Benzene Extraction from Petrol/Gasoline

Benzene is a horrible substance. My fat chemical hazards reference book, which admitedly lists most substances as horrible poisons (I think the old adage of crying wolf applies here) goes bezerk under the entry of benzene, with the words carcinogen repeated about ten times, a multitude of chemical societies listed which class it as a carcinogen, all its horrible effects on the body catalogued, its a right-to-know substance, minimum allowable concentration 0.5ppm etc. The entry itself occupies about a page and a half, thats 10 times the volume for most other chemicals. At the end theres a warning that there might not be a minimum allowable exposure to a carcinogen. In Australia its classed an S7 poison together with cyanides etc, which means its purchase is registered and controlled.

After reading all this information I was truely shocked and praying I never come into contact with this horrible stuff.

The next shock was finding out that common old petrol, the stuff my garage (as do most garages) whiffs of when a car is being fixed, the stuff I have seen some people siphon with their mouths, contained up to 5% benzene in 1997 (although in the US and now in the EU legislation is coming up to limit its concentration to 1%). 5%, thats more than acetic acid in vinegar, and if you can get a good whiff of acteic acid with a boiling point of 117C from vinegar, then petrol should give you a good whiff of benzene, which boils at only 80C. Nobody informed me of that. It figures, authorities love to show their overt concern for safety by banning and limiting stuff. This costs them nothing, allows them to demonstrate their high standards, while exercising their real love - power. But when money gets involved, when its a case of quantifying their concern with something of value to them, then they start re-evaluating as to how much it really matters to them. Apparently they are worried about benzene less than it costs money to make petrol benzene-free.


I have found reports on the internet claiming the air at service stations is only 25ppb benzene on average. But that is ambient air, which I know from experienece does not smell of petrol. One gets a good whiff of the later only when standing over a nozzle, and especially when replacing the latter as this normally involves spilling some petrol. So one would expect garage mechanics and perhaps people who do motoring in general to show higher rates of myeloid leukemia than some country folk leading a reclusive life, and in any case, an increase in the incidence of leukemia as the use of petrol increased from the 1930's onwards. Yet I have read reports that no such trend has been observed!

I could find no figures on the internet for benzene in Australian petrol, but with all the overt concern for safety here (we are not allowed to climb ladders alone at work) I thought it would be on a par with the EU. In any case I decided to do my own investigation (with such a sensitive issue I dont know how much one could believe published data anyway).

500ml of petrol were fractionally distilled. The most voluminous fraction was that boiling 40-60C (pentanes etc) 57%. The 60-75C fraction was 8%, while the benzene containing fraction 75-100C, which includes heptane, was only 12ml (2.4%). So far so good. However the 60-75C fraction should also include benzene, since the lower boiling point alkanes, of which petrol mostly consists, lower the boiling point of its mixture with them.

I now looked around for a test to separate the aromatic in my two fractions from the aliphatics. Most of the books I looked at suggested nitration. The benzene ring is stable and nitrates easily in a HNO3/H2SO4 nitrating mixture without cleavage, the alkanes are unreactive, while alkenes oxidize. Although none of the books looked at this as a quantitative problem, I decided that by measuring the nitrobenzene formed, and asuming the yield quoted in nitrobenzene preparations, I should get a good idea of how much benzene was present in my original mixture. One concern was as to the separability of the alkene oxidation products from the nitrobenzene. For this I would need to know what happens when alkenes are oxidised with nitric acid. The textbooks are useless here. They mostly appear to be copies of each other. They all seem so logical and compelete when one reads them without a problem at hand, but when you have a specific problem you see the omissions. I think their authors would regard the inclusion of phenomena that are complex and they can not fully explain as interefering with the aim of their presentation, of a subject they are perfectly in control of, and/or would undermine their authority.

