2820
R.N.JONES, I',
[CONTRIBUTION FROM
THE
13TJMPIfRIES,
F.I-TERLING AND KONRAD DORRINER
SLOAN-KETTERING INSTITUTE FOR CANCER RESEARCH AND NATIONAL RESEARCH COUNCIL OF CANADA4]
THE
Vol 74
DIVISIONOF CHEMISTRY OF
THE
Studies in Steroid Metabolism. XII. The Determination of the Structure of the Side Chains of C-21 Steroids by Infrared Spectrometry BY R. NORMAN JONES, P. HUMPHRIES, F. HERLING AND KONRAD DOBRINER Infrared absorption bands associated with the C = O and C=C stretching vibrations of functional groups in the side chains of C-21 steroids are described and discussed. Examples are given illustrating the use of these spectra in the elucidation of the structure of the metabolites of adrenocortical hormones isolated froni urine, a n d of reldted steroids prepared synthetically.
The determination of the structural and stereochemical relationships among the carbonyl, ethylene and hydroxyl groups in the side chains of C-21 steroids often presents a difficult problem. This communication describes a number of correlations between the infrared absorption spectra and the side chain structure which have proved of considerable help in the identification of the side chains of newly isolated or synthesized C-21 steroids. These correlations principally involve the frequency of the absorption maxima of the carbonyl stretching bands. Experimental Methods and Results The spectra were measured on Perkin-Elmer single beam spectrometers, and the frequencies of the absorption maxima of the C=O and C=C stretching bands are listed in Table I . Many of the steroids included in this survey contained carbonyl groups and double bonds in the ring system as well as on the side chain, and the absorption maxima associated with these are italicized in the table. For the majority of the measurements a sodium chloride prism was used, the band positions were determined as described in a previous publications and the estimated accuracy is =t3 cm.-l. Representative compounds of the structural types of major interest were measured also with a calcium fluoride prism and the spectra corrected for water vapor and solvent absorption. The bands so measured are indicated in the table by an asterisk and their estimated accuracy is f l cm.-l. Solvent Effects.-When the sDectra are measured in carbon disulfide or carbon tetrachloride solution, the positions of the carbonyl stretching bands are the same. The characteristic band positions for different types of steroid carbonyl compounds have been discussed3 and summarized. Many of the more highly hydroxylated adrenocortical steroid hormones and metabolites are not sufficiently soluble i n either of these two solvents, even after acetylation. They may be studied in chloroform solution5 but unfortunately the carbonyl frequencies, although still characteristic of the individual carbonyl groups are all displaced to lower frequencies by varying amounts; the bands are broader, arid the separate carbonyl maxima in polycarbonyl compounds are less well resolvrd. The conjugated carbonyl bands i n particular tend to exhibit an asymmetry in chloroform solution which makes :I precise eraluation of the position of the maximum more dificult. I n the following discussion the iiiiportance of thew >olvciit effect> must b c kept iri I n i l l t i .
Discussion The carbonyl bands associated with simple side ( l j Published as Contribution N o . 2683 froni the l.:iboratorics of S a t i o n a l Research Council of Callad&, a n d ~ i r e s c t i t c d i i i p u t , btiore t h e Division of Biochenii*try of the Amrricsn Chrxuii,.il S w i r l y , h p i il 1 I t h , 1950, a t Philadelphia, Penria : 5 ) K. S.Jones. V. %. I Y i l l i a n ~ s 1 , 1 J I\.l!aIer! a n d K. I)ol~riilcr,'rills J U L K N A L . 70: 2024 (1948'1. 1 3 ) li. N. Jones, P. Xunlpiirirs s n d K. IlolJriiirr, rbiJ., 71, 241 (1949); 72, '366 (1950). ( 4 ) I < . Y . Jozies 31111 K. n o h r i n e r , i'i!omiirs a i d Hurrnoiies. 7 , 293 1949).
