New perspectives on THCA decarboxylation and accurate GC–MS quantitation of Total THC in Cannabis using analyte protectants

OG Article

Jerome Mulloor, Walter B. Wilson, Lane C. Sander

06-06-2025

Abstract

Forensic and commercial laboratories rely on well-characterized analytical methods to accurately determine total tetrahydrocannabinol (total THC), which is the sum of decarboxylated tetrahydrocannabinolic acid (THCA) and delta-9-tetrahydrocannabinol (Δ9-THC) in Cannabis sativa samples. The Agriculture Improvement Act of 2018 restricts the level of total THC allowed in Cannabis for classification as hemp for commercial purposes. Gas chromatography with mass spectrometry detection (GC–MS) is frequently employed for Cannabis measurements due to its simplicity and speed of analysis. However, several issues complicate the determination of total THC by GC–MS, which are well-documented but unresolved. In the current study, the origins of potential GC–MS method biases are investigated, and novel approaches are presented to mitigate interferences. The behavior of THCA and Δ9-THC during GC–MS analyses was studied extensively using test solutions containing their isotopically labeled analogs. A plant matrix effect was identified that significantly increased THCA and Δ9-THC responses for Cannabis extracts compared with calibrant solutions. A mechanism is proposed based on the interaction of THCA and Δ9-THC with silanols present on heated inlet surfaces. The use of active site blocking agents, known as analyte protectants, reduced these interactions and achieved suitable conditions for quantitation of total THC by GC–MS. When Cannabis plant extracts and calibrants were processed under the recommended conditions, the results were comparable to liquid chromatography with photodiode array (LC-PDA) analysis. The experimental findings ultimately provide evidence to explain the behavior of cannabinoids in the GC–MS system and offer new options for improving the accuracy of total THC measurements.

Graphical abstract

Introduction

Cannabis sativa is the most commonly used recreational drug in the United States (U.S.) [1], and its increasing prevalence significantly impacts public health, medicine, criminal justice, industry, and forensics [2]. According to a 2022 survey, about 22 % of people (61.9 million) over the age of 12 in the U.S. reported using Cannabis in the past year [3]. Cannabis and its primary psychoactive constituent, delta-9-tetrahydrocannabinol (Δ9-THC), have been classified as illicit controlled substances in the U.S. since the 1970s [4]. However, legislative changes over time resulted in the decriminalization or legalization of Cannabis for medicinal and recreational purposes in many states [5]. Furthermore, the 2018 Agricultural Improvement Act (Farm Bill) defined hemp as Cannabis containing a mass fraction of less than or equal to 0.3 % total THC on a dry mass basis and removed hemp from the controlled substances list [6]. Total THC refers to the sum of decarboxylated delta-9-tetrahydrocannabinolic acid (THCA) and Δ9-THC (see Eq. 1). Consequently, as the legal landscape changed, an emerging Cannabis industry concurrently expanded with the marketing of cannabinoid-containing products, such as edibles, extracts, oils, and vapes [7].TotalTHC=0.877×THCA+Δ9−THC

Accurate and reliable methods for total THC analysis are indispensable to forensic, commercial, and compliance testing laboratories. Forensic laboratories are responsible for providing scientific evidence to support findings regarding the legality of seized evidence containing Cannabis plant material. Historically, qualitative test schemes, such as macro- and microscopic identification of plant features and colorimetric testing, were employed for this purpose [8,9]. After the Farm Bill was established in 2018, forensic laboratories adapted their procedures to measure total THC and determine whether the Cannabis sample is marijuana (a federally controlled substance) or hemp (a legal commodity) according to the 0.3 % total THC threshold [10]. Furthermore, manufacturers of cannabinoid-containing products created for consumers must provide accurate labeling for consumer safety and transparency [7]. Independent testing laboratories offer analytical services to evaluate their clients' products to ensure compliance with state and federal regulations. Therefore, establishing dependable methods for total THC analysis is paramount as the market for Cannabis-derived products grows.

The quantitative analysis of other drugs of abuse is commonplace in forensic chemistry, and the accumulated knowledge applies to Cannabis measurements as well. Liquid chromatography (LC) and gas chromatography (GC) coupled to mass spectrometry (MS) are perceived as “gold standard” methods for quantitation in forensic chemistry [11]. Both chromatographic techniques provide identification through retention time matching and quantitation using integrated peak areas. Mass spectrometry detection offers a high level of selectivity and specificity through the structural information in the mass spectra from the mass-to-charge (m/z) ratios of the fragment ions [11]. Seized drug sections within forensic labs generally prefer GC over LC due to the shorter chromatographic separation times and use of compressed gases instead of costlier solvents for the mobile phase [12]. Because of its sensitivity, robustness, familiarity, simplicity, and speed [13], a reliable GC–MS method for total THC analysis would undoubtedly provide value to forensic laboratories seeking updated quantitative procedures.

