A novel liquid chromatography–tandem mass spectrometry (LC–MS/MS) method with a dual electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) source was developed for analyzing 102 pesticides and five mycotoxins that are regulated by the state of Colorado in hemp. The limit of quantitation (LOQ) of the 102 pesticides and five mycotoxins met Colorado state action limit requirements for these analytes in a hemp matrix. For this study, 88 out of 102 pesticides and all five mycotoxins were analyzed using LC–MS/MS with an ESI source, whereas the remaining 14 pesticides were determined using LC–MS/MS with an APCI source. A simple, fast, green, and cheap acetonitrile solvent extraction method was used to extract the pesticides and mycotoxins from the hemp matrix with good extraction efficiency in the range of 80–120%. A hemp matrix is challenging and causes matrix effects such as ion suppression or enhancement. We optimized the LC method and added 30 internal standards to reduce and compensate for these matrix effects to obtain method accuracy in the range of 70–120%. The ionization mechanism of nonpolar pesticides (normally analyzed by gas chromatography–tandem mass spectrometry [GC–MS/MS]) with an APCI ion source was elucidated.
The Agriculture Improvement Act of 2018 (2018 Farm Bill) authorized the production of hemp, removing hemp and hemp seeds from the Drug Enforcement Administration’s (DEA) schedule of Controlled Substances if they contain less than 0.3 percent tetra-hydrocannabinol (THC), which gives users a high (1). Like any other agricultural crop, pesticides are applied to hemp plants to protect it from pests and improve growth yield. Chronic exposure to pesticides can lead to serious health risks; therefore, pesticide analysis in hemp is important for consumer safety and quality control. Because there are no federal regulations for pesticide analysis in hemp, individual states define and regulate the use of pesticides for the production of hemp. Among these states, Oregon, California, and Florida have set action levels for 59, 66, and 67 pesticides, respectively, in cannabis and hemp (2–4). Outside the United States, Canada has set maximum residue limits (MRLs) for 96 pesticides in cannabis-based products (5). Recently, Colorado state has issued action limits for 102 pesticides (including all found on Canada list, and six more) in hemp (6). In addition to pesticide residues, Canada, Colorado, and other U.S. states also require hemp and cannabis products to be tested for five mycotoxins (four aflatoxins and ochratoxin A) (3–6). Mycotoxin contamination can occur either during the cultivation or storage of hemp. Like pesticides, these mycotoxins are toxic and pose a serious health risk to consumers. As a result, testing for the levels of pesticide and mycotoxins in hemp is important to ensure health of consumers and quality control in the United States and Canada.
In the past, pesticide analysis was carried out using both liquid chromatography–tandem mass spectrometry (LC–MS/MS) and gas chromatography–tandem mass spectrometry (GC–MS/MS) methods, coupled with sample preparation methods such as QuEChERS (quick, easy, cheap, effective, rugged, and safe) and solvent extraction with either dispersive solid phase extraction (dSPE) or solid phase extraction (SPE) cleanup. These reported methods are time-consuming, expensive, slow, less environment-friendly, and show poor extraction efficiency for few analytes such as daminozide and others (7–10). In the past, we published our novel LC–MS/MS to meet regulations for pesticides and mycotoxins regulated by different U.S. states and Canada in different cannabis matrices (11–15). The aim of this study was to validate a novel, cost-effective, and fast LC–MS/MS method to analyze all 102 pesticides and five mycotoxins regulated by Colorado state in hemp matrix. The method uses a simple solvent extraction with extraction efficiency in the range of 80–120% and LC–MS/MS method with a dual electrospray ionization (ESI) and atmospheric- pressure chemical ionization (APCI) source to meet action limits set by Colorado state for 102 pesticides and five mycotoxins in hemp matrix (6). Hemp is a challenging matrix for analysis of pesticides and mycotoxins because of its complex chemical composition and causes matrix effects. Both LC and MS methods were optimized to reduce matrix effects, thereby improving method sensitivity, and 30 internal standards were utilized to compensate for matrix effects to achieve method accuracy in the range of 70–120%.
Pesticide and mycotoxin standards were purchased from Accustandard and Sigma-Aldrich, respectively. To compensate for matrix effects observed in complex hemp matrix, 30 internal standards solution was used from PerkinElmer One Pesticide ISO17034 CRM Reagent Kit was used to improve method recovery and quantitation accuracy.
Chromatographic separation was conducted on a PerkinElmer LX50 UHPLC system, while detection was achieved using a PerkinElmer QSight 420 MS/MS detector with a dual ionization ESI and APCI source, which operated independently with two separate inlets. All instrument control, data acquisition, and data processing was performed using the Simplicity 3Q software platform.