'l'lie
( 5 ) Recently methylene chloride has also been employed t o dissolve t h e more highly h y d r o x y l u t r , l ;trroids: i l IIXSt h e n(l\.antulje o f y r c i l e r cheniical stability
chain acetates and ketones are discussed first. The displacements in the carbonyl band positions associated with hydroxyl-carbonyl and carbonylcarbonyl interactions of certain types of side chain groups are next considered, and finally the effects of introducing unsaturated linkages are noted. 20- and 21-Acetates.-The absorption of the 20acetate group (I) is a t 1734-1736 ern.-' (CS2) and is in the same range as that of the 3-, 6-, 7-, 11-, 12and 17-acetates (1'734-1742 cm.-' (CS2, CCL), 1719-1722 ern.-' (CHCL)). No comparisons between 20a- and 20P-acetates have been made but it would seem unlikely that significant differences would occur. The introduction of the 17a-hydroxyl group has no effect on the carbonyl maximum (11). The data for siniple "-acetates are not available, but for structures I11 and IV a single carbonyl band is observed a t 1749 ern.-' (CS2), 1736-1739 cin.-I (CHC13). This value is significantly high, but whether it is characteristic of the 21-acetate per se, or involves interaction with substituents a t the 17 and 20 positions remains to be determined. CH3
CHs.0,CO.CHs
I
I
HC.O~CO~CH,
HC.OR
I
I, R = H 11, R = 011
111, R = H I V , R = O.CO.CH8
20-Keto-21-methyl Group.-This group (V) absorbs a t 1706-1710 cm.-' in CS2 and CC14 and a t 1698-1702 cni.-' in CHC1,. The band positions for the %ketone group are 1715-1719 cm.-' (CS2, CC14) and 1702-1705 cni.-' (CHC13), so that the 3- and 20-ketosteroids can be differentiated in carbon disulfide or carbon tetrachloride but not in chloroform s o l ~ t i o n . ~
Several types of substituents (VI-X, XII) may be introduced a t C-16, C-17 or C-21 without displacing the frequency of the 20-ketone maximum, and the hand position is also unaffected by stereochemical inversion a t C-17 (XIII, XIV). Bromination at the 17a-position leaves the band position unchanged (VII) but on bromination at (2-21 (XI) the inaxiinurn is displaced t o 1716-1722 C I I I . - ~ ( CHCla).
In the spectra of IB,17-methylene-20-ketones (Xvj the carbonyl band is at 1685 ern.-' (CC14, ( G i Tile 4-, 6-, 7-, 11- and 12-ketosteroids all absorb between 1706 aiid 1719 c m . 3 in CS? and CClr (see reference 3) nncl t h c 1 1 - a n d 12-
k e t o s t v r o i d s a t 1701-170(i c 1 n - 1 CIICI,.
June 5, 1952
STUDIES IN STEROID
282 1
METABOLISM
CS2) indicating that the cyclopropanyl group is exercising a conjugation effect on the 20-ketone similar to that of the cyclopropane ring in i-cholestanone-6 and other i-keto~teroids.~
V.RI H VI; R; = OH VII, R1 = H VIII, R1 = H
CO
XI1
Rs
XIII, R = H
xv
c=o
I
c=o.. .
XVI
J
I
XVIA
3600
H ,
CHI
I
c=o
XVII
(7) M.-L. Josien, N. Fuson and A. S. Cary. THISJ O U R N A L , 1 3 , 4445 (1951).
3400
1800
FREOUENCY
1750
'
1700
1650
( C M - ),
Fig. 1.
would appear that the bonding is weaker in the 17phydroxy-20-ketone than in the 17a-hydroxy isomer. The effect is lost on bromination a t C-21 (XI), and it is not shown by the 20-keto-21-methyl01 compounds (VI) , Fig. 3. Other Hydroxy-Carbonyl Interactions.-Other examples of carbonyl band displacements to lower frequencies, and anomalous hydroxyl absorption attributable to intramolecular hydrogen bonding have been noted in structures involving steroid side chains. One of these (Fig. 4) concerns the 17a-hydroxy group and the carbomethoxy group of the bisnorcholanic acid methyl ester side chain (XVIII). A band would normally be expected for the methyl ester at 1738-1742 em.-' (CS2, CC14) but the principal band actually occurs a t 1719 em.-' (CC14) with only a shoulder a t 1740 cm.-'. Two hydroxyl bands are present, their relative intensities not sensitive to concentration. An equilibrium with XVIIIA may account for both of these effects. OCHo
