OPTICAL ROTATORY DISPERSION STUDIES. LVIII.1 THE

OPTICAL ROTATORY DISPERSION STUDIES. LVIII.1 THE COMPLETE ABSOLUTE CONFIGURATIONS OF STEVIOL, KAURENE AND THE DITERPENE ...
0 downloads 0 Views 347KB Size
3720

COMMUNICATIONS TO THE EDITOR

the pellet of CmF4. The solution was then centrifuged, the supernatant liquid was transferred to a quartz absorption cell and an aliquot taken for radiochemical assay. Care was taken to maintain the solution a t 5 10' during these manipulations. The cell, containers, etc., all were prechilled in a refrigerator prior to use. The spectrum was recorded immediately on a Model 14 Cary Spectrophotometer with sample and reference compartments held a t 10.5'. Cesium fluoride (15 M ) was used as the reference solution. Aqueous CmL44(IV) is not stable with time but disappears within ca. one hour as expected. The diminution of the two principal peaks was followed as a function of time with eleven different readings being obtained for each maximum. -1 zero-order rate constant of -1.3 f O.Z%/rninute was obtained a t 10.5' which is consistent with the predicted alpha-induced reduction and curium halflife as described above. The reduction of Cm(1V) is hastened by some other reaction a t 25' as complete disappearance of the oxidized state was noted within 20 minutes a t that temperature. By extrapolation to the time of assay, estimates of the molar extinction coefficients for the two principal maxima were obtained. The location of the observable maxima (AlOA.) in solution and the previously determined maxima for solid CmF412 are compared in Table I. TABLE I l l a x i m a of aqucoiiy Cm(IV) i n l i If CsF a t 10 5'

3310 A. 3380 3506 3565 3830

4514 e E% 1GO 4620 6650 G96O 7820 8000 5640 c = 130 YlOU

Maxima of anhydroiis CmI'4

3865 A. 4010 4118 4504 4607 6730 6960 T650 7915 8560 9100 10975 16120

1

CmFr?

Note that the molar extinction coefficients of Cm(1V) are considerably greater than those of either aqueous Cm(I1I) or Am(1V). I n fact, the positions and magnitudes of the two principal peaks of Cm(1V) are strongly reminiscent of aqueous Xm(II1) with which it is isoelectronic. After the Cm(1V) maxima had disappeared (within one hour), diminished Cm(II1) maxima were apparent a t ca. 4000 A. However, most of the Cm(II1) formed from the self-reduction is in the insoluble CmFa state and floats on the surface of the sirupy solution. I n addition to the Cm(III), we noted a small amount of Am(II1) ( E = 400) a t 5000 A. present only after the Cm(1V) had disappeared. The curium stock was known to contain some americium

VOl. 83

impurity; therefore, the americium iiiust have been present as Am(1V) in solution (with E = 30 for Am(1V) a t 4559 A., such a small concentration could not be detected spectrophotometrically) until after the Cm(1V) had reduced. Attempts were made to dissolve CmF4 in saturated potassium or rubidium fluoride solution but rapid reduction of the curium was observed. Further work is in progress to examine thoroughly the behavior of Cm(1V) in fluoride media and to obtain a crystalline alkali Cm(1V) fluoride for X-ray studies. Los ALAMOSSCIENTIFIC LABORATORY THOMAS K. K E E V A N Los ALAMOS,NEW MEXICO RECEIVED JULY 21, 1961 OPTICAL ROTATORY DISPERSION STUDIES. LVIII. T H E COMPLETE ABSOLUTE CONFIGURATIONS OF STEVIOL, KAURENE AND T H E DITERPENE ALKALOIDS OF THE GARRYFOLINE AND ATISINE GROUPS2

Sir:

