Page 1 of 1

Mobile phase problem

Posted: Mon Jan 02, 2006 11:20 pm
by vanjas
Hi!!
I have some problem with mobile phase when analyzing bile acids.
The mobile phase are composed of 0.2M KH2PO4 and Acetonitrile with gradient mode.
The dilent for sample solution is methanol.
I guess this composition make some kinds of precipitations in HPLC system, so, the pump pressure become zero and the pump is choked.
Is there something wrong with my mobile phase composition.
The thing is that the all three make precipitations when I dissolve together.
Any help would be thankful for me...

Posted: Tue Jan 03, 2006 1:06 am
by Uwe Neue
Your salt solution is fairly high...

Before running a gradient with a high concentration of a salt solution, you should always make an experiment off-line at the highest concentration of organic solvent so see if there is a precipitation. Otherwise, the salt precipitates inside the instrument, which is not pleasant.

You will probably need to flush the HPLC with water. How to do this specifically will depend on your instrument.

Posted: Tue Jan 03, 2006 8:40 am
by syx
Such high concentration of buffer solution will produce salt precipitation when it meets acetonitrile in the mixer, especially in gradient elution. It is recommended to reduce the concentration to 0.05M or less.

Posted: Tue Jan 03, 2006 2:54 pm
by amaryl
A buffer of molarity 10-50 mM is usually employed.

Posted: Wed Jan 04, 2006 10:47 am
by Rafael Chust
Phosphat buffers is one of those annoying things that are very nasty but you have to learn to live with...

Your system is undoubtely full of salt.

Before you do anything, take the column out and try to pump 100% water.

If you can pump do not find any leak under the pistons, your gaskets and check valves are ok.

Flush about 20 ml of water to waste and, in the meantime, inject water in the injector (you might have salts there too!). Connect again the column and detector and flush with water until pressure stabilizes.

Start again and be careful with phosphate buffers in the future! Some methods are developed so fast just to get peaks that do not take into account column lifetime and system downtime!