I
XVIII
CHI
I
,1708
H
Hydroxy-Carbonyl Interactions The 17a-Hydroxy-20-ketone Group.-The introduction of a 17a-hydroxy group (XVI) alters the 20-ketone carbonyl absorption in a characteristic manner. In Fig. 1 the spectrum of pregnanol30-one-20-acetate is compared with the spectrum of the 17a-hydroxy derivative in CCI4 solution. A small band is observed in the latter a t 1693 cm.-' in addition to the bands at 1735 and 1708 em.-' expected for the 3-acetate and the 20-ketone groups. In the hydroxyl stretching region the absorption of the 17a-hydroxyl group is also atypical; two bands occur, one at 3620 em.-' due to a free hydroxyl and a second band a t 3500 em.-' due to an associated hydroxyl group. The two hydroxyl bands are of about equal intensity and their relative intensities are not significantly altered by threefold dilution (Fig. 1). In dilute solution most monohydroxysteroids show a strong free hydroxyl band near 3600 em.-' and a weak second band at 3500 em.-' the relative intensity of which increases with concentration. To explain these effects an equilibrium between XVI and XVIA is proposed, intramolecular bonding between the hydroxylic hydrogen atom and the ketonic oxygen atom accounting for both the anomalous hydroxyl and carbonyl absorption. In chloroform solution the carbonyl bands are lowered by about 5 em.-' but the positions and relative intensities of the hydroxyl bands are not changed (Fig, 2a). On stereochemical inversion a t C-17 the carbonyl band is still observed a t 1690 em.-' (Fig. 2b) (XVII) but the intensity is lower and the relative intensities of the free and associated hydroxyl bonds change considerably on threefold dilution. Thus it CHI
?,
H &=H & = CHr
R; = H R, = Br R1 = H
XIV, R = CHz
1735
XVIIIA
Another example of hydroxy-carbonyl interaction involves the 12a-acetoxy group and the 20phydroxy group (Fig. 5) (XIX-XIXA), where the 12a-acetoxy band is displaced from 1733-1740 em.-' (CClr) to 1718 em.-' with a shoulder persisting a t 1740 cm.-', again two hydroxyl bands are present, relatively insensitive to concentration. No analogous interaction occurs in XX where the positions of the acetate and hydroxyl groups are
I 80
A
GHa
-.
-
I \ Cti,
I
c a 0
m
I
$ 40-
1
Ca-,D S. Keir aiid K Dobriner, rbi,i 74, so ( i u i 2 1 ( 1 1) T h e h a n d positions for 21-acetuxy-20 k ? t o n e s in CSc ; i i i i l CCii i ~ i ca t 1755 1758 and 1721 1 7 3 6 rm.-1. ( 1 2 , ?'hry lid \ t ' iii.cii rliiciiszcd previously (rclcrriicc 2).
STUDIES IN STEROID METABOLISM
June 5, 1952
2823
1733
8
m
m
s
40-
c o F* p r i s m
SOLVENT
CCI,
\ 3600
I800
3400
FREQUENCY
1750 (CM.'
1700
1650
' ),
FREQUENCY
(CM.-
),
Fig. 4.
Fig. 5.
bands are highly characteristic of the 21-acetoxy-20ketone group.
also depend on the spacial distance between the two oxygen atoms.
CH2.0.CO.CHa
CHI
co
co
I
iXXII CHiR
I
I
/XXIII CHI
I
GO
"IAi?-" 'Y-
XXIV, R = H XXV, R = OH
XXVI
Other Dicarbonyl Interactions.-Several other structures containing two carbonyl groups a t nearby positions are included in Table I. A small but significant band shift occurs in the low frequency band of 17~-acetoxy-20-ketones(XXIII) where the maxima are at 1742 and 1715-1716 cm.-' (CS2), the normal positions being 1735-1742 and 17061710 cm.-l. The 11,20- and 16,20-diketones (XXIV-XXVI) absorb normally. The reasons for these upward frequency displacements in dicarbonyl compounds are not yet clear. -4 simple resonance coupling between the carbonyl vibrations is ruled out, since this would displace the bands in opposite directions. A mutual effect in which the dipole field of each carbonyl diminishes the polarization of the other carbonyl bond may be involved. An effect of this kind would have a vector character; its magnitude would depend pn the angle between the carbonyl bonds, being negligible if the bonds were perpendicular. It would
FREQUENCY
( C M - '1
Fig. 6.
