for thirty minutes. After the usual work-up, there wa5 obtained 1.54 g. (88%) of I -methyl-As-dehydroestradiol with m . p. 129-132", which gave no depression on admixture with the sample prepared according t o (A) and possessed the same rotation and spectrum. Hydrogenation of 1-Methyl- As-dehydroestradiol 3,17Diacetate to 1-Methylestradiol 3,17-Diacetate.-Hydrogenation of the A.6-dehydro diacetate in the usual manner in ethyl acetate solution with 10% palladium-on-charcoal catalyst (barium sulfate supported catalyst was equally satisfactory) afforded 81% of shiny, prismatic blades (from methanol) of I-methylestradiol 3,17-&acetate with in. p. 178-180" (Kofler), [ L ~ ] . ~i -Dl l l " , u. v. inaximumat 268 mp (log E 2.53) and minimum a t 252 rnp (log E 2.443. The so-called 1-methylestradiol 3,17-diarst:it~~describpd i n the literat~re69~ had in. p. 1.38.5-139". d n d . Calcd. Cor C . J I ~ O C C', ~ " 74.W: H , h.lh. Pound: C, 74.50; TI, 8.19. 1-Methylestradiol (IVa) , prepared either by saponification of its diacetate or by hydrogenation of 1-methyl-A6dehydroestradiol (XIIIa), was obtained from etherhexane s a microcrystalline powder, which shrank a t 95" and melted a t 110-116" (Kofler), [ m I z 0 +146", ~ u. v. maximum a t 284 mu (log E 3.28) and minimum a t 250 mp (log E 2.25). The physical constants of this alkalisoluble phenol are in complete contrast to those reported for the so-called 1-methyle~tradiolj~~ : insoluble in alkali, crystallizesreadily, rn. p. 2 5-2.16.50, [ a ]n 185" idioxme). Anal. Calcd. For Cl~1&O,: C . 79.67; H. 9.1 4 . Found: C. 79.87: H. 9.36. The 3,17-&propionate was isolated in nearly quantitative yield on heating 1-methylestradiol (IVa) for one hour with propionic anhydride and py:idine ; colorless plates fro? methanol, m. p. 125.5-127 (Kofter), [ a I P a D i-101.5 , u. v. maximum a t 268 m p (log E 2.58) and minimum a t 254 mp (log E 2.45). A n d . Calcd. for CYSH340t: C, 75.34; H, 8.5'3. Found: C, 75.16; 13, 8.63. 1-Methyl-17-dihydroepuilenin-l7j5 (XIVa) .--This phenol was obtained by both selenium dioxide oxidation of I-methyl- A6-dehydrocstradiol 3J7-diacetate followed by saponification, or by lithium aluminum hydride reduetion of 1-methylequilenin acetate exactly as described for the analog lackiiig the 1 -1iiCthyl group
+
1 -Methyl-17-dihydroequilenin-l7@crystallized as small, prismatic needles from either aqueous methanol or hexaneacetone, m. p. 225-227" (Kofler), [ ~ ] * O D 4 3 2 " (dioxane), t i . r. spectrum Fig. 2. d a a l . Calcd. for C I O H ~ ~C, O ~80.81; : H, 7.85. Found: C . 80.54: H, 7.65. The 3,17-diacetate was obtained as colorless needleos from methanol, m. p. 145-147' (Kofler), [.\*OD -16 . The u. v. spectrum was very similar t o that of l-methylequilenin acetate and is depicted in Fig. 2 . Anal. Calcd. for C~~HZBO,: C, 75.38; H, 7.15. Found: C, 73.58; H,7.12
Summary The dienone-phenol rearrangement of the recently described2b10 A'~4*6-androstatrien-17-ol-3one and the corresponding 3,17-dione (XII) produced 1-methyl-A6-dehydroestradiol and 1methyl- Ab-dehydroestrone (XIII) Dehydrogenation led to the corresponding 1-methylequilenin derivatives (XIV) , while catalytic hydrogenation afforded 1-methylestradiol and l-methylestrone, characterized by alkali solubility and high estrogenic potency. A consideration of the reaction mechanism and earlier work on model compounds6,8indicates that the present compounds are the true 1-methylestrogens in contrast to the rearrangement products of steroidal A1l4-dien-3ones (111), previously believed to have this structure,6 which should now be referred to as "xinethylheterophenols. " The present work, coupled with the results of Woodward and Singh,8 demonstrates that the dienone-phenol rearrangement in the steroid series proceeds in different directions depending on the presence or absence of :in additional, conjugated double bond.
