Using the EZLC Modeler for Cannabinoid Separations–Part 3: What You’re Not Monitoring for Still Matters
24 Mar 2025If you haven’t read the first two installments of this four-part blog series, they can be found here: Optimization of an Existing Method and Using EZLC Software to Monitor Effects of Ammonium Formate Concentrations on Cannabinoid Separations.
While target analytes in the method are the main focus, it is also important to know what could be in your sample that you are not monitoring for. A prime example of this is when a regulatory body rolled out a method in 2019. The disseminated regulations, which later received many revisions, were for cannabinoid testing of dried flower, including non-infused pre-rolls. The criterion for this testing is as follows:
1. Compounds: CBDA, CBG, CBD, THCV, CBN, Δ9-THC, Δ8-THC, CBC, THCA
2. HPLC Column: Restek Raptor ARC-18 2.1 x 150 mm, 2.7 µm (cat.# 9314A62) or an equivalent column that can separate the cannabinoids of interest to achieve a minimum resolution of 1.3
3. Mobile Phases:
-
-
- Mobile phase A: Water with 0.05 volume/volume% formic acid
- Mobile phase B: Acetonitrile with 0.05 volume/volume% formic acid
-
4. Flow Rate: 0.4 mL/min
5. Run time: total 12.00 min (7.00 min + 2.00 min washing period + 3.00 min column re-equilibration)
6. Column Temperature: 35 ⁰C
7. Gradient:
| Time (min) | Flow rate (mL/min) | % Mobile Phase A | % Mobile Phase B |
| 0.00 | 0.4 | 25 | 75 |
| 7.00 | 0.4 | 25 | 75 |
| 7.01 | 0.4 | 0 | 100 |
| 9.00 | 0.4 | 0 | 100 |
| 9.01 | 0.4 | 25 | 75 |
| 12.00 | 0.4 | 25 | 75 |
When labs began testing and validating this method, they determined there was a coelution caused by an analyte not in the method. Due to the regulations, labs are not permitted to change the method parameters but are required to state how/why there could be an interference, along with how they would rectify it.
Below is an example, provided by Tania Sasaki, PhD, of Confidence Analytics. The unmonitored analyte, CBNA, is coeluting with the critical pair, Δ9-THC and Δ8-THC.
Figure 1: Coelution of Δ9-THC and CBNA
Figure 2: Coelution of Δ8-THC and CBNA
By using EZLC modeling software, it is easy to determine if a method will have coelutions. By changing the concentration of ammonium formate and formic acid you can see all analytes are resolved with the desired USP resolution of 1.3 and above.
| Peaks | Experimental tR | Modeled tR | Difference (sec) | Run Time Difference (%) | USP Resolution | |
|---|---|---|---|---|---|---|
| 1. | Cannabidivarinic acid (CBDVA) | 1.47 | 1.32 | 9.12 | 2.2 | |
| 2. | Cannabidivarin (CBDV) | 1.63 | 1.46 | 10.26 | 2.4 | 3.2 |
| 3. | Cannabidiolic acid (CBDA) | 2.03 | 1.81 | 13.26 | 3.2 | 7.1 |
| 4. | Cannabigerolic acid (CBGA) | 2.14 | 1.89 | 15.24 | 3.6 | 1.7 |
| 5. | Cannabigerol (CBG) | 2.27 | 2.01 | 15.78 | 3.8 | 1.9 |
| 6. | Cannabidiol (CBD) | 2.40 | 2.16 | 14.64 | 3.5 | 1.9 |
| 7. | Tetrahydrocannabivarin (THCV) | 2.59 | 2.39 | 11.82 | 2.8 | 2.6 |
| 8. | Tetrahydrocannabivarinic acid (THCVA) | 3.25 | 3.02 | 13.92 | 3.3 | 8.1 |
| 9. | Cannabinol (CBN) | 3.50 | 3.21 | 17.28 | 4.1 | 2.6 |
| 10. | Cannabinolic acid (CBNA) | 4.20 | 3.89 | 18.42 | 4.4 | 6.6 |
| 11. | Δ9-Tetrahydrocannabinol (Δ9-THC) | 4.39 | 4.11 | 16.74 | 4.0 | 1.6 |
| 12. | Δ8-Tetrahydrocannabinol (Δ8-THC) | 4.54 | 4.27 | 16.20 | 3.9 | 1.3 |
| 13. | Cannabicyclol (CBL) | 5.21 | 4.95 | 15.30 | 3.6 | 5.2 |
| 14. | Cannabichromene (CBC) | 5.47 | 5.16 | 18.78 | 4.5 | 1.9 |
| 15. | Tetrahydrocannabinolic acid A (THCA-A) | 5.73 | 5.43 | 17.88 | 4.3 | 1.7 |
| 16. | Cannabichromenic acid (CBCA) | 6.45 | 6.09 | 21.54 | 5.1 | 4.4 |
| Column | Raptor ARC-18 (cat.# 9314A62) | ||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Dimensions: | 150 mm x 2.1 mm ID | ||||||||||||||||||||||||||||
| Particle Size: | 2.7 µm | ||||||||||||||||||||||||||||
| Pore Size: | 90 Å | ||||||||||||||||||||||||||||
| Temp.: | 35 °C | ||||||||||||||||||||||||||||
| Standard/Sample | |||||||||||||||||||||||||||||
| Cannabinoids acids 7 standard, 1000 µg/mL, acetonitrile with 1% DIPEA and 0.05% ascorbic acid (cat.# 34144) | |||||||||||||||||||||||||||||
| Cannabinoids neutrals 9 standard, 1000 µg/mL, P&T methanol, 1 mL/ampul (cat.# 34132) | |||||||||||||||||||||||||||||
| Diluent: | Acetonitrile | ||||||||||||||||||||||||||||
| Conc.: | 50 ppm | ||||||||||||||||||||||||||||
| Inj. Vol.: | 2 µL | ||||||||||||||||||||||||||||
| Mobile Phase | |||||||||||||||||||||||||||||
| A: | Water, 3 mM ammonium formate, 0.1 % formic acid | ||||||||||||||||||||||||||||
| B: | Acetonitrile, 0.1 % formic acid | ||||||||||||||||||||||||||||
|
| Detector | UV/Vis @ 228 nm |
|---|---|
| Flow Cell Size: | 500 nL |
| Instrument | Waters ACQUITY UPLC H-Class |
| Sample Preparation | Working standard was prepared in a 2 mL, 9 mm amber vial (cat. 21142) by diluting 50 µL of each standard into 900 µL acetonitrile and capped with a 9 mm short screw cap (cat. 24497). |
In the coelution above, Δ9-THC is one of the analytes in question. Without adequate separation of Δ9-THC from other analytes in the method, hemp could be misidentified as cannabis. This coelution can also cause inflated THC numbers, leading to potential potency inflation whether intentional or unintentional.
In summary, even if you are not required or do not want to monitor a larger panel of cannabinoids, it is important that during method development you consider all compounds that could be present to ensure accurate identification and quantitative results.
Tune in for the final installment of this series to learn how to harness the full capability of your column.
Resources and Further Reading
H.R.5485 - 115th Congress (2017-2018): Hemp Farming Act of 2018 | congress.gov | library of Congress. (n.d.-a). https://www.congress.gov/bill/115th-congress/house-bill/5485
Check Out the Full Blog Series