Choosing Analytical Tools to Assess Complex Cannabis-Infused Matrices

Published on: 
, ,
Cannabis Science and Technology, March/April 2018, Volume 1, Issue 1

Cannabis is currently one of the most widely studied plants in the world because of the medicinal and pharmaceutical benefits of its natural products. These benefits are primarily attributed to the presence of cannabinoids and terpenes, which can elicit myriad molecular responses. The increased interest in cannabis has led to a deeper knowledge of modern pharmacopeia, not only with respect to the all the possible medical benefits of cannabis constituents and their respective biochemical mechanisms of action, but also in reference to a comprehensive spectrum of its components within the contexts of discovery and quality control. Cannabis products can be found in many shapes and forms ranging from the raw plant material to extracts, which can be consumed directly or incorporated into a variety of products designed for human consumption. Such edible products can encapsulate cannabis constituents in a variety of different matrices, which require different sample preparation considerations because of their variable complexity and profiles of interferences. In many cases, edibles represent a significant challenge because there are many different types that can be prepared, each one of which may require different suitable sample preparation techniques to isolate the cannabinoids for an effective analytical determination. Here, different edibles have been considered based on their composition, and different sample preparation strategies are discussed that may be appropriate for analyzing different products. The choices for different sample preparation methods have been considered based on prior food chemistry literature. This information is adapted to provide recommendations for the targeted determination of cannabinoids within the context of each method, the matrix of interest, and subsequent choice of instrumental determination, such as liquid chromatography or gas chromatography. We focus on the different research approaches used in the past for the comprehensive analysis of cannabis and related products. Challenges and the possible solutions will be highlighted to provide insight into what the future may require in terms of more reliable methods for the characterization of cannabinoids in complex matrices, such as edibles.

As of 2017, 29 states in the United States plus the District of Columbia support the legal use of medicinal Cannabis sativa and Cannabis indica, eight of which also allow recreational use (1). These numbers illustrate the importance of performing proper quality assurance (QA) and quality control (QC) with a comprehensive analysis of all cannabis products that should include the cannabinoids (the main components of the plant that confer its psychoactivity and medical benefits) and terpenes, but also any potential contaminants, such as pesticides, growth regulators, heavy metals, microbes, pests, and residual solvents (Table I) (2-4).

More than 100 different putative cannabinoids have been discovered in cannabis, 10 of which are ubiquitously found in numerous cultivars (Table II lists the more prevalent cannabinoids) (1,5-10). Some of these are well known, such as Δ9-tetrahydrocannabinol (Δ9-THC, the main psychoactive component) and cannabidiol (CBD, has the widest range of reported medical benefits). However, it is believed that the therapeutic effects of cannabis are not exclusively attributable to any single cannabinoid, but rather to their synergistic effects when administered in different combinations and in concert with numerous terpenes (11).

Terpenes are a class of molecules found in cannabis that are primarily responsible for odor and fragrances, while exhibiting antimicrobial, antioxidant, and anti-inflammatory properties. More than 30,000 terpenes exist in the natural world, 100 of which have been detected in cannabis. These molecules all derive from isoprene subunits and can be categorized based on the number of C atoms: hemiterpenes (C5), monoterpenes (C10), sesquiterpenes (C15), diterpenes (C20), and sesteterpenes (C25) (1,12,13). Terpenes from the same categories may be differentiated based on the number of double bonds, but also based on the position of the double bonds or atom connectivity. This compound class is characterized by a large number of isomeric structures, which makes their individual speciation challenging. Another category of terpenes are the so-called “oxygenated terpenoids,” which differ because of the presence of one or more O atoms. Some of the most common terpenes are technically terpenoids, since they contain at least one O atom (for example, linalool belongs to this group) (12).

