958 CHEMISTRY & BIODIVERSITY – Vol.

5 (2008)

Cetalox
and Analogues: Synthesis via Acid-Mediated Polyene Cyclizations

by Roger L. Snowden
Firmenich SA, Corporate R&D Division, P. O. Box 239, CH-1211 Geneva 8
(phone: þ 41 22 780 36 08; fax: þ 41 22 780 33 34; e-mail: roger.snowden@firmenich.com)

Using a novel, acid-mediated cyclization methodology, a direct access to Cetalox2 ((  )-1; a

commercially important ambergris-type odorant) and various structurally related didehydro (i.e., 19, 26,
and 30) and tetradehydro (i.e., 28 and 37/38) analogues is described. Treatment of either (E,E)-14 or (E)-
15 with an excess of FSO3H in 2-nitropropane at  908 stereospecifically afforded (  )-1 in 40 and 42%
yield, respectively. Under similar conditions, cyclization of (E)-18 or 20 furnished 19 in 60 and 64% yield,
respectively. Analogously, using an excess of ClSO3H in CH2Cl2 at  808, 26 is formed with high
stereoselectivity by cyclization of either (E)-24 or (Z)-25 (52 and 31% yield, resp.); in the same manner,
28 was prepared from 27 (22% yield). The same principle was applied to the synthesis of racemic
Superambrox (30), via cyclization of 35, but only with poor selectivity (22%) and low yield (7%).
Another approach via cyclization of (E)-40 under solvolysis conditions (excess TFA in CH2Cl2 at  108)
gave a higher yield (15%) with improved selectivity (43%). Finally, cyclization of 34 (1 : 1
diastereoisomer mixture) afforded 37/38 (10 : 1) in 27% yield. The qualitative organoleptic properties
of 19, 26, 28, 30, and 37/38 (10 : 1) are briefly discussed.

Introduction. – In 1950 [1], Stoll and co-workers at Firmenich reported the

preparation of Ambrox2 (()-1) 1) from sclareol, a naturally occurring diterpene
present in clary sage. Right up to the present day, ()-1 has been highly valued in the
fragrance industry for its dry-woody tonality and exceptional fixative properties [3]. By
the early 1980s, the commercial use of ()-1 was no longer protected by patent, and led
to intense competition for access to sclareol and other structurally related natural
products such as manool or abienol, from which ()-1 can be synthesized (Scheme 1).
It was for this reason that the unknown racemate (  )-1 (afterwards named
Cetalox21)) became a viable synthetic target, as it was realized that its putative access
from a non-natural raw material would avoid one of the problems inherent to ()-1,
namely the latterAs dependence on an unreliable natural source whose price is often
subject to important fluctuations. In 1986, the first reported synthesis of (  )-1 had been
published by Japanese workers [4] (see Scheme 2). Within Firmenich, two laboratory
syntheses of (  )-1 had already been developed by the mid-1980s (see Schemes 3 [5]
and 4 [6]), and had provided sufficient material to show that this ingredient indeed had
commercial potential, displaying a strong ambery-woody tonality. However, none of
these syntheses was industrially feasible, and thus there was a real need for a more
direct access to (  )-1. We describe herein an alternative synthesis of (  )-1 via an
unprecedented acid-mediated cyclization strategy, and also explain how this new

1) Ambrox2 (()-1) and Cetalox2 ((  )-1) [2] are trade names of Firmenich SA.

I 2008 Verlag Helvetica Chimica Acta AG, ZKrich

 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008) 959

Scheme 1. Natural Sources for Ambrox2

methodology was used to prepare structurally-related analogues of Cetalox2, with

similar organoleptic properties.

Scheme 2. First Synthesis of Cetalox2 [4]

i) ClCH2CO2Et, EtONa, 58; cat. AcONa, 2008. ii) CH2(CO2H)2 , Et3N, reflux. iii) 15% Ethanolic KOH.
iv) CF3CO2H, 08. v) NaAlH2(OCH2CH2OCH3 )2 (Vitride2 ), 608. vi) TsCl, pyridine, 08.