The 17a-Hydroxy-20-keto-2l-acetoxy Group.The introduction of the 17a-hydroxy group into the 21-acetoxy-20-ketone group yields a structure (XXVII) which might be expected to combine the spectrographic characteristics of XVI and XXII. CH2.O.CO.CHa II
co XXVII
2824
VOl. 74
I?. N.JONES, 1'. HUMPI-IRIES, F. HERLING AND KONRAD DOBRINER
TABLE I POSITIONS OF C 4 AND C-C STRETCHING BANDSIN THE INFRARED SPECTRA OF C-21 S~EROIDS The compounds are listed in order of the structural formulas given in the text. The absorption bands associated with the carbonyl groups of the ring system, a s opposed to the side chain are italicized unless they are superimposed on side chain bands. Measurements made with a calcium fluoride prism are indicated by a n asterisk (see Experimental section). T h e donors of the compounds are indicated by superscripts t o footnotes a t the end of the table. Compound
Allopregnanol-20a-one-3-acetate' ~regnanol-20a-acetate' Allopregnanediol-l7a.20-one-3-acetate-20' Pregnanetriol-3,17a,20-diacetate-3,20" A 6-Pregnenetriol-3j3,17~~,20-diacetate-3,20 ~4-~regnenetriol-17a,ZO,21-one-3-acetatc-Z1~ A4-Pregnenetriol-17a,20,21-one-3-diacetate-20,2 l".'~' Allopregnanol-3a-one-20' Allopregnanediol-3a,6a-onc-20' Pregnanediol-3a,6a-one-2Ow AK-Pregnenol-3p-one-20" A1*3,610-17-Acetylestratrieno~-3d A4-Pregnenedione-3,20 (progeslerom)" A4, 1-Pregnadienedione-3,20p
A~-Pregneiiol-3l-dione-3,20(desaxycorticosleronc)'
A4-Pregnenediol-llj3,21-dione-3,20 ( corlicosteronc)'n 17-Rromoallopregnanone-20'' Ab- 16a-Methylpregnenol-3p-one-2OC A4-1tia-Methylpregnenedione-3,ZOC AS-1tia-Methoxypregnenol-3~-one-20-acet~lc' A4-16a-Methoxypregneriedione-3,2Oc A~-16a-Ethoxypregnenol-3~-one-2O-acetate' 21 -Bromopregnanediol-3 a ,17a-one-20' 21-Nromopregnanediol-3a,l7a-one-20-form~tc-3'
cs2 CS2
CS? CHCl3 CS2 CHC13 CI1ClI CfxC13 CFICll CIICli CHCl3 CHCli CHClj CS2 CHC13 CHC13
cs2
cs2
CS?
c s2 CS?
c s2 CI-ICl? CC14 CIICl? CIICll CIICI ClLC'l c S? C'S.
CIiCl CS2
I 'i-~soprcgnanol-3a-one-20' I ;-1sopregnano1-3a-one-~0-acetate~ A.i-16-~fcthyl-17-isoprcgnenedionc-3,3~)r
c 5: c S?
A 1- 16,17-Methylenepregnenedione-3,20"
CS
As- l6,17-Methylenepregnenol-3 8-one-20'
CCll CClr C!ICli CEICll CJ3C11 CTICl? CCl, c S2
Allopregnenediol-3 8,17a-one-20-acetate-8' (Krirlzstrin,'s Compound L acetate) i1~-Alloprc.g~ienediol-3 b,1ia-one-20-acetat r .3 ~rrgnanetlioi-~oc, ~a-onc-20' ~regiianeriiol-3~ , 1 7 ~ - o n e - 2 0 ~ 1'regnarie(1iol-:3m, 17~-one-20-fortnatr-:l' ~)rcgnanecliol-Xry, lia-one-20-acetate-3' ~ ' r c g n a n e d i o l -Iia-otic-20-acetate-8"1 ~~, AS-Pregnenediol-3~,17ol-one-20-acctaIc-.'~~ ~'-Preg~renol-l7a-dione-3,20~
Maxima, c m.
holvent
CS2 CS2
1736 1734 1738 1739 1736 1736 1749" 1739" 1700 1700 1703 1702 1700 1698" 1702" 1710
1703" 1704 1703 1706 1709 1736 1706 1736 1720 1725" 1717 1730 1716 1701 1736 1736 1724 1706 1 ?'?sa 1703 1685 1685" 1733' 1720" 1720" 1702 1700
1719
1666 1678" 1662"
1617
1663" 1663" 1677 1664" 1663
1615" 1618"
1615"
1617"
1676 1706 1679 1709
1723 1700 1'ion
1706 1701
1704" 1676
1710" 1710" 1706" 1685 168.5
1690"
272.5"
1710'
1604"
17\76
1696 1693" 1697 1661" 1615"
CIICl.
