Expanding the Length of Carbon Chains

[Theophrastus]

Out of interest, are there any simple reagents or methods that can be applied in order lengthen carbon chains (given the proper functional groups, and so forth)? I say this, in the sense of, "Magically adding carbons." For example:

 

[ce] CH3CH2CH2OH ->[delta{{X}}] CH3CH2CH2CH2OH [/ce]

 

or, in a similar fashion, is there a means to bring this process to work in reverse, subtracting carbons?

 

[ce] CH3CH2CH2CH2CH2OH ->[delta{{X}}] CH3CH2CH2CH2OH [/ce]

[hermanntrude]

yes

 

but not magically.

[Theophrastus]

yes

 

but not magically.

 

Aiya! I didn't mean magic literally!

 

However you did say yes: So whaat methods' date=' and reagents are available circumstantially to do so, in most cases?

 

ps: I used the word magically, thinking on the following thread...

 

not what I meant. I meant if you turned all the hydroxy groups into carboxylic acid groups, what would you have? it's basically citric acid with one hydroxy group removed.

 

That would involve magically adding carbons

 

Good times...

[John Cuthber]

If I remember rightly the classic path is

R-OH

to RBr

to RMgBr

to RCH2OH

I will let you work out the reagents.

[Theophrastus]

Thanks John! So for my initial case, the reaction would be:

 

[ce] CH3CH2CH2OH + HBr -> CH3CH2CH2Br + H2O [/ce]

 

[ce] CH3CH2CH2Br + Mg -> CH3CH2CH2MgBr [/ce]

 

[ce] CH3CH2CH2MgBr + H2C=O -> CH3CH2CH2-CH2-OMgBr [/ce]

 

[ce] CH3CH2CH2-CH2-OMgBr + HBr -> CH3CH2CH2CH2OH + MgBr2 [/ce]

 

Though, I'm sorry my examples limited the output; I'm not purely focusing on alcohols, and am open to any methods. It's purely out of interest.

Edited August 25, 2009 by Theophrastus

[John Cuthber]

You could start with an aldehyde or acid, convert to the alcohol, add a carbon and then convert the alcohol back to an aldehyde or acid. It' a bit of a long way round though.

 

I have a feeling there's another trick with the acid (or acid chloride) and diazomethane.

Anyone else got a better memory than me?

[Theophrastus]

You could start with an aldehyde or acid' date=' convert to the alcohol, add a carbon and then convert the alcohol back to an aldehyde or acid. It' a bit of a long way round though.

[/quote']

 

Yeah, it's pretty long, but in terms of do-ability it wouldn't be too bad, as I have a suitable oxidiser, but sadly no suitably strong reducing agent, which means I would have to start with an alcohol, and if necessary, convert it to either an aldehyde or acid in the end.

 

On a cooler note, your idea of using grignard reagents seems to have inspired me, and I've also thought of something. Wouldn't it also be possible to use ethylene oxide, to extend a chain by two? Practically speaking, my only problem is synthesis. I was thinking of using ethylene glycol as a starting point (which ironically, I made a batch of, only a day back), but I don't really know where to go from there. Here is the proposed reaction mechanism: (For simplicity's sake, I'll use methanol as my starting point)

 

[ce] CH3OH + HCl -> CH3Cl + H2O [/ce]

 

[ce] CH3Cl + Mg -> CH3MgCl [/ce]

 

[ce] CH3MgCl + CH2CH2O -> CH3CH2CH2OMgCl [/ce]

 

[ce] CH3CH2CH2OMgCl + HCl -> CH3CH2CH2OH + MgCl2 [/ce]

 

ps: ethylene oxide is cyclic, thus the deviation and oddity of [ce] CH2CH2O [/ce] which may not seem sound at first.

Edited August 26, 2009 by Theophrastus

[John Cuthber]

That would sort of work, but are you sure you want to play with ethylene oxide? It isn't very nice stuff. Also methyl chloride is a gas which isn't convenient and the Grignard reaction is rather slow with chlorides- I didn't sugest the bromide by accident.

 

For the most part, this sequence was only used for the sort of synthesis that proves the structure of something. For that sort of work it doesn't matter that the yield sucks.

Nobody would use it to make propanol (which is, I think) cheaper than ethylene oxide etc.

[hermanntrude]

there are several methods for adding carbon-carbon bonds to a chain, but yes, the above is the most commonly used.

[Theophrastus]

That would sort of work, but are you sure you want to play with ethylene oxide?

 

What's wrong with ethylene oxide? Though in contrast, I've found the method you were discussing, only discover diazomethane is rather explosive Though more importantly (), I'm really not sure on how the hell, given my limited resources, how I can synthesize it, which is a bit of a downer on the whole idea .

 

http://en.wikipedia.org/wiki/Diazomethane

 

Also methyl chloride is a gas which isn't convenient

 

hence the following:

 

(For simplicity's sake' date=' I'll use methanol as my starting point)

[/quote']

 

the Grignard reaction is rather slow with chlorides- I didn't sugest the bromide by accident.

 

Thanks for the note. I'll take take it iodides wouldn't be too bad (As I lack bromine salts and whatnot).

 

Nobody would use it to make propanol (which is' date=' I think) cheaper than ethylene oxide etc.[/quote']

 

See second address.

Edited August 27, 2009 by Theophrastus

[John Cuthber]

"What's wrong with ethylene oxide?"

It's toxic, flammable, explosive and probably carcinogenic. (A bit like diazomethane).

[Theophrastus]

"What's wrong with ethylene oxide?"

It's toxic' date=' flammable, explosive and probably carcinogenic. (A bit like diazomethane).[/quote']

 

That's... interesting. I suppose grignard's my best bet then.

[dttom]

Actually I have thought of another approach though it is quite a long one, maybe it is not a good approach:

begins with ethanol

1) ethylene

2) 1,2-dibromoethane

3) acetylene

4) sodium acetylide

5) alkyl acetylide

6) sodium 1-alkyne

7) (alkyne group)-CONa

8) (alkyne group)-COH

ends by hydrogenation of the triple bond

In this way, one could add from just one carbon (snap from step 4 to 7 directly by adding formaldehyde) to a definite number (n) of carbon (using CH3(CH2)n-2Br in step 5).

[varnis]

If the above is too long for you, you could consider then CN route, adding sodium cyanide, though you would want to be careful of HCN gas. You can easily yield a Carboxylic Acid of one carbon longer than the starting alcohol in this way (Via a Halogen substituent)

Edited September 28, 2009 by varnis
Clarification