THE CONSTITUENTS OF OLEARIA HETEROCARPA.

THE STRUCTURE
OF OLEARIN, A DITERPENE DILACTONE OF THE
CASCARILLIN GROUP*

By J. T. PINHEY,?
R. F. SIMPSON,?and I. L. B A T E Y ~
[Manuscript received July 12, 19711

Abstract
Chloroform extraction of Olearia heterocarpa S. T. Blake has yielded a new
diterpene dilactone, olearin, the known flavone, xanthomicrol, an unknown
C27H4fi03 hydroxy acid, anda mixture of long-chain (approx. CbO)
esters. On chemical
and spectroscopic evidence olearin (1) has been shown to possess a rearranged
enantio-labdane skeleton of the type displayed by the cascarillins. Correlation of
olearin with a diterpene of known stereochemistry has defined the absolute con-
figuration a t all but one centre.

Our interest in the terpenoid constituents of members of the family Compositae

led us to investigate a number of Olearia species, which are commonly known as
"daisy bush". At that time it appeared that the only member of the genus to have
been chemically investigated was Olearia pa?ziculata, which had been shown to con-
tain ( & )-e-curcumene, ( & )-y-curcumene, aromadendrene, 13-epi-(- ) -manoyl oxide,
and a number of triterpenes.2-4
The milled aerial part of Olearia heterocarpa S. T. Blake, collected a t Whian
Whian National Park, Queensland, was extracted by cold percolation with light
petroleum, chloroform, and finally methanol. The chloroform extract yielded a
considerable amount of waxy material from which a new neutral diterpene, m.p.
213", [elD -120°, was obtained by crystallization from ethyl acetate. This com-
pound has been given the name olearinl and its structure determination is the main
subject of this paper. The amorphous material remaining after removal of olearin
was partitioned between aqueous methanol and benzene-carbon tetrachloride.
From the aqueous methanol layer there was obtained in low yield a hydroxy cer-
boxylic acid, Cz7H42O3,m.p. 264O, +77O, which readily gave a monoacetate and
which could be oxidized to a ketone.
* A preliminary account of some of this material has appeared.1
t Department of Organic Chemistry, University of Sydney, N.S.W. 2006.
1 Pinhey, J. T., and Simpson, R. F., Chem. Commun., 1967, 9.
2 Corbett, R . E., Jamieson, G. A., and Murray, J., J. Sci. Pd Agric., 1963, 14, 349.
3 Corbett, R . E., Young, H., and Wilson, R. S., Aust. J. Chem., 1964, 17, 712.
4 McLean, D. H., and Slater, S. N., J. Soc. chem. Ind., 1945, 64, 28.

Aust. J. Chem., 1972, 25, 2621-37

2622 J. T. PINHEY, R. F. SIMPSON, AND I. L. BATEY

Chromatography of the residue from the chloroform extract on neutral alumina

yielded a relatively non-polar material, m.p. 84-85", which analysis and spectral
data indicated to be a mixture of closely related esters of approximate formula
CsoHloo02. A mass spectral analysis of the hydrolysis products indicated that the
esters were formed from Ce2 and C26 saturated carboxylic acids and C23, C25, and
Cz7 saturated alcohols.
I n another attempt a t fractionation of the material extracted by chloroform,
the residue was dissolved in ether and extracted in turn with 5% sodium bicarbonate
solution and 5% aqueous sodium hydroxide. From the sodium hydroxide extract
there was obtained xanthomicrol (5,4'-dihydroxy-6,7,8-trimethoxyflavone) which,
like the diacetate and monomethyl ether, corresponded in all respects to the
material of Stout and Stout.5

Olearin was shown by analysis and its mass spectrum to have the formula
C20H2605. Its infrared [v,,, (Nujol) 3379, 1787, 1740, 1654, and 1630 cm-l] and
ultraviolet (A 214 nm, E 21000) absorption indicated the presence of two cc,/Aunsatur-
ated ester groupings and a hydroxyl group, and on the evidence presented below
it is shown to have structure (1). I t is thus a member of the cascarillin6 group which
have a rearranged enantio-labdane skeleton.'

Olearin could not be recovered from treatment with alkali and thin-layer
chromatography indicated the formation of numerous products. Also, attempts to
oxidize it to the corresponding ketone, even with chromium trioxide-pyridine,s led
5 Stout, G. H., and Stout, V. I?., Tetrahedron, 1961, 14, 296.
6 Halsall,T. G., Oxford, A. W., and Rigby, W., Chem. Commun., 1965, 218.
7 McEachan, C. E., McPhail, A. T., and Sim, G. A., J. chem. Soc. ( B ) , 1966, 633.
8 Poos, G. I., Arth, G. E., Beyler, R. E., and Sarett, L. H., J. Am. chem. Soc., 1983, 75,
422.
 STRUCTURE OF OLEARIN 2623

to complex mixtures. I t was therefore necessary to protect or reduce one or both of

the double bonds in order to proceed with a systematic degradation of the molecule,
The n.m.r. spectrum of olearin indicated the presence of two trisubstituted
double bonds and the assignments (fine multiplet at S 5.90 for H14 and multiplet
at 6.71) are consistent with the proposed structure. Hydrogenation over Raney
nickel or Adams catalyst led to the formation of a single tetrahydro derivative (2)
with an infrared spectrum (1769 and 1759 cm-1) indicative of two saturated ylactone
rings. This compound had only weak "end absorption" in the ultraviolet and gave
no colour reaotion with tetranitromethane. Hydrogenation of acetylolearin (3) also
proceeded stereospecifically to yield a single compound (4) identical with the product
of acetylation of (2). Attempts to selectively reduce both (1) and (3) failed to produce
a satisfactory route to a dihydro compound even though t.1.c. and n.m.r. spectroscopy
indicated a significant amount of the 3,4-dihydro compound was present in the
mixtures.
An examination of the n.m.r. spectra of olearin (1)and acetylolearin (3)indicated
that the hydroxyl group was secondary since a one-proton multiplet at 6 4-86 in (1)
was shifted to S 5.75 in (3). This was confirmed by chromium trioxide-acetic acid
oxidation of (2) to the ketone (5) in high yield. The presence of the ,%substituted
butenolide ring and its relationship to the hydroxyl group were established in the
following way. Treatment of the ketone (5) with hot dilute aqueous sodium hydroxide
gave formaldehyde, which was characterized as its 2,4-dinitrophenylhydrazone.
Since formaldehyde could not be detected under similar reaction oonditions with
olearin and tetrahydroolearin, the reaotion is dependent on the introduced ketone
group, and may be rationalized as proceeding by a retroaldol reaotion as indicated
in (6). Further evidence for the structure of the side-chain was obtained from the
conversion of tetrahydroolearin (2) into isotetrahydroolearin (7) in high yield by
treatment with acid or base followed by acidification. This compound (7) also con-
tained two y-lactone rings (v,, 1773 and 1765 cm-I), and the n.m.r. spectrum
indicated the presence of the grouping >CH-CH20H, since a broad two-proton doublet
at 6 3.65 gave a clean doublet on addition of DzO. In addition, (7) could be readily
oxidized with chromium trioxide in acetic acid to a carboxylic acid (8). In the
formation of isotetrahydroolearin (7), an epimerization was not occurring, since
tetrahydroacetylolearin (4) was unaffected by dilute sodium bicarbonate solution
under oonditions used to effect the conversion of (2) into (7). Olearin gave a positive
Legal test with alkaline sodium nitroprusside under the conditions outlined by
Elderfield et al.9 for /3-substituted Aapfl-butenolides. The presence of this moiety
was further supported by the n.m.r. spectrum (100 MHz) of acetylolearin, which
showed similar features to those of some steroidal examples.loaJlJ2 I n particular,
the two protons on C16 appeared as a partially resolved AB quartet centred at
8 4-82, in which each line was further split into a doublet. On irradiation a t S 5.75
(H12), this system sharpened to a clear AB quartet ( J g e m 17.8 Hz) in which each
line was a doublet (J14,161.8 HZ), while irradiation a t 8 5.97 (H14) caused it to
9 Paist, W. D., Blout, E. R., Uhle, F. C., and Elderfield, R. C., J. org. Chern., 1941, 6 , 273.
10 Bhacca, N. S., and Williams, D. H., "Applications of NMR Speotroscopy in Organic
Chemistry." ( a )p. 46; (b) p. 80. (Holden-Day: San Francisco 1964.)
11 Carman, R. M., Coombe, R. G., and Watson, T. R., Auet. J. Chem., 1964, 17, 673.
12 Collins, D. J., Hobbs, J. J., and Sternhell, S., Auat. J. Chern., 1963, 16, 1030.
2624 J. T. PINHEY, R. I?. SIMPSON, AND I. L. BATEY

