In searching for information regarding E85 ethanol % and maintenance, I stumbled upon a thread on the EvolutionM forums documenting injector gunk buildup on E85. They referenced a thread on the Colorado DSM forum where a supposed chemist (user thiazole) analyzed it and discovered what it most likely was and why it builds up. Since only the first page of the original thread is accessible via the wayback machine[0], car forums can be somewhat ephemeral, and this information doesn't seem to be common knowledge, I thought it prudent to document the findings here.
The following is a combination of posts from thiazole taken from both the archived CODSM thread and later sections of it quoted on the EvolutionM thread.
CODSM "E85 and injectors" Page 1
thiazole: I suppose if could be gunk from the tank that dissolves, passes the filter, but crashes out during atomization and builds up on the end of the injector. I've actually seen a similar phenomenon in the lab, now that I think about it, where stalactites of compound formed on the tip of a purification column where the solvent exits the column.
JSMCPN: I am definitely not a chemist, can ethanol be converted back into sugar? Could the black gunk be E85 cotton candy?
thiazole: No, sugar is a very complex molecule compared to ethanol. Entropy allows us to go from complex to less complex easily, but it doesn't let us go back without a huge price. Even if I ignored all the chiral centers in sugar, I would find it very challenging to convert ethanol back to sugar in the lab.
Cloud: I'm not sure how much freedom and facilities you have at work thiazole but it would be interesting if you could take a sample of fresh e85 from the pump and see if there was anything in it that would explain the formation of this presumed precipitate on the injector, then maybe take a sample from someone who has the issue pretty badly like Jack and compare it to see if there is anything striking that you can find that may or may not give insight to the phenomenon.
thiazole: Assuming it is organic (which if it is a sludge, it is probably organic), then I can take a sample of the sludge along with a sample of E85 and get proton NMRs of each and tell you if it is coming from the E85. If it is inorganic, I don't have the ability to analyze it, but again, inorganic compounds (think metals or salts) tend to be hard solids. I can take samples from rubber hoses, etc and try to match them as well.
thiazole: Well, this is just what I do for a living - in a nutshell, I take small molecules and piece them together in an orderly way to make large molecules in an attempt to make new pharmaceuticals. I'm sure if you started talking about circuit board design that it would sound like Greek to me.
JackM: I might be going to Denver this weekend or early next week. If I do, I'll bring a nice injector for you to analyze if you wish Let me know.
thiazole: Sure, let me know when you'll be up. I have a 3 day weekend with President's day, so if it is Saturday through Monday, I will likely be able to meet up. I'll bring a vial to collect the material. If anyone is interested in how I plan to analyze these samples, this is what I'll be doing: http://en.wikipedia.org/wiki/NMR_spectroscopy
thiazole: Well, our machines aren't quite as impressive as the the 900MHz monster pictured in Wikipedia. That beast is probably a $10 million machine. Our machines don't have built in ladders or a spiral staircase to get to the top either - we have a small step stool. Really, the sample remains separate from the machine (it is in a glass vial) and it only takes about 10 minutes to analyze, so it isn't a big deal.
These machines use superconductors to create mega magnetic fields. The greater the magnetic field, the better they work. If you walk by our machines with a analogue watch, the hands of the watch will freeze up and will look like they are tweaked. It will also kill pacemakers. That machine in Wikipedia would probably rip a pacemaker out of your chest (well, probably not quite). Seriously, though, when they get that big you have to remove all metal from your body before you get near it or it could rip it clean off.
thiazole: This is what our machines look like. They look like toys compared to that 900MHz machine. http://smct.tnw.utwente.nl/nmr_ms/nmr/400/
thiazole: So I got a proton NMR of the E85 that I'm using, and man, it looks like it is almost all ethanol. It looks like E90+ to me (which is bizarre considering that this is supposed to be E70 right now). There is gasoline in it which is the next biggest part, then maybe 1-2% water, and a few unknown trace compounds. It will be the trace compounds that I will look for in the sludge.
thiazole: Here is the NMR of E85. Sorry it isn't that easy to see. The detail is very fine and some of it didn't get picked up. I guess I need to make a higher resolution image for you guys to really see what is going on. Anyway, the A, B, and C are the hydrogen atoms of ethanol. The gasoline peak is just to the right of peak C (I have it circled and labeled, but you probably can't see it). The unknowns are also circled, but the detail is too small to see in this image.
matthewdesigns: Soooo, forgive my ignorance. A couple of questions about the chart, and how the machine works.
