Tetrahedron Letter* No. 35, pp 3099 - 3102, 1975. Pergamon Room. Printed in Great Britain.

THE STHUCTIJHE CF CIMBOWGONE, A NOVEL THITXtPEZiOD FBCM LEMONXASS

Malcolm Crawford, Steven W. Hannon" end Mouatapha E.S. Koker

Chemistry Department, Fourah Eay College, University of Sierra Leone,

Freetown, Sierra Leone

(Beceired in UK 1 July 1975; accepted for publication 17 July 1973)

During investigations into the constituenta of the leaf wax of lemongraes, Cnnbooooaon

citratug Stapf, two triterpenoids were isolated, a ketone, which we have named cymbopogone,

and an alcohol which has been named cymbopogonol.' The spectroscopic and analytical data for

cymbopogone, which we present in this paper, are consistent with structure I.

Cymbopogone orystallieed from ethanol aa needles (overall yield based on fresh leave8

O*OO2$), m.p. 262-265'. The infrared apectnrm indicated the presence of ketone aarbonyl

(1715 cm-') and carbon, hydrogen and ma88 spectral analysis gave a molecular formula of

C30H500* The mase spectrum showed distinct similarities to that of friedelin (II)2 and

indicated a pentacyclic triterpenoid with a D:A-fried0 structure having the carbonyl group in

the A-ring.'j Dy comparieon of infrared epectra and by mixed m.p. determinations we were able

to show that cymbopogone ie neither friedelin (II) nor filican-3-one (III).

In the n.m.r. spectrum of cymbopogone the lowest field resonance was at 6 2.35 p.p.m.;

in the methyl region only five peaks were visible: b 0*72 (3H), O-84 (one peak of a 3H

doublet), O*S7, O-91 end O*96 (3H) p.p.m. In the n.m.r. ape&rum of friedelin the lowest

field resonance is also at b 2*35 p.p.m., but eight peaks are clearly visible in the methyl

region: i3 0*73 (3H), 0.84 (one peak of a 3H doublet), O-87 (3H), O-92 (one peak of a 3H

doublet), 0.96 (3H), l-01 (6H), 1.05 (3H) and 1.18 (3H) p.p.m.

In order to obtain further information about the etructure of

shift reagent Eu(fod)3 wae employed. The results are given in the

obtained for friedelin (II). The similarity in the induced ahifte

cymbopogone the n.m.r.

table along with those

for the 4, 5, 99 13 and

3099

3100 No. 35

Table m(fod), induced shifts and chemicel ehifte of methyl group*

me*1 group 4 5 9 lP 1P 17 other methyls

inauoed shift 100 65.6 19.9 10*3 12.3 4-6 4'0 3'2 cymWwwe

ahedoal shift o*sG 0-72 0990 0.87 o-91 0.96 0*9& O-8&

induced ehift 100 64.9 19.5 10*4 12-o 5-G 3.3 3-2 friedelin

chemical ehift 0*8& O-73 O-87 l*Ol 1 l 05 id 0.96 l*Ol

filican-j-one chemical ehiftf 0*8e0 O-72 0.92 1-00 0.92 0*82 O& 0*8+

a. The induced shifts me relative to the 4-methyl induced shift equal to 100. 'Ihe

solutions of both cymbopogone end friedelin were ca. 0.03 M in CDC13. Ihe cheaiwl

shift6 ere 5 p.p.m. Aeeignments for cymbopogone and friedelin are bssed on 8

comparison of the observed induced shifts with those Cdtiated taking the 0 to EU

distance as 4 4, with the 0-W exis in the C-Q=0 plane atan Mgle of 14OOto the SxiS

of the C-O U-bond and Cis to C-2.4 1. For cymbopogone 8nd friedelin the assignments

for the 13 and l&methyl groups mey have to be inter&en&. g. Doublet, J I 6.5 Es.

A. Complex multiplet. Doublet 8t high Eu(fod)3 concentrations, J = 6 Hz. 0. This

assigment is not definite. The resonence at 6 1.18 may be due to one of the ZO-methyl

groups. f. Unpublished data kindly eupplied by Prof. B. Ageta.

No. 35 3101

ll)-methyl groups in ogmbopogono and friedelin confinna that oymbopogone he,a a D:A-friek

etructure. Further, it &owe that the carbonyl group ie in the S-position which ia in eny

case the moat likely on biosyntheticr grounds.

!lhe chemical shifts of the methyl groups in cymbopogone and friedelin were calculated by

extrapolation to sero Eu(fod) concentration. 3

The values obtained are given in the table.

The chemical shifte of the methyl groupe in filicen-34ne (III) We also &Yen for comparison.

