- Split View
-
Views
-
Cite
Cite
Nicolas Beauval, Sébastien Antherieu, Mélissa Soyez, Nicolas Gengler, Nathalie Grova, Michael Howsam, Emilie M Hardy, Marc Fischer, Brice M.R. Appenzeller, Jean-François Goossens, Delphine Allorge, Guillaume Garçon, Jean-Marc Lo-Guidice, Anne Garat, Chemical Evaluation of Electronic Cigarettes: Multicomponent Analysis of Liquid Refills and their Corresponding Aerosols, Journal of Analytical Toxicology, Volume 41, Issue 8, October 2017, Pages 670–678, https://doi.org/10.1093/jat/bkx054
- Share Icon Share
Abstract
Electronic cigarette use has raised concern worldwide regarding potential health risks and its position in tobacco cessation strategies. As part of any toxicity assessment, the chemical characterization of e-liquids and their related vapors are among fundamental data to be determined. Considering the lack of available reference methods, we developed and validated several analytical procedures in order to conduct a multicomponent analysis of six e-liquid refills and their resultant vapor emissions (generated by a smoking machine), and compared them with tobacco smoke. We combined several techniques including gas-chromatography, high and ultra-performance liquid chromatography and inductively coupled plasma with mass spectrometry or ultraviolet and flame ionization detection in order to identify the main e-liquid constituents (propylene glycol, glycerol and nicotine), as well as multiple potentially harmful components (trace elements, polycyclic aromatic hydrocarbons (PAHs), pesticides and carbonyl compounds). Regarding propylene glycol, glycerol and nicotine concentrations, the six tested e-liquids comply with the advertised composition and contain only traces of pollutants. Noticeable lower concentrations of trace elements (≤3.4 pg/mL puff), pesticides (<LOQ), PAHs (≤4.1 pg/mL puff) and carbonyls (≤2.11 ng/mL puff) were measured in e-vapors compared to those in cigarette smoke (up to 45.0 pg/mL puff, 8.7 pg/mL puff, 560.8 pg/mL puff and 1540 ng/mL puff, respectively). Although an accurate characterization of electronic cigarette emissions requires further analytical optimizations, our results have shown that vaping exposes the user to lesser amounts of selected toxic components of concern found in some representative French e-cigarette products than does smoking typical conventional cigarettes.
Introduction
Tobacco use is one of the major public health concern, causing nearly 6 million deaths annually worldwide, and a predicted death toll of 1 billion within the 21st century (1). Smoking is cited as the most important risk factor for Chronic Obstructive Pulmonary Disease (COPD), cardiovascular disorders and cancers. The World Health Organization (WHO) estimates that smoking costs more than half a trillion dollars each year (2). Smoking cessation is currently the only effective way to slow down the progression of these diseases and to reduce mortality.
The use of electronic cigarettes (“e-cig”) can be considered as an alternative to smoking. Briefly, consumers, or “vapers”, breathe in vapors (or “e-vapors”) produced by the heating of an e-liquid (a mixture of propylene glycol, glycerol and flavorings, supplemented or not with different concentrations of nicotine). E-cigs are generally considered as less harmful than tobacco and has already been recommended by several practitioners in smoking cessation since few years (3, 4). However, it has been reported than the reliable scientific or regulatory information about e-cigs is not widely known (4, 5). In addition, the absolute safety of such products cannot be guaranteed in the absence of a thorough clinical evaluation and a long-term population surveillance.
The toxicity of the main constituents of e-liquids (except nicotine) has to date been poorly studied. For propylene glycol, animal experiments show minor acute and chronic toxicity per os or by inhalation, except for very high doses (6). Propylene glycol is considered to be of low toxicity for humans and professional exposures can give rise to skin irritation or allergic manifestations (6). Concerning glycerol, its acute toxicity seems to be limited to irritation of the skin, eyes and respiratory tract. Renne et al. (7) showed that rats exposed to chronic inhalations of glycerol exhibited minimal to mild squamous metaplasia of the epithelium lining the base of the epiglottis.
Regarding e-liquid flavorings, thousands are actually available on the market (8), exposing consumers to a broad range of chemicals. These are mostly food flavors and are considered as safe for ingestion by authorities. However, they could be at sufficiently high concentrations in refill fluids to be of toxicological concern by inhalation (9). In recent experimental studies, the cytotoxic effects of e-liquids or their aerosols have been shown to be correlated with, or even restricted to, some flavorings (10). Moreover, several studies report the presence of some potentially carcinogenic substances in e-vapors, including carbonyl compounds, volatile organic compounds and nitrosamines (11), which could come from either contamination or molecular transformation. The presence of several trace elements was also reported in e-cig emissions, likely released by cartomizer components (12).
Considering the potential help that e-cig could represent for smoking cessation, the considerable lack of knowledge about long-term effects of e-cig use on human health and the large number of e-cig users, accurate and robust chemical characterization of e-cig emissions are among critical data to be determined in order to assess e-cig toxicity.
The aim of the present study was to develop and test several analytical methods in order to achieve a multicomponent analysis of six e-liquids and their corresponding aerosols. Our purpose was: (i) to check the concentration of the components indicated on e-liquid labels; (ii) to evaluate the potential contamination of these e-liquids by toxic compounds; and (iii) to analyze and compare the chemical composition of both e-cig and conventional cigarette emissions generated via a smoking machine.
Materials and Methods
E-liquids, e-cigarettes and conventional cigarettes
Six e-liquids and one model of e-cigarette (NHOSS® brand) were obtained from a French manufacturer (Innova SAS, Bondues, France). Two flavored e-liquids among NHOSS® best-sellers and a “control” flavorless e-liquid (unavailable on the market), with and without nicotine, were analyzed: chlorophyll mint flavor without nicotine (CM/Nic−) or with 16 mg/mL nicotine (CM/Nic+); blond tobacco flavor without nicotine (BT/Nic−) or with 16 mg/mL nicotine (BT/Nic+); and a mixture of propylene glycol/glycerol without flavoring, produced by the manufacturer under the same conditions as marketed e-liquids, without nicotine (F-/Nic−) or with 16 mg/mL of nicotine (F-/Nic+). All e-liquids were provided in 20 mL plastic bottles labeled as follows: propylene glycol <65%; glycerol <35%; food flavors; nicotine 0 or 16 mg/mL. The e-cigs used for vapor generation were the second generation “Lounge” model designed with 2.8 Ω coil and 3.6 V power supply. A new clearomiser was used for each experiment after having undergone 16 conditioning puffs and a USB power supply replaced the original battery to prevent age-related impairments. The heating was triggered by suction. Clearomisers were weighted before and after each experiment to measure the precise volume of e-liquid consumed.
Conventional cigarettes used for this study were 3R4F standard reference cigarettes (Lexington, Ky; https://ctrp.uky.edu/).
Vapor and smoke generation
E-vapors and tobacco smoke were generated with a Vitrocell® VC1 smoking machine (Vitrocell, Waldkirch, Germany). For e-cig experiments, the Cooperation Centre for Scientific Research Relative to Tobacco (CORESTA) approach was used, with 55 mL puff over 3 s, twice a minute (13). For 3R4F experiments, the VC1 parameters were those of the ISO 3308:2012 smoking regime (35 mL puff over 2 s, once a minute, 8 puffs per cigarette).
Vapor and smoke collection
E-vapors and tobacco smoke were collected from 96 puffs of e-cigs and 48 puffs of 3R4F cigarettes, respectively. Before each experiment, a thorough cleaning of the Vitrocell® pumping system was performed with ethanol and specific tubing restricted to each e-liquid was installed.
Nicotine, cotinine, propylene glycol, ethylene glycol, glycerol, polycyclic aromatic hydrocarbons (PAHs) and pesticides were collected by two impingers placed in series, containing 50 mL (first) and 25 mL (second) of methanol, and refrigerated at −40°C in an ethanol-dry ice bath.
Carbonyl compounds were trapped using specific silica cartridges coated with 2,4 dinitrophenylhydrazine (DNPH) (LpDNPH S10 Cartridge, Sigma-Aldrich, Saint-Quentin Fallavier, France).
Trace elements were collected using two midget impingers disposed in series, each containing 20 mL of 5% (v/v) nitric acid solution in Milli-Q® water (Merck Millipore, Billerica, MA).