.
LAGUNA MAVKAN 413 MEXICOCITS, 1). F.
[ C O Y r R I B V T I O \ FROM THE DEPARTMEN'T OF CHEMISTRY OF
R E C E I V ~FEERUARY .~ 20, 1950
THEOIIIO STATE UNIVERSITY 1
The Structures of p-Diacetone-D-fructose and 8-Monoacetone-D-fructose I-lu AI. I,. WOLFROM, WXLBUR L. SHILLING' AND W. W. BINKLEY' Fischer2 obtained two crystalline products on treating D-fructose with acetone i a t h e presence of acid and these have come to be designated a-cliact$onefructose and 6-diacetonefructose wherein the prefures were assigned in accordance with the order of isolation and bear no anoineric significance. The structure uf the former can be considered to be adequately established as 1,2:4,5diisopropylidene-~-fructopy~-anose~~~ while that of the latter cannot. tion Fellow (W. L. S.) and Research (1) Sugar Research Fo W. B.) of The Ohio State University Research PoundaAssociate (W. ti6n (Project 190). (2) E. Fischer. Bcr , 28, 1164 (1895); cf. H.Ohle and Use Koller, iW.,67, 1588 (1924,. (3) H. Ohle, ibdd., 60, 1168 (1927). (4) C. G . Anderson, W. Charlton. 'A' N H a r t w l i i irrd l' ,z Uicholson. J C k e m Pr,r 1:i17 t I W + )
It is known that C1 is open in the P-isomer since on oxidation with alkaline permanganate an acid was obtained without acetone removal and this acid on hydrolysis yielded 1-C-carbbxy-b-arabinose (2-keto-D-gluconic acid).6 The main difficulties preventing a solution of this problem have been those encountered in preparing the mom-isopropylidene derivative, essential to the structure determination. Both isopropylidene residues hydrolyze a t nearly the same rate and on partial hydrolysis there is obtained a difficultly separable mixture of starting material, D-fructose and a relatively midl amount of the sirupy mono-isopropylideneaD-fntctose. In the work herein reported the mono derivative has been separated by chrorna,)
I1 Ohle, Be?, 58, 2577 11925): H. Ohle niid Gertrud b r i a u d . 60, 1159 il9271
tography on clay and characterized as a crystal- At this point an equal quantity of crushed ice was added the resultant solution was passed over a 680 X 74 mm. line triacetate. Deacetylation of this non-reduc- and (diam.) column of Duolite A49 anion acceptor resin preing compound yielded a product consuming one cooled to ca. 5" with ice water. The effluent and washings mole of periodate per mole of substance. Only were concentrated under reduced pressure to a mixture of two possible non-reducing monoisopropylidene sirup and crystals. This residue was dissolved in 1 liter 98% 2-propanol (sp. g. 0.786/23 ") and chromatographed derivatives of D-fructose are compatible with this of on 3 kg. of Florex XXX1o-Celitelo ( 5 : l by wt.) in an 8 1. analysis : 1,'2-isopropylidene-D-fructofuranose and pharmaceutical percolator. Development was effected ~,3-isopropylidene-~-fructopyranose. The former with 3.5 1. of 9870 2-propanol, the course of the developis eliminated by the evidence cited above. The ment being followed by applying the Molisch reaction to effluent. This removed the diisopropylidene derivastructure of /3-11ionoacetone-~-fructose is thus rig- the tive (17.5 9.) , isolable on solvent removal. Development orously proven to be ~,3-isopropylidene-D-fructo- nith 9870 2-propanol was continued and the succeeding pyranose (I ) and since hydrolysis by aqueous acid 8 1. of effluent was concentrated under reduced pressure, will not shift the position of an attached isopropyl- treated with decolorizing charcoal and concentrated to a colorless sirup ; yield 5.7-g. of monoisopropylidene derivaidene group, the structure of /?-diacetone-D-fruc- tive, [ a ] 2 * ~$17.5" (c 2.3, water), [aIz4D$28.2" (c 2.9, tose is likewise established as 2,3 : 4.5-diisopropyli- ethanol), after purification through the crystalline triscetate described below. Ohle and Kolleri record the dene-D-fructopyranose (I I),
