4962
ROBERTE. SCHAUR. G E O R G E R.-ILLEN, J R . , AND
11.4RTIN
J. IvEISS
VOl. s1
the compound to be identic:il to that obtained in preparation sodium methoxide (1.15 g.) in methanol (7.8 ml.) and tlie A. mixture was added to a solution of Sa-fluoro-1 lp-hydrosy9 ~ ~ F l u o 1r lp,16~,17~-trihydroxy-1,4-pregnadiene-3,20o lBcu,l7~u-isopropylidenediox~~-pregnene-3,20-dione ( XVIh, 7.8 g.) in t-butyl alcohol (62.4 ml.). The reaction mixdione (Ib).-Fifty flasks, each containing a beef extract, yeast extract, peptone and cerelose medium (100 ml.), ture was kept a t room temperature for 24 hours and wi\ then diluted with a large volume of petroleum ether. Thc were inoculated with 1% of an 8-hour growth of Nocardia corallina (ATCC 999). The flasks were placed on a re- yellow sodium salt which separated was filtered off, wadied ciprocating shaker and incubated a t 28" for 17 hours when a with ether, and dried. Starting material (3.4 g.), n1.p. 240solution of 9a-fluoro-llp,l6a,l7a-trihydroxy-4-pregnene- 24-L0, was obtained from the filtrate. A solution of this salt in water was acidified (congo red) with dilute hydro3,211-dione ( X I I d , 20 mg.) in methanol ( 2 ml.) was added to each flask. The fermentatioii was continued for 11 hours chloric acid (10%) and the precipitated light-yellow solid and the flasks then were harvested and their contents pooled. was collected, washed with water, and dried. Tlie weight The beer was extracted with ethyl acetate (3 X 4 1.) and the of dried product was 4 g. combined extracts were washed with water and dried. T h e above product (620 mg.) and anhydrous potassium The dried solvent was concentrated to a volume of 1 liter, acetate (233 mg.) were dissolved in methanol ( 7 . 5 ml.) and treated with charcoal, and concentrated further until crys- the dark-green solution was cooled in a n ice-bath. A solutals (750 mg.) separated. This material was chromatotion of bromine (190 mg.) in methanol ( 2 nil.) was added graphed on Celite using a partition system of cyclohexane dropwise (1 drop per second) to the stirred solution. \Vlien ( 4 vols.), dioxane (5 vols.) and water (1 vol.) and the prodthe addition was complete, methaiiolic sodium irietlioxide uct (528 mg. after washing with a little ether) crystallized (1.5 ml., 1 N ) and phenol (10 nig.) were added to the alniost from acetone as small prisms, m.p. 280-287' dec., [ a I z 5 D colorless solution and the mixture was heated under reflux 4-62" ( c 0.503, methanol), +41.5" ( c 1.062, pyridine), on the steam-bath for 10 minutes. The cooled renrtion A,, 238 mp (E 15,100); vmax 3509, 3401, 1709, 1667, 1618 mixture was poured into water and the product \vas filtered and 1603 cm.-l. off, washed with wzter, and dried. T h e crude bromo compound weighed 540 mg. !!?tal. Calcd. for C21H2;FO:,(378.43): C, 66.65; H, 7.19; The crude bromo product ( 3 . 5 g.) and s-collidine (70 ml.) F, 5,02. Found: C, 66.86; € I , '7.40; F, 4.81. were heated under reflux for 5 hours. Tlie cooled reaction The acetate, 16a-acetoxy 9a-fluoro-1 Ip, 17cu-dihydrosymixture was diluted with ether, filtered, and the filtrate 1,4-pregn.~diene-3,20-dione( I C ) , was prepared by treating was washed with dilute hydrochloric acid ( 3 N), water, the triol I b with acetic anhydride-pyridine overnight and it and dried. The brown solid (2.085 g.), obtained by recrystallized from ethyl acetatepetroleum ether as needles, moval of solvent, was dissolved in benzene and chromatom.p. 242-244', [ C Y ] ~ D4-27.3' ( c 0.768, methanol), A,,, graphed on neuti-31 alumina (63 g.). The materials eluted 239 mp ( e 16,200); vmzx 3509, 3413, 1736, 1695, 1672, 1629 with 75yoether in benzene, ether and 5% acetone in ether, and 1250 cm.-'. were combined and crystallized from ethyl acetate-petri)leum ether to give 9or-fluoro-llB-h~droxy-16a,l7a-isoproAnal. Calcd. for C23H39FO~ (420.46): C, 6.5.70; H,6.95; F, 4.52. Found: C, 65.69; H, 7.17; F , 4.86. pylidenediox!--4,6-pregnadiene-3,2!)-dioneas needles (79C mg.), m.p. 288-294" dec., [ . Y ] ~ ~ +D l l S " ( c O.931) A,, The isopropylidene derivative, 9a.fluoro-1 la-hyd 281 m p ( e 23,80'1); vmax 3448, 1712, 1658, 1621, 1587, 1081. lfia,l7a - isupropylidenedioxy 1,4 -pregnadiene -3,20 1055 and lCl37 mi.-'. ( X X I j , prepared as for XV above, was obtained as needles, dnrzl. Calcd. for C:'1H31F06(418.49): C, 69.88; €1, 7.47; m.p. 307' dec., from ethyl acetate-petroleum ether: [ a ]2 5 ~ 4-102' ( c 0.975),,,,A 238 m p ( 6 ld,500); vmSx 3333, 1712, F , 4.54. Found: C, 69.06; H, 7.68; F, 4.72. 1667, 1626, 1176, and 1059 C I I I . - ~ [lit.Io1n.p. 308-310O dec., The ~n:iteri:il eluted with 10% acetone in ether was [ a I z 6 D 4-102" ( C 1.0), X,,,238 mp (e 15,500); vmar 3344, crystillized from ethyl acetate to give Sa-fluoru-1 16-hy1709, 1661, 1621, 1603, 1319, 1374, 11-1 and 1057 cni.-l]. drosy- I6a,l7a -isopropylicle~iedio~y1,4 -pregnadiene-O,20d i m e as plates (330 m g . ) , m.p. 308" dec., [ C X ] ~ +l0:3" ~D (c Anal. Calcd. for CYIHBIFOJ (418.49): C, 68.55; H, 7.47; 1.018, pyridine),,,,A 238 mu ( e 15,400). F,4.54. Found: C, 69.01; H, 7.77; F , 4 . 5 4 . The compound 9x1 was identical to that prepared from 9a-Fluoro-l16-hydroxy-l60(, 17~-isopropylidenedioxv-l,4- 9a - fluoro - 116,lGa - 170 -trihydroxy - 1,4 - pregnadiene- 3 , 2 h pregnadiene-3 ,ZO-dione (XXI) and Pa-Fluoro-1 Ip-hydroxydiorie (Ib). 16~~,17a-isopropyIidenedioxy-4,6-pregnadiene -3,20- dione (XXII).-Ethyl oxalate (4.65 g.) was added t o a solution of PEARL RIVER,N. Y. ~
[CONTRIBUTION FROM
THE
ORGANIC CHEMICAL RESE.4RCH SECTION, LEDERLELABORATORIES DIVISION,AMERICAN CYANAMID Co.]