The recently announced3 ititerconvcrsioll of steviol (I)l and the diterpene alkaloid garryfolille (11)5 through ( -)-/3-dihydrokaurene6 (dihydro derivative of 111)-coupled with various rotatory dispersion m e a ~ u r e m e n t s ~on , ~ ~ketonic , ~ ~ ~ degradation products of these diterpenoids-has resulted in mutually consistent, absolute configurational assignments a t all asymmetric centers with the exception of C-9. Earlier summarized, circumstantial a r g u m e n t ~ ~ 8led ~ ~ us ~ 9 to favor a 9,lOa d backbone in garryfoline (11) ; if this is correct, then our experimental intercon~ersion~ requires a similar stereochemical feature in steviol (I) and (-)-kaurene (III).6,8b We should now like to report rotatory dispersion measurements on some new steviol and kaurene derivatives, which settle this remaining stereochemical question for this entire group of naturally occurring diterpenoids. Incidental to these studies, i t was observed that ( +)-mirene6 is actually a difficultly separable mixture of (+)-kaurene (antipode of 111) and phyllocladene; its stereochemistry,8b therefore, is of no further concern. Ozonolysis of steviol (I) methyl ester yields4 the norketol V, whose positive Cotton effect (Fig. 1) is very similar to thatlo of ketonorepiallogibberic acid (1) Paper LVII, C. Djerassi, E. J. Warawa, J. M. Berdahl and E. I. Eisenbraun, J . A m . Chem. Sac., 8 3 , 3334 (19G1). (2) The work at Stanford University was supported by grant No. CRTY-BO61 from the National Cancer Institute of the National Institutes of Health, U. S. Public Health Service. ( 3 ) E . Mosettig. P. Quitt, U. Beglinger, J. A . Waters, H. \'orbrueggen and C. Djerassi, J . A m . Chem. Sac., 8 3 , 3163 (19Gl). (4) F. Dolder, H. Lichti, E . Mosettig and P. Q u i t t , ibid., 84, 246

(ism)

(.(a) i)H . Vorhrueggen arid C. Djerassi, Tetrahedron Lellers, 119 (1961); (b) C. Djerassi, C. R. Smith, A . E. L i p p m a n , S. K. Figdor a n d J . IIerran, J . A m . Chem. Sac., 7 7 , 4801, GG33 (1955). (ti) L. H. Briggs, B. F. Cain, B. R . Davis and J. K. Wilmsliurst. Tcfrakedron L e l l e r s , No. 8, 8 (1939). (7) C . Djerassi, R . Riniker and B. Kiniker, J . A m . Chem. hoc.. 78, G R G 2 (1956). (8) (a) L. H. Briggs, B. F. Cain, B. R. Davis and J. K. Wilmshurst, Tetvahedroir Lelters, No. 8 , 13 (1959); (b) L. H. Briggs, B. F. Cain, R. C. Cambie and B. R . Davis, ibitE., no. 24, 18 (1960). (9) .4.J. Solo and S. U 7 . Pdletier, Chein. e,Znd., 1108 (ISGO). (10) See J. F. Crnve and T. P. C. l l u l h o l l a n d , J . C h e m . Sac., 3007 (1960); J . F. Grove, J lIacAlilIan, T. P. C. Mnlholland and W. B. Turner, ibid., 3049 (1900).

372 1

COMMUNICATIONS TO THE EDITOR

Sept. 5 , 1961

and which possesses the identical C/D absolute configuration, Treatment of the norketol V with potassium t-butoxide (4 hr. reflux in t-butanoltetrahydrofuran) or preferably with butyllithium (3 hr. refluxing in ether-tetrahydrofuran) furnished in over 50y0 yield the isonorketol VI1 (m.p. 197199', [a]D -62.5' (all rotations in chloroform); Anal. found for CZOH3004: C, 71.76; H, 9.42), the change resembling mechanistically (Va) the usual steroid D-homo rearrangement. The isonorketol VI1 exhibited a strong negative Cotton effect (Fig. l), as was the case with isosteviol (VII, hydroxyl group replaced by methyl)' or gibberic acid. lo Periodate cleavage of the norketol V provided the corresponding seco-acid VI (m.p. 178-181°, [ a ] ~ -69.5'; Anal. found for C20H3006: C, 68.34; H, S.85), whose weak positive rotatory dispersion curve (Fig. 1) was essentially the mirror image of that of a 5p-3-keto steroid. l1 Similar oxidation of the isonorketol VI1 led to the iso-seco acid VI11 (double m.p. 141" and 198-200', [a]D -46.4'; Anal. found for C20H3006:C, 68.GO; H, 8.73), which exhibited a negative Cotton effect (Fig. 1)qualitatively the mirror image of the R.D. curve of a 5cr-3-keto steroid."

Y

QH

300

400

Amp. 500

L

1.

700

Fig. 1.-Optical rotatory dispersion studies (methanol solution) of steviol degradation products \'-VIII.

