contamination of HPLC system

[KC]

OK DR I have one for you. The new thread about contaminated TEA made me think of it. We have a contamination in one of our HPLC systems. The contamination points to the aqueous mobile phase because the peak is much larger if the system is allowed to run at the lower concentration of organic for awhile before making an injection. Whatever it is collects at the head of the column and elutes as the gradient ramps up. The component is really quite small but when looking at trace levels that little peak is irritating and in some analyses cannot be ignored. In most cases I can avoid it by putting in a couple test injections prior to initiating the sequence as the method itself during a sequence is not at a low enough organic long enough to collect therefore I don't see it in definitive samples (or only I can see it because it is so small that it does not integrate and only appears as a slight rise in the baseline). I believe it is in the system because changing mobile phases (even the ACN) does not eliminate it and I don't believe its the column because I have seen it with different columns on the same system. I also use only B&J bottled HPLC grade water for mobile phases and do not experience this phenomenon on the 5 other systems so I don't think it is the water. The question is, what can I run through the system that will get rid of this. One other analyst tried a solution with TFA/water/ACN with no luck.
PS- I think its from some GD buffer somebody was running in it, he-he.

[KC]

sorry that my post above sounds directed to one individual, was rude of me but was actaully carrying over from a similiar topic concerning contamination. Certainly any advice on the topic from any individual that can shed light on the problem is welcome.

[Rafael Chust]

Have you tried the typical 6N nitric acid chock treatment to it?

[KC]

No I have not, sounds pretty severe. Like I said, I can get around this problem most of the time and in higher level analyses it is no problem at all. Will the nitric harm peek and/or metal tubing as I believe this system still has a combination of both? I suppose you rinse the system with aqueous on both channels, increase the flow to ~2 mL/min and let it run for a few hours?
By the way, the system is an Agilent 1100 and this component begins to move at about 50-60% ACN, anybody have any idea what it might be?

[Rafael Chust]

Well, the stainless steel will be passivated, which is a good feature, but might interfere with PEEK.

I usually clean HPLC systems with 30 to 50 ml HNO3 (making "fake" injections to clean the injector), followed by HPLC water until pH goes up - with a pH stripe at the waste.

I never had problems doing that with any HPLC - Agilent and Waters included - so I keep recommeding it.

[DR]

Apologies for my curtness, but No, it IS the water. I too have tried bottled LC grade water from various vendors and what I get from a really good TOC polishing unit is usually as good, if not better than what comes from bottles.

Try the Empore extraction disk routine on some water before you make "A" phase with it and you will almost certainly see an improvement. If you see no improvement, replace the lines that run from your bottles to your pumps as this is the only other place I can think of where extractables could leach out of the system and end up collecting on your column head.

Passivation isn't a bad idea, but I have seldom needed to do it. Water quality issues have been responsible for many more baseline problems than contamination from other sources has.