Studies in the Synthesis of Triamcinolone. The Ethoxalylation of 4,9( 11),16-Fregnatriene-3,20-dione and 11a-Hydroxy-4,16-pregnadiene-3,20-dione BY ROBERT E. SCHAUB, GEORGER. ALLEN,JR.,
AND
MARTINJ. M'EISS
RECEIVEDJANUARY 24, 1959 However, I Preferential 21-motio-etliosal~lationof the subject compounds ( I and I X ) was essentially unsuccessful and IX could be converted into 2,21-bis-ethoxalyl derivatives ( 1 1 7 and X, respectively). Bromine treatment of IV gave the corresponding dibrornide 1' which, upon acetolysis followed by dehydrobromination. afforded 21-acetoxy-I ,4,9(11) , I C pregnatetraene-3,20-dione(VII). Bromine treatment of X gave dibromide XI which, on acetolysis followed by dehnlogenation, produced 21-acetoxy-lla-hydro~y-4,16-pregnadiene-3,20-dione( X I I I ) . Compounds VI1 and S I 1 1 previously have been converted into triamcinolone.
The important adrenocorticoid activity of BCY- syntheses for this valuable therapeutic agent. fluoro-110,lGa,17a-21-tetrahydroxy - 1,4-pregnadi- An attractive starting material for this purpose ene-3,20-dionei (Aristocort2 triamcinolone) made (3) Other investigations concerning t h e development of new synit of interest to investigate the development of other theses for triamcinolone are described in an accompanying p a ~ e r (1) S. Bernstein and co-worker?, THISJ O U R N A L , 78, 5693 (195G); 81, 4956 (1959). (2) Aristocort is t h e Lederle Laboratories Division, American Cyanamid Co.. trademark for triamcinolone.
T h e general utility of 16a,l7a-epoxy steroids for t h e synthesis of triamcinolone will be discussed in a forthcoming p u b l i ~ a t i o n . ' ~ (4) (a) I\'. S. Allen, S. Bernstein, L. I. Feldman and M . J. Weiss, paper in PreDaration; (b) G. R. Allen, Jr., and lf. J. Weiss, Trrrs JOURNAL, 81, 4968 (1959).
. ~ ~
Sept. 20, 1959
SYNTHESIS O F
TRIAMCINOLONE
4963
was 4,9(11),16-pregnatriene-3,20-dione(I) which the 21-acetoxypregnatetraene (VII) was atcan be obtained readilysa from the available and tempted by the sequence: (1) acetolysis, (2) broniirelatively cheap 16cu,17cr-epoxyproge~terone~~; tri- nation, (3) deacylation and (4) dehydrobrominaene I required only the introduction of a 21- tion. This sequence was essentially unsuccessful, acetoxy group t o give an intermediate (111)6 and the several intermediates could be obtained only which already had been converted into triamcino- in a crude state. However, treatment of IV with two molar equivalone.' Introduction of this group was undertaken ' folcia the ethoxalylation procedure first reported lents of bromine in methanol solution a t 0 lowed by de-ethoxalylation, gave a sirup which, for a 21-acetoxylation by Ruschig.? Thus, Compound I was treated in benzene solu- after chromatography, afforded the crystalline tion with 1.7 molar equivalents of ethyl oxalate and 2.21-dibromo derivative V16 in 37y0 yield. This also was obtained in a similar manner 1.1 molar equivalents of sodium m e t h o ~ i d e . * ~dibromide ~ The crude sodium enolate, which was presumed to directly from the crude sodium salt of IV. Atbe the sodium salt of 11, was then successively tempts to dehydrobrominate V to give a l-dehydrotreated, without purification of the resulting inter- 21-halo derivative by treatment with lithium mediates, with iodine, sodium methoxide and chloride in diiriethylformamide solution" or with y sodium acetate to give the desired 21-acetoxy- collidine gave unsatisfactory results. However, a triene 111. However, this product was obtained preferential displacement of the C-21 bromine in very poor over-all yield (lOyO) and then only proved possible. Thus, reaction of dibroniide V after arduous purification by partition chromatog- with excess potassium acetate in acetone a t room temperature for three days gave an oily mixture raphy. Combustion analysis of the amorphous free which was resolved by Chromatography into startethoxalyl derivative, obtained by acidification of an ing material V (2970) and the desired 21-acetoxyaqueous solution of the crude sodium salt, indi- 2a-bromopregnatriene (VI) (51y0conversion, 7370 cated that the product contained substantial based on unrecovered V). A six-day potassiuni amounts of the bis-ethoxalyl derivative. The posi- acetate treatment resulted in a 5% recovery of V, tions of the ethoxalyl groups were presumed to be but only a 45% conversion yield of VI. Finally, C-21 and C-2 (structure IV).l0 I n contrast, it is dehydrobromination of the 21-acetoxy-2a-bromo interesting t o note that, under similar ethoxalyla- derivative VI with y-collidine afforded the desired tion conditions (1.1 molar equivalents of ethyl 21-acetoxypregnatetraene (VII) in 46% yield. oxalate) 11-ketoprogesterones is reported to undergo This product was identical with the material preferential mono-ethoxalylation a t C-21 in good which already had been obtained18 by selenium yield.'