Ionone—À Rebours

This is not intended to be read, its intended to be skimmed, because frankly its kind of fucking silly. Like the last related post, Rum Oil Introduced In Full Stereo, we got a glimpse of rum oil components through the mind of perfumer Arcadi Boix Camps. Here we are going to do it again by examining more of the perfume trade. The point is to show you how important these compounds are and how people have chased them to the ends of the earth (and beyond). When rum lost rum oil post Arroyo, in perfume, an entire world formed around it.

There is a high degree of obsession on display here as well as a pursuit of nuance that just doesn’t seem to be found in the distilling trade. Again, chemistry is going to be presented, but the point is to be mystified. It is not even normal chemistry to the average chemist. These people have massive dreams and they are wading into the deepest end of multiple disciplines to realize them.

With distilling, we likely will not get beyond the rum oil components of ionone, damascenone, and damascone. In perfume, their dreams are so large they go—À Rebours.

This wonderful art had held him entranced for a long while, but now he was dreaming of another experiment.

He wished to go one step beyond. Instead of artificial flowers imitating real flowers, natural flowers should mimic the artificial ones. -Joris Karl Huysmans, À Rebours (1884)

This all begins in a chapter of Advances in Flavours and Fragrances: From the Sensation To the Synthesis called Designing Damascone and Ionone-like Odorants by Philip Kraft of Givaudan. My commentary follows in bold.

When it comes to smelling ‘sweet’ or ‘powdery’, ionones are the perfumery materials of choice. Discovered as early as 1893 by Tiemann and Krüger, the ionones became central to perfumery and fragrance chemistry, and thus inspired the synthesis of numerous analoga and derivatives. In a broader sense, β-damascenone and other damascones, Iso E Super®, Koavone®, Timberol® and also Georgywood® can be considered ionone analoga, too. Figure 1 places these odorants in a common timeline.

What we know so far is that ionone is a very key odorant in Bourbon while damascenone (also in Bourbon) is a key odorant in rum, and in the best can be found at extraordinary levels.

Likely they chased natural versions before they started synthesizing nature equivalent versions and beyond. Lets pause this chapter, visiting elsewhere in the text, and draw a deeper connection to rum.

The perfume industry is advanced enough to differentiate between natural versions of these products and synthesized versions using stable isotope methods. There is enough demand for authenticity that methods exist and this may help regulators if these compounds start being chased by rum connoisseurs. If we suspected someone faked a rum, we could know.


Stable isotope methods have gained increasing importance in the authenticity assessment of flavour compounds. Measurement of carbon and hydrogen stable isotope ratios, which required multi-milligram quantities of pure compounds when using “off-line” methods, became more straightforward and convenient with the introduction of commercial “on line” GC-IRMS instruments in 1990. The latest GC-“P”-IRMS instrumentation involves a pyrolysis interface (“P”), which allows one to measure oxygen isotope ratios. Over the 10 past years, a significant amount of data have been published on the differentiation of flavour and essential oil components, from (or versus) their synthetic (or so-called “nature identical”) counterparts. Therefore, these data are considered as an authentication tool only. In this respect, a further example of this possibility is provided by the case of (E)-β-damascenone, a ubiquitous key flavour contributor in natural flavours. When isolated from rum (made from cane sugar), damascenone shows a C13 enrichment that is typical of a C4 plant, whereas the same material obtained by a biotechnological process from a hydroxylated carotenoid precursor present in a C3 plant would display a C13/C12 ratio rather similar to that of synthetic damascenone (Table 1).

Oh wow, not too sure what to make of that, but we see a demand for authenticity and we even see rum presented as an example. Does this mean rum is the largest source of natural damascenone? Are they harvesting it all from Hampden Estates brokered by E&A Scheer?

Of all the rums of the world, who would we be looking at in these unintelligible tables? A Batavia Arrack? A Hampden? Is Barbados or Guyana in on this? Is there a machine somewhere (besides the birectifier), that harvests rum oil from rum? Who handles all this shit? Are there a lot of Ferragamo scarves involved?

I feel this information is key because we are going to start defining rum authenticity by the act of the seeking something. Rum making is seeking and these compounds are the target. If you are not seeking, even though you have the same ingredients, you are not making rum. End of story. Have dreams or get out of the way! These compounds are so high value and tricky to work with that there may not be room to cheat, but in any case, the perfumers have helped blaze a trail to authenticity.

The ionone analoga in Fig. 1 cover a broad range of odour notes from floral (damascenone) via animalic (Timberol) to woody characteristics (Georgywood). In the syntheses of these odorants, ionones served as structural or chemical templates, and thus principles of rational design are well illustrated in this family.