Below is the step-by-step sample preparation procedure with a 10-fold dilution.
The LC method and MS source parameters are shown in Table I.
The detection limit for each pesticide and mycotoxin was determined by pre-spiking dried hemp matrix with different low levels and measuring signal to noise and response reproducibility at each level with the LC–MS/ MS method using either the ESI or APCI source. The limit of quantitation (LOQ) for each pesticide and mycotoxin was determined to be the lowest concentration for pre-spiked hemp matrix, at which S/N = 10 or higher and the response relative standard deviation (RSD) (n = 7) less than 15% was measured for quantifier ion. Colorado’s acceptable limits, method limits of quantification (LOQs), and response RSD for each of the pesticides and mycotoxins in dried hemp are summarized in Tables II and III (both online), respectively. The LC–MS/MS method with dual ESI and APCI sources analyzes the 102 pesticides and five mycotoxins in dried hemp, as required by Colorado regulations with two separate injections using the same instrument platform with a total run time of 29 min. A diverter valve enables quick and automated switching of mobile phase eluent to the APCI source from the ESI source and vice versa to facilitate the fast analysis using this method without requiring the time-consuming step of switching APCI and ESI ion sources in LC–MS/MS systems with a single source. Eighty-eight out of 102 pesticides and five mycotoxins were analyzed using the ESI source and the remaining 14 pesticides regulated by the State of Colorado were measured with the APCI source. As demonstrated in Table II, the LOQs determined in this study met the Colorado action limits for all the analyzed pesticides and mycotoxins in the hemp matrix with a single LC–MS/MS instrument platform to reduce cost and complexity of analysis. The LOQ of approximately 83% and 94% of analytes in dried hemp was less than or equal to 0.01 and 0.02 ppm, respectively.
Different groups have developed complex sample preparation methods using either solvent extraction or QuEChERS extraction followed by either SPE or dSPE with different sorbents (7–10). A number of analytes (such as daminozide, imazalil, spirotetramat, pyridaben, and a few others) showed poor extraction efficiency with such intricate, expensive, and time-consuming sample preparation methods. Because daminozide is too polar to be extracted efficiently with QuEChERS, it remains in the aqueous phase and does not partition into the organic solvent during the salting-out step. The recovery of daminozide from a cannabis matrix with QuEChERS extraction has been reported to be less than 10% (9). The extraction efficiency of other pesticides, such as imazalil, spirotetramat, pyridaben, and a few others, was lower than the ideal value of 80% with sample preparation methods using dSPE or SPE clean up, since these pesticides are retained on sorbents used in these sample preparation methods (10). Recently, another group developed a solvent extraction method followed by an evaporation step to carry out solvent exchange with initial LC mobile phase for extraction of pesticides from a hemp matrix. The extra evaporation step in this sample preparation method resulted in low extraction efficiency or recoveries for some of the pesticides such as abamectin, fenpyroximate, pyridaben, and others, due to their precipitation (16). Because of low extraction efficiency of some of pesticides with time-consuming, expensive, and difficult sample preparation methods such as solvent extraction with evaporation, QuEChERS or solvent extraction followed by dSPE or SPE cleanup to extract pesticides and mycotoxins from cannabis and hemp matrix, we used a simple acetonitrile based solvent extraction method to get good extraction efficiency with high throughput. The fortified dried hemp samples were used to determine pesticides and mycotoxins extraction efficiency. Seven dried hemp samples were spiked at two levels (low and high) of the 102 pesticides (0.02 and 0.2 ppm) and 5 mycotoxins (4 and 40 ppb) standard. The extraction efficiency of all the 102 pesticides and five mycotoxins at two different levels was within an acceptable range of 80–120%, with RSD less than 20% for seven dried hemp samples.
Figure 1 shows the extraction efficiency of 93 analytes (88 pesticides and 5 mycotoxins) and 14 pesticides at high level with ESI and APCI source method, respectively. Note that extraction efficiency for few pesticides at low level was not determined since their LOQ was higher than 0.02 ppm in hemp matrix. The sample preparation extraction efficiency was calculated by taking the ratio of signal of analyte in pre-spiked to post-spiked extract and multiplying this number by 100.