* The amorphous bis-ethoxalylpregnatriene dioxide dehydrogenation of the 2 1-acetoxytriene IV could be obtained consistently in 90% yield on (111). treatment of triene I with 2.2 molar equivalents of The conversion of 2 1-acetoxypregnatetraetie sodium inethoxide and 3.4 to 5.0 molar equiva- VI1 into triamcinolone requires the introduction of lents of ethyl oxalate.'' Since an ethoxalyl group a lGa,l7a-diol system and the elaboration of the a t C-2 was potentially convertible to a C1-C2 double ring C fluorohydrin. The diol introduction has bond, via a 2-bromo derivative," it was reasonable been achieved by osmium tetroxide treatment of to expect that IV could be converted into the VI1 to give, after acetylation, 16a,21-diacetoxy21-acetoxy-1,4,9(11).16- pregnatetraene - 3,20 -dione 17a-hydroxy-1,4,9(11)-pregnatriene-3,2O-dion e (V(VII)13 from which the synthesis of triamcinolone III).1313 AS an alternative to the osniium tetroxide already had been established.l,l3 Therefore, the procedure, the diol system was introduced by a transformation of IV into VI1 was investigated. potassium permanganate hydroxylation. This was Reaction of the bis-ethoxalylpregnatriene IV effected by briefly (3 minutes) treating 21-acetoxywith one molar equivalent of iodine followed by pregnatetraene VI1 with potassium permangamethoxide treatment gave an amorphous product nate in acetone containing. acetic acid to eive a 45% which was presumed to be the 21-iodo-2-ethoxalyl yield of the 16a,l7a-diol."which on acetyyatioii gave d e r i v a t i ~ e . ' ~Conversion of this substance into VI11 in 71% yield.lg ( 5 ) ( a ) S. Bernstein, J , J . Brown, L. I. Feldman and N . E . Rigler, (16) Probably t h e 2 - a epimer. Inasmuch a s t h e mechanism of formTHISJOURNAL, 81, 4956 (1959); (b) P. L. Julian, E. W. Meyer and ation of this compound probably involves a n intermediate carbanion 1. Ryden, ibid., 72, 367 (1950). for the de-ethoxalylation step, it is reasonable t o postulate t h a t t h e ( 6 ) W.S. Allen and S. Rernstein, dbid., 77, 1028 (1955). bromine a t o m a t C-2 has sufficient opportunity t o assume the more (7) H. Ruschig, B e y . , 88, 878 (1955). stable equatorial (2-a) configuration see ref. 4b). However, i t may (8) J. A. Hogg and A . H. N a t h a n , U. S. Patent 2,683,724 (1954). be noted t h a t a comparison of t h e infrared spectra of dibromide V (9) J. A. Hogg and co-workers, THISJOURNAL,77, 4436 (1955). and of monobromide VI with t h a t of deoxytriene I and triene 111, re(IO) Bis-ethoxalylation of 11-ketoprogesterone gives the 2,21spectively, did not clearly show t h e shift in t h e position of t h e 8-cardisubstituted product, 1 1 bony1 band usually resulting from t h e introduction of a Pa-bromo sub(11) J. A. Hogg and co-workers, THISJOURNAL, 77, 4438 (1955). stituent [M. Fieser, M. A. Romero and L. F. Fieser, THISJOURNAL. (12) Selective ethoxalylation a t C-21 also has been reported for 77, 3305 (1955); E. G. Cummins and J. E. Page, J. C h r m . Scc., 3847 (1957)]. On t h e other hand, V and V I did not exhibit t h e batbo1 l B - and 11a-hydroxyprogesteroneo and 16-dehydroprogesterone [A. H.N a t h a n and J. A. Hogg, U. S . Patent 2,719,855 (1955)l. chromic shift associated with a 2@-bromo-A4-3-ketone[B. Ellis and V . (13) L. L. Smith a n d H . Mendelsohn of t h e Chemical Process ImPetrow, ibid., 1179 (195G)l. provement Department of these laboratories, unpublished work. (17) R . P. Holysz, THISJOURNAL, 76, 4132 (1963). (14) It has been reported t h a t a 21-ethoxalyl derivative will react (18) L. L. Smith and S. Fox of t h e Chemical Process Improvement preferentially with iodine in the presence of a 2-ethoxalyl group.15 Department of these laboratories, unpublished work. However, it should be noted t h a t 2-ethoxalyl-l6a,17a-isopropylidene- (19) Potassium oermanganate oxidation of a 10,17-double bond t o dioxy-4,9(1l)-pregnadiene-3,2O-dione reacts with 84-98% of one molar give 16a,17a-diol h a s been reported by Petrow and co-workers ( J . equivalent of iodine (see footnote 21 in ref. 4b) a s determined by titraChem. Soc., 4373 (1955)l. T h e particular procedure used in our inI ion. vestigation was developed b y L. L. Smith and M . Marx of t h e Chemical (15) A. H . Nathan a n d J. A. Hogg, U.S. P a t e n t 2,730,537 (1956). Process Improvement Department of these laboratories.
ROBERTE. SCHALJB, GEORGEK.
4964
; ~ L L E N , JR., AXD
MARTINJ. Wmss
00
t-
0
s1
00
CH,CCOEt II II I
CII,
CH2CCOEt I/ I! I
EtOI!&
Vol.
I
I
- &o-
o&
o&=o-+
CI1,O.k
11.
j. CHZBI
CH,OAc
I
CH,OAc
CH~OAC
I
I
I
c=o
-+ /
\'I11
VI 00 YHzCCOEt I/ II
C;HZOAc I
H0.J
I
co €
3
0
.
