Outpourings of Hot Air

I’m pretty sure we’ve been down this road before. The whole formaldehyde scare – leading to the “10 times more cancer causing chemicals” headlines. That particular study (which started life in the NEJM as a letter to the editor) has been roundly debunked – not least of which because to inhale burning e-liquid, is frankly a rather dumb idea. Any researcher that believes us vapers do that on a regular basis, most definitely need to get out more.

But it seems that these budding researchers have nothing better to do than to come up with new ways to thoroughly burn e-liquid so they can tell the media that “e-cigs are bad mmmkaay?”. In this case, more research (funded by the University of California Tobacco Related Disease Research Program) at Lawrence Berkeley National Laboratory have added more wibble.

From the abstract:

This study quantified potentially toxic compounds in the vapor and identified key parameters affecting emissions.

Yes, by burning the liquid. Don’t these folk get it? Well, no of course they don’t. They are just looking for any outlandish excuse to bolster their ever weakening defense against what is proving to be, a safer alternative to smoking – they can’t have that now, otherwise all their funding goes away!

Let’s skip over the usual introductory stuff – the whole “vaping has grown exponentially in the United States” thing, that’s typical soundbite stuff that those in ‘public health’ like to use when informing coercing legislators. So we’ll skip straight to the heart of the matter.

Due to the growing concerns about such passive exposures, local governments and public and private institutions are increasingly restricting vaping in public spaces.

Oh wait, that’s still part of the “introduction” – nice to see these researchers are well versed in how to phrase things for policy makers to go “well this study says this, so let’s ban it”. The rest of the introduction doesn’t get better either. Although I did enjoy this snippet:

Rechargeable lithium ion batteries offer a variety of voltages and storage capacities and provide between 100 and 300 puffs per charge.

General question to my readers – in your device (whatever that is right now) could you approximate how many “puffs” you get from a fully charged cell? I’m genuinely curious!

The temperature at which the eliquid is vaporized is a function of the power output (as determined by the battery voltage and current through the coil) and geometry.

Erm, what? Geometry? Unless they are specifically talking about how the air-flow is directed over and around the coil, I can’t see geometry having much of an actual effect on the temperature at which liquid is vapourised.

Unlike previous “studies” into burning e-liquid, these researchers splashed out a bit and got themselves a CE4 and an Aerotank, they also made their own “vaping machine” instead of using the more traditional smoking machines. You know, just to be..er.. different.

A simple laboratory made “vaping machine”, described in Figure S2 (Supporting Information) was used to generate consistent vapor emissions. Two different vaporizers and three different e-liquids were used in this study.

I’d be willing to bet you can guess which flavours they chose right?


Yep, bubblicious – or more to the point, bubblegum. Seriously, what is it with these researchers and their fascination with bubblegum? Oh, it’s also worth pointing out that they didn’t know the PG/VG ratio from the other flavours. Bubblicious is available in two ratios – a “drip” variety 30/70 (PG/VG) and a “standard” variety 60/40 (PG/VG). Also, the Mojito Mix is from the same maker, but for some reason, that and the Apollo e-liquid were at 18mg/ml of nicotine, while Bubblicious was at 24. Nothing at all like consistency there.

All vaping experiments were carried out using a Vision Spinner II battery with variable voltage from 3.3 to 4.8 V as the power source.

Oh my. Enter stage left, flaw number one. While there are still plenty of Vision Spinner mods out there with the variable voltage (I had one just under two years ago) it is not the most common on the market right now. In fact, I’m relatively sure that the Vision Spinner II isn’t as widespread as it once was.

The puffing protocol consisted of 5 s duration puffs at 600 mL min (-1) (puff volume 50 mL) and interpuff periods of 25 s. Each puffing cycle included a total of 50 individual puffs, performed over a 25 min period.

Enter stage right, flaw number two. This one is by far the biggest flaw out of the whole study (and that’s saying something). For instance, on my Mini VTC with a Nautilus X at 15W, my average “puff” is around 2.5-3.4 seconds in length. Just for kicks, I’ve just taken a 5 second long puff. Never again. On a tank that is designed for mouth to lung (to mimic the act of smoking), it is actually rather painful (for me at least) to draw for five seconds. To then repeat that every 25 seconds is as far from real world as you can get.

