By Anonymous on Tuesday, November 25, 2003 - 08:21 am:

Hi folks,

I have 6 metformin impurities needed to be separated. I have tried several columns(e.g. Partisil/SCX) and different mobile phases without satisfied results. Some impurities come too early and not separated each other.

Would you mind give us some advice? Thanks.

John

-------------------------------------------------------------------------------------------------------
By zelechonok on Tuesday, November 25, 2003 - 12:25 pm:

Try our mixed mode technology. Primesep 200 or Primesep 100 columns separate compounds based on their pKa and hydrophobicity. You have a good chance to resolve all peaks by changing the amount of ionic and/or organic modifier in your mobile phase. HPLC methods for amidines with similar properties and structure to metformin were developed recently based on this mixed mode technology. Another similar example is separation of guanidine http://allsep.com/makeChr.php?chr=Chr_019
The power of separation of impurities in amino sugar can be seen in the following method: http://allsep.com/makeCmp.php?cmp=Cmp_003
All compounds in the mixture are basic and very polar.

-------------------------------------------------------------------------------------------------------
By Einar Pontén - SeQuant AB on Tuesday, November 25, 2003 - 03:00 pm:

Hi John,
You should try HILIC!

metformin = 1,1-Dimethylbiguanide [657-24-9]

-------------------------------------------------------------------------------------------------------
By Andreas Neumaier on Wednesday, November 26, 2003 - 08:59 am:

Third possible way is a fluorinated phase (like Perfluorphenyl or Fluofix) and mobile phase with water/acn and around 0.1% to 0.5% TFA.

There are some disadvantages when using TFA. You have to purge the column with acn to make sure all TFA is washed from the column. When running a gradient baseline may be not as smooth as when using sulfuric acid for example.

TFA is building strong complexes with polar and basic compounds and these complexes are retained on a fluorinated phase (even TFA is retained).

Good luck and let us know, which system you decide to use.
Andreas

-------------------------------------------------------------------------------------------------------
By amrutesh on Tuesday, May 18, 2004 - 08:37 am:

I would like to know the mobile phase for metformin and how to detect metformin from plasma or serum samples

-------------------------------------------------------------------------------------------------------
By Uwe Neue on Tuesday, May 18, 2004 - 03:06 pm:

Here is a suggestion for a complete method:
1. sample preparation: acidify the plasma or serum sample. Centrifuge to remove a precipitate. Add a suitable internal standard. Load this sample onto an Oasis MCX (mixed mode cation exchange) sample preparation device (you may use 96 well plates, if you need to deal with multiple samples). Wash the adsorbed sample with acidified water with 10% methanol, for example water/ 10% MeOH with 1% acetic acid. Now wash the adsorbed sample with methanol. Finally elute the sample with methanol to which 1% ammonia has been added.
2. HPLC: I recommend to develop a HILIC method on a silica column, such as the Atlantis Silica HILIC column. The reason for this recommendation is that you can inject the sample that you have extracted from the SPE directly onto the HILIC column without having to worry about evaporation and reconstitution. In order to develop a HILIC method, you should run a gradient from 95% acetonitrile to 50% acetonitrile with adding an acid such as formic acid or an ammonium salt. The details of the method depend on your detector. What detector are you using?

If you are using a UV based method, it is possible that we need to improve the SPE washing procedure. There are ways to do that, but the suggested approach is already fairly decent.

-------------------------------------------------------------------------------------------------------
By Anonymous on Tuesday, August 17, 2004 - 08:03 am:

I would like to know method of metformin using Partisil/SCX colum only.only one impurity cyanoguanidine needs to be resolved.
pls do reply