Ethyl acetate in e-liquids: Implications for breath testing

Abstract

Electronic cigarette liquids (e-liquids) can contain a variety of chemicals to impart flavors, smells and pharmacological effects. Surveillance studies have identified hundreds of chemicals used in e-liquids that have known health and safety implications. Ethyl acetate has been identified as a common constituent of e-liquids. Ethyl acetate is rapidly hydrolyzed to ethanol in vivo. Animal studies have demonstrated that inhaling >2,000 mg/L ethyl acetate can lead to the accumulation of ethanol in the blood at concentrations >1,000 mg/L, or 0.10%. A “Heisenberg” e-liquid was submitted to the Laboratory for Forensic Toxicology Research for analysis after a random workplace drug test resulted in a breath test result of 0.019% for a safety-sensitive position employee. Analysis of this sample resulted in the detection of 1,488 ± 6 mg/L ethyl acetate. The evaluation of purchased “Heisenberg” e-liquids determined that these products contain ethyl acetate. The identification of ethyl acetate in e-liquids demonstrates poor regulatory oversight and enforcement that potentially has consequences for breath ethanol testing and interpretations. The accumulation of ethanol in the breath from the ingestion/inhalation of ethyl acetate from an e-liquid used prior to a breath test may contribute to the detection of ethanol.

Introduction

Electronic cigarette (e-cig) use is pervasive around the globe. The liquid formulations (e-liquids) used in an e-cig can contain a variety of chemicals including flavorants, solvents, stabilizers, pharmacologically active compounds and chemicals with unknown purposes (1). In the USA, the US Food and Drug Administration (FDA) regulates e-liquid composition. In 2016, the FDA announced their regulatory control of e-cigs when they issued the “Final Deeming Rule” (2). A flavor ban was instituted in 2020 as an attempt to mitigate vaping appeal to youth (3), and in 2022, the FDA’s regulatory control over e-cigs was expanded to include synthetic nicotine, which manufacturers had been using to circumvent regulations (4). Despite efforts to control the e-cig industry, many products remain available through brick and mortar and online retails shops that are not authorized by the FDA (5).

Most of the chemicals identified in e-liquids are considered generally recognized as safe (GRAS) by the FDA, but that designation only applies to foodstuffs to be consumed orally (6). Hundreds of chemicals identified in e-liquids have known health and safety implications, including cannabinoids, synthetic cannabinoids, synthetic cathinones, dextromethorphan, gamma-butyrolactone and solvents such as methanol, acetone, isopropanol and ethanol (1, 7–11). Lack of e-liquid manufacturing quality assurance and compliance is often demonstrated through inaccurate, incomplete or absent product labeling.

Propylene glycol (PG) and vegetable glycerin (VG) are the most common carriers for e-liquid ingredients. Ethanol has also been identified as a common component in e-liquids, in concentrations of ≥20% w/v (1, 7) but is rarely identified on the label. One study reported 95% of 56 e-liquids contained ethanol, with 1.8% of the products labeled as containing ethanol (7). Ethanol is used as a solvent for other ingredients, a thinning agent, or as an active ingredient. The presence of ethanol in e-liquids has led to concerns regarding ethanol impairment and abstinence testing and forensic, clinical, and workplace drug testing.

The possible effects on roadside impairment tests, the preliminary breath test (PBT), and the evidentiary breath test (EBT) caused by vaping an ethanol containing e-liquid have been evaluated (12). In that study, e-liquids containing 0% or 20% ethanol in a 50:50 PG:VG carrier were used. PG and VG were found not to interfere with either the PBT or EBT. This study also demonstrated that the use of an e-liquid containing 20% ethanol could result in ethanol detection by PBT within 5 min of vaping. Wait periods are usually observed in driving under the influence investigations but are not necessarily observed in other breath alcohol testing scenarios, such as court-mandated testing, workplace testing, substance use treatment and abstinence programs, or vehicle ignition interlock systems.