To satisfy consumers’ different tastes, there are different ways to consume or use cannabis, with the most popular means being through smoking plant material, utilizing sublingual tinctures, and consuming edibles (14). The latter is sometimes preferred because edibles are said to be more discreet for consumers, and the toxins linked to smoking can be avoided (14). Another reason why edibles are becoming so important is because of their longer-lasting psychoactive effects. Edibles usually provide peak effects at 2-4 h after ingestion, in contrast to the peak at 20-30 min following inhalation (14). The reason for this difference is the pathway by which Δ9-THC is metabolized throughout the human body. After consumption and once in the liver, Δ9-THC is hydrolyzed enzymatically (mainly, cytochrome P450) to 11-hydroxy-Δ9-tetrahydrocannabinol (11-OH-Δ9-THC) (14-16). This metabolite confers higher psychoactivity than Δ9-THC, and it is actually believed to be a more active form of the cannabinoid, since it is able to cross the blood-brain barrier more easily (14-16). Analysis performed on blood has shown that 11-OH-Δ9-THC is found at higher levels in the human body after cannabis ingestion compared to inhalation (14). However, this metabolic difference can result in undesirable effects. Because of the delayed onset of Δ9-THC during ingestion, inexperienced consumers can consume greater amounts than what is suggested, leading to adverse effects. As such, properly dosing of cannabis edibles, which is guided by the accurate quantification of the psychotropic cannabinoids, is of paramount importance for individuals who wish to retain the therapeutic benefits of cannabis without the potential pulmonary impacts that can result from smoking.

The scope of this review includes an overview of traditional analytical methods for cannabis natural products and contaminants of concern. Furthermore, different methods for extracting and analyzing cannabis edibles are discussed within the context of matrix complexity. Unfortunately, only a limited amount of literature exists on the topic of cannabis edibles analysis; however, a number of insightful parallels can be drawn from available literature in the food sciences. For a systematic approach, edibles have been classified into different subcategories based on the potential interferences they can contain. For example, gummy bears contain mainly sugars and glycerin, which ought to be well differentiable from cannabinoids, but might cause issues in terms of pesticides analysis. In contrast, cannabis-infused fermented beverages can contain ethanol, which can render the quantification of cannabinoids difficult. Further to this point, baked goods can contain a multitude of fats, proteins, fiber, and other ingredients that can hamper the extraction and analysis of targeted components. Additionally, the inherent hydrophobicity of cannabinoids can drive variable binding to individual compounds of edible products. As such, different matrices—and potential for heterogeneous cannabinoid distribution through an edible product—require different sample preparations and selective analytical platforms can also increase accuracy and precision of determinations. As the scope of products incorporating cannabinoids expands, so too must the analytical methods available to provide reliable quality control of these products.


Currently, there are many different methods used for the characterization of cannabis, but gas chromatography (GC) and liquid chromatography (LC) dominate the scene. In particular, LC is widely used for potency testing (1). Cannabinoids and their metabolites can be analyzed both on GC and LC, coupled with different detectors (for example, mass spectrometry [MS] for trace analysis). LC with ultraviolet (UV) detection is often sufficient for potency testing, especially for samples of direct plant extracts that can contain carboxylated and decarboxylated constituents. However, as the matrix becomes more complicated, a more selective detection for particular potency elements may be required. Tandem mass spectrometry (MS/MS) becomes quite attractive in this case, irrespective of whether analysis is performed by GC or LC. Recently, the potential for use of vacuum ultraviolet absorption spectroscopic detection in combination with GC (GC-VUV) has been demonstrated for cannabinoid speciation (17). GC-VUV offers some advantages for differentiation of isomers and deconvolution of coeluted peaks.

On the other hand, terpenes, as well as residual solvents, are best analyzed by GC because of their high volatility. Different detectors may be used, and usually MS is not required, though it does provide a high degree of qualitative information. Even so, the isomeric nature of terpenes can make them difficult to differentiate based on electron ionization mass spectra. To help alleviate this problem, GC-VUV has also been demonstrated for speciating terpenes (18,19). VUV absorption spectra have been shown in many cases to be highly complementary to mass spectra for the differentiation of isomer species (20).