Acyclic compounds of the pseudoionone series were long known to possess no distinct odour characteristics. Therefore, Jitkow and Bogert first varied the substituents on the cyclohexenyl ring; and they stated in 1941 that the typical violet odour of β-ionone does critically depend on a trimethyl-substituted cyclohexenyl ring. Of these three methyl groups, two have to be adjacent to the 3-oxo-1-butenyl chain, so in position 1 and 5 according to the trivial numbering in Fig. 2. Accordingly, two bulky hydrophobic groups flanking the 3-oxo-1-butenyl substituent seemed necessary. But then again, was a cyclohexenyl ring really crucial?

In the 1940’s, they were speculating on the intersection of chemistry and perception.

In 1962, Sestanj imaginarily cut out ring atom C-2 and C-3 of β-ionone, and proposed 5-isopropyl-6-methyl-3,5-heptadien-2-one as an interesting target molecule that could retain the odour characteristics of the parent β-ionone. He synthesised this 1(2),3(4) diseco-β-ionone starting from 3-isopropyl-4-methyl-1-pentyn-3-ol, and indeed found the compound to possess an odour similar to that of β-ionone. The physical data were also reminiscent to those of β-ionone; however, as one could expect the boiling point of 104-108°C/10 mm was significantly lower: The derived seco-ionone was more diffusive, while close in odour to its parent compound.

They were dreaming way out. We also we a new concept emerge. The slightly lower boiling point makes the compound relatively more diffusive.

Does figure 2 show those dreams taking 21 years to realize?

The idea of Sestanj was taken up in the early 1980s by chemists of IFF, who chose Isoraldeine® (8-methylionone) as parent structure. Employing ethyl methyl ketone instead of acetone in the classical ionone synthesis, 8-methylionone had also been discovered by Tiemann in 1893. It is more intense than the genuine ionones, possesses a very fine odour reminiscent of violets with a slightly woody-vetiver tonality, and became therefore the most popular violet odorant in perfumery.

Dissecting the bonds between C-2(C-3) and C-4(C-5), transposing the C-5 methyl group to C-7, and hydrogenating the endocyclic double bond, leads from Isoraldeine® to Koavone®, which was still quite close in odour to Isoraldeine® and 10-methylionone. This odour similarity was also rationalized by a superposition analysis on computer; one of the first applications of molecular modelling in the flavour and fragrance industry.

These categories of odorants are apparently so important they are going to incredible lengths.


Omitting ‘unnecessary atoms‘, while retaining all structural features that code a given odour note, furnishes more volatile and more flexible molecules. These can provide more diffusivity and impact in a perfume. Often however, more substantivity is desired, and this on the contrary requires higher molecular weight, besides lower thresholds that make up for the lower vapour pressure.

Wow, there are some amazing concepts there and somehow the feature extraction makes me think of the super normal stimuli concept that has made many appearances on the blog. I was even so bold as to claim that all creative linkage in the culinary arts is an attempted to create a supernormal stimuli. They are chasing the Venus of Willendorf.

Introducing additional rings in a molecule is an interesting strategy to devise more substantive odorants. In addition to increasing the molecular weight, a ring system can rigidify the molecule; thus allowing better insight into the receptor geometry. This insight can then inspire further structure modifications, and thereby guide the synthesis of more potent odorants.

More insight into their ability to connect complex molecule shapes to perception. Amazing.

This polished visualization downplays that they are likely smoking raw opium and reading Italo Calvino’s Invisible Cities to dream this stuff up. They are in the business of obsession.

Fig. 3 presents the initial thoughts and molecular-modeling calculations that preceded the synthesis of 1-{bicyclo[6.4.0]dodec-1(8)-en-10-yl}ethan-1-one. The steric bulk of the gem-dimethyl and the 8-methyl group of Isoraldeine® was mimicked by fusing a six membered ring. The remaining 5-methyl group was then found to be best superimposed by a cyclooctene ring. This simplified the synthetic access by Diels-Alder reaction of but-3-en-2-one with bis(methylene)cyclooctane, speculating that the double-bonds were more of conformational than electronical importance. Indeed, the designed target compound showed the typical fruity-woody, violet odour of β-ionone.

These molecules seem like Huysmans’ monsterous flowers:

The gardeners brought still other varieties which had the appearance of artificial skin ridged with false veins, and most of them looked as though consumed by syphilis and leprosy, for they exhibited livid surfaces of flesh veined with scarlet rash and damasked with eruptions. Some had the deep red hue of scars that have just closed or the dark tint of incipient scabs. Others were marked with matter raised by scaldings. There were forms which exhibited shaggy skins hollowed by ulcers and relieved by cankers. And a few appeared embossed with wounds, covered with black mercurial hog lard, with green unguents of belladonna smeared with grains of dust and the yellow micas of iodoforme.

