Chem. Senses 24: 171-178,
1999
© Oxford University Press
A New Electronic–Topological Investigation of the Relationship between Chemical Structure and Ambergris Odour
Institute of Chemistry, Academy of Sciences, Kishinev, Republic of Moldova, MD 2028 1 Quest International, Ashford, Kent TN24 0LT, UK
Correspondence to be sent to: M.Yu Gorbachov, Institute of Chemistry, Academy of Sciences, Kishinev, Republic of Moldova, MD 2028
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionSummaryReferences |
|---|
An electronic–topological approach has been used to define an active ambergris fragment (AAF) which correctly describes the presence (or absence) of the ambergris odour of all 181 compounds investigated. The AAF consists of one oxygen atom and three carbon atoms (
,ß,
) which are separated by certain key distances and which possess certain
atomic
charges. The C atom must bear at least one hydrogen atom
(H) which is located at a certain distance from one of the unshared electronic
pairs of the oxygen
atom. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionSummaryReferences |
|---|
Since ancient times, ambergris has been one of the most highly valued perfumery materials. It is secreted in the stomach or intestinal tract of the sperm whale and released into the sea in the form of a grey to black stone-like mass. When exposed to sunlight, air and sea water, the material gradually fades to a light grey or creamy yellow colour and, at the same time, the main component, the odourless triterpene alcohol ambrein (1, Figure 1), is oxidatively degraded. Some of the products resulting from this chemical process are responsible for the organoleptic properties of ambergris. Since these (2–3) and the early synthetic ambergris odorants all contain a trans-decalin system, its presence was originally believed to be one of the prerequisites for an ambergris odour and most of the studies into the relationship between chemical structure and ambergris odour have been based solely on this group of compounds (Ohloff, 1971
;
Vlad et al., 1985a;
Winter, 1990;
Dimoglo et al., 1995).
|
However, a few cis-decalins smell of ambergris (e.g. compound 4, Figure 2) (Ohloff et al., 1990
),
and the number of ambergris odorants which do not even contain a
decalin system has rapidly expanded over the past 15 years. Examples include Timberol ( 5)
(Klein and Rojahn, 1979;
Brunke et al., 1980),
cyclohexyltetrahydrofuran (6) (Vladet al., 1985b),
4-isopropenyl-1,3,3,5,5-pentamethylcyclohexanol (7) (Hoffman and Pauluth,
1985),
Karanal (8) (Newman et al., 1988),
Ambercore (9) (Koshino et al., 1991)
and a mixture of 2-(2-bornyloxy)-ethyl-1-ethanols
(10) (Boden et al., 1994)
(Figure 2). Thus the aforementioned structure–odour
relationships for ambergris are
clearly incomplete.
|
Dimoglo et al. (1995
)
have attempted to use the rules which they
derived from decalin-type systems to explain the odour properties of some monocyclic ambergris
molecules, including 4-isopropenyl-1,3,3,5,5-pentamethylcyclohexanol ( 7) and
3-(2-ethyl-1,3,3-trimethylcyclohexyl)-2-methyl tetrahydrofuran (11, Figure
3b). In the
decalin-type compounds they found that many of the active materials shared two common
molecular fragments (Figure 3a). Fragment I includes an oxygen atom
and four carbon atoms
from the rings A and B. The distances between these carbon atoms (Ci, Cj, Ck and Cl) and the oxygen atom are ~5.15, 5.90, 4.90 and 4.10
Å respectively. Fragment II contains two methyl groups (Cm and Cn) which have the same stereochemical orientation, are attached to quaternary carbon atoms
and are separated by a distance of ~3.45 Å. The effective atomic charges of the oxygen
atom and carbon atoms i–n are also important, but for the sake of simplicity are not
discussed here. Dimoglo's application of these rules to the monocyclic compounds 7 and
11 is shown in Figure 3b. For some reason Dimoglo has only
considered the trans-isomer of 4-isopropenyl-1,3,3,5,5-pentamethylcyclohexanol (7) despite the findings of
Hoffman and Pauluth (1985)
that it is, in fact, the cis-isomer which possesses an ambergris note.