A tricky class of compounds to address is pesticides. The methodology for the analysis of more than 500 pesticides in a vast array of food products and matrices exists and are used routinely worldwide. Virtually an entire conference, the North American Chemical Residues Workshop (, is dedicated to discussing and advancing the state-of-the-art in pesticide analysis. In the cannabis industry, pesticide testing is dictated by state regulations, and thus, lists of required pesticides can vary from state to state, with some lists still being developed and refined. To address the full complement of pesticides of concern worldwide, both LC- and GC-based methods are required because of the wide range of physicochemical properties, such as volatility, exhibited by different pesticide classes. With that being said, even restricted lists of target pesticides will require testing laboratories to have both LC and GC instruments available to test for an adequate range of compounds with the necessary sensitivity (19,21,22). Although it is important to verify the lack of pesticides in the original plant if it is to be consumed directly, further thought should be given to how the plant is processed to produce other products, and how different pesticides may transfer through the process. For example, many pesticides are thermally labile and can degrade if heat is used during the extraction process (23,24).

Heavy metals and microbes are also often tested in cannabis plants and extracts using techniques such as inductively coupled plasma-mass spectrometry (ICP-MS) for the former, and traditional biochemical methods or matrix-assisted laser desorption-ionization mass spectrometry (MALDI-MS) for the latter (1,25). Heavy metals accumulated from the environment can be concentrated from plant matter into different products during processing. On the other hand, a large majority of microbes do not survive many extraction or processing steps, especially if they involve organic solvents. Contamination by different fungi can be an alternate consideration, because these microorganisms can produce small-molecule toxins, which are more robust than pesticides and can be transferred through different extraction or processing steps (26).

Therefore, because of the different components that can be found in the plant, it is important to remember that one analysis method is not enough for comprehensive speciation of the different components of interest, because each analyte requires different preparation.


Edibles are growing in importance among cannabis consumers, not only for their easy preparation and consumption, but also for their longer-lasting effects (14). There are many different categories of edibles, several of which are prepared with cannabis oil or butter. Cannabis oil is prepared by extracting cannabis flowers with different solvents, such as naphtha, petroleum ether, ethanol, or olive oil (27,28). Romano and Hazekamp (27) performed an extraction comparison with these four different solvents, considering not only the recoveries of both cannabinoids and terpenes, but also the residual solvents. Their conclusion was that olive oil was the most efficient extraction solvent, not only because of its lack of toxicity and nonflammability, but also because it extracted the highest amount of terpenes.

The primary issue with the characterization of cannabis natural products in edibles is because of the varying matrix effects that can be associated with different final products. In edibles, all the normal food’s components must be taken into consideration, including fatty acids, sugars, sugar alcohols, proteins, fiber, and so on; therefore, effective sample preparation is mandatory before analysis.

A primary reason for the analysis of edibles is to verify label accuracy of products. In a 2015 research study, Vandrey and colleagues (29) found that Δ9-THC and CBD content was within 10% of the labeling of 75 analyzed products. They also found that 17% of the products were correctly labeled, 23% were underlabeled, and 60% were overlabeled with regard to Δ9-THC content. In earlier work, Pellegrini and colleagues (30) also analyzed different types of edibles (beer, liquor, oil) all derived from hemp. They extracted all the samples with hexane-isopropanol (9:1), looking exclusively at Δ9-THC, CBD, and cannabinol (CBN) content. They found that Δ9-THC had the highest concentration in all the samples, contrary to the prior literature. The authors suggested that this expression of Δ9-THC was more indicative of a cannabis phenotype than a hemp phenotype (30).


A different approach was taken by Zoller and colleagues (31). They analyzed different matrices comparing two different extraction methods. Hempseed oil and hemp tea were prepared from a methanol extract, and hempseed, biscuits, and herb were prepared from a 9:1 methanol-dichloromethane extraction mixture. They compared the analysis with high performance liquid chromatography (HPLC)-UV and HPLC with fluorescence detection, with the latter being extremely selective, making this detector a viable tool for the detection of Δ9-THC (31). Furthermore, none of the extracted samples showed interference peaks in the chromatogram, meaning that the extraction steps were effective.