4
Another convenient starting material for this gum which was considered to contain a substantial investigation was 11a-hydroxy--3,16-pregiiadiene-amount of the corresponding 2a,21-dibromide 3,20-dione ( I X ) ?O which is also readily available'Ib (XI).I6 Crude XI then was converted into the from 16a,1Sa-epoxyprogesterone. Conversion of desired 21-acetoxytriene I11 cia the known 21IX into 21-acetoxytriene I11 was accomplished by acetoxy-lla -hydroxy-4,1G-pregnadieiie-3,%O-dione 21-acetoxylation followed by introduction of the (XIII)4ain the described manner. Treatment of XI C 9 - c ~double bond. Attempted 21-acetoxylation with sodium iodide and then potassium acetate via preferential 21-mono-ethoxalylation of IX was gave an amorphous product which was assumed to apparently unsuccessful. Thus, treatment of IX be a mixture of the 2-iodo- and 2-broiiio-21-acetoxy by the reported mono-ethoxalylatioii conditions8 derivatives (XII). U'ithout purification, this magave a crude product, the combustion values for terial was dehalogenated with chromous chloridez3 which were intermediate between those for a mono- to give a crude preparation of XIII,'a which was and those for a bis-ethoxalyl derivative. Alkaline converted into its crystalline 1la-tosylate, a pre~~ compound. Treatment of this hydrolysis of this product did not afford a crystal- v i o ~ s l yprepared line glyoxylic acid8 and titration with standard tosylate with sodium acetate in refluxing acetic iodine solution resulted in an uptake corresponding acid afforded '21-acetoxytriene 111 as previously to 55% of one molar equivalent.14 Although not described,4a The over-nil yield of 111 from IX conclusive by any n e a n s , these data led to an by this procedure was lSyo. inference t h a t the above preparation was a mixAcknowledgment.-SITe are indebted to Dr. S. ture possibly consisting of 21-ethoxalyl, 2ethoxalyl Bernstein for suggesting the potential utility of and 2,21-bis-ethoxalyl derivatives of IX.zl This progesterone oxide as a starting material for the conclusion was consistent with our observations synthesis of triamcinolone. We are grateful to concerning the ethoxalylation of triene I . Drs. N. Bohonos, L. Feldman and P. Shu and Mr. As an alternative, IX was treated under the bis- C. Pidacks of the Biochemical Research Section for ethoxalylation conditions described above furnish- generous supplies of 1 la-hydroxy-l6a, lia-epoxying in yield aii airiorphous product which progesterone. l y e wish t o thank \Ir. 11.. S. was presumed t,o be crutlc 2 , 2 1 -his-cthoxalyl-1 1a - ,Allcii for technical :Ldvicc, Mr. alntlioriy Pellicario h!-tlroxy-4, I C,-r)regii~tclieiie-:3,"(I-dione(X). This o i l of the Preparations Laboratory for the 1:irge scale i'cactioii with two molxr cc!uivalcnts of broiniiie, prepration of ccrtaiii iiitcriiiediates ant1 l l r . C . follonctl by de-ethos~Ll!-lntioiitrc;ttiiient, aff(.)rtled;i Pitlacks aiid staiT for the partition chro:iia.tog( 2 ( J j 13. J Xlagcrlein, 1) .\. I . y t t l e nil I < . 11. l,cvit,, J . O i f i ~ l !, ~ raphy ~ ~ ! rcportctl ill this paper. Microanalyses were ~
20, 1709 (19%). 121) I t has since been shown t h a t under t h e same conditions proxesterone reacts w i t h ethyl oxalate t o give a similar mixture of ethoxalyl derivatives.22 ( 2 2 ) G. R. Allen, J r . , and h l . J. Weiss, unpublished results.
(33) C. Rosenkranz and co-workers, i b i d . , 72, 1077 (1'3303. (21) This compound w a s first prepared b y V. Origoni and S. J. Fox of t h e Chemical Process Improvement Department of these laboratories.
SYNTHESIS OF TRIAMCINOLONE
Sept. 20, 1959
done by Mr. L. Brancone and staff, and the spectroscoDic and Dolarimetric determinations were done by Mr. W. F'ulmor and staff.
Experimentalz6 Attempted Preparation of Sodium Salt of 21-Ethoxalyl4,9( 11),16-pregnatriene-3,2O-dione(II).-A solution of 6.6 cc. of 1 N methanolic sodium methoxide (1.1 moles) in 40 cc. of anhydrous reagent benzene was freed from methanol by azeotropic distillation, the temperature of the distillate rising from 58 to 80". To the sludge was added 1.37 cc. of redistilled ethyl oxalate (1.7 moles) and the resulting solution was added to a stirred solution of 1.85 g. of 4,9(11),16pregnatriene-3,20-dione (1)s in 35 cc. of anhydrous reagent benzene. I n a few minutes the solution became turbid and a n amorphous solid separated. The mixture, protected from moisture, was stirred for 2 hours; then 35 cc. of ether was added and stirring was continued for 1 hour. After a further 135 cc. of the anhydrous ether was added, stirring was continued for another hour and the mixture was filtered t o yield 1.63 g . of yellow amorphous powder. T h e compound gave a deep red color with 1% alcoholic ferric chloride solution. In another experiment, the yield was 2.1 g. Conversion to the free enolate was achieved from a solution of 250 mg. of the sodium enolate in 10 cc. of water by the addition of 5 cc. of 10% hydrochloric acid. This gave 166 mg. of a n amorphous solid, the combustion values for w-hich were between those calculated for a mono-ethoxalyl derivative I1 and those calculated for a bis-ethoxalyl derivative IV. Anal. Calcd. For C2&408 (bis): C , 68.22; H, 6.71; for C25H3205 (mono): C, 72.79; H , 7.82. Found: C, 69.12; H , 6.96. Attempted Preparation of 21-Iodo-4,9( 11),16-pregnatriene-3,20-dione.-To a stirred solution of 1.39 g. of the sodium salts of the above prepared mixture of ethoxalyl derivatives of 4,9( 11),16-pregnatriene-3,20-dionein 25 cc. of absolute methanol cooled to -15 to -20" was added dropwise, during 1 hour, a n absolute methanol solution (25 cc.) containing 935 mg. of iodine. After stirring for a n additional hour, 3.95 cc. of 1 N methanolic sodium methoxide was added dropwise and stirring was continued a t 0" for another hour. After the dropwise addition of 100 cc. of water, which caused a solid to separate, the mixture was kept at 0" for 24 hours. The buff-colored amorphous solid was collected on a filter and washed with water; yield 1.07 g. of crude iodo derivative. 21-Acetoxy-4,9( 11), 16-pregnatriene-3,ZO-dione(111). A. From Crude 21-Iodo-4,9( 11),16-pregnatriene-3,2O-dione.A mixture of 300 mg. of the preceding crude preparation, which presumably contained some 21-iodo-4,9( 11),16-pregnatriene-3,20-dione, 15 cc. of reagent acetone and a sludge prepared (in a mortar) from 1.6 g. of potassium bicarbona t e and 0.96 cc. of glacial acetic acid was refluxed for 12 hours. After filtration, the solution was clarified with Norit activated carbon and evaporated t o dryness leaving 266 mg. of a glass. Extraction of the residue with hot benzene and evaporation as before afforded 150 mg. of a glass which gave a positive blue tetrazolium test. This glass was subjected to partition chromatography on Celite26 in the following manner. The mobile phase consisted of three parts ethyl acetate and 2 parts petroleum ether ( b . p . 90-10O0), whereas the stationary phase consisted of three parts methanol and two parts water. Both phases were equilibrated with each other. The solid was dissolved in the stationary phase, slurried with Celite and the mixture added to the column. Chromatography mas then initiated with the mobile phase and 10 fractions of 20 ml. volume each were collected. Frac(25) Melting points were determined in a capillary tube and are uncorrected. All infrared spectra were determined in potassium bromide disks. Ultraviolet spectra were determined on a Cary recording svectrophotometer and t h e infrared spectra were determined on a Perkin-Elmer spectrophotometer (model 21). T h e petroleum ether used was t h a t fraction boiling a t 60-70' unless otherwise specified. All concentrations were carried o u t under reduced pressure on t h e steam-bath. (26) T h e adsorbent was specially treated Celite "545" which had been washed with 0 A ' hydrochloric acid, water, and finally with 3 A alcohol, and then dried a t looo. Celite is t h e trademark of Johnshlanville C o . for diatomaceous silica products.