Just for my own curiosity, I asked my followers on Twitter what the average (MTL) puff was, the majority of those that responded were like me – between 2 to 4 seconds. Had the occasional outlier with a 4-5 second puff. I guess the follow-up question to that would be – how often they puffed?

Of course, the researchers needed to know what was in the liquid they were testing – you know, for baseline readings and so forth. So they decided to “incubate” the liquids at different temperatures between 80 and 280 degrees centigrade. This is what they found from the 200 degree incubation:


From the liquid incubation, the researchers deduced that because the flavouring compounds were present in the liquid, and because at 200 degrees the incubation temperature were comparable to those reported near the heated coil (but not the coil itself), the compounds were likely to be in the generated aerosol, and in greater quantities. Let’s not forget that the CT liquid is a standard tobacco flavour.

Unless I’m completely useless, if you take 10ml of liquid and vapourise it, the component parts of that liquid (if it ever returns to liquid form) will remain in the same quantity or less. So with that in mind, how can a portion of the liquid actually create more of itself in aerosol?

This is where it gets interesting. Not only did the researchers incubate the liquid at 200 degrees (apparently close to the temperature measured somewhere near the coil), they also measured the temperature of the vapour itself.


Notice anything “unusual” here? The vapour generated from the CE4 was warmer than that from the AeroTank Mini. Well go figure Einstein. The AeroTank was so named because of the airflow. More air flowing over a coil at a set level (and we’re talking early hardware here) will result in a cooler vape. Less airflow means the vapour is warmer. That’s not exactly rocket science now is it?

In both devices, the vapor temperature increased quickly over the initial 5-10 min, corresponding to the first 20 puffs, and then approached steady state.

Well, there’s a no shit sherlock moment. But are you ready for a facepalm moment?

The emission factors of several vapor constituents, expressed in nanograms of compound emitted per milligram of e-liquid consumed,

Unlike in their simulation, vapers don’t inhale all the vapour. Some of it actually escapes, especially in mouth to lung when some is expelled (at least in my case) prior to the actual inhalation. So to say that the nanograms emitted per millilitre of e-liquid consumed is being inhaled, as is being implied here, is ridiculous. It is of course completely different for those that like a direct to lung hit, but even then a lot of the vapour still escapes – usually through the airflow.

But it gets better.

Aldehyde concentrations were determined with a precision better than 10% in most cases for the EGO device operating at 3.8 V or higher settings. However, lower concentrations measured at the lowest setting of 3.3 V had significantly higher errors.

Now this reminds of another piece of research by, none other than Dr Konstantinos Farsalinos. Dr Farsalinos also presented at the Global Forum on Nicotine and referred to this exact issue. Amazingly, the researchers of this study did cite Dr Farsalinos’ work, but only in relation to the puff cycle and not to the simple fact that if you burn e-liquid you will get aldehydes. Not to mention that in those circumstances, actually inhaling the vapour is incredibly disgusting and painful.

Remember that CT liquid being a standard tobacco flavour? Well, even though there weren’t any additional flavourings, it still produced some nasties at 3.8V and above:


But here’s the thing, the data displayed in Figure 2 (above) is presented as a result of mass of compound emitted per unit mass of e-liquid and not as a mass of compound per puff. Completely different scaling. As we know the AeroTank was a “better” tank than the CE4, not least of which due to the coil design – bottom feed rather than top-feed on the CE4, but the results shown in the study don’t reflect that. It is only reflected in the supplementary information:


A rather significant difference. (h/t Shawn Avery at Daily Drip for highlighting that)

That’s exactly what happens when you get a dry puff – which we already knew. But according to the researchers own words, they had more accurate results at 3.8V. In other words, readings generated from use at 3.3V were ND – not detectable, or not detectable at any significant quantity:


This demonstrates the “detected levels” at the steady state (temperature wise) of a CE4 using the tobacco flavour. Steady state refers to the higher temperature levels (shown in Figure 1 above).

