In this presentation, Sebastian Soulet discusses the chemical processes involved in the conversion of glycerol to aldehydes during vaping. He explains how boiling regimes influence chemical reactions, nicotine delivery, and vaporization efficiency, as well as their toxicological implications.
Transcription:
00:05 - 00:35
[Karin Jacobson]
So we move on to our next talk, which is by Sebastian Soulet. He works for Induscience, a private and independent research laboratory. He has developed several instruments and approaches dedicated to the analysis of physical and chemical properties of vaping products. And the title of his talk is Conversion of Glycerol to Aldehydes Using a Vaping Device, Characterization and Toxicological Implications.
00:42 - 06:21
[Video]
Firstly, I express my heartfelt thanks to the organizers for the invitation at this GFN 2025. My presentation today is a summary of an article under review dealing with conversion of glycerol to aldehydes. It aims to illustrate how boiling regimes also manage this chemical reaction. As a background, we previously illustrated that nicotine delivery is correlated to the mass of liquid vaporized and follow vapor-liquid equilibrium specificities. Then, influence of boiling regime on functioning curve was described and identification of a minimum and a maximum power was discussed accordingly to the onset of boiling point and to the critical heat point. Finally, we correlated a significant increase in aldehyde releasing from the maximum power. Our assumption is that under optimal regime, a liquid is vaporized at its boiling temperature which should induce the same molar conversion rate between glycerol and its main byproduct. Overheating regime controlled by film boiling necessarily implies heating of the gas vaporized above its boiling temperature which should be the onset of a significant increase in the molar conversion rate. To the best of my knowledge, expression of byproduct in mole by mole has never been done before. Most of the literature focuses on toxicological purpose, mainly use mass or concentration units. Through several tests, we aim to explore the conversion rate by varying one-by-one vaping parameters. We started by evaluating conversion rate using a reference condition with the Qubis Automizer and its one-ohm coil filled with pure glycerol and tested under a standardized condition of generation at various powers. Then, we post-rated analysis carried on around 1,100 commercially liquids analyzed for TPD compliance, implying the use of a fixed power. Then we used different atomizers to evaluate the impact of ceramic coil or high power device coupled with high air flow rate. Aldehydes were trapped on DNPH solution and analyzed by chromatography. Data obtained using the QUBIS highlights that under optimal conditions, the conversion rate of formaldehyde is around 80 µmol per mol of glycerol vaporized and increased significantly above its maximum power at around 26 W. A similar observation was done through the commercial e-liquid tested at 15 W. However, for a liquid with a low concentration in glycerol, the rates increase, indicating that 15W might be above their maximal power. At 1.1L per minute, the TFV8 leads to the same result as the QBIS, whereas at 13L per minute, the conversion rate is around 4 times lower but remains almost constant. Finally, VECO+, made with the ceramic coil, exhibits the same 80 µmol per mol, but does not have a significant increase in under-overheating regime. Following the experimental approach, I wanted to illustrate how the constant conversion rates are useful in toxicology. I reported 7 toxicological reference values available in French at the agency. Without going too deep, each one is designed for specific purpose. By considering each values, we can estimate a corresponding maximum quantity of glycerol vaporized such node B exit. The central values are calculated from mean conversion rate, whereas values in bracket are calculated from the 5 to 95% range obtained with commercial eliquid. As illustrated, concerning values of 20 mg to 80 mg by PEF are obtained for chronic exposure and mass of liquid vaporized consistent for high power device used with the highest conversion rate and therefore under overheating condition. Through this summary of our article, we quickly illustrated that the identification of conversion rates show how toxicity related to aldehyde is correlated to the mass of a liquid vaporized, and consequently, it concealed the diversity of device or setting to a central quantity as for nicotine delivery. Then the conversion rate shown very low values under optimal regime, highlighting that they are mainly devices intended to vaporize the liquid and unintentionally leads to a minimal degradation. Experimenting in overheating conditions with settings above maximum power result in significant conversion rate and lead to overestimated risk by many orders of magnitude compared to optimal and relevant conditions. Our assumption seems to be validated. However, there are remaining questions that need deeper investigation. Many thanks for the attention.
