Click here to view the full publication.

Tetrahydrocannabinol acid (THCA) and cannabidiol acid (CBDA), the two crucial organic components in cannabis and hemp, decarboxylate at different rates to their more active neutral forms. Theoretical calculations are used herein to analyze how the remote annulated ring or pendant substituent influences the rate determining steps of the decarboxylation processes. The uncatalyzed keto-enol tautomerization that precedes decarboxylation is found to be extremely slow in both cases albeit with a ten-fold preference for CBDA. A single molecule of methanol dramatically enhances the reaction rates by allowing for tautomerization through a more favorable six-membered ring transition state. Methanol-catalyzed tautomerization is found to be faster in THCA than in CBDA. This difference results from both the larger dipole moment of the THCA scaffold as well as its greater rigidity relative to CBDA. The greater dipole moment leads to a somewhat better binding of methanol. The lower entropic penalty in THCA towards tautomerization leads to faster decarboxylation.

The cannabis and hemp industries have witnessed exponential growth in the past decade due to changes in their legal status around the globe.1,2 With legalization, consumer interest and demands have evolved from predominantly direct flower sales to a broad range of products including cannabinoid concentrates and infused products.3 While both the cannabis and hemp extract industries rely on the cannabis plant, the former requires isolates with high tetrahydrocannabinol (THC) concentration and the latter requires isolates with both high cannabidiol (CBD) concentration and minimal THC contamination.4 Further complicating separation, the naturally occurring form of these cannabinoids are the acid forms, THCA and CBDA.5 These acids are thermally unstable and undergo decarboxylation upon heating to the neutral, and more potent psychoactive molecules THC and CBD, respectively. It has been reported that THCA decarboxylates at a faster rate than CBDA, but no chemical explanation has been proposed to date.6 If the mechanistic foundation of this rate difference can be determined, decarboxylation rates can be exploited to better control and optimize cannabinoid production.

Figure 1. Summary of the state of current knowledge in the decarboxylation of THCA and CBDA via computational studies. The role of remote substitution on the benzoic acid has not been explored and thus the origin of the differing rates of THCA and CBDA decarboxylation have not been elucidated until this work.

There have been a few reported studies on the mechanism of THCA decarboxylation and simplified structural analogs. An early computational study by Ruelle demonstrated that water catalytically lowers the activation energy for the thermal decarboxylation of salicylic acid (Figure 1).7 More recently, Li and Brill explored the role of hydroxyl groups on the aromatic ring in the decarboxylation and found that ortho substitution was critical.8 Chuchev and BelBruno observed the formation of a critical keto-type intermediate and showed the direct role of the o-OH group in the development of the transition state.9 Under acid-catalyzed conditions, Perrotin-Brunel et al. refined this mechanistic proposal and provided further evidence that direct keto-enol tautomerism was the key step in HCOOH-catalyzed decarboxylation of THCA and computed an activation barrier (81 kJ/mol) that is in good agreement with experiment (85 kJ/mol).10 These studies have provided a foundation for understanding the mechanistic pathway of benzoic acid derivatives and even THCA decarboxylation. However, there are no studies that mechanistically explain the discrepancy observed for the rate differences between THCA and CBDA.6 Herein, we report a computational study on the decarboxylation of Δ9-THCA and CBDA that identifies the key mechanistic differences that account for the differing decarboxylation rates.

The difference in kinetic reactivity between two similar systems is often a result of minor electronics or sterics near the reaction site. Simultaneously, the conformational space of complex molecules can be extremely large. It can be beneficial to simplify a computational model to remove structural features that may dramatically increase computational cost while having only a minor influence on the process under investigation. Figure 2 summarizes the key features of both Δ9-THCA and CBDA, which must be maintained: the aromatic core, the dihydropyran ring systems in Δ9-THCA, the acyclic isoprenyl fragment in CBDA, and the local steric environment alpha to the carboxylic acid. Deletions in both 𝑻𝟏 and 𝑪𝟏 involve the removal of the same features: abbreviating the o-pentylsidechain to a methyl group and removing the remote 1-methylcyclohex-1-ene fragment.

Figure 2. Simplified Model of Δ9-THCA (𝑻𝟏) and CBDA (𝑪𝟏). Atoms in blue were deleted to generate simplified models with significantly fewer degrees of freedom without significantly affecting the overall reaction of interest.

DFT calculations were performed to evaluate both the conformational space of the initial reagents, as well as potential reaction pathways. Gas phase calculations were first performed to evaluate the inherent differences in reactivity between THCA and CBDA towards decarboxylation. The role of solvation and explicit methanol in the reaction mechanism were evaluated in subsequent steps.

Fig. 3. Overall view of simple gas phase decarboxylation reaction pathway for Δ9-THCA (top, T1) and CDBA (bottom, C1) models. Reactants are set as 0 kJ/mol in both reaction paths. Relative energies are listed for all intermediates and transition states as Fig. 3. Overall view of simple gas phase decarboxylation reaction pathway for Δ9-THCA (top, T1) and CDBA (bottom, C1) models. Reactants are set as…

Click here to view the full publication.

Join our newsletter

Receive news about our latest research and more!