Attached below are four PDFs of research on electronic cigarette liquid/vapor that I came across. There are some interesting results, but on the whole the tests were not properly managed, and the results are not coherently presented or in a logical - or usable - format. There are still too many questions to be asked.
I wanted to examine this research because there is an ongoing issue with Suppliers publishing various statements about the content of e-cig vapor, apparently without there being any evidence to support these statements, such as, "It's only water vapor", and so on.
Some relevant points follow.
The vapor was analysed before inhalation / filtration, ie as directly output by the e-cigarette.
Some of the test set-ups were photographed, and these arrangements all show the e-cigarette being operated incorrectly, in a manner in which it cannot function correctly or possibly at all (inverted). It may be assumed that those experiments where photos were not provided also featured incorrect operation of the e-cig, since the physical requirements of ease of equipment set-up seemed to have overruled the possibility that an e-cig might not work in such an unusual position: upside down, with the battery end high and liquid reservoir/mouthpiece at the bottom.
An e-cigarette is similar to an electric kettle in operation: a heating element that operates while submerged in a liquid bath. The element must be immersed or a fault condition will exist. In essence, both an electric kettle and an e-cigarette are gravity-fed immersed-element liquid heaters, and operating an e-cigarette while inverted will be the same as operating a kettle while inverted.
Because the liquid reservoir is in the mouthpiece, and the liquid must run downward into the atomizer, a standard e-cigarette must be operated with the battery/tip either level with, or preferably below, the mouthpiece, or the atomizer will run dry. The result of inverted operation as shown will be poor vapor production, leading to zero vapor production and smoke generation from combustion of internal materials, mainly plastics of different types. The atomizer will run too hot as it starts to run dry, causing it to burn off materials from inside the housing, which might consist of burnt cartridge filler, adhesives or coatings of some kind. It is well-known that melted plastic can result from this, as the cartridge body overheats.
If operated in this unusual, inverted fashion by a human operator, use would cease immediately after the first puff or two because the taste and heat would make it impossible to continue, as the 'smoke' resulting would be difficult to inhale and certainly unpleasant. In fact it might well be smoke - instead of the correct water/PG-based mist.
Analysis was not presented as a coherent, full, final percentage result of vapor ingredients, in any of these tests. The most complete analysis showed around 17% as missing percentages not accounted for. This is not acceptable for any kind of usable result, as a full and final analysis needs to show the precise identity of at least 99.5% of the ingredients of the vapor, including of course water if this is present. Omitting the water content even if this is as low as for example 15% will skew the percentages of other ingredients. If the water percentage is fairly high but omitted, the resulting percentages of other materials will be meaningless.
The most complete result showed, approximately:
...and around 17% not accounted for.
It would seem that this particular cartridge contained a PG/alcohol-based liquid [a][b], which is unusual and cannot be considered representative of e-liquids generally, which often have a 80-20 PG-VG base, and perhaps 2% alcohol in some cases though most have zero. Some e-liquids are 100% VG-based, in practice meaning the base is about 80 - 99% VG and 5 - 20% distilled water, plus ethyl maltol in some cases. High levels of ethanol are not just atypical but anomalous, and some investigation would be needed before it could be accepted that such a result does not indicate intra-laboratory contamination of the sample.
a. Unless the 'alcohol' measured resulted from melting of plastics or adhesives within the e-cig body due to the demonstrated incorrect operation.
b. Ethanol or methanol may be added to e-liquid samples to facilitate analysis. This appears to be a standard method when analysing the liquid content of cartomizers: the refill liquid is flushed out of the container using an alcohol, and then analysed. Perhaps there may have been some form of lab contamination here: introduction of materials before analysis that should have been omitted from the result.
Trying to get some sort of meaningful result from these studies is difficult, but on the evidence presented here it seems likely the average vapor content would be two-thirds water (66%), a small amount of PG (3%), a very small amount of VG (around 1 to 2%), a very small amount of nicotine (around 1%), and a significant amount, about 15% or more, of flavorings (the last being my assumption based on the compounds seen in other analyses).