Our own interatomic measurements for trans-7 show that the O–Ci
distance is only 2.42 Å and thus much too short to meet the fragment I criterion of
4.40–5.90 Å. Our measurements for O–Cj, O–Ck, O–Cl and Cm–Cn are in close
agreement with those of
Dimoglo et al. (i.e. 5.74, 4.20, 3.81 and 3.21 Å
respectively).
[Note: interatomic measurements were performed using the Cerius2 version 3
software from Molecular Simulations Incorporated. Geometries were optimized using molecular
mechanics, forcefield cff95_950_1.01
(Dinur and Hagler, 1991).]
|
In one of the most recent investigations into the relationships between structure and ambergris odour, Bajgrowicz and Broger included six active non-decalin compounds in their training set of 23 compounds to generate possible ambergris osmophores (Bajgrowicz and Broger, 1995
).
These osmophores were then tested using a further 121 compounds (45 active, 76
inactive). From six possible osmophores, Bajgrowicz and Broger chose the one which they
thought was most plausible from a chemist's point of view. It consisted of a
hydrogen-bond acceptor, four hydrophobic regions and six excluded volumes [for a recent
review
of this work and of other structure–odour relationship studies of ambergris odorants, see
Rossiter (1996)]. However, unlike the decalin compounds, where, mainly as a result of the extensive studies by Ohloff and co-workers, there is a wealth of odour data for compounds of defined absolute configuration, the odour data for the non-decalin compounds is rather fuzzy and often refers to isomeric mixtures. For example, the individual odour characteristics of the two positional isomers of 2-(2-bornyloxy)-ethyl-1ethanol (10) have not been reported. The influence of chirality also needs further investigation. Thus it is clear that in order to increase our understanding of the structural requirements for an ambergris odour, more detailed organoleptic data is required.
In the Karanal series, the synthesis and organoleptic evaluation of a large number of
analogues enabled the following general requirements for a material in this family to have an
ambergris odour to be defined
(Sell, 1990).
The 5 position of the dioxane ring should be
substituted by a methyl group and a more bulky group (e.g. sec-butyl, n-propyl,
iso-propyl, tert-butyl). The cyclohexane ring should preferably have a double
bond in the 3,4 position, although this is not essential and should carry 1–3 methyl
groups, preferably 2. However, the location of these seems to be more important than the
number.
The structural requirements are further complicated by stereochemical considerations. Karanal (8) is a mixture of eight pairs of enantiomers. The diastereoisomeric pairs have been separated by efficient fractional distillation and it has been shown that only four are organoleptically active (three of strong intensity and one of weak). These findings agree with our gas chromatography olfactometry (GCO) evaluation of the isomeric mixture. GCO is a technique which, provided that the gas chromatography conditions adequately separate the components of a mixture, enables each component to be smelt in an olfactorily pure state at the exit port of a gas chromatography columnWe have been unable to resolve the individual enantiomers of each pair by any standard techniques, including chiral GCO. Chiral synthesis of all eight enantiomers of the four active diastereoisomeric pairs would be laborious and complicated by the fact that the chiral centre at position 2 of the dioxane ring is formed during acetalization and that the proton at C1 of the cyclohexene ring would be racemized during acetalization. Besides, the electronic–topological approach, which we have used to investigate the relationship between structure and ambergis odour, is unable to distinguish between the enantiomers of a racemic pair.
One of the inactive Karanal pairs crystallized on standing, thus enabling us to obtain an
X-ray crystal structure. This, in conjunction with NMR studies, has enabled us to define the
stereochemistry of the active and inactive pairs. These data have been used, along with other nontrans-decalin compounds and all of the compounds used in the investigations of
Vlad et al. (1985) and
Dimoglo et al. (1995),
to develop a structure–odour
relationship which correctly describes the presence (or absence) of the ambergris odour of all 181
compounds investigated.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionSummaryReferences |
|---|
The electronic–topological approach allows the investigation of both the electronic structure and spatial topology of a set of structurally diverse molecules. A detailed account of this approach may be found in the literature (Bersuker et al, 1987
).
