Help with UV "noise" issues in gradient

[peptideguy]

Hi guys

Long time reader, first time poster. I hope that you can help me with a problem that I have been having for several days now. We do a lot of peptide mapping via LC-UV (Agilent 1100 and 1200 systems). We use water/acetonitrile gradients with 0.1% TFA in water and 0.085% TFA in acetonitrile. The problem we are having with several of our system is that we get baseline noise during the early portion of the gradient, and then this noise goes away as the gradient progresses. I do not have a chromatogram to show you - I will try to attach one next week when I am at work.

For example, a typical gradient may go from 3% acetontrile to 60% over 45min. We will see a baseline oscillation, which may range up to 10mAU in amplitude, during the beginning of the run. As the gradient progresses, this noise will get smaller and smaller. Eventually, at ~20-30% acetonitrile, the noise will be negligible. However, the size the oscillations in baseline during the early portion of the run can obscure our ability to analyze the smaller, early-eluting peptides. By the way, we see this oscilaltion at 210 and 214nm, but it goes down as we increase the wavelength. Using refernce wavelengths does not seem to help.

We see this on different HPLC systems, and if we switch the water an acetontrile positions, we still see the same effect. The pump pressure seems fine - there is no significant ripple in pressure, and the system spass leak and detector tests.

I have a feeling that this has to do with build-up of TFA int eh systems - has anyone any similar experiences? Is there a method for cleaning residual TFA out of a system? This is my best hypothesis becuase we went to a system which had never been used for TFA, and within 24hrs we saw the same absorbance oscillations occurring.

Please help. I will try to post a chromatogram on Monday if I do not have an answer by then.

Thank you.

[HPLCCONSULT]

This sounds mechanical as you say it only occurs at the start of the gradient, but here are several questions and comments:

BTW: All runs should be done with the Reference Wavelength feature turned 'OFF'. "TFA" should not clog up the system.

(1) What is the sample wavelength bandwidth setting ? Is it too wide for your wavelength selection ?

(2) Have you tried the same method with a new column ?

(3) TFA does show up in the low UV area. This is normal. Oscillations are not normal, but can be caused by temperature effects (*sometimes from the detector) or a pump issue. If you run an isocratic run at the starting conditions, what is the period or cycle of the oscillations ? This may provide the source.

(4) A fresh batch of 0.1% TFA should not cause any issues. Is your TFA fresh ? Is the water fresh and filtered?

(5) Have you tried the same gradient w/o the TFA ?

(6) The default liquid compressibility setting for the 1100/1200-Series Pump is 100, but you are starting your gradient with water (which has a compressibility of only 46). As the gradient progresses, the amount of ACN is increased and your problem goes away. ACN has a compressibility of 96 which is much closer to the default value of 100. If this is the problem, then I would have expected the ripple to be higher at the start of the gradient than at the end. A long shot, but you cold change the compressibility to 50 and run the method again to see if there is a change. For more info on compressibility, see this web page [http://www.hplctools.com/Tip_110_HPLC_Solvent_Compressibility_Values.htm].

(7) Are your solvents (water and ACN) fully DEGASSED and flushed through the system before you start the analysis ?

(8) What is your injection volume and sample concentration ?

Answers to these questions might narrow down or rule out certain potential problems.

[peptideguy]

Thank you. See my responses below:

This sounds mechanical as you say it only occurs at the start of the gradient, but here are several questions and comments:

BTW: All runs should be done with the Reference Wavelength feature turned 'OFF'. "TFA" should not clog up the system.

(1) What is the sample wavelength bandwidth setting ? Is it too wide for your wavelength selection ?
I do not recall - I will check on Monday. What is an appropriate bandwidth for 210nm?

(2) Have you tried the same method with a new column ?
Yes. This is an on-going problem on several systems and has been occurring for months now.

(3) TFA does show up in the low UV area. This is normal. Oscillations are not normal, but can be caused by temperature effects (*sometimes from the detector) or a pump issue. If you run an isocratic run at the starting conditions, what is the period or cycle of the oscillations ? This may provide the source.

The method has an initial isocratic hold. In addition, when we are equilibrating and running isocratic at 3% B, we see this effect. The effect goes away as the % of aceonitrile increases.

(4) A fresh batch of 0.1% TFA should not cause any issues. Is your TFA fresh ? Is the water fresh and filtered?

The problem has occurrred with multiple batches of TFA over several months. We use DI water - we could try bottled HPLC water, I guess.

(5) Have you tried the same gradient w/o the TFA ?

Good question - No - we have not tried this - we will on Monday.

