Abstract
Background:
Cannabis has a long history of being credited with centuries of healing powers for millennia. The cannabis plant is a rich source of cannabinoids and terpenes. Each cannabis chemovar exhibits a different flavor and aroma, which are determined by its terpene content.
Methods:
In this study, a gas chromatography-flame ionization detector method was developed and validated for the determination of the 10 major terpenes in the main three chemovars of Cannabis sativa L. with n-tridecane used as the internal standard following the standard addition method. The 10 major terpenes (monoterpenes and sesquiterpenes) are α-pinene, β-pinene, β-myrcene, limonene, terpinolene, linalool, α-terpineol, β-caryophyllene, α-humulene, and caryophyllene oxide. The method was validated according to Association of Official Analytical Chemists guidelines. Spike recovery studies for all terpenes were carried out on placebo cannabis material and indoor-growing high THC chemovar with authentic standards.
Results:
The method was linear over the calibration range of 1–100 μg/mL with r2>0.99 for all terpenes. The limit of detection and limit of quantification were calculated to be 0.3 and 1.0 μg/mL, respectively, for all terpenes. The accuracy (%recovery) at all levels ranged from 89% to 104% and 90% to 111% for placebo and indoor-growing high THC chemovar, respectively. The repeatability and intermediate precision of the method were evaluated by the quantification of target terpenes in the three different C. sativa chemovars, resulting in acceptable relative standard deviations (less than 10%).
Conclusions:
The developed method is simple, sensitive, reproducible, and suitable for the detection and quantification of monoterpenes and sesquiterpenes in C. sativa biomass.
Keywords: Cannabis sativa, Terpenes, GC-FID, Validation, Quantitative Determination
Introduction
Cannabis sativa L. is an aromatic medicinal plant belonging to the family Cannabaceae. It is considered a highly promising plant for the development and discovery of new medicinal agents due to its reported medical efficacy in the treatment of several disorders, such as glaucoma, pain, epilepsy, and cancer.1–3 The chemical diversity of constituents of the cannabis plant can be very important in determining its utility, potency, and medicinal effectiveness.4,5
Various components belonging to a variety of chemical classes were found in cannabis samples.6,7 Cannabis contains a unique class of compounds, cannabinoids, but the terpenes, which are non-cannabinoid constituents, play a notable role in the cannabis culture.8 Terpenes are responsible for adding flavor, scent, and other qualities to cannabis and are believed to interact pharmacologically with cannabinoids.9–11 Both cannabinoids and terpenes are produced in the glandular trichomes of the plant.12,13 Major cannabinoids, as well as terpenes, are essential for the determination and differentiation between C. sativa chemovars.3,14,15
Several methods were developed for the quantification of terpenes in cannabis, using gas chromatography methods, including headspace (HS) gas chromatography-flame ionization detector (GC-FID), HS GC-mass spectrometry (MS), two dimensional (GC×GC-qMS), and HS solid-phase microextraction.16–24
Recently, we developed and validated a GC-MS method for the determination of major cannabis terpenes.23
In this study, a GC-FID method was developed, optimized, and validated according to Association of Official Analytical Chemists guidelines for the quantification of terpenes in C. sativa.25
Materials and Methods
Standards and reagents
All reference standards were purchased from Sigma-Aldrich: α-pinene (purity ≥98%), (−)-β-pinene (purity ≥99%), β-myrcene (purity ≥95%), (R)−(+)-limonene (purity ≥97%), terpinolene (purity ≥85%), linalool (purity ≥97%), terpineol (purity ≥90%), β-caryophyllene (purity ≥80%), α-humulene (purity ≥96%), caryophyllene oxide (purity ≥99%), and n-tridecane (purity ≥99%). Their purities were confirmed by GC-MS before the quantification. The chemical structures of the major terpenes are represented in Figure 1.
FIG. 1.
Chemical structures of selected terpenes.
Stock standard solutions
A stock standard solution of each terpene (α-pinene, β-pinene, β-myrcene, limonene, terpinolene, linalool, α-terpineol, β-caryophyllene, α-humulene, and caryophyllene oxide) was prepared in ethyl acetate. The standard terpenes were mixed and the concentration of each terpene was adjusted to be 1.0 mg/mL in the mixture (A). From the stock solution (A), serial dilutions were made to prepare 100 μg/mL (B), 10 μg/mL (C), and 1.0 μg/mL (D) stock standard solutions. These solutions were used to prepare the individual points of calibration curves.
Internal standard preparation
n-Tridecane (C13 hydrocarbon) was selected as the internal standard (IS), and a constant concentration of 100 μg/mL was added to all the calibration standards and sample solutions.
Calibration curves
Six calibration points ranging from 1 to 100 μg/mL were prepared from the previously mentioned stock standard solutions (1, 5, 10, 25, 50, and 100 μg/mL) with the concentration of the IS in each calibration point being 100 μg/mL. These solutions were used to construct individual calibration curves.
Cannabis plant material
Three C. sativa chemovars (high THC, high CBD, and intermediate) were cultivated and grown at the University of Mississippi (indoor facility). Flowering tops were harvested and air-dried for 24 h at a controlled temperature of 50°C in a ventilated oven. All samples were kept in a freezer until analyzed. Confiscated plant samples with different cannabinoid profiles seized by Drug Enforcement Administration (DEA) were submitted to our laboratory for cannabinoid potency monitoring.26 These samples were used in this study for the determination of terpene composition.