Hemp is a challenging matrix and causes matrix effects, such as ion suppression and enhancement, because many of highly concentrated components (such as cannabinoids and terpenes) are also co-extracted with pesticides and mycotoxins. The ion suppression and enhancement effects were determined by checking the signal for spiked pesticides and mycotoxins at a fixed concentration in a dried hemp matrix extract and solvent standard. The matrix effect calculation was performed by taking the percentage difference between the signal of the analyte in the post-spiked hemp extract and clean solvent and dividing it by the signal in clean solvent. About one-third of analytes exhibited absolute matrix effects greater than 20% with the LC–MS/MS method. Because hemp matrix is very hydrophobic, absolute matrix effects greater than 20% were observed mainly at higher elution time for analytes with a reverse phase C18 LC column. Moreover, matrix effects are different in various forms of hemp and cannabis products, such as different types of cannabis and hemp plants and concentrates with different cannabinoids, terpene profiles, and edible products (tinctures, gummies, and chocolates) with high sugar and fat content. Therefore, it is difficult to determine a universal matrix that can represent different forms of cannabis and hemp products for carrying out single matrix-matched calibration for accurate quantitation of pesticides and mycotoxins in different types of cannabis and hemp products. The matrix-matched calibration using different strains of cannabis and hemp flower, types of concentrate and edible products for quantitation of pesticides and mycotoxins in different cannabis and hemp products is not a practical option to achieve high throughput at cannabis testing laboratories because of its complexity and slow cycle time. The practical approach is to use calibration with solvent standards for quantitation of pesticides and mycotoxins in different cannabis and hemp products to improve throughput with less complexity. In this case of solvent calibration, it is vital to add internal standards to both solvent standard calibrants and hemp samples since they would improve method accuracy by compensating for sample matrix effects, such as ion suppression and enhancement effects from hemp matrix.
A solution containing 30 internal standards was spiked in both solvent standards used for generation of calibration curves for quantitation and hemp samples to improve the accuracy of quantitative analysis by compensating for the matrix effects and correcting for any inaccuracies in sample injection in LC. The method accuracy of the LC–MS/MS method is given by following equation:
Seven dried hemp samples were spiked at 2 levels (low and high) of the 102 pesticides (0.02 and 0.2 ppm) and five mycotoxins (4 and 40 ppb) standard. The method accuracy of all 102 pesticides and 5 mycotoxins at two different levels were within an acceptable range of 70–120%, with RSD less than 20% for seven dried hemp samples with internal standards. Figure 2 shows the method accuracy of 93 analytes (88 pesticides and 5 mycotoxins) and 14 pesticides at high level with ESI and APCI source method, respectively, with and without internal standards. Note that method accuracy for a few pesticides at the low level was not determined because their LOQ was higher than 0.02 ppm in hemp matrix. The data illustrates that adding internal standards resulted in 70–120% accuracy for 102 pesticides and five mycotoxins with the LC–MS/MS method. Except for one pesticide, the method accuracy for 101 pesticides and five mycotoxins was in range of 80–120%.
Hemp is a difficult matrix because it shows substantial matrix interference, caused by presence of isobaric compounds, for signaling of some pesticides. To improve the selectivity of pesticide analysis in hemp, it is necessary to have multiple mass transitions (more than two) for a few compounds to find a mass transition that does not have matrix interference. For example, propiconazole shows highest signal based on protonated molecular ion transition in a standard, but the MRM transition (342.1 to 69), based on protonated molecular ion in the hemp matrix, showed poor LOQ of 0.5 ppm because of matrix interference from coextracted compounds isobaric to this pesticide in hemp matrix. Therefore, MRM transition (344.1 to 69) based on the M+2 isotope mass of the protonated molecular ion of propiconazole was determined to reduce matrix interference and achieve a LOQ of 0.02 ppm for propiconazole in the hemp matrix. Similarly, we had to determine optimum MRM transitions for other pesticides such as acequinocyl, prallethrin, pyrethrins, and others to reduce matrix interference.
