Short answer
To rebuild a scent from headspace data, treat the analysis as a map, not a recipe. A headspace sample over-represents the most volatile molecules and under-represents heavy ones, and its biggest peaks are not necessarily what you smell, so the percentages in the report cannot be copied into a formula. Group the identified compounds into available materials, weight them by odor impact and volatility rather than by peak area, replace anything you cannot buy or use, and then correct the reconstruction by smell against the living reference over several versions — checking every material against IFRA and allergen rules as you would any formula.
What headspace is, for readers meeting the term for the first time, is explained in the RUŌOD Journal's what is headspace in perfume. This guide is for the perfumer holding the data.
Why sample the air
- Some flowers give no extract. Lilac, lily of the valley, freesia, gardenia and many others yield no usable oil or absolute, or one that smells nothing like the flower.
- Extraction changes the smell. An absolute is what a solvent pulls from cut flowers; a living flower on the plant smells different, and its scent changes over the day.
- It is non-destructive. A rare flower, a fruit at the market, a forest after rain, can be sampled without harvesting anything.
How the data is produced
- Enclose. A glass or inert film chamber is placed around the flower or object.
- Trap. Air is drawn from the chamber through a tube of adsorbent, or a coated fiber (solid-phase microextraction, SPME) is exposed to the air, for a set time.
- Release and separate. In the lab the trapped molecules are released (by heat or solvent) into a gas chromatograph, which separates them.
- Identify. A mass spectrometer identifies each peak by comparison with spectral libraries and reference standards; trained evaluators may also smell the effluent as it leaves the column (GC-olfactometry), to find which peaks matter to the nose.
- Rebuild. A perfumer composes a reconstruction from the list, adjusting by smell against the original or a memory of it.
The analytical side is introduced in GC-MS for perfumers.
Why the data is not a formula
- Air is not liquid. A headspace sample over-represents volatile molecules and under-represents heavy ones, compared with what a perfume needs in the bottle. A material that is 40% of the air may need to be a small part of the reconstruction, and a heavy one much more.
- Peaks are not odor. A minor peak with a very low odor threshold can define the scent; a large peak may contribute little (odor threshold).
- Not every molecule is available as a commercial aroma chemical, and some are not permitted.
- Living scents change. A morning and an evening sample of the same flower can differ.
So a headspace is a map, not a recipe. The same reasoning applies to reconstructing an essential oil from its published composition (essential oil reconstitution).
What it made possible
Headspace work since the 1970s and 1980s gave perfumery more realistic versions of flowers that could not be extracted, and the habit of reproducing scents of places and objects. The classic results are the muguet and lilac accords (lily of the valley accord, lilac accord), and environmental effects such as rain (the rain accord).
For a small lab
Headspace equipment and analysis are specialist: GC-MS labs and fragrance houses offer it as a service. A small lab can still use the method's logic:
- Read published headspace analyses in the scientific literature for the flower you want, as a starting list.
- Smell the real thing often, at different times, and write precise descriptions (how to describe a scent).
- Build the reconstruction in versions, comparing each against the reference, and record what changed (version control).
- Check every material against IFRA and allergen rules before calling the accord finished.
Records
Keep the source data with the accord: the analysis report or the publication, the date and conditions of sampling, and the decisions you made in turning it into a formula. A reconstruction is a formula like any other, with a code and versions; see the guide to perfume materials by odor family for where such accords sit in a palette.
Frequently asked questions
Is a headspace reconstruction natural?
No. It is built mostly from aroma chemicals, even when it reproduces a natural scent. It can include naturals, but the result is not a natural extract.
Can headspace copy a perfume?
Headspace can sample a perfume's air, but competitive analysis of a perfume is usually done by direct GC-MS of the liquid. Either way the data needs interpretation and does not hand over a formula (fragrance matching).
What is SPME?
Solid-phase microextraction: a fiber coated with an adsorbent is exposed to the air or liquid, collects volatile molecules, and is then inserted into the GC for analysis.
Written and reviewed by the RUŌOD Lab team. This article is general education about perfume formulation and record-keeping; it is not legal, regulatory or safety advice, and the examples are illustrations, not validated commercial formulas. How we write and check these guides.