Ethyl acetate, a carboxylate ester of acetic acid and ethanol, is another solvent identified in e-liquids (13). Studies suggest that ethyl acetate is frequently present in e-liquids, sometimes detected in as much as 95% of analyzed samples (14–17), and concentrations of ethyl acetate in e-liquids have been reported which exceed 10 g/L (16). Ethyl acetate has a number of industrial uses, including use as a solvent (18, 19), an FDA-approved GRAS food additive (20) and a flavoring chemical (21). However, the PubChem database characterizes ethyl acetate as a flammable and an irritant (22). Inhalation is reported as the primary route of absorption responsible for ethyl acetate toxicity. Ethyl acetate damages lung and mucosal tissue, demonstrated by edema and microscopic bleeding (23), as well as the liver and kidney. Other reported health hazards include headache, respiratory and eye irritation, dizziness, nausea, weakness, loss of consciousness and death (22). Ethyl acetate has been reported in e-liquids in concentrations above the Occupational Safety and Health Administration limits (24). Occupational Safety and Health Administration recommends <8 h weighted average of 400 ppm for “recommended and permissible limits,” and suggests that 2,000 ppm is immediately dangerous to life or health (24).

Effects following the ingestion of ethyl acetate are not well documented. In a study involving rats, ethyl acetate was hydrolyzed to ethanol in vivo in 5 to 10 min (25). Inhalation of >2,000 mg/L ethyl acetate led to an accumulation of ethanol in the blood. Blood ethyl acetate concentrations were determined to be <100 mg/L, while blood ethanol concentrations exceeded 1,000 mg/L. The authors contended that while the results could not be extrapolated to humans due to interspecies differences, similar results were likely. A study investigating the metabolism of inspired ethyl acetate in rats and hamsters demonstrated that ethyl acetate hydrolysis was caused by carboxylesterases, which are present in nasal and respiratory tissue (26). A case of an accidental death following acute intoxication of ethyl acetate reported ethyl acetate concentrations ranged from none detected to 2,217 mg/L and ethanol concentrations ranged from 740 to 2,990 mg/L in tissue (23). The highest concentration of ethyl acetate was determined in the testis, while the highest ethanol concentration was determined in lung tissue. The blood ethyl acetate and ethanol concentrations were determined to be 38 mg/L and 2,020 mg/L, respectively. A near fatal ethyl acetate poisoning by intentional consumption of ∼100 mL of 85% ethyl acetate nail polish remover resulted in the rapid release of acetic acid, causing acidosis and hepatocellular damage (27).

The case history and analysis of commercially available e-liquids obtained as part of an investigation following the detection of ethanol by breath test post-vaping are presented. Headspace gas chromatography–flame ionization detection coupled with mass spectrometry (HS-GC–FID–MS) identified and quantitated 1,407 to 1,977 mg/L ethyl acetate in the e-liquids.

Case history

An adult male working in a safety-sensitive position was randomly drug tested. While driving to the testing site, a Heisenberg “The Berg” Menthol 0 mg nicotine e-liquid was reported as being used for the duration of the trip. A Phoenix^® 6.0BT Evidential Workplace Breath Alcohol Tester (Lifeloc Technologies, Wheat Ridge, CO, USA) identified a breath alcohol concentration (BrAC) of 0.019%, resulting in work suspension. No information was provided as to the amount of e-liquid vaped or the time between last use of the e-cigarette and ethanol breath test. A duplicate of the e-liquid product used on the day in question was submitted to the Laboratory for Forensic Toxicology Research (LFTR) at Virginia Commonwealth University for analysis. Four other identical e-liquids with different lot numbers and one other “Heisenberg” e-liquid were purchased. The identification of ethyl acetate in the e-liquids served to revoke the work suspension.

