Total synthesis and olfactory evaluation of 5β,10-dimethyl-des-A-18-/ nor-androstan-13β-ol: A potential human pheromone?

Article abstract of DOI:10.1002/ejoc.200400577

5β,10-Dimethyl-des-A-18-nor-androstan-13β-ol (Limdrostanol, 11) was suspected to be the underlying parent steroid responsible for the interesting urinous-animalic, woody olfactory properties of the commercial odorant Timberol (6-9), and the captives Norlimbanol (7) and Limbanol (10), and so could constitute a potential human pheromone. We report the first synthesis of 11, starting with treatment of the bis-Grignard reagent of 1,4-dibromobutane (15) with γ-butyrolactone (16), Appel-Lee bromination of the resulting diol 17 with elimination of the tertiary hydroxy group, and transformation of the obtained bromo alkene 18 into the corresponding triphenylphosphonium salt 13. This was subjected to a Schlosser-Wittig reaction with the γ,δ-unsaturated aldehyde 14, prepared in turn by Grignard treatment of ethyl methacrylate (19) and subsequent Saucy-Marbet reaction of the resulting dimethyl carbinol 20 with ethyl vinyl ether (21). Cascade cyclization of the Schlosser-Wittig product 23 with methanesulfonic acid in dichloromethane at 0 °C afforded the tricyclic alkene 24, which was transformed into the target structure 11 by epoxidation with 3-chloroperbenzoic acid and subsequent reduction with lithium triethylborohydride. In addition to 11, the corresponding 14α-isomer 26 was obtained, and the olfactory properties of both are discussed. The high odor thresholds of 11 and 26, as well as the distinct differences in odor with 6-10, make it very unlikely that these des-A-18-nor-androstanols are the underlying odorous principle of 6-10, or that they function as human pheromones. An alternative synthesis of 11 by cyclization of 23 at 0 °C with 0.8 equivalents of methanesulfonic acid in formic acid as terminating nucleophile is also discussed, but gave only unsatisfactory yields. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejoc.200400577

FULL PAPER
Total Synthesis and Olfactory Evaluation of 5β,10-Dimethyl-des-A-18-nor-
androstan-13β-ol: A Potential Human Pheromone?
[a]
[a][‡]
Philip Kraft* and Kasim Popaj
Keywords: Cyclization / Olfactory properties / Pheromones / Steroids / StructureϪactivity relationships
5β,10-Dimethyl-des-A-18-nor-androstan-13β-ol
(Limdros-
methanesulfonic acid in dichloromethane at 0 °C afforded
tanol, 11) was suspected to be the underlying parent steroid the tricyclic alkene 24, which was transformed into the target
responsible for the interesting urinous-animalic, woody ol-
structure 11 by epoxidation with 3-chloroperbenzoic acid

factory properties of the commercial odorant Timberol (6−9), and subsequent reduction with lithium triethylborohydride.
and the captives Norlimbanol (7) and Limbanol (10), and so In addition to 11, the corresponding 14α-isomer 26 was ob-
could constitute a potential human pheromone. We report the
first synthesis of 11, starting with treatment of the bis-Grig-
nard reagent of 1,4-dibromobutane (15) with γ-butyrolactone
tained, and the olfactory properties of both are discussed.
The high odor thresholds of 11 and 26, as well as the distinct
differences in odor with 6−10, make it very unlikely that
these des-A-18-nor-androstanols are the underlying odorous
principle of 6−10, or that they function as human phero-
mones. An alternative synthesis of 11 by cyclization of 23 at
0 °C with 0.8 equivalents of methanesulfonic acid in formic
(
16), Appel−Lee bromination of the resulting diol 17 with
elimination of the tertiary hydroxy group, and transformation
of the obtained bromo alkene 18 into the corresponding tri-
phenylphosphonium salt 13. This was subjected to
a
Schlosser−Wittig reaction with the γ,δ-unsaturated aldehyde acid as terminating nucleophile is also discussed, but gave
14, prepared in turn by Grignard treatment of ethyl meth-
only unsatisfactory yields.
acrylate (19) and subsequent Saucy−Marbet reaction of the
resulting dimethyl carbinol 20 with ethyl vinyl ether (21).
Cascade cyclization of the Schlosser−Wittig product 23 with
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
Androst-16-en-3-one (1), which emanates a pronounced,
persistent, and very intense urinous-animalic odor, has also
[
4]
Target and Retrosynthesis
been identified, together with the corresponding alcohols 2
[
5]
and 3, in human axillary sweat. While the androsten-3α-
ol 2 emanates a strong, pleasant, and typical sandalwood
odor, its 3β-isomer 3 is weaker and possesses a more ani-
malic-urinous sandalwood note (Figure 1).
Sex has turned out to be quite a successful strategy for
evolution, even though, biochemically, cell division is less
troublesome and more efficient than union. The necessary
differentiation into male and female individuals is far from
simple, as the underlying genetic information is too com-
plex to be coded only in the sex-determining chromosomes.
Hormones (Greek: ορµαν, to trigger) are therefore essential
to switch on certain genes, not only in sexual differentiation
but also in the regulation of reproductive cycles. Obviously,
sex becomes more successful if information concerning
presence and fertility is transferred between the sexes. The
The pleasant odor of 2 and the occurrence of 1Ϫ3 in
human sweat and urine had resulted in speculations about
the likelihood of them being human pheromones.[ The
first observation of the existence of human pheromones was
the so-called McClintock effect. Martha McClintock[ no-
ticed during her studies at Wellesley college in Massachu-
setts that the menstrual cycles of female room-mates living
together became synchronized. This menstrual synchrony,
the convergence of the onset date of the menstrual flow, has
been experimentally replicated a number of times,[ and
was recently shown to be at least partly mediated by the
5]
6a]
messenger substances that take on this role are called pher-
omones (Greek: ϕερειν, to transfer),[2] and are often chemi-
6b]
cally closely related to hormones. Probably the most promi-
nent example of a pheromone in mammals is 5α-androst-
[
7]
smell of 5α-androst-16-en-3α-ol (2). The frequency of the
pulsatile secretion of lutheinizing hormone (LH) in the fol-
licular phase is significantly decreased by exposing women
16-en-3-one (1), which is secreted in the saliva of the boar
and elicits the characteristic immobilization response (lor-
dosis) of the estrous sow to the advance of her mate. 5α-
[
3]
[
7]
to 2. While one can debate about the biological use of
this, it demonstrates the existence of substances with a pher-
omone effect on humans.
[
[
a]
‡]
Givaudan Schweiz AG, Fragrance Research,
Überlandstrasse 138, 8600 Dübendorf, Switzerland
Fax: (internat.) ϩ 41-44-8242926
There may be more than one human pheromone. After a
E-mail: philip.kraft@givaudan.com
[1]
[8a]
See ref.
study by Jacob and McClintock on mood effects of ster-
Eur. J. Org. Chem. 2004, 4995Ϫ5002
DOI: 10.1002/ejoc.200400577
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4995

P. Kraft, K. Popaj
FULL PAPER


Figure 2. Timberol components 6Ϫ9, Limbanol (10), and the
des-A-18-nor-androstanol target molecule 11, our so-called
Limdrostanol, which was proposed to be the scaffold of 7 and 10
1
.7% in the woody foug e` re ‘‘Minotaure’’ (P. Picasso, 1992),
Figure 1. Steroids 1Ϫ5 with putative pheromone function in although probably the most typical use was in ‘‘Basala’’
humans
(
5
Shiseido, 1993), which Dominique Preyssas built around a
.7% content of Timberol (6Ϫ9).
When Schulte-Elte and co-workers at Firmenich inves-
[
14]
oids on men and women, research is currently focusing on
estra-1,3,5(10),16-tetraen-3-ol (4) as the prime putative tigated the commercial product, they discovered that the
pheromone of females and androsta-4,16-dien-3-one (5) as main cis-configured component 6 (63.8%) possesses only a
the prime putative pheromone of males.[
8b]
In our own weak, rather complex odor of mainly floral tonality, quite
experiments, we have indeed found that men described 4 as different from the overall woody, animalic odor profile of the
sweet, woody, and powdery, while women described it as a mixture. This woody, animalic note was found to be mainly
malodorous note of sweaty, obnoxious character. The odor due to the trans isomer 7, at 13.4% rather a minor constitu-
threshold that we determined was not dependent on sex, ent. In pure form, 7 also emanates quite strong urinous fa-
however, and at 8.0 ng/L air was quite high. Androsta-4,16- cets that to Schulte-Elte et al. recalled the odor of steroids.