(6) The default liquid compressibility setting for the 1100/1200-Series Pump is 100, but you are starting your gradient with water (which has a compressibility of only 46). As the gradient progresses, the amount of ACN is increased and your problem goes away. ACN has a compressibility of 96 which is much closer to the default value of 100. If this is the problem, then I would have expected the ripple to be higher at the start of the gradient than at the end. A long shot, but you cold change the compressibility to 50 and run the method again to see if there is a change. For more info on compressibility, see this web page [http://www.hplctools.com/Tip_110_HPLC_Solvent_Compressibility_Values.htm].

We have absolutely no ripple pressure problems - the pumps appear to be performing beautifully. The only problem is the UV absorbance.

(7) Are your solvents (water and ACN) fully DEGASSED and flushed through the system before you start the analysis ?

We use the in-line degasser that came with the system. Maybe we could try a new degasser? Our 1200 is only ~1yr old, and it has this problem.

(8) What is your injection volume and sample concentration ?

Injection volume and concentration do not appear to matter - we see this with blank 0 vol injection runs, and when the system is simply running isocratically and no samples are going.

Answers to these questions might narrow down or rule out certain potential problems.

Thank you for your response. I will try to attach some chromatogram on Monday. One other observation - this problem is independent of flow cell. We have replaced the flow cell with a brand-new, never used flow cell and the problem is still there.

[HPLCCONSULT]

Thank you for the extra info. Sounds like a very irritating problem.

*210nm. Bandwidth of 5 to 10 nm max should be fine.

Your degassing module should be fine. I would not suspect it.

As I mentioned before, a picture with labeled 'X' and 'Y' scales of the noise period/cycle of the oscillations shown would really help define what this might be. If you can, post this when you are ready.

Please tell me exactly which detector module you have on your 1200-Series (G____). Thermal issues can be a problem. Esp with pure water at very low wavelengths. That is the other reason I would like to see the actual noise over-time. Also, what sampling rate is it set to (aka: response time in Hz for Agilent systems)?

[Andy Alpert]

These symptoms can also be caused by a mixing problem. You're mixing two solvents of different viscosity. This sometimes results in oscillations during the early part of the gradient that you're seeing them in. You can rule this in or out by putting an additional mixer in place. Here's an easy way to do that: Take 50 cm of metal capillary tubing and wrap it tightly in coils around a pipe that's 1/2 inch in diameter (slightly more than 1 cm). Leave the ends straight so you can make normal connections. Slide the coiled tubing off the pipe and install it between the point where the solvents are blended and the injector valve. This mixer will decrease baseline oscillation about 50% at a flow rate of about 1 ml/min. If that's what you see, then just get a more effective mixer for your system.

[peptideguy]

Here is some data to illustrate my problem. The top is the gradient profile, then 2nd image is the pressure profile, the 3rd image is the UV absorbance, and the 4th is a close-up of the UV absorbance.




Please let me know if you can help with this very annoying problem.

Thank you

[tom jupille]

Blow up the pressure profile and see if it has fluctuations that "sync" with the UV. If it does, you have a flow rate problem. If it doesn't, it's a mixing problem as suggested by Andy Alpert.

If I were betting, I'd lay money on mixing.

[HPLCCONSULT]

Just checked in and see that you posted a few pictures of the "problem". OK, there are a few possibilities.

First, more questions.
(1) Which Agilent Pump are you using (Binary or Quat) ? If Binary pump, is the standard mixer in-line ?
(2) Which Agilent detector (G1314, G1315...) and which flow cell (10mm, 6mm, 3mm...) are you using ? Temperature issue ?

The cycling under the isocratic portion does hint at either a mixing problem or switching valve issue (AIV). I will temporarily rule out a valve issue as you indicated this problem appears on ALL HPLC instruments used, not just one. Pure ACN is easy to pump at low wavelengths. Pure water is not.

Next, what is the stroke setting on your pump ? Is it set to default "AUTO" or an actual value ? You might want to change it to a lower value such as 50ul and see if the rate of cycling changes. It should change the cycle rate.

What is the compressibility set to on the pump ? Is it set to 100 (Default) or another number. You could adjust it to a lower number close to the value of water (used at your initial condition) to see if it improves things. Perhaps a value such as 50.

DID you try it w/o the TFA yet ? TFA shows up well on the low UV and it could be magnifying the noise, esp if "dirty". Run a blank to check. Monitor the noise and cycle rate.

The more details you provide the better.

[peptideguy]

FIXED!!

thanks for your responses, but the answer was so simple I curse myself for not thinking of it earlier. We had been running in minimum volume mode, bypassing the dampener and small mixing column that comes with the system. As soon I reset the plumbing tot he original config, the oscillations went away. However, due to the increase in dead volume from the dampener and mixing column, my retention times are shifted outward. But, all in all, I consider that an aceptable change, considering how badly the oscillations were interfering with chromatography.

So, thanks again.