Odor Activity Values: From Concentration to Impact

Before you start

This lesson builds on partial pressures, the vocabulary of odor intensity and odor value, and the logic of GC-olfactometry, where dilution-to-threshold methods come from.

From partial pressure to concentration

For an ideal gas, moles per volume are p/RT, so mass per volume is

C = p · M / (R · T)

With p in pascals, M in g/mol, R = 8.314 J/(mol·K) and T in kelvin, C comes out in g/m³. Continuing the worked blend from the Raoult's law lesson: material A at 0.0102 kPa = 10.2 Pa, M = 136 g/mol, 25 °C:

C = 10.2 × 136 / (8.314 × 298.15) = 0.559 g/m³ = 559 mg/m³.

The odor activity value

OAV = C / OT, where OT is the odor threshold in air, in the same units. If A's threshold were 0.01 mg/m³ (an invented figure for the arithmetic), its headspace OAV would be about 56,000. That is not an intensity: it says only that A is very far above the level at which it can be detected. Ranking materials by OAV is the classic way to shortlist key odorants, and in food chemistry it is paired with aroma recombination and omission, as Grosch's review describes.

Two caveats are built in. First, OAVs for the same compound can differ by orders of magnitude depending on which published threshold you choose, because thresholds depend on method, panel and medium. Report the threshold's source with every OAV, and prefer thresholds measured in air for a headspace calculation. Second, an OAV is calculated per component; a perfume is not smelled one component at a time.

Intensity is not proportional to concentration

Above threshold, perceived intensity ψ grows approximately as a power of concentration, after Stevens:

ψ = k · Cⁿ

For odor, n is typically well below 1. With n = 0.5, doubling C raises ψ by 2^0.5 ≈ 1.41, and a tenfold increase by about 3.16; with n = 0.3, doubling gives only about 1.23. This is why overdosing a material yields diminishing returns in loudness while changing balance, and why an OAV of 56,000 against one of 5,600 does not mean "ten times stronger".

Where it misleads

• Mixture suppression. In mixtures, components commonly suppress each other's perceived intensity, and humans can identify only a few components of a complex mixture (Laing and Francis). A high-OAV material can be masked; a low-OAV one can carry a character.
• Thresholds are not exponents. Two materials with equal OAV can have different power-law exponents, so they grow differently with dose.
• Headspace is not what reaches the nose. Evaporation and air dilution between surface and nose lower every concentration, by factors you rarely know.

In professional practice

OAV reasoning is used to prioritize: which trace components of a natural might explain its character, which impurity might explain an off-note, which materials in a reconstruction deserve the most precision. It is a hypothesis generator. The decision about whether a component matters is made by smelling with and without it.

Open questions

Whether a mixture's perceived intensity can be predicted from its components' intensities and exponents is still argued over; several combination rules have been proposed, and none holds generally. Perfumes, with dozens of components far above threshold, are a hard test case that the literature has barely touched.

Common mistakes

• Mixing threshold media. A threshold in water is not a threshold in air.
• Treating OAV as loudness. It ranks potential; Stevens' exponent governs loudness.
• Omitting without replacing the weight. Then every other material's share changes too, and you are testing two things.

Safety

Omission tests are smelled on strips, never on skin. Read the SDS for every material and follow the safe handling lesson; volunteers are briefed as in the ethics lesson.

Key terms

Air concentration

Mass of a substance per volume of air, here in mg per cubic meter. From a partial pressure p, it is pM/RT.

Omission test

A sensory test that removes one component from a reconstituted mixture and asks whether panelists can detect the difference. It tests importance directly, where an OAV only predicts it.

Power-law exponent

The exponent n in Stevens' law, ψ = kC^n. For odor it is usually well below 1, so intensity grows much more slowly than concentration.