Predicting tea tree oil distillate composition using portable spectrometric technology

Abstract Commercial tea tree oil (TTO) has to conform chemically the international standard ISO4730: 2017, which defines acceptable ranges for 15 main terpenoids therein. Lack of in‐house capacity technology typically required terpenoid quantification (GC‐FID) necessitates that Australian TTO producers outsource analyses commercial service providers. This is associated with high cost and slow turnaround times, have been recognized as limiting factors process optimization in areas quality assurance, distillation blending. paper describes testing two custom portable Raman devices different lasers (785 1064 nm) relative components using model‐based predictions. Initial showed 1064‐nm device provided superior data Machine‐learning algorithms were trained spectral from corresponding GC‐FID 214 samples. Hold‐over validation (HV) correlations between actual predicted values up 53 unseen samples r 2 at or above 0.92 10 0.95 6 specified terpenoids. The highly abundant key compound terpinen‐4‐ol an HV 0.96 a root mean square error (RMSE) 0.37%, whereas lowly 1,8‐cineole 1.00 RMSE 0.13%. These results indicated strong potential spectrometry provide real‐time product sites enabling direct feedback control optimisations.