A bottle of spearmint oil distilled from leaves, a bottle of caraway oil distilled from seeds. On the analysis sheet, the main constituent of both reads carvone, C₁₀H₁₄O, 150.22 g/mol. The two smells are still plainly different, and the reason lies in a detail of geometry that the nose can read.
Spearmint and caraway seed
Spearmint leaves (Mentha spicata, Mentha cardiaca) are steam distilled; a tonne of fresh leaf gives roughly three to five kilograms of oil. Most of what ends up in the bottle is carvone, 50 to 65 percent in Mentha spicata and about 61 to 72 percent in Mentha cardiaca. The next largest part is limonene, 11 to 21 percent. The smell opens sweet and green, closer to chewing gum than to the sharp cold of peppermint.
Caraway (Carum carvi) is the small, slightly curved seed of a biennial herb of northern Europe and western Asia. The oil distilled from the seed holds 47 to 60 percent carvone and 37 to 49 percent limonene. Its smell is the smell of rye bread, of sausages and sauerkraut in German, Austrian and Hungarian kitchens: warm, spicy, a little earthy.
Put two smelling strips side by side and nobody confuses one bottle with the other. Yet read only the ingredient list by name, and the two bottles are almost the same: a ten carbon ketone making up more than half, a ten carbon terpene making up most of the rest. The difference is not in the list of names. It is in the shape.
Dill (Anethum graveolens) also carries carvone, 27 to 53 percent in European dill seed oil. The Apiaceae, the family of dill, caraway, anise and coriander, is a family rich in carvone. Spearmint belongs to the Lamiaceae, a quite different family, yet it arrives at the same molecule by a biosynthetic route of its own.
Same formula, different shape
A carvone molecule is a six carbon ring carrying a ketone group, a methyl group and an isopropenyl branch. Ten carbons, fourteen hydrogens, one oxygen. The formula holds for spearmint carvone and for caraway carvone alike. Same mass, same boiling point, same infrared spectrum.
The difference sits at a single carbon on the ring, the one carrying the isopropenyl branch. That carbon is bonded to four different groups. The four groups can be arranged in space in two ways, and the two arrangements are mirror images of each other. No rotation brings one to lie exactly on the other.
Chemistry calls this property chirality, and calls each form an enantiomer, also known as an optical isomer. The form in spearmint is R-(-)-carvone. The form in caraway is S-(+)-carvone. The letters R and S give the absolute configuration of the chiral carbon; the plus and minus signs give the direction in which a sample turns the plane of polarised light.
In most ordinary physical measurements the two enantiomers cannot be told apart. The balance does not see it. The thermometer does not see it. A gas chromatograph with a standard column lets both forms out at the same moment. Only tools that are themselves chiral can separate them: polarised light, a chromatography column with a chiral stationary phase, or a protein receptor in the nose.
Left hand, right hand
Two hands are the most familiar image of chirality. Same five fingers, same order, same joints. A left glove still does not fit a right hand. A chiral molecule meeting a chiral environment behaves the same way: a protein is built from amino acids of one handedness only, so the two enantiomers do not fit it in the same way.
Plants do not make both hands in equal amounts either. Plant enzymes are chiral proteins, so biosynthesis usually leans strongly towards one form. Spearmint makes almost only R-carvone; caraway makes almost only S-carvone. A fifty fifty mixture of the two enantiomers, called a racemic mixture, does not rotate polarised light, and it is rarely found in genuine essential oils.
This has consequences in the laboratory. Carvone made by ordinary synthesis is usually racemic. A spearmint oil topped up with synthetic carvone will show an enantiomer ratio that drifts from the plant's typical ratio. Chiral gas chromatography and optical rotation are the two parameters used to catch that drift; the spearmint page in the library records this check.
The two hands are not quite equal on skin either. In a patch test at 5 percent on 541 dermatitis patients, 2.8 percent reacted to (-)-carvone, and the safety literature records this form as more skin reactive than the (+) form. Regulation reads both hands under one name: the IFRA Standard on carvone covers both CAS numbers, and the carvone entry in the EU list of fragrance allergens to be labelled names both the 5R and the 5S forms.