collapse to a simple AB quartet. The signal due to H 14 appeared as a doublet of

triplets (J14,16 1 . 8 H Z ; J12,14 1 . 0 HZ) which collapsed to a triplet on irradiation a t
8 5.75 (H 12). Confirmation of the presence of a methylene group a t the 11-position
came from a similar decoupling experiment. The H12 signal was a very broad
doublet which, on irradiation a t 8 5 a97 ( H 14), gave a doublet of doublets (Jll,,12
9.2 H z ; Jllb,i2 2.5 HZ).
To obtain information regarding the substitution a t C 9, a number of attempts
were made to introduce -an 11,12 double bond. Olearin was unaffected by acid under
mild conditions; however, with hydrochloric acid in dioxan a t reflux an isomeric
acidic compound was produced in fair yield. This compound, which was unstable
on long exposure to air, has been assigned structure (9). Its ultraviolet absorption
(A, 237 nm, e 10000) and that of the 2,4-dinitrophenylhydrazonederivative of the
methyl ester (10) (Am,, 251,286, and 376 nm; E 15500,8900, and 27900 respectively),
obtained by treatment of (9) with diazomethane or directly from (1) with methanolic
hydrogen chloride, were in accord13 with the proposed structure. I n the n.m.r. spec-
trum of (9) the two C 14 protons gave rise to a broad singlet (WH 3 HZ)a t 8 3.36,
which excludes the alternative structure (11) for the aldehyde acid. Attempted
oxidation of (9) to the corresponding dicarboxylic acid by silver oxide in ethanol
was unsuccessful.
I n other approaches to introduction of an 11,12 double bond, tetrahydroolearin
methanesulphonate (12) and olearin methanesulphonate (13) were prepared and
their elimination in hot dimethyl sulphoxidel4 was investigated. The tetrahydro
derivative (12) gave a mixture of tv7o less polar compounds (t.1.c.) which could be
separated by preparative layer chromatography but which could not be obtained
crystalline. This approach was, however, not pursued further since i t was found that
the elimination of the corresponding derivative of olearin (13) proceeded readily to
yield a single product which has been assigned the structure (14). Although this
substance could not be obtained crystalline its spectra and behaviour on t.1.c. indicated
that i t was a single compound. Also, i t gave a high yield of the crystalline 3,4-epoxy
derivative (15) on reaction with rn-chloroperbenzoic acid in chloroform. I n agreement
with these structural assignments, (14) and (15) showed high ultraviolet absorption
a t 259 nm, while their infrared spectra (see Experimental) indicated they both
contained two ylactone rings. The n.m.r. spectra of anhydroolearin (14) and the
epoxide (15) were particularly informative. For both (14) and (15) the signals due to
H 11 and H 12 were doublets with splittings of 16.8 Hz, indicating a trans-ethylenic
link.15 I n addition, since the resonances for the protons of the 11,12 double bond
were doublets, C9 did not bear a proton and hence a rearranged labdane skeleton
of the type first proposed for columbinl~was indicated for olearin. I n further support
of the structure assigned to the epoxide (15), the multiplet a t 6 6.75 due to H 3
13 Scott, A. I., "Interpretation of the Ultra-violet Spectra of Natural Products." p. 77.
(Pergamon: Oxford 1964.)
14 Jones, D. N., and Saeed, M. A., J . chem. Soc., 1963, 4657.
1 5 Jackman, L. M.,and Sternhell, S., "Applications of Nuclear Magnetic Resonance
Spectroscopy in Organic Chemistry." p. 301. (Pergamon: London 1969.)
16 Barton, D. H. R., and Elad, D., J. chem. Soc., 1956, 2085 and 2090; Overton, K. H.,
Weir, N. G., and Wylie, A., J. chem. Soc. ( C ) , 1966, 1482; Cheung, K. K., Melville, D.,
Overton, K. H., Robertson, J. M., and Sim, G. A., J . chem. Soc. ( B ) ,1966, 853.
 STRUCTURE O F OLEARIN 2625

in the n.m.r. spectrum of anhydroolearin (14) is replaced in the spectrum of (15)

by a broadened doublet with a 6 Hz splitting a t S 3.38.
As with anhydroolearin (14) the 3,4 double bond of olearin was found to be
considerably more reactive to electrophilic attack than that in the butenolide ring,
and this led eventually to the development of a successful method of stepwise degrada-
tion of the molecule. The unusual reactivity of the 3,4 double dond in olearin and its
derivatives is no doubt due to the strain in this part of the molecule, which is relieved
considerably u lien C 4 becomes an sp3 carbon.

R X Y
. 2" (16) H M O OH
.kc
(17) Ac OH OH

Attempted epoxidation of olearin with perbenzoic acid gave rise to a mixture

of compounds; however, with performic acid acetylolearin (3) gave the diol mono-
formate (16) in almost quantitative yield. Hydrolysis of this compound with dilute
sulphuric acid produced the diol(l7) but attempts to cleave the 3,4 bond with periodic
acid or to oxidize i t to the 3-ketone wtth chromium trioxide-acetic acid were unsuccess-
ful. The position of the formyl group in (16) followed from the n.m.r. spectrum in
which H 3 appeared as a partly obscured narrow multiplet a t 6 5.0. The assignment
of the axial configuration to the formyl group is based on the half-height width of
5 Hz for the signallob due to H 3 a t 6 3.64 in the spectrum (DzO exchange) of the
2626 J. T. PINHEY, R . F. SIMPSON, AND I. L. BATEY

diol (17), which is in agreement with the expected geometry from stereo-electronic
considerations.17 A single dihydro derivative (18) could be obtained in fair yield by
hydrogenation of (17) over Adams catalyst, but hydrolysis to the diol did not give
a crystalline product and a projected lead tetraacetate cleavage was not attempted.
Among other unsuccessful oxidative approaches to the degradation of the molecule,
ozonolyses of olearin and the epoxide (15) failed to yield any isolable products, while
similar results were obtained with permanganate-acetone oxidations of olearin and
anhydroolearin (14). I n addition, the ketone (5) was resistant to Baeyer-Villiger
oxidation with both nz-chloroperbenzoic acid and trifluoroperoxyacetic acid.17
Bromination of olearin with bromine in carbon tetrachloride also proved to be
selective, and gave a high yield of a dibromo derivative (19) from which (1) could be
regenerated with zinc dust in ethanol. The 1l.m.r. spectrum (see Experimental)
of this compound clearly indicated that addition to the 3,4 double bond had occurred.
Although the signal due to H 3 was obviously a narrow multiplet, i t was partly
obscured and a rigorous proof of the C 3 axial bromine configuration was obtained
from the half-height width (5 Hz) of the H 3 resonancelob a t 6 5.02 in the spectrum
of the oxidation product (20) below. Treatment of (19) with ozone, or more satis-
factorily sodium periodate with a trace of potassium permanganate,ls gave the
dibromo acid (20) in fair yield. The infrared spectrum of this substance showed that
it contained a y-lactone ring (v,,, 1760 cm-I), while the n.m.r. spectrum (see
Experimental) indicated the presence of a secondary methyl group and a tertiary
methyl group. Debromination of (20) was examined in an attempt to form the
a,P-unsaturated lactone (21) required for oxidative degradation of ring A . Reaction
with sodium iodide in acetonelggave none of (21), while zinc dust in ethanol, followed
by methylation with diazomethane, furnished only a low yield of non-crystalline
material with the expected spectral features.
A more promising approach to the degradation of (20) was its fragmentation
in dilute alkali to furnish the diene dicarboxylic acid (22) in fair yield and formalde-
hyde which was isolated as its dimedone derivative. The formation of (22) from
(20) no doubt proceeds by a Grob type fragmentation20 depicted in (23), which
shows the favourable trans-diaxial arrangement21 of the bromine and hydroxymethyl
group, followed by dehydrobromination (Scheme 1). Decarboxylative debromination
is another fragmentation theoretically possible for this molecule under these condi-
tions. However, the stereo-electronic arrangement (cis-equatorial, axial) of the two
groups is clearly less favourable for this reaction, and none of the expected product
could be isolated. I n the n.m.r. spectrum of (22) the vinylic proton ( H 3 ) appeared
as a multiplet a t 6 6.64. The chemical shift of the secondary methyl was similar to
17 House, H. O., "Jlodern Synthetic Reactions." p. 105. (Benjamin: Kew York 1965.)
18 Lemieux, R . U., and von Rudloff, E., Can. J. Chem., 1955, 33, 1701; Meinwald, J., and
Gassman, P. G., J. A m . chem. Soo., 1960, 82, 2857.
19 Allinger, N. L., Angyal, S. J.,Eliel, E. L., andMorrison, G. A., "Conformational Analysis."
p. 93. (Interscience: New York 1965.)
20 Grob, C. A., "Proceedings and Discussions of the Kekule Symposium." p. 114. (Butter-
worths: London 1959); Grob, C. A., Qazz. chim. ital., 1962, 92, 902.
21 Grob, C. A., Angew. Chem. i m t . Edn, 1965, 4, 440; Wharton, P. S., and Hiegel, G. A.,
J . org. Chem., 1965, 30, 3254.
 STRUCTURE OF OLEARIN 2627

that for the dibronlo acid (20) whereas the tertiary methyl Ras considerably deshielded,
which is consistent with the proposed structure. The ultraviolet absorption of (22)
(A, 281 nm, E 2407) was in good agreement with values obtained for a number of
methyl esters of cyclohexadiene carboxylic acids.22