Collecting materials were positioned at the smoking machine outlet in order not to alter e-vapor and smoke generation.
Blank collections were performed for all experiments using the smoking machine working without e-cig or conventional cigarette connected to and were taken into account for data analysis (Supplementary Table S2).
Analytical methods
Propylene glycol and ethylene glycol
E-liquids: 300 μL of water-diluted e-liquid (1/1000) were derivatized with 300 μL of phenylboronic acid (0.05 mol/L in a 75/25 (v/v) mixture of dichloromethane/acetone) in the presence of 100 μL of 2,3-butanediol (1 g/L in dichloromethane) used as an internal standard. After centrifugation (5 min, 25,155 g), 1 μL of the organic phase was injected into a gas chromatograph 7890 A series (Agilent Technologies, Santa Clara, CA), coupled with a tandem mass spectrometer Quattro MicroTM GC MICROMASS® (Waters, Milford, MA). Samples were injected with 700 split ratio at 260°C. Chromatographic separation was performed on CP-Sil 8Cb-MS column (30 m × 0.25 mm × 0.25 μm—Agilent Technologies). Analysis was performed in selected ion recording mode.
Vapor extracts: Methanolic extracts were 10-fold water-diluted and then processed as e-liquids above.
Glycerol
E-liquids: 100 μL of water-diluted e-liquid (1/5,000) were mixed with 100 μL of D8-glycerol used as an internal standard (100 mg/L in methanol). The mixture was dried under a nitrogen-stream at 40°C and resuspended with 50 μL of BSTFA/TMCS (N,O-Bis(trimethylsilyl)trifluoroacetamide/trimethylchlorosilane 99/1). After a 20-min heating at 60°C, 1 μL was injected with 400 split ratio at 300°C into the same analyzer used to quantify glycols. Chromatographic separation was achieved with a Grace AT-5 ms Heliflex column (30 m × 0.25 mm × 025 μm—Grace Davidson Discovery Science, Columbia, MD). Analysis was performed in multiple reaction monitoring (MRM) mode (m/z 218.2 → 73.1 and 218.2 → 147.2 for glycerol; m/z 222.2 → 147.2 for D8-glycerol).
Vapor extracts: Methanolic extracts were 20-fold diluted with methanol and then processed as e-liquids above.
Nicotine
E-liquids: Samples were 4-fold diluted in an isopropanol solution containing N-ethylaniline used as internal standard. Analyses were carried out by gas-chromatography with flame ionization detection using a Clarus 580 instrument (Perkin Elmer, Waltham, MA) and operated with a 35 split ratio on a Supelco SLB 5 MS column (5% diphenyl, 95% methyl polysiloxane, 30 m × 0.250 mm × 0.5 μm—Sigma-Aldrich).
Vapor extracts: 100 μL of 40-fold water-diluted methanolic extract were supplemented with 50 μL of D4-nicotine (100 μg/L) as an internal standard. Ten microliter of this solution was injected in full-loop mode into an ultra-performance ACQUITY UPLC liquid chromatographic system coupled with a triple quadrupole ACQUITY TQD detector used with an ACQUITY UPLCr HSS T3 1.8 μm column (Waters) maintained at 50°C. The mobile phase was a gradient mixture of water and acetonitrile. Detection was performed with positive ion electrospray ionization in MRM mode (m/z 163.10 → 117.10 and 163.10 → 130.10 for nicotine; m/z 167.20 → 84.10 and 167.20 → 134.10 for D4-nicotine).
Trace elements
An exhaustive list of the trace elements analyzed is provided in Supplementary data, section 1.1.
E-liquids: Standards and samples were prepared and analyzed using an inductively coupled plasma—mass spectrometer VARIAN 820-MS with SPS3 autosampler (Bruker, Billerica, MA), as described previously (14).
Vapor extracts: Water extracts were analyzed using the same ICP-MS instrument. Standard solutions were prepared with 1.67% (v/v) nitric acid solution in Milli-Q® water containing 0.2% (v/v) butane-1-ol, 0.1 % (v/v) Triton® and 50 μg/L of gold. Collected solutions were 3-fold diluted with a specific diluent in order to obtain the same final concentrations of nitric acid, butane-1-ol, triton and gold as the standard solutions.
Pesticides
An exhaustive list of the pesticides analyzed is provided in Supplementary data, section 1.2.
The method used for the determination of 50 pesticides in e-liquids and methanolic extracts was adapted from a method previously published by Hardy et al. (15) for the analysis of urinary pesticides, using gas-chromatography coupled with tandem mass spectrometry (GC–MS-MS).
Briefly, an aliquot of 500 μL (e-liquid or methanolic extract) supplemented with 5 μL of an internal standard solution (stable isotope labeled analogs) and 7.6 mL of 1 M phosphate buffer at pH 7 were analyzed with direct immersion solid-phase microextraction (SPME) at 60°C for 80 min and desorption for 10 min in the gas chromatograph injector. The analysis was performed with a 7890 gas chromatograph equipped with a HP-5 MS capillary column (30 m × 0.25 mm × 0.25 μm) coupled to a 7000 A triple quadrupole mass spectrometer (Agilent Technologies) operating in negative chemical ionization mode. In the case of e-liquids, 240 μL of acetonitrile were added into the vial to increase the transfer of the chemicals onto the SPME fiber.
Polycyclic aromatic hydrocarbons
An exhaustive list of the PAHs analyzed is provided in supplementary data, section 1.3.
The determination of PAHs in e-liquids was based on the analytical method previously described by Grova et al. (16), using the aforementioned GC–MS-MS instrument operating in electron impact ionization mode.
E-liquids: E-liquid samples (500 μL) were firstly supplemented with 10 μL of a mixed solution of internal standards at 0.1 mg/mL of the 16 PAHs investigated (Dr Ehrenstorfer, LGC Standards, Molsheim, France). Then, 1.5 mL of water and 2 mL of cyclohexane (CH) were added. After agitation and centrifugation, the supernatant was collected and evaporated under nitrogen-stream at 37°C. The residue was dissolved in CH and applied to an Envi-Chrom P solid-phase extraction column (Sigma-Aldrich, Bornem, Belgium) previously conditioned with CH. The PAHs were then eluted with 2 mL of ethyl acetate-CH (50:50; v/v) and the extract evaporated under nitrogen flow. Finally, 1 μL of the supernatant was injected into the analyzer.
Methanolic extracts: Methanolic extracts (2 mL) supplemented with 10 μL of 0.1 mg/L PAH internal standard solution, 0.5 mL of water and 2.5 mL of CH were agitated for 20 min at room temperature, centrifuged for 5 min at 1,800 g, and the two layers were separated. The CH layer containing PAHs was evaporated to a volume of 50 μL and 1 μL was then injected into the analyzer.
Carbonyl compounds
Formaldehyde, acetaldehyde and acrolein were analyzed.
Vapor extracts: Exposed LpDNPH S10 cartridges (Supelco, Sigma-Aldrich) were desorbed with 5 mL of acetonitrile. Twenty-five microliter were injected into a high-performance liquid chromatographic system coupled with a diode array detector which consists in a liquid chromatograph (1260 Infinity) coupled with a DAD detector (G1315C) (Agilent Technologies). Chromatographic separation was achieved on a Symmetry C18, 250 mm × 4.6 mm × 5 μm column (Waters). The mobile phase was a gradient mixture of acetonitrile and water.
Analytical method validation
In spite of the lack of regularity guidelines, reference methods or published critical values for electronic cigarettes, the validation procedure of these methods complies with recommendations for the validation of new analytical methods (17).
Results
Tables I and II describe the chemical analysis of e-liquids and e-cig/3R4F emissions, respectively. For e-vapors and smoke, results were expressed in mass of analyte per milliliter of emission (/mL puff) in order to better evidence differences between smoke and e-vapor chemical compositions and to facilitate data comparison with the literature, considering the large variability of puff generation protocols currently used.