dien-3-one (5) was described as sweaty, algae-like by both The other by-products of the industrial Timberol syn-
male and female subjects, and we determined an odor thesis, such as the open-chain alcohol 8 or 10-ethyl-7,8-di-
threshold of 0.002 ng/L air for both sexes. Lundström et hydro-β-ionol (9), do not contribute much to the overall
al.[ recently reported an absolute detection threshold of odor of the commercial material.
9]
211 µ in propylene glycol for 5, but could not determine
The interesting steroid-type, urinous-animalic, woody
a threshold for 4. Although 4 and 5 are already used in characteristics prompted the Firmenich researchers to de-
[
14]
perfumes claimed to attract the opposite sex, such as ‘‘Re- velop an industrial synthesis of the trans isomer 7, which
alm’’ and ‘‘Realm Men’’ (Erox Corp., 1993),[ sound evi- has been in captive use under the name of Norlimbanol
10]

dence for such an effect is still lacking. Recently, Turner et since the mid 1980s. The prefix ‘‘nor’’ in the name Norlim-
al.[ discovered in 5α-androst-16-en-7-one a new andros- banol (7) refers to the finding that the irone analogue 10
11]

tenone analogue that was reported also to smell sandal- had an even more pronounced urinous-animalic odor of
[
15]
wood-like, as 3, but much more weakly.
woody-powdery tonality.
Compound 10 was thus con-
Astonishingly, Ohloff et al.[ reported a urinous-animalic sidered by Ohloff et al. to be the parent structure of this
12]
[12]
note reminiscent of the odor of steroids such as 5α-androst- steroid-type woody odorant family, and was given the trad-


1
6-en-3-one (1) to be present in the prominent perfumery ename of Limbanol (10). Limbanol (10) is also manufac-


raw material Timberol (6Ϫ9, Figure 2). Timberol (6Ϫ9) tured on an industrial scale and has been used as a captive
[
15]
was discovered in 1979 by Klein and Rojahn of Dragoco in perfumery,
though to a lesser extent than Norlim-
and soon became a popular perfumery banol (7). This, however, is only due to its more complex
ingredient, especially in masculine fine fragrances. It can, synthesis, and a correspondingly higher price.
for instance, be found at 1.5% in the woody chypre ‘‘Fah- The superior steroid-type olfactory properties of Lim-
renheit’’ (Dior, 1988) created by Jean-Louis Sieuzac, or at banol (10) were interpreted by Ohloff et al. as originat-
[
13]

(now Symrise),

[12]
4996
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4995Ϫ5002

Limdrostanol: A Potential Human Pheromone?
FULL PAPER
ing from a steroid-type folding, which should be the result Results and Discussion
of an all-equatorial configuration of the substituents at C-
In the synthesis of the Wittig salt 13 (Scheme 2), we fol-
lowed an efficient strategy that Schore and Knudsen[ had
applied in a synthesis of fused triquinanes by direct ad-
3
, C-5, and C-6 (ionone numbering, see Figure 2). In this
16]
conformation, depicted in Figure 2, the cyclohexyl ring
corresponds to an androstane B ring with the additional 3-
methyl substituents of 10 indicating an A ring fusion, while
rings C and D should be formed by an appropriate folding
of the hydroxyhexyl side chain, despite all the steric hin-
drance this would imply. Accordingly, 5β,10-dimethyl-des-
A-18-nor-androstan-13β-ol (11) was proposed by Ohloff et
[
17]
dition of α,ω-bis-Grignard reagents to lactones.
Thus,
treatment of 1,4-dibromobutane (15) with an excess of mag-
nesium in THF, followed by slow addition of a γ-butyrolac-
tone (16) solution in THF at reflux provided 1-(3Ј-hydroxy-
propyl)cyclopentanol (17) in 83% yield after purification by
flash chromatography (FC). Schore and Knudsen[ also
wanted to halogenate the primary hydroxy group of the diol
17 with concomitant elimination of the tertiary hydroxy
group, but upon subjection of 17 to AppelϪLee chlori-
16]
al.[
12]
as the parent steroid responsible for the interesting
olfactory properties of 7 and 10, and likewise also of the
commercial product Timberol (6Ϫ9). However, this ‘‘Lim-

drostanol’’ (11), as we are going to call it below, was never
synthesized, even though, if the chain of reasoning were
correct, we might expect 11 to possess not only a potent,
highly interesting odor, but possibly also some pheromone-
[
18]
nation
conditions with triphenylphosphane at reflux in
carbon tetrachloride, the desired chloro alkene was ob-
[
16]
tained under the best conditions only in 36% yield, while
1-oxaspiro[4.4]nonane was isolated in 39% yield as the
major product of this reaction. We found, however, that the
like impact. To examine Ohloff et al.’s[
12]
speculations in
greater detail we tackled the synthesis of Limdrostanol (11).
On reflection, ‘‘Limdrostanol’’ (11) appears to be quite a
challenging target. As C-8 (steroid numbering, see Figure 2)
is not quaternary-substituted, a topological retrosynthetic
strategy through cationic cyclization initially appeared un-
favorable to us, but this still seemed to be the shortest route
and we decided to give it a try. Strategic disconnection of
the fused ring system gives the triene 12, which could un-
dergo cationic cyclization with trapping of the positive
charge by a nucleophile at the desired position C-13 (steroid
numbering; see Figure 2). SchlosserϪWittig disconnection
of the central E-configured double bond of 12 then reveals
[
19]
AppelϪLee bromination
of the diol 17 went smoothly
and with selective elimination of the tertiary hydroxy group
on treatment of 17 with two equivalents of triphenylphos-
phane and carbon tetrabromide in toluene at 60 °C, and the
primary bromo alkene 18 was obtained in 61% yield after
purification by flash chromatography on neutral aluminum
oxide. An alternative synthesis of 18 starting from 1-allyl-
cyclopentene had been reported by Brown and Salunkhe,[
but without experimental details.
20]
(3Ј-cyclopent-1-enylpropyl)triphenylphosphonium bromide
(13) and 4,5-dimethylhex-4-enal (14) as the two building
blocks for this convergent approach (Scheme 1).
Scheme 2. Synthesis of the 3Ј-cyclopent-1-enylpropyl Wittig salt 13
from 1,4-dibromomethane (15) and γ-butyrolactone (16)
The transformation of 18 into the corresponding phos-
phonium salt 13 was less straightforward, and standard re-
action conditions with triphenylphosphane at reflux in tolu-
ene or xylene afforded 13 in a maximum 20% yield. When
the reaction was carried out in acetonitrile or nitromethane
no phosphonium salt 13 was formed at all, but use of 2-
[
21]
propanol as solvent began to improve the yields. The best
results were obtained at reflux in ethanol, and after two
days of heating of 18 with triphenylphosphane in ethanol,
the Wittig salt 13 was isolated in 66% yield after recrystalli-
zation from toluene.
The synthesis of the second building block, the γ,δ-un-
saturated aldehyde 14, was achieved in two steps by making
[
22]
use of a SaucyϪMarbet reaction (Scheme 3). A standard
Grignard reaction between ethyl methacrylate (19) and
methylmagnesium iodide furnished the starting material 20
(92% yield) for this Claisen-type rearrangement. While in-
Scheme 1. Retrosynthetic analysis of the target compound 11
Eur. J. Org. Chem. 2004, 4995Ϫ5002
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4997

