In short
Reconstruction is most valuable as research: each version of a rebuilt natural or a classic structure is a hypothesis about what carries its smell, and omission and addition tests tell you which hypotheses hold. Built from a published analysis and tested blind against the real material, it teaches more about a material than years of smelling it alone. The line that matters is between study and copying: reconstructing a natural is ordinary research; reproducing a commercial perfume for sale is a different act.
What you will learn
- Frame a reconstruction as a set of hypotheses about what carries a smell
- Build a first model from published compositional data and odor activity reasoning
- Test it by omission and addition, and record what each test showed
- Keep the ethical and legal line between study and copying
Before this lesson: Originality, Signature and the Luxury Brief, Studying a Fragrance Without Copying It, Analytical Reconstruction: Study Guide, Specialist Project: An Analytical Dossier
Before you start
This builds on studying a fragrance without copying it, reconstruction and its limits and the Analytical track's analytical dossier.
The method
- Start from data. A published GC-MS analysis of the material, with its source, origin and method. Area percent is a starting proportion, not an odor recipe.
- First model. The major components at their reported proportions, with every component you can buy.
- Hypotheses. For each gap between the model and the real material, write what you think is missing and why: a trace compound with a low threshold, an isomer, an unidentified peak.
- Test by omission and addition. Remove a component, or add a candidate, one change per version, compared blind with the previous version and the real material.
- Record. The hypothesis log is the product. The final reconstruction is a by-product.
This is the recombination-and-omission logic Grosch describes for food aromas, applied to a perfumery material. Its central lesson transfers directly: the most abundant components are not always the important ones, and some high-OAV components turn out to be suppressed.
Worked example (outline)
Bergamot oil: published analyses report limonene, linalyl acetate and linalool as the major volatile components. A three-component model smells recognizably citrus-lavender but flat and soapy beside the real oil. Hypotheses: the missing freshness comes from minor terpenes; the missing "peel" from trace aldehydes; the roundness from components not volatile enough to appear prominently. Each is a version and a test. Whatever the result, the log now says something specific about bergamot.
Classic structures
The same method applies to historical accords documented in published formulas — the formula library on this site has examples. Rebuilding a documented structure, then omitting parts, teaches architecture directly. That is study: of published, attributed material, for learning.
The line between study and copying
Analyzing a commercial perfume to learn its structure is common practice. Reproducing it to sell is another matter. In 2006 the Dutch Supreme Court held, in Kecofa v Lancôme, that the scent of Trésor could be protected by copyright and was infringed by a product sharing most of its ingredients; other jurisdictions have taken different views, and the question is not settled across Europe. Nothing here is legal advice; the point is that reconstruction for sale carries risks that reconstruction for study does not.
Limitations
Published analyses vary with origin, harvest, method and column; a reconstruction matches one analysis, not "the" material. Enantiomers and coeluting peaks are invisible in many datasets. And a natural's character is partly components no one has identified.
Research application
A published hypothesis log for the reconstruction of a common natural — which components were tested, which mattered, which surprised — is rare and valuable. Comparing two reconstructions of the same material from two different analyses would show how much "the" material is an average.
Practical exercise: Reconstruct one natural material from published data
- Choose an essential oil or absolute you own, and find one published analysis of it. Record its source, origin and method.
- Build the first model from the major components you can obtain, at reported proportions.
- Write at least five dated hypotheses for the differences you smell, then test each with a single change.
- Compare the final reconstruction blind with the real material, with at least three people.
You have done this well if the model starts from cited data, each version changes one thing and has a dated hypothesis, every comparison was blind, and your log states plainly which hypotheses failed.
Common mistakes
- Treating area percent as an odor recipe.
- Changing several things per version. Then no hypothesis is tested.
- Reconstructing a commercial perfume for sale. Study is not reproduction.
Safety
Reconstructions are smelled on strips. Several citrus components are restricted or phototoxic; read every SDS and follow the safe handling lesson.
Key terms
- Reconstitution
- A blend of identified components built to approximate a natural material, from analytical data. It tests which components carry the smell.
- Hypothesis log
- A dated record of what each change to a reconstruction was meant to test and what it showed, kept so the learning survives the reconstruction.
Knowledge check
Written and reviewed by the RUŌOD Lab team. This is general education, not legal, regulatory, medical or safety advice. Example dilutions and exercises are for learning, are not validated products, and have not been safety-assessed. Completing a lesson confers no qualification. How we check these lessons.