Treating the Air, Not the SkinThe ambient layer

Article 2 runs two households through its decision framework and leaves two situations unanswered: Hugo asleep with the window cracked, and Delphine asleep with her child and the windows open. In both, the verdict is the same and it is unsatisfying: the boundary layer is not sufficient, whatever composition is chosen.

It also leaves an intuition it flags explicitly and does not correct: that a treated space works like a shield. Both belong to the same layer, the one this article covers: the air, rather than the skin.

A topical repellent acts on a surface. An ambient treatment acts on a volume: it raises the concentration of a volatile molecule in the air of a space, enough to disrupt host-seeking. These are two different mechanisms, and they do not degrade in the same way.

A surface has an edge. A volume does not. Effective concentration falls off with distance from the source, and it falls off faster the more the air is replaced. Someone sitting at the far end of the table is therefore not standing just outside a shield: they are in the same treated zone as everyone else, simply at a lower concentration, with a correspondingly smaller effect.

The practical consequence is counter-intuitive and worth holding onto: the dominant variable is not the oil, it is the air volume and whether it stays put. The figures in the next section say nothing else.

Article 2 notes that the strongest data on geraniol comes from ambient diffuser trials, and that it says nothing about how geraniol behaves on skin. This is where it becomes relevant: this layer is the question those trials were answering.

Müller et al. (2009) measured the same molecules across two devices and two environments. The gaps matter more than the values:

Two structural readings. First, the device outweighs the molecule: citronella goes from 14% to 68% without changing oil, simply by moving from a candle to continuous release. That is a wider gap than the one between citronella and geraniol at equal device. Second, enclosure outweighs both: the same diffusion loses two thirds of its effect moving from a room to a terrace.

One last reading, a practical one: that row comes in a bottle. Geraniol is a molecule, but the oil carrying the most of it is rose geranium, at roughly 26% geraniol alongside 40% citronellol, the two best-documented repellent compounds together in a single bottle. That is what places it, with citronella, among the most consistently active whole oils in comparative testing.

Three independent reasons rule the ambient layer out of the night scenario, and none of them is corrected by a better blend.

An open window is not an enclosed volume. Every high figure in Section 3 comes from a room that holds its air. A cracked window is precisely the condition that removes the variable the effect depends on. Continuous overnight diffusion is not a recommended practice in its own right, independently of mosquitoes: the common rule remains 20 to 30 minutes in a ventilated room. A device that should not run for eight hours cannot cover an eight-hour exposure. And finally, nobody is awake to adjust anything, which is exactly the constraint Axis D in Article 2 was measuring.

The answer to this scenario is physical: a net over the bed, or a screen on the window. It carries no dilution ceiling, no reapplication burden, no sensitisation risk and no population restriction. It works all night, it works for a child, and across this whole series it is the protection with the strongest evidence base.

Ending three articles of aromatic chemistry on “put up a net” is faintly deflating, and that is exactly why it needs saying. A framework that can never conclude that the right answer lies outside its own subject is not a framework, it is a sales argument.

As with skin, there is a regulated reference standard above aromatic diffusion. Plug-in insecticide vaporisers, coils and mats are pyrethroid-based and are biocidal products: they carry an authorisation number and reviewed claims. Essential oil diffusion is not in that category and does not carry those claims. Combustion also has a cost of its own: burning a coil in a closed room introduces a particulate exposure that has nothing to do with mosquitoes.

In areas where mosquito-borne illness is a real risk, neither this layer nor the previous one is the subject: public health guidance is.