A Preliminary FET Headspace GC-FID Method for Comprehensive Terpene Profiling in Cannabis
This application note describes an FET headspace GC-FID method that was developed in hops for the analysis of terpenes in cannabis. Good chromatographic separation allowed quantification of critical compounds across the volatility range, including α-pinene, β-myrcene, α-humulene, β-caryophyllene, and caryophyllene oxide.
In addition to cannabinoids, cannabis contains a suite of compounds known as terpenes. Terpenes are not only responsible for the characteristic aromas of cannabis strains, but they also are suspected to contribute to the therapeutic properties of cannabis. By themselves, terpenes have anti-inflammatory and anti-microbial properties, and they also reportedly contribute to an “entourage effect” with cannabinoids, modulating and/or enhancing their activity [1,2].
Because terpenes may contribute to the therapeutic effects of cannabis, there is a growing demand for analytical methods that profile terpenes in marijuana samples. In addition to analyzing terpenes for therapeutic purposes, terpenes can also be used as differentiators among cannabis strains and terpene profiles can be used for strain identification.
While relatively few terpenes have been studied for therapeutic purposes, cannabis strains can contain dozens of terpenes in varying levels. Of these, the primary compounds of interest include α-pinene, β-myrcene, α-humulene, and β-caryophyllene [2,3]. Accurately profiling these analytes and other emerging terpenes of interest depends heavily on separating them from potentially interfering compounds. When an interfering terpene, or other compound, coelutes with a terpene of interest, quantification will be compromised and, since many terpenes have the same molecular weight and share fragment ions, mass spectrometry cannot be relied upon to distinguish a terpene of interest from a coeluting interference terpene. The only way to accurately identify and quantify terpenes is to ensure that the terpenes of interest are chromatographically separated from all interfering compounds. GC is an excellent technique for accomplishing this.
Here we present a headspace gas chromatography–flame ionization detection (GC-FID) method for a comprehensive set of 38 terpenes found in cannabis. Since cannabis is illegal in Pennsylvania where this work was done, we developed the method using hops as a model system since they are related to cannabis and contain a similar suite of terpenes [2,3,4]. The headspace method presented here utilizes full evaporation technique (FET) sample preparation because cannabis product matrices are extremely varied and plant material will not dissolve in solvent. FET involves the use of a very small sample amount (10–50 mg), which effectively creates a single-phase gas system in the headspace vial at equilibrium, making it ideal for this application [5,6,7]. Figure 1 illustrates the basic principle of headspace gas chromatography using FET. To achieve chromatographic separation, a 30 m x 0.25 mm x 1.4 µm Rxi®-624Sil MS column was used. This column was chosen based on several factors. First, and most importantly, the cyano-based stationary phase of the Rxi®-624Sil MS has excellent selectivity for terpenes, making it ideal to effect a good separation for a large suite of these compounds. Second, in addition to its excellent selectivity for terpenes, the maximum temperature of this column is 320 °C, which allows for elution of some of the less-volatile terpenes and matrix compounds that may be present in the headspace sample. Third, this GC column phase is also well-suited for residual solvent analysis, potentially minimizing the number of columns and instruments required by labs to test cannabis.
Figure 1: Setup and Basic Principle of FET Headspace Injection Coupled With GC-FID Analysis
Pelletized hops from three strains (UK East Kent Golding, Citra, and Cascade) were purchased from HopUnion. The pelletized hops were first ground to a fine powder using an IKA® mill. Because the hops were already ground and pelletized, very little grinding was necessary. For cannabis plant material, it is recommended that samples be frozen prior to grinding or that grinding occur under liquid nitrogen. This keeps the samples cold during the grinding process, reducing loss of the more volatile terpenes such as α-pinene. 10 mg samples of each strain were then placed in headspace vials (Figure 2). An incubation temperature of 140 °C was used to ensure volatilization of all terpenes and terpenoids in the sample. This temperature was chosen because it is also sufficient to melt samples of cannabis concentrates. An incubation time of 30 minutes was used to ensure the establishment of equilibrium during incubation, which is required for reproducible, quantitative results.
Figure 2: Grinding samples maximizes and normalizes surface area from sample to sample, increasing sensitivity and reproducibility.
Gas Chromatographic Conditions
Samples were analyzed on an Agilent® 6890 gas chromatograph equipped with a Tekmar® HT-3 headspace autosampler. A 30 m x 0.25 mm x 1.4 µm Rxi®-624Sil MS column was installed based on its selectivity for terpenes and because it could also be used for analysis of residual solvents in cannabis concentrates. A 1 mm straight Restek Premium inlet liner was used to limit the volume in the GC inlet. For headspace instruments, reducing the inlet volume increases efficiency by reducing band broadening during sample introduction. Greater efficiency maximizes peak separation, which is essential for this analysis. Complete chromatographic conditions are presented in Figure 4.
To aid in peak identification, a multi-component terpene standard was prepared with each compound at approximately 0.02% wt/vol. 10 µL of this standard solution was injected into a capped headspace vial and analyzed by FET headspace GC-FID. Standards were analyzed under the same conditions as the samples in order to eliminate the potential for discrimination across the volatility range (e.g., more volatile terpenes may show higher responses than less volatile terpenes). Since any discrimination effect would be the same in both the sample and standard, analytes were quantified based on their relative response factor compared to the standard as shown in Equation 1. This normalizes the values between sample and standard, ensuring accurate quantification across the full range of volatility for terpenes. Note that while the relative response factor technique improves accuracy, the semi-quantitative preparation of the standard and lack of well-characterized certified reference materials for terpenes limits the overall quantitative accuracy that can be obtained for this analysis. Additionally, the lack of pure, neat standards available to prepare a more concentrated standard resulted in a standard well below the level of many of the terpenes detected in this work. For accurate quantification, a calibration curve encompassing the expected concentration range of all analytes is required. The data presented in this article should be considered semi-quantitative.
Equation 1: Sample Concentration Calculation.
Results and Discussion
The purpose of this study was to develop an FET headspace GC-FID method for the analysis of terpenes in cannabis using hops as a model system. The terpenes found in our samples matched well with literature descriptions of the terpenes present in hops . High levels of terpenes were found across the volatility range, indicating that the FET headspace GC-FID technique was appropriate and that analysis of the standard adequately normalized any discrimination between the more and less volatile terpenes (Figure 3). Due to the starting concentration of some of the commercially available terpene standards, the maximum concentration at which the mixed terpene standard used for quantification could be prepared was 0.02% wt/vol, which is significantly lower than the concentration of some of the more prevalent terpenes in hops and cannabis. The use of a more concentrated standard solution is recommended to improve quantification of the higher concentrations found in these samples.
Figure 3: Terpene Profiles of Pelletized Hops
Figures 4–7 show individual chromatograms for the standard and each sample profiled for terpenes. Note that α-pinene, β-myrcene, α-humulene, β-caryophyllene, and caryophyllene oxide are well separated from interferences. For complex matrices, such as hops and marijuana, excellent chromatographic efficiency and selectivity are required to separate terpenes from one another and from other volatile matrix components in order to obtain accurate quantification. The selectivity of the Rxi®-624Sil MS column used here provided good separation of most terpenes and the small bore configuration (0.25 mm internal diameter) improved column efficiency, ultimately resulting in greater resolution between closely eluting terpenes than would be obtained using a wider bore column.
Figure 4: A 0.02% wt/vol multi-component terpenes standard analyzed on an Rxi®-624Sil MS column (30 m x 0.25 mm x 1.4 µm) demonstrates that this column provides the selectivity and efficiency needed to separate key terpenes using a simple FET headspace GC-FID method. (View larger)