. | Compounds . | LOQ (sample) . | F-/Nic− . | F-/Nic+ . | BT/Nic− . | BT/Nic+ . | CM/Nic− . | CM/Nic+ . |
---|---|---|---|---|---|---|---|---|
Constituents (mg/mL) | Nicotine | 2 | ~ | 15.8 | ~ | 16.0 | ~ | 15.9 |
Propylene glycol | 31.25 | 613 | 634 | 588 | 639 | 598 | 541 | |
Glycerol | 12.5 | 307 | 341 | 307 | 306 | 314 | 302 | |
Ethylene glycol | 62.5 | ~ | ~ | ~ | ~ | ~ | ~ | |
Trace elements (ng/mL) | Aluminum (Al) | 4 | 12 | 15 | 11 | 11 | 10 | 11 |
Antimony (Sb) | 0.1 | 1.5 | 1.5 | 1.2 | 1.2 | 1.3 | 1.4 | |
Arsenic (As) | 1 | 1.5 | 1.5 | ~ | ~ | ~ | ~ | |
Beryllium (Be) | 0.1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Cadmium (Cd) | 0.4 | ~ | ~ | ~ | ~ | ~ | ~ | |
Chromium (Cr) | 3.7 | 4.1 | 4.6 | 4.6 | 4.7 | 7.7 | 5.5 | |
Cobalt (Co) | 0.1 | 0.27 | 0.22 | ~ | ~ | ~ | ~ | |
Copper (Cu) | 20 | ~ | ~ | ~ | ~ | 32 | 25 | |
Lead (Pb) | 1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Manganese (Mn) | 1.6 | 3.1 | 3.3 | ~ | ~ | ~ | ~ | |
Mercury (Hg) | 4 | ~ | ~ | ~ | ~ | ~ | ~ | |
Nickel (Ni) | 16 | ~ | ~ | ~ | ~ | ~ | ~ | |
Thallium (Tl) | 0.1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Vanadium (V) | 0.4 | 0.64 | 0.44 | ~ | ~ | ~ | ~ | |
Zinc (Zn) | 200 | ~ | ~ | ~ | ~ | ~ | ~ | |
Pesticidesa (pg/mL) | Chlorpyrifos ethyl | 20 | ~ | ~ | 32.1 | ~ | 66.3 | 46.0 |
Trifluralin | 20 | ~ | ~ | ~ | ~ | 25.3 | 24.7 | |
PAHsa (ng/mL) | Acenaphthene | 0.20 | ~ | ~ | 0.52 | 1.12 | ~ | 0.43 |
Acenaphthylene | 0.02 | ~ | 0.03 | 0.03 | 0.03 | ~ | 0.05 | |
Benzo[a]pyrene | 0.02 | ~ | ~ | ~ | ~ | ~ | 0.02 | |
Benzo[b]fluoranthene | 0.02 | ~ | 0.02 | ~ | ~ | ~ | 0.03 | |
Benzo[g,h,i]perylene | 0.05 | ~ | ~ | ~ | ~ | ~ | 0.07 | |
Chrysene | 0.02 | 0.02 | 0.02 | ~ | ~ | ~ | 0.03 | |
Fluoranthene | 0.05 | 0.05 | 0.06 | ~ | ~ | 0.09 | 0.08 | |
Fluorene | 0.20 | 0.29 | 0.43 | ~ | 0.47 | 0.21 | 0.57 | |
Naphthalene | 0.20 | ~ | 19.1 | 18.9 | 61.8 | 4.24 | 32.8 | |
Phenanthrene | 0.20 | 3.45 | 3.83 | 2.66 | 2.43 | 3.53 | 3.81 |
. | Compounds . | LOQ (sample) . | F-/Nic− . | F-/Nic+ . | BT/Nic− . | BT/Nic+ . | CM/Nic− . | CM/Nic+ . |
---|---|---|---|---|---|---|---|---|
Constituents (mg/mL) | Nicotine | 2 | ~ | 15.8 | ~ | 16.0 | ~ | 15.9 |
Propylene glycol | 31.25 | 613 | 634 | 588 | 639 | 598 | 541 | |
Glycerol | 12.5 | 307 | 341 | 307 | 306 | 314 | 302 | |
Ethylene glycol | 62.5 | ~ | ~ | ~ | ~ | ~ | ~ | |
Trace elements (ng/mL) | Aluminum (Al) | 4 | 12 | 15 | 11 | 11 | 10 | 11 |
Antimony (Sb) | 0.1 | 1.5 | 1.5 | 1.2 | 1.2 | 1.3 | 1.4 | |
Arsenic (As) | 1 | 1.5 | 1.5 | ~ | ~ | ~ | ~ | |
Beryllium (Be) | 0.1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Cadmium (Cd) | 0.4 | ~ | ~ | ~ | ~ | ~ | ~ | |
Chromium (Cr) | 3.7 | 4.1 | 4.6 | 4.6 | 4.7 | 7.7 | 5.5 | |
Cobalt (Co) | 0.1 | 0.27 | 0.22 | ~ | ~ | ~ | ~ | |
Copper (Cu) | 20 | ~ | ~ | ~ | ~ | 32 | 25 | |
Lead (Pb) | 1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Manganese (Mn) | 1.6 | 3.1 | 3.3 | ~ | ~ | ~ | ~ | |
Mercury (Hg) | 4 | ~ | ~ | ~ | ~ | ~ | ~ | |
Nickel (Ni) | 16 | ~ | ~ | ~ | ~ | ~ | ~ | |
Thallium (Tl) | 0.1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Vanadium (V) | 0.4 | 0.64 | 0.44 | ~ | ~ | ~ | ~ | |
Zinc (Zn) | 200 | ~ | ~ | ~ | ~ | ~ | ~ | |
Pesticidesa (pg/mL) | Chlorpyrifos ethyl | 20 | ~ | ~ | 32.1 | ~ | 66.3 | 46.0 |
Trifluralin | 20 | ~ | ~ | ~ | ~ | 25.3 | 24.7 | |
PAHsa (ng/mL) | Acenaphthene | 0.20 | ~ | ~ | 0.52 | 1.12 | ~ | 0.43 |
Acenaphthylene | 0.02 | ~ | 0.03 | 0.03 | 0.03 | ~ | 0.05 | |
Benzo[a]pyrene | 0.02 | ~ | ~ | ~ | ~ | ~ | 0.02 | |
Benzo[b]fluoranthene | 0.02 | ~ | 0.02 | ~ | ~ | ~ | 0.03 | |
Benzo[g,h,i]perylene | 0.05 | ~ | ~ | ~ | ~ | ~ | 0.07 | |
Chrysene | 0.02 | 0.02 | 0.02 | ~ | ~ | ~ | 0.03 | |
Fluoranthene | 0.05 | 0.05 | 0.06 | ~ | ~ | 0.09 | 0.08 | |
Fluorene | 0.20 | 0.29 | 0.43 | ~ | 0.47 | 0.21 | 0.57 | |
Naphthalene | 0.20 | ~ | 19.1 | 18.9 | 61.8 | 4.24 | 32.8 | |
Phenanthrene | 0.20 | 3.45 | 3.83 | 2.66 | 2.43 | 3.53 | 3.81 |
“~”, <Limit of quantification; “F-”, unflavored e-liquid; “BT”, blond tobacco flavored e-liquid' “CM”, chlorophyll mint flavored e-liquid; “Nic−”, e-liquid without nicotine' “Nic+”, e-liquid with 16 mg/mL nicotine. aAre only presented compounds which were quantified in at least one sample.
