Big change in Retetion time of Metformin

[hplc_newbie]

Hi everyone! This was a strange problem in my analysis with Metformin. Here is the method:
Buffer: 30g NH4(H2PO4)/1L water then adjust to pH 3.0. The flow rate is 1ml/min. The RT of Cyanoguanidine = 2.0 . Melamin = 4.5 and Metformin = 10. The peak shape of Metformin always is very poor with tailing is always higher than 2.0 even with new column whereas the other two peaks is still good and sharp.
Today, I incidentally prepared the buffer with half Concentration and I got a big surprise. The RT of Melamin and Cyanoguanidine didn't change much but RT of Metformin changed a lot - it was double and peak shape was better. Normally, I found that RT of Metformin can be changed by changing pH or using IP reagent. But changing Concentration to change RT is new to me. And I am also surprised when the protocol requires a big concentration for HPLC (according to me it should be form 0.005M to 0.1M).
Why RT increased and peak shape of Metformin was better when Buffer Concentration was halved? I hope to get your answer. Thanks very much

[Vlad Orlovsky]

we have a method for metformin and melamine with good peak shapes, it just needs to be optimized for the mixture of your compounds:
http://www.sielc.com/Compound-Metformin.html
http://www.sielc.com/Compound-Melamine- ... amine.html

Regarding retention time change: metformin can interact with residual silanols by cation-exchange mechanism, changing buffer concentration and pH will cause retention time shift depending on the way you change these parameters.

[hplc_newbie]

Regarding retention time change: metformin can interact with residual silanols by cation-exchange mechanism, changing buffer concentration and pH will cause retention time shift depending on the way you change these parameters.

Thanks for your website with a good method but I am afraid that I can't use that because I must follow the approved procedure which means I can't add organic solvent to my Mobile phase.
Is the interaction between Metformin and residual silanols called secondary interaction? If so, it means if I use high concentration buffer, the secondary interaction will keep Metformin in column longer than usual (Increse RT) and because of secondary interaction my peak is broaden. But to have the second interaction, silanols must be ionized and to ionize silanols pH is from 5-7. In my experiment, pH is 3.0 so how can the secondary interaction happen? (if I am not right please tell me the pH range which can ionize Silanols).

[Vlad Orlovsky]

Silanols are not uniformed. Some impurity inclusions will make them more cidic (sometimes it is even pKa 2).