Scheme 1

I n further agreement with the assigned structure (22), its dimethyl ester gave
the tetralin dimethyl ester (24) in 95% yield d e n refluxed with dichlorodicyano-
quinone in toluene. I n the n.m.r. spectrum of (24) the $-axial tertiary methyl signal
was a t 6 1.28 and the pattern of signals for the aromatic protons was consistent with
a 1,2,3-trisubstituted benzene. The tetralindicarboxylic acid (25),formed on alkaline
hydrolysis of (24), showed ultraviolet absorption maxima a t 234 and 283 nm ( E 6260
and 1300 respectively) which mas very similar t o t h a t (A,, 235 and 285 nm, E 6000
and 1300 respectively) of 5,6,7,8-tetrahydronaphth-1-oic acid.23 Additional evidence
for the structures proposed for the dicarboxylic acids (22) and (24) was obtained
from a palladium-on-charcoal dehydrogenation of (22), which afforded 1,2-dimethyl-
naphthalene. The point of attachment of one carboxylic acid group was indicated
by the formation of 1,2,5-trimethylnaphthalene together with 1,2-dimethylnaphthalene
when a similar dehydrogenation was performed on the non-crystalline diol prepared
by lithium aluminium hydride reduction of (24).

On biogenetic grounds a trans A/B ring junction would be predicted for olearin,
but, while a number of diterpenes with a rearranged labdane skeleton (e.g. clerodin24
and cascarillin6~7)have been shown t o possess this stereochemistry, there are com-
pounds such as columbinl6 and thelepogine25 which have cis-fused rings. From a
study of the 100-MHz n.m.r. spectrum of olearin acetate i t was possible t o show
that olearin was a member of the first group and had the relative stereochemistry
a t positions 5, 9, and 10 shown in (1). The C 18 protons appeared as an AB quartet
22Berchtold, G. A., Ciabattoni, J., and Tunick, A. A., J. org. Chem., 1965, 30, 3679.
23Fenton, S. W., De Wald, A. E., and Arnold, R. T., J . A m . chem. Soc., 1955, 77, 979.
24 Barton, D. H. R., Cheung, H. T., Cross, A. D., Jackman, L. If., and Martin-Smith, M.,
J . chem. Soc., 1961, 5061.
25 Fridriohsons, J., and Mathieson, A. McL., Acta crystallogr., 1963, 16, 206; Tetrahedron
Lett., 1960, No. 26, 18.
2628 J. T. PINHEY, R. F. SIMPSON, AND I. L. BATEY

(J8 Hz) centred a t 6 4.10, although further long-range coupling was evident in the
upfield proton (H 18a) (see structure (A))a t
u 6 3.92, which a t high resolution showed a
I I
splitting of 1- 5 Hz. From an examination
of Dreiding models constructed with both
cis- and ~ Y G ~ S - A / B ring junctions it was
clear t h a t four-bond ("W") coupling26 of
significance was only possible between
H18a and H 6 axial when the A/B ring -
junction was trans (as in (A)). The angles
made by the H 18a-C 18 bond and the H6ax-C6 bond with the C 18-C5-C6 plane
in (A) are approximately 205" and 180" respectively, for which a coupling constant
of about 1 . 2 Hz would be expected.26 I n accord with the expectation t h a t the
side-chain a t C9 was equatorial, IVH for the C9 methyl signal a t 6 0 . 6 5 was 1 . 2 Hz
a t a resolution a t which the signal for tetramethylsilane had WH 0.45 Hz, which
indicated an axial configuration.27

To investigate the absolute stereochemistry of olearin an attempt was made

to reach a compound of type (26)for 0.r.d. measurement. The projected route involved
a reduction of the 3,4-epoxide (15) t o the 3-hydroxy compound which we thought
might not dehydrate due to strain in the compounds with a 3,4 double bond. How-
ever, chromous acetate28 and chromous chloride28 reductions gave anhydroolearin
as the only isolable product. Olearin was eventually shown t o have the absolute
stereochemistry displayed in (1)by conversion into the carboxylic acid (27) obtained
by Jefferies and Payne29 in structural studies on the Dodonaea diterpene (29). Hydro-
genolysis of the dibromo acid (20) over palladium-on-charcoal and calcium carbonate
gave the non-crystalline acid lactone (28) which was characterized as the amide.
Extension of the side-chain of (28) by an Arndt-Eistert reaction30 yielded the methyl
26 Barfield, X., J , chem. Phys., 1964, 41, 3825, and references therein.
27 Shoppee, C. W., Johnson, F. P., Lack, R. E., and Sternhell, S., Chern. Cow~murz.,1965,
347; Ohochuku, N. S., and Powell, J. W., Chem. Commurz., 1966, 422, and references
therein.
28 Cole, W., and Julian, P. L., J. org. Chem., 1954, 19, 131.
29 Jefferies, P . R., and Payne, T. G., Tetrahedron, Lett., 1967, 4777.
30 Polyakova, A. BL, Preobrazhenskii, N. A., and Preobrazhenskii, V. A., Ber. dt, chern.
Ges., 1936, 69, 1314.
 STRUCTURE OF OLEARIN 2629

ester (30)29* which could be hydrolysed to the carboxylic acid (27).29 Both (27) and
(30) were identical with samples kindly provided by Professor P. R. Jefferies.

Nelting points were determined on a Kofler hot stage and are uncorrected. Optical rotations
were measured in chloroform unless otherwise specified. Ultraviolet spectra were determined in
purified ethanol on a Perkin-Elmer 4000A spectrophotometer, and infrared spectra were recorded
on a Perkin-Elmer 221 spectrophotometer. S.m.r. spectra were obtained with either a Varian
A60 or a Varian HA100 instrument using approximately 10% solutions in deuterochloroform
unless otherwise specified. N.m.r. data are given in the following manner: chemical shifts (6)
are in p.p.m. from internal reference tetramethylsilane. Multiplicity: s, singlet; d, doublet;
t, triplet; q, quartet; dq, doublet of quartets; dd, doublet of doublets; m, multiplet; b, broad;
(exch.), exchanges on shaking with DzO. Relative intensities are given in numbers of protons,
e.g. 3H denotes a relative intensity of three protons. Mass spectra were run on a GEC-AEI
MS902 instrument operated a t 70 eV. The peak intensity is given as a percentage of the base peak.
Silica gel refers to Davidson's grade 923 (100-200 mesh), while alumina refers to Peter
Spence type H and neutral alumina to Woelm (grade 111). Thin-layer chromatography (t.1.c.)
and preparative layer chromatography were carried out on silica gel plates. Analytical gas-liquid
chromatography (g.1.c.) was conducted with an F&Id 402 gas chromatograph and preparative
g.1.c. was carried out using a n Aerograph Autoprep 705 instrument.
Light petroleum refers to the fraction of b.p. 60-80". Organic extracts were dried over
anhydrous magnesium sulphate. Analyses were performed by the Australian Microanalytical
Service, Melbourne.

