Synthetic Cooling Agents in Australian-Marketed E-cigarette Refill Liquids and Disposable E-cigarettes: Trends Follow the U.S. Market

Abstract Introduction E-cigarettes are becoming increasingly popular in Australia, especially amongst the younger population. The synthetic cooling molecules WS-3 and WS-23 have been identified in e-cigarette products from the United States and Europe. The extent of inclusion of these synthetic coolants in Australian e-liquids is unknown, particularly in newer disposable e-cigarettes. Aims and Methods E-cigarettes and e-liquids were purchased within Australia and anonymously donated by Australian users. Nicotine, WS-3, WS-23, and menthol were quantified in the e-liquids using gas chromatography-mass spectrometry (GC-MS). Results WS-23 and nicotine were detected in all of the disposable e-cigarettes with WS-23 often present in high concentrations. There was no correlation between cooling terms in the flavor name and the inclusion of cooling agents. Only three bottled e-liquids were found to contain WS-23 while none contained WS-3 above the limit of detection. Conclusions Synthetic coolants were a common addition in disposable e-cigarettes while rarely added to e-liquid bottle refills. Their inclusion in these products is reflective of trends observed in United States and European e-cigarette products. Implications The increase in synthetic cooling agents as components of e-liquids, particularly disposable e-cigarette devices, has been observed within Australian samples across a range of brands and flavors. WS-23 was present in every disposable e-cigarette analyzed in this study, often in relatively high concentrations. Its inhalational toxicology should be considered when evaluating the safety of these products.


Introduction
Electronic cigarettes (e-cigarettes) are increasingly popular devices that allow users to inhale a vaporized liquid (e-liquid), often containing nicotine and offering potential as a smoking cessation device. 1 Currently, e-cigarettes and e-liquids containing nicotine are banned in Australia for nontherapeutic use, with Australian retailers (retail shops and Australian online suppliers) only allowed to sell nicotine-free vaping products. 2Individuals requiring access to nicotinecontaining e-liquids may import up to 3 months' supply of nicotine-containing e-liquids from overseas providers with a prescription from a medical practitioner. 2The Therapeutic Goods Administration's Standard for Nicotine Vaping Products (TGO110), which commenced on October 1, 2021, sets out minimum safety and quality requirements for nicotine vaping products that are not registered in the Australian Register of Therapeutic Goods and are imported into, supplied in or exported from Australia. 3This includes, packaging requirements and limitations on content such as prohibition of chemicals with known inhalational toxicities.The TGO110 only applies to nicotine-containing products, with nicotine-free products being exempt.This split regulation has resulted in a wide range of unlabeled nicotine-containing e-liquid products being easily accessible on the black market or obtained over the internet.
Within Australia, 9.3% of individuals aged 18 years or older had ever used an e-cigarette in 2021 . 4The popularity of e-cigarettes can be partially attributed to the extensive range of flavors available, including fruity, dessert, and tobacco flavors. 5In addition to the suite of molecules composing each of these flavors, some may also include the presence of cooling agents which produce a cold mouth-feel when inhaled. 6wo of these synthetic cooling agents, WS-3 (N-ethyl-pmenthane-3-carboxamide) and WS-23 (2-isopropyl-N,2,3trimethylbutyramide), provide a cooling sensation without the minty flavor that is generally associated with menthol. 6,7he addition of these cooling agents may also be an attempt to reduce the throat irritation experienced by the user in products with a high concentration of nicotine, 6,8 particularly among nicotine-naïve individuals.Recent recommendations in Australia have promoted the limitation of e-cigarette products to unflavored or tobacco-flavored only in an attempt to reduce the appeal and uptake of these products among youth. 9here is a dearth of research on the inclusion of these synthetic cooling compounds in e-cigarette products available on the Australian market.1][12][13] A recent article by Jabba, et al. 10 found that WS-3 was a common addition in e-liquid refills while WS-23 was commonly found in PuffBar branded disposable e-cigarettes in the United States.Similarly, Omaiye, et al. 12 found that synthetic coolants were common in newer disposable devices compared to e-cigarette products purchased in the United States between 2016 and 2019.Since disposable e-cigarettes are emerging as a popular device type among youth in Australia 14 it is essential to investigate if similar trends are identified in products sold on the Australian market.