A number of hydrophobic and non-polar pesticides (quintozene, chlorfenapyr, methyl parathion, iprodione, fenvalerate, and others) found on the Colorado list are traditionally analyzed by GC–MS/MS because they do not ionize effectively when analyzed using LC–MS/MS with an ESI source. These analytes either have low proton affinity or lack acidic functional groups and are difficult to ionize by an ESI source in either positive or negative ion mode. An APCI ion source is much better suited for ionization of these hydrophobic and non-polar analytes, and it was used to determine the detection limits of these hydrophobic analytes in hemp. Using a eleven-minute LC–MS/MS method with an APCI source, the LOQs of 14 non-polar pesticides in dried hemp was in the range of 0.005–0.1 ppm. Figures 3a and 3b show an excellent S/N ratio for two representative pesticides (quintozene and acequinocyl) pre-spiked at a level of 0.01 ppm in the dried hemp matrix using an LC–MS/MS method with an APCI source. This demonstrates the extreme sensitivity of the APCI method. Most analytes with basic or acidic functional groups get ionized in ESI and APCI source with either protonation in positive ion mode or deprotonation in negative ion mode. Since some of the analytes analyzed using the APCI method could not be ionized with above mentioned ionization (protonation/deprotonation) mechanism, we had to determine other ways of ionizing these compounds using APCI ion source. For ionization of compounds in APCI source, different ionization mechanisms such as proton attachment, proton abstraction, anion adduction, electron capture, and dissociative electron capture have been proposed in the past (17). It was demonstrated that chlorinated nitrobenzene compounds can form phenoxide ions under negative APCI conditions (18). Similarly, we proposed the mechanism for ionization of quintozene with APCI source in negative ion mode (12,13,15). The different ionization mechanisms of pesticides analyzed in hemp matrix with APCI ion source were determined by collecting first quad scan mass spectra over range of 50–500 Da of different analytes and are given below.
Where [M] stands for different pesticides, [A]– stands for an anion, [F]– stands for a fragment ion, and [N] stands for a neutral fragment. Note that we demonstrated the ionization mechanism for two more pesticides (captan and chlordane) not on Colorado’s list, but present in cannabis regulations for other U.S states, such as California and Florida.
A number of non-polar pesticides do not have low detection limits using LC–MS/MS systems with a conventional ESI source. Because we are using heated sheath of gas around electrospray source plume or droplets, we were able to ionize these compounds with good sensitivity in dried hemp. The detection limits of these non-polar pesticides (allethrin, bifenthrin, cyfluthrin, cypermethrin, deltamethrin, kinoprene, lambda-cyhalothrin, methoprene, MGK-264, permethrin, phenothrin, resmethrin, tetramethrin, and a few others) was in range of 0.01–0.2 ppm in dried hemp to meet Colorado action limits in hemp.
In this study, a novel LC–MS/MS method with a dual ESI and APCI source using simple solvent extraction for sample preparation was validated for analyzing 102 pesticides and five mycotoxins regulated by Colorado state in dried hemp. This method enabled the identification and quantification of all analytes with the LOQ in the range of 0.002 to 0.2 ppm, which meets the action limits set by the state of Colorado in dried hemp matrix. Unlike other published sample preparation methods, simple solvent extraction showed excellent extraction efficiency in the range of 80–120%, with RSD less than 20% for all of 102 pesticides and five mycotoxins. Hemp matrix is challenging and causes matrix effects and interference. We minimized matrix interference by optimizing both LC–MS/MS method and added 30 internal standards to our method to compensate for matrix effects and this helped us in attaining method accuracy in the range of 70–120%, with RSD less than 20%, for all of 102 pesticides and five mycotoxins. The method demonstrated good robustness, precision, high throughput, and reproducibility. The different ionization mechanisms of non-polar pesticides with APCI ion source was elucidated. The LC–MS/MS method worked well for wide range of polar and non-polar analytes including the analytes normally analyzed by GC–MS/MS. The ability to screen and quantitate all 102 pesticides and five mycotoxins, including the compounds analyzed on a GC–MS/MS system, makes this LC–MS/MS method a novel way to screen and quantitate pesticides and mycotoxins in dried hemp with a single instrument platform.
(1) The Farm Bill, Hemp Legalization and the Status of CBD: An Explainer. https://www.brookings.edu/blog/fixgov/2018/12/14/the-farm-bill-hemp-and-cbd-explainer/ (accessed 2022-07-27).
(2) Oregon Administrative Rules 333–007–0400 Pesticide Analytes and Their Action Levels. https://www.oregon.gov/oha/PH/DISEASESCONDITIONS/CHRONICDISEASE/MEDICALMARIJUANAPROGRAM/Documents/rules/333-007-0400_0410_0415_0425_Exhibit_A_eff_March312022.pdf (accessed 2022-07-27).
(3) California Code of Regulations, Title 16, Division 42, Bureau of Cannabis control, Article 5, § 5719. https://bcc.ca.gov/law_regs/readopt_text_final.pdf (accessed 2022-07-27).
(4) Florida State Regulations for Pesticide and Mycotoxin Analysis in Cannabis Flower. https://www.flrules.org/gateway/ruleNo.asp?id=64ER20-9 (accessed 2022-07-27).