In summary, the
following sequence of computational calculations are performed:
- Conformational analysis using the MM2 method
(Allinger and Burkert, 1982).
- Quantum-chemical calculation of electronic properties using the CNDO/2 method
(Segal, 1977).
- Formation of electronic–topological matrices of contiguity (ETMCs). Each
compound
is represented by an n2 quadratic matrix, n being the number of
atoms. Since this square matrix is symmetrical with respect to its diagonal elements, only the
upper half is usually shown. In this study, the diagonal elements are the effective atomic charges
and the non-diagonal ones are the optimized interatomic distances. Unlike
Dimoglo et al. (1995),
for the triand polycyclic trans-decalin compounds, we took into account
not only the distances between the oxygen atom of ring C and the carbon atoms of rings A and B
(Figure 3a), but also the distances between the carbon atoms of ring C
and those of rings A and
B.
- Computerized comparison of the ETMCs and activity feature selection using pattern
recognition theory
(Rosenblit and Golender, 1983)
of the atomic charges and interatomic
distances in N-dimensional space. The number N is equal to the total number
of the atomic charges and interatomic distances in the largest ETMC.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionSummaryReferences |
|---|
The electronic–topological investigation involved two stages. For the first step, a small restricted set of compounds was chosen: 4-isopropenyl-1,3,3,5,5-pentamethylcyclohexanol (7), the active Karanal stereoisomers, the four active molecules (12–15, Figure 4) and the inactive one (16, Figure 4). Compound 7 was chosen because the organoleptic properties of the two geometric isomers (Hoffman and Pauluth, 1985
)
are known and because, as a consequence of its symmetry and size, there are only a small number
of interatomic distances. Compound 11
(Figure 3b) was chosen because it has the shortest average distance
between its oxygen atom and
carbon atoms. The Karanal isomers were chosen because of our particular interest in this series
of compounds, they provide good examples of active and inactive non-trans-decalin
compounds and they have never been studied before in an ambergris odour–structure
relationship. Compounds 12–16 were included as representatives of the trans-decalin family.
|
A computerized comparison of the ETMCs of these 10 compounds revealed that all nine active materials contained the following structural fragment, hereafter referred to as the active ambergris fragment (AAF). The AAF contains one oxygen atom and three carbon atoms (
,
ß, ) which are separated by certain key distances and which possess certain atomic
charges. The C atom cannot be a quaternary carbon since it must bear at least
one hydrogen atom (hereafter referred to as H). This is nicely illustrated by the
comparison of the active and inactive compounds, 13 and 16 respectively
(note: for inactive
structures the letters , ß and are replaced by a, b and c respectively). The upper
limits of the distances of the AAF are provided by the active stereoisomers of Karanal and
compound 15.
The study was extended to include, in addition to the 10 compounds studied in step 1, the
four inactive Karanal diastereoisomers and all of the compounds used by
Vlad et al.(1985a)
and Dimoglo et al. (1995).
The chemical structures and associated odour data
are listed in these references. The extended data set of 181 compounds allowed the distances and
charge criteria for the AAF to be defined more precisely (Figure 5). The
distance between the H atom and one of the unshared electronic pairs of the
AAF's oxygen
atom (L), and the steric accessibility of H and of the oxygen
atom's active electronic pair (AEP) were also found to be important.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionSummaryReferences |
|---|
The AAF correctly describes the presence (or absence) of the ambergris odour for all of the 181 compounds investigated. Its use is best illustrated with a few examples. In these examples the numerical values adjacent to the
, ß and carbon atoms are the distances
between these carbon atoms and the AAF's oxygen atom. The latter, in the case of
polyoxygenated compounds, is marked by an asterisk. The distances between the carbon atoms
are listed below each structure and are labelled Lß, L and Lß for active compounds, and Lab, Lac
and Lbc for inactive compounds, where carbon atoms a, b and c are those which most
closely fit the requirements for the active carbon atoms , ß and respectively. The
distance (L) for active molecules is also provided where appropriate.