+
value [ C Y ] ~ D 27.6" (c 1.5, ethanol). The D-fructose fraction remained on the top of the column and was isolable by elution with 2 liters of S0/20 (by vol.) ethanol/ water with subsequent solvent removal; yield 3.2 g. of sirup. 2,3-Isopropylidene-~-fructopyranoseTriacetate.-An amount of 7.8 g. of the above sirupy 2,3-isopropylidene-~fructopyranose ( p-monoacetone-D-fructose) was acetylated for five hours a t 75" with acetic anhydride (40 ml.) bH H Mq-0 H and anhydrous sodium acetate (1 g.). The cooled reacI* 11 tion mixture was poured into 250 g. of ice and water and * Anomeric configurations arbitrdry. adjusted, with stirring, to PH 6 with sodium bicarbonate (60 g . ) . The dried chloroform extract was concentrated The prefixes CY and /? are then no longer re- to a sirup under reduced pressure and acetic acid was required to distinguish the known isopropylidene moved by co-distillation with toluene under reduced pressure. The product was crystallized from ether-petroleum derivatives of D-fructose and the definitive ra- ether (b. p. 3040"); 10.6 g. (two crops, 937()? m: p. tional names should replace the older trivial terms. 55.556" (cor.) unchanged on further recrystallization, I t is interesting that in the formation of both of [a]"D $18" (c 5 , abs. ethanol), [ C X ] ~ ~f D7 " (c 5 , chlorothese diisopropylidene derivatives of D-fructose, form). The substance crystallized as thick prisms or plates. It was readily soluble in the common solvents this ketosehas reacted under acid conditions in the thin escept water and petroleum ether. pyranoid ring conformation. The properties of Anal. Calcd. for C~H~O~.(CH~)PC.(COCH~)~: C, 52.02; the three crystalline isomeric monoisopropylidene- H. 6.40: CHXO. 9.9 ml. 0.1 N NaOH Der 100 m e . : D-fructose triacetates are compared in Table I. CHs-C, 4males/rnde. Found: C, 51.92; H, 6.41; CI%& CO, 9.5 ml.; CHI-C, 3.7 moles. TABLE I Periodate Assay on 2,3-Isopropylidene-~-fructopyranose (8-Monoacetone-D-fructose) Purified through its PROPERTIES OF THE ISOMERIC MOSOISOPROPYLIDENE-DCrystalline Triacetate.-To 0.1745 g. of the above crysFRUCTOSE TRIACETATES talline ~,4,5-triacetyl-2,3-isopropylidene-~-fructop~ano~ [a]22-% dissolved in 10 ml. of dry methanol was added a solution c < 10, of 7.5 mg. of sodium in i . 5 ml. of dry methanol. After bl. p . . OC. ethanol Ref. Substarice standing overnight a t 10 , the solution was made up t o 1,2-Isopropylidene-D-fruc- 99-101 - 135,5" 6 100 ml. with 20 ml. of 0.046 iM sodium metaperiodate, 1.6 ml. of 0.1 N hydrochloric acid, and water. The solution topyranose triacetate (cor.) was maintained at 20" and aliquots showed the following 2,3-Isopropylidene-1~-fruc-55.6-56 18 This periodate consumption in moles of oxidant per mole of topyranose t riave tate (cor.) work substance: 0.95 a t 1.5 hr.; 0.96 a t 4.5 hr.; 1.0 a t 11.8 hr. H
I
*(
+
",3-Isopropylideiic-D-Eruc-
tofuranosc triace title
55
- 8
7
Experimental Preparation of 2,3-Isopropylidene-D-fructopyranose ( 8 Monoacetone-D-fructose) .-This substance was prepared essentially according to the procedure of Ohle and Koller2 but was purified by chromatographic technics. An amount of 32.4 g. of finely powdered p-diacetonea-fructose* was hydrolyzed for thirteen hours (approximately ~ basis the point of smallest levorotation; [ a ] -20", starting material) a t 22" in 1100 ml. of N sulfuric acid. (6) E. Fischer and H. Noth, &r., 61, 346 (ISIS). (7) L. Zervas and P. Sessler, ibid., 66, 1698 (1933). (8) E. Pacsu, E
