OK, in talking with all these brainiacs, none of them were surprised by the delta between these two graphs. Although Dr F for example says we should not use "per puff" values at all since the
juice consumption per puff can be so variable depending on the device (#of coils, wick efficiency, etc), and Dr Kurt also recommends mg\g as the standard. It levels the playing field.
The 1st graph is mg\g of formaldehyde per gram of ejuice,
The 2nd graph is mg\g of formaldehyde per puff. (Which upon discussion I realize is too variable to benchmark)
If you look at the actual data there is some variation of weight from puff to puff, just like in real life, but the averages show the picture. Even though the sample method was the same in all tests, ie 55ml of aerosol drawn over a 3-4 sec period, the average of 5 puffs for 100%PG was 33g, the avg of 90/10 was 30g, and the avg of 50/50 was 23g.
Since the 50/50 juice weighs less, less juice (by weight) was consumed per puff. This explains why the 50/50 shows lower per puff vs the 100%VG in the mg\g chart, the 90/10 was at the bottom of both because of the much lower boiling point.
Now I know the common perception is that PG has less formaldehyde than VG, although the data I measured doesnt back that up, and in fact neither does the Wang or Kurt data. VG has more Acrolein which is even worse than formaldehyde. What Wang measured was total Carbonyl compounds (which includes Acrolein), what Kurt measured was 50/50 juice by device. Neither separated formaldehyde alone by type of juice.
So in any event, I havent seen an apples to apples study that confirms PG is less formaldehyde than VG. It may be less total aldehydes though.
I have 2 more tests I want to do.
- A test of 100% PG by temp
- A test of a carbonized coil over time using 50/50.
Unfortunately, my employer want his scale back. I can probably stall until Monday and get the PG test done. The gunked coil test may have to wait. I am scouting ebay for an Analytical Scale but havent found anything affordable yet. I need a 320g x .001g scale to do the work I am doing. If any of you hear of a crackhead deal on one, please PM me. There are some cheap Chinese clones out there but I read bad things about their standard deviation.
In any event, I have proven that nasties are greatly increased at high temps, and really thats all I set out to do. The exact specific numbers will change from device to device, and vape style to vape style, but i have validated the phenomena on real vape gear, under real conditions. I think its getting time to hang up my research hat and turn this over to the pros, who are all procrastinating about publishing due to commercial concerns.
Oh, and I did get a kind offer from one researcher:
I do have an HPLC (an Agilent 1290 Infinity II) you are welcome to come
play with it any time you like.
Unfortunately it would be 450 mile commute each way.