. | Compounds . | LOQ (sample) . | F-/Nic− . | F-/Nic+ . | BT/Nic− . | BT/Nic+ . | CM/Nic− . | CM/Nic+ . |
---|---|---|---|---|---|---|---|---|
Constituents (mg/mL) | Nicotine | 2 | ~ | 15.8 | ~ | 16.0 | ~ | 15.9 |
Propylene glycol | 31.25 | 613 | 634 | 588 | 639 | 598 | 541 | |
Glycerol | 12.5 | 307 | 341 | 307 | 306 | 314 | 302 | |
Ethylene glycol | 62.5 | ~ | ~ | ~ | ~ | ~ | ~ | |
Trace elements (ng/mL) | Aluminum (Al) | 4 | 12 | 15 | 11 | 11 | 10 | 11 |
Antimony (Sb) | 0.1 | 1.5 | 1.5 | 1.2 | 1.2 | 1.3 | 1.4 | |
Arsenic (As) | 1 | 1.5 | 1.5 | ~ | ~ | ~ | ~ | |
Beryllium (Be) | 0.1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Cadmium (Cd) | 0.4 | ~ | ~ | ~ | ~ | ~ | ~ | |
Chromium (Cr) | 3.7 | 4.1 | 4.6 | 4.6 | 4.7 | 7.7 | 5.5 | |
Cobalt (Co) | 0.1 | 0.27 | 0.22 | ~ | ~ | ~ | ~ | |
Copper (Cu) | 20 | ~ | ~ | ~ | ~ | 32 | 25 | |
Lead (Pb) | 1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Manganese (Mn) | 1.6 | 3.1 | 3.3 | ~ | ~ | ~ | ~ | |
Mercury (Hg) | 4 | ~ | ~ | ~ | ~ | ~ | ~ | |
Nickel (Ni) | 16 | ~ | ~ | ~ | ~ | ~ | ~ | |
Thallium (Tl) | 0.1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Vanadium (V) | 0.4 | 0.64 | 0.44 | ~ | ~ | ~ | ~ | |
Zinc (Zn) | 200 | ~ | ~ | ~ | ~ | ~ | ~ | |
Pesticidesa (pg/mL) | Chlorpyrifos ethyl | 20 | ~ | ~ | 32.1 | ~ | 66.3 | 46.0 |
Trifluralin | 20 | ~ | ~ | ~ | ~ | 25.3 | 24.7 | |
PAHsa (ng/mL) | Acenaphthene | 0.20 | ~ | ~ | 0.52 | 1.12 | ~ | 0.43 |
Acenaphthylene | 0.02 | ~ | 0.03 | 0.03 | 0.03 | ~ | 0.05 | |
Benzo[a]pyrene | 0.02 | ~ | ~ | ~ | ~ | ~ | 0.02 | |
Benzo[b]fluoranthene | 0.02 | ~ | 0.02 | ~ | ~ | ~ | 0.03 | |
Benzo[g,h,i]perylene | 0.05 | ~ | ~ | ~ | ~ | ~ | 0.07 | |
Chrysene | 0.02 | 0.02 | 0.02 | ~ | ~ | ~ | 0.03 | |
Fluoranthene | 0.05 | 0.05 | 0.06 | ~ | ~ | 0.09 | 0.08 | |
Fluorene | 0.20 | 0.29 | 0.43 | ~ | 0.47 | 0.21 | 0.57 | |
Naphthalene | 0.20 | ~ | 19.1 | 18.9 | 61.8 | 4.24 | 32.8 | |
Phenanthrene | 0.20 | 3.45 | 3.83 | 2.66 | 2.43 | 3.53 | 3.81 |
. | Compounds . | LOQ (sample) . | F-/Nic− . | F-/Nic+ . | BT/Nic− . | BT/Nic+ . | CM/Nic− . | CM/Nic+ . |
---|---|---|---|---|---|---|---|---|
Constituents (mg/mL) | Nicotine | 2 | ~ | 15.8 | ~ | 16.0 | ~ | 15.9 |
Propylene glycol | 31.25 | 613 | 634 | 588 | 639 | 598 | 541 | |
Glycerol | 12.5 | 307 | 341 | 307 | 306 | 314 | 302 | |
Ethylene glycol | 62.5 | ~ | ~ | ~ | ~ | ~ | ~ | |
Trace elements (ng/mL) | Aluminum (Al) | 4 | 12 | 15 | 11 | 11 | 10 | 11 |
Antimony (Sb) | 0.1 | 1.5 | 1.5 | 1.2 | 1.2 | 1.3 | 1.4 | |
Arsenic (As) | 1 | 1.5 | 1.5 | ~ | ~ | ~ | ~ | |
Beryllium (Be) | 0.1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Cadmium (Cd) | 0.4 | ~ | ~ | ~ | ~ | ~ | ~ | |
Chromium (Cr) | 3.7 | 4.1 | 4.6 | 4.6 | 4.7 | 7.7 | 5.5 | |
Cobalt (Co) | 0.1 | 0.27 | 0.22 | ~ | ~ | ~ | ~ | |
Copper (Cu) | 20 | ~ | ~ | ~ | ~ | 32 | 25 | |
Lead (Pb) | 1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Manganese (Mn) | 1.6 | 3.1 | 3.3 | ~ | ~ | ~ | ~ | |
Mercury (Hg) | 4 | ~ | ~ | ~ | ~ | ~ | ~ | |
Nickel (Ni) | 16 | ~ | ~ | ~ | ~ | ~ | ~ | |
Thallium (Tl) | 0.1 | ~ | ~ | ~ | ~ | ~ | ~ | |
Vanadium (V) | 0.4 | 0.64 | 0.44 | ~ | ~ | ~ | ~ | |
Zinc (Zn) | 200 | ~ | ~ | ~ | ~ | ~ | ~ | |
Pesticidesa (pg/mL) | Chlorpyrifos ethyl | 20 | ~ | ~ | 32.1 | ~ | 66.3 | 46.0 |
Trifluralin | 20 | ~ | ~ | ~ | ~ | 25.3 | 24.7 | |
PAHsa (ng/mL) | Acenaphthene | 0.20 | ~ | ~ | 0.52 | 1.12 | ~ | 0.43 |
Acenaphthylene | 0.02 | ~ | 0.03 | 0.03 | 0.03 | ~ | 0.05 | |
Benzo[a]pyrene | 0.02 | ~ | ~ | ~ | ~ | ~ | 0.02 | |
Benzo[b]fluoranthene | 0.02 | ~ | 0.02 | ~ | ~ | ~ | 0.03 | |
Benzo[g,h,i]perylene | 0.05 | ~ | ~ | ~ | ~ | ~ | 0.07 | |
Chrysene | 0.02 | 0.02 | 0.02 | ~ | ~ | ~ | 0.03 | |
Fluoranthene | 0.05 | 0.05 | 0.06 | ~ | ~ | 0.09 | 0.08 | |
Fluorene | 0.20 | 0.29 | 0.43 | ~ | 0.47 | 0.21 | 0.57 | |
Naphthalene | 0.20 | ~ | 19.1 | 18.9 | 61.8 | 4.24 | 32.8 | |
Phenanthrene | 0.20 | 3.45 | 3.83 | 2.66 | 2.43 | 3.53 | 3.81 |
“~”, <Limit of quantification; “F-”, unflavored e-liquid; “BT”, blond tobacco flavored e-liquid' “CM”, chlorophyll mint flavored e-liquid; “Nic−”, e-liquid without nicotine' “Nic+”, e-liquid with 16 mg/mL nicotine. aAre only presented compounds which were quantified in at least one sample.