Let us first consider the transand cis-isomers of
1-[2,2,6trimethylcyclohexanyl]hexan-3-ol (5a and 5b respectively, Figure 6).
Ohloff (1984,
1985)
has demonstrated that it is only the trans-isomer which possesses an
ambergris–woody odour and that the cis-isomer has a much weaker, undefinable
floral character. NMR evidence suggests that the conformational arrangement of the
cis-isomer, at least at room temperature and in solution, is such that the hexan-3-ol side
chain is in the axial position and the methyl group at C6 is in the equatorial position
(Ohloff et al., 1985).
In this conformation (5b-I), the distance between the oxygen and Cb atom is
too long to satisfy the distance criterion of the AAF (4.8Å as opposed to
3.8–4.6Å). Molecular modelling has shown that the reverse
cis-conformational arrangement (5b-II) is also energetically favourable (23.1
kcal/mol compared with 23.5 kcal/mol for 5b-I), but again the AAF distance criteria are
not met, with the distances
between the oxygen atom and the carbon atoms b and c being too long. However, in the
trans-isomer (5a) the absence of steric hindrance between the axial methyl
group at C6 and the hydrogen atoms of the first carbon atom of the oxygencontaining side chain
allows this chain to rotate such that the distances between the oxygen atom and carbon atoms b
and c become shorter (4.6 and 5.1Å respectively) and the AAF distance criteria are
satisfied.
|
For the trans-isomer of 4-isopropenyl-1,3,3,5,5-pentamethylcyclohexanol (7b, Figure 6) one finds that the distance between carbon atoms a and b is too short to meet the distance requirements of the AAF (3.4 Å as opposed to 4.3–5.5 Å). This material does not have any ambergris character but is weak, woody and earthy. On the other hand, the cis-isomer with the axial hydroxyl group contains the AAF and possesses, according to Hoffman and Pauluth (1985
),
an ambergris, woody, patchouli odour. Some more examples of structurally related active and inactive compounds which illustrate the importance of the AAF's distances are provided in Figure 7. The distances which do not meet the AAF criteria are underlined. For example, in compound 24 the distance between atoms Ca and Cc is too short compared with that in compound 23 (4.2 versus 5.4 Å), and in compound 26 the distances between atoms Ca and Cb and between Ca and Cc are too long compared with those in compound 25; hence compounds 24 and 26, which have no ambergris character, do not contain the AAF.
|
The L
ß and L distances are also important in
the Karanal isomers (8a–8h, Figure 8). All four active
pairs of diastereoisomers (
8a–8d) have a cis-configuration about the dioxane ring, with the
cyclohexenyl
substituent equatorial and the sec-butyl group axial. In this configuration, the distances
between the ß and carbon atoms of the sec-butyl
group and the methylene group of the cyclohexenyl group (i.e. Lß and L respectively) satisfy the distance criteria of the
AAF. However, in the trans-configuration these distances (Lab and Lac) become too long because of the equatorial position of the sec-butyl group and,
as a consequence, the odourless isomers 8e–8h no longer possess the AAF. For
the
inactive isomers 8f and 8g, both oxygen atoms satisfy the requirements for the
AAF, which in
turn leads to two candidate alpha carbon atoms (a or a1). However, neither of these
fulfil the Lab or Lac criteria. In the case of the weak smelling ambergris
stereoisomer (8b), molecular modelling shows that there are two low energy conformers
which
result from rotation about the bond joining the two cyclic systems. The lowest energy one (29.33
kcal/mol ) is postulated to be inactive because the distance Lac is too
long, whereas the second one (8b, Figure 8, energy = 29.84
kcal/mol) meets all of the
AAF's requirements. Since the energy barrier to rotation is relatively high (7.6 kcal/mol),
the population of the active conformation (8b) will be low; this is consistent with the
observation that 8b only has a weak ambergris odour. It is worth pointing out here that,
for
flexible molecules, all possible low energy conformations need to be investigated to determine
whether or not the molecule possesses the active ambergris fragment. For example, in the case of
the active Karanal stereoisomer (8c), MM2 conformational analysis shows that rotation
about
the bond joining the two rings gives rise to three local energy minimum conformations of
energies 29.39, 29.86 and 30.43 kcal/mol. Two of these satisfy the AAF's requirements;
the one with an energy of 29.86 kcal/mol is shown in Figure 8. The
conformation which does not
possess the AAF is, in fact, the lowest energy one.
