Thesis
A chemometric approach to rapid quantification of terpenoids and cannabinoids using Raman and NIR spectroscopy
Southern Cross University
Master of Science (MSc), Southern Cross University
2021
Metrics
80 File views/ downloads
97 Record Views
Abstract
Natural products derived from Tea Tree (Melaleuca alternifolia) and Cannabis Sativa have witnessed considerable growth across global therapeutic markets in recent years, benefiting producing countries such as Australia. Tea Tree oil (TTO) is governed by international standards ISO 4730:2017, defining the acceptable ranges of 15 terpenoids. C. sativa is regulated in accordance with the concentrations of the psychoactive cannabinoid Δ9-Tetrahydrocannabinol (Δ9-THC). Additionally, the medicinal Cannabis industry examines up to a dozen cannabinoids in female inflorescence.
Typically, chromatographic analysis is used to identify and quantify target compounds in TTO and C. sativa. Despite being state-of-the-art, cost and turn-around times for chromatographic analyses are prohibitive for large-scale applications. Spectroscopic techniques such as Raman spectroscopy (RS) and Near-Infrared spectroscopy (NIRS) have become feasible alternatives for a scalable, cost-effective, real-time solution to compound quantification and identification.
In this thesis, RS and NIRS coupled to chemometrics were investigated for their capacity to predict terpenoid and cannabinoid content in TTO and C. sativa. Two RS devices of different laser wavelengths (785 nm and 1064 nm) were explored for TTO analysis. Using 214 TTO samples with corresponding gas-chromatography data showed that 1064 nm provided superior predictions. Ten out of 15 target terpenoids showed r2 values above 0.9 between predicted and actual values in hold-over validation (HV), with key terpenoids terpinen-4-ol and 1,8-cineole predicting r2 above 0.95. For C. sativa, NIRS was explored. NIRS data were used to build predictive chemometric models for 12 cannabinoids using 249 samples. Four major cannabinoids, including Δ9-THC, displayed r2 values above 0.9 in HV, while r2 values ranged between 0.7 and 0.9 for most minor cannabinoids.
RS - chemometrics reliably predicted TTO composition for distillation optimisation and quality control. Statistic metrics for HV were inferior to the cross-over validation (CV), with most notable discrepancies observed for α- and γ-terpinene with a CV r2 of 1 and a HV r2 at 0.70 and 0.65 respectively. NIRS - chemometrics showed potential to quantify major, and minor cannabinoids in C. sativa dried flower materials. However, limitations with low HV r2 values were observed for minor C3-alkyl, in particular cannabidivarin (CBDV) and cannabidivarinic acid (CBDVA) with HV r2 values of 0.03 and 0.5, respectively.
Overall this thesis demonstrates the use of portable spectroscopic technologies aided with machine learning platforms as a viable analytical tool for phytochemical detection and analysis that is cost-effective, robust and efficient for phytochemical screening.
Details
- Title
- A chemometric approach to rapid quantification of terpenoids and cannabinoids using Raman and NIR spectroscopy
- Creators
- Francine Gloerfelt-Tarp
- Contributors
- Tobias Kretzschmar (Supervisor) - Southern Cross UniversityAmitha Hewavitharana (Supervisor) - Southern Cross University
- Awarding Institution
- Southern Cross University; Master of Science (MSc)
- Theses
- Master of Science (MSc), Southern Cross University
- Publisher
- Southern Cross University
- Number of pages
- xvi, 279
- Identifiers
- 991013098513202368
- Copyright
- © Ecofibre Ltd 2021
- Academic Unit
- Faculty of Science and Engineering; School of Environment, Science and Engineering
- Resource Type
- Thesis