Quality control samples
Samples from three chemovars of C. sativa (high THC chemovar, intermediate chemovar, and high CBD chemovar) were dried for 24 h at 50°C in a ventilated oven and then ground in a stainless-steel coffee grinder. Triplicate of powdered samples (1.0 g each) was weighed in a 15-mL centrifuge tube and extracted with 10 mL of the extraction solution (100 μg/mL of the IS solution in ethyl acetate) by sonication for 15 min. The mixture was centrifuged at 3000 rpm for 5 min and the clear supernatants (without filtration) were used for analysis. An aliquot of 2 μL of each sample solution was then injected into the GC-FID.
Samples
For extraction, 1 g sample was weighed in a 15-mL centrifuge tube. A 10 mL of the extraction solution was added to the sample, and then the sample was mixed on a vortex mixer, sonicated for 15 min, and centrifuged at 3000 rpm for 5 min. A part of the clear supernatant was transferred to a GC vial and a 2 μL aliquot was injected into the GC-FID.
GC-FID instrumentation and conditions
GC-FID analysis was performed on an Agilent 7890B GC system fitted with an autosampler 7693. Separation was performed using a DB5-MS (30 m×0.25 mm internal diameter, 0.25-μm film thickness; J&W Scientific, Inc., Agilent Technologies) column. Helium was used as the carrier gas at a flow rate of 1.2 mL/min and the FID make-up gas. The inlet was configured in split mode with a 15:1 split ratio and a temperature of 250°C. The oven time program began at 70°C for 2 min before ramping at a rate of 3°C/min to 85°C.
The oven temperature was increased at a rate of 2°C/min to 165°C and held for 1 min before ramping at a rate of 30°C/min to 250°C, where it was held for 20 min. The total run time was ∼60 min. The detector temperature was set at 300°C and the hydrogen, air, and make-up flow rates were 40, 500, and 27 mL/min, respectively. Data analysis was performed using Agilent ChemStation® software (rev. B.04.02). The injection volume was 2 μL. All terpenes were recognized in samples by comparing their retention times with authentic references.
Identification of peaks
Identification of all constituents was performed by comparing the retention times of the peaks in the samples with those of authentic reference standards.
Results and Discussion
Extraction optimization and IS
The extraction procedure was optimized regarding the extraction solvent and recovery. Before method validation, different extraction solvents such as methanol, ethanol, ethyl acetate, and hexanes were tried. It was found that ethyl acetate was the optimum extraction solvent for the target terpenes from C. sativa L. samples.
The IS method with a six-point calibration curve was applied to quantify each terpene, using the regression equation of each curve, with each sample determined in triplicate.
Internal standard
Utilizing n-tridecane (C13 hydrocarbon) was found experimentally to be the ideal IS, with a retention time falling between the monoterpenes and sesquiterpenes, and it is not present in C. sativa plant material (Fig. 2).
FIG. 2.
GC-FID chromatogram of standard terpene solution at 100 μg/mL. GC-FID, gas chromatography-flame ionization detector.
Method validation
The method was validated according to AOAC guidelines concerning linearity, accuracy (recovery), selectivity, repeatability, intermediate precision, the limit of detection (LOD), and the limit of quantification (LOQ).25
Linearity
Six-point standard calibration curves were used to evaluate linearity. Calibration curves were determined by plotting the peak area ratio (y; the peak area of each terpene to the peak area of IS) versus the terpene concentration (x) (Fig. 3). The concentration-peak area ratio relationship was used to construct the calibration curve, where it was found to be linear over the range of 1–100 μg/mL for all terpenes. Regression equation parameters and the calibration curves of the selected terpenes are shown in Table 1 and Figure 3, respectively.
FIG. 3.
Calibration curves of selected terpenes.
Table 1.
Regression Equation Parameters of the Selected Terpenes
| Parameters | α-Pinene | β-Pinene | β-Myrcene | Limonene | Terpinolene | Linalool | α-Terpineol | β-Caryophyllene | α-Humulene | Caryophyllene oxide |
|---|---|---|---|---|---|---|---|---|---|---|
| Retention time (min) | 10.173 | 12.002 | 12.270 | 14.191 | 16.801 | 17.346 | 21.967 | 32.103 | 33.573 | 38.238 |
