LC pump problem

[atram]

Hello everybody,

We have an Agilent 1200 with quaternary pump, and these days we have some problems with the chromatogram appareance. We think there is something wrong with the pump and I hope your experience can help us to solve the problem.

The chromatogram obtained is:

Appart of to obtein peaks with this appareance (notice that baseline is OK), the retention time has doubled.

Could it be the outlet valve? the inlet valve? anything with the seals?
Any idea?

thank you in advance.

[project44]

Hi there,

First of all it is important to give some more information about the analysis you run.

1. Is it an isocratic or gradient method?
2. What kind of column do you use (C18, Cyano, polar embedded e.t.c)
3. What is the nature of the analytes (acidic, basic..pKa)?
4. Flow rate, mobile phase composition
5. Is the system pressure stable? What is the usual back pressure?

[atram]

Sorry, I was not very explicit before…

We use Size exclusion Chromatography to analyse glycosaminoglycans.
It is an isocratic method where mobile phase is 100% phosphate buffer pH=6.
The column used is a BioSep-SEC-3000 of 300 mm x 7.8 mm, and with a flow rate of 1 m/min, the system presents a pressure of 800 psi.

Here is a zoom of our peak of interest or analyt. Maybe now you can see better the steps that are present among the peak and which I am worry about.



Thanks

[Consumer Products Guy]

Are you bypassing the multi-channel valve? If not, remember to purge all four channels even if you're only using one.

[HPLCCONSULT]

Some ideas:

(1) If your flow rate is half of what it was before the problem (your retention times are doubled), then you probably have one broken pump piston (piston #1).

(2) If you method uses solvent from more than one bottle, then you may have nothing but air running into the MCGV from one channel.

(3) One solenoid in the MCGV could be stuck closed.

Blast out all solvent channels with pure solvent (ACN or MeOH would work well) and make sure the system is 100% primed.
Open up the prime purge valve and check flow from each channel, one at a time, with a graduated cylinder or 10 ml volumetric flask to determine actual flow rate.

Update 9/16/10: I went back and looked at your info again. You stated before that your retention times have doubled. This is a very important comment (see my #1 note above). This has to mean that you are pumping half the flow rate you were before. Measure it and compare it to what the instrument read-out is.
You also now mention you are running in isocratic mode so that should rule out the MCGV as a problem (only one channel being used 100% of the time). Perhaps your problem is a simple one. Are you degassing your mobile phase ? Cavitation can cause the saw tooth profile and wrong flow rate. Your Active Inlet valve may be suspect, but first you must supply the pump with degassed mobile phase before you trouble shoot that part.

[HW Mueller]

Why would this not show up in the baseline? I also don´t know what to think of the mirror image type relation of the teeth on the up and down sides of the peak.

[unmgvar]

what detector are you using?

is it a UV, RI, PDA?

If you think the pump is the problem,
did you acquire the pump pressure chrom. as well?

[benW]

Hello,

This looks like an electronics or data handling problem, especially with the mirror image effect of the "teeth" on the up and down slope (good observation HW Mueller).

Is the run actually taking as long as the chromatogram indicates?

It shows that the data channel is duplicating the chromatogram, then setting back and then recording the data "again". Therefore, it is not evident in the flat baseline section but very obvious on the sloping peak sides and explains the doubling of the retention time.

With an isocratic method it would be unusual to see such variation in detector response with pump/flow problems, and these would not be mirrored on the peak as here.

It will be interesting to see what the problem is, but I am guessing it has to do with the detector/timing relationship. Check where the timing is controlled... actually I am thinking that the data file is created on the instrument control PC, and the timing is controlled there...?

Try a different PC...

Let us know how you get on,
Cheers,
Ben

[atram]

Thank you for all your ideas and suggestions.

We are using an isocratic method (only one channel) and the detector is a PDA.

As HPLCCONSULT says, I also think that having a retention time doubled (half flow) is related with a pump problem.

Yesterday I removed the pump head assembly, cleaned the plungers with MeOH and changed the seals. Any deformation was present at these seals, but there were exchanged for last time far ago.
After reinstalling the pump head assembly, I check the system and recorded news chromatograms. And by now, the system works correctly: retention time and flow are correct and teeth have disappeared. Could it be a leak on the seals?