Isolation of Olearin and Other Compounds

The dried milled aerial part (12.3 kg) of Olearia heterocarpa S. T. Blake collected a t Whian
TVhian National Park, Queensland, was exhausted by cold percolation with chloroform and
methanol and the extracts concentrated. Ether (500 ml) was added to the chloroform extract
(630 g) and the mixture kept in the refrigerator for several weeks. The waxy solid which deposited
was collected and triturated with a large volume of hot light petroleum to remove waxes. After
filtration of the residue in chloroform through a short column of alumina and several crystalliza-
tions fromethylacetate there was obtained colourless needles (14.24 g) of olearin (I),m.p. 212-213",
[a]n - 120" (c. 1.0) (Found: C, 69.5: H, 7.6. C20H2605 requires C, 69.3; H , 7.6%). A 214 nm
( E 21000); vmax (Sujol) 3379, 1787, 1740, and 1654 om-1; n.m.r. spectrum: 6 0.62 (3H, s ;
C9-CH3), 0.80 (3H, d, J 6 Hz; C 8-C&), 3.45 (1H (exch.),d, J 5 Hz; OH), 3.80 ( l H , d, J 8 Hz;
H 18a), 4.27 ( l H , d, J 8 Hz; R 18b), 4.86 and 4.89 (2H, ABq, J 18 Hz; C 16 protons), 4.86
(partly obscured) ( l H , m ; H 12), 5.90 (lH, fine m ; H 14), and 6.71 ( I H , m ; H 3 ) ; mass spec-
trum: m/e 346 (3), 345 (15), 344 (66), 343 (3), 331 (5), 330 (20), 329 (loo), 328 (5),283 (4), 211 (13),
and 183 (10).
Part of the material (2.35 g) remaining after removal of olearin was partitioned between
methanol (300 ml)-water (100 ml) and benzene (100 m1)-carbon tetrachloride (100 ml). After
several crystallizations from methanol the material from the latter fraction yielded small needles
of a hydroxy acid, m.p. 263-264", [aID +77" (c, 1.7 in pyridine) (Found: C, 78.2; H, 10.5.
C27H4203 requires C, 78.2; H, 10.2%). Amax 284 ( 6 278), and 310shnm (€217); vmax (Nujol)
3390 and 1682 cm-1; n.m.r. spectrum (DMSO-d6): 6 0.6-2.3 (38H, complex envelope of CH,
CH2, and CH3 groups), 2.98 ( l H , m), 5.12 ( l H , m), 5.43 ( l H , bm), and 12.0 (IH, b).
Treatment of the hydroxy acid (300 mg) with acetic anhydride (3 ml) and pyridine (3 ml)
a t room temperature gave an acetyl derivative, colourless prisms from ethyl acetate-light petroleum,
m.p. 291-292", [ a ] ~+57O (c, 0.84 in pyridine) (Found: C, 76.5; H , 9.8. C29H4404 requires
C, 76.3; H, 9.7%). Amax 284 and 310 nm ( E 278 and 217 respectively) ; vmax (CHC13) 1713, 1695,
and 1265 om-1; n.m.r. spectrum (DMSO-d6): 6 0.7-2.3 (37H, complex envelope of CH, CH2,
* Personal communication from Professor P. R. Jefferies. Compound (27) prepared during
earlier work29 was found to have m.p. 84-87', [aID -5'.
 2630 J. T. PINHEY, R. F. SIMPSON, AND I. L. BATEY

and CH3 groups), 1.99 (3H, s; CH&OO), 4-40 (2H, m), 5.17 ( l H , m), and 12.50 ( l H , b ) ; mass
spectrum: rn/e 456, 438, and 248 (base peak).
I n another attempt a t separating the material extracted with chloroform, a portion of the
residue (10 g) was chromatographed on neutral alumina (200 g). The benzene eluate afforded
material which crystallized from benzene in oolourless needles, m.p. 84-85' (Found: C, 82.0;
H, 13.8. CsoHlooO2 requires C, 81.9; H, 13.8%. C52H10402 requires C, 82.0; H, 13.8%).
vmax 1724, 1170, 730 and 720 om-1; n.m.r. spectrum: 6 0.94 (9H, m), 1.26 (approx. 87H, bs),
2 - 2 8 (2H, m), and 4.04 (2H, m ) ; mass spectrum: m/e 788, 774, 760, 744, 732, 718, 704, 690, 676.
The above mixture of esters (13 mg) was heated under reflux for 3 hr in 50% ethanol
(9 ml) containing sodium hydroxide (0.5 g). The mixture was acidified and the precipitated
product collected and examined by mass spectroscopy. The spectrum showed peaks (mle) a t 448,
420, 396, 368, 364, 340, and 336.
The combined residues (400 g) from the chloroform extract of 0. heterocarpa were dissolved
in ether (1 1.) and extracted in turn with 5% sodium bicarbonate solution (2 x 200 ml), 5% sodium
carbonate solution (2 x 200 ml), and 5% sodium hydroxide solution (2 x 200 ml). Addition of
ethanol to the sodium hydroxide extracted material (10 g) deposited a yellow solid which crystal-
lized from ethanol to give xanthomicrol5 (5,4'-dihydroxy-6,7,8-trimethoxyflavone)as pale
yellow needles, m.p. 231-233' ( l k 5 227-230") (Found: C, 62.6; H, 4.9. Calc. for C18H1607:
C, 62.8; H, 4.7%). Amax 234, 282, and 292sh ( E 25320, 22240, and 18510); vmax (Nujol) 3270,
1651, 1649, 1600, 1585, and 1559; n.m.r. spectrum (DMSO-ds): 6 3.52 ( l H , s), 3.88 (3H, s),
3.97 (3H, s), 4.08 (3H, s), 6.80 ( l H , s), AA'BB' system centred a t 7.03 and 7.93 (4H, JAB
9 Hz), 10.39 (0.5H, b), and 12.80 (0.5H, s ) ; mass spectrum: m/e 345 (22), 344 (84), 330 (20),
329 (loo), 211 (lo), 183 (5), and 172 (5).
The diacetate, prepared according to the method of Stout and Stout,5 was obtained as
colourless needles, m.p. 124-125" (lit.5 127-129") (Found: C, 61.8; H, 5.0. Calc. for CzzHzo09:
C, 61.7; H, 4.7%).
The monomethyl ether* of xanthomicrol was obtained by the method of Stout and Stout5
as pale yellow needles (methanol), m.p. 177-178" (undepressed on admixture with authentic
material kindly provided by Professor G. H. Stout).

Acetylolearin ( 3 )
Olearin (106 mg) in pyridine (1 ml) containing acetic anhydride (1 ml) was kept a t room
temperature overnight. The acidified reaction mixture was extracted with chloroform to afford
material which, after filtration in chloroform-benzene (1 : 1) through a short column of neutral
alumina, crystallized from ethyl acetate-light petroleum to give acetylolearin (3) in colourless
needles, m.p. 193-195', [ a ] n - 121" (c, 1.0) (Found: C, 68.2; H, 7.3. CzzHz806 requires
C, 68.0; H, 7.3%). X211nm ( E 19000); vmax (CHC13) 1775, 1748, 1655, and 1635om-1; v,,,,
(Nujol) 1785, 1771, 1754, 1736, 1657, 1627, and 1238 cm-l; n.m.r. spectrum: 60.65 (3H, s ;
C9-CH3), 0.87 (3H, d, J 6 Hz; C8-CH3), 2.14 (3H, s ; OCOCH3), 3.82 ( l H , part of ABq,
JAR 8 H Z ; H lsa), 4.28 ( l H , part of ABq, JAB 8 H Z ; H 18b), 4.82 (2H, long-range coupled
AB quartet (see Discussion), C 16 protons), 5.75 ( l H , dd, J 9.2 and 2.5 Hz; H 12), 5.97 ( l H ,
fine m ; H 14), and 6.75 ( l H , dd, J 2 a s , 3 7 . 1 HZ, Jzeq,32.4 Hz; H 3 ) .

Tetrahydroolearin (2)
Olearin (132 mg) in ethyl acetate (25 ml) was hydrogenated over Adams catalyst (16 mg)
a t room temperature and atmospheric pressure until the uptake of hydrogen had ceased. The
product crystallized from ethyl acetate-light petroleum to give tetrahydroolearin (2) as colourless
needles (110 mg), m.p. 174-175", [ o l ] ~- 6.7' (c, 0.9) (Found: C, 68.5; H, 8.5. C20H300j requires
C, 68.5; H, 8.6%). X 214 nm ( E 132); vmax (CHC13) 3490, 1769, and 1759 om-1; vmaX (Nujol)
3427, 1776, 1734, and 1702 om-I; n.m.r. spectrum: 6 0.56 (3H, s ; C9-CH3), 0.79 (3H, d,
J 6 Hz; C8-CHs), 2.55 (3H, m), 3.12 (1H (exch.), d, J 6 Hz; OH), 3.83 ( l H , m ; H12), an
* This compound, named tangeretin (m.p. 176-177"), has been isolated from a Citrus
species.31
31Chaliha, B. P., Sastry, G . P., and Rao, P. R., Tetrahedron, 1965, 21, 1441.
 STRUCTURE OF OLEARIN 2631

AB quartet (partly obscured) a t 4.23 and 4.29 (2H; C 18 protons) and 4.30 (2H, partly obscured
bm; C 16 protons).
Tetrahydroolearin failed to give a colour reaction with C(N02)4 in chloroform.