P. Kraft, K. Popaj
FULL PAPER
itial SaucyϪMarbet reaction conditions[22] with the di-
methyl carbinol 20 and ethyl vinyl ether (21) in the presence
of phosphoric acid as catalyst resulted in yields around or
below 20% (GC/MS), heating of 20 in ethyl vinyl ether (21)
in an autoclave at 160 °C/17 bar in the presence of catalytic
amounts of 2,6,7-trioxa-4-aza-1-phosphabicyclo[2.2.2]-
[
23]
octane 1-oxide (22), provided our second building block
4 in 54% yield after chromatographic purification.
1
Scheme 4. WittigϪSchlosser olefination of 14, acid-catalyzed cycli-
zation of the resulting triene 23, epoxidation, and subsequent re-
duction of 25 to the target structure 11
Scheme 3. Synthesis of 4,5-dimethylhex-4-enal (14) by
SaucyϪMarbet reaction
or hydrolysis, but the target structure 11 was evidently not
present in the fractionated mixtures.
Searching for more drastic conditions, we then applied
Next on the agenda was the central Wittig reaction be- fluorosulfonic acid in nitropropane according to a pro-
[
28]
tween the phosphonium salt 13 and the γ,δ-unsaturated al- cedure used by Snowden et al.
for the synthesis of (Ϯ)-
dehyde 14, in which a high E selectivity was hoped for Ambrox from linear trienols or monocyclic dienols. Treat-
(
Scheme 4). Initial Wittig reactions between 13 and 14 un- ment of 23 with fluorosulfonic acid in 2-nitropropane at
[24]
der FitjerϪQuabeck conditions
with the use of potass- Ϫ90 °C afforded two new products with very similar chro-
ium tert-butoxide as a base at reflux in tetrahydrofuran af- matographic retentions and mass spectra resembling those
forded, as suspected, exclusively Z-configured 23, but in of the starting material 23. By repeated chromatography,
disappointingly moderate yields of only around 30%. The one of these was enriched and tentatively assigned the struc-
Schlosser modification[
21,25]
of the Wittig reaction should, ture of the tricycle 24 (Scheme 4), so apparently the fluoro-
however, permit the construction of E-disposed olefins sulfonate group had eliminated under the reaction con-
from aldehydes and non-stabilized phosphorus ylides. By ditions with the formation of 24 and an isomeric com-
the general method of Bestmann et al.,[ 13 was then pound. Although 24 was unintended, all that remained to
transformed into its ylide by treatment with one equivalent convert it to the target structure 11 was the introduction of
of phenyllithium in THF, and then treated with 14 at Ϫ80 a hydroxy function at C-13.
21]
°
C. Deprotonation of the obtained lithio betaine species
We decided to optimize this step, and found that treat-
with an additional equivalent of phenyllithium furnished, ment of 23 with an excess of methanesulfonic acid in di-
after quenching with methanol, 1-(7Ј,8Ј-dimethylnona- chloromethane at 0 °C increased the yield to 62%. Epoxid-
3
Ј,7Ј-dienyl)cyclopent-1-ene (23), with an E/Z ratio of 8:2, ation of 24 with 3-chloroperbenzoic acid in dichlorometh-
in 61% yield. ane at 0 °C then furnished, after workup and purification
The two building blocks 13 and 14 having been success- by flash chromatography, the corresponding epoxide 25, al-
fully stitched together, our effort was now directed towards beit in a moderate yield of only 28%. Since all attempts
the cascade cyclization of the triene 23 to provide the final to achieve reductive opening of 25 with lithium aluminum
ring system of our target molecule 11, the most critical step hydride were in vain, even at reflux in tetrahydrofuran or
of our synthetic plan (Scheme 1). First experiments were dioxane, we employed lithium triethylborohydride (Super-
carried out in formic acid.[ At room temperature, no con- Hydride ) as the most powerful hydride nucleophile avail-
26]

version was discernible, but treatment of 23 with formic able to us. Treatment of the epoxide 25 with three equiva-

acid at 100 °C and hydrolysis with aqueous potassium hy- lents of Super-Hydride in tetrahydrofuran at reflux finally
droxide provided a complex mixture that could not be sepa- provided our molecular target Limdrostanol (11), ac-
rated or further characterized, except that it also contained companied by its regioisomer 26. Despite similar chromato-
a secondary alcohol. Next, the triene 23 was treated with graphic retentions, we were able to separate the two isomers
trifluoroacetic acid in dichloromethane at 0 °C,[ but again by flash chromatography, and isolated 11 and 26 in 30%
a complex product mixture was obtained. This could not and 38% yields, respectively.
27]
be characterized adequately to allow the identification of
The structures and relative configurations of 11 and 26
its components either after chromatographic fractionation were assigned by 2D NMR spectroscopy, including the use
4998
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4995Ϫ5002

Limdrostanol: A Potential Human Pheromone?
FULL PAPER

of INADEQUATE experiments to establish all CϪC con- (11) nor its 14α-isomer 26 were reminiscent of Limbanol
nectivities. In the NOESY spectrum of 11, a cross-peak be- (10), Norlimbanol (7), or Timberol (6Ϫ9), nor did they
tween 10-Meax and 5-Me substantiated the β-equatorial possess any typical urinous-animalic woody odor. 5β,10-Di-
orientation of the latter methyl group, while cross-peaks be- methyl-des-A-18-nor-androstan-13β-ol (11) was certainly
tween 10-Meax and 6-H, as well as between 10-Meax and woody in smell, but in an earthy direction, reminding one of
8-H and between 8-H and 13-OH, established the β-axial patchouli oil rather than of steroids, and in addition it had
configurations of all these substituents. Cross-peaks be- some sweet, almost chocolate-type character. Earthy aspects
tween 9-Hax and 5-, 12- and 14-H then unambiguously es- were also present in the 14α-isomer 26, which also ema-
tablished the α-axial configurations of these hydrogen nated a pleasant ambery note. With an odor threshold of 46
atoms, since the transfusion of rings B and C was apparent ng/L air, 26 was somewhat more intense than the target
from a 1D slice extracted from the HSQC spectrum at δ ϭ compound 11, for which we measured a threshold of 90 ng/

3
6.2 ppm, which displayed a quadruplet with an all-axial L air. Both 11 and 26 are far weaker than Limbanol (10),
coupling constant of 12.0 Hz for 8-H at δ ϭ 1.38 ppm, thus Norlimbanol (7), or Timberol (6Ϫ9), and they are even
determining the relative configurations of all stereocenters farther away from the thresholds of the putative human
in 5β,10-dimethyl-des-A-18-nor-androstan-13β-ol (11).
pheromones delineated in Figure 1. Neither the odor
The relative stereochemistry of the 14α-isomer 26 was de- character nor the thresholds of 11 and 26 are gender-
termined similarly, with cross-peaks between 14-OHax and specific, and this altogether makes it highly improbable that
9
-H , between 14-OH and 7-H , and between 14-OH
they could function as human pheromones. Since both 11
ax
ax
ax
ax
and 12-Hax on the α-face, and cross-peaks between 10-Meax and 26 differ distinctly from 6Ϫ10 in their odor profiles
and 6-, 8- and 11-H in β-axial orientations. The β-equatorial and are of weaker intensity, they do not even constitute
configuration of 5-Me was again derived from a cross peak appropriate templates for the design of new woody odor-
with 10-Me . The high stereoselectivity, with all-trans- ants.
ax
fusions of ring B, C and D in both 11 and 26, reflects a
So the dream of the fictitious fragrance chemist Henri
concerted course of the cyclization of 25 to 26 according to Biotte in a short story by Roald Dahl has not come true:
the StorkϪEschenmoser hypothesis.[
29]
‘‘What I intend to do,’’ he said, ‘‘is to produce a perfume
Although we had reached our synthetic goal, we still which will have the same electrifying effect upon a man as
wondered if the yield of 11 might be improved if formic acid the scent of a bitch in heat has upon a dog! One whiff and
[
30]
were present as a terminating nucleophile in the cyclization that’ll be it!’’ Considering the consequences Henri Biotte
reaction of 23 with methanesulfonic acid (Scheme 5). Sub- suffers in this story, we should perhaps thank our lucky
sequent alkaline hydrolysis of the crude products might stars that 11 was not an active human pheromone.
then open up a direct route to our target molecule 11, cir-
cumventing the formation of 26. Thus, 23 was treated with
0
.8 equivalents of methanesulfonic acid in formic acid at 0 Experimental Section
°
C, but Limdrostanol (11) was isolated by flash chromatog-
IR: PerkinϪElmer Spectrum One, Bruker VECTOR 22/Harrick
SplitPea micro ATR, Si. NMR: Bruker ARX 300, Bruker AV-
ANCE DPX 400, TMS int. (δ ϭ 0 ppm). MS: Finnigan TSQ 700
raphy in a disappointingly low yield of only 3% after hy-
drolysis of the resulting product mixture with ethanolic so-
dium hydroxide. The circuitous route via the epoxide 25 was
still superior to the direct conversion of 23 into the target
molecule 11, and in addition the olfactory evaluation of the
(
3
ESI), Finnigan SSQ 700 (CI: NH ), Finnigan MAT 95 (EI: 70 eV),
HP Chemstation 6890 GC/5973 Mass Sensitive Detector. FC:
Merck Kieselgel 60 (40Ϫ63 µm). TLC: Merck Kieselgel 60 F254
14α-isomer 26 in comparison with 11 was highly interesting.
(
5 cm ϫ 7.5 cm on aluminum); visualization reagent: PMA spray
soln. for TLC, Merck 1.00480.0100. Melting points: Mettler FP5
melting point apparatus. Elemental analyses: Mikroanalytisches
Laboratorium Ilse Beetz, 96301 Kronach, Germany. Unless other-
2
wise stated, all reactions were performed under N with reagents
and solvents (puriss. or purum) from Fluka, used without further
purification. The odor thresholds were determined by
GCϪolfactometry: Different dilutions of the sample substance are
injected into a gas chromatograph in descending order of concen-
tration until the panelist fails to detect the respective substance at
the sniffing port. The panelist smells in blind and presses a button
on perceiving an odor. If the recorded time matches the retention
time, the sample is further diluted. The last concentration detected
at the correct retention time is the individual odor threshold. The
Scheme 5. Direct synthesis of the target compound 11 from the
triene 23
Olfactory Evaluation and Conclusions
With both 5β,10-dimethyl-des-A-18-nor-androstan-13β- reported threshold values are the geometrical means of the individ-
ual odor thresholds of the different panelists.
ol (Limdrostanol, 11) and its 14α-isomer 26 to hand, we
return to our initial starting point, of whether or not 11 is
1
-(3Ј-Hydroxypropyl)cyclopentanol (17): A solution of 1,4-dibromo-