And/or a bit porn star like:

“Sapristi!” he exclaimed enthusiastically.

A new plant, modelled like the Caladiums, the Alocasia Metallica, excited him even more. It was coated with a layer of bronze green on which glanced silver reflections. It was the masterpiece of artificiality. It could be called a piece of stove pipe, cut by a chimney-maker into the form of a pike head.

1-{Bicyclo[6.4.0]dodec-1(8)-en-10-yl}ethan-1-one was, however, not only interesting for giving an idea about the molecular dimensions of the ionone receptor, but also as a structural link to the woody family, e.g. Iso E Super®. An interesting new reaction led regio- and distareoselectively to a potent woody-ambery odorant, reminiscent of Georgywood® and Iso E Super® (Fig. 4). In the presence of Wilkinson’s catalyst, 4-methylenespiro[2.7]decane reacted with 3-methylbut-3-en-2-one to provide (r-9,0-10)-1-{9,10-dimethylbicyclo[6.4.0]dodec-1(8)-en-10-yl)ethan-1-one, which possesses an intense woody-ambery odour with an excellent threshold of 0.01 ng/L air. By introduction of just two methyl groups one can switch the odour note from typical violet-like to woody ambery. Further derivatives provided additional insight into the structure-odour correlation of these two odour notes.

They brought a new batch of monstrosities from the wagon: Echinopses, issuing from padded compresses with rose-colored flowers that looked like the pitiful stumps; gaping Nidularia revealing skinless foundations in steel plates; Tillandsia Lindeni, the color of wine must, with jagged scrapers; Cypripedia, with complicated contours, a crazy piece of work seemingly designed by a crazy inventor. They looked like sabots or like a lady’s work-table on which lies a human tongue with taut filaments, such as one sees designed on the illustrated pages of works treating of the diseases of the throat and mouth; two little side-pieces, of a red jujube color, which appeared to have been borrowed from a child’s toy mill completed this singular collection of a tongue’s underside with the color of slate and wine lees, and of a glossy pocket from whose lining oozed a viscous glue.

Do not worry about not understanding any of this. We are only walking through someone else’s world and admiring their quest and obsessions.


Iso E Super®, as a better quality of Isocyclemone E® was called later by IFF, was for instance used in 18% in »Trésor« (Lancôme, 1990). in 25% in »Fahrenheit« (Dior, 1988), in 35% in Déclarations (Cartier, 1998), and even in 43% in »Feminité du Bois« (Shiseido, 1992). It is synthesized by Diels-Alder reaction of myrcene with 3-methylpent 3-en-2-one and subsequent acid-catalyzed cyclization as outlined in Fig. 5.

I don’t understand the nomenclature why which they arrive at a percentage.

However, GC-olfactometry at Givaudan showed that not the main product of Fig. 5, but a ca. 5% constituent determines the characteristic woody-ambery odour of the commercial material. While the threshold of the main compound of Iso E Super® is about 500 ng/L, this minor constituent possesses an odour threshold of only 0.005 ng/L; and thus, became known as “Iso E Super Plus” at Givaudan internally.

I think I interpret this that some of these trade marked products are a collection molecules and not simply an individual alternative to ionone or damascenone.

By reaction of commercial Iso E Super® with peracetic acid and repeated column chromatography, this powerful minor constituent could be isolated, and INADEQUATE as well as NO DIFF experiments led to the assignment of the structure depicted in Fig. 6. This structure was proved by a directed synthesis starting from α-ionone. Its formation in the [pages not available].

So they are going to great lengths. Next we get a glimpse of work on damascone.

[…] synthesis of (S)-(-)-α-damascone, which could be carried out on industrial scale. That this elegant process, which utilises (1R,2S)-(-)-N-isopropylephedrine as a chiral auxiliary in an enantioselective protonation, still has not been employed technically, might be due to the fact that both antipodes smell relatively similar. The (R)-(+)-isomer is certainly more apple-like, and also possesses an unpleasant cork-stopper nuance, but both enantiomers share the same main odour characteristics, and their thresholds differ ‘just‘ by a factor of 70.

It should therefore still be possible to find more powerful damascone odorants, for which one would expect a higher chiral discrimination on the receptor. If a chiral odorant is ‘perfectly‘ complementary to the chiral proteinogenic receptor site, its enantiomer should be odorless, or weaker by at least a factor of 1000.

Who knows what much of this means, but is nature maxed out? Or, are some of these hypothesized variations discoverable? Any probability we could find them in wine or spirits?

