Evaluating Different C18 Columns for the LC Separation of Δ9-THC and Other Related Cannabinoids

Abstract The best separation possible of delta-9-tetrahydrocannabinol (Δ 9 -THC) in complex samples via liquid chromatography (LC) is particularly challenging due to the potential for interference from other cannabinoids. Over 100 cannabinoids have been detected in Cannabis sativa plant samples. The separation of Δ 9 -THC in Cannabis -derived finished products was believed to be easier than plant extracts to analyze at one time because the acidic cannabinoids are converted to their neutral cannabinoids during material preparation. However, the emergence of synthetic or semi-synthetic cannabinoid products, such as delta-8-tetrahydrocannabinol (Δ 8 -THC), has led to more chromatographic interferences due to the presence of synthetic by-products. The original LC separation method implemented in the Chemical Sciences Division (CSD) at the National Institute of Standards and Technology (NIST) was not acceptable when these chromatographic interferences were present. Within this context, the work presented here explores initial separation of an 11 cannabinoid mixture using different monomeric and polymeric octadecylsilane (C 18 ) columns via liquid chromatography. These columns were characterized using the Standard Reference Material 869b as a three-component column selectivity test mixture to determine if an LC C 18 column is classified as monomeric or polymeric. Monomeric C 18 columns (NexLeaf C 18 , ACE 3 C 18 , and ACE Super C 18 ) provided better separations of the 11 cannabinoids, and baseline separations were obtained in less than 13 min after minor adjustments to the mobile phase program. Using the NexLeaf C 18 LC-UV method, mixtures of Δ 9 -THC and four known chromatographic interferences were analyzed. Cannabinolic acid (CBNA) could not be separated from Δ 9 -THC; however, CBNA has drastically different absorbance spectra from Δ 9 -THC, and the co-elution of these cannabinoids can easily be recognized using photodiode array detection. When CBNA is present, the sample can be reanalyzed using an alternate NexLeaf C 18 LC-UV method developed here, which baseline-resolves Δ 9 -THC and CBNA in 60 min. These two LC-UV methods will be further evaluated at NIST CSD through quantitative comparisons in future publications, enabling their use in the development of reference materials for Cannabis plants and/or Cannabis -derived finished products.