Materials and methods

Materials

Type 1 water was generated in-house using a Millipore Direct-Q3 system. Air, helium, hydrogen and nitrogen gases were purchased from AirGas (Richmond, VA, USA). HPLC-grade acetone and ethyl acetate and Optima grade isopropanol and methanol were purchased from Fisher Scientific (Hanover Park, IL, USA). Ethanol, 200-proof, was purchased from Decon Labs (King of Prussia, PA, USA). Tertiary butyl alcohol (t-butanol) was purchased from Honeywell Riedel-de Haën (Seelze, Germany). United States Pharmacopeia grade PG and VG were purchased from Wizard Labs (Altamonte Springs, FL, USA). Amitriptyline, diazepam, fluoxetine, methadone, nicotine, nordiazepam, norfluoxetine, nortriptyline, paroxetine and trazodone-certified reference materials were acquired from Cerilliant (Round Rock, TX, USA) to create an in-house made quality assurance test mix, combined at 20 µg/mL in methanol. Residual Solvents #1 (Cat# 34105) and Class 3—Mix A (Cat# 36013) standards were purchased from Restek (Bellefonte, PA, USA) to use as interference standards.

The case sample, one Heisenberg “The Berg” Menthol 0 mg nicotine e-liquid, was received by the LFTR for testing. Four other identical e-liquids (different lots) and one other “Heisenberg” e-liquid manufactured by Innevape were purchased from four different online vendors by the LFTR for characterization and comparison to the case sample (Table I).

Table I.

Results of ethyl acetate quantitation

					% Difference
	Lot	Average EtAC (mg/L)	SD	%CV	Within lot	Total average	Menthol average
Heisenberg “The Berg” Menthola	70002996	1,488	6	0.4	N/A	−7	−2
Heisenberg “The Berg” Menthol	70003481	1,629	45	2.8	3	2	7
Heisenberg “The Berg” Menthol	70003481	1,584	2	0.1		−1	4
Heisenberg “The Berg”	70003475	1,976	90	4.5	N/A	24	30
Heisenberg “The Berg” Menthol	70002882	1,407	9	0.6	N/A	−12	−7
Heisenberg “The Berg” Menthol	No stamp	1,482	17	1.1	N/A	−7	−2

^aCase sample.

Methods

Product investigation

Photographs were taken of all packaging and associated components of the samples. Manufacturer and third-party retailer websites were evaluated for advertising claims about the product. Associated certificates of analysis and other product information, including customer reviews, were also documented.

GC–MS and HS-GC–dual FID analysis

The e-liquids were analyzed using previously published methods (1). A Shimadzu QP-2020 gas chromatograph–mass spectrometer (GC–MS) was used to identify carriers, flavoring chemicals and other ingredients. Chromatographic quality was confirmed using a lab-made test mix combination of 10 different drugs. Volatiles (acetone, ethanol, isopropanol and methanol) analysis was accomplished using a headspace gas chromatography–dual flame ionization detector (HS-GC–dual FID) employing a Shimadzu HS-20 headspace sampler attached to a Nexis 2030 GC–dual FID controlled by LabSolutions software (Shimadzu Corp., Kyoto, Japan).

Identification and quantitation of ethyl acetate by HS-GC–FID–MS

Analysis was performed using a modified version of the HS–GC-dual FID method, adapted for use on a Thermo Fisher Scientific TriPlus 500 headspace autosampler coupled with a Trace 1310 gas chromatograph, an instant-connect FID, and an ISQ 7000 MS (Thermo Fisher Scientific, Austin, TX, USA). Samples were incubated for 15 min at 70°C with the shaker on “fast” mode. The sample loop temperature was set to 70°C, and samples were injected over 0.5 min using a 1:15 split and a 200°C injection temperature. Chromatographic separation was achieved using a Thermo Fisher TG-ALC Plus II column (30 m × 0.32 mm id × 0.60 μm) with a column flow of 15 mL/min and an oven temperature of 50°C. Following separation, samples entered a splitter and proceeded to the FID and MS. The FID was set to 300°C with a collection rate of 15 Hz. The MS transfer line and ion source temperatures were set to 200°C and 250°C, and the scan range was m/z 30–300. Data were processed using Thermo Fisher Chromeleon chromatography data system version 7.2.10 and a National Institute of Science and Technology (NIST) library for mass spectral identification.

