Study titled “Benzene formation in e-cigarettes” found that air has more benzene than e-cigs


By Dr Farsalinos

A study was recently published in the journal PLoS One, titled: “Benzene formation in e-cigarettes”. In what should be considered very fashionable in e-cigarette research, it was accompanied by a press statement discussing about the huge risks associated with benzene exposure from e-cigarettes. It is very interesting to once again see the inconsistency between the study findings and the press statement.

First of all, the study was performed based on the finding of high levels of benzoic acid in JUUL (about 45 mg/mL or 4.5%). JUUL is a prefilled e-cigarette with very high levels of nicotine (they found 6% nicotine, i.e. 60 mg/mL) which protonates nicotine with the use of acid (benzoic acid). It is unique for JUUL to use so high levels of benzoic acid; the authors mention they tested commercial liquids and found benzoic acid at levels of 0.02-2 mg/mL (0.002-0.2%). Benzoic acid could be transformed to benzene, but the study found NON-DETECTED levels of benzene in JUUL despite the use of 5 seconds puff duration.

JUUL is not a variable-wattage device, so it was impossible to stress it to high enough levels so that benzene would be formed (and other chemicals, from melting the device). So, they tested some variable-wattage devices (EVOD 1.8 Ohm single coil atomizer and Subtank Nano 1.2 Ohm coil atomizer), using custom-made liquids containing benzoic acid and benzaldehyde. Again they used 5 seconds puffs (which is very high puff duration for mouth-to-lung vaping). EVOD was stressed to 13 W (for 5-second puffs), which is an extreme value. Interestingly, the authors mentioned that the recommended setting for EVOD was 6 W. Subtank was stressed to 25 W, which may be within the recommended power setting but NOT for 5-second puffs. Even 4-second puffs is very high for Subtank at those power settings. Of note, they report liquid consumption of 24 mg per puff at 25 W, which would be unbearable for a vaper doing mouth to lung vaping. For these two device, despite reporting that in commercial liquid they found 0.02-2 mg/mL benzoic acid, they prepared liquids with 9 mg/mL benzoic acid. They also added 10 mg/mL benzaldehyde, a well-known flavoring.

Going back to basic physics, it is the energy (W x s = Joule) that defines whether the liquid will overheat and generate dry puffs. I explained that in detail in the study evaluating aldehyde emissions in realistic and dry puff conditions. In fact, I performed that study as a response to the famous “Hidden formaldehyde” research letter, which was authored by the same researchers as this benzene study. So, although there is no excuse for researchers to forget basic laws in physics, this is twice inexcusable because we repeated the physics in our paper quite recently (just 2 years ago).

The results of the study are very interesting. At the recommended setting of 6 W with the EVOD (I repeat, at 5 second puffs), they found from non-detected levels (in most of the samples) to 0.16 μg/g liquid consumption (keep these numbers). At 13 W (dry puffs), they found up to 24 μg/g. With the Subtank, they found either non-detected or up to 0.19 μg/g even at the extreme power settings.

What do the findings mean? It depends on how you want to look at it. The authors calculated the concentration of benzene in inhaled air, and report levels up to 5000 μg/m3 air (at dry puffs of course). Compared to the ambient levels of 1 μg/m3 of benzene, everyone would think that this is a disaster.

But not really, this methodology suffers from a major problem. Humans take about 12 breaths per minute, i.e. 17,000 (thousand) breaths per 24 h. The volume of air inhaled in 24 h is 20 m3. So, the daily exposure to benzene from ambient air is 20 μg. Even if you assume that Subtank at 25 W with 5-second puffs represent realistic conditions (they are not), you need to consume 105 mL e-liquid per day in order to be exposed to the same levels of benzene as breathing ambient air. For the EVOD under normal vaping conditions, you need to vape 125 mL e-liquid per day.

The press statement mentions that: “The power levels used in the study were still far below those accessible to users on some devices, which can exceed 200 watts”. This statement is similar to saying that: “We crashed with a car in Trafalgar Square with a speed of 100 mph, but still that was far below the 150 mph speed that cars can reach”. Another similar statement would be: “Eating 5 kg of vegetables in one meal can lead to death, but that is still below the tens of kg available in grocery store where customers buy their vegetables”.

I understand it is frustrating to desperately try to find a problem but fail. And this is not the first time, we’ve seen it in the past (and recently) with formaldehyde and other toxic aldehydes (just wait for a couple of papers that will be published soon). However, this still does not prevent the mispresentation of evidence and science. Also, scientists completely ignore the dry puff phenomenon and instead of them verifying realistic conditions in their experiments, they consider theoretical the criticism they get for their own omission! What a disappointment for the scientific community...

In conclusion: “Benzene formation in e-cigarettes: it does not exist…”



Background Image

Header Color


Content Color