Brand new medicinal cannabis and adult-use markets are emerging in North America.
The consumption of CBD and cannabis concentrate products such as isolates, vape products, beverages, topicals, edibles and waxes continue to grow in popularity.
Concentrates and concentrate derivative products are projected to exceed half of the consumer market by 2022, according to recent market research.1
This growth, along with the variety of sample types, poses an analytical challenge for testing laboratories. Concentrate matrices have a considerable effect on the analytical process due to greater effects in sample matrices and increased concentration levels (up to 95%/wt) of cannabinoids in the sample.
As this effect affects the response of several pesticide molecules, laboratories need to validate a pesticide method specific to the sample matrix type.
This article outlines an LC/MS/MS method for the analysis of 66 pesticides, including five mycotoxins. It also explores the investigation of hydrophobic and chlorinated pesticides usually analyzed by GC/MS/MS.
The method uses a cannabis concentrate matrix and features a simple solvent extraction. This is followed by analysis via an LC/MS/MS instrument with dual APCI and ESI sources.
The analysis in this study generated exceptional recoveries and detection limits for all analytes - well under the limits specified by California’s state cannabis regulations.
Experimental
Hardware and Software
A PerkinElmer QSight® LX50 UHPLC system was employed for chromatographic separation. Subsequent detection was completed using a PerkinElmer QSight 420 MS/MS detector with dual APCI and ESI ionization sources. These sources were operated independently with two separate inlets.
The Simplicity 3Q™ software platform was utilized for instrument control, data acquisition and data processing.
Sample Preparation
Sample preparation involved a 25-fold dilution for the APCI source and a 50-fold dilution for the ESI source. The procedure was as follows:
- Approximately 5 g of cannabis concentrate was measured out as a representative sample for each sample batch.
- A measure of 1 g of sample was taken and placed into a 50 ml centrifuge tube.
- A 10 ml measure of LC/MS grade acetonitrile with 0.1% formic acid was added to the tube, which was then capped.
- The tube was placed in a multi-tube vortex mixer and allowed to vortex for 10 minutes.
- The extract was centrifuged in the tube for 10 minutes at 3000 rpm.
- The solvent was transferred into a 10 ml glass amber vial and capped.
- The bottle was labeled with the sample ID.
For the APCI method, 400 µL of extracted sample was transferred into a 2 ml HPLC vial. It was spiked with 10 µL of internal standard and then diluted with 590 µl of LC/MS grade acetonitrile with 0.1% formic acid and mixed.
The ESI method involved transferring 200 µl of extracted sample into a 2 ml HPLC vial. It was spiked with 10 µl of internal standard and then diluted with 790 µl of LC/MS grade acetonitrile with 0.1% formic acid and mixed.
The sample was injected for LC/MS/MS analysis, utilizing pesticide methods.
Results and Discussion
Detectability and Reproducibility
The majority of laboratories currently employ several analytical instruments (such as LC/MS/MS and GC/MS/MS) and laborious sample preparation methods (including QuEChERS) to attain the low pesticide regulatory limits for a variety of food matrices.
As a possible solution to this complex process, a proven LC/MS/MS analytical method incorporating fast solvent extraction is outlined in this study.
A PerkinElmer liquid chromatograph coupled to a tandem mass spectrometer was utilized to carry out a complete analysis of all 66 pesticides and five mycotoxins listed in California’s state regulations for cannabis concentrates.
Every compound sampled was analyzed with a QSight 420 dual source mass spectrometer which had been equipped with ESI and APCI ionization probes.
Cypermethrin, ethyl parathion, pentochloronitrobenzene (quintozene) and other pesticides that were usually analyzed by gas chromatography were all detected using this single platform LC/MS/MS system.
LC Method and MS Source Conditions
The source parameters for the LC and MS methods are shown in Table 1.
Table 1. LC Method and MS Source Conditions. Source: PerkinElmer Cannabis & Hemp Testing Solutions
| LC Conditions |
| LC Column |
PerkinElmer Quasar™ SPP Pesticides (4.6 × 100 mm, 2.7 μm) |
Mobile Phase A
(ESI method) |
2 mM ammonium formate + 0.1% formic acid (in LC/MS grade water) |
Mobile Phase B
(ESI method) |
2 mM ammonium formate + 0.1% formic acid (in LC/MS grade methanol) |
Mobile Phase A
(APCI method) |
LC/MS grade water |
Mobile Phase B
(APCI method) |
LC/MS grade methanol |
Mobile Phase
Gradient |
The run time for the optimized gradient elution method, including analytical column re-conditioning, was 18 minutes for the ESI method, and 12 minutes for the APCI method. The final method ensured separation of the bulk cannabis matrix from the analytes for improved quantitation. |
Column Oven
Temperature |
30 ºC |
Auto sampler
Temperature |
20 ºC |
| Injection Volume |
3 μL and 10 μL for LC/MS/MS method with ESI and APCI source, respectively. |
| MS Source Conditions for ESI Source and APCI Source |
ESI Voltage
(Positive) |
+5100 V |
ESI Voltage
(Negative) |
-4200V |
APCI Corona
Discharge |
-3 μA |
| Drying Gas |
150 arbitrary units |
| Nebulizer Gas |
350 arbitrary units |
Source Temperature
(ESI Method)) |
315 ºC |
Source Temperature
(APCI Method) |
250 ºC |
HSID Temperature
(ESI Method) |
200 ºC |
HSID Temperature
(APCI Method) |
180 ºC |
| Detection Mode |
Time-managed MRM™ |
The LOQ (limit of quantification) and the response reproducibility at the LOQ for each of the mycotoxins and pesticides (Category 1 and 2) in the cannabis concentrate sample are outlined in Tables 2, 3 and 4.