| Calibration range (μg/mL) | 1–100 | 1–00 | 1–100 | 1–100 | 1–100 | 1–100 | 1–100 | 1–100 | 1–100 | 1–100 |
| LOD (μg/mL) | 0.300 | 0.300 | 0.300 | 0.300 | 0.300 | 0.300 | 0.300 | 0.300 | 0.300 | 0.300 |
| LOQ (μg/mL) | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 |
| Regression equation (Y)a: | y=0.0100x−0.0041 | y=0.0100x−0.0042 | y=0.0092x−0.0009 | y=0.0103x−0.0023 | y=0.0092x−0.0138 | y=0.0087x – 0.0012 | y=0.0083x – 0.001 | y=0.0106x+0.0008 | y=0.0106x – 0.0018 | y=0.0096x+0.0014 |
| SD of the slope (Sb) | 4.96×10−4 | 4.22×10−04 | 2.03×10−04 | 3.40×10−04 | 7.29×10−04 | 1.01×10−04 | 8.75×10−05 | 9.20×10−04 | 4.70×10−04 | 1.77×10−04 |
| RSD% of the slope | 4.96 | 4.24 | 2.21 | 3.30 | 7.92 | 1.17 | 1.06 | 9.20 | 4.70 | 1.84 |
| Confidence limit of the slopeb | (9.42×10−03)–(1.05×10−02) | (9.48×10−03)–(1.04×10−02) | (8.97×10−03)–(9.43×10−03) | (9.89×10−03)–(1.07×10−02) | (8.37×10−03)–(1×10−02) | (8.54×10−03)–(8.76×10−03) | (8.18×10−03)–(8.38×10−03) | (1.05×10−02)–(1.07×10−02) | (1.06×10−02)–(1.07×10−02) | (9.41×10−03)–(9.81×10−03) |
| Intercept (a) | −0.0041 | −0.0042 | −0.0009 | −0.0023 | −0.0138 | −0.0012 | −0.001 | 0.0008 | −0.0018 | 0.0014 |
| SD of the intercept (Sa) | 2.33×10−02 | 1.98×10−02 | 9.55×10−03 | 1.60×10−02 | 3.42×10−02 | 4.74×10−03 | 4.11×10−03 | 4.32×10−03 | 2.21×10−03 | 8.30×10−03 |
| Confidence limit of the interceptb | (−3.05×10−02)–(2.23×10−02) | (−2.67×10−02)–(1.83×10−02) | (−1.18×10−02)–(9.88×10−03) | (−2.04×10−02) –(1.58×10−02) | (−5.26×10−02)–(2.50×10−02) | (−6.61×10−03)–(4.14×10−03) | (−5.62×10−03)–(3.70×10−04) | (−4.09×10−03)–(5.71×10–03) | (−4.27×10−03)–(7.41×10−04) | −(7.96×10−03)–(1.09×10−02) |
| Correlation coefficient (r2) | 0.9985 | 0.9988 | 0.9997 | 0.9993 | 0.9958 | 0.9999 | 0.9999 | 0.9999 | 0.9999 | 0.9998 |
y=bx + a (y=area ratio, b=the slope, x=concentration, and a=the intercept).
LOD, limit of detection; LOQ, limit of quantification; RSD%, percentage relative standard deviation; SD, standard deviation.
LOD and LOQ
LOD and LOQ are expressed as LOD=3.3σ/S and LOQ=10σ/S, where σ=standard deviation of the response of each terpene and S=slope of the calibration curve of each terpene. The LOQ and LOD for each terpene were determined to be 0.3 and 1 μg/mL, respectively (Table 1).
Repeatability and intermediate precision
The method precision was evaluated by measuring the quantification of individual terpenes in three different C. sativa chemovars. The analysis of samples was made in six replicates on 3 separate days. The intraday and interday precision (repeatability and intermediate precision) were determined in terms of percentage relative standard deviation (RSD%) (Tables 2–6).
Table 2.
Intraday Precision Parameters for High CBD Chemovar (Sample No. 1371)
| Compound | Day 1 (n=6) |
Day 2 (n=6) |
Day 3 (n=6) |
||||||
|---|---|---|---|---|---|---|---|---|---|
| w/w (mg/g) (mean, n=6) | SD | Precision (RSD%) | w/w (mg/g) (mean, n=6) | SD | Precision (RSD%) | w/w (mg/g) (mean, n=6) | SD | Precision (RSD%) | |
| α-Pinene | 0.0241 | 0.0008 | 3.1975 | 0.0227 | 0.0011 | 5.0357 | 0.0233 | 0.0008 | 3.3783 |
| β-Pinene | ND | ND | ND | ND | ND | ND | ND | ND | ND |
| β-Myrcene | 0.015 | 0.001 | 4.779 | 0.016 | 0.001 | 7.449 | 0.015 | 0.001 | 4.205 |
| Limonene | 0.020 | 0.002 | 8.874 | 0.018 | 0.001 | 4.876 | 0.016 | 0.001 | 5.242 |
| Terpinolene | ND | ND | ND | ND | ND | ND | ND | ND | ND |
| Linalool | 0.023 | 0.002 | 9.887 | 0.022 | 0.002 | 9.336 | 0.020 | 0.001 | 4.810 |
| α-Terpineol | 0.056 | 0.002 | 3.751 | 0.054 | 0.003 | 6.095 | 0.050 | 0.001 | 2.261 |
| β-Caryophyllene | 0.329 | 0.005 | 1.522 | 0.339 | 0.003 | 1.026 | 0.333 | 0.004 | 1.080 |
| α-Humulene | 0.111 | 0.003 | 2.747 | 0.116 | 0.004 | 3.216 | 0.111 | 0.001 | 1.092 |
| Caryophyllene oxide | 0.241 | 0.006 | 2.653 | 0.227 | 0.006 | 2.759 | 0.237 | 0.002 | 0.946 |
n, number of replicates; ND, not detected.
Table 3.
Interday Precision Parameters for High CBD Chemovar (Sample No. 1371)
| Compound | w/w (mg/g) (n=18) | SD | Precision (RSD%) |
|---|---|---|---|
| α-Pinene | 0.023 | 0.0007 | 3.182 |
| β-Pinene | ND | ND | ND |
| β-Myrcene | 0.015 | 0.0003 | 2.264 |
| Limonene | 0.018 | 0.002 | 9.888 |
| Terpinolene | ND | ND | ND |
| Linalool | 0.022 | 0.001 | 6.211 |
| α-Terpineol | 0.053 | 0.003 | 6.102 |
| β-Caryophyllene | 0.334 | 0.005 | 1.462 |
| α-Humulene | 0.112 | 0.003 | 2.440 |
| Caryophyllene oxide | 0.235 | 0.007 | 2.954 |
Table 4.