In any case 30 ml of the second fraction were nitrated, proceeding as in Vogel (or in the nitrobenzene thread in this forum, fortunately I did not have fuming HNO3 available, else I would have fallen into the same trap as described there). I noticed during the reaction that as well as an orange oil a green particulate substance which disolved with shaking was formed, and was sure this was oxidation products from other components of petrol. None of the textbooks make mention of it. Yet reading postings on the forum, I see this same phenomenon mentioned starting with, presumably, real benzene. As I said before I see this as a really is a advantage of this forum over so-called professional publications. People here do not mind mentioning things they do not fully understand.

At the separation stage three layers were formed. The upper layer, 17ml, was lighter than water and consists presumably of the alkane/unreacted portion. The middle layer, 13ml was orange, oily, and heavier than water. This must contain the nitrobenzene, and due to its orange rather than yellow colour I would have thought it to contain a good portion of unknown alkene oxidation products, however reading the nitrobenzene thread, I see that a similar result can be obtained with relatively pure benzene. The green stuff was contained in the densest acid layer, and became more apparent when water was added to it.

I washed the nitro layer with water and carbonate several times, and it became opaque orange-brown and emuslified. Disposing of washings down the sink I caught a very definite, pleasant, medicinal-almond odour, I could not distinguish it from benzaldehyde. Reading the same on the forum, I am now confident I have nitrobenzene in good yield.

Now as to the benzene content in Australian winter petrol. Going by the unreacted fraction, if we assume the rest was benzene, we have 13ml out of 30ml in the second fraction. I.e. 43%, while 2nd and 3rd fractions make 10% of petrol. All in all petrol contains 4.3% benzene. Going by the amount of nitrobenzene formed assuming its essentially pure, we have 13ml nitrobenzene, which equates to 11 ml pure benzene initially present. I.e. 36% of the fraction, or 3.6% in petrol. However if we take into account the 85% yield reported in textbooks for the nitration reaction, this becomes 4.3% again!

Overall a very bad picture for Australia, on the fronts of both safety and hypocricy.

P.S. The final step would be to distil the dryed nitrobenzene, but Im hesitant to do this. The books say it forms explosive hazards whe heated with many other organic substances, while the distilations performed in textbooks would not contain even trace amount of some of the substances present in my case. Furthermore I have no definite idea of what these could be. I also noticed the formation of a yellow percipitate if the washing (containing acid) were allowed to stand overnight (even in the cold). This presumably is dinitrobenzenes, which are explosive when heated. Again this observation was mentioned in the thread, while books say nothing on this. They all talk about a vigorous nitration environment and 100C needed to form dinitro. Has anyone had any experience distilling impure nitrobenzene?

[Edited on 18-7-2006 by len]

Edited title. Chemoleo.

[Edited on 18-7-2006 by chemoleo]

There is a huge thread here called "Benzene synthesis." It will likely answer all your questions. Use the search function to find it.

I used a large glass test tube, and yes the molten NaOH ruins it after a few runs (cracks). Others have gone to metal containers at a semi-industrial scale.

nitorarenes

About 7 years ago I played around with 2 or 3 nitroarenes. I believe I made ortho & para nitrotoluenes and chloronitrobenzene. Just the sweet odors and the literature convinced me of there ominously poisonous nature so I have not messed with them since. I may in the future but...they are scary I found a quite satisfactor way of drying the compounds though. I dissolved them in chloroform rapidly stirred the solution with anhydrous calcium chloride. An orange cloudy solution became crystal clear orange or yellow.

As far as distilling nitrobenzene, I assume it can be done safely but you will have to sit down near it and monitor the temperature closely. Be careful!

[Edited on 7/20/2006 by chloric1]

Thanks chloric, thats helpful. Ill follow your procedure with chloroform and report on the results.

As far as sitting next to it closely and monitoring the temperature, I think you might be trying to get rid of me.

I agree about closely monitoring the temperature but I am not ready to die in the name of establishing that Aussie petrol is full of benzene. In fact devising a way to keep track of the distillation while being a good distance away from it is what is occupying me now.

