Splitted peaks - what do I need to do?

[dragosb]

Hi,

I am running a series of samples and I see that I get splitted peaks but not for all samples.
I get splitted peaks for two samples then for the third it is Ok and then again I get splitted peaks for a couple of samples and then again a good one...

What could be the problem is such a case?

I changed the guard collumn before launching the samples.
Should I perform a reversed cleaning procedure?

Thanks

[vahur]

Is there any difference in the sample compositions or pH?

[kyleb]

If your sample compositon is stronger than your mobile phase, that could be your problem.

What is your mobile phase composed of vs the sample.

Kyle

[tom jupille]

The first thing to look at: do you have more than one peak in your sample? If you do, do *all* the peaks show the same kind of splitting? If they do, then you are almost certainly looking at a column inlet flow profile anomaly. From the fact that the problem is intermittent, I would tend to suspect an occasional air bubble (which would suggest checking for leaks, inadequate degassing, etc.).

If only *some* peaks are split and others look OK, then you are almost certainly looking at a chemical problem. As Kyle suggested, this could be a "strong diluent effect". Other matrix-related problems are also possible.

These kinds of things are common enough that I actually did a "chalk talk" on the subject:
viewtopic.php?f=31&t=18426

[juddc]

A few questions:

Is the issue sample specific?

If you repeatedly inject the same sample, will it always behave either well or badly?

If you have an intermittent situation upon repeated injection of the same sample, then it's likely your system. If one sample's always good and the next always bad, then it's likely a function of the sample, as others have suggested.

Is there any change in behavior if you vary your injection volume?

@ Tom - Chalk? What is this stuff of which you speak?

CJ

[tom jupille]

Chalk? What is this stuff of which you speak?

Among other uses, a bit of powdered chalk on my fingers gives me a better grip on my slide rule.

[juddc]

OT I know, but slide rules are cool and bring back memories.

My dad was an archetypal mechanical engineer, rarely seen without his slide rule and mechanical pencil, he always wore a jacket and tie to work regardless of what he was doing (which often involved wrenching on military aircraft), and he always drove something interesting. He cut quite a figure in his Jaguar Mk2...

[dragosb]

Hello,

Thanks for the answers.
I use an acetic acid 96% (0.58 ml/L)/ NaOH (10 mg/L) mobile pahse ph 3.5.
The derivatization solution is:
20 mg OPA, 1.8 mL MeOH, 20 µL 3MPA, 0.2 mL Borate buffer pH 9.5 and htere I add 20 µL sample + 30 µL internal standard

As for the splitting it appears in almost all the peaks that I have

[Gerhard Kratz]

Now all Peaks are splitted, than the packing bed is damaged! Maybe, depending on the injection volume with such a highly Basic sample solution you got partially hydrolysis and so called channels in the packing bed. If you run your column in reversed flow you will get Minimum 1 run without splitted Peaks, but maybe not more.
Sorry, ist time to order a new column.

[sandy1989]

Are u using the autosampler or manual injection?
and if it is autosampler try to extend the run time and check !
And if it is manual try to change ur mobile phase composition!
and i think the way ur query was either ur column is already damaged or about to!
try to change the column and repeat the separation?
and how old ur column is? and how good ur sample stability is?

[M. Farooq Wahab]

Hi,

I am running a series of samples and I see that I get splitted peaks but not for all samples.
I get splitted peaks for two samples then for the third it is Ok and then again I get splitted peaks for a couple of samples and then again a good one...

What could be the problem is such a case?

I changed the guard collumn before launching the samples.
Should I perform a reversed cleaning procedure?

Thanks

Among other reasons suggested by others (mainly blocked frits) - check if you are not overloading the column. I am wondering because if the column were damaged, all samples would have shown splitting. But as you say, some samples show splitting and the rest don't. If the sample concentration is high, you may see split peaks.

Regards.