How do you arrive at a percentage-based breakout of all the values shown? IE, as you pointed out, gasoline is a tiny peak as compared to "C", so do you add all the peak values together to find "100%", then divide it out?
How does the machine distinguish the hydrogen in ethanol from that of the gasoline? Do you define what you are looking for prior to analysis, so other compounds are displayed as such, instead of as molecular "parts".
thiazole: The machine integrates the peaks so that you get a relative quantity measurement between them. The calculation is a little tricky because a molecule of ethanol is much lighter than a molecule of gasoline (octane) but it also has fewer hydrogens. Ethanol has a molecular weight of about 46 and has 6 hydrogens and octane has a molecular weight of about 114 and has 18 hydrogens. If I ignore everything else, the machine is showing that the 6 hydrogens of ethanol account for about 93.5% of the area and the 18 hydrogens of gasoline account for 6.5% of the area. So from there is is easy to calculate the mole %. First of all, since there are 3X as many protons in gasoline, the 6.5 area % is 3X its mole %, so that would be only 2.17 mole %. But, since gasoline weighs 2.48X more than ethanol, that pushes it up to 5.4 weight %, or about E95.
There is an obvious problem with that, though, and that is that I don't know what all the mystery peaks are. They aren't calling it E85 because it is 85% ethanol and 15% gasoline necessarily. Perhaps it is 85% ethanol, 5% gasoline, and a 10% mixture of other crap. Or, some of the gasoline protons could be buried underneath ethanol peak C and I'm counting them as ethanol when they are really gasoline. Anyway, here is a higher resolution picture of the NMR:
The picture here was not present in the archive.
thiazole (in reponse to matthewdesigns last paragraph): The machine doesn't do any of that. It just gives raw data. There are rules for determining where peaks will show up based on the hydrogens and to be able to interpret an NMR, you just have to know those rules (something I learned in school). Below is a basic set of rules for where different hydrogens show up, but in the end, it is much more complicated than that and there is a lot more than I want to get into (here is an nice NMR tutorial, though: http://www.cem.msu.edu/~reusch/Virtu.../nmr1.htm#nmr1 ).
thiazole (in response to himself, "First of all, since there are 3X as many protons in gasoline"): I should point out, to organic chemists, a proton is the same as a hydrogen atom (which is just a proton and an electron). I tried not to use the term "proton" for a hydrogen, but old habits die hard. We never talk about nuclear chemistry, so it doesn't confuse us to use that term, but I can see it being very confusing to anyone else.
thiazole: I think I see xylene in the mixture. That would be consistent with the unknown that shows up at about 7.4 and at 2.4.
EvolutionM "E85, Injector Gunk & Seafoam." Quotes
The following parts are quotes taken from the evolutionm forum not present on the first page of the original codsm thread. Note that not all of these are from thiazole, but the quotes weren't labelled with usernames.
Alright, I finished my analysis of this stuff, and my finding is very unexpected. I think it will probably surprise everyone else as well. This doesn't appear to be forming because ethanol is "such a good solvent" but because ethanol is a poor solvent. I suppose it should have been obvious when others said that it "washes right off with gasoline". Why would something that ethanol is selectively dissolving wash off with gasoline? If this were something in rubber or from our fuel tanks, wouldn't that imply that gasoline would dissolve it even more readily than E85?
Alright, so what is this stuff? It is a appears to be a very large petroleum based hydrocarbon, similar to Vaseline. There isn't a single hetero-atom in the molecule (ie, the entire molecule is comprised of hydrogens and carbons), but the molecule is very large. It is also completely aliphatic (ie, only single bonds in the structure - no double or triple bonds). Where did it come from? I can only think of two different sources it could be coming from. It is either something that is mixed in with the rubber hoses that is meant to dissolve away in the gasoline, or it is a trace impurity in the 15% gasoline that is in E85 that wasn't separated during the fractional distillation process. Because it is such a large molecule, it wouldn't be very soluble in ethanol and could easily crash out of solution at the injector.