4965
tions 4, 5 and 6 gave a single spot at 18.5 ctn. on paper chromatographic analysis.27 Fractions 4, 5 and 6 on evaporation to dryness gave 90 mg. of a glass which could not be crystallized and was, therefore, rechromatographed as above, except t h a t 5 parts of petroleum ether instead of 2 parts were used. Fraction 4 from the second column gave a single spot at 18.5 cm. and on evaporztion afforded 43.3 mg. of product as a gum which crystallized from acetonepetroleum ether. Recrystallization from the same solvent pair gave material of m.p. 123-125". Admixture of this material with authentic 21-acetoxy-4,9( l l ) , 16-pregnatriene-3,20-dione6 did not depress the melting point. The infrared spectra for the two samples were identical; umax 2920, 1760, 1680, 1440, 1380, 1240, 1220 cm.-l. B. From 21-Acetoxy-l1~u-p-toluenesulfonyloxy-4,16-pregnadiene-3,20-dione.-Using the procedure described previouslj-,4a a solution of 1.500 g. of 21-acetoxy-1 la-p-toluenesulfonyloxy-4,16-pregnadiene-3,20-dione(prepared as described below b y tosylation of X I I I ) , 1.500 g. of sodium acetate and 35 ml. of glacial acetic acid was allowed to reflux for three hours to give, after recryst-dllization from acetonepetroleum ether, 0.540 g. (55% yield) of long needles, m.p. A mixture of this material with an authentic 130-131'. sample6 of the triene melted at 129-131 . The material had [ c x ] ~ ~ +l89" D (c 2.04, chloroform) and X",'B.."" 239 mp ( e 26,500). Reported6 values are [Lu]''D +166" (chloroform) and A:tgH 239 mfi ( e 24,000). Moreover, the infrared spectra of this preparation and of a n authentic sample were identical. Sodium Salt of 2,21-Bis-ethoxalyl-4,9( 11),16-pregnatriene-3,20-dione (IV).--A solution containing 144 cc. of 1 N methanolic sodium methoxide (2.2 moles) and 450 cc. of anhydrous reagent benzene was freed from methanol by azeotropic distillation, the temperature of the distillate rising from 58 t o 80". T o the sludge was added 48 cc. ( 6 moles) of ethyl oxalate and the resulting solution was added to a stirred solution of 4,9( 11),16-pregnatriene-3,20-dione6 ( I ) (20 g.) in 400 cc. of anhydrous reagent benzene. I n a few, minutes the solution became turbid and a n amorphous solid separated. The mixture, protected from moisture, was stirred for 2 hours, 600 cc. of ether then was added and stirring was continued for an additional hour. The yellow solid was filtered; yield 39.2 g. (109%). 2,2 1-Bis-ethoxalyl-4,9( 11), 16-pregnatriene-3,20-dione (IV).--A solution of 39.2 g. of sodium salt of 2,21-bis-ethoxaly1-4,9( 11), 16-pregnatriene-3,20-dione,prepared above, in 400 cc. of water was clarified by filtration, and then acidified with 100 cc. of 5% aqueous hydrochloric acid. T h e precipitated amorphous solid was collected and washed well with water; yield 29.9 g. (goyo from triene I ) . In a pilot 242 mv run, the yield of product was 288 mg. (74%)H: : A: ( e 16,900) and 315 my ( e 13,300), A::,""."' 249 mp ( ~ 2 0 , 0 0 0 ) and 315 mb ( E 10,400); vmax 2940, 1740, 1140 and 1270 cm.-'. Anal. Calcd. for C29Ha408: C, 68.22; H,6.71. Found: C, 68.31; H , 6.71. 2-Ethoxalyl-2 1-iodo-4,9( 11), 16-pregnatriene-3,20-dione. -To a stirred solution of 564 mg. of 2,21-bis-ethoxalyl-4,9(11),16-pregnatriene-3,20-dione (11;) and 2 g. of anhydrous sodium acetate in 30 cc. of reagent methanol, cooled to - 15' in a n acetone-Dry Ice-bath, was added, dropwise, a solution of 9.33 cc. of absolute methanol containing 280 mg. of iodine. Stirring was continued for 3 hours. After this period a n aliquot, titrated with 0.01 AT sodium thiosulfate in the presence of starch indicator, showed a negligible amount of free iodine present in the solution. After the addition of 2.26 cc. of 1 N methanolic sodium methoxide, stirring was continued at 0' for 1 hour and then the solution was neutralized with dilute acetic acid. The dropwise addition of 135 cc. of water precipitated an amorphous solid which was collected on the filtcr and washed well with water; !+Id 526 mg. (88Yc),XP;!" 242 nip ( e 17,700), NnoH 228 m p ( E 21,6(10) m d A,, 340 iiip ( e 7800), ::A: IiC' 2-48 m p (E 17,700); vmn, ?%o, 1740, 1650, 1600, 1440 and 1280 cm.-'. This solid W:LS prcsunicd 127) This method was essentially t h a t of I. E . Bush, Biiichrm J., SO, 370 (1952). T h e chromatography was carried o u t on untreated Whatman No. 1 paper with a toluenepetroleum ether (h.p. 913.100")methanol-water ( 9 : G:9.75:5 . 2 5 ) solvent system, and was allowed t o run a t 27' for two hours. T h e steroids were detected with a n alkaline blue tetrazoleum spray.