Calibrators were prepared at 100, 200, 500, 900, 1,500, 3,000 and 5,000 mg/L of ethyl acetate in water. A standard was prepared at 10,000 mg/L of ethyl acetate in water to evaluate the upper limit of linearity (ULOL). Matrix matched quality controls were prepared at 300, 750 and 4,000 mg/L ethyl acetate in-house using a 30:70 PG:VG mixture. Calibrators, controls in triplicate, blank matrix with and without internal standard, and samples were prepared by adding 100 µL to a headspace vial with 1 mL internal standard (90 mg/L t-butanol in water). Interferences were investigated by injecting standards of Restek Class 3—Mix A and Restek Residual Solvents #1 reference materials. Analysis was performed using the HS-GC–FID–MS method described above.

Method validation for the quantitation of ethyl acetate was performed by analyzing calibrators, controls in triplicate and interference standards over 3 days. The HS-GC–FID–MS method for the quantitation of ethyl acetate was determined to have a linear range of 100–10,000 mg/L and a coefficient of determination (r²) value ≥0.9990. The limit of quantitation was administratively set at 100 mg/L. The controls, ULOL standards, and control dilutions were determined to be within ±9.6%, ±2.9% and ±6.8% of their expected concentrations. The intra-day and inter-day precision (percent coefficient of variation, %CV) was determined to be within ±1.9% and ±6.1%. Carryover was assessed by injecting a negative control without internal standard after the highest calibrator, ULOL standard, interference standards, and between controls and product samples. No carryover was detected in any negative control. No interferences were detected in the Restek Residual Solvents #1 or Class 3—Mix A standards, which contained 20 and 24 analytes at 3,000 mg/L and 5,000 mg/L, respectively (Supplementary Figures 1 and 2). Peak co-elution was not noted in the MS or FID chromatographs. Calculated ethyl acetate concentrations in both standards were within 20% of the manufacturers reported values.

Results

The product submitted for analysis was labeled as “0 mg nicotine,” however, the back label of the product listed “Vegetable Glycerin, Propylene Glycol, Nicotine & Artificial Flavorings” as ingredients (Supplementary Figure S3). No further information was available from the manufacturer, as the manufacturer website was an inactive webpage (28). Product claims listed on third-party websites stated that the size of the e-liquid bottle had been increased to leave 20 mL of excess room to allow the consumer to “mix in whatever you need!” Analysis by GC–MS identified PG, VG, menthol and other minor flavorants (Figure 1). Nicotine was not identified. Analysis by HS-GC–dual FID did not identify acetone, ethanol, isopropanol or methanol. Ethyl acetate was identified on the HS-GC–dual FID by retention times and use of an ethyl acetate reference standard. Ethyl acetate was identified on the HS-GC–FID–MS using the NIST library and an ethyl acetate standard that allowed for the comparison of the retention time and mass fragmentation patterns (Figures 1 and 2).

Figure 1.

(A) Total ion chromatogram from untargeted GC–MS analysis; (B) chromatograph from volatiles analysis using HS-GC–FID–MS, with mass spectra of unknown peak and an ethyl acetate standard.

Figure 2.

Case sample and ethyl acetate standard chromatograms from HS-GC–dual FID.

The analysis of the purchased Heisenberg e-liquids produced results similar to those of the case sample, with consistent product composition between lots except for the lack of menthol in the non-mentholated product. The front product labels and websites indicated that the purchased e-liquids were nicotine-free formulations, and the back labels of the products listed nicotine as an ingredient. Nicotine was not identified in any purchased product. Acetone, ethanol, isopropanol and methanol were not detected in any product. Ethyl acetate was identified in every purchased product.