Synthesis of (  )-1 via Acid-Mediated Cyclization. – The synthesis of (  )-1 in 40

and 42% yield, respectively, via the acid-mediated cyclization (fluorosulfonic acid
(FSO3H),  908) of either (E,E)-homofarnesol ((E,E)-14) or the monocyclic
homoallylic alcohol (E)-15, its constitutional isomer, is shown in Scheme 5 [7].
960 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008)

Scheme 3. First Firmenich Synthesis of Cetalox2 [5]

i) SnCl4 , CH2Cl2 , 108. ii) NaH, allyl bromide, DMF, r.t.; xylene, reflux. iii) CaCl2 · 2 H2O, DMSO, reflux.
iv) MeMgI, Et2O. v) O3 , MeOH,  208. vi) NaBH4 , MeOH, r.t. vii) Cat. TsOH · H2O, MeNO2 , 308.

Scheme 4. Second Firmenich Synthesis of Cetalox2 [6]

i) LiAlH4 , Et2O, r.t. ii) MsCl, pyridine, 08. iii) Pyridinium chlorochromate (PCC), AcONa, CH2Cl2 , r.t.
iv) 1,5-Diazabicyclo[4.3.0]non-5-ene (DBN), toluene, r.t. v) KCN, NH4Cl, DMF/H2O, r.t. vi) KOH,
MeOH/H2O, reflux. vii) MeLi, THF/H2O,  608 to r.t. viii) Cat. TsOH · H2O, toluene, reflux. ix) Cat.
TsOH · H2O, MeNO2 , 308.
 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008) 961

Scheme 5. Synthesis of Cetalox2 via Acid-Mediated Cyclization [7]

i) FSO3H (10 mol-equiv.), 2-nitropropane,  908.

These kinetically controlled, stereospecific reactions represent rare examples of

acid-mediated polyene cyclizations in which the initiating group is an alkene, and where
termination is effected internally by a OH group 2 ). Mechanistically, the cyclization of
(E,E)-14 can proceed by either a concerted process via chair-like nascent cyclohexane
rings and anti-addition across the C¼C bonds [9], or a stepwise process via a
conformationally frozen cyclohexyl cation with a kinetic preference for CC bond-
formation [10]. In contrast, the fact that the cyclization of (E)-15 leads to a trans-
decalin skeleton means that this cyclization is non-concerted. Thus, stereoselective
axial protonation of the cyclohexenyl bond is followed by anti-addition across the side-
chain C¼C bond with kinetically controlled equatorial CC bond formation.
Subsequent work [11], involving the cyclization of Me-substituted analogues of (E)-
15, reinforced our mechanistic hypothesis and demonstrated that the efficiency of this
reaction is independent of the nature of the OH group, which may be primary,
secondary, or tertiary.

Synthesis of Ambrox
(()-1) via Enantioselective Acid-Mediated Cyclization. – It

should be mentioned that there were many fruitless attempts within Firmenich to
perform the cyclization of (E,E)-14 enantioselectively, the goal being a possible direct
access to Ambrox2 (()-1). However, several years ago, it was Yamamoto and co-
workers who published the transformation of (E,E)-16, the triethylsilyl ether of (E,E)-
14, to ()-1 (75% ee, ca. 50% yield) using an excess of a chiral Lewis acid based on
binaphthol, in conjunction with a Brønsted acid, CF3COOH, in EtNO2 at  708 [12]
(see Scheme 6). However, to our knowledge, this elegant approach has not yet been
commercialized.

2) For other examples of this reaction, the pioneering work of Smit and co-workers, and Vlad et al.
deserves special mention [8].
962 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008)

Scheme 6. Synthesis of Ambrox2 via an Enantioselective Acid-Mediated Cyclization

Synthesis of Cetalox
Analogues. – In parallel to the central activity of developing

an industrial synthesis of Cetalox2 ((  )-1), there was also a research program directed
towards the search for new fragrance ingredients even more powerful than (  )-1. In
this context, it was realized that this new acid-mediated cyclization reaction could also
be applied to the preparation of closely related analogues of (  )-1 which satisfied the
structural requirements of the Ptriaxial ruleA, a useful empirical theory of the 1970s
related to the ambergris-type odor [13]. In this rule, a multi-point interaction between a
dimethyl-trans-decalin skeleton and the complementary receptor site is considered to
be associated with the axial orientation of the substituents R’, R’’, and Ra , as shown in
structure A (Scheme 7). Accordingly, it was decided to prepare the racemates of 19 and
26, two didehydro analogues of (  )-1.