the odorous principle behind Limbanol (10), Norlim-
butane (15, 216 g, 1.00 mol) in dry THF (500 mL) was added drop-
banol (7), and Timberol (6Ϫ9), and how likely it is that wise over 3.5 h to magnesium turnings (72.9 g, 3.00 mol), the reac-

it constitutes a human pheromone. Neither Limdrostanol tion being initiated by occasional heating with a heat gun. The
Eur. J. Org. Chem. 2004, 4995Ϫ5002
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4999

P. Kraft, K. Popaj
FULL PAPER
reaction mixture was stirred at reflux for an additional 3 h, and a
ϩ
MS (ESI): m/z (%) ϭ 371 (100) [M Ϫ Br] ppm. C26H28BrP
solution of γ-butyrolactone (16, 94.6 g, 1.10 mol) in dry THF
(451.39): calcd. C 69.18, H 6.25; found C 69.18, H 6.24.
(
1.3 L) was then added dropwise with stirring at reflux temp. over
a period of 12 h. The reaction mixture was then cooled to room
temp., and subsequently quenched with satd. aq. NH Cl solution
700 mL). The aqueous layer was extracted with Et O (3 ϫ
00 mL), and the combined organic extracts were washed with
SO ), and concentrated under reduced
pressure. The residue was purified by silica gel FC (Et O/MeOH,
8:2; R ϭ 0.23) to afford the title compound 17 (120 g, 83%) as a
slightly yellow oil. IR (film): ν˜ ϭ 3333 (s, ν OϪH), 1058/981 (m, ν
2
2
3
,3-Dimethylbut-3-en-2-ol (20): A solution of methyl iodide (284 g,
.00 mol) in Et O (700 mL) was added dropwise over a period of
h to a stirred suspension of magnesium turnings (53.5 g, 2.20
O (80 mL), with the reaction being initiated by oc-
2
4
(
6
2
mol) in Et
2
casional heating with a heat gun. The reaction mixture was then
stirred at reflux for an additional 3 h, and cooled to room temp. A
solution of ethyl methacrylate (19, 114 g, 1.00 mol) in Et
brine (300 mL), dried (Na
2
4
2
2
O
9
f
(
700 mL) was added dropwise over 2.5 h, and the mixture was
Ϫ1
1
stirred at reflux for an additional 16 h. After the reaction mixture
had cooled down to room temp., it was quenched by addition of
satd. aq. NH
tracted with Et
were washed with brine (250 mL), dried (Na
trated under reduced pressure. The crude product was distilled over
a 15-cm Vigreux column at 48Ϫ50 °C/20 mbar to provide the title
compound 20 (92.3 g, 92%) as a colorless liquid. IR (film): ν˜ ϭ
CϪO), 1449 (m, δ CϪH) cm . H NMR (CDCl
m, 2 H, 2Ј-H ), 1.52Ϫ1.58 (m, 2 H, 1Ј-H ), 1.60Ϫ1.67 (m, 2 H, 2-,
), 1.69Ϫ1.72 (m, 4 H, 3-,4-H ), 1.78Ϫ1.83 (m, 2 H, 2-,5-H ),
.33 (s, 2 H, 1-,3Ј-OH), 3.64 (t, J ϭ 6.0 Hz, 2 H, 3Ј-H
): δ ϭ 23.8 (2t, C-3,-4), 28.1 (t, C-2Ј), 38.5 (t, C-1Ј),
9.7 (t, C-2,-5), 63.0 (t, C-3Ј), 82.0 (s, C-1) ppm. MS (EI): m/z
3
): δ ϭ 1.43Ϫ1.49
(
5
3
2
2
4
Cl solution (800 mL). The aqueous layer was ex-
O (3 ϫ 500 mL), and the combined organic extracts
SO ), and concen-
-H
b
2
a
1
3
2
2
) ppm.
C
2
4
NMR (CDCl
3
3
ϩ
ϩ
ϩ
2 2 5
(%) ϭ 144 (3) [M ], 126 (5) [M Ϫ H O], 97 (54) [M Ϫ C H OH],
ϩ
ϩ
8
6 4 7
5 (76) [C H13 ], 55 (100) [C H ].
3
3
380 (s, ν OϪH), 1449 (s, δ OϪH), 1374 (s, δ CH ), 1163 (s, ν
CϪO), 960/933 (s, ν CϭCϪH), 1449 (s, δ CϪH), 3091 (m, ν Cϭ
1
1
-(3Ј-Bromopropyl)cyclopentene (18): Triphenylphosphane (432 g,
Ϫ1
1
CϪH) cm . H NMR (CDCl
3 3
): δ ϭ 1.34 (s, 6 H, 1-H , 2-Me),
.65 mol) was added portionwise at room temp. to a stirred suspen-