The average ethyl acetate concentration in the case sample was determined to be 1,488 ± 6 mg/L. The other Heisenberg e-liquids were determined to have ethyl acetate concentrations ranging from 1,407 to 1,977 mg/L. Results of each product can be found in Table I.

Discussion

The e-liquids were labeled on the front “0 mg nicotine,” and on the back “PG, VG, nicotine and flavorings.” Nicotine was not identified in any of the e-liquids. No pharmacologically active controlled substances, including ethanol, were identified in the e-liquids. Unlabeled ingredients included ethyl acetate. Other compounds identified were flavoring chemicals and carriers that have been previously identified in other e-liquids (1). The carriers were PG and VG, which do not interfere with breath testing instruments (12).

Ethanol can accumulate in the blood following ingestion of ethyl acetate. Ethyl acetate in e-liquids may contribute to a positive breath ethanol result, depending on the amount vaped, the time between vaping and the administration of the breath test, and the rate of metabolism of ethyl acetate to ethanol via carboxylesterases and other non-specific esterases present in lung tissue (26, 29). Inhalation of 5,000 mg/L ethyl acetate for 5 h resulted in 0.06% ethanol in rat blood (25). Inhalation of 10,000 mg/L ethyl acetate resulted in more rapid ethanol accumulation, with >0.10% ethanol in blood, and the death of one animal attributed to severe respiratory depression. Additionally, a postmortem case noted that lung tissue contained greater concentrations of ethanol than in other tissues following ethyl acetate hydrolysis (23). Lung tissue concentrations of 35 mg/L ethyl acetate and 2,999 mg/L (0.299%) ethanol were determined.

Holt et al. demonstrated that soon after vaping (<5 min), 1 and 10 puffs of a vaped 20% ethanol e-liquid can produce a positive BrAC result of 0.03% and 0.074%, respectively (12). Oral consumption of the small amount of ethyl acetate in the e-liquid would not lead to detectable BrAC following absorption of ethyl acetate and subsequent metabolism and distribution. However, germane to this case is the hydrolysis of ethyl acetate directly in lung tissue, which could allow for the near immediate (~5 min) production of detectable ethanol in exhaled breath prior to absorption and equilibration.

The positive breath test result may have been influenced by ethyl acetate in the breath sample. A previous study has shown that low concentrations (<0.10 mg/L) of ethyl acetate could influence ethanol readings without triggering an “interfering compound” message using a Dräger 7110 Evidential breath analyzer (30). The addition of 0.10 mg/L of ethyl acetate to a known ethanol concentration increased the ethanol results by 0.07 mg/L. Ethyl acetate from inhaled glue has been reported to have possibly contributed to a positive breath sample (31). These authors are unaware of any ethyl acetate cross-reactivity studies for the Phoenix 6.0 Life Loc-BT.

Conclusion

The detection of unlabeled ethyl acetate in the “Heisenberg” e-liquids demonstrates poor labeling practices in the e-cigarette industry. The evaluation of several “Heisenberg” e-liquids determined that these products contain ethyl acetate, which can convert to ethanol in vivo. This provided a case for an affirmative defense of a positive breath alcohol result of 0.019%, even though the “Heisenberg” e-liquid used was not analyzed. Studies are needed to elucidate the correlation between inhaled ethyl acetate and in vivo ethanol concentrations and to determine the impact of vaping ethyl acetate on ethanol testing.

Supplementary data

Supplementary data is available at Journal of Analytical Toxicology online.

Data availability

The data underlying this article are available in the article and in its online supplementary material.

Funding

This work was supported by the National Institute of Justice (2019-MU-MU-0007) and the National Institute of Health: National Institute on Drug Abuse (P30 DA033934). The opinions, findings and conclusions or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect those of the Department of Justice.