD SUTTON, J . Am. Chern. SOC.,72,5524 (1950). (15) LINSTEADAND THOMAS, J . Chem. SOC.,1130 (1940). (16) FIESER AND W. S. JOHNSON, J . Am. Chewz. SOC,61,1847(1939);BACHMAKN AND STRUVE, J . Org. Chem., 4, 456 (1939).