Many other studies have been performed on cannabis oil, always using a solvent extraction protocol to quantify the amount of Δ9-THC (32,33). However, none of these studies aimed to determine impurities (such as pesticides) or trace cannabinoids. If the final aim is to focus more on the determination of trace impurities to guarantee safety of the product, then a more sensitive analysis method and a more efficient extraction technique is needed. A valid solution is the use of QuEChERS (quick, easy, cheap, effective, rugged, and safe) kits, a combined liquid-liquid extraction (LLE) and dispersive solid-phase extraction (dSPE) method that is already widely used for the analysis of pesticides (23,34,35). This kit also uses an extraction salt to improve the efficiency in case of water-soluble analytes, which also improves the cleanliness of the extract. Unfortunately though, the acetonitrile layer still contains matrix components such as fats, sugars, and pigments. For this reason, it is important to couple the extraction step with a sensitive and specific method that will allow the detection of low-concentration cannabinoids or pesticides. Either LC-MS/MS or GC-MS/MS (or both) could be viable or needed, depending on the specific targets.

In general, it could be useful to differentiate foods based on their matrix composition to choose the most suitable extraction technique. Tanner and colleagues (36) suggested the use of a triangle based on the normalized content of the three main components of food—namely fats, proteins, and carbohydrates (Figure 1)—to describe different food compositions. (See upper right for Figure 1, click to enlarge.) This scheme was then divided into nine sectors, each one with different concentrations of those components. The idea is that if one extraction method is suitable for a type of food, then it would be suitable for all the foods in the same sector.

Marazuela and colleagues (37) and Kinsella and colleagues (38) provided comprehensive reviews about the different extraction methods for food analysis, ranging from off-line extraction to sample purification. One particularly promising and popular technique is pressurized liquid extraction (PLE). This automated technique is carried out at higher temperatures than the boiling point of the solvent, by keeping it in the liquid phase with high pressure. This approach results in increased throughput and low solvent consumption, but the required equipment is fairly expensive. Furthermore, even if the conditions are optimized, some other matrix components may be extracted as well, therefore requiring a cleanup step before the analysis. The most used techniques in this case are solid-phase extraction (SPE), pre-PLE, and matrix solid-phase dispersion (MSPD). Pre-PLE could be performed with a nonpolar solvent to eliminate the hydrophobic interferents, while if the sample has a high concentration of fats, fat-retaining sorbents can be used (such as alumina or silica gel) (38).

Microwave-assisted extraction (MAE) is another common food extraction technique. MAE uses microwave energy to heat a solvent and extract the analyte from the matrix into the solvent in a confined vessel at a specified temperature (37). An average extraction is 15 min, and this approach offers higher throughput (parallel processing with many systems) and low solvent consumption (38). There are two modes of operation for MAE (focused or closed) that use open and closed vessels, respectively. The former works at atmospheric pressure, and the latter is performed under high pressures, making it similar to PLE, since the pressure can be increased with temperature (38). A drawback of this extraction is the difficulty of extracting analytes in matrices with more than 30% of water because of the limited diffusion of the solvent (37). Furthermore, even in this case, an additional cleanup step is usually required.

Another extraction technique that is making a resurgence because of the renewed availability of commercial instrumentation is supercritical fluid extraction (SFE). SFE uses the properties of a supercritical fluid (solvability and density of a liquid, with the viscosity and diffusivity of a gas) to extract the compound of interest (38). This method is useful because the solvating power can significantly improve by slightly changing the temperature and the pressure of the fluid. Usually, carbon dioxide (CO2) is widely used as a supercritical fluid because of its inertness, low cost, high purity, low toxicity, and low critical point conditions. However, if the analyte is strongly bound to the matrix, a polar additive can be added to the supercritical fluid. It is very common to add organic modifiers, such as methanol, to increase the extractability of more polar compounds (38). It has been shown by Veress that this technique extracts CBD faster than Δ9-THC, because it is dissolved more easily by supercritical CO2 (39,40). Diaz-Maroto and colleagues showed that the extraction of terpenes does not require the use of any modifier, as was expected because of the low polarity of these molecules (41).