. | Compounds . | LOQ (e-vapors) . | F-/Nic− . | F-/Nic+ . | BT/Nic− . | BT/Nic+ . | CM/Nic− . | CM/Nic+ . | LOQ (smoke) . | 3R4F . |
---|---|---|---|---|---|---|---|---|---|---|
Constituents (μg/mL puffa) | Nicotine | 0.0038 | ~ | 0.55 ± 0.033 | ~ | 0.56 ± 0.012 | ~ | 0.30 ± 0.036 | 0.0119 | 1.03 ± 0.024 |
Propylene glycol | 3.0 | 17.2 ± 2.7b | 22.8 ± 0.4 | 21.0 ± 0.4 | 22.0 ± 0.3 | 16.1 ± 1.5 | 10.0 ± 1.4 | 9.3 | ~ | |
Glycerol | 2.4 | 12.3 ± 2.2 | 15.7 ± 0.5 | 12.2 ± 0.6 | 12.9 ± 0.3 | 10.3 ± 0.8 | 7.1 ± 1.1 | 7.4 | ~ | |
Ethylene glycol | 5.9 | ~ | ~ | ~ | ~ | ~ | ~ | 18.6 | ~ | |
Trace elementsc (pg/mL puff) | Antimony (Sb) | 0.11 | 0.47 ± 0.30 | 0.28 ± 0.17 | 0.19 ± 0.09 | ~ | 0.14 ± 0.01 | ~ | 0.36 | ~ |
Arsenic (As) | 0.23 | ~ | ~ | ~ | ~ | ~ | ~ | 0.71 | 4.21 ± 0.18 | |
Cadmium (Cd) | 0.02 | ~ | 0.04 ± 0.01 | ~ | ~ | 0.14 ± 0.05 | ~ | 0.07 | 44.98 ± 1.90 | |
Chromium (Cr) | 2.1 | ~ | 3.4 ± 0.6 | 3.3 ± 0.5 | 2.9 ± 0.7 | ~ | ~ | 6.6 | ~ | |
Lead (Pb) | 0.23 | 1.2 ± 0.4 | 1.3 ± 0.8 | 1.6 ± 0.5 | 1.0 ± 0.2 | ~ | ~ | 0.71 | 16.5 ± 0.6 | |
Thallium (Tl) | 0.06 | ~ | ~ | ~ | ~ | ~ | ~ | 0.18 | 1.02 ± 0.14 | |
Pesticidesc (pg/mL puff) | Chlorpyrifos ethyl | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 0.9 |
Cyhalothrin | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 0.9 | |
Cypermethrin | 0.38 | ~ | ~ | ~ | ~ | ~ | ~ | 1.19 | 2.4 | |
Trifluralin | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 7.3 | |
α-endosulfan | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 2.2 | |
β-endosulfan | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 8.7 | |
PAHs (pg/mL puff) | Acenaphthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 92.2 |
Acenaphthylene | 0.09 | 0.21 | 0.30 | 0.19 | 0.37 | 0.24 | 0.25 | 0.30 | 384 | |
Anthracene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 106 | |
Benz[a]anthracene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 22.6 | |
Benzo[a]pyrene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 17.4 | |
Benzo[b]fluoranthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 9.90 | |
Benzo[g,h,i]perylene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 14.5 | |
Benzo[k]fluoranthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 11.1 | |
Chrysene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 33.0 | |
Dibenz[a,h]anthracene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 1.15 | |
Fluoranthene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 117 | |
Fluorene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 367 | |
Indeno[1,2,3-cd]pyrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 7.51 | |
Naphthalene | 0.47 | 2.39 | 4.01 | 1.79 | 4.10 | 1.84 | 3.07 | 1.49 | 561 | |
Phenanthrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 289 | |
Pyrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 115 | |
Carbonyls (ng/mL puff) | Acetaldehyde | 0.05 | 0.44 ± 0.12 | 0.40 ± 0.10 | 0.16 ± 0.08 | 0.48 ± 0.33 | 0.76 ± 0.59 | 0.96 ± 0.25 | 1540d | |
Acrolein | 0.05 | ~ | 0.17 ± 0.05 | 2.11 ± 0.39 | 1.72 ± 0.41 | 0.11 ± 0.15 | ~ | 171d | ||
Formaldehyde | 0.05 | 0.95 ± 0.40 | 0.70 ± 0.09 | 0.37 ± 0.02 | 0.78 ± 0.48 | 1.29 ± 0.88 | 1.48 ± 0.32 | 82d |
. | Compounds . | LOQ (e-vapors) . | F-/Nic− . | F-/Nic+ . | BT/Nic− . | BT/Nic+ . | CM/Nic− . | CM/Nic+ . | LOQ (smoke) . | 3R4F . |
---|---|---|---|---|---|---|---|---|---|---|
Constituents (μg/mL puffa) | Nicotine | 0.0038 | ~ | 0.55 ± 0.033 | ~ | 0.56 ± 0.012 | ~ | 0.30 ± 0.036 | 0.0119 | 1.03 ± 0.024 |
Propylene glycol | 3.0 | 17.2 ± 2.7b | 22.8 ± 0.4 | 21.0 ± 0.4 | 22.0 ± 0.3 | 16.1 ± 1.5 | 10.0 ± 1.4 | 9.3 | ~ | |
Glycerol | 2.4 | 12.3 ± 2.2 | 15.7 ± 0.5 | 12.2 ± 0.6 | 12.9 ± 0.3 | 10.3 ± 0.8 | 7.1 ± 1.1 | 7.4 | ~ | |
Ethylene glycol | 5.9 | ~ | ~ | ~ | ~ | ~ | ~ | 18.6 | ~ | |
Trace elementsc (pg/mL puff) | Antimony (Sb) | 0.11 | 0.47 ± 0.30 | 0.28 ± 0.17 | 0.19 ± 0.09 | ~ | 0.14 ± 0.01 | ~ | 0.36 | ~ |
Arsenic (As) | 0.23 | ~ | ~ | ~ | ~ | ~ | ~ | 0.71 | 4.21 ± 0.18 | |
Cadmium (Cd) | 0.02 | ~ | 0.04 ± 0.01 | ~ | ~ | 0.14 ± 0.05 | ~ | 0.07 | 44.98 ± 1.90 | |
Chromium (Cr) | 2.1 | ~ | 3.4 ± 0.6 | 3.3 ± 0.5 | 2.9 ± 0.7 | ~ | ~ | 6.6 | ~ | |
Lead (Pb) | 0.23 | 1.2 ± 0.4 | 1.3 ± 0.8 | 1.6 ± 0.5 | 1.0 ± 0.2 | ~ | ~ | 0.71 | 16.5 ± 0.6 | |
Thallium (Tl) | 0.06 | ~ | ~ | ~ | ~ | ~ | ~ | 0.18 | 1.02 ± 0.14 | |
Pesticidesc (pg/mL puff) | Chlorpyrifos ethyl | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 0.9 |
Cyhalothrin | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 0.9 | |
Cypermethrin | 0.38 | ~ | ~ | ~ | ~ | ~ | ~ | 1.19 | 2.4 | |
Trifluralin | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 7.3 | |
α-endosulfan | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 2.2 | |
β-endosulfan | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 8.7 | |
PAHs (pg/mL puff) | Acenaphthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 92.2 |
Acenaphthylene | 0.09 | 0.21 | 0.30 | 0.19 | 0.37 | 0.24 | 0.25 | 0.30 | 384 | |
Anthracene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 106 | |
Benz[a]anthracene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 22.6 | |
Benzo[a]pyrene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 17.4 | |
Benzo[b]fluoranthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 9.90 | |
Benzo[g,h,i]perylene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 14.5 | |
Benzo[k]fluoranthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 11.1 | |
Chrysene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 33.0 | |
Dibenz[a,h]anthracene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 1.15 | |
Fluoranthene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 117 | |
Fluorene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 367 | |
Indeno[1,2,3-cd]pyrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 7.51 | |
Naphthalene | 0.47 | 2.39 | 4.01 | 1.79 | 4.10 | 1.84 | 3.07 | 1.49 | 561 | |
Phenanthrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 289 | |
Pyrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 115 | |
Carbonyls (ng/mL puff) | Acetaldehyde | 0.05 | 0.44 ± 0.12 | 0.40 ± 0.10 | 0.16 ± 0.08 | 0.48 ± 0.33 | 0.76 ± 0.59 | 0.96 ± 0.25 | 1540d | |
Acrolein | 0.05 | ~ | 0.17 ± 0.05 | 2.11 ± 0.39 | 1.72 ± 0.41 | 0.11 ± 0.15 | ~ | 171d | ||
Formaldehyde | 0.05 | 0.95 ± 0.40 | 0.70 ± 0.09 | 0.37 ± 0.02 | 0.78 ± 0.48 | 1.29 ± 0.88 | 1.48 ± 0.32 | 82d |
“~”, <Limit of quantification; “F-”, unflavored e-liquid; “BT”, blond tobacco flavored e-liquid' “CM”, chlorophyll mint flavored e-liquid; “Nic−”, e-liquid without nicotine; “Nic+”, e-liquid with 16 mg/mL nicotine; “3R4F”, Kentucky 3R4F standard reference cigarettes (Lexington, Ky; https://ctrp.uky.edu/).
aResults are presented in mass of analyte per milliliter of emission (/mL puff).
bResults are presented as mean +/− standard deviation of triplicate experiments, when available.
cAre only presented compounds which were quantified or higher than experimental blanks in at least one sample.
dExtrapolated values from Eldridge et al. (18).