C02H

'I

III

IVa, R = HCOCH,, b,R=HOH

Va

c,R=O

IX

x

The base-catalyzed conversion (Va) of the norketo1 (V) to the isonorketol (VII) involves only carbon atoins 8 and 13, while the configuration a t C-0 must have remained unchanged. Consequently, if steviol possesses a 9,lO-anti backbone, the derived seco acids must be represented by the pair VI and VIII, while a 9,lO-syn stereochemistry in steviol would lead to the C-8isomeric seco acids I X and X. As has already been noted earlier,I2 (1 1) C . Djerassi, "Optical R o t a t o r y Dispersion," McGraw-Hill Book Co.,New York, N. Y., 1900. (12) C. Djerassi, M. Cais and L. A. hlitscher, J . A m . Chem. Soc., 81, 2386 (1959).

the octant rule11313predicts a weak positive Cotton effect for VI and a negative one for VI11 in agreement with the experimental results (Fig. 1). Turning to the C-9 isomeric pair, IX would be expected to have a much stronger positive Cotton effect than VI, but most importantly, X would exhibit a positive Cotton effect in contrast to the observed negative one, which according to the tenets of the octant rule13is only consistent12 with VIII. ( -)-Isokaurene6 (15-16 double bond isomer of 111) upon permanganate oxidation provided the methyl ketone IVa (m.p. 140"; Anal. Found for C20H3203: C, 74.44; H, 10.08; neut. equiv., 322), which was subjectedL4to Bayer-Villiger oxidation followed by saponification. The resulting hydroxy acid IVb (m.p. 254-2%' ; Anal. Found for ClgHaoOs: C, 73.27; H, 10.40) was oxidized to the keto acid IVc (m.p. 169-170'; Anal. Found for C18H2803: C, 73.92; H, 9.77), whose positive rotatory dispersion curve (peak a t [ a]306Me0H +150°) was similar to that (Fig. 1) of the steviol seco-acid VI and antipodal to that" of a 5p-3-keto steroid. These rotatory dispersion studiesIb remove the last doubt concerning the absolute configuration of this group of diterpenoids. The above conclusions from the steviol (I) and (--)-kaurene (111) series are directly applicable to garryfoline (11), because of their mutual experimental interconversion. (13) W. Mofitt, R. B. Woodward, A . hloscowitz, W. Klyne and C. Djerassi, ibid., 83. October (1961). (14) For analogous reaction sequence in t h e phyllocladene series, see P. K. G r a n t and R. Hodges, Tetraheduon, 8 , 2til (1960). (1.5) We are indebted to Mrs. T. Nakano and Mrs. R u t h Record: for t h e O.R.D. measurements.

3722

COMMUNICATIONS TO THE

Furthermore, garryfoline (11) has been relatedbbto the other Garrya alkaloidsi6 as well as to members of the atisine class," so that the complete absolute stereochemistry of these alkaloids is now known. (112 K. Wiesner and Z. Valenta in L. Zechmeister's "Progress in the Chemistry uf Organic X a t u r a l Products," Springer, Vienna, 1958, Vol. XVI, pp. 26-89, (17) S W. Pelletier, J . A m . Chein. Soc., 82, 23'38 (19GOj.

DEPARTMEST OF CHEMISTRY STAXFORD UXIVERSITP STANFORD, CALIFORSIA CARLD JERASSI SATIONAL INSTITUTE OF ARTHRITIS A N D METABOLIC DISEASES S A T I O S A L ISSTITUTES O F HEALTH PETERQUITT BETHESDA,MARYLASD ERICHMOSETTIG DEPARTMENT OF CHEMISTRY R . C. CAMBIE t-SIVEBSITY OF AUCKLAXD P. S. RUTLEDGE AUCKLAND, XEK Z E A L A S D L. H. BRIGCS RECEIVED JULY 17, 1961

s;z

EDITOR

~701.

,0

0

h

0

0 trans-I €3

cis-I 5

1

I!