I have meanwhile found a few worrying reactions in the literature were alkenes (and even some especially reactive alkanes) actually add an NO2
under the influence of nitric acid, rather than being simply oxidised, as my earlier sources have indicated. Aliphatic nitro compunds are explosive. I do not know their solubility in water though. Maybe they have been washed away (the washings were all dark red). Many of the oxidation products should certainly have been washed away as they would have an OH group, the NO2 Im not sure about.

Its interesting to see UK petrol is only 1.5% benzene. We are a technological backwater here unfortunately. We try to cast a good image of ourselves but frequently dont have the technology to follow up on it with actions. I am beginning to agree with unionised that the danger of benzene is overrated. It should not be S7. The human statistics I have mentioned, and what he said do not support the carcinigenicity it is purported to have.

P.S. I have not established the 4.3% benzene in petrol with absolute certainty yet. I believe that would come with distillation of my product and comparison of its boiling and melting points.

And yes, thanks Magpie I did not know how to search this site properly at first and so missed the large benzene synth page. I got one page of results on benzene and thought that was all. I am not entirely to blame though, there were no page numbers at the bottom of the page to indicate more responses to my search were available.

[Edited on 21-7-2006 by len]

[Edited on 21-7-2006 by len]

The esters of nitronic acids can be quite explosive when distilled to dryness, but aliphatic nitro compounds are merely flammable It will take a column of very many theoretical plates to separate benzene from gasoline, a paper on exactly how it's done might be around somewhere

Thanks not_important, you obviously know quite a lot about it, pretty impressive to comunicate with someone like that.

Meanwhile I have done the reaction with KMnO4, so can answer some of my own questions, and the degree of unsaturation in petroleum ether (I used the low bp fraction as I had none of the benzene fraction left).

I shook 10 ml of the 40-60bp fraction in the cold with dropwise addition of KMnO4 (1.4gms in 30 ml H2O acidified with 0.6ml H2SO4). It reacts with unsaturateds in petroleum ether in the cold but requires vigorous shaking to mix the layers, a brown percipitate of MnO2 was formed, which surprised me, I though in acidic solution Mn went all the way to Mn2+. Just the fumes of petroleum ether are enough to turn, the tip of an inserted pipette with KMnO4 solution brown. Its a bit hard to tell the end point of the titration, because a fine layer of MnO2 percipitate sticks to the walls of the flask. But when the liquid splashed against fresh walls by the shaking no longer showed a purple tinge, I knew all KMnO4 reacted and added more.

I was able to add 1.2 gms KMnO4 this way, which (assuming all alkenes react) gives the degree of unsaturation of the petroleum fraction as 14%. If the same is true of the benzene fraction, this is quite abit higher than I hoped . More concerned now about an explosion during distilation.

I hope everyone wishing to try this realises that to distill petrol one does no use a bunsen burner. A heating mantle and quickfit flask/condenser are required.

[Edited on 22-7-2006 by len]

[Edited on 22-7-2006 by len]

Sorry, you are right not_important I did not do the stoichiometry. 1mol H+ per 1mol MnO4- is not enough to give Mn2+. However I think my degree of unsaturation is still right (at least it cant be less than the already fairly high figure) since the MnO4- was used up, and I assumed it all went to MnO2. The Mn couldnt reduce to a higher oxidation state than MnO2.

I think you are right, with the amount of alkenes there I have to repeat my experiment. Before that though, Ill distill the ArNO2 (with suitable preliminaries) since I cant thrown it away after all the work Ive done. It'll take some time, come to think of it, since I dont want to expose the indoors to toxic (carcinogenic?) nitrobenzene fumes and an explosion hazard, and I can't distill it outside lest the neighbours think this is on direct assignment from bin Laden.

Your proposal to analyse the oxidation products sound interesting - but to separate the individual products Ill need and MS and GC wont I. Iff there are N unsaturateds in petrol, then after oxidation I should get something like 3N products right?

I can just see the petrol distillation on a good fireplace. You wont need many logs to keep warm that night! Mind you some of the things I read on this site made me think that someone might just try something like that. And then there will be trouble. I wish our winters were like Canadian summers. Weve had a few -5C at night lately!

[Edited on 23-7-2006 by len]