4966
ROBERTE SCIIAUB, GEORGEI-infrared and mixed m.p. stir fnr one hour. The mixture was filtered to givc 0.37116. T h e product then was eluted n-ith 3 1. of lo:& ether-benzene and was crystallized from ether ti, give 6.4 g. (51% or Y:$yobased on unrecovered I-)of 21-:icetusy-2o/-broni(i-4,~( 11), 16-prcgnatriene-3 ,2(]-diniie (1-1) , ! s i m .p. 168-172' der. 12ecrystalliz:ition from ncctclrie-petrnleuni ctlier affnrded white crystxlq, m . p . 171-17~5° cicc., [ ~ I Z ~ +188" D (1Yc iti CHCla), A$,:" 250 1 1 1 , ~ ( E 24,300); vmag 1760, 16!10, 1640, water, and tlie slightly turbid solution was filtered through 1625, 1595, 1220 cm.-I. n bcd o f Celite. The cleltr yellow filtrate was treated with Anal. Calcd. for C23€12104Br:C , 61.77; 11, 6.09; Br, 2 nil. of 5c& hydrochloric acid solution, and the precipitated 17.87. F o ~ n d : C, 61.98; H, 6.34; Br, 17.91. solid was collected by filtration and wxh.ed with water to 2 1-Acetoxy-l,4,9(11),16-pregnatetraene-3 ,20-dione (VII). give 0.2?3 g. i6t?cpj-idd) nf ivory-colnrcd solid. The -.I mixture of 2 g. of 21-:rcetox~--2~~-l~ru1no-4,R(I 11.10- in:iteriil liad hE;fl' L"i2 n i p ( e 18,300). 313 111p ( E 6.5001; i~~~~ 2.18 m,u ( E 20,000).305 nip ( 6 4.380 ") :;;:A 2.18 111p ( e pregnatriene-3,20-dione (TI)and 10 cc. of r-collidine was - stirred a t 155' in an oil-bath for 2.5 hours. sulution being complete on warming. The cooled mixturz was diluted with ( 2 8 ) T h e chromtltographg was carried o u t on untreated \\'hatman No. 1 paper with a rxtroleum ether (h.p. 90-10O3)-methano1-watrr 100 cc. rif e t h e r and filtercd froin collidine hydrobromide ( 1 0 ' 8 : Z ) solvent srstein and w a s allowed to run a t 27' for t w o hour, g.'. The filtrate. w:ishecl with %yo sulfuric acid and (1.06 T h e steroids were dctrcted with a blue tetrazo!ium s p r a y . thrii with w:iter, w:ts dried with magncsiuiii sulfate :iiid ev:ipt o be 2-e thoxalyl-2 l-iodo-4,9( 11), IF-pregiiatriene-3,20-di-
cine.14
Sept. 20, 1959
SYNTHESIS OF TRIAMCINOLONE
4967
250 mp ( e 18,100) and 335 mp ( e 20,100); a vmSx 3450, 19,300), 335 mM ( e 8900)29; YE:: 3430, 1740, 1665, 1630 ~",.,p 1740, 1660 (inflection), 1627, 1264(broad) c ~ n . - l . ~ ~ and 1225 cm.+ (broad). Anal. Calcd. for C25H310~(monoethoxalyl): C, 70.07; A n d . Calcd. for CngHasOg: C, 65.89; H, 6.87. Found: H, 7.53. Found. C, 67.18; H, 7.17. C, 63.48, 63.53; H, 6.78,6.93. This solid (250 mg., 0.5 mniole) was suspended in 15 nil. 23a,2 1-Dibromo-1la-hydroxy4,16-pregnadiene-3,20-diof benzene and 0.5 ml. of 1 N methanolic sodium metlioxide one (XI).'"-% solution o f 4.582 g. (8.7 mmoles) of 2,21was added. A11 solid dissolved immediately, and within one ethoxalgl-1 l a - h y d r o x y l , IR-pregnadiene-3,20-dione (X), minute solid precipitated. The mixture was diluted with 3.500 g. (34.8 mmoles) of potassium acetate and 100 ml. 10 ml. of ether and filtered to give 0.200 g. (76% yield) of of methanol was chilled iti a n ice-bath with continuous stirthe sodio derivative. This material gave a clear yellow soluring. T o this solution there was added over a period of one tion with mater; this solution gave a red color when treated hour 30 ml. of 0.58 hf bromine in carbon tetrachloride with a n alcoholic ferric chloride solution. solution, The resulting yellow solution now was treated Anal. Calcd. for Cn&lOGNa (monoethoxalyl): C, with 17.5 nil. of 1 N methanolic sodium methoxide. Some 65.64; H , 6.94: Na, 5.11. Found: C,62.18; H,6.69; N a . solid separated at this time. The mixture was distributed between 200 ml. of methylene chloride and 200 ml. of water. 5.93. The organic solution was separated, washed with saturated Attempted Preparation of 1la-Hydroxy-3,20-dione-4,16- saline solution, dried over magnesium sulfate and taken to pregnadiene-21-glyoxalic Acid.-'\ solution of 0.425 g. ( 1 .0 dryness to give 4.126 g. of gummy residue. This material mmole) of 21-ethoxalgl-1 la-hydroxy-3,16-pregnadiene-3,20- gave two blue tetrazolium-pl)si'i~espots at 14.5 and 1'3.0 dione and 2 nil. of 1 ,V potassium hydroxide in methanol was cm. on paper ~ l i r o i n a t o g r a p l iand ~ ~ ~i t had A%:" 244 m p allowed to stand at room temperature for 45 minutes. The ( e 14,400); vmSx 3480, 1750, 1677, 1627, 1582, 1241 and solution was diluted with 2 ml. of water and acidified with 1210 em.-'. The bands a t 1750, 1241 and 1210 em.-' 5% hydrochloric acid solution. The precipitated amorsuggested that this crude material contained some etlioxalJ.1phous solid was collected by filtration t o give 0.386 g. (97% substituted steroid. yield). This material gave a red color with alcoholic ferric chloride solution. The material had XE'"242 mu ( e 17.100). Anal. Calcd. for C?lH&r?Oa: C, ,5137; 13, 5.39; R r , 315 nip ( e 6000); ::A: 248 mp ( e 18,%), 305 ki, ( 6 4200); 32.94. Found: C , 49.87; H, 5.39; Br,27.59. A:$" 248 mp ( E 18,700), 333 inp (E 8600); Y",: 343(J, 1740, I n other preparations of this material, attempts to crystal1665, 1630 and 1225 cni.-'(broad). lize it or to purify i t by chromatography on silica gel were Anal. Calcd. for C2sH,sOs (monoethoxalyl): C, 68.98; unsuccessful. H, 7.05. Found: C, 66.8'3; H, 7.36. 2 1-Acetoxy-1 la-hydroxy-4,16-pregnadiene-3 ,20-dione (XII1j.--A solution of 1.85 g. (0.012 mole) of sodium iodide Attempted Preparation of 1la-Hydroxy-2 l-iodo-4,16-pregnadiene-d,ZO-dione.