Acetylolearin (100 mg) in ethyl acetate (30 ml) was hydrogenated as above to give tetra-
hydroacetylolearin (4), which crystallized from ethyl acetate-light petroleum in colourless needles
+
(96 mg), m.p. 221-222", [ a r ] ~ 1.9' (c, 1.1) (Found: C, 67.3; H, 8.1. C22H3206 requires
C, 67.3; H, 8.2%). X 209 nm ( E 235): vmax (Nujol) 1773, 1754, 1724, and 1229 cm-1; n.m.r.
spectrum: 6 0.56 (3H, s; C9-CHs), 0.82 (3H, d, J 6 Hz; C8-CH3), 2.06 (3H, s; OCOCH3),
2.54 (3H, m), two overlapping AB quartets with centres of doublets a t 4.12, 4.19, 4.23, and 4.30
(4H, C 16 and C 18 protons), and 5.29 ( l H , m ; H 12).
Acetylation of tetrahydroolearin as for (1) above also afforded the acetyl derivate (4).

A solution of chromium trioxide (26 mg) in water (0.1 ml) and acetic acid (0.1 ml) was
added to tetrahydroolearin (119 mg) in acetic acid (2 ml) a t room temperature and then heated
a t 50" for 15 min. The reaction mixture was diluted w ~ t hwater and extracted with chloroform
to yield dehydrotetrahydroolearin (5), which crystallized from ethyl acetate-light petroleum in
oolourless needles (113 mg), m.p. 192-194", [a]= +3.6' ( c , 1.1) (Found: C, 69.0; H , 8.1.
CzoHzs06 requires C, 68.9; H, 8.1%). Xmax 288 ( E 42); vmax (Nujol) 1770, 1747, and 1705 om-1;
n.m.r. spectrum: 6 0.60 (3H, s ; C9-CH3), 0.85 (3H, d, J 6 Hz; C8-CH3), 2.72 ( l H , m), 3.52
( l H , m), 4.17 and 4.22 (2H, ABq, JAB9 Hz; H 18a and H 18b), and 4.32 (2H, m ; C 16 protons).

Attempted Partial Hydrogenation of Olearin

(A) Olearin (71 mg) in ethyl acetate (5 ml) was hydrogenated a t atmospheric pressure
over 5% palladium-on-charcoal (14 mg) until 1.2 mol of hydrogen had been taken up. A t.1.c.
examination indicated that in addition to olearin and tetrahydroolearin there was a third com-
pound of intermediate polarity present. Although accounting for approximately 50% of the
mixture, this compound could not be obtained pure even by preparative t.1.c. The n.m.r. spectrum
of the mixture showed a strong absorption a t 6 5.90, while the signal a t 6 6.71 in the spectrum
of olearin was quite weak.
(B) Olearin (40 mg) in benzene (20 ml) was hydrogenated a t atmospheric pressure over
Raney nickel. The uptake of hydrogen was, however, very rapid and the only product isolated
was tetrahydroolearin.

Action of Allcali o n (A) Dehydrotetrahydroolearin (j),(B) Tetrahydroolearin (Z), and (c) Olearin
(A) Dehydrotetrahydroolearin (42 mg) was dissolved in 15% aqueous sodium hydroxide
solution (10 ml) and the reaction mixture distilled slowly mto 2,4-dinitrophenylhydrazine-per-
ohloric acid reagent. The fine precipitate which formed on standing was collected and crystallized
from benzene-light petroleum to give formaldehyde 2,4-dinitropheuylhydrazone(3 mg, lo%),
m.p. and mixed m.p. 161-162', identical with authentic material by paper chromatography.32
(B) When tetrahydroolearin (45 mg) was treated as above there was no formation of a
DNP derivative.
(c) Treatment of olearin (50 mg) as in (A) failed to produce a DNP derivative.

Isotetrahydroolearin ( 7 )
A saturated solution of sodium bicarbonate (16 ml) was added to tetrahydroolearin (120 mg)
in 75% ethanol (21 ml) and the solution heated a t reflux for 1 hr. The solution was then acidified
with dilute sulphuric acid and extracted with chloroform. The product crystallized from ethyl
acetate-light petroleum to give isotetrahydroolearirt (7) as colourless needles (90 mg, 75%), m.p.
178-17g0, [ a ] -57'
~ (c, 1.0) (Found: C, 68.5; H , 8.6. C20H3005 requires C, 68.5; H, 8.6%).
32 Huelin, F. E., Aust. J . scient. Res. ( B ) ,1952, 5, 328.
2632 J. T. PINHEY, R. F. SIMPSON, AND I. L. BATEY

X213nm ( E 156); vmax (CHC13) 3640, 1773, and 1765cm-1; vmax (Nujol) 3442, 1771, and
1750 om-1; n.m.r. spectrum: 6 0.60 (3H, s; C9-CH3), 0.83 (3H, d, J 6 Hz; C8-CH3), 2.84
(1H (exch.), b ; OH), 3.65 (2H, bd which collapsed to a sharp doublet on exch.; C 16 protons),
4.20 and 4.30 (2H, ABq, JAB 9 H Z ; C 18 protons), and 4.48 ( l H , m ; H 12).
Isotetrahydroolearin (7) may also be obtained in good yield by heating tetrahydroolearin
(2) with 5~ hydrochloric acid in dioxan a t 100' for 16 hr.
The acetyl derivative of (7), prepared in the same way as acetylolearin, crystallized from
ethyl acetate-light petroleum in colourless needles, m.p. 157-158", [a]= -38' (0, 1.2) (Found:
C. 67.4; H, 8.5. C22H3206 requires C, 67.3; H, 8.2%). X213nm (6 189); vmax (Nujol) 1788,
1771, 1756, 1749sh, and 1243 cm-1.

Treatment of Tetrahydroacetylolearin (4) with Sodium Bicarbonate

To a solution of tetrahydroacetylolearin (10 mg) in 50% aqueous ethanol (9 ml) a saturated
aqueous solution of sodium bicarbonate (5 ml) was added and the mixture heated a t reflux for
45 min. The ethanol u7as removed under reduced pressure. the solution acidified with dilute
sulphuric acid and then extracted with chloroform. The residue was crystallized from ethyl
acetate-light petroleum to give (4), and shown to be identical with the starting material (mixed
m.p. and infrared spectrum).

Ozidation of Isotetrahydroolearin to the Carboxylic Acid (8)

The oxidation of isotetrahydroolearin (55 mg) with chromium trioxide (34 mg) in acetic
acid (2 ml) was carried out as for the preparation of dehydrotetrahydroolearin (5) above. Crystal-
lization of the product from ethyl acetatelight petroleum gave the carbozylic acid (8) as colourless
needles (51 mg, go%), m.p. 99-100' (with decomposition due to ethyl acetate of crystallization,
which could be removed by heating under vacuum a t 80°), [ o l ] ~-72' (c, 1.0) (Found: C, 65.6;
H, 8.1. CzoHzsOe requires C, 65.9; H, 7.7%). h 216 ( c 175); vmax (Sujol) 3120 (broad), 1789,
1762, 1745 (ethyl acetate), and 1730 (broad sh) cm-1; n.m.r. spectrum: 6 0.60 (3H, s ; C 9-CH3),
0.83 (3H, d, J 6 Hz; C 8-CH3), 2.88 (3H, m), 4.21 and 4.29 (2H, ABq, J A B 9 Hz; C 18 protons);
4.66 ( l H , m; H 12), and 10.86 (1H (exch.) b; COOH).

Action of Acid on Olearirz

Hydrochloric acid ( 4 ~ 15 , ml) was added to olearin (165 mg) in dioxan (13 ml) and the
solution heated a t reflux for 17 hr. Most of the dioxan was removed under reduced pressure, the
mixture extracted with chloroform, and the residue separated into neutral and acidic fractions
by means of sodium bicarbonate. The acidic material crystallized from ethyl acetate-light
petroleum to give the aldehyde acid (9) as colourless needles (75 mg, 46%), m.p. 170-l'i2°, [a]=
-89' (c, 1.6) (Found: C, 68.9; H, 7.5. CzoHz605 requires C, 69.3; H, 7.6%). hmax 237 nm
( E 10000); vmax (Nujol) 3000 (very broad), 1750, 1730, 1693, 1668, and 1643 cm-1; n.m.r. spec-
trum: 6 0.72 (3H, s ; C 9-CH3), 0.88 (3H, d, J 6 Hz; C 8-CH3), 3.36 (2H, s; C 14 protons),
3.92 and 4.26 (2H, ABq, J A 9~H Z ; C 18 protons), 6.73 (2H, m ; H 3 and H 12), 9.57 ( l H , s ;
CHO), and 10.36 (1H (exch.), b ; COOH).
Treatment of the aldehyde acid (9) in methanol with diazomethane in ether a t 0' afforded
the methyl ester (10) (Amax 224 nm ( 6 10040)),an oil, which readily gave a 2,4-dinitrophenylhydra-
zone derivative, orange needles (80% yield) from benzene, m.p. 158-160' (Found: C, 59.8;
H, 6.2. Cg7H32N408 requires C, 60.0; H , 6.0%). Amax 251, 286, and 376 nm (c 15500, 8900,
and 27900 respectively); vmax (Sujol) 3270, 1760, and 1730 om-1.
When olearin (84 mg) was heated a t reflux for 3 hr in methanolic hydrogen chloride, and the
solvent removed under reduced pressure, the above methyl ester (59 mg, 66%) was obtained.
The identity was confirmed by conversion into the 2,4-dinitrophenylhydrazone,m.p. and mixed
m.p. 158-160".