In 1978, even before the introduction of Koavone (Fig. 2), 2-methyl-3-isopropylhepta 2,5-dien-4-one, a 2(3),4(5)-diseco-β-damascone, had already been described in a patent of Takasago. Grignard reaction of diisopropyl ketone with μ-(3-butyn-2-olato)bis(magnesium bromide) furnished 5-isopropyl-6-methylhept-3-yne-2,5-diol, which was transformed into the target structure by means of a Rupe rearrangement (Fig. 10). This seco-damascone possesses fruity, damascone- and tobacco-like odour characteristics, which are indeed quite close to the parent β-damascone. However, 2-methyl-3-isopropylhepta-2,5 dien-4-one was never introduced into perfumery.

It was probably like that Wu-Tang album that the Fed’s confiscated from Martin Shkreli.


We found this interesting seco-damascone by GC-olfactometry as a ca. 1.1% trace constituent of a very complex, crude reaction product. At first, no reasonable structure could be proposed for this trace constituent on the basis of the chemistry involved or the GC/MS data. Thus, a ca. 370 μg sample was isolated by preparative GC, but HMBC and NOESY NMR spectra did not allow the unambiguous assignment of the structure. While there was no doubt about a hepta-2,5-dien-4-one skeleton with one tetrasubstituted double bond bearing a methyl and an isopropyl group, the relative positions of the latter remained unclear. Besides 2-methyl-3-isopropylhepta-2,5-dien-4-one, 5,6,7-trimethylocta-2,5-dien 4-one was therefore also a possible structure for this trace constituent. When we synthesised both compounds, it turned out that the trace constituent of the reaction mixture was indeed the 2-methyl-3-isopropylhepta-2,5-dien-4-one that had been patented by Takasago (Fig. 10). Yet, when we finally had also the 5,6,7-trimethylocta-2,5-dien-4-one in hand, we found that this also possessed a damascone odour, and in fact was much stronger and by far superior to that of the trace constituent.

Figure 11 details our synthesis of 5,6,7-trimethylocta-2,5-dien-4-one, which commenced with the preparation of 1-ethoxypropene from 2-chloro-1,1-diethoxyethane. After treatment with sodium amide in liquid ammonia, the resulting sodium acetylide was alkylated with iodomethane. Addition of 1-ethoxypropene to methyl isopropyl ketone in the presence of boron trifluoride diethyl etherate furnished in 59% yield the ethyl 2,3,4 trimethylpent-2-enoate. This was converted in the next step to the target compound by a Grignard reaction with prop-1-en-1-ylmagnesium bromide and in-situ trapping of the resulting dienone as lithium enolate. (2E,5E/Z)-5,6,7-Trimethylocta-2,5-dien-4-one possesses a very powerful and diffusive, fruity-rosy, β-damascone-like odour, reminiscent of apples, plums, raisins, and other dried fruits. Contrary to what one would expect with the previous seco-structures in mind, these interesting odour characteristics are however mainly due to the (2E,5Z)-configurated cis-isomer, while the (2E5E)-configurated trans isomer is almost not detectable for some people.

We have to assume, there are people somewhere that understand this.

This figure is interesting because we can see the degrees of nuance they are messing with.

Therefore, 5,6,7-trimethylocta-2,5-dien-4-one can not be considered a seco-damascone, anymore, and it thus was very exciting to explore the structure odour correlation of related oct-2-en-4-ones. In Figure 12, four analoga are shown, together with their respective odour descriptions. Even the partially hydrogenated derivative still possesses a fruity-rosy odour, reminiscent of raspberries and dried fruits. In the dienone analogs, the strength of the (5E)-isomers increases when the C6-substituent becomes more bulky, and finally the (5E)-isomers become even stronger than the (5Z)-isomers.

However, (2E,5Z)-5,6,7-trimethylocta-2,5-dien-4-one, discovered by serendipity, still constitutes the most powerful and most pleasant damascone-like odorant of this series. As with the powerful impurity in Iso E Super®, serendipity therefore continues to play an essential part in fragrance chemistry.

I kept going because I loved how they ended on a note of serendipity.

“These plants are amazing,” he reflected. Then he drew back to let his eye encompass the whole collection at a glance. His purpose was achieved. Not one single specimen seemed real; the cloth, paper, porcelain and metal seemed to have been loaned by man to nature to enable her to create her monstrosities. When unable to imitate man’s handiwork, nature had been reduced to copying the inner membranes of animals, to borrowing the vivid tints of their rotting flesh, their magnificent corruptions.

1 thought on “Ionone—À Rebours

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

search previous next tag category expand menu location phone mail time cart zoom edit close