(5) Health Canada, Mandatory Cannabis Testing for Pesticide Active Ingredients—List and Limits. https://www.canada.ca/en/public-health/services/publications/drugs-health-products/cannabis-testing-pesticide-list-limits.html (accessed 2022-07-27).
(6) Colorado Department of Public Health and Environment Regulations for Colorado Wholesale Food, Industrial Hemp, and shellfish 6 CCR 1010–21. https://drive.google.com/file/d/15kEVNcFFdP07OFaalZ8D5OG4KEStUxce/view (accessed 2022-07-27).
(7) Stenerson K. K.; Oden, G. Improved Workflow in the Analysis of Pesticide Residues in Cannabis by GC–MS/MS and LC–MS/MS. Cann. Sci. & Tech. 2018, 1 (1), 48–53.
(8) Kowlaski, J.; Dahl, J. H.; Rigdon, A.; Cochran, J.; Laine, D.; Fagras, G. Evaluation of Modified QuECHERS for Pesticide Analysis in Cannabis. LCGC North Am. 2017, 35 (5s), 8–22.
(9) Wang, X.; Mackowsky, D.; Searfoss, J. Telepchak, M. Novel LC–MS/MS Method with a Dual ESI and APCI Ion Source for Analysis of Califor- nia-Regulated Pesticides and Mycotoxins in Medium-Chain Triglyceride (MCT) Oil Cannabis Tinctures. LCGC North Am. 2016, 34 (10), 20–27.
(10) Moulins, J.R.; Blais, M.; Montsion, K.; Tully, J.; Mohan, W.; Gagnon, M. et al. Multiresidue Method of Analysis of Pesticides in Medical Cannabis. J. AOAC Int. 2108, 101 (6), 1948–1960. DOI: 10.5740/jaoacint.17-0495
(11) Dalmia, A.; Cudjoe, E.; Astill, T.; Jalali, J.; Weisenseel, J. P. et al. LC–MS/MS with ESI and APCI Sources for Meeting California Cannabis Pesti- cide and Mycotoxin Residue Regulatory Requirements. Cann. Sci. & Tech. 2018, 1 (3), 38–50.
(12) Dalmia, A. Systems and Methods for Pesticide Detection Using Mass Spectroscopy. US 10914713, 2021.
(13) Dalmia, A.; Johnson, C.; Hariri, S.; Jalali, J.; Cudjoe, E; Kingstad, J.; Qin, F. Novel LC–MS/MS Method with a Dual ESI and APCI Ion Source for Analysis of California-Regulated Pesticides and Mycotoxins in Medium-Chain Triglyceride (MCT) Oil Cannabis Tinctures. Curr Trends Mass Spec. 2020, 18 (3), 22–29.
(14) Dalmia, A.: Cudjoe, E.; Jalali, J.; Wu, J.; Hariri, S.; Guthrie, M. et al. A Novel ESI and APCI LC/MS/MS Analytical Method for Meeting the Canadian Cannabis Pesticide Residues Regulatory Requirements. 2019. https://resources.perkinelmer.com/lab-solutions/resources/docs/app_69954-cannabisconcentrates_final.pdf (accessed 2022-07-27).
(15) Dalmia, A.; Cudjoe, E.; Jalali, J.; Qin, F. A LC–MS/MS Method with Electrospray Ionization and Atmospheric Pressure Chemical Ionization Source for Analysis of Pesticides in Hemp. J. Cannabis Res. 2021, 3, 50. DOI: 10.1186/s42238-021-00106-9
(16) Michlig, N.; Lehotay, S.J.; Lightfield, A.R.; Beldoménico, H.; Repetti, M.R. Validation of a High-Throughput Method for Analysis of Pes- ticide Residues in Hemp and Hemp Products. J. Chromatography A 2021, 1645, 462097. DOI: 10.1016/j.chroma.2021.462097
(17) McEwen, C.N.;Larsen, B.S. Ionization Mechanisms Related to Negative Ion APPI, APCI, and DART. J. Am. Soc. Mass. Spectrometry 2009, 20, 1518–1521. DOI: 10.1016/j.jasms.2009.04.010
(18) Carroll, D.I.; Dzidic, I.; Stillwell, R.N.; Horning, M.G.; Horning, E.C. Subpicogram Detection System for Gas Phase Analysis Based Upon Atmospheric Pressure Ionization (API) Mass Spectrometry. Anal. Chem. 1975, 47 (8), 1308–1312. DOI: 10.1021/ac60342a009
Avinash Dalmia is a Senior Principal Application Scientist with PerkinElmer, in Shelton, Connecticut. Direct correspondence to: firstname.lastname@example.org