. | Compounds . | LOQ (e-vapors) . | F-/Nic− . | F-/Nic+ . | BT/Nic− . | BT/Nic+ . | CM/Nic− . | CM/Nic+ . | LOQ (smoke) . | 3R4F . |
---|---|---|---|---|---|---|---|---|---|---|
Constituents (μg/mL puffa) | Nicotine | 0.0038 | ~ | 0.55 ± 0.033 | ~ | 0.56 ± 0.012 | ~ | 0.30 ± 0.036 | 0.0119 | 1.03 ± 0.024 |
Propylene glycol | 3.0 | 17.2 ± 2.7b | 22.8 ± 0.4 | 21.0 ± 0.4 | 22.0 ± 0.3 | 16.1 ± 1.5 | 10.0 ± 1.4 | 9.3 | ~ | |
Glycerol | 2.4 | 12.3 ± 2.2 | 15.7 ± 0.5 | 12.2 ± 0.6 | 12.9 ± 0.3 | 10.3 ± 0.8 | 7.1 ± 1.1 | 7.4 | ~ | |
Ethylene glycol | 5.9 | ~ | ~ | ~ | ~ | ~ | ~ | 18.6 | ~ | |
Trace elementsc (pg/mL puff) | Antimony (Sb) | 0.11 | 0.47 ± 0.30 | 0.28 ± 0.17 | 0.19 ± 0.09 | ~ | 0.14 ± 0.01 | ~ | 0.36 | ~ |
Arsenic (As) | 0.23 | ~ | ~ | ~ | ~ | ~ | ~ | 0.71 | 4.21 ± 0.18 | |
Cadmium (Cd) | 0.02 | ~ | 0.04 ± 0.01 | ~ | ~ | 0.14 ± 0.05 | ~ | 0.07 | 44.98 ± 1.90 | |
Chromium (Cr) | 2.1 | ~ | 3.4 ± 0.6 | 3.3 ± 0.5 | 2.9 ± 0.7 | ~ | ~ | 6.6 | ~ | |
Lead (Pb) | 0.23 | 1.2 ± 0.4 | 1.3 ± 0.8 | 1.6 ± 0.5 | 1.0 ± 0.2 | ~ | ~ | 0.71 | 16.5 ± 0.6 | |
Thallium (Tl) | 0.06 | ~ | ~ | ~ | ~ | ~ | ~ | 0.18 | 1.02 ± 0.14 | |
Pesticidesc (pg/mL puff) | Chlorpyrifos ethyl | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 0.9 |
Cyhalothrin | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 0.9 | |
Cypermethrin | 0.38 | ~ | ~ | ~ | ~ | ~ | ~ | 1.19 | 2.4 | |
Trifluralin | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 7.3 | |
α-endosulfan | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 2.2 | |
β-endosulfan | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 8.7 | |
PAHs (pg/mL puff) | Acenaphthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 92.2 |
Acenaphthylene | 0.09 | 0.21 | 0.30 | 0.19 | 0.37 | 0.24 | 0.25 | 0.30 | 384 | |
Anthracene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 106 | |
Benz[a]anthracene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 22.6 | |
Benzo[a]pyrene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 17.4 | |
Benzo[b]fluoranthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 9.90 | |
Benzo[g,h,i]perylene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 14.5 | |
Benzo[k]fluoranthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 11.1 | |
Chrysene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 33.0 | |
Dibenz[a,h]anthracene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 1.15 | |
Fluoranthene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 117 | |
Fluorene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 367 | |
Indeno[1,2,3-cd]pyrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 7.51 | |
Naphthalene | 0.47 | 2.39 | 4.01 | 1.79 | 4.10 | 1.84 | 3.07 | 1.49 | 561 | |
Phenanthrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 289 | |
Pyrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 115 | |
Carbonyls (ng/mL puff) | Acetaldehyde | 0.05 | 0.44 ± 0.12 | 0.40 ± 0.10 | 0.16 ± 0.08 | 0.48 ± 0.33 | 0.76 ± 0.59 | 0.96 ± 0.25 | 1540d | |
Acrolein | 0.05 | ~ | 0.17 ± 0.05 | 2.11 ± 0.39 | 1.72 ± 0.41 | 0.11 ± 0.15 | ~ | 171d | ||
Formaldehyde | 0.05 | 0.95 ± 0.40 | 0.70 ± 0.09 | 0.37 ± 0.02 | 0.78 ± 0.48 | 1.29 ± 0.88 | 1.48 ± 0.32 | 82d |
. | Compounds . | LOQ (e-vapors) . | F-/Nic− . | F-/Nic+ . | BT/Nic− . | BT/Nic+ . | CM/Nic− . | CM/Nic+ . | LOQ (smoke) . | 3R4F . |
---|---|---|---|---|---|---|---|---|---|---|
Constituents (μg/mL puffa) | Nicotine | 0.0038 | ~ | 0.55 ± 0.033 | ~ | 0.56 ± 0.012 | ~ | 0.30 ± 0.036 | 0.0119 | 1.03 ± 0.024 |
Propylene glycol | 3.0 | 17.2 ± 2.7b | 22.8 ± 0.4 | 21.0 ± 0.4 | 22.0 ± 0.3 | 16.1 ± 1.5 | 10.0 ± 1.4 | 9.3 | ~ | |
Glycerol | 2.4 | 12.3 ± 2.2 | 15.7 ± 0.5 | 12.2 ± 0.6 | 12.9 ± 0.3 | 10.3 ± 0.8 | 7.1 ± 1.1 | 7.4 | ~ | |
Ethylene glycol | 5.9 | ~ | ~ | ~ | ~ | ~ | ~ | 18.6 | ~ | |
Trace elementsc (pg/mL puff) | Antimony (Sb) | 0.11 | 0.47 ± 0.30 | 0.28 ± 0.17 | 0.19 ± 0.09 | ~ | 0.14 ± 0.01 | ~ | 0.36 | ~ |
Arsenic (As) | 0.23 | ~ | ~ | ~ | ~ | ~ | ~ | 0.71 | 4.21 ± 0.18 | |
Cadmium (Cd) | 0.02 | ~ | 0.04 ± 0.01 | ~ | ~ | 0.14 ± 0.05 | ~ | 0.07 | 44.98 ± 1.90 | |
Chromium (Cr) | 2.1 | ~ | 3.4 ± 0.6 | 3.3 ± 0.5 | 2.9 ± 0.7 | ~ | ~ | 6.6 | ~ | |
Lead (Pb) | 0.23 | 1.2 ± 0.4 | 1.3 ± 0.8 | 1.6 ± 0.5 | 1.0 ± 0.2 | ~ | ~ | 0.71 | 16.5 ± 0.6 | |
Thallium (Tl) | 0.06 | ~ | ~ | ~ | ~ | ~ | ~ | 0.18 | 1.02 ± 0.14 | |
Pesticidesc (pg/mL puff) | Chlorpyrifos ethyl | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 0.9 |
Cyhalothrin | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 0.9 | |
Cypermethrin | 0.38 | ~ | ~ | ~ | ~ | ~ | ~ | 1.19 | 2.4 | |
Trifluralin | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 7.3 | |
α-endosulfan | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 2.2 | |
β-endosulfan | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 8.7 | |
PAHs (pg/mL puff) | Acenaphthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 92.2 |
Acenaphthylene | 0.09 | 0.21 | 0.30 | 0.19 | 0.37 | 0.24 | 0.25 | 0.30 | 384 | |
Anthracene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 106 | |
Benz[a]anthracene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 22.6 | |
Benzo[a]pyrene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 17.4 | |
Benzo[b]fluoranthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 9.90 | |
Benzo[g,h,i]perylene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 14.5 | |
Benzo[k]fluoranthene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 11.1 | |
Chrysene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 33.0 | |
Dibenz[a,h]anthracene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 1.15 | |
Fluoranthene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 117 | |
Fluorene | 0.09 | ~ | ~ | ~ | ~ | ~ | ~ | 0.30 | 367 | |
Indeno[1,2,3-cd]pyrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 7.51 | |
Naphthalene | 0.47 | 2.39 | 4.01 | 1.79 | 4.10 | 1.84 | 3.07 | 1.49 | 561 | |
Phenanthrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 289 | |
Pyrene | 0.19 | ~ | ~ | ~ | ~ | ~ | ~ | 0.60 | 115 | |
Carbonyls (ng/mL puff) | Acetaldehyde | 0.05 | 0.44 ± 0.12 | 0.40 ± 0.10 | 0.16 ± 0.08 | 0.48 ± 0.33 | 0.76 ± 0.59 | 0.96 ± 0.25 | 1540d | |
Acrolein | 0.05 | ~ | 0.17 ± 0.05 | 2.11 ± 0.39 | 1.72 ± 0.41 | 0.11 ± 0.15 | ~ | 171d | ||
Formaldehyde | 0.05 | 0.95 ± 0.40 | 0.70 ± 0.09 | 0.37 ± 0.02 | 0.78 ± 0.48 | 1.29 ± 0.88 | 1.48 ± 0.32 | 82d |
“~”, <Limit of quantification; “F-”, unflavored e-liquid; “BT”, blond tobacco flavored e-liquid' “CM”, chlorophyll mint flavored e-liquid; “Nic−”, e-liquid without nicotine; “Nic+”, e-liquid with 16 mg/mL nicotine; “3R4F”, Kentucky 3R4F standard reference cigarettes (Lexington, Ky; https://ctrp.uky.edu/).
aResults are presented in mass of analyte per milliliter of emission (/mL puff).
bResults are presented as mean +/− standard deviation of triplicate experiments, when available.
cAre only presented compounds which were quantified or higher than experimental blanks in at least one sample.
dExtrapolated values from Eldridge et al. (18).