LOQs were established by studying the signal of the quantifier ion (S/N > 10) and making certain that the ratios of the product ion were within the 30% tolerance windows of the anticipated ion ratio.
The response RSD for every single pesticide and mycotoxin at its LOQ level in the cannabis matrix was found to be under 20%.
LOQs determined in this study were well under the state of California’s action limit by a factor of 1.2 to 1000 for all listed pesticides and mycotoxins (Tables 2, 3, and 4).
This shows the sensitivity and reproducibility of the method for pesticide and mycotoxin analysis in cannabis concentrate samples, in support of the sample’s adherence to California’s state regulatory program.
Table 2. LOQs for California Category II Pesticides with LC/MS/MS in Cannabis Concentrate. Red/Green: Pesticides typically analyzed by GC/MS/MS. Of those, analytes highlighted in red were analyzed on the QSight by ESI, and those in green were analyzed on the QSight by APCI. Pesticides in black were analyzed on the QSight by ESI. Source: PerkinElmer Cannabis & Hemp Testing Solutions
| S. No. |
Category II
Residual Pesticide |
LOQ |
Action
Level (μg/g) |
Action
Level/LOQ |
LC/MS/MS
(μg/g) |
%CV
(n=7) |
| 1 |
Abamectin |
0.08 |
14.0 |
0.1 |
1.2 |
| 2 |
Acephate |
0.01 |
4.5 |
0.1 |
10 |
| 3 |
Acequinocyl |
0.05 |
10.8 |
0.1 |
2 |
| 4 |
Acetamiprid |
0.01 |
3.3 |
0.1 |
10 |
| 5 |
Azoxystrobin |
0.005 |
11.9 |
0.1 |
20 |
| 6 |
Bifenazate |
0.005 |
15.2 |
0.1 |
20 |
| 7 |
Bifenthrin |
0.005 |
5.3 |
0.5 |
100 |
| 8 |
Boscalid |
0.005 |
14.2 |
0.1 |
20 |
| 9 |
Captan |
0.5 |
13.0 |
0.7 |
1.4 |
| 10 |
Carbaryl |
0.005 |
7.4 |
0.5 |
100 |
| 11 |
Chlorantraniliprole |
0.01 |
10.0 |
10.0 |
1000 |
| 12 |
Clofentezine |
0.01 |
14.4 |
0.1 |
10 |
| 13 |
Cyfluthrin |
0.9 |
16.0 |
2.0 |
2.2 |
| 14 |
Cypermethrin |
0.15 |
7.8 |
1.0 |
6.66 |
| 15 |
Diazinon |
0.01 |
7.5 |
0.2 |
20 |
| 16 |
Dimethomorph |
0.005 |
17.4 |
2.0 |
400 |
| 17 |
Etoxazole |
0.01 |
5.4 |
0.1 |
10 |
| 18 |
Fenhexamid |
0.05 |
8.3 |
0.1 |
2 |
| 19 |
Fenpyroximate |
0.01 |
6.9 |
0.1 |
10 |
| 20 |
Flonicamid |
0.01 |
6.4 |
0.1 |
10 |
| 21 |
Fludioxonil |
0.005 |
11.9 |
0.1 |
20 |
| 22 |
Hexythiazox |
0.005 |
10.1 |
0.1 |
20 |
| 23 |
Imidacloprid |
0.01 |
9.9 |
3.0 |
300 |
| 24 |
Kresoxim-methyl |
0.05 |
4.8 |
0.1 |
2 |
| 25 |
Malathion |
0.005 |
6.1 |
0.5 |
100 |
| 26 |
Metalaxyl |
0.005 |
3.3 |
2.0 |
400 |
| 27 |
Methomyl |
0.01 |
12.3 |
0.1 |
10 |
| 28 |
Myclobutanil |
0.005 |
5.4 |
0.1 |
20 |
| 29 |
Naled |
0.05 |
13.0 |
0.1 |
2 |
| 30 |
Oxamyl |
0.01 |
4.1 |
0.2 |
20 |
| 31 |
Pentachloronitrobenzene |
0.025 |
10.2 |
0.1 |
4 |
| 32 |
Permethrin |
0.05 |
6.8 |
0.5 |
10 |
| 33 |
Phosmet |
0.01 |
12.0 |
0.1 |
10 |
| 34 |