1.75 (s, 1 H, OH), 1.80 (s, 3 H, 3-Me), 4.75 (t, J ϭ 1.5 Hz, 1 H, 4-
sion of Celite (100 g), tetrabromomethane (547 g, 1.65 mol), and
1
for 3 h at 60 °C. The reaction mixture was cooled to room temp.,
and diluted with hexane (1 L). The precipitated triphenylphos-
phane oxide was filtered off and washed with hexane (200 mL),
and the filtrate was concentrated on a rotary evaporator at reduced
1
3
H
(Z)), 4.99 (t, J ϭ 1.5 Hz, 1 H, 4-H(E)) ppm. C NMR (CDCl
3
):
7 (119 g, 825 mmol) in toluene (3 L), and stirring was continued
δ ϭ 19.1 (q, 3-Me), 28.8 (q, C-1, 2-Me), 73.1 (s, C-2), 108.4 (d, C-
ϩ
4
), 152.0 (s, C-3) ppm. MS (EI): m/z (%) ϭ 100 (8) [M ], 85 (100)
ϩ
ϩ
ϩ
[M
Ϫ CH
3 2 3 3 6
], 67 (11) [M Ϫ H O Ϫ CH ], 59 (64) [C H O ].
4,5-Dimethylhex-4-enal (14): A mixture of 20 (72.1 g, 720 mmol)
pressure. The crude product was purified by FC (hexane/Et
8:2; R ϭ 0.54) on neutral aluminum oxide to furnish the title
compound 18 (95.2 g, 61%). IR (film): ν˜ ϭ 657 (s, ν CϪBr), 1436
2
O,
and 2,6,7-trioxa-4-aza-1-phosphabicyclo[2.2.2]octane 1-oxide (22,
.09 g, 7.21 mmol) in ethyl vinyl ether (21, 114 g, 1.58 mol) was
stirred for 3 h in an autoclave at 160 °C/17 bar. After the reaction
mixture had cooled to room temp., it was diluted with Et
9
f
1
Ϫ1
1
(
m, δ CϪH), 3043 (w, ν CϭCϪH), 1650 (w, ν CϭCϪC) cm . H
NMR (CDCl ): δ ϭ 1.85 (dt, J ϭ 7.0, 7.0 Hz, 2 H, 2Ј-H ), 2.00
), 2.19Ϫ2.24 (m, 4 H, 1Ј-,5-H ),
), 5.37 (t,
2
O
3
2
(
500 mL), washed with HCl solution (0.5, 100 mL), satd. aq.
(dt, J ϭ 7.0, 7.0 Hz, 2 H, 4-H
2
2
NaHCO
3
solution (2 ϫ 100 mL), and brine (100 mL). The organic
SO ), and the solvent was evaporated on a
rotary evaporator. The residue was purified by silica gel FC (pen-
tane/Et O, 20:1; R ϭ 0.34) to furnish the odoriferous title com-
pound 14 (49.1 g, 54%). IR (film): ν˜ ϭ 1725 (s, ν CϭO), 2719 (s, ν
HϪCϭO), 1450 (s, δ CϪH), 1375 (s, δ CH ), 3043 (m, ν CϭCϪH),
44 (w, δ CϭCϪH) cm . H NMR (CDCl ): δ ϭ 1.64 (s, 6 H, 5-
Me ), 1.66 (s, 3 H, 4-Me), 2.35 (t, J ϭ 6.5 Hz, 2 H, 3-H ), 2.44 (td,
J ϭ 6.5, 2.0 Hz, 2 H, 2-H ), 9.77 (t, J ϭ 2.0 Hz, 1 H, 1-H) ppm.
): δ ϭ 15.4 (q, 4-Me), 20.2/20.6 (2q, C-6, 5-Me),
7.0 (t, C-3), 42.5 (t, C-2), 125.4 (s, C-5), 125.7 (s, C-4), 202.7 (s,
2.27Ϫ2.32 (m, 2 H, 3-H
2
), 3.39 (t, J ϭ 6.5 Hz, 2 H, 3Ј-H
2
layer was dried (Na
2
4
13
J ϭ 2.0 Hz, 1 H, 2-H) ppm. C NMR (CDCl
3
): δ ϭ 23.4 (t, C-4),
9.5 (t, C-2Ј), 30.8 (t, C-3), 32.4 (t, C-3Ј), 33.5 (t, C-1Ј), 35.0 (t, C-
), 124.5 (d, C-2), 142.7 (s, C-1) ppm. MS (EI): m/z (%) ϭ 190
2
5
2
f
ϩ
ϩ
ϩ
ϩ
(
(
6
3) [M ], 188 (3) [M ], 109 (5) [M Ϫ Br], 82 (18) [C H10 ], 67
100) [C
3
ϩ
7
5
H
].
Ϫ1 1
8
3
2
2
(3Ј-Cyclopent-1-enylpropyl)triphenylphosphonium Bromide (13): A
2
solution of 18 (95.1 g, 502 mmol) in EtOH (200 mL) was added
dropwise at room temp. over 20 min to a solution of triphenylphos-
phane (132 g, 502 mmol) in EtOH (800 mL). The reaction mixture
was stirred at reflux for 2 days and was then cooled down to room
temp., and the solvent was then evaporated on the rotary evapor-
ator. The resulting residue was taken up in toluene (200 mL), and
the phosphonium salt was precipitated by dropwise addition of
Et O (150 mL). The precipitate was vacuum filtered, suspended at
2
reflux in toluene (500 mL) for 1 h, and then stirred at room temp.
for a further 2 h. The resulting crystals were filtered off under suc-
13
C NMR (CDCl
3
2
ϩ
ϩ
C-1) ppm. MS (EI): m/z (%) ϭ 126 (40) [M ], 111 (25) [M
CH
Ϫ
ϩ
ϩ
ϩ
3
2 6 5 9
], 108 (35) [M Ϫ H O], 83 (29) [C H11 ], 69 (35) [C H ], 55
ϩ
(
4 7
100) [C H ].
(3ЈE/Z)-1-(7Ј,8Ј-Dimethylnona-3Ј,7Ј-dienyl)cyclopent-1-ene (23): A
phenyllithium solution in cyclohexane (2 , 73.5 mL, 147 mol) was
added dropwise over 20 min to a suspension of 13 (63.2 g,
140 mmol) in dry THF (600 mL). The resulting deep red solution
tion, washed with toluene (100 mL), and dried to constant weight was cooled to Ϫ80 °C, and a solution of 14 (17.7 g, 140 mmol) in
at room temp./2 mbar to afford the title compound 13 (149 g, 66%).
M.p. 208.1Ϫ210.2 °C. IR (KBr): ν˜ ϭ 1435/1113 (s, ν PϪPh), 693
dry Et
2
O (200 mL) was added over a period of 20 min. After
15 min of stirring at Ϫ80 °C, the reaction mixture was warmed
(
s, δ CϪH oop, arom.), 1482 (m, δ CϪH), 1586 (m, ν CϭC, arom.), up to Ϫ30 °C and treated dropwise with additional phenyllithium
Ϫ1
1
3
052 (m, ν CϪH, arom.), 3043 (w, ν CϭCϪH) cm . H NMR solution (2 , 73.5 mL, 147 mol). After 10 min of stirring at Ϫ30
OD): δ ϭ 1.82 (quint., J ϭ 7.5 Hz, 4 H, 2Ј-,4-H ), 2.10Ϫ2.14 °C, MeOH (21 mL) was added carefully, upon which a precipitate
m, 2 H, 3-H ), 2.27Ϫ2.33 (m, 4 H, 1Ј-,5-H ), 3.39Ϫ3.45 (m, 2 H, of triphenylphosphane oxide was formed. The resulting suspension
Ј-H ), 5.38 (br. s, 1 H, 2-H), 7.73Ϫ7.92 (m, 15 H, 2ЈЈЈ-HϪ6ЈЈЈ-H) was stirred for an additional 2 h at room temp., poured into water
(CD
3
2
(
3
2
2
2
13
ppm. C NMR (CD
3
OD): δ ϭ 21.7/21.8 (2t, C-3Ј), 22.1 (t, C-1Ј),
4.3 (t, C-4), 32.3/32.5 (2t, C-1Ј), 33.2 (t, C-3), 35.0 (t, C-5), 119.2/ organic layers were dried (Na
20.4 (2s, C-1ЈЈЈ), 126.5 (d, C-2), 131.4/131.5 (2d, C-3ЈЈЈ,-5ЈЈЈ), duced pressure. The resulting residue was purified by silica gel FC
34.6/134.8 (2d, C-2ЈЈЈ, Ϫ6ЈЈЈ), 136.2 (d, C-4ЈЈЈ), 143.3 (s, C-1) ppm. (pentane/Et
(300 mL), and extracted with Et
2
O (3 ϫ 400 mL). The combined
2
1
1
2
SO
4
) and concentrated under re-
2
O, 98:2; R
f
ϭ 0.65) to furnish the title compound 23
Eur. J. Org. Chem. 2004, 4995Ϫ5002
5
000
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org