|
It is generally accepted that the conformation of a molecule responsible for triggering an odour response does not necessarily have to correspond to any of the theoretical minimum energy conformations. This phenomenon is also well accepted in drug–receptor interactions. Burgen et al. (1975
)
suggest that when a flexible drug molecule binds to a receptor, it is
probable that a nucleation complex is first formed by association of part of the drug with a subset
on the receptor. The half-life of dissociation at this subset is long enough to allow
conformational
rearrangement of the drug molecule favouring binding at subsites. Hence the need for thorough
conformational analysis in all structure–activity relationship (SAR) work.
The examples which we have discussed so far have illustrated how structural modifications
which alter the AAF's distances give rise to compounds which no longer possess an
ambergris odour. Let us now turn to the remaining requirements of the active ambergris
fragment: the atomic charges of the oxygen atom, C, Cß and C; the steric accessibility of the H atom; and the steric
accessibility of oxygen atoms' active electronic pair (AEP).
The active and inactive compounds, 27 and 28 respectively (Figure 9), highlight the
importance of the AAF's electronic characteristics. Although both compounds fulfil all of
the other requirements for the active ambergris fragment, only compound 27 has the correct
atomic charge on the C atom (0.0). In the inactive compound 28), the
charge
(Q) is -0.05 a.u., which just falls outside the required range of -0.04 to 0.16 a.u.
|
In compounds 29 and 30 (Figure 10), the H
atom and the AEP
respectively are sterically shielded by a methyl group; thus, although both of these compounds
meet the other requirements of the AAF, they do not possess a strong ambergris odour.
Compound 29 has a weak anisic and medicinal odour (Winter, 1989)
and compound 30, to those who can smell it, has a very weak ambergris odour
(Ohloff et al., 1980).
Twelve out of the 30 subjects who assessed compound 30 as a 1% ethanolic solution
from a
smelling strip could not detect an odour. All of the remaining subjects perceived it as weak and
ambergris,
except for one female who described it as strongly camphoraceous, slightly sweaty and
urine-like. Replacing the methyl group adjacent to the ether oxygen atom by a hydrogen reduces
the steric shielding of the oxygen atom's active electronic pair and, as a result, compound
31, which is a strong ambergris odorant, completely satisfies the AAF's
requirements.
|
|
Summary |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionSummaryReferences |
|---|
The electronic–topological approach has been used to define an AAF which correctly describes the presence (or absence) of the ambergris odour of all 181 compounds investigated. The AAF consists of one oxygen atom and three carbon atoms (
, ß, ) which are
separated by certain key distances and which possess certain atomic charges. The C atom must bear at least one hydrogen atom (H), which is located at a
certain distance from one of the unshared electronic pairs of the oxygen atom. This investigation into the relationship between structure and ambergris odour has used the largest data set of closely related active and inactive compounds to date, and has included several non-trans-decalin examples, such as the ambergris-smelling and odourless diastereoisomers of the aroma chemical registered under the tradename of Karanal. The development of structure–activity relationships is an iterative process. The AAF will be used to aid the design of novel ambergris molecules and the results from this work used to further refine the model if exceptions to the rules are found.