Piperonylbutoxide |
0.15 |
4.0 |
3.0 |
20 |
| 35 |
Prallethrin |
0.08 |
14.4 |
0.1 |
1.2 |
| 36 |
Propiconazole |
0.005 |
4.5 |
0.1 |
20 |
| 37 |
Pyrethrins |
0.37 |
5.1 |
0.5 |
1.3 |
| 38 |
Pyridaben |
0.01 |
10.4 |
0.1 |
10 |
| 39 |
Spinetoram |
0.008 |
9.0 |
0.1 |
12.5 |
| 40 |
Spinosad |
0.01 |
10.7 |
0.1 |
10 |
| 41 |
Spiromesifen |
0.05 |
5.2 |
0.1 |
2 |
| 42 |
Spirotetramat |
0.005 |
12.5 |
0.1 |
20 |
| 43 |
Tebuconazole |
0.01 |
13.8 |
0.1 |
10 |
| 44 |
Thiamethoxam |
0.005 |
8.5 |
4.5 |
900 |
| 45 |
Trifloxystrobin |
0.005 |
4.9 |
0.1 |
20 |
Table 3. LOQs for California Category I Pesticides with LC/MS/MS in Cannabis Concentrate. Red/Green: Pesticides typically analyzed by GC/MS/MS. Of those, analytes highlighted in red were analyzed on the QSight by ESI, and those in green were analyzed on the QSight by APCI. Pesticides in black were analyzed on the QSight by ESI. Source: PerkinElmer Cannabis & Hemp Testing Solutions
| S. No. |
Category I
Residual Pesticide |
LOQ |
Action
Level
(μg/g) |
Action
Level/LOQ |
LC/MS/MS
(μg/g) |
%CV
(n=7) |
| 1 |
Aldicarb |
0.025 |
9.5 |
0.1 |
4 |
| 2 |
Carbofuran |
0.005 |
8.5 |
0.1 |
20 |
| 3 |
Chlordane |
0.08 |
15.3 |
0.1 |
1.2 |
| 4 |
Chlorfenpyr |
0.05 |
18.0 |
0.1 |
2 |
| 5 |
Chlorpyrifos |
0.05 |
8.5 |
0.1 |
2 |
| 6 |
Coumaphos |
0.01 |
15.7 |
0.1 |
10 |
| 7 |
Daminozide |
0.05 |
11.3 |
0.1 |
2 |
| 8 |
DDVP (Dichlorvos) |
0.025 |
4.2 |
0.1 |
4 |
| 9 |
Dimethoate |
0.005 |
5.1 |
0.1 |
20 |
| 10 |
Ethoprop(hos) |
0.01 |
12.5 |
0.1 |
10 |
| 11 |
Etofenprox |
0.01 |
8.6 |
0.1 |
10 |
| 12 |
Fenoxycarb |
0.005 |
5.5 |
0.1 |
20 |
| 13 |
Fipronil |
0.005 |
9.8 |
0.1 |
20 |
| 14 |
Imazalil |
0.005 |
19.3 |
0.1 |
20 |
| 15 |
Methiocarb |
0.005 |
10.9 |
0.1 |
20 |
| 16 |
Methyl Parathion |
0.05 |
3.0 |
0.1 |
2 |
| 17 |
Mevinphos |
0.01 |
8.1 |
0.1 |
10 |
| 18 |
Paclobutrazol |
0.01 |
10.2 |
0.1 |
10 |
| 19 |
Propoxur |
0.01 |
11.8 |
0.1 |
10 |
| 20 |
Spiroxamine |
0.01 |
6.3 |
0.1 |
10 |
| 21 |
Thiacloprid |
0.005 |
6.5 |
0.1 |
20 |
Table 4. LOQs for Mycotoxins with LC/MS/MS in the Cannabis Concentrate. Source: PerkinElmer Cannabis & Hemp Testing Solutions
| S. No. |
Category II
Mycotoxin |
LOQ |
Action
Level (μg/g) |
Action
Level/LOQ |
LC/MS/MS
(μg/g) |
%CV
(n=7) |
| 1 |
Ochratoxin A |
0.0125 |
12.6 |
0.020 |
1.6 |
| 2 |
Aflatoxin B1 |
0.003 |
12.4 |
NA |
NA |
| 3 |
Aflatoxin B2 |
0.003 |
13.0 |
NA |
NA |
| 4 |
Aflatoxin G1 |
0.004 |
8.2 |
NA |
NA |
| 5 |
Aflatoxin G2 |
0.005 |
10.5 |
NA |
NA |
| 6 |
Aflatoxin (B1+B2+G1+G2) |
0.015 |
NA |
0.020 |
1.33 |
Recovery Studies With Solvent Extraction
Sample preparation in cannabis concentrate testing is frequently pinpointed as the key bottleneck associated with the analysis of mycotoxins and pesticides.
