Chromatography, Detection, and Software Tools to Aid in the Exploration of Delta-8 THC Distillates

The expanding use of delta-8 tetrahydrocannabinol (Δ⁸-THC) which is manufactured from hemp derived CBD is causing safety concerns in the United States (US). The conversion of CBD to Δ⁸-THC can lead to multiple reaction byproducts which need to be characterized to enhance understanding of the chemical components produced and their potential risks to consumer safety.

The following work describes the analysis of Δ⁸-THC distillates using High-Performance Liquid Chromatography (HPLC) with photodiode array (PDA), single quadrupole mass spectrometry (MS) detection and tools within Empower™ Chromatography Data Software that aid in sample exploration.

Several known cannabinoids were identified in the distillate sample analyzed based on retention time and UV spectra, including Δ⁸-THC and Δ⁹-THC. However, several unidentified peaks were also detected in the UV data. The UV spectra indicated that there may be structural similarities between the unknown components and the primary compound in the distillates, Δ⁸-THC. In the MS analysis, the software highlighted m/z 315 as the base peak for several unknowns. Cannabinoid spectral libraries were generated using authentic standards. The spectra and retention times, were used to aid in confirming the identity of the target components and highlighting potential structurally related unknowns by spectral matching. 

Benefits

• Enhanced confidence in the detection of cannabinoids using both PDA and MS detection
• PDA spectral libraries increase confidence in compound assignments and enable possible isomer identification
• Peak Purity can be used to aid in the identification of co-eluting components
• Additional structural information using in-source fragmentation of precursor to product ions

Expanding cannabis legalization has driven innovation in mechanisms of consumption for cannabis and its derived products. The importance of correctly identifying cannabinoids is of increasing significance as the safety of various routes of ingestion are unknown. Delta-9 tetrahydrocannabinol (Δ⁹-THC) is the main intoxicating component in the cannabis plant. Its isomer, Δ⁸-THC, is also intoxicating and naturally occurs in the cannabis plant at trace levels. The Δ⁸-THC found in consumer products is produced from hemp-derived cannabidiol (CBD), which is a non-intoxicating cannabinoid. Interpretation of the Unites States Agriculture Improvement Act (2018 US Farm Bill) has led to varying discourse about Δ⁸-THC legality; many feel that Δ⁸-THC is legal, leading to a steady increase in US sales. Regulations governing the use of synthetic components derived from hemp are not clearly addressed which has created a growing market for Δ⁸-THC production and use. Synthetic conversion of CBD to Δ⁸-THC leads to multiple reaction byproducts which need to be characterized to enhance both the chemical understanding of the resulting components and their safety profiles.^1–7

Recently, there have been reports of product labeling inaccuracies associated with Δ⁸-THC consumer products such as incorrect labels and undeclared reaction byproducts, including isomers.^4–6 There are several known double bond isomers of Δ⁹-THC with differing stereochemistry that can be present, which could create a considerable separation challenge. This could also cause legal issues if the analytical identification of the regulated Δ⁹-THC is ambiguous.⁸

The following work describes the analysis of Δ⁸-THC distillates using HPLC with PDA and single quadrupole MS detection and tools within Empower Chromatography Data Software that aid in sample exploration.

The chromatographic separation of an authentic standard mixture of ten neutral cannabinoids (Figure 1) using PDA detection at 228 nm is shown in Figure 2. The mixture contains several isomers of Δ⁹-THC including CBD, exo-THC, Δ⁸-THC, CBL, and CBC. 

Following the analysis of a cannabis distillate sample, several compounds were identified based on retention times (t_R) recorded in the Empower processing method: CBD, CBN, exo-THC, Δ⁹-THC, and Δ⁸-THC. Two components eluting after the main Δ⁸-THC peak with Area% values of 2% and 12%, respectively, were integrated but not identified by the processing method.

The main component, identified as Δ⁸-THC (at t_R 5.351 minutes), was calculated to have an Area% of 66% (Figure 3).