Our findings suggest that most of the additional energy provided by increasing the voltage in the range of 3.3−4.3 V was used primarily to evaporate larger amounts of e-liquid, but further voltage increases to 4.8 V led to a marginally higher amount of evaporation combined with greatly enhanced decomposition of the main e-liquid constituents.

Well there’s another no shit sherlock moment. The more energy applied, the greater the vapour generation. It’s not as if we didn’t already know that. But there’s more.

The amount of acrolein formed at the maximum voltage was an order of magnitude higher than that formed at the lowest voltage. These observations are consistent with the hypothesis that these compounds were formed as byproducts of the thermal decomposition of the e-liquid main constituents, a process that is extremely sensitive to thermal conditions at the coil.

Well of course your going to find more acrolein at higher voltages than at lower ones – it’s known as the dry puff, you should know you referenced Dr Farsalinos’ work!

In a separate experiment, a single device was used repeatedly at 4.8 V over nine consecutive 50 puff cycles. This test allowed for examination of changes in the emissions associated exclusively with device aging (without cleaning) under conditions that may be commonly used by many vapers.

So use a CE4 for 5 seconds at a time, every 25 seconds for 25 minutes, then repeat that nine times and not clean the coil to mimic “conditions that may be commonly used by many vapers”. No. Never in a gazillion fucking years would I, or any other vaper do that. To properly research a subject you must have a basic understanding of the subject matter – in this case, us. Something that is quite clearly lacking here.

Since harmful chemical emissions are primarily due to thermal decomposition of e-liquid constituents, reducing these temperatures is a promising approach to limiting the harm caused by e-cigarettes.

In case you didn’t know, these researchers have no idea that the industry and technology involved with vapour products has already evolved to that point. But it gets even more interesting.

To explore the hypothesis that harmful emissions originate from thermolysis of the main e-liquid constituents, additional experiments were carried out using neat PG and glycerin instead of the e-liquid, and measuring emission factors for the VOCs and aldehydes identified earlier in the study.

So after seriously abusing a single coil with 5 second puffs at 25 second intervals for 25 minutes nine times the researchers then decided to “vape” pure PG and pure VG – at 3.8V – where they’ve been “finding” the most nasties and this is what they found:


Remember, thats nanograms per milligram. Toxicology 101 folks, the dose does not equal poison.

But why worry about understanding the subject matter at hand, after all it’s not as though this particular research hasn’t been done before is it?

(Image credit Diego Cervo/Shutterstock.com)

10 thoughts on “Outpourings of Hot Air

  1. I have no idea how many puffs I get from my fully charged RX200. Fucking hundreds. Lots. More puffs than several expert county people could realistically count.


  2. It’s funny, the scientists were just looking for ways to make it safer… Media ran with it… In a different direction…

    but…. they tested outdated devices…. why not test a temperature control device at 220f see what happens…


    • If only they were looking to make the devices safer. This is merely a repeat of the NEJM shenanigans to sow confusion.

      But I would like to see them try the experiment on a temp control device – but I’m not going to hold my breath for them to do that ;-)


      • It WOULD be nice, if someone actually did an analysis like this using up to date hardware, and real world settings/use. To do that, though, they’d actually have to talk to VAPERS (OH, the horror!) Publish the results alongside the data from a similar analysis of cigarette smoke, and give us the opportunity to make an honestly informed decision


        • Oh I totally agree – Dr Farsalinos did just that – admittedly on older hardware just to counter the NEJM shenanigans – included vapers as part of the study – which, the authors of this one cited, but clearly didn’t read ;)


  3. Wow! This article is longer than I expected, but I like it and you certainly have a way with words: “Well, there’s a no shit sherlock moment. But are you ready for a facepalm moment?” ;-)


  4. They did the research that way because they were paid to do it that way to make sure they got the results the customer wanted because that’s how science works these days.

    Fortunately Galileo, Copernicus, and Einstein weren’t funded by the National Science Foundation.


  5. I cannot find the original article about the research. I found only related news and some article which I can access for 40$. Do you have the link to the original paper?


Comments are closed.