Methanesulphonate of Tetrahydroolearin
Methanesulphonyl chloride (1 ml) was added to a n ice-cold solution of tetrahydroolearin
(100 mg) in pyridine (1 ml) and benzene ( 1 . 5 ml), and the mixture kept a t 0' overnight. The
reaction mixture was poured onto ice and extracted with chloroform to give tetrahyd~oolearin
 STRUCTURE OF OLEARIN 2633

methanesulphonate (12), which crystallized from ethyl acetate-light petroleum as colourless

needles (110 mg), m.p. 149-15O0, [MID - 11' (c, 1 . 1 in pyridine) (Found: C, 59.1; H, 7.6; S, 7.5.
C21H3207S requires C, 58.9; H, 7.5; S, 7.3%). X 213 nm ( e 159); vmax (Nujol) 1778sh, 1767,
1325, 1181, and 1163 om-1.

Olearin Methunesulphonate (13)

Methanesulphonyl chloride (2 ml) was added to an ice-cold solution of olearin (334 mg)
in pyridine (2 ml) and benzene (2 ml), and the mixture kept a t 0' overnight. The reaction mixture
was poured onto ice and extracted with chloroform. The product crystallized from ethyl acetate
to give olearin methanesulphonate (13) as colourless needles (357 mg), m.p. 172-173O, [MID -126'
( c , 1.5) (Found: C, 59.2; H, 6.8; S, 7.6. CzlHz8O7Srequires C, 59.4; H, 6.7; S, 7.5%).
X 220n.m ( e 14550); vmax (Nujol) 1770sh, 1742, 1729sh, 1656, 1637, 1346, 1176, and 1154 cm-1;
n.m.r. spectrum: 6 0.67 (3H, s; C9-CH3), 0.85 (3H, d, J 6 Hz; C8-CH3), 3.08 (3H, s ;
OSO&H3), 3.95 and 4.29 (2H, ABq, J A B 8 H Z ; C 18 protons), 4.96 and 5.01 (2H, ABq, J A B 17.5
Hz; C 16 protons), 5.75 ( l H , d d ; H 12), 6.21 ( l H , narrow m ; H 14), and 6.77 ( l H , dd; H 3 ) .

Anhydroolearin (14)
The above methanesulphonate (13) (115 mg) was heated a t 120-125' for 75 min in dimethyl
sulphoxide (3 ml). The reaction mixture was diluted with water and extracted with chloroform
to afford a non-crystalline product (85 mg, 96%), a single compound by t.l.c., [aID - 104' (c, 0.9);
Xmax 259111x1 ( E 21510); vm,, (CHC13) 1761, 1745, 1640, and 1588 om-1; n.m.r. spectrum: 6 0.77
(3H, d (partly obscured); C8-CHa), 0.82 (3H, s ; C9-CH3), 3.95 and 4.33 (2H, ABq, J A B 8 H Z ;
C 18 protons), 4.99 and 5.01 (2H, ABq, J A B 16 Hz; C 16 protons), 5.93 ( l H , narrow m ; H 14),
5.94 and 6.34 (2H, ABq, J A B 16.8Hz; H12 and H 1 1 respectively), and 6.75 ( l H , d d ; H3).

Epoxidation of Anhydroolearin,
Anhydroolearin (14) (200 mg) in chloroform (8 ml) was treated with m-chloroperbenzoic
acid (82%, 352 mg) and the mixture kept a t room temperature for 68 hr. The reaction mixture
was diluted with chloroform and washed with dilute solutions of potassium iodide, sodium thio-
sulphate, and sodium bicarbonate. Removal of the solvent yielded a residue which crystallized
from ethyl acetate to give 3j3,4-epoxyanhydroolearin(15) as colourless needles (167 mg, go%),
m.p. 255-256', [a]D -53" (c, 1.1 in pyridine) (Found: C, 69.5; H, 6.9. C20H2405requires
C, 69.8; H, 7.0%). Amax 259 nm ( E 26790); vmax (Sujol) 1793, 1740, 1647, and 1596 om-1;
n.m.r. spectrum (DMSO-d6): 6 0.75 (3H, d partly obscured; C8-CH3), 0.77 (3H, s ; C9-CH3),
3.38 ( l H , bd, J 3 , z a 6 Hz; H 3 ) , 4.44 (2H, approximate s ; C 18 protons), 4.94 and 4.97 (2H,
ABq, J A B 17 Hz; C 16protons), 5.82 and 6.30 (2H, ABq, J A B 16 Hz; H 12 a n d H 11 respectively),
and 5.92 ( l H , narrow m ; H 14).

Performic Acid Oxidation of Acetylolearin (3)

To a solution of acetylolearin (170 mg) in 99% formic acid (1.7 ml) was added hydrogen
peroxide (87% w/w, 112 mg) and the solution was kept a t room temperature for 22 hr. The
reaction mixture was poured into water, extracted with a large volume of chloroform, and the
chloroform solution washed with water until the washings were free of peroxide. Crystallization
of the product from ethyl acetate-light petroleum gave the diol monoformate (16) as colourless
needles (190 mg, 97%), m.p. 239-240°, [&ID -37' (c, 1.2 in pyridine) (Found: C, 61.7; H , 6.6.
C23H3009 requires C, 61.3; H, 6.7%). h 215 nm ( e 10200); vmax (Nujol) 3483, 1780, 1744, 1718,
1642, and 1248 om-1; n.m.r. spectrum (DMSO-d6): 6 0.57 (3H, s; C 9-CH3), 0.77 (3H, d, J 6 Hz;
C8-CHs), 2.08 (3H, s ; CHaCOO), 4.26 and 4.31 (2H, ABq, J a B 9 Hz; C 18 protons), 4.93 and
4.95 (2H, ABq, J A B 17 H Z ; C l 6 protons), 5.00 ( l H , m ; H 3 ) , 5.65 ( l H , m ; H12), 6.10 ( l H ,
narrow m ; H 14), 6.19 (1H (exch.), s ; OH), and 8.18 ( l H , s ; OCHO).

Hydrolysis of the Diol Monoformate (16)

Sulphuric acid (lo%, 1.5 ml) was added to the diol monoformate (46 mg) in 95% ethanol
(3 ml) and the mixture heated on the water bath for 12 min. Dilution with water and extraction
2034 J. T. PINHEY, R. F . SIMPSON, AND I. L. BATEY

with a large volume of chloroform afforded material which crystallized from chloroform to give
the diol (17) as colourless needles (25 mg), m.p. 245-246", [a]= -41' (c, 1.1 in pyridine) (Found:
C, 62.2; H, 7.3. CzzH3oOs requires C, 62.5; H, 7.2%). h 210nm ( e 12730); ,v (Nujol) 3455,
3438sh, 1782, 1742, 1709, 1635, and 1243 om-1; n.m.r. spectrum (DNSO-ds): 6 0.55 (3H, s ;
C9-CH3), 0.77 (3H, d, J 6 H z ; C8-CH3), 2.07 (3H, s ; 0COCH3), 3.65 (IH, m ; H 3 ) , 4.01
and 4.34 (2H, ABq, J A B8 H Z ; C 18 protons), 4.89 (2H, bs; C 16 protons), 5.12 (1H (exch.),
d, J 5 Hz; C 3-OH), 5.39 (1H (exch.), s; C P O H ) , 5.61 ( l H , m ; H 12), and 6.06 (IH, narrow m ;
H 14).

Hydrogenation of the Diol Monoformate ( 1 6 )

The diol monoformate (16) !189 mg) was hydrogenated over Adams catalyst (26 mg) in
ethyl acetate (30 ml). The product crystallized from ethyl acetate-light petroleum to give the
dihydro diol monoformate (18) as colourless needles, m.p. 207-20g0, [a]= -24" (c, 1.4 in pyridine)
(Found: C, 61.0; H, 7.1. CZ3H32o9requires C, 61.1; H, 7.1%). X 220 nm (e 262); vma, (Nujol)
3399, 1773, 1752, 1731, 1708, and 1239 om-I; n.m.r. spectrum (DMSO-as): 60.54 (3H, s;
C9-CH3), 0.77 (3H, d, J 6 Hz; C8-CH3), 2.04 (3H, s ; OCOCHs), 4.14 (2H, m ; C 16 protons),
4.26 and 4.30 (2H, ABq, JAB 8 H Z ; C 18 protons), 5.10 (2H, m ; H 3 and H 12), 6.17 (1H (exch.),
s ; C4-OH), and 8.18 ( l H , s; OCHO).
Attempted hydrolysis of (18) to the dihydro diol, under the conditions used above for
hydrolysis of (16) to (17), afforded a mixture which mas inseparable by preparative layer
chromatography.