Additional analytical method parameters are available in Supplementary data (Table S3).
Propylene glycol, ethylene glycol, glycerol and nicotine
Nicotine was not detected in e-liquids labeled as nicotine-free nor in their respective vapors. In nicotine-labeled e-liquids, sold as containing 16 mg/mL nicotine, concentrations were 16.0 and 15.9 mg/mL for BT/N+ and CM/N+ samples, respectively (Table I). Propylene glycol and glycerol concentrations in e-liquids ranged from 541 to 639 and 302 to 341 g/L, respectively, corresponding to the advertised propylene glycol/glycerol ratios of 65/35. In e-vapors, concentrations of propylene glycol, glycerol and nicotine varied within e-liquids (Table II) but remained consistent when related to the mass of e-liquid consumed during experiments (data not shown). Their relative proportions, previously measured in e-liquids, were also maintained. Ethylene glycol was not detected in any e-liquid or e-vapor sample.
Trace elements
Low concentrations of all analyzed trace elements were found in e-liquids, mostly below or close to their LOQs (<32 μg/L for all elements, except for Zn <200 μg/L, Table I). In e-vapors, Be, Cu, Hg, V and Zn were not quantified in any collected solution. Quantifiable concentrations of Al, Co, Mn, Ni and Pb were found in several e-vapor samples, but were comparable with experimental blank concentrations. The presence of the aforementioned elements in the blank samples likely results from contamination by the smoking machine or by the collecting or storage systems. Only Cd, Cr and Sb were present in some e-vapors, up to 0.14, 3.4 and 0.47 pg/mL puff, respectively. As, Cd, Pb and Tl were quantified in 3R4F smoke, from 1.02 pg/mL puff for Tl to 44.98 pg/mL puff for Cd (Table II).
Pesticides
Most of the 50 pesticides tested were not detected in any e-liquid or e-vapor samples. Only chlorpyrifos ethyl and trifluralin were found in some e-liquid samples, with concentrations slightly higher than the LOQ (up to 66.3 pg/mL for CM/Nic-, Table I). While no pesticide was quantified in e-vapors, chlorpyrifos ethyl, cyhalothrin, cypermethrin, trifluralin, α-endosulfan and β-endosulfan were present in cigarette smoke (from 0.9 to 8.7 pg/mL puff, Table II).
Polycyclic aromatic hydrocarbons
Ten of the 16 targeted PAHs were detected in e-liquids with concentrations close to or below LOQs. Naphthalene and phenanthrene were the major PAHs present in e-liquids, from <0.2 to 61.8 ng/mL and from 2.43 to 3.83 ng/mL, respectively (Table I). Only naphthalene and acenaphthylene were systematically detected in e-vapors at very low concentrations (up to 4.10 and 0.37 pg/mL puff, respectively), compared to 3R4F cigarette smoke (561 and 384 pg/mL puff, respectively). Trace concentrations of naphthalene, phenanthrene and fluorene were quantified in experimental blanks, though negligible compared to their concentrations in smoke (Supplementary Table S2). All PAHs were quantified in conventional cigarette smoke, at variable concentrations, from 1.15 pg/mL puff for dibenz[a,h]anthracene to 367 pg/mL puff for fluorene (Table II).
Carbonyl compounds
Carbonyls were only studied in e-vapors (Table II). Relatively similar concentrations of formaldehyde and acetaldehyde, ranging from 0.16 to 1.48 ng/mL puff, were quantified in all e-vapors, whereas acrolein was almost only found in BT-related vapors without and with nicotine (2.11 and 1.72 ng/mL puff, respectively). As 3R4F smoke collections demonstrated signs of saturation, by consumption of all available DNPH, even after processing only 8 puffs (data not shown), carbonyl compounds were not considered in 3R4F smoke.
Discussion
Propylene glycol and glycerol are the base components of e-liquid mixtures and their proportions vary depending on the brand and the manufacturer. Labeled as containing <65% propylene glycol and <35% glycerol, the e-liquids tested here complied with the advertised composition, were relatively similar to each other, and did not contain ethylene glycol (Table I).
Even if a portion of e-vapors was stagnating above the impinger's liquid level, indicating that collection efficiency in our experiment was not optimal, particular attention was given to the harmonization of all protocols with a view to perform comparative studies. When related to a single puff, Pellegrino et al. (19) reported similar concentrations as those measured here, whereas higher concentrations were reported by Breiev et al. (20) using an online proton transfer reaction mass spectrometry approach.
Nicotine is one of the major concerns for “vapers”, especially in a smoking cessation attempt, and is likely the major toxic compound contained in e-liquids and vapors. In contrast with results described in recent studies (21, 22), we demonstrated that the e-liquids tested here contained nicotine concentrations corresponding closely to their labeling.
In e-vapors, nicotine yields should be controlled to ensure consumer safety and effective nicotine withdrawal, if desired. While expected nicotine mass emitted from 3R4F smoke was not reached during experiments, suggesting some underestimation, results obtained for e-vapors (Table II) were comparable to those from previous studies (23–25). When related to 1 mL of vapor, extracted nicotine concentrations were ~2- to 3-fold higher in 3R4F smoke than in vapors from e-liquids initially containing 16 mg/mL nicotine. Related to one puff, nicotine delivery by vaping was almost similar to that by smoking (Supplementary data Table S4), except for CM/Nic+ because of a lower e-liquid consumption during the experiments (data not shown). However, El-Hellani et al. (26) suggested that the bioavailable nicotine could be lower than expected because of the pH of e-liquids, so further investigations are needed to associate nicotine theoretical intake with vapers’ real exposure, regarding conditions of use.
The quantification by ICP-MS of 15 trace elements in a mixture of propylene glycol and glycerol was subject to strong matrix effects and presented a real analytical challenge. Low concentrations of the selected trace elements were found in the e-liquids tested (Table I), similar to those reported by Saffari et al. (27). In a previous study, we analyzed 48 other e-liquids from the NHOSS® brand (14). Neither higher concentrations nor noticeable variability within e-liquids was observed for any sample and elements studied here, except for cherry flavored e-liquids without and with nicotine, which contained higher amounts of Sb (214 and 99.3 μg/L, respectively) and Zn (325 and 510 μg/L, respectively). Thus, further studies should evaluate how flavor components may be considered as a potential source of direct e-liquid contamination by trace elements.
In 3R4F smoke, only As, Cd, Pb and Tl were quantified (Table II). Comparable, although slightly higher, concentrations of As (2.5 vs. 1.18 ng/cigarette in our study), Cd (34.5 vs. 12.6 ng/cigarette) and Pb (9.1 vs. 4.3 ng/cigarette) were reported by Pappas et al. (28), suggesting our smoke collection method is suitable in this context. Compared to 3R4F smoke, e-vapors emitted barely detectable levels of As, Cd, and Tl (mostly below LOQs), but quantifiable levels of Cr and Sb (concentrations near LOQs) in some e-vapor samples (up to 4.36 pg/mL puff and 0.466 pg/mL puff, respectively). Concentrations of Pb observed in this experiment cannot be attributed to e-vapors regarding the probable contamination highlighted by experimental blank results (Supplementary Table S2). Further optimization is then necessary; however, these concentrations remain lower than those in 3R4F smoke. Williams et al. (12) attributed the origin of elemental emissions in e-vapors to several device components. Interestingly, the e-cig tested in the present study contained nichrome wires that could be potential sources of chromium emission, even if we used new clearomisers for each experiment in an attempt to minimize potential effects on emissions of deterioration of the device. Specific concentrations of chromium oxidation states were not determined in this study. As chromium (VI) is carcinogenic according to IARC (29), further studies should explore this issue, as well as the potential impact of device aging and vaping conditions on elemental emissions. Our results remain consistent with the few available studies which quantified trace elements in e-cigarette emissions (12, 30).
