HPLC RI drifing positive baseline

[ressorgimento]

helo, I'm using an Agilent HPLC, a quite new one, only 0.5 year old.
When i'm doing my analysis i got a constant drifting baseline. It's a straight line that's going up. I already purged the RI and the pump several times, and refreshed the solvent cyclohexane. Still my baseline is going up. I checked the RI temperature, and the average temp. is 39.985°C +/- 0.005 (small peaks till 40.0 and 39.975), the signal of the reference cell is constant, the other cell is constantly going down.. I use a flow of 0.4ml/min.

The column i use i made myself and exist out of a freshly calcinated zeolite (so all dirt is gone), wich i packed under N2 atmosphere. The poresize of the zeolite is choosen so that the solvent can't enter the pores. I'm already flushing for 4 days, but still no change.

[Bruce Hamilton]

One obvious solution would be to replace your column with a restrictor or another column, and see if the baseline keeps going up - if it does then it's a problem with the detector, possibly a solvent leak.

Also, does the slope change with the flow rate?. If you take some of the cyclohexane that has passed through the column, and reuse that, does the baseline show any perturbation?. I'm wondering if your zeolite is removing or releasing a small molecule, such as water - but that is a remote possibility. Ensure that your system doesn't have any leaks or slow thermal effects ( such as warming of the detector lines ) from changing ambient conditions.

I'd suspect that the detector doesn't like cyclohexane, as most of the probably causes should have either reached an equilibrium or, at least, produced some change in the slope during the day.

Good luck, and please keep having fun,

Bruce Hamilton

[ressorgimento]

I continued the experiments, and after a long time, the baseline starts to equilibrate, but it takes quite some times, it can be up to several days.
Strange thing is for example, yesterdaynight i got a rather stable baseline, +1000mV in a half an hour. I lowered the flow till 0.1ml/min overnight and thought the baseline would be complete stable today. It was at 0.1 ml/min, but when I raised the flow the baseline went back up. Now I am back again flushing for an entire day, and it's again getting better, but still not good enough for good experiments.

You said you suspect the detector doesn't like cyclohexane, I have a question.. Last week, in an effort to stabilize the baseline I switched to Isopropanol. Because they say it is a better solvent to clean your column. I got very fast a stable baseline unbelievable. Because my experiments utilize cyclohexane, I changed again the solvent, and there it was again, from the moment it was 10 to 1 cyclohexane, it started drifting again.

So what do you mean with my detector doesn't like cyclohexane?



Thanks for the help.

[Noser222]

That your baseline gets better with Isopropyl alcohol makes me think of a moisture issue, as Bruce said.
Maybe you could try "half-saturating" your cyclohexane.

[juddc]

In my experience, changing the flow rate on an RI will in all probability disrupt the baseline.

Also, I've seen my RI equilibrate more quickly with certain solvents than with others. When working with my RI, acetonitrile based mobile phases seem to equilibrate more rapidly than methanol based MP's, for instance. I'm not sure why this is the case, but it appears to be.

Other things to help stabilize an RI baseline:

Make sure that your columns are at the same temperature as your RI cell (or a few degrees warmer) and at at least 5 deg C above ambient temperature.

If it's compatible with your mobile phase, use a nut and ferrule made of PEEK rather than stainless steel to connect the tubing from the column to the detector. If possible, use PEEK for the entire length of tubing. PEEK has much lower thermal conductivity than steel and it does make a difference, I've seen it. If you must use stainless steel, insulate everything from the column to the detector. Some foam and lab tape will do fine, despite looking a bit less than high tech on the front of your shiny new machine.

If you haven't, recirculate your mobile phase. When starting my system up, I'll usually allow 10 or so system volumes to go to waste while purging the cell, then run the line out of the purge port back into the MP reservoir. When I'm running 3 or 4 300mm x7.8mm GPC columns in series at 0.8 mL / min, the initial purge can take all day, then I recirculate all night, and I'm ready to run first thing the next day (about 30 minutes after flipping the purge valve). This way I don't have to touch the flow rate and wait for the RI to calm down. When recirculating, make sure everything is clean and that you shield your mobile phase well from dust accumulation.

Hope this helps!

[Bruce Hamilton]

Cyclohexane has low viscosity ( 0.5 v 2.4 for IPA ), higher volatility, and significantly different RI ( 1.43 v 1.38 for IPA, and 1.33 for H2O ). The low viscosity and different RI will test some detectors that are designed mainly for aqueous systems.

I'd suspect that it's not just a thermal equilibrium, but possibly a small amount of the cyclohexane is evaporating/boiling somewhere and causing a cooling effect. Alternatively, the cyclohexane is warming somewhere in the detector, and you are seeing a change of RI with temperature, but both reference and sample should compensate.

You haven't said what happens to the slope over time when you change the flow rate. I'd want to ensure there is a reasonable backpressure on the refractometer to ensure the flows are constant and, if there is a leak, you'll find it sooner .

With regard to thermal issues, I would agree about insulation, except that cyclohexane isn't that different to IPA with regard to density ( 0.778 v 0.785 for IPA ) or specific heat ( 2.2 v 2.6 for IPA ), the two major parameters for heat transfer.

I'd certainly be looking to ensure that no extraneous heat is appearing from nearby instrument fans or sunlight, but would focus on ensuring that the same amount of cyclohexane is leaving as is entering, and that you aren't seeing the effects of misbalanced internal warming.

Please keep having fun,

Bruce Hamilton

[HW Mueller]

Especially with solvents (nonpolar) which dissolve air easily I have seen these endlessly rising baselines even with UV detectors (via RF changes, scattering....). In those cases evakuated mobile phases had been used, no other de-gassing. They picked up air slowly during runs.