Solid phase and other techniques involve multiple steps and substantial quantities of costly sorbent materials.3
In contrast, solvent extraction offers a means of achieving high extraction recovery that is easy, efficient and allows for high throughput. A solvent extraction method was used in this study for the extraction of mycotoxins and pesticides for this reason.
Spiked cannabis concentrate samples were utilized to corroborate the recovery performance of this technique, and the cannabis concentrate samples were analyzed to confirm the absence of mycotoxins and pesticides prior to the spiking stage.
The concentrate samples of cannabis were spiked at two levels for each contaminant as follows:
- Mycotoxins were spiked at 0.02 µg/g (low) and 0.2 µg/g (high)
- Pesticides were spiked at 0.1 µg/g (low) and 1.0 µg/g (high)
Tables 5, 6 and 7 indicate that the absolute recoveries for all mycotoxins and pesticides at both spiking levels were within the acceptable range of 70% to 120%, with RSD values < 20%.
As their LOQ is higher than 0.1 µg/g, no recovery data could be acquired for cyfluthrin, captan and cypermethrin (all pesticide) at the lower level of 0.1 µg/g.
Table 5. Recoveries of Category II pesticides in cannabis concentrate matrix at two different levels with solvent extraction. Source: PerkinElmer Cannabis & Hemp Testing Solutions
| S. No. |
Category II
Residual Pesticide |
Recovery/% |
RSD/%
(n=3) |
Recovery/% |
RSD/%
(n=3) |
| 1 |
Abamectin |
81.6 |
4.7 |
83.7 |
15.5 |
| 2 |
Acephate |
98.3 |
2.0 |
93.4 |
1.1 |
| 3 |
Acequinocyl |
99.2 |
5.3 |
84.7 |
1.6 |
| 4 |
Acetamiprid |
94.7 |
1.0 |
94.4 |
0.7 |
| 5 |
Azoxystrobin |
93.0 |
2.2 |
98.5 |
5.3 |
| 6 |
Bifenazate |
91.9 |
3.2 |
91.6 |
0.9 |
| 7 |
Bifenthrin |
94.5 |
3.4 |
93.7 |
0.3 |
| 8 |
Boscalid |
82.0 |
3.1 |
98.7 |
10.5 |
| 9 |
Captan* |
- |
- |
96.4 |
18.9 |
| 10 |
Carbaryl |
93.6 |
6.1 |
93.9 |
4.3 |
| 11 |
Chlorantraniliprole |
87.8 |
5.1 |
98.1 |
8.8 |
| 12 |
Clofentezine |
71.9 |
3.3 |
87.1 |
16.4 |
| 13 |
Cyfluthrin* |
- |
- |
95.4 |
5.5 |
| 14 |
Cypermethrin* |
- |
- |
93.4 |
5.5 |
| 15 |
Diazinon |
89.1 |
1.1 |
94.5 |
4.1 |
| 16 |
Dimethomorph |
83.7 |
2.6 |
93.8 |
4.0 |
| 17 |
Etoxazole |
97.6 |
1.9 |
96.9 |
3.1 |
| 18 |
Fenhexamid |
102.8 |
10.6 |
103.0 |
13.5 |
| 19 |
Fenpyroximate |
91.1 |
1.7 |
95.7 |
1.2 |
| 20 |
Flonicamid |
102.6 |
5.7 |
97.8 |
0.9 |
| 21 |
Fludioxonil |
103.3 |
3.9 |
96.1 |
1.6 |
| 22 |
Hexythiazox |
79.8 |
2.7 |
96.6 |
11.7 |
| 23 |
Imidacloprid |
95.9 |
2.4 |
95.4 |
1.2 |
| 24 |
Kresoxim-methyl |
93.4 |
3.0 |
96.1 |
2.5 |
| 25 |
Malathion |
95.5 |
5.2 |
93.6 |
3.4 |
| 26 |
Metalaxyl |
93.2 |
2.8 |
95.1 |
3.7 |
| 27 |
Methomyl |
97.4 |
2.7 |
97.4 |
2.1 |
| 28 |
Myclobutanil |
85.7 |
3.2 |
94.9 |
1.6 |
| 29 |
Naled |
100.0 |
8.2 |
96.9 |
5.0 |
| 30 |
Oxamyl |
98.9 |