If the problem reappears again in a few days, I will tell you.

Thanks!

[HW Mueller]

Hopefully you are rid of the problem, it still leaves me puzzled, though.

[Gerhard Kratz]

Yes, that was a problem of your pump head. Could be that one of the valves did not work properly or you had a leackage. The teeths you saw are typically for pulsation problem. Maintenance of your pump head solved the problem. If you use your pump on a daily base I would recommend to clean the pump head once a month.

[HPLCCONSULT]

"atram": Glad to hear that you are making some progress. As I mentioned before, the problem is clearly solvent delivery path related, not data station related. The key to good troubleshooting is to start with the basics and not ignore obvious clues (retention time doubling). Now that you have replaced the pump head seals and cleaned the pistons, please make sure you also replace the 10 micron PTFE frit (inside the Prime Purge Valve) with a new one. We recommend that you replace the PTFE frit every single month. It protects the entire flow path from fine debris and also bits of pump seal particles which will appear from normal use. The frit collects this debris as it leaves the pump head and stops it from getting into your injector or column. The frit is VERY inexpensive so be sure to have plenty on hand for monthly maintenance. *Do not confuse these with an aftermarket inline filter.

If you have the later model Active Inlet Valve (AIV) on your pump, then please check the cartridge for contamination and replace if needed. The older style did not use the replaceable cartridge. A small obstruction in the AIV can cause cavitation and pulsation problems. This is very rare, but when a system is not maintained or supplied with unfiltered buffers it can cause problems from salt crystals or debris getting stuck in it. A good flush with some filtered, acidified, aqueous solution can sometimes clean it out at high flow rates.

[benW]

Hi all,
Good to hear that you have found the cause and apologies if I sent you down the wrong avenue.

A very interesting symptom and I agree that a real doubling in retention times has to be pump fluidics related – hence the question:
“Is the run actually taking as long as the chromatogram indicates?â€

[HPLCCONSULT]

"benw": You are trying to make the whole thing more complicated than it actually is. This spiked pattern seen is a characteristic of an electronically controlled valve (AIV) and the dual variable stroke/volume piston set in the pump head. The pump is cavitating a bit when it does not get any liquid on one stroke (sent to the second piston) of the piston. The computer can vary the stroke volume to compensate for the gas and you get a hiccup effect. The computer does its best job to compensate for the problem (one piston is not pumping solvent under pressure in this case). I have seen it many times over the past two decades using and servicing various manufacturer's pumps.

A sticking mechanical pump valve usually results in zero flow and no pressure (If you have ever used most early Waters, Hitachi, Beckman, PE, Gilson or Spectra-Physics pumps than you know that you spend most of your time keeping the pumps primed and chasing air bubbles out of the check valves!). The Agilent pump heads found in the 1050, 1100 and 1200 (all are basically 99.5% same design) use an electronically controlled inlet valve (AIV) which really helps keep the system fully pressurized/primed and therefore it is super rare to ever see the one and only mechanical outlet valve cause problems on an Agilent (HP) system. The twenty plus year old design is very robust and a pleasure to use. I have used all of the major brands of pumps over the years and I find my favorite HPLC pumps are the ones that I do not have to always spend time working on. *As you can tell, I appreciate them from an engineering aspect too.

[carls]

HPLCCONSULT: Sounds like you have significant experience so I'd like to pick your brain. I believe what is puzzling to most of us is the lack of baseline noise. In my experience, a malfunctioning check valve gives high noise throughout the chromatogram. In this case the noise is only evident when a peak is eluting. I was leaning towards an electronic problem but if repairing the pump fixed the problem then this is a new one.

Since you have a grip on this problem could you explain how a bad check valve would only shown noise when a peak is eluting, i.e. noise on baseline much lower (non-existent) than that observed on a peak.

I have another idea. Is it possible the reference wavelength is set to a wavelength range where the analyte also absorbs and thus exaggerates the pump noise when a peak elutes?