Noisy Baseline

[mdyo]

Hello everybody

I am trying to analyse environmental samples at very low level using HPLC. I faced a problem of noisy base line. The noise is a long-term one, as it is shown in the chromatogram in the following link:

http://us.f13.yahoofs.com/bc/441aeded_2 ... EB9h83AWqd[/img]

The first 100 min of the chromatogram generated at the normal running conditions for my analysis which are: C18 column (250*4.6mm), 70 % ACN/Water mixed by the low-pressure-mixing pump. Flow Rate 1.5 ml/min, DAD set to monitor at 230nm.

105-160 min The column was disconnected as I suspect it to be the cause of the noise. However, the noise continued.

170-200 min I used premixed 70% ACN/Water and put the 2 inlet channels A and B into the same bottle to check if the problem was due to mixing but the problem continued.

> 200 min I stopped the pump and let the DAD on but I got a continuously raising straight line in an angle of about 45 degree.

The large spikes between these periods were due to stopping the pump.


I rally don’t know what may cause such a problem and how I may solve it!

Any advice or comment is very appreciated.

Thank you all in advance.

[Mark Tracy]

Would you please fix the typo in the link so we can see your chromatogram?

[mdyo]

Thank you Mark and sorry for this technical problem!
I moved the chromatogram to the following link:

http://uk.pg.photos.yahoo.com/ph/chroma ... pg&.src=ph


I am hoping that it works.

[Mark Tracy]

Yahoo has an advert that obscures part of the image, but from what I can see, we can eliminate a few things. Since there is no obvious pattern, we can eliminate things like a mechanical pump problem or a bubble in the pump. The time scale is too large to be a particle or bubble in the detector cell. Sometimes the air conditioning system can cause this kind of thing, but usually it is even longer in time scale and smoother looking. An old lamp in the detector will have unstable output, that can cause this sort of thing. I have also seen cases of cross-talk between the vacuum degasser and the detector; hard to believe but true.

The drift with the flow stopped is quite usual behavior for most detectors.

[ccyting]

We had similar situation before. You just have to remove the column and follow the operation mannual to clean up the whole HPLC system from injector to detector using water, 6 N nitric acid, water, methanol, and water. For an isocratic separation, we always use pre-mixed mobile phase. I don't think the problem is malfunctioning of the instrument, just the system is not clean. When you do the cleaning, please be patient, that is my advice. It is a tedious procedure but hopefully will be able to solve your problem. Good luck.

[mdyo]

Thank you all.

An old lamp in the detector will have unstable output, that can cause this sort of thing

The lamp is now 5 years old with fairly heavy work. Do you think this is a normal lifetime for the lamp.

I don't think the problem is malfunctioning of the instrument, just the system is not clean.

I already tried to clean the flowcell with Nitric Acid but no improvement noticed. The attached chromatogram was after the cleaning

[Bryan Evans]

5 year old lamp may definetely need to be replaced!
We usually were able to get 6 months/3000 hrs or so
out of our lamps.

You're using diode array detector - do you also have a
reference wavelength for your detector?

[HW Mueller]

My guess is that ccyting is right.

[Noser222]

In reference to Mark's comment about the air conditioning, we had a new lab built to be climate controlled and under positive pressure. It turned out that the air conditioning was causing the long-term noise, much worse on the RI detectors than the UV. Our pattern was similar to yours, so I don't know if it is always "smooth." The cycling time would depend on when the air conditioning comes on and shuts off. Do you normally have run times as long as this one? Our problem was hardly noticeable for a more typical run time, say 15 minutes.

[mdyo]

Thanks for all of you.

You're using diode array detector - do you also have a
reference wavelength for your detector?

Could you please explain what you mean by "Reference Wavelength"?

My guess is that ccyting is right.

I think I have cleaned the flowcell well- Do you think some dirts in the other parts of the system may cause such a long term noise?

Do you normally have run times as long as this one?

The longest runtime I use is 80 min. However, in shorter runtimes (10-20 min) the proplem appears as a drift (up in some runs and down in others).

[Bryan Evans]

The VWD has a beam splitter - the reference wavelength is than subtracted from the sample wavelength to account for power surges.
(i.e. power surge from AC, ect.)

You are using a DAD (but are using it for single wavelength detection). Most software packages will allow you to set a reference wavelength for your DAD. Choose a wavelength that you know is not absorbed by any of your samples (i.e. something like 280 nm). Any fluctuations in the voltage from the power supply will be accounted for by subtracting this from the sample wavelength.

Ex: Instrument Response = Response at 230 nm - Response at 280 nm

I would rinse the flow cell again - as this is also a likely culprit, as is
the air conditioning.

[HW Mueller]

mdyo, more so than the flow cell.

[mdyo]

Hello everyone,

Thank you all for your valuable suggestions and comments and sorry for the delay in reporting back.

I replaced the old detector lamp with a new one and tried the cleaning process which has been thankfully suggested above and this is what I have got at the end of the day:


http://uk.pg.photos.yahoo.com/ph/chroma ... pg&.src=ph


This chromatogram ( I hope it is clear enough) was generated at W.L=228 nm with the column disconnected and the pump delivering 100% Acetonitrile through a single inlet line at a flow rate of 0.5 ml/min.

The noise seems to be increased in magnitude. I am really not sure what is happening but I am sure that it will be solved with your help guys.

Thank you in advance.

[mdyo]

Hi everybody

I am still trying to spot the problem in my instrument. I tried here to run the system at different flow rates to check if the pump contributes in the problem and here is what I got in the following links:

http://uk.pg.photos.yahoo.com/ph/chroma ... pg&.src=ph


http://uk.pg.photos.yahoo.com/ph/chroma ... pg&.src=ph


Both chromatograms were done with the column disconnected and the detector at 228nm.

The first one (Noisy_baseline 3) was generated using a flow rate of 1.0 ml/min of 100% water, and the other one at 0.1 ml/min.


Do you think the pump may contribute in such a problem?

Your suggestions are very appreciated!

Thank you in advance.

[Mark Tracy]

Can you identify anything in your laboratory or HPLC system with a 20 minute cycle? The "Noisy_baseline_4" has a very distinct and regular cycle. This could be the air conditioning, the vacuum degasser, etc.

By the way, your pump probably is not stable at low flowrates and no backpressure. The software in the pump depends on the pressure to stabilize the flow. A few feet of 0.12 mm capillary will provide enough backpressure for your pump to operate in a stable mode.