1.7 |
95.1 |
0.9 |
| 31 |
Pentachloronitrobenzene |
92.8 |
4.3 |
96.0 |
3.5 |
| 32 |
Permethrin |
92.8 |
13.1 |
98.9 |
3.0 |
| 33 |
Phosmet |
80.2 |
3.9 |
94.3 |
3.3 |
| 34 |
Piperonylbutoxide |
90.3 |
2.0 |
95.2 |
2.1 |
| 35 |
Prallethrin |
90.5 |
14.4 |
101.7 |
8.3 |
| 36 |
Propiconazole |
81.3 |
1.8 |
93.9 |
12.0 |
| 37 |
Pyrethrins |
109 |
16.9 |
101.0 |
14.4 |
| 38 |
Pyridaben |
91.9 |
3.5 |
95.2 |
2.8 |
| 39 |
Spinetoram |
92.1 |
1.6 |
93.4 |
1.8 |
| 40 |
Spinosad |
95.1 |
8.7 |
97.7 |
3.4 |
| 41 |
Spiromesifen |
99.8 |
5.0 |
99.0 |
5.6 |
| 42 |
Spirotetramat |
95.8 |
2.6 |
94.7 |
1.8 |
| 43 |
Tebuconazole |
96.4 |
2.7 |
94.9 |
1.7 |
| 44 |
Thiamethoxam |
97.6 |
2.4 |
96.7 |
1.7 |
| 45 |
Trifloxystrobin |
92.7 |
3.5 |
97.0 |
0.9 |
Table 6. Recoveries of Category I pesticides in cannabis concentrate matrix at two different levels with solvent extraction. Source: PerkinElmer Cannabis & Hemp Testing Solutions
| S. No. |
Category I
Residual Pesticide |
Low Level 0.1 μg/g |
High Level 1 μg/g |
| Recovery/% |
RSD/%
(n=3) |
Recovery/% |
RSD/%
(n=3) |
| 1 |
Aldicarb |
88.9 |
14.2 |
95.5 |
4.1 |
| 2 |
Carbofuran |
91.9 |
1.5 |
93.8 |
3.9 |
| 3 |
Chlordane |
102.3 |
15.3 |
105.2 |
4.4 |
| 4 |
Chlorfenapyr |
94.8 |
3.0 |
94.7 |
4.8 |
| 5 |
Chlorpyrifos |
108.6 |
4.9 |
97.9 |
13.4 |
| 6 |
Coumaphos |
73.6 |
5.1 |
93.5 |
13.8 |
| 7 |
Daminozide |
95.1 |
6.0 |
95.2 |
1.6 |
| 8 |
DDVP (Dichlorvos) |
92.6 |
3.5 |
95.5 |
1.2 |
| 9 |
Dimethoate |
94.2 |
0.9 |
96.9 |
1.1 |
| 10 |
Ethoprop(hos) |
88.0 |
5.7 |
95.7 |
2.7 |
| 11 |
Etofenprox |
101.3 |
4.4 |
97.4 |
3.7 |
| 12 |
Fenoxycarb |
97.1 |
3.5 |
96.9 |
1.5 |
| 13 |
Fipronil |
98.1 |
3.6 |
95.8 |
3.7 |
| 14 |
Imazalil |
88.5 |
9.8 |
98.0 |
4.5 |
| 15 |
Methiocarb |
94.8 |
4.4 |
101.9 |
1.3 |
| 16 |
Methyl parathion |
94.4 |
4.8 |
95.6 |
6.3 |
| 17 |
Mevinphos |
93.4 |
4.0 |
96.4 |
1.7 |
| 18 |
Paclobutrazol |
94.2 |
2.6 |
97.8 |
1.7 |
| t19 |
Propoxur |
90.9 |
2.7 |
94.3 |
3.8 |
| 20 |
Spiroxamine |
97.3 |
0.9 |
95.9 |
1.9 |
| 21 |
Thiacloprid |
92.9 |
2.5 |
93.6 |
3.0 |
Table 7. Recoveries of mycotoxins in cannabis concentrate matrix at 2 different levels with solvent extraction. Source: PerkinElmer Cannabis & Hemp Testing Solutions
| S. No. |
Category II
Mycotoxin |
Low Level 0.02 μg/g |
High Level 0.2 μg/g |
| Recovery/% |
RSD/%
(n=3) |
Recovery/% |
RSD/%
(n=3) |
| 1 |
Aflatoxin B1 |
92 |
8 |
93 |
5 |
| 2 |
Aflatoxin B2 |
94 |
9 |
92 |
6 |
| 3 |
Aflatoxin G1 |
81 |
18 |
98 |
9 |
| 4 |
Aflatoxin G2 |
96 |
17 |
91 |
10 |
| 5 |
Ochratoxin A |
87 |
12 |
85 |
3 |
Internal Standards
The cannabis concentrate samples exhibited a significant matrix effect due to the significant volume of cannabinoids present – 50% to 95%. This led to 30 internal standards being utilized to improve the quantitative analysis and overall recovery.