Ozonolyses of ( A ) Olearin and (B) Anhydroolearin Epoxide (5)

(A) Olearin (48 mg) in ethyl acetate (20 ml) a t -50' was treated with ozonized oxygen
until t.1.c. indicated that no starting materials remained. The temperature of the solution was
allowed to rise to 0' and was then treated with hydrogen in the presence of 10% palladium-on-
charcoal. Removal of the catalyst and the solvent afforded a non-crystalline residue which was
separated (sodium bicarbonate) into neutral and acidic fractions. Both of these were shown to
be complex mixtures by t.1.c.
(B) Anhydroolearin epoxide (22.5 mg) in ethyl acet'ate (10 ml) was treated with ozonized
oxygen a t - 78' until t.1.c. indicated that no starting material remained. The reaction mixture
was worked up as in (A) above to yield acidic and neutral fractions which t.1.c. indicated to be
complex mixtures.

Attempted Baeyer-Villiger Oxidation of Dehydrotetrahydroolearin (5)

(A) A solution of the ketone (5) (105 mg) in chloroform (3 ml) containing m-ohloroperbenzoic
acid (82%, 125 mg) wa,s kept a t room temperature for 90 hr. Working up in the usual way (see
above) yielded only unchanged starting material (identical by m.p. and infrared).
( B ) Approximately 0 . 0 0 8 ~
trifluoroperacetic acid was prepared by adding trifluoroacetic
anhydride (3 ml) during 10 min to a vigorously stirred solution of hydrogen peroxide (87%,
0.3 ml) in chloroform (10 ml) a t 5-10'.
A volume of the reagent containing trifluoroperacetic acid (0.139 ml) was added dropwise
with stirring over 10 min to the ketone (6) (100 mg) and sodium hydrogen phosphate (93 mg)
in chloroform (4 ml). The mixture was heated a t 30" with stirring for a further 80 min, diluted
with water, and extracted with chloroform. The chloroform extract was washed with water until
free of acid, and the solvent removed to give a high recovery of starting material (m.p. and infrared
identical).

Dibromoolearin ( 1 9 )
Bromine (123 mg) in carbon tetrachloride (1.7 ml) was added slowly to olearin (222 mg)
in chloroform (6 ml) a t room temperature over 30 min. The solvent was removed under reduced
pressure and the residue filtered in chloroform through a short column of neutral alumina. The
eluted material crystallized from ethyl acetate to give dibromoolearin (19) as colourless needles
(280 mg, go%), m.p. 206-207O, [a]* -21' (c, 1 . 1 in pyridine) (Found: C, 47.4; H, 5.3.
CzoHa6BrzOj requires C, 47.5; H, 5.2%). X 210 nm ( e 15460); vmax (Nujol) 3500, 1778, 1531,
 STRUCTURE OF OLEARIS 2635

and 1626 cm-l; n.m.r. spectrum (DMSO-ds): 6 0.64 (3H, s ; C9-CH3), 0.80 (3H, d, J 6 Hz;
C 8-CH3), 4.60 ( l H , m ; H 12), 4.62 and 4.66 (2H, ABq, JAB 9 Hz; C 18 protons), 4.94 and
4.97 (2H, ABq; C 16 protons), 6.06 ( l H , m ; H 3 ) , and 6.00 ( l H , narrow m; H 14).
Dibromoolearin (10 mg) was heated a t reflux for 6 min with zinc dust (30 mg) in 95%
ethanol. Removal of the zinc and solvent yielded a residue which crystallized from ethyl acetate
to give olearin (identified by mixed m.p. and infrared).

Ozidation of Dibromoolearin to the Dibromo Acid (20)

(A) O~ono1y~i~.-Dibr0moo1earin(184 mg) in ethyl acetate (30 ml) was treated with ozonized
oxygen a t - 50' until t.1.c. indicated that no starting material remained. The reaction mixture
was heated with water for 10 min, during which time the ethyl acetate was removed by steam
distillation. The product was extracted into chloroform and separated (sodium bicarbonate)
into neutral (120 mg) and acidic (65 mg) fractions. Sodium metaperiodate (250 mg) in water
(5 ml) was added to the neutral material (120 mg) in methanol (15 ml) and the mixture kept a t
room temperature overnight. Most of the methanol was removed under reduced pressure, the
product extracted into chloroform, and separated into neutral and acidic fractions. The acidic
material crystallized from ethyl acetate-light petroleum to give the dibromo acid (20) as colourless
needles (33 mg), m.p. 230-232", [aID +3O (c, 0.7 in pyridine) (Found: C, 44.2; H , 5.3.
Cl~HzzBr204requires C, 43.9; H, 5.3%). X 222 nm ( E 9830); vmsx (Nujol) 3195 (broad), 1760,
and 1716 cm-1; n.m.r. spectrum (DMSO-d6): 6 0.65 (3H, s; C9-CHs), 0.87 (3H, d, J 6 Hz;
C 8-CHs), 2.30 (2H, bs; C 11 protons), 4.59 and 4.64 (2H, ABq, J A B 9 H Z ; C 18 protons), and
5.02 ( I H , narrow m, W E 5 H Z ; H 3 ) .
The residues from the neutral fraction and the acidic fraction were combined and treated
with chromium trioxide (30 mg) in acetic acid (3 ml) a t room temperature for 30 min. The reaction
was worked up in the usual way to yield a further quantity (47 mg) of the dibromo acid (20).
(B) Periodate-permangamte mixture.-Dibromoolearin (250 mg) in dioxan (5 ml) was added
with stirring to a solution of sodium periodate (2.5 g), potassium permanganate (150 mg), and
potassium carbonate (2.0 g) in water (30 ml). The mixture was stirred a t room temperature for
28 hr, and then the permanganate and excess periodate were decomposed by the addition of a dilute
acidic solution of sodium bisulphite. The resulting mixture was extracted with ether and the
extracted material was separated into neutral and acidic fractions by means of sodium bicarbonate.
The acidic material crystallized from ethyl acetate-light petroleum to give the dibromo acid
(20) (120 mg, 56%), m.p. and mixed m.p. 230-232' (infrared identical).

Reaction of the Dibromo Acid (20) with Alkali

The dibromo acid (250 mg) was heated a t reflux for 2 hr in water (25 ml) containing sodium
hydroxide (1.25 g). Part of the reaction mixture (5 ml) was slowly distilled into a saturated
aqueous solution of dimedone to yield the dimedone derivative of formaldehyde (2 mg) (identified
by m.p., mixed m.p., and infrared). The reaction mixture was acidified with dilute sulphuric acid
and extracted with ether. The residue crystallized from ethyl acetate to give the diene dicarboxylic
acid (22)as colourless needles (45 mg, 28%), m.p. 251-252", [ a ]+41°
~ (c, 0.7 inpyridine) (Found:
C, 68.1; H , 7.8. C15H2004 requires C, 68.2; H , 7.6%). Amax 281 nm ( E 2410); v m a x (Nujol)
2950 (very broad), 1688, 1650sh, and 1590 om-1; n.m.r. spectrum (DMSO-de): 6 0.88 (3H, d,
J 6 Hz; C8-CH3), 0.96 (3H, s ; C9-CHs), 2.38 (2H, m), 6.64 ( l H , m ; H 3 ) , and 11.7 (2H, very
broad; 2 x COOH).
The dicarboxylic acid (22) afforded a dimethyl ester with diazomethane in the usual way.
This compound was non-crystalline and was used in the following dehydrogenation without
full characterization. I t was a single compound by t.1.c. and the n.m.r. spectrum showed signals
a t 60.95 (3H, d, J 6 H z ; C8-CHs), 1-02 (3H, s ; C9-CHs), 2.51 (2H, bs; C l l protons), 3.61
(3H, s ; COOCH3), 3.73 (3H, s ; COOCH3), and 6.78 ( l H , m ; H 3 ) .

