Method Development: anilines and positional isomers

[PJ8]

I am trying to develope a method for analysing 4-iodo, 2-methylaniline. It is a starting material in our process, and I need to analyse it for impurities formed during it's synthesis, the most likely of which are positional isomers around the ring, di-iodo versions or ortho-toluidine without the iodo group.

I was wondering if anyone had developed methods for anything like this before and could point me in the right direction. The literature I have read has covered a lot of bases, RP and NP using everything from bare basic silica with Na ion modification to C18's, ODS, cyclodextrin chiral columns and silica gel columns etc.

Therefore if anyone has done anything like this and can tell me what things they tried and how successful it was I'd greatly appreciate it.

Cheers, Pat Jackson.

[SIELC_Tech]

Dear Pat,

I think that you will get few advices from this board. I would like to propose that you use mixed mode approach. Your compounds are slightly basic and hydrophobic so Log P and Log D are slightly different due to different substitution positions. Mixed mode approach will allow you to use two mechanisms to separate your compounds. This approach is described in the following graph:

http://www.sielc.com/Technology_2D_Properties.html

We separated few isomers (four for people on this forum) using this approach. Here are methods for separation of isomers with basic and hydrophobic properties:

Separation of Toluidine Isomers on Primesep 100 and Primesep 200 Columns (you are using reverse phase and cation exchange mechanisms):
http://www.sielc.com/compound_214.html

Separation of Aminobiphenyls on Primesep D column (you are using reverse phase and ion-exclusion mechanisms):
http://www.sielc.com/application_088.html

If you are not sure if this approach will work for you please consider sending us a sample for method development (at no charge or obligation to you).

Regards,

Vlad

[PJ8]

Much food for thought there, many thanks!
I am awaiting one or two more potential impurities to arrive for me to spike into my main analyte before I begin method development, but thanks for supplying me a few places to start. I will update with progress as and when I can.

Cheers, Pat!

[Rob Burgess]

Posistional isomers are notouriously difficult to separate, but aren't impossible by RP-gradient LC. I would try a DryLab modelling approach to cut down the guesswork. Contact me if you need further details....

Another option I have used in the past for Fluro posistional analogues is fluro stationary phases. They may work for your iodo isomers as well?

[PJ8]

Thanks Rob. I was expecting a difficult time with this method development, the first thing I did was run a system suitability sample containing 4 positional isomers down a generic column screen we had and 3 of the six columns seperated all four!

Luna C18 with MeCN and 0.05 % v/v TFA is looking the most promising so far, which goes to show the old school simple stuff is the best thing to try first before getting overcomplicated. I have a couple more compounds to spike now thanks the trainee synthetic chemists so we'll see how things hold up.

[Noser222]

might even consider a phenyl phase

[PJ8]

Zorbax SB phenyl and Luna phenyl hexyl were included in the screen, zorbax didn't score too well and Luna Ph-hex lost one of the positional isomers consistently.

Luna C18 performed the best with Zorbax SBC8 and Zorbax bonus as back-ups. Also have the Luna C18 mimic columns Sunfire and Eclipse Plus (although we are encouraged to stick with Luna C18 by the powers that be...)

[Bryan Evans]

Hi PJ8 -

Here are some aniline and toluidine positional isomer separations using our Unison UK-C18 and Unison UK-Phenyl columns:

http://www.imtakt.com/TecInfo/TI208E.pdf
http://www.imtakt.com/TecInfo/TI087E.pdf

Hope that helps!

[PJ8]

Not sure if this will work as I haven't posted images before.
If it does this is what the seperation looks like on a Luna C18.

[zoom]

Your results look pretty good. Have you had a chance to test this method with the di-iodo byproducts?

[PJ8]

That's the next plan zoom. I have a really dirty sample from one of the suppliers which looks to have some in and they elute between 20 and 25 minutes (2 peaks above 1 % area each.) This sample unlike the others dissolves to a purple clear liquid (others are clear colourless) and has an extra absorbance in it's UV at 536.6nm. This is not due to free iodine (apparently, I didn't run that test) so di-iodo species are suspected. A chemist is working on di-iodo samples for me to use as spikes.

Since we have two suppliers are 99.5 + percent and one with di-iodo supplying at 93% we probably wont have a problem but it will be nice to confirm that di-iodo is separated by this method.