Characterization of Commercially-Available Herbal Incense Products Using QuEChERS Extraction and GCxGC-TOFMS Amanda Rigdon* 1 , Jack Cochran 1 , Paul Kennedy 2 . 1 Restek Corporation, Bellefonte, PA, 2 Cayman Chemical, Ann Arbor, MI Introduction In the past few years, a new product called ‘K2’, ‘spice’, or ‘legal marijuana’ has gained popularity. Although there are many products marketed as herbal incense, all are mixtures of herbs such as pink lotus, marshmallow, and baybean. These mixtures are not marketed for human consumption, but there have been many reports of cannabis-like highs when these products are smoked. Even though these blends are marketed as containing only natural products, some blends, such as Spice Gold, have been found to contain synthetic cannabinoids such as JWH-018, JWH-072, HU-210, as well as several other analogs. Because there is a large number of synthetic cannabinoids available, when a ‘spice’ sample is analyzed, it may not be known which analyte should be targeted for quantitation. Also, since new synthetic cannabinoids are being synthesized, seized samples may not contain any target analytes, but could contain other previously unknown synthetic cannabinoids. Since herbal incense is a mixture of unknown dried plant material, the matrix for sample analysis is complex, which could lead to problems identifying and quantifying analytes of interest. Several years ago, QuEChERS (Quick, Easy, Cheap, Effective, Rugged, Safe) was adopted by the food safety industry for the extraction of a wide range of pesticides and other residues from food matrices. While QuEChERS is relatively new to the field of forensics and forensic toxicology, this approach can also prove useful in the extraction of a wide range of drug products from complex matrices. When coupled with a comprehensive two-dimensional gas chromatographic (GCxGC) system and a time-of-flight mass spectrometer (TOFMS), drug compounds can be efficiently separated from complex matrices, and a large amount of information can be gathered on both the sample matrix, target analytes, and other unknown analytes that may be of interest. Instrumentation and Extraction Method During this project, 7 different samples of herbal incense from two vendors were extracted using QuEChERS and analyzed by GCxGC-TOFMS. QuEChERS is an acronym that stands for Quick, Easy, Cheap, Effective, Rugged, and Safe. QuEChers is a relatively new extraction method that has been recently adopted by the food safety industry for the analysis of pesticide residues in food. Because this extraction method was designed to extract a wide range of analytes from complex food matrices, it is useful for the extraction of possible unknown analytes from complex matrices such as incense and brownies. The extracts that are produced from the QuEChERS method can still be relatively dirty, so a dSPE (dispersive SPE) cleanup step can be employed to remove co- extracted matrix compounds, such as oils, sugars, and pigments. dSPE tubes contain sorbents such as C18, graphitized carbon black (GCB), and/or PSA (primary and secondary amine). C18 sorbents remove oils and hydrophobic interferences, GCB removes pigments and planar interferences, and PSA removes sugars and other compounds capable of hydrogen bonding. The QuEChERS extraction and dSPE cleanup process used for the herbal incense mixtures in this project is illustrated below: Weigh 1g of sample and place in 50mL centrifuge tube Wet sample with 9mL water and spike if necessary Add 10mL Acetonitrile, shake for ~30 sec. Add QuEChERS Salts. Shake for 1 min Centrifuge for 5 min @ 3000 x g If dSPE cleanup is required, place 1mL of extract into a 2mL dSPE tube containing the desired sorbent. Shake for 1 min and centrifuge for 5 min @ 3000xg. Figure 1. QuEChERS Extraction and Cleanup Process for Incense Samples The extracted samples were analyzed on a comprehensive two-dimensional gas chromatographic (GCxGC) system coupled with a time-of-flight mass spectrometer (TOFMS). The first dimension column used for analysis was a polar Rxi-17SilMS (equivalent to 50% phenylmethyl polysiloxane/ 50% dimethyl polysiloxane) and the second dimension column was an Rxi-1MS (100% dimethyl polysiloxane). This column combination allowed for good retention of the synthetic cannabinoids on both the first and second dimensions, while separating the more polar matrix compounds in the second dimension. Chromatography is illustrated in Figure 2. Conditions are shown below: Instrument: LECO Pegasus® GCxGC-TOFMS 1 st Dimension Column (in GC oven): Rxi-17SilMS, 15m x 0.25mm x 0.25μm 2 nd Dimension Column (in modulator): Rxi-1MS, 0.5m x 0.15mm x 0.15μm Transfer Line: Rxi-1MS, 0.2m x 0.15mm x 0.15μm Carrier Gas: Helium @ 2.2mL/min Injection Mode: Splitless @ 250°C Injection Volume: 1μL Liner: Single Gooseneck without wool Primary Oven Program: 90°C (hold 1 min), 5°C/min to 325°C (hold 3 min) Secondary (Modulator) Oven Parameters: +20°C relative to primary oven, modulation period 3 sec, hot pulse 0.90 sec, cold pulse time 0.60 sec. Transfer Line Temperature: 290°C Source Temperature: 225°C Scan Range: 75 – 550 amu Table 1. GCxGC-TOFMS Conditions Experimental Results Synthetic Cannabinoid Content: Significant levels of previously-characterized synthetic cannabinoids were detected in all of the samples analyzed. The level of each analyte was measured in a semi-quantitative manner against a mixture of reference standards offered by Cayman Chemical. Levels found in each sample can be found in Table 2, as well as levels of other synthetic cannabinoids found at much lower levels. Efficacy of dSPE cleanup: Three types of dSPE cleanup methods were investigated: 1) MgSO 4 + PSA; 2) MgSO 4 + PSA + C18; and 3) MgSO 4 + PSA + GCB. Although the sample matrix was extremely complex, due to the consistency of the sample (large, leafy pieces) and the fact that the synthetic cannabinoids were most likely applied to the surface of the sample, much of the matrix was not co- extracted during the QuEChERS extraction. This resulted in relatively ‘clean’ extracts. Additionally, the underivatized analytes retain extremely well on GC stationary phases, which acts to separate them from the matrix chromatographically. Because of this, it is not absolutely necessary to perform additional cleanup on extracted samples. That being said, the mixture of MgSO 4 + PSA + GCB cleaned a large amount of pigment out of extrcts, while the mixture containing C18 reduced the level of Vitamin E present in some of the samples. However, since the Vitamin E did not chromatographically interfere with any of the compounds of interest, cleanup of this compound is not strictly necessary. Presence of “Uncharacterized” Synthetic Cannabinoids: Perhaps the most interesting portion of this project was the characterization of several presumably undocumented synthetic cannabinoids. The presence of the undocumented cannabinoids is described in Figure 3. Matrix Elution Region Analyte Elution Region JWH-018 JWH-019 JWH-073 JWH-015 Unk. 1 Unk. 2 HU-331 (-) – CP 47, 497 HU-308 (+/-)-CP 47, 497 C8 Homologue CP 55, 940 JWH-250 JWH-015 JWH-018 JWH-019 JWH-200 Chromatogram of Standard Supplied by Cayman Chemicals JWH-073 Co-extracted matrix is mostly fatty acids Figure 2. Chromatogram of Voodoo Child Sample Table 2. Synthetic Cannabinoid Content for Incense Blends Sample Compounds Found Level (mg/g) Notes K2 Summit JWH-018 34.82 Also Contains Caffeine Puff JWH-073 0.01 JWH-018 12.30 K2 Standard JWH-018 15.23 Also Contains Vitamin E (+-)-CP 47, 497 C8 Homologue 11.05 JWH-018 0.02 Hypnotic JWH-015 0.01 Contains Green Dye that did not chromatograph JWH-073 0.08 JWH-018 34.80 Tribal Warrior (+-)-CP 47, 497 C8 Homologue 0.38 Also Contains Vitamin E JWH-018 39.96 Voodoo Child JWH-015 0.03 JWH-073 0.24 JWH-018 60.84 Investigation of Unknowns and Additional Work Investigation of Unknowns: Five unknown substances were detected in four of the incense samples. While not close to the level of the main active compound in the sample, the level of the unknowns was significant (2 – 12 mg/g). The spectra for four of the six unknowns were similar to the spectra of the known synthetic cannabinoids. The spectra of these compounds and comparison to known similar synthetic cannabinoids is shown in figure 3 below. Additional Work: Instead of using incense blends to experience cannabis-like highs, some users vaporize or otherwise ingest neat synthetic cannabinoids. According to users, the ingestion of high amounts of cannabinoids results in a psychadelic high rather like LSD, instead of the sedated state commonly experienced while smoking marajuana. Neat synthetic cannabinoids are readily available on the internet, and with the price of neat synthetic cannabinoid at ~ $35 USD/g, and a normal dose of 1 – 3 mg, the potential for abuse of neat substance is very high. Several samples of neat synthetic cannabinoids, along with two ‘proprietary blends’ were purchased, diluted, and analyzed in the same manner as the incense blends. All of the named synthetic cannabinoid samples contained the compounds as advertized at very close to 100% purity. The proprietary blends contained unknown compounds, some of which were similar or identical to the unknown compounds found in the incense samples. Voodoo Child Unknown 1 NOW720 MW 272 Voodoo Child Unknown 2 -Elutes immediately after JWH-018 JWH-018 Spectrum from Standard Puff Unknown 2 JWH-250 Spectrum from Standard Hypnotic Unknown 1 JWH-200 Spectrum from Standard Tribal Warrior Unknown 1 CP47, 497 Spectrum from Standard Figure 3. Investigation of Unknown Compounds