Dehydrogenation of the Dimethyl Ester of (22) with Dichlorodicyartoquinone

The dimethyl ester of (22) (220 mg) was heated for 4 hr a t reflux in toluene (30 ml) con-
taining 2,3-dichloro-5,6-dicyanoquinone (183 mg). The reaction mixture was diluted with ether
and washed with dilute sodium hydroxide solution (briefly) and water until the washings were
2636 J. T. PINHEY, R. F. SIMPSON, AND I. L. BATEY

colourless. The product (207 mg) which was non+crystallinewas a single compound by t.1.o.
and was shown to be the tetralin dimethyl ester (24)by the n.m.r. spectrum which exhibited signals
a t 8 1.00 (3H, d, J 7 Hz; C 8-CHz), 1.28 (3H, s ; C9-CHs), 2.74 (2H, bs; C 11 protons), 3.12
(2H, m ; C 6 protons), 3.50 (3H, s; COOCHs), 3.87 (SH, 8 ; COOCHs),and 7.04-7.83 (3H,aromat-
ic protons).
The dimethyl ester (24) (207 mg) was heated a t reflux for 1 hr in methanol (5 ml) and
2~ sodium hydroxide (25 ml). Isolation in the usual way gave an acidic fraction which crystallized
from ethyl acetate-light petroleum to give the tetralindicarboxylic acid (25) as colourless needles
(140 mg, 75%), m.p. 180-182", [aIn +105O (c, 1 . 0 in pyridine) (Found: C, 68.9; H, 7.1.
CljH1804 requires C, 68.7; H, 6.9%). Amax 234 and 283 nm ( e 6260 and 1300 respectively);
v,,, (Yujol) 2920 (very broad), 1688, 1628, 1590, 1388, 1268, and 917 om-1.

Palladium Dehydrogenation of the Diene Dicarboxylic Acid (22)

An intimate mixture of the dicarboxylic acid (22) (1.55 g) and 10% palladium-on-charcoal
(1 g) was heated a t 280' for 3 hr. The mixture was then extracted with chloroform and separated
into neutral and acidic fractions by means of sodium bicarbonate. The neutral fraction (350 mg)
was chromatographed on alumina (20 g) to give, on elution with pentane, 1,2-dimethylnaphthalene
(90 mg) (identical by g.1.c. and n.m.r. spectrum). Treatment of part of this material with 1,3,5-
trinitrobenzene in ethanol gave 1,2-dimethylnaphthalene trinitrobenzene adduct, m.p. and
mixed m.p. 144-145'.

Palladium Dehydrogenation of the Product of L i A l H 4 Reduction of (24)

The tetralin dimethyl ester (24) (450 mg) in tetrahydrofuran (30 ml) was added to a suspen-
sion of lithium aluminium hydride (1g) in tetrahydrofuran (20 ml) a t room temperature, and the
mixture was then heated a t reflux for 5 hr. The non-crystalline product (340 mg), which had no
carbonyl absorption in the infrared, was used without further purification in the next step.
A mixture of the above diol (340 mg) and 10% palladium-on-charcoal (0.8 g) was heated
a t 310-320' for 3 hr. The reaction mixture was extracted with chloroform and the product
chromatographed on alumina (10 g). The pentane-eluted material (100 mg) was shown by g.1.c.
(10% Apiezon L on Chromosorb W) to be a mixture of two major components. A separation was
achieved by preparative g.1.c. (10% Apiezon L on Chromosorb W) to give 1,2-dimethylnaphthalene
(28 mg) (m.p. and mixed m.p. of trinitrobenzene adduct 144-145') and 1,2,5trimethylnaphthalene
(13 mg) (m.p. and mixed m.p. of trinitrobenzene adduct 157-158'). Both naphthalenes were
identified further by g.1.c. retention times and infrared spectra of the trinitrobenzene adducts.
Authentic 1,2,5-trimethylnaphthalenewas prepared by the method of Hosking and Ruzicka.33

Reaction of the Epoxide ( 1 5 ) with Chromous Salts

(A) Chromous acetate28 (200 mg) was added to anhydroolearin epoxide (60 mg) in acetone
(3 ml) and acetic acid (3 ml), and the mixture stirred a t room temperature under nitrogen for
60 hr. The mixture was then diluted with water (30 ml) and extracted with chloroform. The
extract was washed with sodium bicarbonate solution and water to give a product consisting of
two components (t.1.c.) separable by preparative layer chromatography to give anhydroolearin
(20 mg) (infrared identical) and the epoxide (10 mg).
( B ) The epoxide (15) (50 mg) in acetic acid (2 ml) was added to a solution of chromous
chloride (0. Six, 2 m1)" under nitrogen and the mixture stirred a t room temperature for 30 min.
The reaction was worked up in the manner described above to give a mixture of a t least two
components (t.1.c.). Separation by preparative layer chromatography afforded anhydroolearin
(15 mg) (infrared identical) and a more polar compound (5 mg) which had no lactone carbonyl
absorption in the infrared.

Hydrogenolysis of Dibromo Acid (20)

The dibromo acid (50 mg) was dissolved in methanol (10 ml) and shaken a t room tempera-
ture with 10% palladium-on-charcoal (25 mg) and calcium carbonate (25 mg) in a n atmosphere
33 Hosking, J. R., and Ruzicka, L., Helv. chim. Acta, 1930, 13, 1402.
 STRUCTURE OF OLEARIN 2637

of hydrogen until the uptake of hydrogen had ceased. Removal of the catalyst and solvent yielded
an acidic product which could not be induced to crystallize, but which behaved as a single com-
pound on t.1.c. (Found: mol. wt., 280. C16H2404 requires mol. wt., 280). vmax 1762 and 1703 cm-1.
The above acid (28) (40 mg) was heated under reflux with thionyl chloride (1 ml), the
excess reagent removed under reduced pressure, and concentrated ammonia solution (15 ml)
added. The product crystallized from aqueous methanol to give the amide of (28) as the mono-
hydrate (10 mg), m.p. 142-143' (Found: C, 64.4; H, 9.3; N, 4.9. C16H85K03,H20requires
C, 64.6; H, 9.2; K, 4.7%). vm,, 3500, 3250, 1760, and 1690 om-1.

Arndt-Eistert Reaction30 on the Acid (28)

The acid chloride of the carboxylic acid (28) (110 mg) was prepared as described above,
dissolved in dry ether (2 ml), and added carefully to excess ethereal diazomethane a t 0'. After
keeping a t room temperature overnight, the ether was removed and the residue dissolved in
anhydrous methanol and stirred with finely divided silver oxide (600 mg) a t 50' for 1 hr. Removal
of the precipitate and the solvent afforded a product which crystallized from hexane to give the
methyl ester (30) as colourless needles (18 mg), m.p. 84-85' (undepressed on admixture with
authentic materialeg), [ o l ] ~ -4.7' (c, 1 . I ) ; infrared spectra (Nujol) of the two samples were
identical.
The methyl ester (30) (10 mg) was heated a t reflur with aqueous sodium hydroxide ( 0 . 5 ~ ,
1ml) for 15 min. The solution was acidified with concentrated hydrochloric acid and warmed
for 10 min on the steam bath. Extraction with ether yielded material which crystallized from
ethyl acetate-light petroleum to give the carboxylic acid (27), m.p. 155-157" (undepressed on
admixture with a n authentic sample) (lit.29 156-158').

Attempted Conversion of the Dibromo Acid (20) into the a$-Unsaturated Acid (21)
(A) The dibromo acid (60 mg) in acetone (5 ml) was heated a t reflux with sodium iodide
(150 mg) for 4 hr. Dilute acidified sodium bisulphite solution was then added and the mixture
extracted with ether. The product was shown (t.1.c.) to be a complex mixture of many components.
(B) The dibromo acid (172 mg) in 95% ethanol (2 ml) was heated a t reflux for 30 min in the
presence of zinc dust (150 mg). Removal of the zinc dust and the solvent gave a non-crystalline
residue which was dissolved in chloroform and separated into neutral and acidic fractions by
means of sodium bicarbonate. The acidic material (98 mg) was methylated with diazomethane
in ether in the usual way, and the product chromatographed on silica gel (5 g). Elution with
chloroform gave a non-crystalline fraction (58 mg) which t.1.c. indicated to be a single compound.
This material had [ a ]-~75' (c, 1.5) ; h 210 nm ( e 2140) ; vmax (CHC13)1764, 1721, and 1656 cm-1 ;
n.m.r. spectrum: 6 0.65 (3H, s ; C9-CH3), 0.93 (3H, d, J 6 Hz; C8-CH3), 2.44 (2H, bs), 3.68
(3H, s, COOCHa), 3.94 and 4.27 (2H, ABq, J A B 8 HZ), and 6.81 ( l H , m).

The authors wish to thank Professor P. R. Jefferies, Department of Organic

Chemistry, University of Western Australia, for providing samples of compounds
(27) and (30). One of us (R.F.S.) gratefully acknowledges receipt of a Commonwealth
Postgraduate Award.