Limdrostanol: A Potential Human Pheromone?
FULL PAPER
(
(
18.7 g, 61%) as an isomeric mixture with an E/Z ratio of 8:2. IR
film): ν˜ ϭ 1445 (s, δ CϪH), 965 (s, δ CϭCϪH), 1375 (s, δ CH
), MS (EI): m/z (%) ϭ 234 (56) [M ], 219 (57) [M Ϫ CH ], 216 (75)
3 3
8), 41.6 (d, C-5), 44.6 (d, C-9), 67.8 (s, C-14), 71.8 (s, C-13) ppm.
ϩ
ϩ
Ϫ1
1
ϩ
ϩ
ϩ
1
651 (m, ν CϭCϪC) cm . H NMR (CDCl
), 1.83 (quint., J ϭ 7.0 Hz, 2 H, 4-H
m, 4 H, 2Ј-,5Ј-H ), 2.03Ϫ7.09 (m, 2 H, 6Ј-H
), 2.20Ϫ2.32 (m, 4 H, 5-,1Ј-H ), 5.35Ϫ5.37 (m, 1 H, 2-H),
3
): δ ϭ 1.63 (br. s, 9 [M Ϫ H
2
2 3
O], 201 (100) [M Ϫ H O Ϫ CH ], 134 (67) [C10H14 ],
ϩ
H, 7Ј-Me, 8Ј-Me
(
H, 3-H
2
2
), 2.01Ϫ2.06 97 (78) [C H14 ]. C16H26O (234.4): calcd. C 81.99, H 11.18; found
7
2
2
), 2.11Ϫ2.18 (m, 2 C 81.99, H 11.14.
2
2
.40Ϫ5.43 (m, 3Ј-,4Ј-H) ppm. (3ЈE)-Isomer: 1 C NMR (CDCl
3
5β,10-Dimethyl-des-A-18-nor-androstan-13β-ol (Limdrostanol, 11)/
5
3
):
5β,10-Dimethyl-des-A-18-nor-androstan-14α-ol (26): A solution of
δ ϭ 18.3 (q, 7Ј-Me), 20.1 (q, 8Ј-Me), 20.5 (q, C-9Ј), 23.4 (t, C-4),
3
5
(
2
5 (610 mg, 2.60 mmol) in dry THF (12 mL) was added dropwise
1.0 (t, C-2Ј), 31.2 (t, C-5Ј), 31.3 (t, C-3), 32.4 (t, C-6Ј), 34.7 (t, C-
), 35.2 (t, C-1Ј) 123.3 (d, C-2), 124.1 (s, C-8Ј), 127.4 (s, C-7Ј), 129.9