06:23 - 06:25
[Karin Jacobson]
Sebastian, the floor is yours.
06:25 - 09:59
[Sebastian Soulet]
Yes, so thank you. What I wanted to explain is that we investigated how emissions are generated using vaping devices in the literature. And what we face is that there is an increasing use of high-powered devices to generate aerosol. And most of the time, these devices are not used under what they are required for, but under what authors of these papers selected as a condition valuable for their testing. And what was interesting is that we We've seen that there is no justification for how these devices are used. And most of the time, they are used under the worst conditions that could be set on batteries. So it means that if your batteries can be set at 200 watts or 100 watts, most of the time this power would be set as setting parameters for emission generation, despite the atomizer is not required for this. And what we wanted to show here is how The power and boiling regime manage the conversion rate, because if you set under a required condition, you will see that the aldehyde released would set on a minimal level, whereas when you go above a maximal power, you will enhance some chemical reaction, some degradation, like the one we wanted to show here. What I don't highlight and point is that the scale of the graph was logarithmic scale, so there were many order of magnitude that was reached under overheating condition. And the second point that we don't show here is that here we focus on glycerol degradation. But as there is also cotton, the degradation that was discussed regarding the tobacco leaves are also available. And this is commonly what we call dry eating. But dry eating is mainly an end process of pyrolysis of cellulose. And what would be added under an over-eating condition is all the compounds that was discussed with the tobacco leaves, some byproducts that are not relevant of normal condition of use, but that results from the degradation of the cellulose. We and Roberto will emphasize this later this afternoon is that under laboratory testing it is very easy to generate aerosols that are irrelevant of any normal usage and that are mainly based on the let's say, is the worst case, but which is irrelevant just because there is a lack of understanding of how they are operating and how they are functioning. Thank you.
10:01 - 10:29
[Karin Jacobson]
Thank you, Sebastian. Do we have any questions from the audience? Maybe I was thinking what you were saying. So very often what we see in papers when they do it in the laboratory, they do it under these conditions that wouldn't be used by the consumers. If the consumers were to use above optimal, how would that be? How would this experience be then?
10:30 - 12:25
[Sebastian Soulet]
Like what was saying for the tobacco leaves, it means that when you over it, the system, the first step is that you will dry locally the cotton and by By this drying, you will allow the first step of the pyrolysis of the cotton. And so you will firstly break the cellulose structure, and you will generate some what's called levoglucosone, nevertheless. And then you will also break all these compound, and you will generate what is called furan, which are compounds which are very, sensible test of burning tests i don't know how to say it english but it will give an unpleasant taste and what is also interesting is that if you continue if you do and if you keep the same condition you will always have this smelling because the structure would be degrade, and you will continue to smell the same thing. This is something which is well experienced by user. However, if you do experiment in laboratory, vaping machine are not able to smell and to feel what is generated. And what will be done in the laboratory testing is that the experiment will continue under the same testing. So you will emphasize the degradation, and you will, by the end, collect something which is far irrelevant, whereas the user because it will be, the structure will be degraded. You will felt the unpleasant test and you will necessarily have to change the device to remove this test because it will signal that the structure is over.
12:26 - 12:32
[Karin Jacobson]
Thank you. You're welcome. And we have a question over here, yes.
12:33 - 12:43
[George Cassels-Smith]
Are you seeing much of a difference between an open system and a closed system? testing under the parameters of normal use?
12:43 - 14:16
[Sebastian Soulet]
The main problem, so physically, as you have the same components in knitting system and elements that is used for the liquid retention, like cotton or ceramic, you can face the same problem. But what is different between open and closed system is that with open system, most of the time, you have the capacity to adjust the power supply. Whereas in closed system, most of the time, the power is fixed. And most of the time, I suspect, because I'm not a device manufacturer, And I suspect that all these tests are done by, let's say, empirical tests. I suspect that in Chinese manufacturing, there are internal tests with the device. And some colleagues test the device, and they fix a requirement based on what they felt. But this point is totally unclear. However, there is some range recommended in some device. And it would be really interesting to see how they fix this range. But this is mostly for open system, closed system, or mostly fixed. And this adjustment has mainly been done in internal test of device manufacturer and manufacturer.