As one of the most widely used sample preparation techniques, SPE is available in different formats, the most common being packed particle beds. As an alternate format, sorbent-impregnated disks are available that can withstand high flow rates, but these have been primarily used for environmental analysis (such as water sampling). It can be difficult to obtain consistent flow for automated cleanup (37,38). Dispersive-SPE, where functionalized solid particles are dispersed on solution, was mentioned previously as a common first step for QuEChERS extraction (38). Dispersive-SPE can be effective on its own, but requires a filtration step after extraction to remove the particles. Balancing the functionality of the sorbent with the physicochemical nature of the analytes (or interferences) desired to be removed is important as different edible compositions are considered.

Another type of SPE involves the use of molecularly imprinted polymers (MIPs), which are crosslinked polymers designed to exhibit high affinity toward specific compounds or a class of compounds (38). One of the biggest drawbacks is the need to prepare the highly selective material. It has also been reported that the potential for leaking can contaminate sample extracts (37), but this potential problem is likely because of the format of the material in some specific designs. Another alternative, matrix solid-phase dispersion was introduced in 1989. It combines homogenization, disruption, extraction, and cleanup in one process. The sample is blended with a dispersing agent, then packed into a column for extraction and cleanup (37). As of now, this technique has been mainly used in environmental, clinical, and food analysis, primarily for the analysis of antibacterial residues (42,43).

Restricted access material (RAMs) is another form of solid-phase extraction. RAM uses porous chromatographic supports to achieve extraction of small molecules, in the presence of protein or large molecule interferences (44,45). High-molecular-weight species are excluded from the pores based on a size-exclusion mechanism; macromolecules are eluted with the dead volume, while small molecules penetrate the pores and are retained by functional units (such as reversed phase or ion exchange) contained therein. As with traditional SPE, care needs to be taken in the use of RAM regarding the coextraction of interferences, such as fats, when targeting hydrophobic cannabinoid compounds.

Another molecular weight cut-off cleanup is ultrafiltration (UF), which is used to separate macromolecules, such as proteins, from the analyte of interest (38,46). Although they are effective, these membranes can be subject to fouling because of the precipitation of proteins, microorganisms, and fats, and therefore need to be constantly cleaned (47,48).

Turbulent flow chromatography (TFC), like RAM, has been primarily designed as an online sample preparation format. Thus, TFC gives both high sample throughput and high reproducibility because of automation (49,50). It consists of columns packed with large particle size particles, with which small analytes will interact. Turbulent flow is used to limit the interactions by large molecules (slow diffusion coefficients) with the sorbent. The drawback of this technique is the consumption of a high volume of solvent (38).

Finally, the last cleanup technique to consider is dialysis (38). It is not selective, but the extraction cell is easy to build, and the technique is efficient for the removal of macromolecules, since only small molecules are allowed to pass through the membrane (51,52). This technique can be used in edibles that have a high abundance of proteins, such as protein bars and smoothies. It is generally based on the removal of salts and sugars, but could be modified by using a membrane that is impermeable to the analyte of interest but permeable to interfering matrix components.

