[SEL] semi OT - acetone in gas for cleaner, better running?

[John Culp]

> As I understand it the problem of knocking is a burn rate problem. Too 
> quick of a flame propagation/burn flexes the top of the piston and we 
> hear the knocking sound.
> I've been told all gasoline is the same with the differences in 
> regular, mid, and premium being the additive packages that slow the 
> burn rate.
> We hear these called octane boosters. What exactly is an octane 
> booster?
> It is hard to imagine that all gasoline is perfect C(8) carbon chains. 
> I'm guessing that it is a mixture of volatiles of lower carbon chains 
> and oils of higher carbon chains, with the average being C(8) long. Do 
> octane boosters make an attempt to tie shorter chains together to 
> reduce the volatility? Or is there some other mechanism at work to 
> reduce knocking/slow burn rate?

Curt, knocking isn't due to too-rapid combustion. It's due to too-easy 
autoignition of a heated flammable mixture (compression providing most 
of the heat), which causes ignition of a large amount of the mixture at 
once rather than an orderly spreading flame front from the point of 
ignition at the spark plug. The very rapid rate of pressure rise is 
what causes the explosion-like knock.

Combustion of fuel-air mixtures consists primarily of a series of free 
radical chain reactions. Typically it's started by a fuel molecule 
breaking at some point, the ruptured bond being split into an unpaired 
electron on each piece of the molecule or radical. Now, if this fuel 
radical bumps into an oxygen (O2) molecule, one of the oxygen atoms in 
that molecule would rather pair with the carbon or hydrogen atoms in 
the fuel radical, so it divorces itself from its oxygen partner and 
hooks up with the fuel radical. Since the oxygen atoms shared two bonds 
(pairs of electrons), that often produces an oxidized radical that 
STILL has an unpaired electron on it, while turning a free single 
oxygen atom (a most reactive free radical) loose to go attack a fresh, 
unbroken fuel molecule. That's an example of a chain-branching reaction 
that produces more free radicals, speeding up the rate of the reaction 
by increasing the number of reactive particles flying around. Another 
is an H2 molecule running into an O2 molecule and making two OH 
radicals, which are vigorous oxidizers and very common in hydrocarbon 
flames. Some free radical reactions sustain the reaction, releasing 
energy but producing just as many free radicals at the end as at the 
start. Still others are chain terminating reactions, like one of those 
free oxygen atoms running into another one and recombining into O2.

Some molecules are slow to break down under heating into highly 
reactive free radicals, or if they do, they tend to take part in 
chain-sustaining or chain-terminating reactions more than 
chain-branching ones, so there's a slow process of oxidation but it 
doesn't rapidly explode into an intense flame. When a flame front full 
of energetic, highly reactive free radicals contacts the mixture, 
though, it's overwhelmed and happily participates in the orgy of 
molecular splitups and reattachments. That's your high octane fuels. 
They burn just as fast as the low octane ones once the flame gets to 
them, they just don't autoignite so rapidly. Hydrocarbons with very 
highly branched structures with many short branches are really good 
this way, also ones with aromatic rings, because they can rearrange 
internally under heating without immediately taking part in external 
reactions.

Other molecules readily break down in chain branching reactions. They 
readily, rapidly self ignite and knock like crazy. That's a great thing 
in Diesel fuel, but a bad thing in a high octane spark ignition engine 
where the fuel and air are mixed before compression. Long straight 
chain hydrocarbons are especially prone to this type of behavior. When 
they get heated and break, they've got unpaired electrons hanging right 
off the ends of the fragments, ready to grab onto any passing molecule.

Octane boosters are mainly substances that "quench" chain-branching 
free radical reactions in one way or another, to delay autoignition. 
The "oxygenates," aromatic hydrocarbons, aniline, tetraethyl lead, 
iodine, bromine, and water vapor all do this by different mechanisms, 
but the results are similar: less knocking.

> I've been trying to correlate physical properties to knocking. At 
> first I was wondering if there was a relationship between flash point 
> and knocking. But kero has a much higher flash point than premium 
> gasoline and yet it is prone to knocking, so that kills that 
> correlation.

You're in good company, as early researchers tried to do the same. They 
gave up. But we have a legacy of that in avgas specifications for a 
Reid vapor pressure below some certain number. That's just because of 
an empirical observation with certain gasolines in the 1930s having 
better knocking properties at lower vapor pressures. I've got a book on 
airplane engines written by the guy who wrote the specification. He 
went on at great length about the history of antiknock research as 
applied to aircraft engines.

> I've also been wondering if there is a relationship between knocking 
> and the explosive range of a fuel. For example, propane has a fairly 
> narrow mixture range at which it will explode. Gasoline has a moderate 
> range. On the other extreme is hydrogen. With hydrogen almost any mix 
> ratio with air will result in an explosive mixture. I don't think I've 
> ever heard a propane engine knock (forklifts for example).

Interesting observation, but the real reason is that hydrogen and 
oxygen is a mixture extremely prone to producing branching free radical 
reactions, whereas propane and oxygen are far less so.

> Your thoughts? What physical properties of a volatile represent burn 
> rate? If we know this, perhaps Bill's question about acetone as a 
> decent non-knocking fuel can be answered.

Ketones and aldehydes can do a lot of internal rearranging under 
heating without reacting with surrounding molecules. They therefore 
tend to be pretty high octane, and they also are often the leftovers at 
the end of combustion when the mixture gets lean and the chain 
terminating reactions outnumber the chain initiating or chain branching 
ones. A lot of those chain terminating reactions produce an aldehyde, 
or sometimes a ketone. Most of what you smell in Diesel exhaust is 
aldehydes left over from partially oxidized hydrocarbons. Same with the 
stinky exhaust of a cold gasoline engine.

Acetone's a pretty good fuel, but it's too expensive for practical use, 
is made from hydrocarbons that would be better used directly, and is 
too good a solvent to be trusted around rubber and plastic fuel system 
parts in high concentration.

John Culp
Bristol, Tennessee, USA