The differences in the chemical shifta of both the 13 and 14 methyl groupa in oymbopogone and

friedelin are noteworthy but can be explained by the difference8 in the preferred conform-

ations of the two compounds. In cymbopogone the D-ring takes up a ohair conformation and the

E-ring a boat conformation whereas in friedelin it is probable that both the D and E-rings

take up boat confonr&ions.5

At low concentrationa of Eu(fod)3, i.e. with molar equivalent.6 of Eu(fod)3 below 0.15,

no peaks were vieible for the 19 end 2O-methyl groups in cymbopogone. On increaeing the

Eb(fod)3 concentration the other overlapping methyl resonances were ehifted to lower field

away from these methyl groups which were then seen to resonate aa a broad multiplet. Only

at molar equivelente of Eu(fod) 3 over O-5 did the 19 and 20-methyl resonance8 appear ae

oVerlapping doublets. This is in agreement with previous finding6 for the vicinal methyl

groupsin theureane E-ring. 67 Theae methyls form an ADX3Y3 system with the 19 and 20-

methine protons which results in complex resonances being produced. Since however their

calculated induced shifts differ markedly, at high Eu(fod) concentrations the two methine 3

proton reeonanaee are shifted sufficiently apart relative to their coupling conetant for the

methyl reclonances to be amenable to firstdrder analysis. It should be noted that in the

n.m.r. spectra of B':A'-neoganrmacerane derivatives8 and lupane derivatives 7.9 the ieopropyl

methyl groups reeonate aa well-defined doublets.

A compound with structure I is also highly likely on bioeynthetic grounds, indeed ite

existence was forecast twenty yeera ago. 10 Only three groups of pentacyclic triterpenoide

have friedo-derived etructurelr, these are: friedooleananee (D, D:C, D:B and DrA-fried0

Compoundlr me known), friedo-B'zA'-neogammaoeranee (DrC, D:B and DtA-fried0 compounds are

knoWl) and friedoureanee (until now only D and DlC-fried0 compoundlr were known)." Hence

WmbopOSOneP which C8n be neighed systematically aa D:A-friedoura-j-one, is the misting malogue

of D:A-friedoolean-3-e (friedelin) and DrA-friedo-B'rA1-n Ver-34ne (filican-34ne).

3102 No. 35

Acknowleduements

We wish to thank Dr. E.J. Thomas for his assistance with the n.m.r. and maas spectra,

Dr. T.G. Halsall for providing a sample of friedelin, Prof. H. Age&a snd Prof. A. Marsili

for providing samples of filican-3-one snd the Chemical Society for a Research Fund grant.

References and Notes_

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

F.A. Menezee, Ph.D. Thesis, University of E&t Africa, 1968.

The mass spectrum of cymbpogone shove the following major peaks: m/e 426 (4y16), 411

(6%), 383 (3%), 341 (3%), 302 (6%), 286 (9%)~ 273 (15%), 246 (11%)~ 231 (11%)~ 2'8

(Id), 205 (2%), 191 (11%), 179 (I@%), 163 (15%), 149 (16%), 137 (23%)~ 123 (10%) and

109 (4%). The mass spectrum of friedelin shows the following major peeks: m/e 426

(2996), 411 (II%), 341 (e%), 302 (2%). 287 (12%)~ 273 (44%). 246 (33%)~ 231 (2%)~ 218

(37%), 205 (44%), 191 (32%), 179 (42%), 163 (4s%), 149 (33%)1 137 (46%), 123 (90%) ana

109 (100%)#

H. Bdzikiewicz, J.M. Wilson snd C. Djerassi, J. Amer. Chem. SOO., 85, 3688 (1963).

D.J. Chsdwiok and D.R. Williams, J.C.S. Comm., 128, (1974)s P. Finocchiaro,

A. Recta, P. Maravigna and G. Montauao, ~trshedron, Zp, 4159 (1974).

S.K. Talapatra, S. Sengupta and B. Talapatra, $etrahebon Letters, 5963 (1968).

J. Katliner and C. Djerassi, J,ora.. , 11, 1945 (1966).

H.T. Cheung and D.G. Williamson, B, 2% 119 (1969).

K. Nishimoto, M. Ito, S. Natori and T. Ohmoto, Tetrahedrop, & 735 (1968).

J.M. Lehn and G. Ouriaeon, &&I. Sot- Chj&& l , 1137 (1962).

E.J. Corey and J.J. Ureprung, w., a, 3668 (1955).

For this system of triterpenoid nomenclature, see: S. Allard and G. Ourisson,

Tetrahedron, 1, 277 (1957).