While sample preparation is an important step to avoid interferences, the instrumental analysis step can also be chosen so that it provides sufficient sensitivity and specificity. Because of the difficulties in removing interferences solely by sample preparation techniques, Leghissa and colleagues introduced the use of GC-triple-quadrupole mass spectrometry (GC-QqQ-MS) for the analysis of seven underivatized cannabinoids and eight silylated cannabinoids, reaching limits of detection (LODs) in the order of picograms on column (53). This technique is based on the fact that the focus of the analysis is on the specific transitions that each cannabinoid creates in the MS, so that the noise and impurities (compounds of non-interest) will not be detected. The advantage of using this method is the ability to analyze low-level cannabinoids even in a complex and dirty matrix, making it a viable tool for the analysis of impurities. As it was stated previously, the analysis of cannabinoids via GC-MS is easy and leads to high sensitivity and selectivity, but it must follow a derivatization reaction to protect the carboxyl groups of the acidic cannabinoids. The most common derivatization technique is silylation (53). The study of the multiple reaction monitoring (MRM) transitions of these compounds (both underivatized and in their sylilated forms) highlighted how cannabinoids undergo similar fragmentation pathways. As is shown in Figure 2, the most common fragmentation that underivatized cannabinoids undergo is the Retro-Diels Alder reaction followed by the loss of the C side chain. (See upper right for Figure 2, click to enlarge.) Because of the different starting molecular weights, this reaction leads to different m/z values. For the derivatized cannabinoids, on the other hand, the most recurrent steps are the loss of a trimethylsilyl group and of the C side chain (53).

Another way of analyzing cannabinoids is via LC-MS, which does not require a derivatization reaction before analysis, but is not suitable for the analysis of other components of cannabis that may be of interest (such as terpenes). Even with this instrument, high specificities can be achieved with the use of MRM mode.

Lastly, another approach that is gaining a lot of recognition in the pharmaceutical field is SFC-MS/MS because of its lower solvent use and wide application range. When coupled with SFE, this technique can also be used for the simultaneous analysis of analytes with different polarities, because of the possibility to adjust the extraction capability of supercritical CO2 by adding an organic modifier

Future Directions

The future of cannabis analysis still has an unknown fate, partially because of the fact that each state has its own regulations. In some cases, these regulations are still being formulated as the legality of medicinal and recreational cannabis use continues to evolve. Since edibles are becoming more prevalent among cannabis consumers, it is important for producers and laboratories to develop standardized tests not only for their potency, but also for all the other analytes of interest. Unfortunately, as of now there are no guidelines that specifically apply to edibles, and even cannabis regulations vary from state to state. The possibility for laboratories to use an approved protocol for the quality control and potency test is the first step for the manufacturing of cannabis products; however, additional efforts should be spent on the analysis of metabolites and pharmacokinetics upon ingestion, so that a greater degree of confidence can be had with edible versions of this new frontier of medicine.

Allegra Leghissa and Kevin A. Schug are with the Department of Chemistry and Biochemistry at The University of Texas at Arlington in Arlington, Texas. Zacariah L. Hildenbrand is with Inform Environmental, LLC, in Dallas, Texas. Direct correspondence to:


  1. A. Leghissa, Z.L. Hildenbrand, and K.A. Schug, J. Sep. Sci. 41, 398-415 (2018).
  2. M. Holmes, J.M. Vyas, W. Steinbach, and J. McPartland, Microbiological Safety Testing of Cannabis (Cannabis Safety Institute, 2015).
  3. Washington State Legislature, Chapter 314-55: Marijuana Processor License Extraction Requirements (Washington State Legislature, 2014).
  4. Chapter 467 “Residual Solvents” in United States Pharmacopeia 40-National Formulary 35 (United States Pharmacopeial Convention, Rockville, Maryland, 2007).
  5. H.N. Eisohly, C.E. Turner, A.M. Clark, and M.A. Eisohly, J. Pharm. Sci. 71, 319-1323 (1982).
  6. A.T. El-Alfy, K. Ivey, K. Robinson, S. Ahmed, M. Radwan, D. Slade, I. Khan, M.A. ElSohly, and S. Ross, Pharmacol. Biochem. Behav. 95, 434-442 (2010).
  7. F. Grotenhermen and K. Müller-Vahl, Dtsch. Arztebl. Int. 109, 495-501 (2012).
  8. J.A. Beutler and A.H. Marderosian, Econ. Bot. 32, 387-394 (1978).
  9. R. Brenneisen, Chemistry and Analysis of Phytocannabinoids and Other Cannabis Constituents (Humana Press, Totowa, New Jersey, 2007).
  10. E.T. Wargent, M.S. Zaibi, C. Silvestri, D.C. Hislop, C.J. Stocker, C.G. Stott, G.W. Guy, M. Duncan, V. Di Marzo, and M.A. Cawthorne, Nutr. Diabetes 3, E68 (2013).
  11. E.B. Russo, Br. J. Pharmacol. 163, 1344-1364 (2011).
  12. E. Breitmaier, Terpenes: Flavors, Fragrances, Pharmaca, Pheromones (John Wiley & Sons, Hoboken, New Jersey, 2006).
  13. T.K. Devon, Handbook of Naturally Occurring Compounds (Elsevier, New York, New York, 2012).
  14. D. Barrus, K. Capogrossi, S. Cates, C. Gourdet, N. Peiper, S. Novak, T. Lefever, and J. Wiley, Tasty THC: Promises and Challenges of Cannabis Edibles (Methods Rep RTI Press, 10.3768/rtipress.2016.op.0035.1611, 2016).
  15. C. Iribarne, F. Berthou, S. Baird, Y. Dréano, D. Picart, J.P. Bail, P. Beaune, and J. François Ménez, Chem. Res. Toxicol. 9, 365-373 (1996).
  16. T. Matsunagal, Y. Lwawaki, K. Watanabe, K. Yamamotol, T. Kageyama, and H. Yoshimura, Life Sci. 56, 2089-2095 (2012).
  17. A. Leghissa, J. Smuts, C. Qiu, Z.L. Hildenbrand, and K.A. Schug, Sep. Sci. Plus 1, 37-42 (2018).
  18. C. Qiu, J. Smuts, and K.A. Schug, J. Sep. Sci. 40, 869-877 (2017).
  19. K.A. Schug, I. Sawicki, D.D. Carlton, H. Fan, H.M. Mcnair, J.P. Nimmo, P. Kroll, J. Smuts, P. Walsh, and D. Harrison, Anal. Chem. 86, 8329-8335 (2014).
  20. L. Skultety, P. Frycak, C. Qiu, J. Smuts, L. Shear-Laude, K. Lemr, J.X. Mao, P. Kroll, K.A. Schug, A. Szewczak, C. Vaught, I. Lurie, and V. Havlicek, Anal. Chim. Acta 971, 55-67 (2017).
  21. J.W. Wong, M.G. Webster, D.Z. Bezabeh, M.J. Hengel, K.K. Ngim, A.J. Krynitsky, and S.E. Ebeler, J. Agric. Food Chem. 52, 6361-6372 (2004).
  22. L. Alder, K. Greulich, G. Kempe, and B. Vieth, Mass Spectrom. Rev. 25, 838-865 (2006).
  23. M. Anastassiades, S.J. Lehotay, and D. Stajnbaher, “Quick, easy, cheap, effective, rugged, and safe (QuEChERS) approach for the determination of pesticide residues,” presented at the 18th Annual Waste Testing and Quality Symposium in Arlington, Virginia, 2002, pp. 231-241.
  24. D.G. Hayward, J.W. Wong, F. Shi, K. Zhang, N.S. Lee, A.L. Dibenedetto, and M.J. Hengel, Anal. Chem. DOI: 10.1021/ac400481w (2013).
