Added to thread as a postscript
There are no toxic compounds that can be found in ecig vapour at significant levels when tested within normal parameters: that means if the tests exactly replicate normal use by a human user. However, by use of machine testing, it is possible to obtain results that indicate a user would be rapidly poisoned by toxic smoke, and if they survived would have been exposed to significant levels of carcinogens.
Machine tests of ecigs are useless (and sometimes deliberately so) unless:
1. A human user checks the output of the ecig to ensure it is satisfactory (machines cannot detect toxic smoke created by out of parameter operation). Humans don't continue to inhale smoke produced by an ecig that has obviously run out of liquid or is being operated inverted (which produces the same result: smoke). Only a human can detect if an atomiser is working properly, no machine can do this. (Although machines can probably do this, by extrapolation: if no toxic compounds are detected then the ecig is being operated correctly; because if no toxic compounds are present then there is no smoke, so the machines have determined a correct result.)
2. Until photos of the lab testing set-up are checked to ensure that the device was operated at an angle well below the horizontal, since ecigs are a gravity-fed liquid-feed device that do not work upside down or even at an angle above the horizontal (very much like an electric kettle). Many photos show a set-up where the ecig was either above the horizontal or even inverted completely. The guaranteed result with such erroneous operation is toxic smoke.
3. Until the operating schedule was checked to ensure that it did not exceed a set number of puffs of a set length at a set vacuum with a set timegap between puffs with a set number of total puffs per cartridge or liquid refill size at a set max atomiser temperature. (In other words, there is a set inhale length, which must be at a set strength of draw suction, and a set inter-puff interval time, and a set max number of puffs per minute or other set time period, and a set maximum puffs per head refill, and the atomiser temperature must be monitored to ensure it never overheats.)
These are referred to as the test protocols. Unless the protocols are precisely defined then the tests are worthless. No protocols have been defined as yet, so all tests appear to use speculative protocols that have no agreed value for determining measurement of compounds declared. It is extremely easy to produce results that appear to indicate ecig vapour is toxic. However, no human user could inhale the smoke produced from these faulty (and sometimes deliberately faulty) tests, which are clearly the result of running the head dry and burning up the atomiser. A human can detect the instant the vapour goes out of parameter. Machines just continue to test the smoke that results from faulty usage.
Additional notes
Atomisers normally run at 60 - 70 C. If it is deliberately run dry (or runs dry as a result of faulty protocols such as inverted operation) then it will climb to 300 C or more. No human can possibly inhale the product of such faulty operation - smoke is formed by the melting of internal components. To a machine, the result is normal and contains a multitude of toxic compounds. None are present in normal vapour, or are only just detectable (the dose makes the poison). Therefore, a lab test where the atomiser temperature was not constantly monitored is a useless test, since the atomiser might have run out of parameter. Humans stop inhalation when operation goes out of parameter.
Formaldehyde is present in everyday air, in differing amounts according to location and other factors including human presence. Human cell operation requires formaldehyde; formaldehyde is exhaled. New houses generally contain high levels of atmospheric formaldehyde. People exhale it and it is not toxic except at high levels. In order to measure formaldehyde in ecig vapour, the following procedure must be used:
a. Air is drawn through the test ecig with an empty carto attached, so that the room air formaldehyde level can be measured first.
b. Then the ecig is operated normally according to agreed protocols, and the formaldehyde level in the vapour is measured.
c. Level (a) is subtracted from (b) to give the ecig-generated formaldehyde level.
I have not seen any lab tests that accord with that operating protocol. Results published are spurious and technically indefensible since there are multiple areas of possible error. For example the room air level was not subtracted, or the ecig test protocols were not demonstrably acceptable.
In order to measure and attribute room air formaldehyde levels to ecig or e-liquid use, the following tests must be carried out:
a. Measure the test chamber atmosphere after the subject or subjects have been present for 15 minutes without ventilation.
b. Ventilate the chamber.
c. Measure formaldehyde levels after 15 minutes of ecig use.
d. Subtract (a) from (b), and that is the amount generated by ecig use.
There is no other valid measurement for formaldehyde. No clinical studies of room air formaldehyde levels after vaping appear to have used that method.
Acrolein cannot be detected in vapour when supplied as-used by a human operator. It can easily be created when an ecig is run on a test rig in a way that no human would inhale the resulting product (hot, unpleasant dry vapour verging on smoke).
As an example of this, Intellicig (Ecopure, CN Creative) were the first to widely use and promote glycerine as the main excipient (as against PG). They are a large and well-funded operation, and their production is overseen by qualified academics from the local university and further afield. They were the first to run extensive testing on their hardware and refill products as a requirement of medical licensing (this is around three years ago). They tested their 100% glycerine liquid for acrolein at normal atomiser temperatures, then at far higher temperatures, and found no measurable acrolein. If they had found any it would have been difficult for them to continue with their business. (Note: Intellicig were not testing the vapour, the university staff including professors of chemistry were doing the testing.)