For each setting, the vape pen was operated until approximately 100 mg of VEA had been consumed; this consumption was typically achieved within 10–20 puffs. In instances where more puffs were required, the vape pen was allowed to rest at 20 puffs for 10–20 minutes to prevent overheating of the battery. Condensed emission products were dissolved in 1 mL of ACN, with 10 μL of 1, 3, 5-TCB solution added to each sample as an internal standard for chemical analysis. Emissions were analyzed immediately after collection or stored at -80˚C to prevent any aging effects.To determine the impact of the device on the degradation of e-liquids, pure pyrolysis of VEA oil was simulated using a tube furnace reactor system . The schematic of the set up for these experiments is shown in S2 Fig in S1 File. An alumina crucible containing 100 mg of VEA standard oil was weighed, and then placed into a high temperature quartz tube furnace capable of reaching temperatures as high as 1200˚C. The tube furnace was initially set to 23˚C, then ramped to each temperature setting at a rate of 10˚C min-1, and then held at the target temperature for 75 minutes to allow for VEA oil to be evenly heated. Inert argon gas was flowed through the system at a rate of 0.18 L min-1 to carry the VEA pyrolysis products into cold trap apparatus kept at -40˚C. After 75 minutes, the tube furnace was programed to return to room temperature before the alumina crucible was removed and re-weighed to determine the amount of VEA that was consumed. Pyrolysis products condensed in the cold trap were dissolved in 1 mL of ACN and concentrated to 100 μL using a gentle N2 gas stream. Then, 10 μL of 1, 3, 5-TCB solution was added to each sample as an internal standard for chemical analysis.Large molecular weight and non-polar degradation products were analyzed by directly injecting 2 μL of sample into an Agilent J&W DB-5MS column for separation. A solvent delay of 6 min was used; the GC was initially set to 60˚C for 1 min, then ramped to 150˚C at a rate of 3˚C min-1, held at 150˚C for 2 min, ramped to 310˚C at a rate of 20˚C min-1, and then held at 310˚C for 5 min.
Smaller molecular weight,outdoor cannabis grow polar degradation products were analyzed by directly injecting 2 μL of sample into a Rtx-VMS fused silica column . A solvent delay of 6 min was used. The GC was set to 35˚C for 1 min, ramped to 240˚C at a rate of 10˚C min-1, and held 4 min. The detailed procedures for the operation of GC/MS can be found in a previous publication.Degradation products were identified using the NIST 2008 mass spectral database. Compounds with probability 50% and match factor scores 800 were considered as good matches. For compounds that were suspected to be present in our spectra but could not be identified using the NIST library due to lack of available standards, Quantum Chemistry Electron Ionization Mass Spectrometry was used to simulate theoretical EI mass spectra of molecules.The detailed procedures for QCEIMS calculations can be found in the supporting information. Peak abundances were normalized to the 1,3,5-TCB internal standard for quantification.Fig 1 shows the temperature profiles of the e-cigarette coil and VEA oil in the cartridge operated at each voltage setting. Peak coil temperature at each voltage setting was fairly consistent between each measurement with no significant increase after consecutive use, which agrees with previous reports. Though the starting temperature after 1 min of rest increased slightly with subsequent measurements, the starting temperature never exceeded 33˚C. In contrast, the temperature of the oil in the cartridge increased with each subsequent measurement until seeming to plateau. The peak temperatures of both the coil and the oil were then taken and plotted as a function of voltage, as shown in Fig 2. Coil temperature showed a strong positive linear relationship with applied voltage , whereas oil temperature increased linearly with voltage until 41˚C , where the peak temperatures at 4.3 and 4.8 V do not significantly differ. This is likely due to the specific heat capacity of VEA; at higher voltages. Visible discoloration to the oil and wick could be seen during temperature measurements, indicating that the specific heat capacity of the oil in the cartridges may have been exceeded and part of the stored VEA may have been transformed before it is vaped .The total ion chromatographs obtained from GC/MS analysis of VEA vaping emissions produced at each temperature setting are shown in Fig 3. Overall, clear temperature dependent degradation of VEA vaping emissions can be seen as the amount and abundance of degradation products substantially increases with increasing coil temperature.
Analysis of the GC/MS results revealed 19 compounds that were able to be tentatively identified based on consistent NIST MS spectral library match scores of 800 or greater. One other compound, 1-pristene, was not found in the NIST library and thus was identified based on comparison with previously reported mass spectra and a mass spectrum generated with the QCEIMS program that found signature fragments of m/z 266, 111, and 126, which are consistent with our results . A summary of the identified compounds and chemical information identified from PubChem can be found in the supporting information . Many of the products described here, such as phytol, 2,3,5-trimethyl-1,4-benzenediol and 2-hydroxy-4-methoxy-3,6-dimethyl benzaldehyde, have not been previously detected from VEA vaping to our knowledge. An isomer of 2,3,5-trimethyl-1,4-benzenediol has also recently been identified as a substantial VEA degradation product at temperatures 220˚C. Authentic standards were purchased for 2-methyl-1-heptene, phytol, and 2,3,5-trimethyl-1,4-benzenediol to confirm identities of observed products . Other compounds, such as vitamin E, DQ, DHQ, 1-pristene, and 3,7,11-trimethyl-1-dodecanol, have been consistently identified as VEA decomposition products . Several products, such as DHQMA or ketene, that have been previously reported in VEA vaping emissions could not be found in our spectra, likely due to the limitations of the emission collection and analysis method described in section 3.4. A heat map of the mass fractions of degradation products generated at each temperature is shown in Fig 4. Products that contribute to the majority of the observed VEA degradation were separated from the total heat map to better visualize the change in each concentration as a function of temperature. VEA, 1-pristene, and 3,7,11-trimethyl- 1-dodecanol were found to be the most dominant vaping emission products at all of temperature settings, while other compounds, such as duroquinone, durohydroquinone, and 2-methyl-1-heptene steadily increase in concentration as temperature increases. Furthermore, certain compounds including 2,3,5-trimethyl-1,4-benzenediol, 2,6-dimethyl-1,6-heptadiene, 3,7-dimethyl-1-octene, and 3-methyl-1-octene are not produced in concentrations above the detection limit of our instrument until 322˚C, which suggests a potential risk that users who operated vaping devices at lower temperatures would not be exposed to. However, while most identified compounds appear to increase in concentration as temperature increases, phytol and 2,6,10-trimethyl-dodecane are produced at detectable levels at 176 and 237˚C but cannot be found at higher temperatures.