1779 1704*
1707 1708" 1710 1087"
~regnanol-17a-trione-3,11,20' A~-17-Isopregncnol-l7@-dione-3,20~
CHCIa CHC13
1705 1703"
1685 1690"
i15.30,17u-Diliydroxybisnoreholerlic acid methyl ester'
CCl, CC14 CCl, CCl, CSI Cs?
1740" 1718" 1735" 1733" 1736 1736
1719"
CS c Si
-1
1719
17n-7 1 )
1688" 1688"
1663" 1613"
t~regnanetriol-3a,l2a,20@-acetate-1~' ~~~-e~~~netriol-3ol.l2~,20@-diacetate-3,12' I'I-~giiaiietriul-:(Lu,l"u,30B-tliacetate-3,arJ'
i'rcgnaiietriol-Ra, 12a,~~a-acetate-12' Prcgi lane tr io14 a , 1 2 e,20 a-aceta t c -3Jf
1718"
June 5, 1952
STUDIES IN STEROID
2825
METABOLISM
TABLE I (Continued) Compound
Structure
XXII~
Allopregnanediol-38-21-one-20-acetate-21' Allopregnanol-21-dione-3,20-acetate' A1-Allopregnenol-21-dione-3,20-acetated Pregnanediol-3a,21-one-20-acetate-21r Pregnanetriol-3a,lla,21-one-20-diacetate-l1,21'
Pregnanetriol-3a,12a,21-one-20-acet ate-21' Pregnanol-21-dione-3,20-acetated AS-Pregnenediol-3 j3,21-one-20-acetate-21c~' A'-Pregnenol-21-dione-3,2O-acetate' (desoxycorticosteroneacetate) A4-Pregnenediol-1 113,21-dione-3,2O-acet ate-21 (corticosterone)
XXIII
A'-Pregnenol-21-trione-3,11,20-acetate" (dehydrocorticosterone acetate) Allopregnanediol-3j3,17a-one-20-diacetateC A6-Pregnenediol-3j3,17j3-one-20-diacetateq
1746 1755 1756 1749 1750 1755 1756" 1746 1755 1745 1747"
CS2 CHCla
1754 1745
1732 1724
cs2
1759"
1733"
Cs2
1736 1742 1733 1742" 1730
1716
cs2 CHC13 CS2 CHCls
XXIVb
XXV XXVI XXVII
XXVIII XXIX
xxx
XXXI XXXII
XXXIII
CS2 CS2 CS2 CS2 CHCla
cs2 CHCls CHCla CHCls CS2 CS2 CHC13 CHCla CHCh CHClz
CHCla CHCl,
CS2 c s 2
CS2 CS2
1723 1732 1733 1727 1723 1736 1724" 1720 1732 1721 1723"
1736 1722 1715" 1714
CS2 Csl CSZ CHC1t
1739 1742 1736 1739 1739 1738 1668
A16-Pregnenediol-3j3,12p-one-20-diacetated
CS2
1740
1671
AS**6-Pregnadienedio1-2,3 j3-one-20-diacetateo A4-1e-Pregnadienedione-3,20c A6-16-16-Methylpregnadienol-3 j3-one-20-acetate'
CSZ CHClr
1743 1662" 1736
1617"
CS2
cs2
CS2
A4-17a-\'inylandrostenol-l7
1670
1719 1663" 1615" 1675 1663 1620 1709" 1678"
1716 1712 1677" 1666
1742 1750 1750 1749 1753 1756 1762 1736 1656
cs2
1719 1683
1617 1709" 1737' 1709" 1709 1676 1713 1703 1707 1663 1620 1706 1739 1745" 1724" 1745 1736 1724 1746" 1724" 1705" 1736 1753 1752 1732 1733 1720 1750 1748" 1725" 1708" 1744" 1725" 1706" 1746" 1728" 1660' 1615" 1748" 1728" 1705" 1668" 1617" 1748 1728 1665 1615 1742 1676 1739
1671 1587"
1661 1652
CHCls 1724 CClr 30W A4-17a-Ethynylandrostenol-17j3-one-3' CHCli 3308"' A5-17-Ethynylandrostenediol-3j3,17-acetate-3~ 3310"' CCl, A113f6:101 7-Ethynylestratrienediol-3,l7-acetate-3' CHCla 3310"' Determinedwith calcium fluorideprism. * For additionalexamples in CS,solution see references2 and 3.