In all structure–odour relationship work, the greatest potential for error is
undoubtedly the organoleptic data. In order to investigate a large number of compounds we have
used data obtained both by ourselves and by other workers. Care must always be taken when
combining odour data from different sources. Discrepancies between reported odour
characteristics for the same compound are not uncommon. Factors contributing towards these
discrepancies include the absence of a universal odour language, the subjective nature of odour
perception, different methods of sensory assessment and the presence of trace impurities. Many
materials which smell of ambergris also have some woody or animal character and vice versa.
The presence of more than one olfactory note may stem from the presence of isomers
(Ohmoto, 1991);
for flexible
molecules, from the detection of different conformers; or simply just subjectivity
(Schulte-Elte et al., 1991).
Many ambergris odorants also provoke specific anosmia, that is, either the
inability to smell a particular odorant or a dramatically reduced sensitivity to that odorant. The
phenomenon of anosmia can, for example, lead to the same compound being described by one
person (who is prone to `ambergris' anosmia) as purely woody and by another
(who is prone to `woody' anosmia) as purely ambergris. This highlights the need
to use, wherever possible, an expert panel and a standard glossary of descriptors to minimize the
problems associated with subjectivity. Despite these drawbacks, the ambergris family is one of
the most suited odour types for structure–odour relationship studies because its odour is
one of the least difficult to define and many molecules belonging to this family are relatively
rigid. The next challenge is not only to be able to classify molecules into those which have an
ambergris odour and those which do not, but to be able to predict the intensity of the active ones
and to account for differences between the organoleptic properties of optical isomers.
Chemists are always tempted to interpret their observations on SARs in mechanistic terms
and to postulate ways by which biological activity could be brought about by some feature of
molecular structure or chemical activity that seems significant in the SAR. It is also tempting to
do the reverse, that is, to postulate a biological mechanism and then work backwards to structural
features that might satisfy it. Despite the significant advances made over the past decade in the
biochemistry of olfaction, our understanding of the mechanism of odour perception is still rather
limited
(reviewed in Sell, 1997).
However, SAR can be a useful tool in the design of new
odorants.
|
Acknowledgments |
|---|
We wish to thank Dr C.S. Sell for his comments on this manuscript. We are also indebted to Dr D.J. Watkin at the University of Oxford, UK, Mr S. Bociek of Quest and Drs D. Smith and C. Brown at the University of Kent, UK for their X-ray and NMR studies respectively of the Karanal isomers.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionSummaryReferences |
|---|
Allinger, N.L. and Burkert, U. (1982) Molecular Mechanics. ACS Monograph 177. American Chemical Society, Washington, DC.
Bajgrowicz, J.A. and Broger, C. (1995) Molecular modelling in design of new odorants; scope and limitations. . In Baser, K.H. C. (ed.), Flavors, Fragrances and Essential Oils. Proceedings of the 13th International Congress of Flavors, Fragrances and Essential Oils, Istanbul, Turkey, 15–19 October 1995; AREP Publishers, Istanbul, Vol., 3, pp. 1–15.
Bersuker, I. B., Dimoglo, A.S. and Gorbachov, M.Yu. (1987) Electron– topological approach in studying structure–activity relationships. Inhibition of thymidine phosphorylase by uracil derivatives. Bioorg. Chem. , 13 , 38 –44.
Boden, R.M., Schreiber, W.L., McGhie, J.A. and Geiger, K.A. (1994) United States Patent 5366959.
Brunke, E.J., Rojahn, W. and Warnecke, H.U. (1980) Chemical structure and aroma effect in the tetrahydroionol series. Dragoco Rep. , 27 , 199 – 200, 202–203.
Burgen, A.S.V., Roberts, G.C.K. and Feeny, J. (1975) Binding of flexible ligands to macromolecules. Nature , 253 , 753 –755.[Medline]
Dimoglo, A.S., Vlad, P.F., Shvets, N.M., Coltsa, M.N., Güzel, Y., Saraçoglu, M., Saripinar, E. and Patat, S. (1995) Electronic–topological investigations of the relationship between chemical structure and ambergris odour. New. J. Chem. , 19 , 1 –10.