In the region of the chromatogram between 4.4 to 5.3 minutes (the peaks preceding the Δ⁸-THC peak at t_R 5.351 minutes), the UV and MS spectral data for the unknown component peaks showed close similarities to those of the identified components. The protonated ion [M+H]⁺ m/z 315 was observed for each of the identified components: exo-THC, Δ⁹-THC, and Δ⁸-THC. The base peak for each of the unknown component peaks at t_R 4.512 min, 4.754 minutes and 5.229 minutes was also m/z 315 Figure 4. These data suggest that the components may be structurally related and possible isomers.

The UV spectra for the two components eluting after the main Δ⁸-THC peak at t_R 6.151 min and 6.530 minutes differed significantly from the six components described in Figure 4, despite also having a base peak of m/z 315 (Figure 5).

The UV spectra for all integrated peaks can be compared in a single view using the Empower Spectrum Index plot (Figure 6). The clear spectral differences between the last two eluting components and the previous six can be observed in Figure 6.

Empower library searching can be used to match unknown spectra from peaks with PDA or mass spectral data stored in a library. The library spectra are first collected using authentic standards and saved. A cannabinoid PDA spectral library was compiled using available authentic standards for both acidic and neutral components. PDA spectra and retention times are recorded in the library. Libraries can be searched manually or automatically as part of the processing method.

When the cannabinoid library was searched with a retention time window of 5% set in the processing method, one library match was proposed for each of the unknown components at t_R 6.151 minutes and 6.530 minutes.

The library match results indicate the best possible match for the unknown component at 6.151 minutes to be (6aR,9S)-Δ¹⁰-THC (highlighted in Figure 7). Using the spectral contrast algorithm in Empower, the Match 1 Angle was observed to be less than the Match 1 Threshold.^9–13 The smaller the match angle, the better the match. In the Results window, the PDA Triple Plot, displays the spectrum from the unknown component along with the best possible spectral match, and the difference between them (lower spectrum, Figure 7). To aid visual inspection of the spectra, they can be superimposed to highlight any spectral differences. Figure 8 shows the superimposed spectral for the unknown component at 6.151 minutes and the library spectrum, indicating an observed close match.

The library search for the unknown at 6.530 minutes indicated the closest match to be (6aR,9R)-Δ¹⁰-THC. However, the Match 1 Angle was greater that the Match 1 Threshold, in this case. In addition, there was a purity flag which required further investigation.

Peak Purity

The Purity Plot can be used for evaluating co-eluting components.^14–15 Each spectrum across the integrated peak is evaluated. In the Purity Plot (Figure 9), the Purity 1 Angle is greater than the Purity 1 Threshold indicating the presence of an impurity. The vertical line annotated with the letter M indicates the point in the peak with the maximum value for the ratio of purity angle/purity threshold. In this case, the leading edge of the peak.

Comparison of the t_R and UV spectra between authentic standards of (6aR,9S)-Δ¹⁰-THC and (6aR,9R)-Δ¹⁰-THC and the unknown components in the distillate sample showed close similarities providing further confidence in the identifications. 

In-source Fragmentation

MS data adds confidence to component identification. Precursor and product ion data are frequently used together to further increase confidence in the identification of unknowns. In-source fragmentation experiments with a single quadrupole mass spectrometer can provide additional structural information. However, unlike tandem mass spectrometry, this mode of fragmentation is non-specific and can include contributions from multiple components in the event of compound co-elution.

The method can be configured to simultaneously acquire data using two MS scan functions at two different cone voltages: a low value (15 V) and a high value (45 V). The high cone voltage function induces fragmentation due to ion acceleration and collision with gaseous species including the desolvation gas and solvent vapor.

During the analysis of the distillate sample using the low cone voltage experiment, the precursor ion [M+H]⁺ m/z 315 for the peak identified as (6aR,9R)-Δ¹⁰-THC was observed. During the high cone voltage experiment, several product ions with m/z typical of Δ⁹-THC and its isomers were observed including m/z 259, 193, 135, and 123. There was good agreement between the fragmentation patterns observed for the high cone voltage experiment when an authentic standard of (6aR,9R)-Δ¹⁰-THC was compared with the peak in the sample at t_R at 6.530 minutes (Figure 10).

