Two principal methods are available to determine the oil concentration: steam distillation or solvent extraction. A specific case of steam distillation that has the biomass immersed in the boiling water is sometimes termed hydrodistillation. More specialised methods such as head space analysis are usually confined to detailed experiments. Each of the two methods has its own advantages. Steam distillation can accommodate the larger samples that arise when all positions on a single plant or a number of plants in an agronomically sized plot are sampled. The method mimics the commercial process and gives oil of a similar chemical composition. It is important that the system incorporates a cohobation or reflux return to obtain complete recovery of the alcohol components of the oil (Kawakami et al. 1990; Murtagh 1991a).
On the other hand, solvent extraction followed by quantitative gas chromatographic analysis can use much smaller samples that can vary from a single leaf up to at least 5 g. Solvent extraction also avoids the conversion of precursor compounds in young leaves to the major constituent, terpinen-4-ol, as occurs with steam distillation (Southwell and Stiff 1989; Cornwell et al. 1995), and thereby suits biochemical investigations. It is the system of choice when there are a large number of samples, as in plant breeding programs, and in such sampling the issue of extracts having a different oil composition to the commercial product, which is distilled, can be avoided by not using young growth (Baker 1995). Tea tree has a sufficiently high oil concentration to enable a direct chromatographic analysis of the solvent extract without the need to concentrate the extract by evaporation, thus avoiding the accompanying losses that can be a drawback of the solvent extraction method (Charles and Simon 1990).
In a comparative study with tea tree (Baker et al. 1995), the two methods gave a similar recovery of monoterpenes, but more sesquiterpenes with solvent extraction. A major problem in doing the comparison was to obtain equivalent samples of very different size.
Either fresh or air dried samples can be distilled without affecting the result (Murtagh and Curtis 1991). Although dried samples distill more slowly, distillation is virtually complete within two hours (Murtagh and Smith 1993). Similar delays occur with solvent extraction (Doran et al. 1996). Because of oil losses, samples should not be oven dried before distilling. Losses increase with the drying temperature, and also vary between samples (Curtis and Murtagh 1989). The widest variation they measured between samples ranged from no oil loss at 45°C and 50% loss at 125°C in samples from one tree, to 33% loss at 45°C and 93% loss at 125°C from another tree. Points of note in these results were the consistent pattern of relatively high or low losses across all temperatures within a batch, and the lack of a relation between the relative batch loss and the moisture content of the samples before drying.
Was this article helpful?