Synthesis of 19. – The racemic allenic alcohols (E)-18 and 20 were prepared from
unsaturated ketones (E)-19 and 2, respectively, by using a modification of a reported
procedure [14]. These two compounds were then submitted to identical conditions
(FSO3H, 2-nitropropane,  908 to  308), resulting in the stereoselective (88 and 84%
stereoselectivity, resp.) formation of 19 in 60 and 64% yield, respectively [15]
(Scheme 8). These acid-mediated polyene cyclizations, involving an alkene as the
initiating group and an allenic alcohol as the terminating group, are unprecedented. As
before, the stereochemical course of cyclization is directed by the conformation of an
intermediate cyclohexyl cation, and the process is non-synchronous. Moreover, it is
interesting to note that the catalytic hydrogenation of 19 shows complete b-face
selectivity, affording tricyclic ether 21 in 92% yield, and thus offers a novel access to this
powerful, woody odorant 3 ).

3) In comparison to its stereoisomers, ()-21 was reported to possess the strongest scent and lowest
threshold [16]; for a stereoselective synthesis of ()-21, see [17].
 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008) 963

Scheme 7. The PTriaxial RuleA for Ambergris-Type Odor [13]

Scheme 8. Synthesis of 19 via Acid-Mediated Cyclization [15]

i) Prop-2-yn-1-ol, KOH, THF, r.t. ii) Ac2O, Et3N, r.t. iii) Ethyl vinyl ether, cat. TsOH · H2O, toluene,
 208. iv) LiAlH4 , THF, toluene, r.t. v) FSO3H (2.3 mol-equiv.), 2-nitropropane,  908 to  308. vi) H2 ,
cat. 10% Pd/C, toluene, r.t.
964 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008)

Synthesis of 26. – Bicyclic homoallylic alcohols (E)-24 and (Z)-25 were prepared
from d-pyronene (1,1-dimethyl-2,3-dimethylidenecyclohexane; 22) as shown in
Scheme 9. These two compounds were then submitted to identical conditions
(chlorosulfonic acid (ClSO3H), CH2Cl2 ,  808), resulting in the stereoselective (74
and 91% selectivity, resp.) formation of 26 in 52 and 31% yield, respectively [17].

Scheme 9. Synthesis of 26 via Acid-Mediated Cyclization [18]

i) Br2 , CH2Cl2 , 08. ii) Dimethyl 3-oxoglutarate, NaHCO3 , H2O/CH2Cl2 , r.t. iii) NaCl, H2O, DMSO, 128 –
1358. iv) [Ph3P(CH2 )3OH] + Br 7, BuLi (2 mol-equiv.), THF,  208 to r.t. v) NaH, THF; then tBuMe2SiCl,
52 – 548. vi) Pyridinium dichromate (PDC) (4 mol-equiv.), tBuO2H (3.1 mol equiv.), toluene, r.t. vii)
TsNHNH2 , EtOH, reflux. viii) Catecholborane (2 mol-equiv.), CHCl3 , 58; AcONa · 3 H2O, reflux. ix)
ClSO3H (5 mol-equiv.), CH2Cl2 ,  808.

Synthesis of 28/29 (1.4 : 1) via Acid-Mediated Cyclization. – Bicyclic allenic alcohol

27, prepared from 23 as shown in Scheme 10, was submitted to the cyclization conditions
indicated (ClSO3H, 2-nitropropane,  808) to afford 28/29 (1.4 : 1) in 37% yield [18].
Separation of these two stereoisomers was possible by chromatography.

Organoleptic Properties of 19, 26, and 28. – In comparison with (  )-1, 19, despite a
lower substantivity, has a similar intensity, exhibiting the same characteristic amber and
woody notes. Concerning 26 and 28, due to the small amounts of material available,
only qualitative odor evaluations were effected; both 26 and 28 were perceived as
possessing a tenacious, amber odor fully comparable to that of (  )-1.

