Maybe someone can help me out with this vexing problem. My system is an HP 1100 series hplc (w/ degasser, dad, etc), linked with a Finnigan LCQ. My problem is with the LC. I am running a gradient of 60:40 H2O:ACN 2mL/min held 2 min, then ramped up to 100 ACN over 25 minutes on a C18 column. Monitoring at 254nm. When I run a blank of ACN the baseline is very erratic and at 16 minutes there is a very strange, broad "lump" in the chromatogram that increases until 26 minutes, where it goes back down to just the erratic baseline. Analyzing the region with the mass spectrum shows no material coming off (at least between 50-400 m/z). It doesn't show up in the PDA scan either.

I have had this problem with another NP column as well. I have tried re-equilibrating the column with ACN, but it is still there. All solvents are HPLC grade. I doubt as to whether it is a mixing or air bubble problem, but it might be. If I run a sample, the lump is still there, though not as noticable since it is in the baseline. It seems to be just a baseline problem.

If anyone has any insight into this problem, I would appreciate it. Thanks.

It sounds like an artifact of the gradient. The sudden rush of acetonitrile has a different absorbance and the sytem eventually equilbrates and goes back down. Many gradients we use will ramp up and back down during the run and a ramp like that would be seen. We have validated the method to allow baseline subtraction to give a smooth look and it aids somwhat in the chromatography.

Thanks anon. I did a few more runs with blanks and it looks like it probably is from the gradient. Running just ACN doesn't give the lump, but the baseline is still not as flat as I would like. Would increasing the run time prevent that sudden rush of acetonitrile?

Since when is an increase of 2.4% per minute in a gradient called sudden rush. ACN at 254 nm has very little (if any) UV absorbance so the previously proposed explination really does not make too much sense. It is much more probable that you have contaminated water or maybe you are stripping phase material off of your column.

I've thought about both of those explanations too. I've eliminated the contaminated water possibility by putting new (Aldrich HPLC grd.) water online. I could be stripping the column, but are these conditions really that detrimental to the column? It's about 2 years old, but has not been used for many runs.

I just tried running this gradient on my instrument, also an HP1100 with DAD. I have a Waters Pico-Tag column (C18) installed.

I set signal A to 254 +/-4 nm with a 350 +/-50 nm reference, and signal B to 254 +/-4 nm with no reference signal. Signal A is close to flat. It's perfectly flat for 3 minuted, then starts decreasing to -0.8 mAU at 23 min, then recovering to zero prior to reequilibration at the starting composition.

Signal B shows somewhat more movement, rising to about 0.2 mAU over 3 min, then droppign to -1.2 mAU at about 21 min before recovering.

These are probably refractive index changes. In any event, they bear little resemblance to what Ross reports.

I think the evidence of the MS shouldn't be ignored. If there is material eluting, it should be detectable. (It might be worth checking the UV spectrum of the lump to see whether it supports the MS data.)

The question I have for Ross regards degassing. What sort of degassing is being used? Usually the HP1100 membrane degasser works quite well, but no doubt it could fail. The noisy baseline is suggestive of air bubbles, but without seeing the chromatogram I can't be sure. Conceivably the lump could relate to a combination of air and a change in the solvent, hence changing the properties of air bubbles.

Otherwise, contamination of the water or acetonitrile is worth investigating. What grades are being used? Have different lots been tested?

Bruce-

Thanks for checking that out. The degasser I use is the HP 1100. Perhaps I should check and make sure it is working properly. I don't receive any error messages from it, though. How can I test how well it is working?

Again, the solvents are HPLC grade, but I haven't tried different lots.

I'll try a few other things...

For what it's worth, I run an acetonitrile-water gradient to wash my columns, and I see a similar effect. It happens with every column that I have. Have you tried running the same gradient with a different type (C8, phenyl, etc.) of column?

Ross:

Sorry, offhand I don't know an easy way to check that a degasser is functioning properly. You MIGHT try running aerated water (i.e., water shaken with air) and aerated acetonitrile through two different channels of the thing and see whether bubbles develop in the effluent. These two solvents, when aerated, usually put out copious bubbles when mixed.

Acetone has a much greater absorbance in the UV (cut-off around 340 nm) than does acetonitrile (cut-off around 200 nm), so would be expected to give a considerable rise during wash-out.

However, this suggests a question: In your original posting did you mean that the baseline ONLY recovers AFTER you return to the original conditions? (I may not have understood you correctly.) If so, then your acetonitrile may be contaminated with a UV-absorbing compound. In this case, using a fresh lot of HPLC-grade acetonitrile could be the cure.

Bruce-

My graphs show that when I ran the gradient using an acetonitrile blank, the baseline began rising steadily at ~16 min, then suddenly came back down at ~26 minutes, so the gradient was still running up to 100 acetonitrile. At 27 minutes the gradient ran 100% acetonitrile until 35 min. It looks like the lump part is when the water and ACN are running together, but it doesn't start until the last ten minutes of the 25 minute gradient.

Could some kind of temperature fluctuation be the cause of either the lump or the erratic baseline at least? I have a column heater and keep it as close to room temperature as I can.

I sounds like a contaminant is collecting on the head of the column and is washed off when the AN concentration gets high enough. I can't explain the MS data, though. Is the baseline problem consistent regardless of how long the system is held at the initial parameters or does it get worse if these parameters are held longer? You might compare different equilibration times to see if the problem is consistent or varies depending on how long the system is held at the initial parameters.

Ross,
Okay, that does not sound like a simple contamination of the acetonitrile with a UV-absorbing chemical.

Russ may be right about something accumulating then washing off. The lack of MS evidence might simply be due to a very small amount of contaminent with a very high extinction coefficient.

Russ makes a very good point. If his hypothesis is correct and if your contamination is present in the initial mobile phase, then the lump will be bigger the longer you hold the system at the initial conditions before running the gradient. I've seen this many times. In fact, in an automated system, it is sometimes possible to tolerate lumps like this because they "go away" after the first injection or after the first few injections. (Just make sure to run blank injections before your standards and samples.)

Since you have a DAD, check the spectrum of the lump and see whether this provides any clue.

Remember that the contamination could be in ANY mobile phase ingredient. I'd suggest trying fresh reagents, or even different manufacturers. If you're using water from a lab water system, try using bottled HPLC water instead.

Temperature fluctuation is usually cyclical -- say a 5-minute cycle or so. It cause waviness of the baseline that can interfere with integration. I've never seen it cause a lump like you describe.
That said, trying to control a column oven near room temperature is usually a mistake. First, columns are generally more efficient run warm. Second, column ovens generally don't control well near RT. Check your manual to find the minimum differential from RT the oven can reliably control.

Bruce and Russ-

Thanks for the great suggestions. I haven't been able to test them out yet because I've been doing maintenance on the MS, but I will try it out and let you know.

You can test the functioning of the HP1100 degasser by using a milivoltmeter and measuring the signal from the pressure transducer. There is an aux jack on the back of the degasser and the transducer signal is available across two of the pins, details are in the degasser manual.