Intraday Precision Parameters for High THC Chemovar (Sample No. 1440)
| Compound | Day 1 (n=6) |
Day 2 (n=6) |
Day 3 (n=6) |
||||||
|---|---|---|---|---|---|---|---|---|---|
| w/w (mg/g) (mean, n=6) | SD | Precision (RSD%) | w/w (mg/g) (mean, n=6) | SD | Precision (RSD%) | w/w (mg/g) (mean, n=6) | SD | Precision (RSD%) | |
| α-Pinene | 0.0328 | 0.0009 | 2.8794 | 0.0323 | 0.0015 | 4.5034 | 0.0300 | 0.0010 | 3.4863 |
| β-Pinene | 0.019 | 0.001 | 4.266 | 0.019 | 0.001 | 6.562 | 0.018 | 0.001 | 4.628 |
| β-Myrcene | 0.0122 | 0.0001 | 1.0719 | 0.0104 | 0.0008 | 7.2759 | 0.0110 | 0.0004 | 3.5414 |
| Limonene | ND | ND | ND | ND | ND | ND | ND | ND | ND |
| Terpinolene | ND | ND | ND | ND | ND | ND | ND | ND | ND |
| Linalool | 0.025 | 0.001 | 5.195 | 0.030 | 0.001 | 4.224 | 0.027 | 0.002 | 6.379 |
| α-Terpineol | 0.035 | 0.002 | 5.477 | 0.034 | 0.002 | 5.478 | 0.034 | 0.002 | 5.323 |
| β-Caryophyllene | 0.63 | 0.01 | 0.85 | 0.69 | 0.01 | 0.92 | 0.65 | 0.01 | 1.61 |
| α-Humulene | 0.191 | 0.002 | 0.834 | 0.215 | 0.003 | 1.306 | 0.199 | 0.011 | 5.290 |
| Caryophyllene oxide | 0.213 | 0.003 | 1.486 | 0.221 | 0.005 | 2.351 | 0.218 | 0.005 | 2.379 |
Table 5.
Interday Precision Parameters High THC Chemovar (Sample No. 1440)
| Compound | w/w (mg/g) (n=18) | SD | Precision (RSD%) |
|---|---|---|---|
| α-Pinene | 0.0317 | 0.0015 | 4.6507 |
| β-Pinene | 0.019 | 0.001 | 2.822 |
| β-Myrcene | 0.011 | 0.001 | 8.302 |
| Limonene | ND | ND | ND |
| Terpinolene | ND | ND | ND |
| Linalool | 0.027 | 0.002 | 9.209 |
| α-Terpineol | 0.034 | 0.001 | 2.272 |
| β-Caryophyllene | 0.66 | 0.03 | 4.68 |
| α-Humulene | 0.202 | 0.012 | 6.189 |
| Caryophyllene oxide | 0.217 | 0.004 | 1.993 |
Table 6.
Intraday Precision Parameters for Intermediate Chemovar (Sample No. 1363)
| Compound | Day 1 (n=6) |
Day 2 (n=6) |
Day 3 (n=6) |
||||||
|---|---|---|---|---|---|---|---|---|---|
| w/w (mg/g) (mean, n=6) | SD | Precision (RSD%) | w/w (mg/g) (mean, n=6) | SD | Precision (RSD%) | w/w (mg/g) (mean, n=6) | SD | Precision (RSD%) | |
| α-Pinene | 0.195 | 0.007 | 3.752 | 0.172 | 0.005 | 2.685 | 0.19 | 0.01 | 3.77 |
| β-Pinene | 0.051 | 0.002 | 2.996 | 0.046 | 0.001 | 1.722 | 0.049 | 0.002 | 3.238 |
| β-Myrcene | 0.0102 | 0.0003 | 2.8948 | 0.0106 | 0.0007 | 6.1295 | 0.0109 | 0.0006 | 5.6851 |
| Limonene | 0.0640 | 0.0024 | 3.6901 | 0.0594 | 0.0021 | 3.5278 | 0.0619 | 0.0014 | 2.3109 |
| Terpinolene | ND | ND | ND | ND | ND | ND | ND | ND | ND |
| Linalool | 0.0399 | 0.0017 | 4.1960 | 0.0383 | 0.0002 | 0.5632 | 0.0394 | 0.0009 | 2.3984 |
| α-Terpineol | 0.087 | 0.004 | 4.742 | 0.076 | 0.003 | 3.547 | 0.084 | 0.002 | 2.942 |
| β-Caryophyllene | 0.67 | 0.02 | 2.76 | 0.64 | 0.02 | 2.55 | 0.70 | 0.01 | 2.08 |
| α-Humulene | 0.25 | 0.01 | 4.47 | 0.25 | 0.01 | 3.19 | 0.270 | 0.004 | 1.657 |
| Caryophyllene oxide | 0.33 | 0.01 | 2.59 | 0.30 | 0.01 | 4.36 | 0.33 | 0.01 | 2.68 |
In the high CBD chemovar, the measured repeatability ranged from 0.94% to 9.88% and the intermediate precision ranged from 1.46% to 9.88% (Tables 2 and 3) for all terpenes. The high THC chemovar showed 0.83–7.27% and 1.99–9.20% for repeatability and intermediate precision, respectively (Tables 4 and 5). For the intermediate chemovar, the measured values ranged from 0.56% to 6.12% for repeatability, and 2.14% to 6.48% for intermediate precision, as shown in Tables 6 and 7. In all cases, precision was found to be less than 10%, indicating that the method is precise.
Table 7.
