When a fragrance diffuser is used, fragrance compounds enter the room air. But in an ordinary occupied home, how difficult is it to measure only the increase attributable to that product?
Thomas Warburton and colleagues, in a study published in 2023, compared air during periods with and without a plug-in fragrance diffuser in 60 households in Ashford, UK. Their main objective was to determine how much one product with known emissions added to concentrations of volatile organic compounds, or VOCs, in real homes.
Repeatedly collecting three days of air
Participating households were selected from a fragrance-industry panel of untrained evaluators. Two groups of 30 households were studied at different times, with air collected in the main living space at least four times per home: before use, twice while the diffuser was operating, and after it had been switched off.
Sampling containers slowly admitted small amounts of air, filling over 72 hours from Friday morning to Monday morning. After excluding incomplete collections and other unsuitable samples, 259 samples were analyzed and 47 VOCs quantified. Residents were also asked about how often they used other household products and about room characteristics.
The study used one diffuser model with a constant output and one fragrance formulation. What it called total VOC was the sum of the 47 measured compounds, not every organic compound in the air. The researchers also summed six measurable compounds known to be in the formulation as an indicator of fragrance-related changes. This did not represent the entire formulation either.
Some compounds increased, but there was not necessarily one cause
VOC concentrations varied enormously between homes. Across the dataset, some compounds, including α-pinene and eucalyptol, and the sum of measured compounds increased significantly while the diffuser was operating. Other compounds decreased. The result was not that nothing changed across the households.
However, not every change could be attributed to the diffuser. Butane, propane, and ethanol accounted for much of the measured quantity, and indoor spaces contain many other sources. Substances also found in the diffuser can be emitted by other products or materials.
A separate analysis counting whether concentrations were higher or lower during use in each home likewise did not show all six fragrance compounds increasing together. There were also few clear relationships between residents’ reported frequency of product use and individual compound concentrations. The authors considered that substantial fluctuations in occupied homes made the contribution of one product difficult to isolate.
An increment was visible in homes inferred to have less ventilation
The researchers therefore focused on air exchange. Carbon dioxide concentrations measured in a subset of samples were positively related to total VOC concentrations. On that basis, homes with higher total VOC concentrations before diffuser use were classified as potentially having less ventilation. Ventilation rates were not measured directly in every household.
In the quarter of homes with the highest pre-use total VOC concentrations, the sum of the six measured fragrance compounds approximately doubled during diffuser use, a statistically significant increase. α-Pinene also increased significantly. Differences between use and non-use periods tended to be smaller in the other groups.
The authors interpreted this as evidence that a product-related increment can be easier to detect under conditions of low air exchange, even with background sources present. However, high VOC and carbon dioxide concentrations can reflect occupant activities and combustion appliances as well as ventilation. The classification therefore retains this uncertainty.
Estimating the increment from emissions and room conditions
Separately from the air measurements, emissions of individual compounds were calculated from the diffuser’s weight loss and formulation information. These values were combined with room volume and ventilation estimates derived from carbon dioxide and other information to estimate the increase in airborne concentrations.
The predicted α-pinene increment was broadly within the same order of magnitude as the measured change. In the model, the same emissions produce higher concentrations in a smaller room with less air exchange. The authors suggested that this could inform product guidance for use in very small, poorly ventilated spaces.
However, the model treats a room as one well-mixed space and omits oxidation reactions and uptake by surfaces such as walls. It does not reproduce a real home completely, and its calculations do not guarantee concentrations or safety in an individual dwelling.
What about other substances formed by reactions?
VOCs differ not only in their initial amounts but also in their tendency to react in air and form other substances. The study also calculated an indicator of secondary-product formation potential by weighting concentrations using existing coefficients. For fragrance compounds, this covered α-pinene, β-pinene, and benzaldehyde, for which coefficients were available.
The fragrance compounds’ contribution was larger in this indicator than in the simple mass sum. However, across all households, diffuser use did not significantly increase the indicator. This was a calculation using coefficients and measured concentrations, not a measurement of the secondary products themselves. It did not establish that secondary products do not increase or that health effects are absent.
Conclusion
The study shows that measuring diffuser emissions alone does not establish how concentrations change in homes where people are exposed. Real homes contain many sources and differ in ventilation and room size. Within that variability, the increment in measured fragrance compounds was clearer in households inferred to have less ventilation.
An important implication is to evaluate product emissions together with the conditions of the space in which a product is used. Comparing measurements with a simple model helps identify settings where concentrations may rise more readily. However, this study covered one model and one formulation, and 72-hour sampling does not adequately capture short concentration peaks. Symptoms and individual health risks were not measured, and the study does not provide a blanket judgment that these products are safe or dangerous.
Research funding came from Givaudan and Bath & Body Works, and employees of both companies were coauthors. The work also received national research support. The authors stated that measurements and data analysis were performed by the University of York and that there were no restrictions on publication.
References
Warburton, T., Grange, S. K., Hopkins, J. R., Andrews, S. J., Lewis, A. C., Owen, N., Jordan, C., Adamson, G., & Xia, B. (2023). The impact of plug-in fragrance diffusers on residential indoor VOC concentrations. Environmental Science: Processes & Impacts, 25(4), 805–817. DOI: 10.1039/D2EM00444E / Full text. Original article: CC BY 3.0.