The internal standards were used to compensate for matrix ion suppression effects and were corrected for any analyte loss during sample preparation.
As seen in the experimental results shown in Figure 1, the utilization of internal standards considerably increased the overall recovery of coumaphos.
This was calculated based on evaluating the extracted concentration of pre-spiked analyte against the neat or unextracted solution concentration (from 56% to 86%) owing to correction of matrix effects and analyte loss during the extraction step.
The overall recoveries of 70% to130 % were achieved for all 66 pesticides and five mycotoxins with the addition of 30 internal standards to the cannabis concentrate matrix.
Figure 1. (a) Overlay of the response of coumaphos in solvent (red) and coumaphos (green) pre-spiked in the cannabis concentrate matrix, without an internal standard. The response ratio (RR) of coumaphos in cannabis extract to solvent standard was 0.56. (b) Overlay of the response of coumaphos (green) and coumaphos-D10 internal standard (red) in the pre-spiked cannabis concentrate matrix with a response ratio (RR) of 2.17 for the analyte to internal standard. (c) Overlay of the response of coumaphos (green) and coumaphos-D10 internal standard (red) in the solvent with a response ratio (RR) of 2.56 for the analyte to internal standard. Image Credit: PerkinElmer Cannabis & Hemp Testing Solutions
Analysis of Pesticides Typically Analyzed by GC-MS/MS, Utilizing LC/MS/MS With Dual ESI and APCI Ion Sources
A range of pesticides regulated for use in cannabis production by California and other states are usually analyzed utilizing GC/MS/MS with an ESI source.
This is because such pesticides display either low proton affinity, which results in low ionization efficiency with the ESI source, or they cannot be ionized by the ESI ion source used in traditional LC/MS/MS systems.
Pesticides typically analyzed by GC/MS/MS include cypermethrin, captan, naled, cyfluthrin, permethrin, pentachlornitrobenzene, chlordane, chlorfenapyr, methyl parathion and pyrethrins.
To realize the necessary sensitivity and low detection limits for some of these pesticides (cypermethrin, cyfluthrin, naled, captan, permethrin, chlorpyrifos, coumaphos, prallethrin and pyrethrins), the selected MRMs and source conditions such as temperature and flow were optimized with a heated electrospray source.
The remaining pesticides (chlorfenapyr, pentachlornitrobenzene, chlordane and methyl parathion) were measured at low limits in the cannabis concentrate matrix utilizing the APCI source in the LC/MS/MS instrument.
These analytes had LOQs in the range of 0.05 to 0.9 µg/g, which are well under California’s action limits.
A sample chromatogram of cannabis concentrate spiked at a level of 0.1 µg/g with the pesticides chlorfenapyr and naled, which were analyzed by LC/MS/MS with APCI and ESI sources, respectively, can be seen in Figure 2.
Figure 2. Sample chromatogram of (a) naled and (b) chlorfenapyr spiked at a level of 0.1 μg/g in a cannabis concentrate matrix using an LC/MS/MS system with an ESI and APCI source, respectively. Image Credit: PerkinElmer Cannabis & Hemp Testing Solutions
Method Optimization to Overcome Matrix Ion Suppression Effects From a Challenging Cannabis Concentrate Matrix
Cannabis concentrates are prepared by the extraction of cannabis flowers, which means they usually display 3 to 5 times higher levels of cannabinoids (CBD and THC) than cannabis flower raw materials.
With their higher concentration of cannabinoids (in the range of 50% to 95%), cannabis concentrate matrices can cause a significantly more challenging matrix in comparison to cannabis flower samples.
This is further complicated by the low concentration levels of mycotoxins and pesticides in the samples. For pesticide analysis in a cannabis flower sample, the matrix ion suppression effects are minimized by utilizing an overall dilution factor with solvent in the range of 10-20x.
Pesticide analysis in cannabis concentrate with this LC/MS/MS method, however, involved a considerably higher overall dilution factor of 50x for the ESI source and 25x for the APCI source to minimize matrix effects.
A previous study evaluated the analysis of four pesticides in a cannabis flower matrix using a fast 6-minute LC gradient with the APCI source.4
The same 6-minute LC gradient method was utilized for the analysis of the same four pesticides in a cannabis concentrate matrix. A sizable ion suppression matrix effect was noted, which resulted in a much lower signal and reduced sensitivity.
As well as the higher dilution factor for the cannabis concentrate matrix, a 12-minute slower LC gradient method with an APCI source was created to separate cannabinoids and pesticides on the LC column and reduce the ion signal suppression effects.
When compared to the 6-minute fast LC gradient method, the signal-to-noise result for PCNB in the cannabis concentrate matrix was enhanced by a factor of 60 using the 12-minute optimized LC gradient method (Figure 3).