Unfortunately, several factors complicate GC-based measurements of total THC. Gas chromatography inherently requires high temperatures for sample introduction and elution. Under these conditions, THCA decarboxylates to Δ9-THC through the loss of CO2, accounted for by the conversion factor in Eq. 1 (see Fig. 1) [14]. At first glance, the conversion appears to simplify the analysis since a single chromatographic peak would determine the total THC (i.e., Δ9-THC). However, it is widely reported that THCA is incompletely converted to Δ9-THC by GC methods [15]. Dussy et al. reported 70 % conversion efficiency using a packed glass wool injector, with the extent of conversion varying with the injector temperature [16]. These observations naturally lead to the following questions: what happens to the remaining THCA if it's not fully converted to Δ9-THC? Since the liner material and inlet temperature are known to affect the decarboxylation rate [17], by what mechanisms does the liner interact with THCA? Given these issues, what instrumental conditions and analytical method approaches are needed to consistently produce accurate quantitation? Unless the decarboxylation process is fully understood, and all species are quantitatively accounted for, the potential for measurement bias cannot be eliminated.

In the present study, a GC–MS method was first developed to separate Δ9-THC from eight other cannabinoids in under ten minutes. Next, isotopically labeled analogs of THCA and Δ9-THC were utilized to study the decarboxylation of THCA under different instrumental conditions. During these experiments, a plant matrix effect for both compounds was discovered that was causing biases affecting quantitation. A new hypothesis was proposed that THCA and Δ9-THC adsorb to active sites in the GC system, which was mitigated by the presence of the Cannabis plant matrix. Further studies using analyte protectant compounds to block the active sites provided additional evidence supporting the proposed theory. Finally, a proof of concept was demonstrated for solving the challenges encountered in GC–MS measurements of total THC by utilizing analyte protectants and accurately quantifying total THC in three Cannabis plant sample extracts.

Section snippets

Chemicals and materials

Ampouled calibration stock solutions of cannabidivarin (CBDV), cannabicyclol (CBL), cannabidiol (CBD), tetrahydrocannabivarin (THCV), cannabichromene (CBC), delta-8-tetrahydrocannabinol (Δ8-THC), delta-9-tetrahydrocannabinol (Δ9-THC), cannabigerol (CBG), and cannabinol (CBN) were obtained from Cerilliant (Round Rock, TX, USA) for method development. Cannabinoid stock solutions (1 mg/mL) of THCA in acetonitrile (ACN), THCA-d3 in ACN, Δ9-THC in methanol (MeOH), and Δ9-THC-d3 in MeOH were obtained

GC–MS method development

This research was undertaken to support the development of an accurate and rapid GC–MS method to quantify Δ9-THC in forensic laboratories requiring high throughput analyses. Priority was given to the separation of Δ9-THC from eight other commonly encountered neutral cannabinoids at time of study, including CBDV, THCV, CBL, CBC, CBD, Δ8-THC, CBG, and CBN [14,23]. A stock mixture containing equal quantities of each cannabinoid (20 μg/mL) was prepared in MeOH for method development. For

Conclusions

The accuracy of methods for the GC–MS analysis of Cannabis is influenced by potential biases resulting from analyte adsorption on active surfaces and the quantitative conversion of THCA to Δ9-THC through decarboxylation. Improved measurement accuracy is achieved by implementing measures that reduce or eliminate surface activity often attributed to silanols. A test solution containing Δ9-THC-d3 and Δ9-THC-d9 provided unique capabilities to study adsorption and decarboxylation phenomena since all

CRediT authorship contribution statement

Jerome Mulloor: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Walter B. Wilson: Writing – review & editing, Funding acquisition. Lane C. Sander: Writing – review & editing, Methodology.

Disclaimer

Certain commercial equipment or materials are identified in this paper to adequately specify the experimental procedure. Such identification does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the materials or equipment identified are necessarily the best available for the purpose.

Funding

This work was supported by the NIJ-DOJ (Grant No. DJO-NIJ-20-RO-0009).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We thank Frances Scott and Megan I. Chambers, colleagues at the National Institute of Justice, Office of Justice Programs, and U.S. Department of Justice (NIJ-DOJ), for their support and assistance with this publication. We also thank Catherine A. Rimmer, colleague at NIST.