SXXIV XXXV
a
Pregnanol-3a-dione-11,20r Pregnanol-3a-dione-11 ,20-acetater A'-Pregnentrione3,l 1,ZO' Allopregnanediol-3j3,2l-dione-l1,2Ow A4-Pregnenol-21 -trione-3,11,20" Pregnanol-3B-dione-16,20-acetate0 Allopregnanetriol-3j3,l7a-21-one-2O-diacetate-3,2lc Allopregnanetetrol-3j3,llj3,17a,21-one-20-diacetate-3,21r Allopregnanetriol-3j3,170,21-dione-l 1,20-diacetate-3,21a Pregnanetriol-3a, 17a,21-one-20-diacetate-3,21' Pregnanetriol-3j3,17a,21-one-2O-diacetate-3,21' Pregnanetetrol-3a,llj3,17a,21-one-20-diacetate-3,21v Pregnanediol-17a,21-trione-3,1 1,20-acetate-218 Pregnanetriol-4,17a,21-trione-3,11,20-diacetate-4,21' A4-Pregnenediol-17a,21-dione-3,20-acetate-21k (Reichstein's Compound S acetate) A'-Pregnenediol-17a,21-trione-3,11,20-acetate-2 la (cortisone acetate) A'-Pregnenetriol-llj3,17a,21-dione-3,20-acetate-21u (Kendall's Compound F acetate) A4b1720-Pregnadienol-21-one-3-acetate' A18-Allopregnenediol-3j3,20j3-diacetate0 A16-Pregnenediol-3 j3,20j3-diacetate0 A1~'M-Allopregnenediol-3&20-diacetate (cis and trans-isomers)' A17 20-Pregnenediol-3a,20-diacetate' A6J7:m-Pregnadienediol-3 8,20-diacetatef A17:m-Pregnenetriol-3j3,11j3,20-triacetate' A17:20-Pregnenetriol-3a,12a,20-triacetate' A17~-2l-Benzalpregnenetriol-3a,12a,20-triacetate' A16-Allopregnenol-3 j3-one-20-acetatez A18-Pregnenol-3a-20z~o
Maxima, c m - 1
Solvent
CHCla CS2 CS2 CHC1, CHC13 CS2 CS2 CHCls CS2 CHCls CHCli
P-one-a$
C.Djerassi
2826
R. N.
JONES,
Vol. 74
P. HUMPHRIBS, F. HERLING AND KONRAD DOBRINER
and G. Rosenkranz, Syntex, S. A. Mexico City, Mexico. d C. Djerassi and C. R. Scholz, Ciba Pharmaceutical Products T. F. Gallagher, Sloan-KetterInc., Summit, N. J. e L. F. Fieser and M. Fields, Harvard University, Cambridge, Mass. ing Institute, New York, N. Y . G . A. Grant, Ayerst, McKgnna and Harrison Ltd., Montreal, P. ,Q. * E. B. Hershberg, J. R. Jamieson and E. The Schering Corp., Bloomfield, N. J. W. H . Hoehn, G. A. Breon and Co., lQns+s Cit Mo. Lozinski, Charles E. Frosst and Co., Montqeal, P. Q. k P . L. Julian, The Glidden Co., Aicago, Ill. 0. Kamm, Parke, Davis and Co., Detroit, Mich. M. H. Kuizenga, The Upjohn Co., Kalamadoo, Mich., H. L. Mason, Mayo Clinic, Rochester, Minn. R. E. Matker, Pennsylvania State College, State College, Pa. p H. B. MacPhillamy, Ciba Pharmaceutical T. Reichstein, University of Products, Inc., Summit, N, J. * P. A. Plattner, Eidg. Tech. Hochschule, Zurich, Switz. L. H. Sarett, Merck and Co., Inc., Rahway, N. J. 1 C. R. Scholz, Ciba Pharmaceufical Products Basel, Basel, Switz. E. Schwenk, The Schering Corp., Bloomfield, N. f . C. M. Suter, Sterlihg-Winthrop Research Inc., Summit, N. J. R. Turner, Harvard University, Cambridge, Mass. R. B. Wagner, Pennsylvania State COlInst., Rensselaer, N. Y. or C=C-H carbon-hydrogen Compound isolated a t the Sloan-Kettering Institute. a C=C-H lege, State ColIege, Pa. stretching vibration. J
(I