!t

NON-CHAIR CONFORMATIONS. EQUILIBRATION OF CIS- AND T R A S S -

2,5-DI-t-BUTYL-1,4-CYCLOHEXANEDIONE Sir: tjans-I 3 Lib-I 7 Equilibrations of cis- and trans-l,3-di-t-butylcyclohexane' and of cis- and trans-2,4-di-t-butylcyclohexanone2demonstrate that in these molecules, the t-butyl groups prefer the equatorial orientation R o w 0 in chair conformations. This communication de- 0-0 scribes the equilibration of cis- and trans-2,5-di-tbutyl-1,4-cyclohexanedione (cis-I and trans-I), k i hlthough trans-I can exist in a chair conformation trans-I 4 cis-1 8 with both t-butyl groups equatorial ( l ) , the equiR i E t-Bu librium favors cis-I. We conclude that cis-I presolution 0.58 X in hydrogen chloride (prepared froni fers a nonchair conformation such as 7. The diones, cis-I and trans-I, were prepared ster- 5.00 ml. of concentrated hydrochloric acid plus eospecifically in good yield by the Jones oxidation3 sufficient glacial acetic acid to bring the total volume to 100.0 ml.) are given in Table I. of two isomeric 2,S-di-t-butyl-1,4-cyclohexanediols, diol A (t-butyl groups cis) and diol B (t-butyl TABLE 1 groups presumed tra.ns). Oxidation of diol A,4m.p. EQUILIBRATIOS: t i a m - I i-2 c.ib-1 157.5-158.5 O, yielded cis-I, m.p. 110-140.5'. Rapid 7', 'l'imc. , . , reaction of 3 moles of hydrogen with 2,S-di-tA/' I< b a C. hr. butylhydroquinone in acetic acid solution (con0 . 2 -0.94 i 0 . 0 4 &9 37 81.5 i 0 . 4 4.41 taining one drop of concentrated hydrochloric acid) 85.0 3.0 7 9 . 1 k O . 2 3 . 7 8 f 0 . 1 - 0 . 9 5 f 0 . 0 2 with platinum oxide catalyst a t 75-80' under 2-4 .ltialyses by gas cliromatography were carried out in atm. pressure gave a product mixture from which a duplicate a t 180' x i t h a 10 ft. 0.25 in. copper colunin 2,5-di-t-butyl-I ,4-cyclohexanediol, diol B, m.p. packed with 21170 silicone gum rubber on 60-80 mesh firebrick. The analL-ses were calibrated against known mix220-221 ', was isolated by fractional recrystalliza- tures, one containing 79.1yc ris-I. * From a graph of 111 K tion in ca. 5% yie1d.j Oxidation3 of diol €3 yielded i function of 1/T, the :ipprosirriate values of the enthalpy trans-I, m.p. 151.5-152°.6 1 cntrnp!- of equilibration ivere deterniined: A H , - 0.8'7 Acid catalyzed equilibrations a t 25-100' of the + 0.3 kcd.jmole; AS,0.2 i I1.T e.u. diones, cis-I and trans-I, 0.1-0.2 .14 solutions in For cii-I and tuum-I, the three different possible acetic acid-water, 0.1-1 AT in hydrogen chloride, chair conformations (1, 2 and 5 ) and three of the and in carbon tetrachloride, 0.1 ,2i in hydrogen nine different possible boat conformations (3, 6 and chloride, gave mixtures containing 20 f 10% 7) are illustrated. The other possible boat conforniatruns-I and SO =I= 10% &-I a t equilibrium. The tions are predicted to have higher energies than 3, 6 results of equilihratioris in an acetic acid-water and 7 because of stronger repulsions between noti. h m & .SOL.,83, 2.3!U (1) N. I,. Allinger a n d I, A. I'reiberg. J . ~ l v i C bonded groups. In addition, other conformations, (I900). such as twist7 conformations 4 and 8, require con(2) N.I,. Allinger and H. 11.Blatter, { b i d . , 83, 994 (1961). (:O R . G . Curtis, I . Heilbron, E. R . H . Junes a n d G . F . I V < i d s , sideration. J . < ' h e m . SOC, 457 (1953). ,-1 simple argument can be given in support of the (4) R. D . Stolow. J . A ? u ( ' h e i n . Soc., 8 8 , 2592 (1901). Assignment conclusion that cis-I prefers a non-chair conformarif t h e c i s configuration t o t h e I-butyl groups in diol A is considered tion. If a t-butyl group preferred an equatorial oriento h e unequivocal, and is based upon t h e observation t h a t diol A tation when 1,4-cyclohexanedione was in the chair exhibits intramolecular hydrogen bonding. ( 5 ) Diol B does not exbihit intramolecular hydrogen bonding. conformation, then 1 (the diequatorial chair con(ti) T h e diones, cis-I a n d l?a?is-I, gave acceptable carbon and formation of trans-I) would be more stable than 5 . hydrogen analyses a n d have been characterized f u r t h e r hy their (

infrared, ultraviolet and nuclear well a s h y gas chromatography.

magnetic resonance spectra, a,

,!..'I

*

(7) W. S . Johnson, V. J. Bauer, J . I.. l I a r g r a v e , &I. A . Frisch. T,, H. Dreper and W. N. H u h h a r d . .I. A i i i . CIiPm Soc., 83, GO6 (1961).