Figure 3. (a) Response for PCNB spiked at level of 1 μg/g in a cannabis concentrate matrix using a fast 6-min LC gradient method, coupled with the QSight MS/MS system with APCI source. (b) Response for PCNB spiked at a level of 1 μg/g in a cannabis concentrate matrix using an optimized 12-minute LC gradient method, coupled with a QSight MS/MS system with APCI source. Image Credit: PerkinElmer Cannabis & Hemp Testing Solutions
Selectivity of PCNB Analysis and Mechanism of PCNB Ionization With APCI Source
PCNB does not have a hydrogen atom to lose, so it cannot be ionized utilizing an ESI source in negative ion mode. PCNB is further precluded from forming ions utilizing an ESI source in positive ion mode for the following reasons:
- Low proton affinity
- The nonpolar nature of the compound
- The inability to form adducts with ammonia and other metal ions
These issues meant the ESI source could not be employed to detect PCNB, so the APCI source in negative ion mode was used for selective analysis of PCNB in a number of cannabis matrices.
Figure 4. PCNB response in a blank cannabis concentrate matrix (a), and from spiked level of 0.1 μg/g in cannabis concentrate matrix (b). Image Credit: PerkinElmer Cannabis & Hemp Testing Solutions
The response for PCNB in a blank cannabis concentrate matrix and in a cannabis concentrate matrix spiked with 0.1 µg/g of PCNB can be seen in Figure 4.
The FDA’s method validation guidelines relating to the selectivity of analysis indicates that matrix blanks should be free of any matrix interference peaks at the analyte’s retention time.5
PCNB’s matrix response in the blank cannabis concentrate indicates low background signal with no matrix interference peak at the retention time of PCNB. This demonstrates that the measurement of PCNB in the cannabis concentrate matrix is highly selective (Figure 4a).
The PCNB’s signal-to-noise ratio spiked at California’s action limit of 0.1 ug/g in the cannabis concentrate matrix shows that PCNB can be determined using an APCI source in LC/MS/MS systems possessing good sensitivity and selectivity.
Figure 5. Linearity of PCNB response over 3.5 orders of magnitude in 25 times diluted cannabis concentrate concentrate. Image Credit: PerkinElmer Cannabis & Hemp Testing Solutions
The excellent linearity of the PCNB response over a concentration range of 1 to 3000 ppb (corresponding to 25 to 75000 ppb in cannabis concentrate) in the 25x diluted cannabis concentrate extract, and with a regression fit (R2) of 0.9999, is outlined in Figure 5.
With a PCNB regression fit value of greater than 0.99, the result meets the California Bureau of Cannabis Control’s requirements, which stipulate that regression fits be higher than 0.99.6
The calibration curve’s accuracy was confirmed by comparing back-calculated concentrations from the calibration curve with known concentrations of PCNB, making sure that the stringent criterion of a maximum deviation of 10% was met for all concentration levels.
The literature states that a PCNB analysis with an APCI LC/MS source is not selective.
It may require a quadratic calibration curve receptive to a poor correlation coefficient, but it should be stated that this experimental work outlined a robust APCI method that exhibited excellent selectivity, sensitivity and linearity of a PCNB analysis in a cannabis sample.7
Previous studies have proposed other mechanisms for negative APCI ionization, such as anion adduction, proton abstraction, electron capture and dissociative electron capture.8
It has been shown that nitrobenzene compounds that have been chlorinated can form phenoxide ions under negative APCI conditions.9 The following mechanism for ionization of PCNB was put forward in a previous publication, with the APCI source in negative ion mode (where M is PCNB):10
O2 + eˉ → O2 ̶
M + O2 ̶→ [M – Cl + O]ˉ + ClO
It is possible to attribute the formation of the superoxide ion (O2-) by electron capture to [M-Cl+O]-, followed by its chemical reaction with PCNB.
By analyzing the Q1 scan data for PCNB infusion into the APCI source, this mechanism can be explained further by the Q1 scan data demonstrating a monoisotopic base peak at a nominal mass of 274 dalton.
PCNB’s nominal monoisotopic mass is 293 dalton.
The mass loss of 19 dalton from an intact molecule of PCNB can be explained as follows: chlorine, with a nominal monoisotopic mass of 35, has been lost and an oxygen atom, with a nominal monoisotopic mass of 16, has been added to the PCNB molecule to form a negatively charged ion.
An experimentally observed isotope ratio or pattern of the PCNB ion was discovered to closely match the theoretical isotope pattern of an ion with four chlorine atoms. This provides further proof that PCNB loses one chlorine atom in the APCI ion source.
The APCI ion source’s low mass spectra was examined to confirm the formation of the superoxide reagent ion species, which could interact with PCNB to ionize it.