Synthesis of (  )-Superambrox (30) via Acid-Mediated Cyclization. – Although

not strictly an example of the Ptriaxial ruleA, lacking an axial substituent at C(5a), the
didehydro analogue of Ambrox2, ()-30, aptly named Superambrox by Firmenich
perfumers, due to its extremely powerful, ambergris tonality, was first synthesized in the
 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008) 965

Scheme 10. Synthesis of 28/29 (1.4 : 1) via Acid-Mediated Cyclization [18]

i) Prop-2-yn-1-ol, KOH, THF, r.t. ii) Ac2O, Et3N, r.t. iii) Ethyl vinyl ether, cat. TsOH · H2O, toluene,
 208. iv) LiAlH4 , THF, toluene, r.t. v) ClSO3H (9 mol-equiv.), 2-nitropropane,  808.

1980s by Escher, starting from either (þ)-larixyl acetate or (þ)-sclareolide ((þ)-5) [19]
(see Scheme 11) 4 ).

Scheme 11. First Synthesis of Superambrox [19]

As in the case of Cetalox2, it was thus decided to develop a synthesis of the

unknown racemate 30, a potential odorant whose accessibility would also not be
dependent on a natural source. Accordingly, homoallylic alcohol (E)-35, prepared from
dehydrolinalool (3,7-dimethyloct-6-en-1-yn-3-ol; 31) as outlined in Scheme 12, was
submitted to the cyclization conditions indicated (ClSO3H, 2-nitropropane,  808) to
afford a chromatographically separable 3.8 : 1 mixture of the tricyclic ethers 36 and 30,
in 30% yield [21]. Unfortunately, this kinetically controlled cyclization gives the
desired product 30 with poor selectivity, a result which is readily explained by
comparison of the MM2 energies of the respective transition states. A side-product
from the aforegoing synthesis of (E)-35 was the allenic alcohol 34 (1 : 1 diastereoisomer
mixture; see Scheme 12), whose acid-mediated cyclization under conditions identical to
those employed for (E)-35 furnished 37/38 (10 : 1) in 27% yield. Once again, this result
may be rationalized by a kinetically controlled, stepwise process.

4) For a later synthesis of ()-30, see [20].

966 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008)

Scheme 12. First Synthesis of (  )-Superambrox via Acid-Mediated Cyclization [20]

i) 2-Methoxypropene, cat. hydroquinone, cat. TsOH · H2O, petroleum ether, 808. ii) 3-[(Trimethylsilyl)-
oxy]prop-1-yne, BuLi, THF,  908; 5% aq. HCl, r.t. iii) Ac2O, Et3N, cat. 4-(dimethylamino)pyridine
(DMAP), CH2Cl2 , 28. iv) Ethyl vinyl ether, cat. TsOH · H2O, toluene,  108 to 08. v) LiAlH4 , THF, 208.
vi) Ac2O, Et3N, cat. DMAP, toluene, r.t. vii) LiAlH4 , THF, toluene, reflux. viii) ClSO3H (6 mol-equiv.),
2-nitropropane,  808.
a
) Compound 33 was obtained as a 1 : 1 : 1 : 1 diastereoisomer mixture. b ) Compound 34 was obtained as a
1 : 1 diastereoisomer mixture.

Another cyclization route to 30 was also elaborated (see Scheme 13). This second
approach is fundamentally different from the first approach in that, although
cyclization termination is effected internally by a OH group, the initiating group is
 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008) 967

now an allylic alcohol rather than an alkene 5 ). The targeted cyclization substrate, diol
(E)-40, was synthesized from b-cyclocitral (2,6,6-trimethylcyclohex-1-ene-1-carbalde-
hyde; 39) in eight steps (17% overall yield). Not unexpectedly, the acid-mediated
cyclization of (E)-40 using ClSO3H in 2-nitropropane at  808 was unsuccessful,
affording complex mixtures containing only small amounts (ca. 5%) of tricyclic ethers
36 and 30. Because solvolysis of the allylic CO bond of (E)-40 to the cyclohexane-
based allylic cation is required to trigger the cyclization, the best conditions were found
to be the use of a large excess of CF3COOH (TFA) in CH2Cl2 at  108. Subsequent
workup afforded a 1.3 : 1 mixture of 36 and 30 in 35% yield. The selectivity (43%) is low
for the desired product, but, as discussed previously, can be explained by inspection of
the respective transition states leading to 36 and 30.

