Bubbles

[HPLCSlave]

Hi Everyone!

I would like to ask two rather simple questions.
How can I eliminate gas from my eluent? There are some small tiny bubbles all over the tubing that leads to the pump...Apparently they form while the pump absorbs the eluent from the bottle.
I can try helium but I also have problem with it because there are small bubbles adsorbed on the bottle walls that eventually enter the tubes for the eluent...
My second question is whether such a problem could be responsible for the following bizzare chromatogram, recorded on a BECKMAN System GOLD HPLC...


Thanks for any reply!

[mbulentcakar]

Hi,
You may put your eluent in an ultrasonic bath for ten to thirty minutes, preferably applying vacuum.
The procedure is called degassing, so I suggest you to search the form by phrases such as, degas, degasser, degassing.
Good luck,
Bulent

[zokitano]

When forming of bubbles occures during the absorption of eluent from the bottle I'd rather check the solvent filter placed at the end of the solvent tube immersed in the eluent in the bottle. Occlusion of the filter often produces bubbles in the tube when the pump sucks out mobile phase from the solvent bottle. I suggest you to place the filter into a glass with ethanol or methanol and sonicate it for about 15 minutes. Then put the filter back on the top of the solvent tube, rinsed before with water with HPLC quality.
I hope that will solve most of your problems

Best regards

[Mark Tracy]

The other alternative is an inline vacuum degasser; most modern HPLC systems use this technique. Rheodyne makes a nice one, and there are others. If at all possible, get disposable filters for the mobile phase; your labor is more valuable than a piece of frit.

[HW Mueller]

Also, if your mobile phase is composed of more than a pure solvent the composition may change under vacuum.

[HPLCSlave]

Thank you all very much for the precious advices!
I hope that these would solve the problem!

Nevertheless I wanted to post a strange chromatogram, but I think nobody have seen it...
I tried to upload it with the img parameter...but I haven't succeeded.
Anyway it can be found under:

http://www.funpic.de/fotoalbum/foto,151860,0.htm

Provided I'm not directed there because of my IP

Thank you once again for your kind help!

[sgkhayade]

Dear friend if you are using an isocratic system by mixing two or more solvents in one reservior , you ensure the complete mixing of the solvents otherwise incomplete mixing will produce bubbles and inconsistent retention of the peaks.

[JM]

Looking at your chromatogram , it does not seems to be cos of air bubble.

This kind of uniform ripples may be due to flow problem or detector . please check if you get same kind of ripples with no flow?? if yes detector is the culprit, if no than your pump flow problem.

JM

[HW Mueller]

I have seen plenty of saw like baselines, due to bubbles, but nothing like this. There are no piston pumps in this lab so I wonder if that could be bubbles in the detector + periodic surging (starving) of the pump? (Doesn´t sound so hot, used to work with piston pumps long ago, can´t remember periodic pumping problems). Strange wiggle at the apex... I would really be surprised if that was electronic. Is this really a baseline without any manual modification of any kind? The HPLC was just running?

[Mark Tracy]

Now that I can see your "chromatogram" I must say that it is strange indeed. Definitely not bubbles; the period is too regular. It appears that the period of the oscillation is around 7.5 minutes. Is there any process with that period? Look closely at your system and its surroundings because this probably will be something you didn't expect.

[Bruce Hamilton]

That is novel modern chromatographic art.

It appears to me like a temperature control issue or a detector with low signal strength. Ensure that your instrument isn't near an air-conditioning duct, airflow from another instrument, or fumehood that is cycling. Is this problem specific to a method, or all methods?. If the former, I'd be first looking at the method in close detail, such as ensuring the detector is seeing plenty of energy.

What happens when you halve the pump flow rate, does the frequency change?. If it stays the same, you probably can exclude the pump.
If you have a column oven, turn it off ( don't just take the column out, turn off the heating control of the oven ).

If the pattern is still there, and you are using a UV detector, change the detector wavelength, increasing it by about 30 nm to a more transparent region of your solvent. Ensure your lamp still produces plenty of energy, and hasn't exceeded design life.

If you are using a Refrective Index or other type of detector that is heated, turn off the temperature control. You will see gradual baseline drift, but does the sine wave disappear?.

What are the analytical conditions, especially the mobile phase and detector type ( and wavelength )?.

Good luck,

Bruce Hamilton

[HPLCSlave]

First of all this type of chromatogram is occuring only accidentally, apparently with no connection with any change in parameters...So I can hardly try to reproduce the conditions to see whether it occurs again...
It's a simple baseline...not manually modified at all.
The UV lamp was recently replaced...

I can hardly imagine a cyclic process that takes 7,5 minutes...
At a flow rate of 0,8 ml/min the pump sucks every ~15 seconds...

The mobile phase is pH3 water with sulfuric acid...The detector is a BECKMAN 168 Detector System Gold...
This works with a Photo Diode Array and when this problem occurs, it occurs for all wavelenghts...

This chromatogram is the most regular of a series of "failed" chromatograms that occur in the following situations:

- changing the solvent (water to methanol, water pH3 to water pH4)
- leaving the system for a longer time without runs

The system generally recovers after a long running time, but I've always bubbled Helium in between...so I'm not quite sure if it would recover without it...

Thank you once again for the great help...It really gives hope!

[danko]

Hi HPLCslave,

It looks to me that you have an air bubble cycling in the flow-cell. You can easily find out if it is the case and just as easily remove it.

First to the confirmation: Just use one of the eluents e.g.â€

[JM]

To get rid of bubble formation/air intake in detector cell connect restrictor coil in the detector outlet.

JM

[Bruce Hamilton]

Given the new information, I'd agree that it's probably an issue with undissolved gases.

The easiest solution would be the option suggested by Mark Tracy, an inline vacuuum degasser. If you don't want to do that, then I'd suggest making a 1:1 mix of methanol:water, degassing that, and using using it for few minutes as a transition between the water and methanol.

As the problem occurs rarely and on standing, I'd also suggest that you ocassionally flush the detector with a solvent that makes the surfaces less hydrophobic - as well as reducing the dissolved gas content of the mobile phase and putting some backpressure onto the detector cell as suggested above. 15 minutes of pumping IPA usually works for me, but you may need to perform a nitric acid clean, depending on the mobile phases and samples.

FYI, I've had HPLC control loops giving cyclic baseline perturbations with cycle times of between 5 - 10 minutes, the most recent being a failing D2 lamp on an Agilent 1100 detector used at a low wavelength.

Others have included gradient mixing valves, older pump units, detector thermal controllers, and column ovens. They are usually only visible on a UV if the detector is seiing low energy, but can be more obvious on RI.

As you say, the pump stroke is of a much faster frequency, but mixing controllers and valves seem to produce cyclic frequencies of their own that affect baselines, without noticeably affecting pressure.

Please keep having fun.

Bruce Hamilton