• Several unidentified components with a base peak of m/z 315 were observed eluting in the region between 4.0 minutes and 7.0 minutes in the distillate sample. HPLC-PDA-MS data suggest that they are possible structural isomers of Δ⁹-THC.
• Unknown components were tentatively identified using a cannabinoid PDA spectral library with compound assignment based on PDA spectral matching.
• The Purity Plot can be used for evaluating peak purity, indicating the presence of co-eluting components.
• Authentic standards of (6aR,9S)-Δ¹⁰-THC and (6aR,9R)-Δ¹⁰-THC were used to confirm the identifications based on t_R, precursor, and product ions.
• Multiple unknown components were detected in the distillate samples. Characterization of components with unknown biological activity or toxicological data is important to enhance understanding and ensure consumer safety.

1. Erickson, BE. (2021). C&En News. 99(31). "Delta-8-THC Craze Concerns Chemists - cen.acs.org." 30 Aug. 2021, https://cen.acs.org/biological-chemistry/natural-products/Delta-8-THC-craze-concerns-chemists/99/i31.
2. Hudalla, C. (2021). “We Believe in Unicorns (and Delta-8)”. The Cannabis Scientist. https://thecannabisscientist.com/testing-processing/we-believe-in-unicorns-and-delta-8.
   
3. Webster, GRB., Sarna, LP., Mechoulam, R. US Patent Application Publication. Pub. https://patents.google.com/patent/US20040143126A1/en.
   
4. Golombek, P., et al. (2020). "Conversion of Cannabidiol (CBD) into Psychotropic Cannabinoids Including Tetrahydrocannabinol (THC): A Controversy in the Scientific Literature." Toxics 8(2).
   
5. Kiselak, T. D., et al. (2020). "Synthetic Route Sourcing of Illicit At Home Cannabidiol (CBD) Isomerization to Psychoactive Cannabinoids Using Ion Mobility-Coupled-LC-MS/MS." Forensic Sci Int 308: 110173.
   
6. Meehan-Atrash, J., and I. Rahman (2022). "Novel Δ8-Tetrahydrocannabinol Vaporizers Contain Unlabeled Adulterants, Unintended Byproducts of Chemical Synthesis, and Heavy Metals." Chemical Research in Toxicology 35(1): 73–76.
   
7. Watanabe, K., et al. (2007). "Conversion of Cannabidiol to Δ9-Tetrahydrocannabinol and Related Cannabinoids in artificial gastric juice, and their pharmacological effects in mice." Forensic Toxicology 25(1): 16–21.
8. Helander, A., et al. (2022). "Analytical and medico-legal problems linked to the presence of delta-8-tetrahydrocannabinol (delta-8-THC): Results from urine drug testing in Sweden." Drug Testing and Analysis 14(2): 371–376.
   
9. Gorenstein, MV., Li, JB., Van Antwerp, J., Chapman, D., Detecting Coeluting Impurities by Spectral Comparison. https://preprod1.waters.com/webassets/other/SpectralContrastAlgorithm.PDF.
   
10. Lander. N., Get Empowered: Creating, Searching, and Restoring UV Libraries | Tip #144, Creating a UV Library - Waters.
    
11. Lander, N., Get Empowered: Creating, Searching, and Restoring UV Libraries | Tip #145, Manually Searching a UV Library - Waters.
    
12. Lander, N., Get Empowered: Creating, Searching, and Restoring UV Libraries | Tip #146, Performing an Automated Search in a UV Library - Waters.
    
13. Lander, N., Get Empowered: Creating, Searching, and Restoring UV Libraries | Tip #147, Back up and Restore PDA Libraries - Waters.
    
14. Lander, N., Get Empowered: Tip #251 | Peak Purity - Waters.
    
15. Lander, N., The Math Behind Peak Purity Calculations | Waters Blog.