to a solution of lithium triethylborohydride (Super-Hydride ) in
THF (1 , 7.8 mL, 7.80 mmol), and the reaction mixture was
stirred for 2 days under reflux. The reaction mixture was cooled to
1
3
d, C-3Ј), 130.3 (d, C-4Ј), 144.5 (s, C-1) ppm. (3ЈZ)-Isomer:
C
3
NMR (CDCl ): δ ϭ 18.3 (q, 7Ј-Me), 20.1 (q, 8Ј-Me), 20.5 (q, C-
room temp., and then was taken up in Et
with satd. aq. NH Cl solution (2 ϫ 100 mL), water (100 mL), and
brine (100 mL). The ethereal extract was dried (Na SO ) and con-
centrated under reduced pressure. The resulting residue was puri-
fied by silica gel FC (hexane/Et O, 99:1) to afford 11 (186 mg, 30%)
2
and 26 (237 mg, 38%).
2
O (800 mL) and washed
9
Ј), 23.4 (t, C-4), 25.7 (t, C-2Ј), 26.0 (t, C-5Ј), 31.3 (t, C-3), 32.4 (t,
C-6Ј), 34.7 (t, C-5), 35.2 (t, C-1Ј) 123.3 (d, C-2), 124.1 (s, C-8Ј),
27.4 (s, C-7Ј), 129.5 (d, C-3Ј), 129.6 (d, C-4Ј), 144.4 (s, C-1) ppm.
4
2
4
1
ϩ
ϩ
MS (EI): m/z (%) ϭ 218 (2) [M ], 203 (6) [M Ϫ CH
3
], 162 (9)
ϩ
ϩ
[M
Ϫ C
4 8 6
H ], 83 (100) [C H11 ]. C16H26 (218.4): calcd. C 88.00,
H 12.00; found C 87.96, H 11.99.
Data for 11: R
f
ϭ 0.10 (hexane/Et
2
O, 9:1). IR (Film): ν˜ ϭ 3474 (s,
),
6 6
196 (m, ν CϪO) cm . H NMR (C D ): δ ϭ 0.52 (br. s, 1 H, 13-
OHax), 0.62 (dt, J ϭ 10.5, 5.0 Hz, 1 H, 9-Hax), 0.71 (s, 3 H, 10-
Meax), 0.79 (2t, J ϭ 11.0, 6.0, 1 H, 14-Hax), 0.85 (d, J ϭ 6.5 Hz, 3
5
β,10-Dimethyl-des-A-18-nor-androst-13(14)-ene (24): A solution of
methanesulfonic acid (8.65 g, 90.0 mmol) in CH Cl (50 mL) was
added at 0 °C over 30 min to a solution of 23 (3.93 g, 18.0 mmol)
in CH Cl (150 mL). After stirring for 1 h at 0 °C, the reaction
mixture was poured into Et O (900 mL) and washed with water (2
ϫ 150 mL), satd. aq. NaHCO solution (2 ϫ 200 mL), and brine
100 mL). The organic layer was separated, dried (Na SO ), and
concentrated on a rotary evaporator. The resulting residue was
purified by silica gel FC (hexane/Et O, 99:1; R ϭ 0.66) to afford
the title compound 24 (2.44 g, 62%) as a colorless oil. IR (film):
ν OϪH), 1449 (s, δ CϪH), 1389 (s, δ OϪH), 1365 (m, δ CH
1
3
2
2
Ϫ1 1
2
2
2
H, 5-Meeq), 0.88 (m
J ϭ 6.0, 5.0 Hz, 1 H, 12-Hax), 1.17 (m
.5 Hz, 1 H, 6-Hax), 1.27 (tt, J ϭ 7.5, 2.0 Hz, 1 H, 17-Hax), 1.38
q, J ϭ 12.0 Hz, 1 H, 8-Hax), 1.39 (m , 1 H, 6-Heq), 1.42 (m , 1 H,
, 2 H, 16-Hax,17-Heq), 1.55 (m , 2 H, 11-H ), 1.62
, 1 H, 15-Heq), 1.67 (q, J ϭ 3.0 Hz, 1 H, 7-Heq), 1.78 (quint.,
J ϭ 3.5 Hz, 1 H, 16-Heq), 1.80 (t, J ϭ 3.0 Hz, 1 H, 12-Heq) ppm.
c
, 1 H, 7-Hax), 0.91 (s, 3 H, 10-Meeq), 0.96 (2d,
3
c
, 1 H, 5-Hax), 1.20 (2d, J ϭ
(
2
4
4
(
1
c
c
2
f
5-Hax), 1.50 (m
c
c
2
(m
c
Ϫ1
1
ν˜ ϭ 1455 (s, δ CϪH), 1372 (s, δ
NMR (CDCl ): δ ϭ 0.66 (br. s, 3 H, 10-Meax), 0.72 (s, 3 H, 5-Me),
.86 (s, 3 H, 10-Meeq), 1.12Ϫ1.19 (m, 1 H, 5-H), 1.32Ϫ1.38 (m, 4
3
CH ), 1651 (w, ν CϭC) cm . H
s
3
1
1
H, H NOESY: 10-Meax ϫ 5-Meeq, 8-Hax ϫ 13-OHax, 10-Meax
-Hax, 14-Hax ϫ 12-Hax, 12-Hax ϫ 9-Hax, 5-Hax ϫ 9-Hax, 14-Hax
ϫ 9-Hax, 10-Meax ϫ 6-Hax ppm. 13C NMR (C
): δ ϭ 14.2 (q,
0-Meax), 16.8 (q, 5-Meeq), 21.1 (t, C-16), 21.9 (t, C-11), 26.0 (t, C-
ϫ
0
8
H, 6-,7-H
H
2
2
), 1.58Ϫ1.61 (m, 2 H, 11-H
), 1.96Ϫ1.99 (m, 2 H, 8-,9-H),
) ppm. 13C NMR (CDCl
): δ ϭ 13.8
2
), 1.84Ϫ1.91 (m, 2 H, 16-
6
D
6
2
), 1.92Ϫ1.97 (m, 2 H, 12-H
.09Ϫ2.17 (m, 4 H, 15-,17-H
2
1
1
3
2
3
5), 27.1 (q, 10-Meeq), 31.1 (t, C-6), 32.6 (t, C-7), 36.1 (d, C-8),
6.2 (s, C-10), 36.9 (t, C-12), 39.4 (t, C-17), 42.8 (d, C-5), 52.8 (d,
(
(
q, 10-Meax), 16.4 (q, 5-Me), 26.7 (q, 10-Meeq), 22.3 (t, C-16), 23.6
t, C-11), 27.1 (t, C-12), 31.3 (t, C-6), 32.6 (t, C-7), 36.0 (t, C-15),
C-9), 54.4 (d, C-14), 77.8 (s, C-13) ppm. MS (EI): m/z (%) ϭ 236
37.1 (s, C-10), 37.7 (d, C-8), 38.3 (t, C-17), 42.2 (d, C-5), 51.1 (d,
ϩ
ϩ
ϩ
(
1
10) [M ], 218 (12) [M Ϫ H
2
O], 203 (22) [M Ϫ H
2 3
O Ϫ CH ],
C-9), 133.9 (s, C-13), 138.5 (s, C-14) ppm. MS (EI): m/z (%) ϭ 218
ϩ
94 (100) [C13H22O ]. C16H28O (236.4): calcd. C 81.29, H 11.94;
ϩ
ϩ
ϩ
(45) [M ], 203 (100) [M Ϫ CH
3
], 162 (34) [M Ϫ C
4 8
H ], 109 (31)
found C 81.12, H 11.96. Odor: Woody, earthy, sweet. Odor thresh-
old: 90 ng/L air.
ϩ
[C H13 ]. C16H26 (218.4): calcd. C 88.00, H 12.00; found C 87.98,
H 12.05.
8
Data for 26: R
f 2
ϭ 0.12 (hexane/Et O, 9:1). M.p. 55.3Ϫ58.1 °C. IR
5
β,10-Dimethyl-13,14-epoxy-des-A-18-nor-androstane (25): A solu-
tion of 3-chloroperbenzoic acid (70%, 2.61 g, 10.6 mmol) in
CH Cl (40 mL) was added dropwise at 0 °C to a stirred solution
of 24 (2.31 g, 10.6 mmol) in CH Cl (50 mL), and the resulting sus-
(KBr): ν˜ ϭ 3505 (s, ν OϪH), 1450 (s, δ CϪH), 1389 (s, δ OϪH),
1364 (s, δ CH ), 1139 (m, ν CϪO) cm . H NMR (C D ): δ ϭ
3 6 6
Ϫ1
1
2
2
0.57 (br. s, 14-OHax), 0.66 (s, 3 H, 10-Meax), 0.85 (d, J ϭ 6.5 Hz,
3 H, 5-Meeq), 0.91 (s, 3 H, 10-Meeq), 0.92 (dt, J ϭ 4.0, 3.0 Hz, 1
2
2
pension was stirred for an additional 4 h at 0 °C. The reaction H, 11-Hax), 0.98 (quint., J ϭ 4.0 Hz, 1 H, 13-Hax), 1.10 (2d, J ϭ
mixture was then diluted with Et O (800 mL) and washed with aq. 11.0 Hz, 1 H, 8-Hax), 1.13 (m , 1 H, 5-Hax), 1.17 (d, J ϭ 4.0 Hz, 1
Na SO solution (10%, 2 ϫ 70 mL), satd. aq. NaHCO (3 ϫ H, 9-Hax), 1.18 (m , 1 H, 15-Hax), 1.19 (m , 1 H, 6-Hax), 1.38 (t,
00 mL), and water (2 ϫ 150 mL). The organic extract was dried J ϭ 4.0 Hz, 2 H, 7-Hax,12-Hax), 1.40 (m , 1 H, 6-Heq), 1.50 (m , 1
Na SO ) and concentrated to dryness on a rotary evaporator. Re- H, 7-Heq), 1.51 (quint., J ϭ 3.0, 1 H, 16-Hax), 1.54 (m , 2 H, 12-
peated silica gel FC (hexane/Et O, 99:1) of the crude material fur- eq,15-Heq), 1.57 (m , 2 H, 17-H ), 1.72 (quint., J ϭ 3.5 Hz, 1 H,
nished the title compound 25 (695 mg, 28%; R ϭ 0.12). IR (film): 16-Heq), 1.74 (m , 1 H, 11-Heq) ppm. H, H NOESY: 10-Meax
ν˜ ϭ 1455 (s, δ CϪH), 890/920 (s, ν CϪOϪC), 1372 (s, δ CH
), 8-Hax, 10-Meax ϫ 11-Hax, 10-Meax ϫ 5-Meeq, 10-Meax ϫ 6-Hax, 9-
2
c
2
3
3
c
c
1
c
c
(
2
4
c
2
H
c
2
1
1
f
c
ϫ
3
Ϫ1
1
13
3
3
5
050 (w, ν CϪH, epoxide) cm . H NMR (CDCl
H, 10-Meax), 0.77Ϫ0.80 (m, 1 H, 9-H), 0.81 (d, J ϭ 6.0 Hz, 3 H,
-Me), 0.88 (s, 3 H, 10-Meeq), 1.12Ϫ1.19 (m, 1 H, 5-H), 1.31Ϫ1.37 (t, C-12), 26.5 (t, C-11), 27.1 (q, 10-Meeq), 27.2 (t, C-7), 28.2 (t, C-
3
): δ ϭ 0.61 (s,
H
ax ϫ 14-OHax, 7-Hax ϫ 14-OHax, 12-Hax ϫ 14-OHax
.
C NMR
(C ): δ ϭ 13.9 (q, 10-Meax), 16.8 (q, 5-Meeq), 21.1 (t, C-16), 25.4
6
D
6
(
m, 2 H, 6-H 17), 30.8 (t, C-6), 36.2 (s, C-10), 37.3 (t, C-15), 42.4 (d, C-5), 45.7
2
), 1.38Ϫ1.43 (m, 2 H, 7-H
2
), 1.57Ϫ1.58 (m, 4 H, 12-
,
1
17-H ), 1.62Ϫ1.65 (m, 2 H, 11-H ), 1.69Ϫ1.76 (m, 2 H, 16-H ), (d, C-8), 46.4 (d, C-9), 48.8 (d, C-13), 80.6 (s, C-14) ppm. MS (EI):
2
2 2
), 1.93Ϫ1.97 (m, 2 H, 8-,9-H) ppm. 13C m/z (%) ϭ 236 (25) [M ], 218 (3) [M Ϫ H
ϩ
ϩ
ϩ
.79Ϫ1.85 (m, 2 H, 15-H
2
2
O], 203 (9) [M
2 3
H O Ϫ CH ], 194 (100) [C13H22O ]. C16H28O (236.4): calcd. C
Ϫ
ϩ
NMR (CDCl
3
): δ ϭ 13.4 (q, 10-Meax), 16.5 (q, 5-Me), 19.7 (t, C-
1
2
6), 22.9 (t, C-11), 26.7 (q, 10-Meeq), 26.7 (t, C-15), 28.7 (t, C-12), 81.29, H 11.94; found C 81.26, H 11.92. Odor: earthy, ambery.
9.6 (t, C-17), 30.8 (t, C-6), 32.4 (t, C-7), 35.6 (s, C-10), 37.4 (d, C-
Odor threshold: 46 ng/L air.
Eur. J. Org. Chem. 2004, 4995Ϫ5002
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5001