Interday Precision Parameters for Intermediate Chemovar (Sample No. 1363)
| Compound | w/w (mg/g) (n=18) | SD | Precision (RSD%) |
|---|---|---|---|
| β-Pinene | 0.19 | 0.01 | 6.48 |
| β-Pinene | 0.05 | 0.003 | 5.85 |
| β-Myrcene | 0.0106 | 0.0004 | 3.3186 |
| Limonene | 0.0617 | 0.0023 | 3.7598 |
| Terpinolene | ND | ND | ND |
| Linalool | 0.0392 | 0.0008 | 2.1431 |
| α-Terpineol | 0.08 | 0.01 | 7.03 |
| β-Caryophyllene | 0.67 | 0.03 | 4.05 |
| α-Humulene | 0.26 | 0.01 | 4.36 |
| Caryophyllene oxide | 0.32 | 0.02 | 5.59 |
Accuracy (recovery)
To determine the accuracy (recovery) of terpenes, triplicates of the stock standard solution of each terpene were spiked to 1 g of plant material at three different concentration levels: low, medium, and high (0.05, 0.25, and 0.50 mg/g, respectively). Two plant materials were used: indoor-growing high THC chemovar and placebo (cannabis biomass free from terpenes, obtained by exhaustive solvent extraction with hexane followed by ethanol). Each plant material was analyzed before and after spiking according to the above sample preparation method.
The %recovery (accuracy) of each terpene is calculated as follows: (amount after spiking − amount before spiking)/(spiked amount)×100%. The average recoveries of all terpenes were measured using placebo cannabis samples after spiking and found to be 103% for α-pinene, 104% for β-pinene, 101% for β-myrcene, 97% for limonene, 94% for linalool, 103% α-terpineol, 89% for β-caryophyllene, 92% α-humulene, and 90% for caryophyllene oxide (Table 8).
Table 8.
Recovery (Accuracy of Selected Terpenes) in Placebo Cannabis Sample (n=6)
| Compound | Spiked amount w/w (mg/g) | Found amount w/w (mg/g) | SD | %Recovery |
|---|---|---|---|---|
| α-Pinene | 0.05 | 0.052 | 0.0046 | 103% |
| 0.25 | 0.259 | 0.018 | 104% | |
| 0.50 | 0.512 | 0.032 | 102% | |
| β-Pinene | 0.05 | 0.052 | 0.0050 | 104% |
| 0.25 | 0.260 | 0.018 | 104% | |
| 0.50 | 0.517 | 0.034 | 103% | |
| β-Myrcene | 0.05 | 0.050 | 0.0036 | 100% |
| 0.25 | 0.254 | 0.014 | 102% | |
| 0.50 | 0.510 | 0.033 | 102% | |
| Limonene | 0.05 | 0.050 | 0.0065 | 100% |
| 0.25 | 0.239 | 0.019 | 96% | |
| 0.50 | 0.476 | 0.031 | 95% | |
| Terpinolene | 0.05 | 0.043 | 0.0046 | 85% |
| 0.25 | 0.186 | 0.016 | 74% | |
| 0.50 | 0.370 | 0.032 | 74% | |
| Linalool | 0.05 | 0.047 | 0.0037 | 94% |
| 0.25 | 0.236 | 0.012 | 94% | |
| 0.50 | 0.470 | 0.028 | 94% | |
| α-Terpineol | 0.05 | 0.051 | 0.0039 | 103% |
| 0.25 | 0.255 | 0.016 | 102% | |
| 0.50 | 0.520 | 0.027 | 104% | |
| β-Caryophyllene | 0.05 | 0.044 | 0.0066 | 89% |
| 0.25 | 0.215 | 0.017 | 93% | |
| 0.50 | 0.419 | 0.028 | 84% | |
| α-Humulene | 0.05 | 0.046 | 0.0051 | 91% |
| 0.25 | 0.232 | 0.017 | 92% | |
| 0.50 | 0.460 | 0.033 | 92% | |
| Caryophyllene oxide | 0.05 | 0.046 | 0.0052 | 92% |
| 0.25 | 0.224 | 0.014 | 90% | |
| 0.50 | 0.434 | 0.028 | 87% |
The recoveries of target terpenes were also measured using indoor-growing plant material after spiking with known standard amounts at three control levels: low, medium, and high. The average recoveries were found to be 110% for α-pinene, 107% for β-pinene, 99% for β-myrcene, 94% for limonene, 90% for terpinolene (it was 78% in placebo samples), 103% for linalool, 111% α-terpineol, 107% for β-caryophyllene, 102% for α-humulene, and 96% for caryophyllene oxide (Table 9).
Table 9.
