Rapid determination of patchouli oil quality as a tropical plant using infrared technology

Abstract

Patchouli alcohol (PA) content is an important indication in the world patchouli oil trade. Testing the PA content of patchouli oil usually takes a long time and expensive, so that it greatly inhibits the activity of buying and selling patchouli oil in the community. Therefore the main objective of this study is to develop NIRS technology to estimate the PA content in Aceh patchouli oil purified rapidly and appropriately using PLSR and PCR methods. The sample in this study was patchouli oil which was purified into 20 types based on PA content. Laboratory test using GC-MS. Near infrared (NIR) spectrum of Patchouli oil samples were acquired and recorded using self-developed FT-IRIPTEK T-1516. The results showed that NIRS technology with PLSR and PCR methods provides a solution to solve the problem of PA content prediction in Aceh's patchouli oil. SNV pretreatment is proven to improve prediction accuracy and robustness of the prediction model that built. The PLSR-SNV method succeeded in obtaining a prediction model that was categorized as very good model performance, while PCR-SNV method succeeded in obtaining a prediction model that was categorized as good model performance.