--A solution of 0.450 g. (1.0 mmole) in 100 ml. of acetone was added to 4.00 g. (8.2 mmoles) of of the presumed sodio salt of 21-ethoxalyl-lla-hydroxy-4,16- 2a,21-dibromo - I la- hydroxy -4,16- pregnadiene-3,20 -dione pregnadiene-3,20-dione in 5 ml. of methanol was chilled to (XI). A solid began precipitating immediately; the mixture was warmed on the steam-bath for 15 minutes and filtered. -15' during magnetic stirring. A solution of 0.134 M iodine in methanol then was added dropwise until iodine The filtrate was allowed to reflux with roiit i*iuous stirring for 18 hours with a n intimately ground mixture of 21 g. was no longer absorbed; this required 4.10 ml. (0.55 mmole). of potassium bicarbonate and 12.5 ml. of glacial acetic acid. Iodine then was added up to a total of 4.75 i d . of the soluThe mixture was distributed between 200 ml. of mater and tion. The solution was stirred for one hour, but the excess iodine had not been consumed. The excess iodine then was 300 ml. of methylene chloride. The organic layer was separated, washed with saturated sodium thiosulfate solution destroyed by titration with 1 N methanolic sodium methox(250 ml.) and water ( 2 X 250 ml.), dried over magnesium ide; 0.65 ml. of this solution was required. An additional sulfate and taken to dryness to give 2.92 g. of a glass. 0.55 ml. of 1 iV methanolic sodium methoxide was put in This material was dissolved in 50 ml. of glacial acetic acid, and the solution stirred at 0 ' for one hour. Water (30 ml.) and the solution was swept with carbon dioxide. While now was added dropwise over one hour and the mixture treated with 2 g. of sodium chloride and stirred at 0" for under carbon dioxide atmosphere, the solution was treated wit.h 50 ml. of approximately 0.4 N chromous chloride 2.5 hours. The misture was filtered to give 0.262 g. of solution during 15 minutes. The green solution was disamorphous solid. It was not possible to crystallize this matributed between 100 ml. of water and 200 ml. of methylene terial nor to obtain satisfactory analyses. 2 , 21-Bis-ethoxa I:+ 11a-hydroxy-4,16-pregnadiene-3,20- chloride. The organic layer was separated, washed with water (2 X 100 ml.), saturated sodium bicarbonate solution dione (X).-A solution o f 21 ml. of 1 N rneihanolic sodium (2 X 100 ml.) and again with water (2 x 100 ml.), and taken methoxide in 100 ml. of benzene was distilled until 62 ml. of to dryness. The residue was dissolved in 100 ml. of ethyl distillate was collected. The cooled residual mixture was treated with 0.500 g. (0.034 mole, 4.6 ml.) of ethyl oxalate acetate, dried over magnesium sulfate and taken to dryness to give 1 .*56g, of a glass which had ";A: 240 mfi (E 20,900) during magnetic stirring; all solid dissolved immediately. and an infrared spectrum which was identical with that of a n A slurry of 3.28 g. (0.010 mole) of lla-hydroxy-4,16-pregauthentic sample.4a nadiene-3,~n-dioiie ( I S ) in 50 rnl. of dry benzene was added, a n additional 10 ml. of benzene being used to aid Anal. Cnlcd. for C23H3006: C, 71.48; H, 7.82. Found: in the transfer. All solid dissolved to give a yellow solution C, 68.86, 68.85; H, 8.30, 5.42. from which a n amorphous solid began separating within one minute. Stirring was continued during 24 hours. The 2 1-Acetoxy-1 la-p-toluenesulfonyloxy4,16-pregnadie nemixture then was diluted with 100 ml. of dry ether, and 3,20-di0ne.~~?24-A solution of 1.500 g. (4.65 nimolrs) o f stirring vas continued for a n additional hour. The solid 2 1-acetoxy-1 la-hydroxy-4,16-pregnadiene-3,2O-dic~ne ( S II I ) was collected by filtration and washed with 50 ml. of ether and 0.950 g. (5.00 mmoles) of p-toluenesulfonyl chloride in to give 5.73 g. ( 100yc yield) of the disodium salt of 2,2117 ml. of pyridine was allowed t o stand at 5" for 18 11ours. bis-ethox~lq.l-lla-hydroxy-4,16-pregnadiene-3,20-dione (X) the solution then was distributed between 101) ml. of water as crude amorphous yellow solid. and 100 ml. of methylene chloride. The organic solutio11 was washed with three 50-ml. portions of 57Ghydrochloric Anal. Calcd. fur C29H34085a2: C, 60.83; H , 5.97; Na, acid solution and two 50-ml. portions of water, dried o v w 8.03. Found: C, 66.45, 56.06; H, 6.77, 6.35; "a, 7.68. magnesium sulfate and taken to dryness to give 2.17 g. c i f This solid was dissolved in 250 ml. of water to give a clear residue. This material was dissolved in the minimum qualisolution which was acidified with a 5y0 hydrochloric acid tity of benzene and adsorbed onto 50 g. of silica gel (coluinn solution. The amorphous solid which precipitated was colsize: 2.8 X 20 cm.). The column was washed with 500 ml. lected by filtration and dried under reduced pressure over of a 5% ether-in-benzene solution, and these washings were phosphorus pentoxide to give 4.73 g. (90% yield). The discarded. Elution with 3250 mi. of a 20y0 ether-in-benmaterial had"::A: 244 mp ( e 15,600) and 313 mp ( e 12,900), zene solution gave a small amount of glass which was discarded. The column then was washed with a 50% ether-in(29) For t h e acid and base spectra, a methanolic solution was diluted benzene solution; 250-1111. fractions were collected. The 1:l with 0.1 iV hydrochloric acid and 0.1 AT sodium hydroxide, respecmaterial eluted in fractions 2-12 was combined and n'tively. crystallized from acetonepetroleum ether to give 1.7811 g .
1068
GEORGEli. ALLEN,JR.,
AND
MARTINJ. ~ V E I S S
Vol. b l
(77Ycyield) o f hard crystals, n1.p. 158-160' dec. alone nr in Crilcd. for CaoHs&S: C, 66.64; H, 6.71; S, irii.cture with 2n authentic sample.4a The infrared spectra 5.93. Found: C, 66.75; H, 6.89; S, 6.18. for the two samples were identical. The material had Conversion of this compound to pregnatriene III is de[ n ] 2 4+~7 I o (c 1.06, chloroform)andhz:.",0H229mo (€30,800). Iceported 4a values are m.p. 158-159", [ a ] " D +69" (chloro- scribed above. timn) and"::A: 231 r n M ( e 31,400). PEARL RIVER,N.Y.
[CONTRIBUTION FROX THE ORGANIC CHEMICAL RESEARCH SECTION,
LEDERLE LABORATORIES DIVISION, AMERICAK
CYANAMID Co.]