Scheme 13. Second Synthesis of (  )-Superambrox via Acid-Mediated Cyclization [20]

i) 3-Chloro-2-methylprop-1-ene, Mg, Et2O, reflux. ii) NaH, tBu(Me)2SiCl, THF, reflux. iii) 1,9-
Borabicyclo[3.3.1]nonane (9-BBN), THF, r.t.; 2.3n aq. NaOH soln., 35% aq. H2O2 soln., 458. iv) Oxalyl
chloride, DMSO, Et3N, CH2Cl2 ,  558 to r.t. v) (MeO)2P(O)CH2CO2Me, MeONa/MeOH, toluene, r.t.
vi) 1,8-Diazobicyclo[5.4.0]undec-7-ene (DBU), 95 – 988. vii) LiAlH4 , THF, reflux. viii) Bu4NF · 3 H2O,
THF, reflux. ix) CF3CO2H (32 mol-equiv.), CH2Cl2 ,  108.

Our two syntheses of 30 via acid-mediated cyclizations had provided enough

material for a positive organoleptic evaluation, but due to the lack of stereocontrol,
were evidently unsuitable for further development. Fortunately, research work within
Firmenich continued, and recently the first stereocontrolled synthesis was reported by
Fehr [23]. This synthesis, starting from aldehyde 3, is outlined in Scheme 14.

Organoleptic Properties of 30 and 37/38 (10 : 1). – Due to the limited amounts of
compounds available, only qualitative odor evaluations were possible. Nevertheless, in
comparison with Cetalox2, both 30 and 37/38 (10 : 1) were perceived as possessing a
similarly strong, ambery-woody tonality.

5) For literature precedents for this type of cyclization, see [22].

968 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008)

Scheme 14. First Stereocontrolled Synthesis of (  )-Superambrox [22]

i) EtAlCl2 (1 mol-equiv.), CH2Cl2 , 08. ii) Cl3CCHO (2 mol-equiv.), 08. iii) Lithium diisopropylamide
(LDA; 1.25 mol-equiv.), THF,  58; Me3SiCl (2.4 mol-equiv.). iv) m-Chloroperbenzoic acid (mCPBA;
1.1 mol-equiv.), CH2Cl2 , 08. v) Ac2O/H2O (10 : 1), cat. TsOH · H2O, r.t. vi) LDA (1.25 mol-equiv.), THF,
 708. vii) SOCl2 (5 mol-equiv.), pyridine,  308. viii) LiAlH4 , Et2O, r.t. ix) Cat. [RuH(h5-C8H11)2 ] +
[BF4 ] 7, CH2Cl2 , r.t. x) Et3SiH (2 mol-equiv.), Amberlyst2 15, r.t.

Conclusion. – With the aim to discover new, commercially important fragrance

ingredients, synthetic access to several didehydro and tetradehydro analogues of
Cetalox2, the racemate of the norlabdane tricyclic ether Ambrox2 (()-1), has been
achieved using novel, Brønsted acid-mediated cyclization reactions. These polyene
cyclizations are particularly of interest, as they represent rare examples in which the
initiating group is an alkene, and the internal terminating group is an alcohol. Although
these cyclizations do not always occur with high stereoselectivity and proceed in only
fair yields, this methodology is often suited to access highly strained tricyclic di- and
tetrahydrofurans that would be difficult to prepare by other means.