In fact, the introduction of the 17a-hydroxy group into XXII does not give rise t o any new band analogous to the 1685-1695 cm.-' band of XVI (Fig. 6). In the hydroxyl stretching region compounds containing XXVII possess a strong free hydroxyl band a t 3600-3625 cm. -l, a weaker associated hydroxyl band a t 3500 cm.-l and, in some cases, a third bandla between 3400 and 3450 cm.-l. Such hydroxyl bands in the spectra of acetylated adrenocortical steroids or their metabolites are highly suggestive of the presence of the 17ahydroxyl group. In these compounds positions l l p - and 17a- are the only ones at which hydroxyl groups resistant to mild acetylation are normally encountered, and these two positions can be readily distinguished, since mild oxidation converts the 110-hydroxyl group to an 11-ketone and a new maximum appears a t 1710-1716 cm.-' (CS2, CCL), 1702 cm.-l (CHCls). Unsaturated Groups Unsaturated Acetates.-In the unsaturated acetates XXVIII and X X I X the ,$?-double bonds have no effect on the carbonyl frequency, but in the enol acetate XXX the band is displaced from 17351742 cm.-l to 1749-1756 cm.-l (CS2) and a further displacement to 1762 cm.-' (CS2) is noted in the cross conjugated system XXXI. An absorption maximum between 1750 and 1765 cm.-' (CS2, CCld) is seen in many compounds containing the -C=C-0-C=O structure (e.g., phenolic acetate~~). CH,.O.CO.CHI
I
CH
CHa
I
HC-O.CO.CH1
other conjugated ketones. l4 In chloroform solution the differentiation among these conjugated ketones is rendered more difficult by the broadening of the band, but the C=C stretching bands of conjugated ketones are readily observed16and serve to identify . the A16-20-ketonegroup. l6 The carbonyl frequency is not altered on introduction of a 16-methyl group XXXIII. CHI !
/-'
17-Vinyl and 17-Ethynyl Groups.-The 17avinyl group in x;rwEIV gives a weak =C-H stretching vibration at 3085 cm.-l but the C=€ stretching band has not yet been detected. .The ethynyl group in XXXV is easily recognized by the strong narrow CsC-H carbon-hydrogen stretching band a t 3340 em.-'. Typical curves illustrating these characteristic ethylenic and acetylenic bands in steroid spectra have been published elsewhere." In the ethynyl compounds a C=C stretching band should occur between 2080 and 2180 cm.-l but has not yet been detected. The wide variation in the intensity of this band in acetylenic compounds has been commented on by Wotiz and Miller. l8 CHz
II
CH
'Y IT XXXIV
XXVIII
XXIX
CH3
C€I=CH.CsHj
C-O.CO.CH8
C.O,CO.CHa !I
I
'.;1 I!
/-
xxx
I
\]A
1 1 /-
XXXI
The A16-20-ketone Group.-In this group XXXII the carbonyl maximum position differs from that in (13) This third band is not necessarily associated with an hydroxyl vibration; it may be a harmonic carbonyl band, since most of the compounds containing XXVII possess several ring carbonyl groups and the carbonyl intensity win be strong.
E"
XXXV
Elucidation of Structure These correlations have been applied to several problems concerned with the identification of naturally occurring and syhthetic steroids, and a few examples may be briefly noted. (14) A16-2O-kclone.1666-1670 cm. -1; A4-3-ketone, 1674-1678 cm. -1; A6-7-ketone, 1677 cm. -1; AI-3-ketone, 1680-1684 cm. -1; AB:11-lZketone, 1680-1684 cm.-I; Ae:1'-15-ketone. 1705 cm.-1; A16-17-ketone, 1716 cm.-1 (all in CSr or CClr solution). Al8-2O-kdonc, 1652-1662 cm.- 1 ; A4-3-ketone, 1860-1668 cm. - 1 ; AWLketone, 1670 ern.-' (CHCla solution). (15) R.N.Jones, P. Humphries, E. Packard and K.Dobriner, THIS JOURNAL, 12, 86 (1950). (16) A1%'0-kctonc, 1585-1590 cm.- 1 ; 81-3-ketone, 1604-1609 cm. -1; A9:11-12-ketone, 1607 cm. -1; A'-&ketone, 1815-ldlO crn.-1 (all in CHCli solution). (17) R. N. Jones, Chcmislry in Canada, 2, 26 (94) (1950). (18) J. H.Wotiz and F. A. Miller, T R I S JOURNAL, 01, a441 (1949).