In circumstances when the mobile phase was changed from 75:25 methanol:water with 0.1% formic acid and 2 mm ammonium formate to just 75:25 methanol:water, it was noted that the superoxide ion (O2-) and PCNB signal increased by a factor of approximately 300 and 30 respectively.
This offered more evidence that superoxide ions play an important role in the ionization of PCNB in the APCI source.
Conclusions
This study demonstrated an LC/MS/MS method with dual APCI and ESI sources for the analysis of various mycotoxin and pesticide residues in cannabis concentrates, which was unique, rapid, quantitative and reliable.
In the procedure outlined above, 62 pesticides and 5 mycotoxins were analyzed with an ESI source and a run time of 18 minutes. An additional 4 pesticides were analyzed with an APCI source and a run time of 12 minutes.
The solvent extraction method as proposed, with 30 internal standards, is suitable for labs analyzing samples in accordance with California regulations.
It is also suitable as the overall recovery of all mycotoxins and pesticides from the cannabis concentrate matrix was in the acceptable range of 70% to 130%, with an RSD of less than 20%.
The method permitted the identification and quantification of all 66 pesticides and 5 mycotoxins in cannabis concentrate samples at levels (0.005 to 0.9 µg/g) below California’s state action limits.
It was also shown that the analysis of PCNB (a pesticide normally analyzed by GC/MSMS with an ESI source) using an APCI source is both sensitive and selective, with excellent linearity.
The method’s capacity screening and quantifying the 66 pesticides and five mycotoxins, which include the hydrophobic and chlorinated compounds normally analyzed by GC/ MSMS, eliminated the requirement of utilizing both an LC/MS/MS and GC/ MS/MS instrument for such an analysis.
This method demonstrates the use of LC/MS/MS as a unique, efficient and cost-effective process for screening and quantifying mycotoxins and pesticides in a cannabis concentrate matrix with a single LC/MS/MS instrument.
References
- https://bdsanalytics.com/wp-content/uploads/2019/01/BDS-Analytics-Top- 10-Trends-2019.pdf.
- United States Department of Agriculture Food Safety and Inspection Service, Office of Public Health Science,” Screening for Pesticides by LC/ MS/MS and GC/MS/MS,” 2018, available from https://www.fsis.usda. gov/wps/wcm/connect/499a8e9e-49bd-480a-b8b6-d1867f96c39d/CLG- PST5.pdf?MOD=AJPERES.
- X. Wang, D. Mackowsky, J. Searfoss and M. Telepchak, LCGC, 34(10), 20-27 (2016).
- A. Dalmia, E. Cudjoe, T. Astill, J. Jalali, J.P. Weisenseel, F. Qin, M. Murphy, and T. Ruthenberg, Cannabis Science and Technology 1(3), 38-50 (2018).
- “Bioanalytical Method Validation Guidance for Industry”, May 2018, available from https://www.fda.gov/.
- https://bcc.ca.gov/law_regs/cannabis_order_of_adoption.pdf.
- Macherone, A. (2019). Tackle Emerging Cannabis Regulations with Confidence. Agilent Application Brief. Retrieved from https://www. agilent.com/cs/library/applications/application-cannabis-hemp-pesticide- 6545-1290-infinity-5994-1127en-agilent.pdf
- C. N. McEwena and B. S Larsen, J. Am. Soc. Mass. Spectrometry, 20, 1518-1521 (2009).
- I. Dzidic, D. I. Carroll, R. N. Stillwell and E. C. Hornig, Anal. Chem., 47(8), 1308-1312 (1975).
- A. Dalmia, US patent application number 20190227041, Jan 23 2019.
Acknowledgments
Produced from materials originally authored by Avinash Dalmia, Saba Hariri, Jacob Jalali, Erasmus Cudjoe, Toby Astill, Charlie Schmidt, and Feng Qin from PerkinElmer; and Charles Johnson, Joey Kingstad and Kevin Smith from Napro Research.
About PerkinElmer Cannabis & Hemp Testing Solutions
With the cannabis and hemp markets continuing to grow rapidly and regulations strengthening, labs increasingly need streamlined access to best-in-class testing solutions geared toward the unique requirements of the industry. Whether your lab is well established or just starting up, PerkinElmer is a single-source vendor for instruments, methods, reagents, and consumables on hand to help enhance your testing capacity and get ahead of the competition.
They help drive analytical best practices and operating procedures and commit to ensuring your laboratory has maximum uptime. Learn about their various instruments, testing methods, and applications for cannabis analyses. Let them work with you to build an efficient workflow, so you can focus on growing your business.
Sponsored Content Policy: AZO Life Science publishes articles and related content that may be derived from sources where we have existing commercial relationships, provided such content adds value to the core editorial ethos of AZO Life Science, which is to educate and inform site visitors interested in medical research, science, medical devices, and treatments.