Water distillation

The raw material goes into the still with water, usually fully submerged. The still is heated from below by a direct flame, or indirectly by a steam jacket in larger operations.

Water and raw material boil together. The rising steam carries the essential oil, passes through the condenser and separates in the receiver, as in every other form of distillation.

Distillation usually takes longer than in the other two methods, and for some materials such as patchouli or vetiver it can last a whole day.

For rose there is an additional step called cohobation: the water left after separating the oil is distilled again. The reason is that 2-phenylethanol, the compound behind much of the rose character, dissolves fairly well in water and stays there instead of rising with the oil. Without this step, rose oil lacks its most important part.

This is the most important thing to know about this method. A raw material soaked in boiling water for hours is exposed to a chemical environment quite unlike steam that merely passes through it.

Esters are hydrolysed: linalyl acetate splits into linalool and acetic acid. So the same batch of lavender or clary sage, distilled in water, gives a lower ester share and a higher share of free alcohols than steam distillation. Both species are rich in linalyl acetate, so the difference is visible straight away on an analysis.

With low-ester materials the trace is faint or absent. Lemongrass, for example: its leading compound is citral, an aldehyde, so ester hydrolysis says almost nothing here. The difference is measurable on an analysis, and it comes not from the plant but from the equipment.

The water also becomes slightly acidic during distillation, which pushes hydrolysis further still. The result is a softer, less sharp odour profile, and for some species one that is genuinely different rather than merely weaker.

Material touching the hot wall of the still can scorch, sending a burnt smell into the whole batch. This is the most common fault with direct firing on a wood stove.

Long run times and temperatures that cannot be controlled independently of the raw material make batch-to-batch variation much larger. Two batches of the same material from the same distiller can still give two different analyses.

It is unsuitable for ester-rich materials whose esters you want to keep, such as lavender or clary sage.

On an analysis, the suggestive sign is an unusually high share of free alcohols relative to esters in species that are naturally ester-rich. For lavender, a linalool-to-linalyl-acetate ratio skewed strongly towards linalool is an indicator.

Measurable free acetic acid is also a sign, although it is usually not reported on an ordinary commercial analysis.

To be clear: this is a suggestive indicator, not proof. Cultivar, harvest time and growing region all affect the same ratio.

This is the form of distillation Persian alchemists used from the tenth century, and in principle it has hardly changed since.

In Vietnam, the directly fired copper still is a familiar sight in the distilling of lemongrass, cajuput, mint and star anise oils. The craft exists in Hung Yen, Nghe An, Lang Son and many other areas, mostly at household scale.

It remains common not because it is backward but because it makes sense: low capital, equipment that can be repaired on the spot, and results that are good enough for many materials. But knowing which method you use is a precondition for reading your own analyses correctly.