Methods
This study examined the content of bottled e-liquids, N = 33 (brands: Air Factory, Aqua, Bam Bam's, Boardroom, Burst, Cloud Nurdz, Hi Drip, HILIQ, Juice Head, Liquid Nicotine Wholesalers, Salt Bae, and VaporFi) and disposable e-cigarettes, N = 37 (brands: Switch, HQD, Gunpod, IGET, PuffBar, Xtra, Trip, and Vaporlax) either purchased from Australian retailers, purchased online from stores which ship to Australia, or donated by Australian e-cigarette users.The e-liquids E1-E24 and disposables D1-D6 were purchased online from Liquid Nicotine Wholesalers in January 2021.The disposables D7 and D8 were purchased from CTC Bulli (Wollongong, Australia) in September 2021.The e-liquids E25-E29 were purchased from VaporFi in January 2020.The e-liquids E30-E33 and disposables D9-D38 were donated anonymously to the University of Wollongong in 2021 with purchased origin provided when specified by donor.For sample details see Tables 1 and 2.

Gas Chromatography-Mass Spectrometry (GC-MS)
Mass spectra were obtained using a Shimadzu QP2020 GC-MS single quadrupole system with electron ionization (Japan).Samples were introduced onto an SH-Rxi-5Sil MS capillary column (30 m × 0.25 mm × 0.25 µm) by injecting 1 µL of solution.The final method had a total run time of 13.67 minutes with a solvent cut time of 1.8 minutes.The injection temperature was 270°C with a split ratio of 1:10.The oven had an initial temperature of 60°C and was increased to 250°C at a rate of 15°C/min and held for 1 minute.Helium was used as the carrier gas at a flow rate of 0.95 mL/min.The mass spectrometer ion source temperature was 250°C and the interface temperature was set to 270°C.The mass spectra were obtained in scan mode over the 50-1000 m/z range.
Quantification of nicotine, menthol, and coolant molecules within e-liquids and disposables was achieved through creation of internal calibration curves using quinoline (Thermofisher Scientific Australia, North Ryde) as the internal standard (25 µg/mL).Analytical standards (≥98% purity) of WS-3 (N-ethyl-p-menthane-3-carboxamide), nicotine, and menthol were purchased from Sigma Aldrich (Castle Hill, Australia) and WS-23 (2-isopropyl-N,2,3trimethylbutyramide) was purchased from Tokyo Chemical Industry (Chuo City, Japan).Individual stock solutions of each quantified compound were prepared in methanol (Thermofisher Scientific Australia, North Ryde) at 100 mg/mL.Volumes of these 100 mg/ml stock solutions were combined to obtain a mixed stock from which the working standards were obtained.Calibration curve concentration ranges were selected to prevent detector saturation and were as follows: Nicotine: 8-400 µg/mL, WS-23: 5-250 µg/mL, WS-3: 4-200 µg/mL, and menthol: 4-200 µg/mL.Six working standards were then analyzed by GC-MS to create the respective calibration curves.

E-liquid Sample Preparation
The dilution factors for the e-liquids were selected to prevent saturation of the GC-MS detector during analysis and optimize the peak intensities for each flavoring molecule.The e-liquids (N = 33) and disposable samples (N = 37) were all diluted 1:200 (v/v) in methanol and spiked with quinoline (internal standard) at 25 µg/mL before analysis.All samples were prepared and analyzed in triplicate.

