Unpacking the Smell of Wood: Hints of Cucumber, Cheese, and Vanilla

The smell of wood can be resinous, slightly sweet, and calming. It may come as a surprise that it also contains compounds reminiscent of cucumber, cheese, and vanilla.

A 2018 study examined Scots pine wood using both the human nose and chemical analysis. It detected 44 odor-active compounds. The smell of wood was formed not just by familiar terpenes but by an overlapping range of components with different odor qualities.

Separating what evaporates from what we can smell

Scots pine, scientifically known as Pinus sylvestris, is widely used for furniture and building materials. The volatile compounds released from its wood into the air have been studied, but not all of them necessarily produce an odor detectable by the human nose.

The researchers wanted to go beyond an inventory of substances in wood to identify compounds that actually have an odor and connect them with the impressions people get when smelling the wood.

The samples came from three different trees in a forest near Nuremberg, Germany. The wood was dried and cut into thin shavings for sensory evaluation of the wood itself and analysis of its extracted odor compounds. The material studied was wood, not the air of an entire forest or pine essential oil.

A resinous whole with subtler impressions beneath

First, 30 untrained participants rated the intensity and pleasantness of one wood sample. On a scale from 0 to 10, mean intensity was 7.8 and pleasantness was 6.2. The odor was strong and, on average, somewhat pleasant.

Next, 11 assessors who had trained for at least six months described the aromas of the three samples in detail. They agreed on 12 descriptors and rated the intensity of each.

For sample 1, described in detail in the paper, the strongest impression was resinous, with a rating of 7.2. Carpenter’s workshop or sawdust and frankincense impressions followed, alongside descriptors such as wood glue, herbs, pepper, and citrus.

There were also cucumber, cheese or sweat, and pencil impressions, but these received low ratings. This does not mean the wood smelled like food overall; rather, subtle elements surrounded its dominant resinous impression.

People smell the components as they are separated

Volatile compounds extracted from the wood were separated by gas chromatography. A person sniffed part of the outlet stream as the compounds emerged one after another, recording where odors appeared and what they smelled like.

The extracts were then progressively diluted by factors of 3, 9, 27, and so on to determine how far each compound remained detectable. The last dilution at which it could be detected was its flavor dilution, or FD, factor. This helps identify candidates whose odors persist even in a diluted extract.

Across the three samples, 44 odor-active compounds were detected, and every compound was smelled in at least two samples. Structures were assigned to 39 compounds through comparison with reference substances and other evidence.

The confidence of identification was not uniform, however. Twenty-three compounds were confirmed using evidence including mass-spectral comparison with reference substances, and another was matched to a mass-spectral library. The remainder included tentative identifications based on elution position and odor, because low quantities or other constraints prevented sufficiently informative mass spectra from being obtained.

Candidates for resin, cucumber, cheese, and vanilla notes

Woody, resinous α-pinene appeared as a large analytical peak and remained detectable at high dilution. This agrees with the strong resinous impression of the whole wood and suggests that it is an important contributor.

Other compounds had the following recognizable odor qualities.

• Cucumber-like: (E,Z)-2,6-nonadienal. A candidate for the faint impression of freshly cut cucumber reported in the sensory evaluation.
• Cheese-like: Butanoic acid, 3-methylbutanoic acid, and related compounds. Separating the wood’s odor into its components reveals these food-like elements as well.
• Vanilla-like: Vanillin. In at least one sample, it remained detectable after the extract had been diluted 729-fold.
• Honey-like: Phenylacetic acid. One of the compounds with a sweet impression.
• Pencil-like: Thymoquinone. A candidate contributor to a woody smell reminiscent of pencils.
• Coconut- or peach-like: Lactones. The wood’s odor also contained compounds with fruit-like qualities.

Even when the same compound is present, its prominence in a mixture can vary. For example, isolated thymoquinone smells pencil-like, but this impression was weak in the whole wood studied here. The authors discuss the possibility that stronger components masked it.

Lipids and woody tissue, as well as resin, can be sources of odor

The researchers also considered the origins of the compounds they found. One major group consisted of substances thought to derive from fatty-acid degradation, with a wide range of odors including grassy, fatty, cheesy, and coconut-like notes.

When the paper grouped the odor compounds by their origins, this group accounted for 69%. That figure describes the proportion of compound types. It does not mean 69% of the wood’s odor intensity or 69% of the amount of volatile material.

Another group consisted of terpenes found in resin. These included α-pinene, citrus- and floral-smelling linalool, and balsamic, peppery α-bisabolol. Aromatic compounds such as vanillin were also identified as degradation products related to lignin, a constituent of wood cell walls.

The components supporting the smell of wood thus appear to arise not just from resin but from several materials within the wood and their transformations.

Reconstructing the overall smell of wood remains a next step

The carpenter’s workshop or sawdust, frankincense, and wood-glue impressions were not linked to a single specific compound in this analysis. Mixing odors may produce impressions that are difficult to predict from the individual components.

Testing those relationships requires accurate measurement of each compound’s quantity and experiments combining them to reconstruct the wood’s aroma. The FD factors in this study help prioritize candidates; they do not directly establish the size of each compound’s contribution to the whole wood’s odor.

The samples were also limited to three trees. Nevertheless, using the human nose to find trace components that instruments may struggle to detect established a foundation for studying wood odor in terms of specific molecules. It can also support research on how processing and storage change that odor.

Summary

Scots pine wood had a predominantly resinous impression while containing compounds reminiscent of cucumber, cheese, vanilla, honey, and other odors. Forty-four odor-active compounds were detected in samples from three trees, and structural evidence was obtained for many of them.

A woody smell is not made of a single “wood-odor molecule.” Seeing it as the overlap of diverse compounds derived from resin, lipids, and woody tissue brings new depth to the smell of familiar timber.

The paper acknowledges assistance with sample provision from Staedtler Mars and others. Publication was supported by the German Research Foundation and Friedrich-Alexander-Universität Erlangen-Nürnberg. The authors declared no competing interests.

References

Schreiner L, Bauer P, Buettner A. (2018). Resolving the smell of wood – identification of odour-active compounds in Scots pine (Pinus sylvestris L.). Scientific Reports, 8, 8294. Original article and open-access full text