Another recent study has also detected production of phytol when vitamin E were heated in a microchamber/thermal extractor at 250˚C . It is possible that at these compounds are stable at lower temperatures but begin to break down into degradation products themselves as the temperature increases. Another important pattern to note is the increase in compounds that may pose a risk of oxidative damage to lungs, such as DQ and 2,3,5-trimethyl-1,4-benzenediol, at higher concentrations. While not investigated in this study, prior research has shown that increased temperature may result in the enhanced emission of carbonyl-containing compounds when vaping e-liquids containing propylene glycol and glycerin. Thus, vaping VEA at greater temperature settings may also carry the risk of exposure to highly electrophilic molecules and subsequent oxidative lung injury. In order to better understand the interactions between temperature and the generated emission products, a Pearson correlation analysis was performed . Overall,vertical grow system all but four of the identified compounds were strongly correlated with temperature . Compounds such as DQ, 1-pristene, 2-methyl-1-heptene, 2-hydroxy-4-methoxy-3,6-dimethyl benzaldehyde, and 2,6-dimethyl-1,6-heptadiene, were very well correlated with temperature , indicating a strong increase in concentration as temperature increases. VEA and phytol, in contrast, were strongly anti-correlated with temperature , while VE and 2,6,10-trimethyl-dodecane were moderately anti-correlated with temperature . In addition, VEA was found to be weakly to strongly anti-correlated with all degradation products excepting phytol and VE, which demonstrate a strong positive correlation . These results support our analysis of the mass fractions, indicating that as temperature increases, thermal decomposition of VEA is heightened. Further analysis of the correlations between degradation products shows that phytol is strongly anti-correlated with all VEA degradation products with the exception of 2,6,10-trimethyl-dodecane, which was found to have a strong positive correlation with phytol . Phytol was also found to be strongly correlated with VEA , likely because as more VEA was evaporated during the vaping process, the greater the chance of degradation into phytol. These relationships further suggest that while some degradation products may be stable at high temperatures, phytol may further decompose into shorter-chain alcohols, alkanes, and alkenes and enhance the production of VEA vaping emission products. Phytol is known both as a precursor for the synthesis of VE and vitamin K12, as well as a byproduct of chlorophyll degradation . Inhalation of aerosolized phytol has previously been shown to induce lung injury in exposed rats. In addition, phytol is a long chain alkyl alcohol compound, meaning that it has the potential to induce damage to the membrane of cells in a biological system. Overall, the toxicity of phytol raises questions about the safety of vaping not only VEA but cannabis-containing vape products that may result in phytol production.
These results clearly indicate that the product distributions of VEA vaping emissions are highly dependent on the operating temperature of the vape pen. As a result, the exposure for vape users operating the same e-cigarette products at different temperatures may differ significantly.Previous reports of VEA pyrolysis indicate that VEA begins to degrade starting at ~200–240˚C. However, our results clearly demonstrate degradation of VEA and formation of products such as DQ at 176˚C, indicating that the device itself may play a larger role in the decomposition of VEA than initially anticipated. Previous study in our lab has also found substantial formation of DQ at 218˚C–several hundred degrees lower than what has been predicted. To further understand if the device itself may impact the thermal degradation of VEA, pure pyrolysis of VEA oil was carried out using a tube furnace reactor. After 75 minutes, the average mass loss of VEA heated at 176, 237, 322, and 356˚C was found to be 0.11 ± 0.091, 0.37 ± 0.11, 3.7 ± 0.072 and 7.1 ± 0.0016 mg of VEA consumed. At 176 and 237˚C, VEA was fairly stable; substantial consumption of VEA oil was not observed until the two higher temperatures, despite clear consumption at all temperatures during the vaping collection. Fig 6 demonstrates the product distribution of VEA degradation products collected and analyzed using GC/MS. Here, we did not observe substantial thermal decomposition of VEA when heated at 176˚C for 75 minutes, which greatly contrasts with the degradation of VEA at 176˚C for only 4 s during the vaping collection. At 237˚C, the parent VEA molecule was the only detectable emission product, indicating that VEA again did not degrade at this lower temperature, though 237˚C was enough to evaporate VEA so that it could be collected in the cold trap. Degradation products were only detectable from samples collected at 322 and 356˚C, though the number of products and abundance of observed peaks are drastically reduced when compared to the vaping emissions. It should be noted that the tube furnace is capable of heating VEA at more accurate and consistent temperatures than the vape pen itself, which often saw temperature fluctuations that may influence results. The stark difference in product distribution provides evidence that VEA vaping emissions may not be the result of pure pyrolysis alone. Instead, external factors such as the device elements themselves or environmental interactions may play a role in the catalysis of VEA degradation.