opinion: noisy baseline

[rochelle]

Hi there!

I just want to ask you guys if you encountered analyzing samples even if the signal is higher than your usual signal in FID/PID detectors or in any detectors. What will be the cause and effect in our sample or for the detectors or instruments if we continue analyzing sample even if the baseline is not good. As a chromatographer, will you allowed this? or is it ok to run samples even the signal to the detector is not stable? I need your intelligent opinion guys.

Thanks

[zokitano]

Dear Rochelle,

All chromatographers (or most of them) tend to obtain stabilized baseline before running the samples. Noisy baseline can affect chromatograms and also the integration step and quantification.
If you have noisy baseline or it wanders so much you'll have to obtain first stable baseline before you run your analyses. Factors that can influence the baseline are various: column, detectors etc. Also when using temperature programme during analyses, with the raise of the temperature, one often observes increasing of the baseline. If you run your analyses at temperatures close or above the maximum column temperature, the noisy baseline can occur due to the column bleed. Always run your analyses at temperatures below the maximum column temperature to prolong the column lifetime.
Contamination of the detector can also produce variation in the baseline.

Best wishes

[chromatographer1]

zokitano summed it up well.

A noisy baseline indicates that there is a hardware problem, column, pneumatics, or electronic.

It is an indication that the instrument is NOT working as it should be.

It is also an indication that the results are not to be trusted.

Would you want to trust data collected from an instrument that OBVIOUSLY was not working correctly?

Sigh........ Sometimes we have to do maintenance. I don't like it either, but it comes with the territory.

best wishes,

Rod

[HW Mueller]

One usually runs a standard regularly to check on deterioration.

[Bruce Hamilton]

Sadly, my world isn't ideal, and I often have to compromise, and
my response depends on what you are defining as noise.

I'm assuming the instrument is behaving normally, and the noise isn't due to some easily-identifiable and correctable instrumental misbehaviour.

If you started out with noisy system, and it stays the same, the following techniques should keep you safe... The important aspect is to define the point when the noise adversely affects your assay, and either modify the assay or fix the problem, perhaps by improved sample preparation or instrument maintenance.

If the signal is higher, and all the conditions are constant, but your samples are well above the noise ( lowest standard/reported concentration must be at least 10X noise ), then, provided your integration can handle the noise, and you use regular standards thoroughout the sequence, you can get away with noisy baselines. Ideally, your method should include an internal standard so you can ascertain if the noise is interfering with quantitation.

If it is just a gradual increase in background because of junk from your samples that is eluting, then you definiately should use an internal standard and have regular standards in the sequence. I also like to run some of the early samples at the end to ensure the result is the same.
Ideally, changing the method to eliminate the junk, or the GC programme to elute the junk is preferable. You need to ensure that your integration is coping with the junk.

Please keep having fun,

Bruce Hamilton