This applies to the mainstream vapor, i.e. before it enters the lungs, and not after being exhaled. It is likely that the particulate matter would be lower in the exhaled vapor, and the water content higher. This is because some materials would be absorbed by the body, and also because water vapor is present (added to) in all exhaled air, meaning that even if some water was absorbed on inhale, it would be added to the exhaled air, plus additional carbon dioxide.
It must be stressed this is simply a guess. The results of experiments carried out and presented like this may well be wrong.
Real research needs to be done, where the researchers have an advisor present to ensure that (a) the e-cigarette is operated properly, (b) the atomizer is flushed and pre-run before the experiment starts (this has to be done, to avoid polluted primer or burnt-off production adhesives being present in the vapor, as will normally be the case with new atomizers), and so on.
6. Smoking machines
Some labs have a smoking machine used for testing tobacco cigarette smoke. These cannot be used for testing e-cigarette vapor as there are multiple issues; an important reason is that the machine can easily be cleaned after tobacco smoke tests, by flushing with clean air - 'air blast' (and there are set protocols for this).
That method does not work with a water-based vapor as the material clings to the machine's internals and cannot be removed by air throughput - the machine must be flushed in a different way, and so far an efficient and non-damaging method has not been published as a suggested protocol. The first sample tested will contaminate all further samples (as has been demonstrated). This seems to indicate that ad hoc flask systems, perhaps using a large syringe as the vacuum, might be used (for example). We await some sort of agreement on a testing system, usage protocols, and a cleaning protocol.
PG = propylene glycol = propane-1,2 diol
VG = [vegetable] glycerine = propane-1,2,3 triol
You can safely ignore the content of MSDS sheets as regards toxicity unless they quote a very small figure for the toxic amount (ex: LD50 = 60mg). The other info is simply in there for legal purposes and is essentially meaningless, despite the alarmist content. A person would need to drown in the material for it to be harmful in the case of the vast majority of materials for which the MSDS sheet starts out sounding as if it can be used in chemical warfare.
Take propane-1,2,3 triol for example, the MSDS sheet makes it sound as if it is some sort of deadly poison. But you're talking about VG, which has pharmaceutical licenses for inhalation, ingestion and topical application, and is a food and skincare product. You can inhale it, drink it, bathe in it, do whatever with it. It's a harmless everyday product consumed in large quantities by people everywhere, and broken down by the body into carbohydrates. The only thing you wouldn't want to do with it is burn it at high temperature and inhale the result (acrolein), but despite this being the only practical way it can be harmful, this isn't addressed by the MSDS sheet.
Exactly the same goes for propane-1,2 diol, or PG as we know it, you can even inject yourself with significant quantities of it without harm since it is effectively inert (as is done - it's used as the liquid carrier for injectable drugs that don't mix with water). The main difference between PG and glycerine, in this area, is that PG has 70 years' history of safe use and clinical research. PG is used in asthma inhalers, and in the nebulizers used by lung transplant patients. (Contrast this latter use with the MSDS general tone.)
Note that both these materials are approved and licensed by the FDA for inhalation, ingestion and application to the skin, and, in the case of PG, for injection. They are approved to GRAS level, aka Acceptably Safe, and generally regarded as harmless. You will note that this conflicts with the MSDS, which are to be ignored.
It appears that the main purpose of an MSDS is for transport contractors and bulk chemical storage operators. If staff come into contact with it or there is a fire on the premises, the MSDS can be given to the services responding. As far as the use of an MSDS for everyday purposes is concerned, they have no relevance unless the toxic quantity given is very small. As an example the MSDS for water would include this statement:
IRRITANT TO THE EYES AND LUNGS. In case of contact, the treatment advised is extended dehydration.
...and so on. This is the legally-required content - everything is toxic, in an MSDS. Even pure air would require an MSDS if transported or stored. The contents of the MSDS would make it appear a dangerous material to the untutored eye.
4 x PDFs
[PDF 16-3 had to be edited as the 6MB filesize upload failed. All gfx were removed, leaving the text only. It was edited down from 167pp to 15pp. The gfx edited out were GC-MS readouts and not relevant for this purpose]