Dinur, U. and Hagler, A.T. (1991) New approaches to empirical force fields. Rev. Comput. Chem. , 2 , 99 –164.
Hoffman, H.M.R. and Pauluth, D. (1985) Terpenes and terpenoid compounds, 14. Synthesis of oxygenated derivatives in the tetramethyllimonene series. New odorants of the wood–ambergris type. Liebigs Ann. Chem. , 2 , 396 –402.
Klein, E. and Rojahn, W. (1979) German Patent DE 2807584. Chem. Abstr. (1979), , 92 , 11114.
Koshino, J., Fujikura, Y., Utsunomiya-shi, T., Toi, N., Sakura-shi, C., Yuki, R. and Miyabe, H. (1991) European Patent 0467290A2.
Newman, C.P., Rossiter, K.J. and Sell C.S. (1988) European Patent 0276998. Chem. Abstr. (1988) , 110 , 13401.
Ohloff, G. (1971) Relationship between odor sensation and stereochemistry of decalin ring compounds. In Ohloff, G. and Thomas, A.F. (eds), Gustation and Olfaction. Academic Press, London, New York, pp. 178–183.
Ohloff, G. (1984) European Patent 118809. Chem. Abstr. (1984) , 102 , 79179.
Ohloff, G., Vial, C., Wolf, H. R., Job, K., Jégou, E., Polonsky, J. andLederer, E. (1980) Stereochemistry–odor relationships in enantiomeric ambergris fragrances. Helv. Chim. Acta , 63 , 1932 –1946.
Ohloff, G., Schulte-Elte, K.H., Wolfgang, G., Winter, B. and Pamingle, H. (1985) Diastereoselective odor perception of alcohols in the ionone series. Helv. Chim. Acta , 68 , 1961 –1985.
Ohloff, G., Escher, S., Giersch, W., Niclass, Y. and Bernardinelli, G. (1990) Configuration–odor relationships in 5ß-Ambrox. Helv. Chim. Acta , 73 , 1935 –1947.
Ohmoto, T. (1991) European Patent 456932. Chem. Abstr. (1991), , 116, 128234.
Rosenblit, A.B. and Golender, V.E. (1983) Logical–combinatorial methods in drug design. Chem. Abstr. (1983) , 101 , 144155f.
Rossiter, K.J. (1996) Structure–odor relationships. Chem. Rev. , 96 , 3201 –3240.[Web of Science][Medline]
Segal, G.A. (ed.) (1977) Semiempirical Methods of Electronic Structure Calculation, Part A: Techniques. Plenum Press, New York.
Sell, C.S. (1990) The chemistry of ambergris. Chem. Ind. , 16 , 516 –520.
Sell, C.S. (1997) On the right scent. Chem. Britain , 33 , 39 –42.
Schulte-Elte, K.H., Margot, C., Chapuis, C., Simmons, D.P. andReichlin, D. (1991) European Patent 457022. Chem. Abstr. (1991) , 117 , 8242.
Vlad, P.F., Bersuker, I.B., Dimoglo, A.S., Gorbachov, M.Yu. andKoltsa, M.N. (1985a) Structural and electronic origin of ambergris odor of cyclic compounds. New. J.Chem. , 9 , 211 –218.
Vlad, P.F., Morary, A.F. and Koltsa, M.N. (1985b) Russian Patent SU 1169971. Chem. Abstr. (1986) , 104 , 95284.
Winter, B. (1989) Ring-opened analogues of Ambrox®: synthesis and structure–odour relationships. Helv. Chim. Acta , 72 , 1278 –1283.
Winter, B. (1990) Studies related to ambergris-type odorants. Pure Appl. Chem. , 62 , 1377 –1380.
Accepted December 9, 1998
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  F. Yoshii, Y. Yamada, T. Hoshi, and H. Hagiwara The Creation of a Database of Odorous Compounds Focused on Molecular Rigidity and Analysis of the Molecular Features of the Compounds in the Database Chem Senses, June 1, 2002; 27(5): 399 - 405. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||