NIR spectroscopy for rapid measurement of moisture and cannabinoid contents of industrial hemp (Cannabis sativa)

Authors: SU, KAI - University Of Kentucky; Maghirang, Elizabeth; TAN, JIA WEN - University Of Kentucky; YOON, JU YOUNG - University Of Kentucky; Armstrong, Paul; KACHROO, PRADEEP - University Of Kentucky; HILDEBRAND, DAVID - University Of Kentucky

Submitted to: Industrial Crops and Products
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/25/2022
Publication Date: 5/13/2022
Citation: Su, K., Maghirang, E.B., Tan, J., Yoon, J., Armstrong, P.R., Kachroo, P., Hildebrand, D. 2022. NIR spectroscopy for rapid measurement of moisture and cannabinoid contents of industrial hemp (Cannabis sativa). Industrial Crops and Products. https://doi.org/10.1016/j.indcrop.2022.115007.
DOI: https://doi.org/10.1016/j.indcrop.2022.115007

Interpretive Summary: In 2018 the industrial production of hemp was legalized in the United States but it is required that the tetrahydrocannabinol (THC) content be less or equal to 0.3% for the US and 0.2% for Europe. Since that time the market for various cannabinoids including cannabidiol (CBD), THC, cannabinol (CBN), cannabigerol (CBG), and cannabichromene (CBC) have flourished. There is a critical need for a rapid and accurate method to measure these constituents for regulatory compliance, determination of usage, and for quality control. Current measurement methods such as ultra-high performance liquid chromatography and gas chromatography requires sophisticated instruments, are time-consuming, and require highly-trained personnel. Near-infrared spectroscopy (NIR) was looked at as a method for rapid measurement of moisture and cannabinoid contents of ground and whole hemp. Calibrations for predicting moisture content (MC) and cannabinoids were developed for a commercial NIR instrument. This work discovered there is good potential for measuring MC, CBD, and THC <0.5% for ground hemp, but measurement of CBG and CBN were poor. For whole hemp, measurement was good for MC, CBD and CBC while poor accuracies were obtained for THC, CBG, and CBN. Results have also shown that industrial hemp can be classified as legal or illegal based on specified limits, such as those for the United States and Europe (=0.3% and =0.2% THC, respectively). NIR spectroscopy was shown to be effective for identification of legal hemp and measure some hemp constituents. The concurrent prediction of MC allows real-time adjustments for MC of cannabinoid contents. These are important tools that will benefit the hemp industry and regulators.

Technical Abstract: With the legalization in 2018 of industrial hemp in the United States, although requiring a tetrahydrocannabinol (THC) content of =0.3% in the U.S. and =0.2% in Europe, the market for various cannabinoids including cannabidiol (CBD), THC, cannabinol (CBN), cannabigerol (CBG), and cannabichromene (CBC) have flourished. There is a critical need for a rapid method to measure these constituents for regulatory compliance, determination of usage, and for quality control considering that current measurement methods such as ultra-high performance liquid chromatography and gas chromatography requires sophisticated instruments, which are time-consuming and require highly-trained personnel. In this work, the potential of using near-infrared spectroscopy (NIR) (950 to 1650 nm) for rapid measurement of moisture and cannabinoid contents of ground and whole hemp was evaluated. Calibration models for predicting moisture content (MC) and cannabinoids were developed using partial least squares (PLS) analysis. The reference values for MC were obtained from a nuclear magnetic resonance instrument calibrated (R2 = 0.99) for hemp MC while the cannabinoids were measured using gas chromatography. For ground hemp, there is good potential for measuring MC, CBD, and THC <0.5 as shown by independent validation for MC (R2 = 0.91, SEP = 1.28), CBD (R2 = 0.85, SEP = 0.92), and THC <0.5 (R2 = 0.81, SEP = 0.03). However, poor predictions were obtained for CBG and CBN of ground hemp. For whole hemp, constituents with good potential for NIR prediction included MC (R2 = 0.94; SEP = 1.24), CBD (R2 = 0.89, SEP = 0.60), and CBC (R2 = 0.79, SEP = 0.02). Poor predictions were obtained for THC, CBG, and CBN of whole hemp. Discriminant analysis showed that industrial hemp can be classified as legal or illegal based on specified limits, such as those for the United States and Europe (=0.3% and =0.2% THC, respectively). Calibration models developed showed accuracies ranging from 93.0% to 94.0% and 90.1% to 90.8% with correct classifications for independent validation at 83.9% to 87.5% and 84.4% to 85.6% for limits set by the U.S. and Europe, respectively. NIR spectroscopy technique can identify legal hemp and measure some hemp constituents. The concurrent prediction of MC allows real-time adjustments for MC of cannabinoid contents. These are important tools that will benefit the hemp industry and regulators.