REFERENCES
[1] M. Stoll, M. Hinder, Helv. Chim. Acta 1950, 33, 1251; M. Hinder, M. Stoll, Helv. Chim. Acta 1950, 33,
1308.
[2] C. Chauffat, A. Morris, Perfum. Flavor. 2004, 29(2), 34.
[3] G. Ohloff, in PFragrance ChemistryA, Ed. E. T. Theimer, Academic Press, New York, 1982, p. 535.
[4] T. Kawanobe, K. Kogami, M. Matsui, Agric. Biol. Chem. 1986, 50, 1475.
[5] G. BKchi, H. WKest, Helv. Chim. Acta 1989, 72, 996.
[6] R. L. Snowden, S. M. Linder, Tetrahedron Lett. 1991, 32, 4119.
[7] R. L. Snowden, J.-C. Eichenberger, S. M. Linder, P. Sonnay, C. Vial, K.-H. Schulte-Elte, J. Org.
Chem. 1992, 57, 955.
[8] G. E. Muntyan, M. Kurbanov, V. A. Smit, A. V. Semenovskii, V. F. Kucherov. Izv. Akad. Nauk SSSR
Ser. Khim. 1973, 633; P. F. Vlad, N. D. Ungar, V. B. Perutskii, Khim. Prir. Soedin 1986, 793; P. F.
Vlad, N. D. Ungar, V. B. Perutskii, Khim Geterotsikl Soedin SSSR 1990, 26, 896.
 CHEMISTRY & BIODIVERSITY – Vol. 5 (2008) 969

[9] G. Stork, W. A. Burgstahler, J. Am. Chem. Soc. 1955, 77, 5068; A. Eschenmoser, L. Ruzicka, O.
Jeger, D. Arigoni, Helv. Chim. Acta 1955, 38, 1890.
[10] W. S. Johnson, Trans. N.Y. Acad. Sci. 1967, 29, 1001; M. Nishizawa, H. Takenaka, Y. Hayashi, J. Am.
Chem. Soc. 1985, 107, 522; M. Nishizawa, H. Takenaka, Y. Hayashi, J. Org. Chem. 1986, 51, 806.
[11] R. L. Snowden, J.-C. Eichenberger, W. Giersch, W. Thommen, K.-H. Schulte-Elte, Helv. Chim. Acta
1993, 76, 1608.
[12] K. Ishihara, S. Nakamura, H. Yamamoto, J. Am. Chem. Soc. 1999, 121, 4906; S. Nakamura, K.
Ishihara, H. Yamamoto, J. Am. Chem. Soc. 2000, 122, 8131; K. Ishihara, H. Ishibashi, H. Yamamoto,
J. Am. Chem. Soc. 2002, 124, 3647.
[13] G. Ohloff, F. NQf, R. Decorzant, W. Thommen, E. Sundt, Helv. Chim. Acta 1973, 56, 1414.
[14] J. S. Cowie, P. D. Landor, S. R. Landor, J. Chem. Soc., Perkin Trans. 1 1973, 720.
[15] R. L. Snowden, J.-C Eichenberger, S. Linder, P. Sonnay, Helv. Chim. Acta 2004, 87, 1711.
[16] G. Ohloff, W. Giersch, W. Pickenhagen, A. Furrer, B. Frei, Helv. Chim. Acta 1985, 68, 2022.
[17] L. A. Paquette, R. E. Maleczka Jr., J. Org. Chem. 1991, 56, 912.
[18] R. L. Snowden, S. Linder, Helv. Chim. Acta 2005, 88, 3055.
[19] R. L. Snowden, S. D. Escher, to Firmenich SA, Eur. Pat. EP 558928, 1993 (Chem. Abstr. 1994, 120,
107408).
[20] M. G. Bolster, B. J. M. Jansen, A. de Groot, Tetrahedron 2001, 57, 5663.
[21] R. L. Snowden, S. Linder, Helv. Chim. Acta 2006, 89, 3071.
[22] M. E. Garst, Y.-F. Cheung, W. S. Johnson, J. Am. Chem. Soc. 1979, 101, 4404; M. Wolf, U. MQtzel, E.-
J. Brunke, E. Klein, Tetrahedron Lett. 1979, 20, 2339.
[23] C. Fehr, I. Farris, Angew. Chem., Int. Ed. 2006, 45, 6904.

Received December 19, 2007