Recovery (Accuracy of Selected Terpenes) Using Indoor-Growing High THC Chemovar (Sample No. 1441)
| Compound | Analysis w/w (mg/g) (n=3) | Spiked amount w/w (mg/g) (n=6) | Found amount w/w (mg/g) | SD | %Recovery | RE% |
|---|---|---|---|---|---|---|
| α-Pinene | 0.398±0.019 | 0.050 | 0.456 | 0.025 | 117% | 17% |
| 0.250 | 0.660 | 0.023 | 105% | 5% | ||
| 0.500 | 0.941 | 0.022 | 109% | 9% | ||
| β-Pinene | 0.221±0.013 | 0.050 | 0.278 | 0.016 | 114% | 14% |
| 0.250 | 0.479 | 0.018 | 103% | 3% | ||
| 0.500 | 0.744 | 0.018 | 105% | 5% | ||
| β-Myrcene | 0.515±0.012 | 0.050 | 0.564 | 0.049 | 98% | −2% |
| 0.250 | 0.744 | 0.011 | 92% | −8% | ||
| 0.500 | 1.045 | 0.068 | 106% | −6% | ||
| Limonene | 0.057±0.001 | 0.050 | 0.107 | 0.010 | 99% | −1% |
| 0.250 | 0.286 | 0.017 | 92% | −8% | ||
| 0.500 | 0.508 | 0.013 | 90% | −10% | ||
| Terpinolene | 0.349±0.013 | 0.050 | 0.395 | 0.031 | 92% | −8% |
| 0.250 | 0.565 | 0.003 | 86% | −14% | ||
| 0.500 | 0.806 | 0.037 | 91% | −9% | ||
| Linalool | 0.094±0.005 | 0.050 | 0.148 | 0.009 | 109% | 9% |
| 0.250 | 0.346 | 0.018 | 101% | 1% | ||
| 0.500 | 0.588 | 0.008 | 99% | −1% | ||
| α-Terpineol | 0.050±0.001 | 0.050 | 0.105 | 0.004 | 110% | 10% |
| 0.250 | 0.336 | 0.025 | 115% | 15% | ||
| 0.500 | 0.585 | 0.011 | 107% | 7% | ||
| β-Caryophyllene | 1.007±0.099 | 0.050 | 1.056 | 0.065 | 97% | −3% |
| 0.250 | 1.283 | 0.073 | 110% | 10% | ||
| 0.500 | 1.585 | 0.055 | 115% | 15% | ||
| α-Humulene | 0.282±0.023 | 0.050 | 0.331 | 0.017 | 98% | −2% |
| 0.250 | 0.544 | 0.032 | 105% | 5% | ||
| 0.500 | 0.800 | 0.010 | 104% | 4% | ||
| Caryophyllene oxide | 0.109±0.006 | 0.050 | 0.161 | 0.008 | 103% | 3% |
| 0.250 | 0.342 | 0.015 | 93% | –7% | ||
| 0.500 | 0.565 | 0.006 | 91% | –9% |
RE%=relative recovery percentage.
Carryover
To determine the degree of carryover, one ethyl acetate blank was injected between each calibration standard or sample extract. The blank did not show peaks for the analytes or the IS at signal–to–noise ratio of ≥3.
Selectivity
The resolution of terpene peaks in the GC chromatogram represents the selectivity. Based on this method, the peaks of the analyzed terpenes showed good resolution (Rs >1.9) as presented in (Fig. 2).
Repeatability and intermediate precision
The precision was evaluated by the quantification of individual terpenes in three different C. sativa chemovars. The analysis of samples was made in six replicates on 3 separate days. The intraday and interday precision (repeatability and intermediate precision) were determined in terms of RSD% (Tables 4–6).
In the high CBD chemovar, the measured repeatability ranged from 0.94% to 9.88% and the intermediate precision ranged from 1.46% to 9.88% (Tables 2 and 3) for all terpenes. The high THC chemovar showed 0.83–7.27% and 1.99–9.20% for repeatability and intermediate precision, respectively (Tables 4 and 5). For the intermediate chemovar, the measured values ranged from 0.56% to 6.12% for repeatability and 2.14% to 6.48% for intermediate precision, as shown in (Tables 6 and 7). In all cases, precision was found to be less than 10%, indicating that the method is precise.
GC-FID analysis of confiscated samples
This method was applied to the analysis of plant samples submitted to our laboratory for cannabinoid potency monitoring by the DEA.26 The IS method with a six-point calibration curve was applied to quantify each terpene and each sample was analyzed in triplicate.
Representative chromatograms of this method are given in Figures 2, 4, and 5 for a standard mixture of target terpenes and an ethyl acetate extract, respectively. The concentration of each terpene was expressed as mg/g of plant material. The sesquiterpenes were the most predominant group in all samples with β-caryophyllene and α-humulene being the main representatives. The samples showed large qualitative and quantitative variation that could be attributed to different factors such as the part of the plant (buds vs. leaves), the time of sample collection, and possibly the sample's origin. The quantitative data are reported in Table 10. The low levels of monoterpenes indicate that the drying of the plant material plays an important role in determining the terpene content of cannabis samples.27
FIG. 4.
GC-FID chromatogram of indoor-growing high THC chemovar plant material sample.
Table 10.