  25. I.C. Santos, Z.L. Hildenbrand, and K.A. Schug, Analyst 141, 2827-2837 (2016).
  26. H.S. Hussein and J.M. Brasel, Toxicology 167, 101-134 (2001).
  27. L.L. Romano and A. Hazekamp, Cannabinoids 1, 1-11 (2013).
  28. C. Citti, G.Ciccarella, D. Braghiroli, C. Parenti, M.A. Vandelli, and G. Cannazza, J. Pharm. Biomed. Anal. 128, 201-209 (2016).
  29. R. Vandrey, J.C. Raber, M.E. Raber, B. Douglass, C. Miller, and M.O. Bonn-Miller, J. Am. Med. Assoc. 313, 2491-2493 (2015).
  30. M. Pellegrini, E. Marchei, R. Pacifici, and S. Pichini, J. Pharm. Biomed. Anal. 36, 939-946 (2005).
  31. O. Zoller, P. Rhyn, and B. Zimmerli, J. Chromatogr. A 872, 101-110 (2000).
  32. C. Carcieri, C. Tomasello, M. Simiele, A. De Nicolò, V. Avataneo, L. Canzoneri, J. Cusato, G. Di Perri, and A. D’Avolio, J. Pharm. Pharmacol. DOI: 10.1111/jphp.12845 (2017).
  33. R. Pacifici, E. Marchei, F. Salvatore, L. Guandalini, F.P. Busardò, and S. Pichini, Clin. Chem. Lab. Med. 55, 1555-1563 (2017).
  34. S.J. Lehotay, J. Tully, A.V. Garca, M. Contreras, H. Mol, V. Heinke, T. Anspach, G. Lach, R. Fussell, K. Mastovska, and M. Poulsen, J. AOAC Int. 90, 485-520 (2007).
  35. A.K. Srivastava, P. Trivedi, M.K. Srivastava, M. Lohani, and L.P. Srivastava, Environ. Monit. Assess. 176, 465-472 (2011).
  36. J.T. Tanner, W.R. Wolf, and W. Horwitz, in Quality and Accessibility of Food-Related Data, H. Greenfield, Ed. (AOAC International, Sydney, Australia, 1993) pp. 99-104.
  37. M.D. Marazuela and S. Bogialli, Anal. Chim. Acta DOI: 10.1016/j.aca.2009.04.031 (2009).
  38. B. Kinsella, J. O’Mahony, E. Malone, M. Moloney, H. Cantwell, A. Furey, and M. Danaher, J. Chromatogr. A 1216, 7977-8015 (2009).
  39. T. Veress, J. Chromatogr. A 668, 285-91 (1994).
  40. C. Radcliffe, K. Maguire, and B. Lockwood, J. Biochem. Biophys. Methods 43, 261-272 (2000).
  41. M.C. Díaz-Maroto, M.S. Pérez-Coello, and M.D. Cabezudo, J. Chromatogr. A 947, 23-29 (2002).
  42. S. Bogialli and A. Di Corcia, J. Biochem. Biophys. Methods 70, 163-179 (2007).
  43. S.A. Barker, J. Chromatogr. A 885, 115-127 (2000).
  44. S.H. Yang, H. Fan, R.J. Classon, and K.A. Schug, J. Sep. Sci. 36, 2922-2938 (2013).
  45. P. Sadílek, D. Šatínský, and P. Solich, Trends Anal. Chem. 26, 375-384 (2007).
  46. X.U. Yan-yan, J. Fujian Fish 3, 8 (2006).
  47. K.J. Kim, P. Sun, V. Chen, D.E. Wiley, and A.G. Fane, J. Memb. Sci. 80, 241-249 (1993).
  48. M. Kazemimoghadam and T. Mohammadi, Desalination 204, 213-218 (2007).
  49. R.N. Xu, L. Fan, M.J. Rieser, and T.A. El-Shourbagy, J. Pharm. Biomed. Anal. 44, 342-355 (2007).
  50. P. Mottier, Y.A. Hammel, E. Gremaud, and P. Guy, J. Agric. Food Chem. 56, 35-43 (2007).
  51. R. Herraez-Hernandez, N.C. Van de Merbel, and U.T. Brinkman, J. Chromatogr. B Biomed. Sci. Appl. 666, 127-137 (1995).
  52. J.Å. Jönsson and L. Mathiasson, J. Sep. Sci. 24, 495-507 (2001).
  53. A. Leghissa, Z.L. Hildenbrand, F.W. Foss, and K.A. Schug, J. Sep. Sci. DOI: 10.1002/jssc.201700946 (2018).

How to Cite This Article

A. Leghissa, Z.L. Hildenbrand, and K.A. Schug, Cannabis Science and Technology 1(1), 24-35 (2018).