Results
WS-23 was detected in every disposable e-cigarette sample analyzed in this study (Table 1) with concentrations ranging between 3.6 ± 0.2 (D30, Puff Bar O.M.G) and 37 ± 2 mg/ mL (D32, Vaporlax Iced Cola) with a mean of 16.4 mg/mL.WS-3 was identified in only three disposable samples (D14, D22, and D27, 8%), all at a relatively low concentration (<3 mg/mL) compared to WS-23.Menthol was present in 27% of disposable e-cigarette samples (n = 10) with a mean concentration of 2.1 mg/mL, although only two samples (5%) were named as mint flavored (D15 and D27).Interestingly, D27 ("Puff Bar Cool Mint") contained no detectable menthol despite its description as a "mint"-flavored e-liquid.Even though all the disposable devices contained cooling agents, only 38% of samples contained a cooling term in the flavor name (eg, "Iced Lychee").
For the bottled e-cigarette liquid samples analyzed in this study, WS-3 was not detected in any samples and WS-23 was only detected in three samples (4.5 ± 0.2 to 7.36 ± 0.07 mg/ mL; mean 5.6 mg/mL).The liquids which contained WS-23 were all fruit flavored and included a cooling term such as "iced" or "freeze" on the label.
Nicotine was also present in all disposable devices analyzed, with concentrations ranging from 26 ± 1 mg/mL (D27) to 41.8 ± 0.5 mg/mL (D22) and a mean concentration of 36.2 mg/mL (Table 1).The nicotine concentration of the bottled e-liquids was significantly lower than in the disposable devices (p < .001),with concentrations ranging from 2.74 ± 0.05 mg/ mL (E32) to 37 ± 1 mg/mL (E2) and a mean concentration of 14.7 mg/mL (Table 2).In line with this finding, the bottled e-liquids contained mostly free-base nicotine samples, in contrast with the disposable devices, which were nicotine salts.

Discussion
Our study determined the concentration of nicotine, two synthetic cooling agents (WS-3 and WS-23), and menthol in a range of disposable e-cigarettes and e-liquids available to Australian consumers.These synthetic cooling agents have been found to be increasingly used in the newer disposable e-cigarettes in the United States with high concentrations often observed. 10,12Australian disposable e-cigarettes were all found to contain WS-23 in concentrations comparable to those reported by Jabba, et al. 10 and Omaiye, et al. 12 Overall, use of WS-3 was uncommon (4%, N = 70) and only detected in disposable e-cigarettes, while WS-23 was more common in the Australian disposable e-cigarettes (100%, N = 37) than in bottled e-liquids (9%, N = 33).Some differences were observed between this study and previous U.S. studies. 10,12The PuffBar Cool Mint disposable (D27, Table 1), analyzed in all three studies, had a lower concentration of WS-23 in this study.Interestingly, menthol was not detected in D27 in this study and WS-3 was detected in a much lower concentration compared to the previous studies.Most of the samples analyzed in this study were donated as used devices (Australian users, Wollongong area) which may have affected the concentration of some analytes in the residual liquids, particularly volatile molecules.As a result of this, it is likely that the more volatile compounds, including menthol, have been preferentially vaporized leaving little to no concentration of these volatile compounds at the end of the device life.The concentration of menthol measured may have been different in a new device of the same brand and flavor but was not analyzed here for comparison.The Puffbar Pomegranate disposable (D31, Table 1) was found to have a WS-23 concentration of 10.5 ± 0.1 mg/mL in our study.Jabba, et al. 10 did not detect any WS-23 in a similarly labeled U.S. sample while Omaiye, et al. 12 detected WS-23 at a lower concentration.Despite no apparent correlations between nicotine concentration and presence of WS-23 in e-liquid refills, the addition of WS-23 in disposable cigarettes is more widespread than in e-liquid refill solutions.The inclusion of WS-23, in disposable e-cigarettes which all contained high concentrations of nicotine, may be an attempt to reduce the unpleasant throat irritation experienced by vapers upon inhalation of the high nicotine concentrations identified in these devices or allow tolerance to more intense flavors. 8,10This is particularly relevant in samples where there is no reference to "cool," "iced" or "freeze" terminology on the label, where high concentrations of nicotine and WS-23 were still detected.The recent increase in the use of these disposable e-cigarette devices has been noted among nonsmoking youth who may be unaccustomed to the irritation caused by nicotine. 14Thus, it is possible that the inclusion of cooling agents may facilitate the uptake of e-cigarettes among youth.
The similarity between results from the United States and Australia demonstrates that trends in disposable e-liquid composition are comparable between the two countries, likely due to the same manufacturing sources.This study has also confirmed that this trend can be seen in a large range of e-cigarette brands and is not unique to PuffBar disposables.While WS-23 has not been included in the current bans of ingredients by the Australian government, it is concerning that concentrations observed here, are similar to those observed by Jabba, et al. 10 that have been identified as unsafe based on thresholds recommended by the Joint FAO/WHO Expert Committee on Food Additives (JECFA). 15Cytotoxicity studies by Omaiye, et al. 12 found that concentrations of WS-23 below those