Well, here is what I did just so everyone is clear. I filled a 40mL vial with E85 and blew it dry with nitrogen gas and mild heating (about 150*F). After there was no fuel left, I placed it under high vacuum to remove any remaining volatiles for about an hour. I was left with a clear sticky residue that smelled bad - like nasty frying oil. I dissolved this sample in the NMR solvent and analyzed it and it IS the same goo that was on the injector. There was smaller amounts of some other stuff in it as well, but the same peaks I saw in the black goo were in this residue. The black goo IS coming from the E85. It isn't naturally black, though. I suspect it just has soot mixed in with it that is giving it the color.
So the next challenge is figuring out why is this crap in our fuel, and if it is in everyone's fuel (particularly people who aren't having this problem).
Gum in E85!
Ok, it isn't chewing gum, of course. I think gum is a generic term for high MW sticky solids. Anyway, if you look at table 1 in this article, it mentions that there is up to 5mg of "solvent-washed gum content"/100mL and up to 20mg "unwashed gum content"/100mL. It think this might be what is sticking to our injectors.
Later in the article it also mentions that mixing E85 and pump gas WILL cause additives to crash out and stick to the injectors and intake runners. I don't think this is what we are seeing since I saw the molecule in a clean sample of E85, but it does open that possibility for others who are mixing.
(link to doc that didn't work)
Unfortunately, l2r99gst (the user quoting these) removed this link rather than checking the wayback machine. The internet was a very different place in 2009.
So I'm in a brief meeting intermission, but I really am convinced that this stuff is the "gum" mentioned in that article. I found another paper that defined the gum as the residue left after evaporation of the fuel. I'm going to see if I can get ahold of the author of that paper and see if I can figure out exactly what the "gum" is to verify that this is what we are seeing.
Thiazole, I think have a lead on an exact chemical name to follow up on in solving this mystery. Read the following:
"Overuse of additives with E85 may result in poor vehicle operation. RFA has also made certain recommendations about appropriate detergent treatment of E85. Some detergents, such as polyisobutylene amine, have performed poorly in FFV operation. At some blend levels, these additives may precipitate out of the blend resulting in excessive fuel system deposition."
So polyisobutylene amine might be the gum that is precipitating and this is a straight up fuel quality issue and not a fuel environment based one...
It wouldn't surprise me if gasoline evaporates faster than ethanol in which case you could see this stuff crash out (since it isn't soluble in ethanol). I doubt it happens in the gas tank, though, or the entire fuel system would be covered in this gunk, which I haven't seen. What probably happens is after turning the engine off, whatever E85 is left on the tips of the injectors and in the intake manifold evaporates off leaving behind this gum. After doing this several times, you get a significant amount of gum formation.
It is dissolved in the E85, but just barely so that it readily crashes out at the injector tip. You can't filter it. It would be like filtering out hard water to remove hard water stains. It can't be done until after it crashes out, but by then it is too late.
You wouldn't want to filter this out even if you could, anyway. The E85 I'm buying has about 10mg of this gum/100mL which means it has about 6000mg per tank. The amount of gum required to clog up an injector is probably only 20mg. The amount to clog a fuel filter would probably be about 1000mg or less. If you could filter it, you'd be clogging several filters per tank. When we scale up the synthesis of a drug, if there is an intermediate that forms a gum and requires filtration, it can actually kill that synthesis and send us back to the drawing board. Gums SUCK and there really aren't many good ways to deal with them other than just dissolving them away.
Q from someone: I was just thinking that if the gum is a component of the gas that's added to the ethanol to make the E85 (the other 15% that's not ethanol) then adding a little more gas that is more highly refined might push the cross over point far enough that we'd never see it... ?
Yes - I suspect that when they add the gasoline to the ethanol that the gum is at or very near the saturation point. If you consider that 10 gallons of E85 only has 1.5 gallons of gasoline in it, adding another 1.5 gallons of non-gum containing gasoline like racing gas would decrease the relative concentration of the gum to gasoline by 50% away from saturation. I think this alone would make a big difference and be pretty affordable if it works.