Average Terpene Concentration in Confiscated Samples Using the Proposed Gas Chromatography-Flame Ionization Detector Method (mg/g)
| Sample code | α-Pinene | β-Pinene | β-Myrcene | Limonene | Terpinolene | Linalool | α-Terpineol | α-Humulene | β-Caryophyllene | Caryophyllene oxide |
|---|---|---|---|---|---|---|---|---|---|---|
| 78606 | 0.015 | 0.023 | 0.016 | ND | 0.067 | 0.025 | 0.272 | 0.301 | 0.115 | 0.240 |
| 78607 | 0.156 | 0.067 | 0.040 | ND | 0.004 | 0.424 | 0.203 | 0.439 | 0.162 | 0.371 |
| 78612 | 0.015 | 0.042 | 0.013 | 0.459 | 0.019 | 0.725 | 0.262 | 1.481 | 0.597 | 1.113 |
| 78614 | 0.007 | 0.024 | 0.026 | ND | 0.000 | 0.653 | 0.155 | 0.710 | 0.339 | 0.873 |
| 78622 | 0.019 | 0.035 | 0.127 | 0.261 | 0.017 | 0.954 | 0.401 | 0.383 | 0.179 | 0.397 |
| 78777 | 0.018 | 0.013 | 0.034 | 0.050 | ND | 0.053 | 0.039 | 0.753 | 0.241 | 0.244 |
| 78786 | 0.053 | 0.019 | 0.047 | ND | ND | 0.097 | 0.014 | 0.307 | 0.089 | 0.141 |
| 78850 | 0.029 | 0.021 | 0.011 | 0.017 | ND | 0.077 | 0.121 | 0.222 | 0.067 | 0.306 |
| 78868 | 0.094 | 0.039 | 0.011 | 0.014 | 0.013 | 0.036 | 0.070 | 0.495 | 0.134 | 0.281 |
| 78889 | 0.027 | 0.048 | 0.011 | 0.011 | 0.013 | 0.048 | 0.082 | 0.396 | 0.142 | 0.592 |
| 78908 | 0.013 | 0.022 | 0.010 | ND | ND | 0.022 | 0.034 | 0.300 | 0.124 | 0.451 |
| 79037 | 0.301 | 0.125 | 0.053 | 0.058 | ND | 0.051 | 0.047 | 0.558 | 0.078 | 0.340 |
| 79118 | 0.031 | 0.099 | 0.032 | 0.195 | 0.016 | 0.361 | 0.211 | 0.813 | 0.311 | 0.715 |
| 79162 | 0.198 | 0.119 | 0.030 | 0.053 | 0.024 | 0.230 | 0.214 | 0.626 | 0.237 | 0.515 |
| 79672 | 0.024 | 0.026 | 0.012 | 0.010 | 0.026 | 0.164 | 0.190 | 0.112 | 0.056 | 0.206 |
| 79803 | 0.086 | 0.119 | 0.057 | 0.102 | 0.021 | 0.061 | 0.207 | 0.308 | 0.092 | 0.144 |
| 80132 | 0.049 | 0.041 | 0.042 | 0.097 | 0.015 | 0.164 | 0.244 | 0.426 | 0.123 | 0.347 |
FIG. 5.
GC-FID chromatogram of a representative confiscated sample.
Conclusions
An accurate GC-FID method was developed and validated for the simultaneous identification and quantification of the main terpenes in biomass of different cannabis chemovar samples. Both precision and recovery of the method were found to be acceptable for all the terpenes analyzed. The proposed method is used for routine analysis of cannabis samples in our laboratories as it is accurate, reliable, economical, and simple.
In light of differences observed in the terpene content of different cannabis samples, it is recommended that the characterization of the chemical profile of cannabis products be based not only on the cannabinoids content but also on the terpene profile.
Acknowledgment
E.A.I. would like to acknowledge the Egyptian government for the scholarship supervision program no. JS-3576.
Abbreviations Used
- AOAC
Association of Official Analytical Chemists
- CBD
cannabidiol
- DEA
Drug Enforcement Administration
- GC-FID
gas chromatography-flame ionization detector
- HS
headspace
- IS
internal standard
- LOD
limit of detection
- LOQ
limit of quantification
- MS
mass spectrometry
- RSD
relative standard deviation
- RSD%
percentage relative standard deviation
- THC
tetrahydrocannabinol
Authors' Contributions
E.A.I., M.M.R., A.S.W. and M.A.E. conceived the experiments; E.A.I., W.G., and M.A.E. designed the experiments; E.A.I., M.M.R., and A.S.W. performed the experiments; G.M.H., R.A.A., A.K.I., S.A.A., and E.A.I. analyzed the data; E.A.I. drafted the article with input from C.G.M., S.C., and H.L.; and M.A.E., M.M.R., A.S.W. and W.G. supervised the project. All authors reviewed the article and approved its submission.
Author Disclosure Statement
No competing financial interests exist.
Funding Information
This work is supported, in part, by the National Institute on Drug Abuse (contract no. N01DA-15-7793).
Cite this article as: Ibrahim EA, Radwan MM, Gul W, Majumdar CG, Hadad GM, Abdel Salam RA, Ibrahim AK, Ahmed SA, Chandra S, Lata H, ElSohly MA, Wanas AS (2023) Quantitative determination of cannabis terpenes using gas chromatography-flame ionization detector, Cannabis and Cannabinoid Research 8:5, 899–910, DOI: 10.1089/can.2022.0188.