P. Kraft, K. Popaj
FULL PAPER
37, 57Ϫ78. [8b] J. N. Lundström, M. Gon c¸ alves, F. Esteves, M.
J. Olsson, Hormones and Behaviour 2003, 44, 395Ϫ401.
J. N. Lundström, T. Hummel, M. J. Olsson, Chem. Senses 2003,
Direct Synthesis of 5β,10-Dimethyl-des-A-18-nor-androstan-13β-ol
(Limdrostanol, 11) from (3ЈE/Z)-1-(7Ј,8Ј-Dimethylnona-3Ј,7Ј-dienyl)-
[
9]
cyclopent-1-ene (23): Methanesulfonic acid (892 mg, 9.28 mmol)
was added dropwise at 0 °C to a stirred solution of 23 (2.54 g,
28, 643Ϫ650.
[10]
D. Berliner (Erox Corp.), EP 562Ј843, prior. 24 March, 1992,
1
1.6 mmol) in HCOOH (20 mL), and the resulting slightly red solu-
tion was stirred for an additional 6 h at 0 °C. The reaction mixture
was then diluted with Et O (1 L) and washed with water (2 ϫ
00 mL), satd. aq. NaHCO (3 ϫ 150 mL), and brine (200 mL).
The organic layer was dried (Na SO ) and concentrated under re-
[
Chem. Abstr. 1993, 119, 278388, AN 1993:678388].
[11]
J. A. S. Adams, S. Clunas, A. B. Turner, Tetrahedron Lett. 2002,
4
2
3, 9559Ϫ9560.
[
[
[
12]
13]
14]
2
3
K.-H. Schulte-Elte, W. Giersch, B. Winter, H. Pamingle, G.
Ohloff, Helv. Chim. Acta 1985, 68, 1961Ϫ1985.
E. Klein, W. Rojahn (Dragoco AG), DE 2Ј807Ј584, prior. 30
August, 1979, [Chem. Abstr. 1980, 92, 11114, AN 1980:11114].
K.-H. Schulte-Elte, G. Ohloff, B. L. Müller, W. K. Giersch
2
4
duced pressure. The resulting residue was taken up in aq. NaOH/
EtOH (1 , 1:1, 60 mL), and the mixture was stirred at room temp.
overnight. The reaction mixture was then poured into water
(
Firmenich SA), EP 118Ј809, prior. 9 March, 1984, [Chem.
(
150 mL), and the product was extracted with Et
The ethereal extract was washed with brine (150 mL) and dried
Na SO ), and the solvents were evaporated on a rotary evaporator.
The resulting residue was purified by silica gel FC (hexane/Et O,
9:1) to afford 11 (84.9 mg, 3%) as a colorless oil. The spectro-
2
O (3 ϫ 300 mL).
Abstr. 1985, 102, 79179, AN 1985:79179].
[15]
K.-H. Schulte-Elte, H. Pamingle (Firmenich SA), EP 121Ј828,
prior. 12 April, 1983 [Chem. Abstr. 1985, 102, 166981, AN
(
2
4
2
1
985:166981].
[
[
16]
17]
9
N. E. Schore, M. J. Knudsen, J. Org. Chem. 1987, 52, 569Ϫ580.
P. Cannone, G. B. Foscolos, D. B e´ langer, J. Org. Chem. 1980,
45, 1828Ϫ1835.
scopic data for the isolated compound were identical in all respects
to those reported for 11 above.
[
18] [18a]
R. Appel, Angew. Chem. 1975, 87, 863Ϫ874; Angew. Chem.
Int. Ed. Engl. 1975, 14, 801Ϫ821. [
Holmes, J. B. Lee, Chem. Ind. 1966, 900Ϫ901.
P. J. Kocienski, G. Cernigliaro, G. Feldstein, J. Org. Chem.
18b]
I. M. Downie, J. B.
Acknowledgments
[19]
1977, 42, 353Ϫ355.
We are grateful to Professor Dr. Georg Fr a´ ter, Givaudan, and Pro-
fessor Dr. Andrea Vasella, ETH Zurich, for fruitful discussions.
Thanks are also due to Dr. Gerhard Brunner for multidimensional
NMR experiments, to Katarina Grman for GC-threshold determi-
nations, and to Jean-Jacques Rouge for the odor descriptions. Care-
ful proofreading of the manuscript by Professor Dr. Georg Fr a´ ter,
[20]
H. C. Brown, A. M. Salunkhe, Tetrahedron Lett. 1993, 34,
265Ϫ1268.
1
[21]
H. J. Bestmann, P. Rösel, O. Vostrowsky, Liebigs Ann. Chem.
979, 1189Ϫ1204.
1
[
[
22]
23]
R. Marbet, G. Saucy, Helv. Chim. Acta 1967, 50, 2095Ϫ2100.
P. Y. Reshetnikov, L. S. Shevnitsyn, V. M. Romanov, V. V. Lezh-
enin, G. K. Azhigaliev, A. G. Shkuro, V. M. Kozlov, Y. I. Bar-
anov, N. K. Malinin (USSR), SU 80Ϫ2Ј946Ј109, prior. 26 June,
c
Dr. Markus Gautschi, and John Anthony M Stea is also acknowl-
edged with gratitude.
1980, [Chem. Abstr. 1983, 98, 89668, AN 1983:89668].
[24]
L. Fitjer, U. Quabeck, Synth. Commun. 1985, 15, 855Ϫ864.
M. Schlosser, K. F. Christmann, Justus Liebigs Ann. Chem.
1967, 708, 1Ϫ35.
26] [26a]
[
1]
[25]
Givaudan Schweiz AG postdoctoral fellow, Fragrance Re-
search, 8600 Dübendorf, Switzerland, January 2003 until De-
cember 2004.
P. Karlson, M. Lüscher, Nature 1959, 183, 55Ϫ56.
G. C. Perry, R. L. S. Patterson, K. J. H. MacFie, C. G. Stinson,
Anim. Prod. 1980, 31, 191Ϫ199.
[
M. B. Gravestock, W. S. Johnson, B. E. McCarry, R. J.
[
[
2]
3]
Parry, B. E. Ratcliffe, J. Am. Chem. Soc. 1978, 100, 4274Ϫ4282.
[26b]
W. S. Johnson, N. P. Jensen, J. Hooz, E. J. Leopold, J. Am.
Chem. Soc. 1968, 90, 5872Ϫ5881.
[
4] [4a]
[27] [27a]
V. Prelog, L. Ruzicka, Helv. Chim. Acta 1944, 27, 61Ϫ66.
V. Prelog, L. Ruzicka, P. Wieland, Helv. Chim. Acta 1944,
W. S. Johnson, T. M. Yarnell, R. F. Myers, D. R. Morton,
[
4b]
[27b]
Tetrahedron Lett. 1978, 19, 2549Ϫ2552.
Chem. Soc., Rev. 1980, 9, 265Ϫ280.
[28]
K. J. Sutherland,
[
4c]
2
7, 66Ϫ71.
V. Prelog, L. Ruzicka, P. Meister, P. Wieland,
[
4d]
Helv. Chim. Acta 1945, 28, 618Ϫ627.
L. Ruzicka, V. Prelog,
R. L. Snowden, J. C. Eichenberger, S. M. Linder, P. Sonnay, C.
P. Meister, Helv. Chim. Acta 1945, 28, 1651Ϫ1660. [ L. Ruz-
4e]
Vial, K.-H. Schulte-Elte, J. Org. Chem. 1992, 57, 955Ϫ960.
[
[
29] [29a]
icka, P. Meister, V. Prelog, Helv. Chim. Acta 1947, 30, 867Ϫ878.
G. Stork, W. A. Burgstahler, J. Am. Chem. Soc. 1955, 77,
[
5] [5a]
[5b]
[29b]
A. Comfort, Nature 1971, 230, 432Ϫ433 and 479.
J. N.
5068Ϫ5077.
A. Eschenmoser, L. Ruzicka, O. Jeger, D. Ari-
Labows, Perfum. Flavor. 1979, 4(4), 12Ϫ17. [ R. L. Doty,
5c]
goni, Helv. Chim. Acta 1955, 38, 1890Ϫ1904.
30]
Chem. Senses 1981, 6, 351Ϫ376.
R. Dahl, Switch Bitch, Penguin Books, London, 1976, pp.
112Ϫ140; (German edition: R. Dahl) Kuschelmuschel, Rowohlt
Taschenbuch Verlag, 1978, pp. 99Ϫ124.
Received August 13, 2004
[
[
[
6] [6a]
[6b]
M. K. McClintock, Nature 1971, 229, 244Ϫ245.
A.
Weller, L. Weller, Neurosci. Biobehav. Rev. 1993, 17, 427Ϫ439.
K. Shinohara, M. Morofushi, T. Funabashi, D. Mitsushima, F.
Kimura, Chem. Senses 2000, 25, 465Ϫ467.
7]
Early View Article
Published Online November 10, 2004
8] [8a]
S. Jacob, M. K. McClintock, Hormones and Behaviour 2000,
5002
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 4995Ϫ5002