Studies in the Synthesis of Triamcinolone. The Condensation of 16a,17aIsopropylidenedioxy-4,9( 11)-pregnadiene-3,ZO-dione with Ethyl Oxalate BY GEORGER. ALLEN,JR., AKD MARTIN J. L ~ ~ E I S S ' RECEIVED JAXUARY 24, 1969 Etlioxalglatiun of 16a,lT~-isopropylidenedioxy-4,9( 1l)-pregnadiene-3,23-dione (XI), which was prepared from 16a,17aepoxy-1lor-hydroxyprogesterone (11), in the presence of a molar equivalent of sodiurn methoxitie gave exclusively the 2ethoxalyl derivative XVIII, which was converted by a six-step procedure into 9~-fluoro-ll,8,16or,l7a-trihydrosy-1,4-pregnadiene-3,20-dione (XXL'II). The Sa-chloro analog of XXVII also was prcpved. Bis-ethoxalylation of XI led to t h e preparation of the 2a,X-dibromide XXII. Attempts to convert XXII into 16a,21-diacet,xy-17a-h~r~roxy-4,9( 11)-pregnadiene-3,20-dione were unsuccessful.
The importance of Sa-fluoro-llp,l6a,lia,21tetrahydroxy-1,4-pregnadiene-3,20-dione (I) 2a (Aristocort triamcinolonezb) as a therapeutic agent3 made i t desirable t o explore other routes for the synthesis of this material4 An attractive starting point for a synthesis of I is 16a,l7a-epoxy-llahydroxyprogesterone (11) which has been obtained from 16a,17a-epoxyprogesterone by fermentation with Rhizopus nigricans.6 I n the present paper we wish t o describe the conversion of I1 into 1 6 a j l 7 a- dihydroxy-4,9(11) -pregnadiene - 3,20- dione (X), and efforts to transform the latter material and certain of its derivatives into intermediates t h a t have been utilized for the preparation of 1.9 Our preparation of X was accomplished in the following manner. Reaction of 11 with meth(1) T o whom inquiries concerning this paper should be addressed. (2) (a) S. Bernstein, R. H. Lenhard, IT. S . Allen, hZ. Heller, R. Littell, S. M Stoiar, L. I. Feldman and R . H . Blank, THISJ O U R N A L , 78, 8093 (19.56); 81,in press (19593; (b) t h e trademark of American Cyanamid Co. for triamcinolone is Aristocort. (3) L. Hellman, B. Zumoff, hf. K . Schwartz, T . F. Gallagher, C. 4. Bcrntsen and R . H. Freyberg, Abstracts of papers presented a t t h e 3rd Interim Meeting of t h e American Rheumatism Association, Bethesda, M d . , P\-ovember 30, 1956. (4) T h e original method for t h e prepaiation of I utilized 21-acetoxy4,9 (1I ) ,In-pregnatriene-3 ,ZO-dionel,j as t h e key intermediate. This l.itter substance has been prepared f r o m 21-acetoxy-17or-hydroxynrogesterone (Reichstein's Substance S)8 and t h e his-ethylene ketal of cortisone acetate.6 hfore recently, Fried and his co-workers have described t h e preparation of I from Sa-Ruorohydrocortisone; this elegant procedure utilized microbiological fermentation t o introduce t h e required 160-hydroxy function.7 6 )\V, S . Allen and S. Bernstein, THIS JOURNAL,77, 1028 (1955) (6) 1%' S Allcn, S Bernstein and R . Littell, i b i d . , 76, 6110 (1934).
( 7 ) R . \V T h o m a , J . Fried, S . Bonanno and P. Grahovich, i b i d . , 79, 4818 (1957). ( 8 ) D H. Peterson, P.D. RIcister, A . Weintraub, I,. 11. Reinrkc, H. Ei>pstein, 13. C. Murray and H. h l 1, Oshorn. i b i d , , 77, 1128 19s;). (9) Other efforts t o utilize 16a,17a-eooxy.l lor-hydroxyprogesterone for the preparation of triamcinolone are t h e subject of a n accompanying paper.10 T h e general utility of 16a,l7a-epoxy steroids for t h e synth?. sis of triamcinolone will be discussed in a forthcoming publication ( W . S. Allen, S. Bernstein, I-. I. Feldman and M. J. TVeiss) ( I O ) R. E. Schaub, G. R. Allen, J r . , and M. J . Weiss, THISJ O C R N A L , 81, 4962 (1959).
anesulfonyl (mesyl) chloride gave the 11 a-niesyloxy derivative I11 from which 16a,1Ta-epoxy4,9(11)-pregnadiene-3,20-dione (IV) was prepared by elimination of the elements of niethanesulfonic acid. This procedure, which is reported in the patent literature by Bergstroin! l 1 in our hands has afforded a consistent over-all yield of 80%. The conversion of the 10a,lTa-epoxy group of IV into a 16a,17a-diol system was accomplished by using the procedure of Koino and Iiomo de Vivar.I2 Reaction of the oxide 1 1 7 with hydrogen bromide in glacial acetic acid gave the bromohydrin V in 62% yield. Acetylation of this material with acetic anhydride in the presence of p-toluenesulfonic acid afforded the enol diacetate VI1 in 95c!;8 vield. Treatment of VI1 with sodium acetate In refluxing acetic acid produced l(ia-acetoxy-17ochydroxy-4,9(11)-pregnadiene-3,2O-dione(1x1 in 57% yield. This step, wherein the l(iiI-bromo group is displaced, involves participation of t h e neighboring 17-acetoxy group, presumably ciu an intermediate ortho ester type cation." The diol acetate IX also was prepared by two variations oi the above method. Reaction of the oxide I V with sodium iodide in glacial acetic acid gave in 69 yield the iodohydrin V I i 3which was converted into IX via the enol acetate VIII. Finally, I X also was obtained from 16a.1 Toc-epoxy-] la-mesyloxyprogesterone (111). Reaction of this substance with hydrogen bromide in glacial acetic acid gave X I 1 in almost quantitative yield. The bromohydrin XI1 was converted into the enol diacetate XTII which, without isolation, was treated with sodium acetate in refluxing acetic acid to give IX in S!):. yield. Of the above three methods, the last gives the most satisfactory yield, whereas the iiic.thot1 based on the iodohydrin VI gives the p(iores1. Hydrolysis of the diol acetate i X with 0.5 .I7 hydrochloric acid gave l(ioc,lTa,dihvdrusy-4,!'~11)(11) C. G. Bergstroru, U. S. P a t e n t 2 , i O 3 , 7 ! W ( 1 2 ) T. Romo a n d A . Roino de Vivar, J . 0i.p. C h c v . , 21, $102 (1950). (l:3) T h e opening of 16a,l7a-oxides t o yield iodohydrins has heen reported by 4 . Ercoli and P. de Ruggieri [GQZZ.c h i m i t a l , 84, 479 (195411.