June 5, 1952
STUDIES IN STEROID METABOLISM
(A) .-During an investigation of the steroid hormones and metabolites excreted in the urine of patients receiving cortisone or ACTH, l9 A4-pregnenediol-17a,21-trione-3,11,20-acetate-21 (cortisone acetate) and A4-pregnenetriol-l1P,17a121dione-3,20-acetate-21 (Kendall's Compound F acetate) were isolated from acetylated ketonic fractions. 'Those substances were readily identified by cornparison with the infrared spectra of authentic specimens between 950 and 1150 cm.-'. It was anticipated that metabolites of these hormones would also occur in which ring A is reduced. As yet few model compounds of this type are available for direct spectral comparison, but the presence of compounds XXXVI and XXXVIII in human urine was inferred from the spectra of the hydroxyl and carbonyl stretching regions of the spectra of two acetylated steroid metabolites which were isolated (XXXVII, XXXIX). CHJOR
co I O€I\A
cm.-' band were associated with the 11-ketone group, then the hydroxyl group present in the acetylated compound is almost certainly a t the 17aposition. Thus the tentative structure XXXIX was arrived at for this acetylated metabolite and this was subsequently confirmed by oxidation to XL and comparison with the spectrum of the authentic compound which was available.
i Go Fa p r i m
$
+ n.
60-
zm
co
a 40sp
O\CIA--OH
SOLVENT
CHCl,
K
CHzOR
I
I
2827
&-OH
I ' ol/\/-
I l l
2ol
0 4 ' -
3610
XXXVI, R = H XXXVII, R = CO.CH8
XXXVIII, R = H XXXIX, R = CO.CH3
FREQUENCY
(CM,-').
Fig. 7.
The spectrum of one of these corn mnds (XXXIX) possessed maxima in chlorofon solution a t 1752, 1724 and 1708 cm.-'. The enter band was more intense than either of thc other two (Fig. 7), and bands were also present I the hydroxyl region a t 3610 and 3460 cm-' The bands a t 1752 and 1724 strongly suggested tf presence of the 21-acetoxy-20-ketone group (: IIII), and the absence of any carbonyl absorption )elow 1700 ern.-' showed that the A4-3-ketone roup of the hormone had been reduced. The bl ' a t 1708 cm.-' suggested a ketone group a t e her position 3, or 11 (positions 4, 6, 7 and 12) 1: ig excluded from consideration on biochen a1 grounds. I t was considered that the reductic of the A4-3-ketone would more probably yield a 3hydroxyl than a 3-ketone group and the 1708 cm.-' band was tentatively assigned to the 11ketone. An acetylated 3-hydroxyl group would be required by this hypothesis; this would absorb in chloroform solution near 1720 cm.-l and the likelihood of two carbonyl groups contributing to the 1724 cm.-' maximum seemed reasonable in view of its high relative intensity. If the 1708 (19) S. Lieberman, L. B. Hariton, hl. B. Stokem, K Dohriner, Federation PYOC., 10, 315 (1951).
P.E. Studcr and
(B).-In the course of an investigation of the action of perbenzoic acid on the enol acetate XXX a product was obtained by Gallagher and KritchevskyZ0 which exhibited the characteristic absorption of the 17-hydroxy-20-keto-2l-methyl group (XVI) in both the C=O and 0-H regions. An intermediate product, obtained prior to hydrolysis possessed no hydroxyl band and the carbonyl absorption of a normal saturated 3-acetate consistent with the epoxide structure XLI. CHI
I
HC
1 ,\:o
"1
/-- XLI (C)._In a reinvestigation by Gallagher and FukushimaZ1of a reaction described by Markerz2 for the introduction of the 17a-hydroxyl group, the product obtained did not exhibit the characteristic absorption of the 17a-hydroxy-20-keto-21 methyl group, and this observation led subsequently to the correct identification of the compound as a 16 - methoxy - 20 -keto - 21 - methyl derivative by chemical means. (D).-During the synthesis of progesterone labeled a t position 21 with C14, Heard and Yatesz3 ( 2 0 ) T. H. Kritchevsky a n d T F. Gallagher, THISJOCRNAI., 73, 184
(1951). (21) D.K Fukushima and T. F. Gallagher, ibid , 71, 2306 (1950). (22) R. E. Marker, ibid., 71, 4149 (1949). (23) R . D. H. Heard and C Yates. private commiinication.
2s2s
I