Owing to the potential natural origin of several e-liquid compounds like glycerol, nicotine or flavorings, 50 pesticides were investigated in our study. Only two molecules were found in some e-liquid samples, mostly at trace levels (Table I). Chlorpyrifos ethyl and trifluralin were quantified in some flavored e-liquids, both without and with nicotine, suggesting flavorings to be a more likely source of pesticides than nicotine. However, unquantifiable amounts were also found in unflavored e-liquids, making conclusion on pesticide origin difficult.
In e-vapors, only chlorpyrifos ethyl and trifluralin were detected, but were not quantifiable. In contrast, various levels of α-endosulfan, β-endosulfan, chlorpyrifos ethyl, cyhalothrin, cypermethrin and trifluralin were quantified in 3R4F smoke (Table II). Considering the products tested, our results suggest that pesticide exposure by vaping does not represent a major risk, at least compared to smoking.
PAHs are part of principal compounds released from the tobacco combustion and many of them are classified by the International Agency for Research on Cancer (IARC) as possibly, probably or proven carcinogenic to humans (29). Ten PAHs were detected in e-liquid samples but at low levels, mostly near or below LOQs (Table I). Except for the F-/Nic- e-liquid, the major quantified PAH was naphthalene, classified as possibly carcinogenic to humans, with maximum concentration of 61.8 ng/mL. Kavvalakis et al. (31) failed to detect PAHs in any e-liquid sample, while Han et al. (32) reported quantifiable levels of the PAHs we analyzed in a large range of products. All these data raise the question of the quality of e-liquid manufacture and refer to incoming regulations.
In all analyzed e-vapors, we detected quantifiable amounts of naphthalene and acenaphthylene, but at 136- to 2018-fold lower concentrations than in 3R4F smoke (Table II). In fact, we measured high concentrations of all 16 studied PAHs in 3R4F smoke, comparable to other previously published data (33). Our results are in agreement with Margham et al., (34) who reported similar observations, and suggest that there is no evidence for a substantial exposure to PAHs through vaping, and certainly not when compared to tobacco smoking.
According to some authors, formaldehyde, acetaldehyde or acrolein may be generated by oxidative (35) or pyrolytic (36) reactions during vaping cycles. In particular, formaldehyde classified as carcinogenic to humans, has been described in several studies, at varying levels depending on the experimental conditions. To facilitate comparisons between published data and this study, the following reported concentrations have been converted or extrapolated into the units we used. High concentrations of formaldehyde of up to 1.4 μg/mL puff were reported by Uchiyama et al. (35) who emphasized that carbonyl emissions appear only after several puffs, and varied depending on the e-cig model used. Hutzler et al. (37) reported concentrations of up to 0.9 μg/mL puff, increasing with the decrease in the e-liquid level in the cartridge. Concentrations varying from 1 to 927 ng/mL puff were reported by Gillman et al., (36) depending on the battery voltage and the device used. Kosmider et al. and Geiss et al. respectively reported relatively lower formaldehyde concentrations, up to 24 ng/mL puff with high battery voltage (38) and from 0.5 to 33.2 ng/mL puff depending on the battery power output (39). Thus, while vaping conditions seem to strongly affect carbonyl generation, Farsalinos et al. highlighted that they must remain realistic. They reported that high carbonyl concentrations are only generated in “dry puff” conditions, leading to an unpleasant taste which is not desired by users (40).
In our study, with cartridges fully filled before each puff series, we found concentrations of formaldehyde and acetaldehyde close to 1 ng/mL puff for e-vapor samples (Table II). These results are comparable with Goniewicz et al., who found 0.3–5.3 ng/mL puff (30). No noticeable difference was observed between flavorings, except for acrolein concentrations for which the BT flavored samples gave much higher results than the other samples tested. Although the stability of acrolein may be affected by the sampling protocol used (41), this observation agrees with Khlystov and Samburova findings which suggest that flavorings and their concentrations in e-liquids could affect carbonyl emissions in e-vapors (42). While DNPH cartridges present some advantages when studying e-vapors (43), the model used was easily saturated with cigarette smoke, in part because of very high concentrations of acetaldehyde (data not shown). Concentrations of formaldehyde, acetaldehyde and acrolein in 3R4F smoke reported by Eldridge et al., using the same ISO smoking regime but employing liquid traps, can be extrapolated as follows: 82, 1,540 and 171 ng/mL puff, respectively (18)—concentrations that are far higher than those found in e-vapors in this study.
Our data, taken together with the highly variable published data, highlight the need to develop optimized protocols and to use controlled and realistic vaping conditions to assess carbonyl emissions by vaping.
While our analytical methods are considered reliable, we assume that vapor transfer and collection could suffer from some limitations. Due to potential adsorption or release phenomena in the smoking machine tubing or potential losses with collecting materials, collecting yields for each analyte could not be accurately determined. However, we optimized and harmonized many experimental conditions in order to minimize risks of variation and make e-vapor versus smoke comparison possible including specific inert tubing, consumed e-liquid mass tracking, representative vapor/smoke samples and realistic vaping/smoking conditions.
This study was designed with a few e-liquids and a single e-cigarette model in order to perform concomitant in vitro analyses, so the extrapolation of these results should be made with caution. Considering the ongoing toxicity assessment approach, analyses should be extended to other materials, especially to other e-cig coil designs and generations, as well as to other target compounds including tobacco specific nitrosamines or flavorings.
Conclusion
This study provides a chemical characterization of both a selection of e-liquid refills and their resultant emissions in e-vapors. Reliable analytical methods were used and emission generation and collection were optimized in order to compare both electronic and conventional cigarette emissions. Our results demonstrate on the one hand that the e-liquids we tested comply with their labeled composition and contain only a few pollutants at trace levels and, on the other hand, that their respective vapors contain fewer potentially toxic compounds than tobacco smoke, including trace elements, pesticides, PAHs and carbonyls. Considering existing operating limitations and the importance of such analyses in the assessment of e-cigarette toxicity, harmonized protocols and guidelines are clearly needed to accurately estimate consumers’ chemical exposure and make the results comparable. A short-term exposure in vitro toxicity evaluation of e-vapors generated from the same e-liquids and in the same experimental conditions we used in the present study suggests a lower toxicity of e-vapors compared to cigarette smoke in human bronchial epithelial cells (44). These preliminary works constitute a baseline for further in vivo studies to allow the highly needed long-term toxicity assessment of e-cig. Moreover, other investigations are also ongoing upon the influence of conditions of use of electronic cigarettes on e-vapor composition.
Supplementary Data
Supplementary data is available at Journal of Analytical Toxicology online.
Funding
Analytical fees were jointly covered by Université Lille 2 and INNOVA SAS that markets the NHOSS® electronic cigarette brand. The authors themselves received no financial or other consideration from the electronic cigarette or tobacco industry, while the study was partially funded by INNOVA SAS.
References
WHO | Tobacco WHO. http://www.who.int/mediacentre/factsheets/fs339/en/ (accessed March 23, 2017).
WHO Report on the global Tobacco Epidemic: Enforcing Bans on Tobacco Advertising, Promotion and Sponsorship (2013) World Health Organization. http://apps.who.int/iris/bitstream/10665/85380/1/9789241505871_eng.pdf (accessed July 8, 2016).
E-cigarettes Around 95% Less Harmful than Tobacco Estimates Landmark Review—GOV.UK. https://www.gov.uk/government/news/e-cigarettes-around-95-less-harmful-than-tobacco-estimates-landmark-review (accessed March 23, 2017).
Propylène-glycol (FT 226) - Fiche toxicologique - INRS. http://www.inrs.fr/publications/bdd/fichetox/fiche.html?refINRS=FICHETOX_226 (accessed March 23, 2017).
IARC Monographs- Classifications. http://monographs.iarc.fr/ENG/Classification/ (accessed July 28, 2016).