References
- 1. McPartland JM, Russo EB. Non-Phytocannabinoid Constituents of Cannabis and Herbal Synergy. In: Handbook of Cannabis. Oxford University Press: Oxford, UK, 2014; pp. 280–295. [Google Scholar]
- 2. Flores-Sanchez IJ, Verpoorte R. Secondary metabolism in cannabis. Phytochem Rev 2008;7(3):615–639; doi: 10.1007/s11101-008-9094-4. [DOI] [Google Scholar]
- 3. Pertwee RG. Cannabinoid pharmacology: The first 66 years. Br J Pharmacol 2006;147 (suppl 1):S163–S171; doi: 10.1038/sj.bjp.0706406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Bostwick JM, Reisfield GM, DuPont RL. Clinical decisions: Medicinal use of marijuana. N Engl J Med 2013;368(9):866–868; doi: 10.1056/NEJMclde1300970. [DOI] [PubMed] [Google Scholar]
- 5. Elzinga S, Fischedick J, Podkolinski R, et al. Cannabinoids and terpenes as chemotaxonomic markers in cannabis. Nat Prod Chem Res 2015;3(4):1–9; doi: 10.4172/2329-6836.1000181. [DOI] [Google Scholar]
- 6. ElSohly MA, Radwan MM, Gul W, et al. Phytochemistry of Cannabis sativa L. In: Phytocannabinoids. Springer: Berlin/Heidelberg, Germany, 2017; pp. 1–36. [DOI] [PubMed] [Google Scholar]
- 7. Zager JJ, Lange I, Srividya N, et al. Gene networks underlying cannabinoid and terpenoid accumulation in cannabis. Plant Physiol 2019;180(4):1877–1897; doi: 10.1104/pp.18.01506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. ElSohly MA. Marijuana and the Cannabinoids. Springer Science & Business Media; Humana Press: Totowa, New Jersey, USA; 2007. [Google Scholar]
- 9. Gallily R, Yekhtin Z, Hanuš LO. The anti-inflammatory properties of terpenoids from cannabis. Cannabis Cannabinoid Res 2018;3(1):282–290; doi: 10.1089/can.2018.0014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Gonçalves EC, Baldasso GM, Bicca MA, et al. Terpenoids, cannabimimetic ligands, beyond the cannabis plant. Molecules 2020;25(7):1567; doi: 10.3390/molecules25071567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Hanuš LO, Hod Y. Terpenes/terpenoids in cannabis: Are they important? Med Cannabis Cannabinoids 2020;3(1):25–60; doi: 10.1159/000509733. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Livingston SJ, Quilichini TD, Booth JK, et al. Cannabis glandular trichomes alter morphology and metabolite content during flower maturation. Plant J 2020;101(1):37–56; doi: 10.1111/tpj.14516. [DOI] [PubMed] [Google Scholar]
- 13. Conneely LJ, Mauleon R, Mieog J, et al. Characterization of the Cannabis sativa glandular trichome proteome. PLoS One 2021;16(4):e0242633; doi: 10.1371/journal.pone.0242633. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Davis MP. Cannabinoids for symptom management and cancer therapy: The evidence. J Natl Compr Canc Netw 2016;14(7):915–922; doi: 10.6004/jnccn.2016.0094. [DOI] [PubMed] [Google Scholar]
- 15. Hill KP. Medical marijuana for treatment of chronic pain and other medical and psychiatric problems: A clinical review. JAMA 2015;313(24):2474–2483; doi: 10.1001/jama.2015.6199. [DOI] [PubMed] [Google Scholar]
- 16. Ferioli V, Rustichelli C, Pavesi G, et al. Analytical characterisation of hashish samples. Chromatographia 2000;52(1):39–44; doi: 10.1007/BF02490790. [DOI] [Google Scholar]
- 17. Careri M, Mangia A.. Spectrometry Analysis of Flavors and Fragrances. In: Current Practice of Gas Chromatography-Mass Spectrometry. Marcel Dekker, New York, USA, 2001; p. 409. [Google Scholar]
- 18. Lockwood G. Techniques for gas chromatography of volatile terpenoids from a range of matrices. J Chromatogr A 2001;936(1–2):23–31; doi: 10.1016/s0021-9673(01)01151-7. [DOI] [PubMed] [Google Scholar]
- 19. Omar J, Olivares M, Amigo JM, et al. Resolution of co-eluting compounds of Cannabis sativa in comprehensive two-dimensional gas chromatography/mass spectrometry detection with multivariate curve resolution-alternating least squares. Talanta 2014;121:273–280; doi: 10.1016/j.talanta.2013.12.044. [DOI] [PubMed] [Google Scholar]
- 20. Da Porto C, Decorti D, Natolino A. Separation of aroma compounds from industrial hemp inflorescences (Cannabis sativa L.) by supercritical CO2 extraction and on-line fractionation. Ind Crops Prod 2014;58:99–103; doi: 10.1016/j.indcrop.2014.03.042. [DOI] [Google Scholar]
- 21. Romano LL, Hazekamp A. Cannabis oil: Chemical evaluation of an upcoming cannabis-based medicine. Cannabinoids 2013;1(1):1–11. [Google Scholar]
- 22. Sexton M, Shelton K, Haley P, et al. Evaluation of cannabinoid and terpenoid content: Cannabis flower compared to supercritical CO2 concentrate. Planta Med 2018;84(4):234–241; doi: 10.1055/s-0043-119361. [DOI] [PubMed] [Google Scholar]
- 23. Ibrahim EA, Wang M, Radwan MM, et al. Analysis of terpenes in Cannabis sativa L. using GC/MS: Method development, validation, and application. Planta Med 2019;85(5):431–438; doi: 10.1055/a-0828-8387. [DOI] [PubMed] [Google Scholar]
- 24. Giese MW, Lewis MA, Giese L, et al. Method for the analysis of cannabinoids and terpenes in Cannabis. J AOAC Int 2015;98(6):1503–1522; doi: 10.5740/jaoacint.15-116. [DOI] [PubMed] [Google Scholar]
- 25. Horwitz W. AOAC Guidelines for Single Laboratory Validation of Chemical Methods for Dietary Supplements and Botanicals. AOAC International: Gaithersburg, MD, USA; 2002; pp. 12–19. [Google Scholar]
- 26. ElSohly MA, Chandra S, Radwan M, et al. A comprehensive review of cannabis potency in the United States in the last decade. Biol Psychiatry Cogn Neurosci Neuroimaging 2021;6(6):603–606; doi: 10.1016/j.bpsc.2020.12.016. [DOI] [PubMed] [Google Scholar]
- 27. Wanas AS, Radwan MM, Chandra S, et al. Chemical composition of volatile oils of fresh and air-dried buds of cannabis chemovars: Their insecticidal and repellent activities. Nat Prod Commun 2020;15(5):1–7; doi: 10.1177/1934578X20926729. [DOI] [Google Scholar]