The two forms of limonene
Limonene is a ten carbon terpene with no oxygen, found in almost every citrus oil. It too has one chiral carbon and two hands. (+)-Limonene, that is R-limonene, is the form of orange and lemon peel: fresh orange, lightly sweet. In cold pressed lemon oil this form makes up 57 to 76 percent.
(-)-Limonene, that is S-limonene, smells lightly of pine with a faint mint, far less citrus. It appears as a minor constituent in some pine and mint oils. The Lê Mai glossary uses this very pair as its example under enantiomer.
The two hands of limonene also share one fate when they meet air. Left long in an opened bottle, limonene oxidises into limonene 1,2 oxide, carveol, hydroperoxides, and carvone as well. EU cosmetic labels name both forms under one shared name, Limonene, together with a requirement that the peroxide value stay below 20 mmol per litre.
One detail is worth pausing on. Spearmint and caraway both contain (+)-limonene, the same hand. What separates their two smells is carvone, and only the handedness of carvone. The two oils stand at either side of a mirror, with limonene standing still in the middle.
Shape or vibration: two hypotheses
A molecule drifting into the nose carries several properties at once: size, shape, functional groups, the way it vibrates. The question of which property decides smell received at least two major answers in the twentieth century.
The first answer rests on shape. In 1952 the British biochemist John Amoore proposed that smell follows molecular shape and size, fitting receptor sites like a key in a lock. In 1963, in a paper in Nature, he named seven primary odours. The number seven did not hold, but the receptor idea was partly borne out in 1991, when Linda Buck and Richard Axel described a large gene family encoding odorant receptors. Humans have about 400 working odorant receptors, and each odour is coded by a combination of them.
The second answer rests on vibration. Malcolm Dyson proposed in 1928 that the nose senses the vibrational frequency of molecules; Robert Wright extended the idea in 1954. In 1996 Luca Turin revived the hypothesis in the journal Chemical Senses, with a specific mechanism: the receptor reads molecular vibration through inelastic electron tunnelling.
The test for the vibration hypothesis is to replace hydrogen with deuterium: the shape stays nearly the same, the vibration changes. The results do not agree. In 2004 Keller and Vosshall reported in Nature Neuroscience that people could not tell ordinary acetophenone from deuterated acetophenone. In 2013 Gane and colleagues reported that people could tell a large ring musk from its fully deuterated form. In 2015 Block and colleagues reported in PNAS that the human musk receptor OR5AN1 responds similarly to the ordinary, deuterated and carbon 13 forms.
Carvone is a hard case for the vibration hypothesis. The two enantiomers have identical vibrational spectra, yet one smells of spearmint and the other of caraway. Supporters of the vibration hypothesis have proposed that the two hands may stay in a receptor for different lengths of time and so produce signals of different strength. That is an added explanation; it concedes that shape still takes part.
A nose that reads geometry
A molecule does not know what it smells of. It only has a shape, and that shape meets some four hundred kinds of receptor in the lining of the nose. The two carvone enantiomers activate combinations of receptors that are not quite the same, and the brain reads the two combinations as two names: spearmint, caraway.
The ability is not only human. Squirrel monkeys have also been found to tell the two carvone enantiomers apart. That suggests reading left and right hands is a feature of the olfactory system in general, not a special skill of a trained human nose.
A 2025 review in Chemical Senses concludes that no universal relation between odour quality and the physical or chemical parameters of molecules has been found. Carvone is one of the rare cases where the relation shows itself neatly: everything stays the same except one direction in space, and the smell changes completely.
For anyone working with scent the lesson is practical. A bottle that says carvone has not said enough; the question is which hand, and from which plant. Spearmint oil and caraway oil cannot stand in for each other even though they share the name of their main constituent. The essay Four Hundred Receptors and One Molecule goes further into how those receptors work.
The same ten carbons, the same oxygen.
A leaf and a seed stand on either side of the mirror.
The nose was the first instrument for geometry that people had.