Hello,
We are observing a trend on the ELSD Waters2424 detector. Signal increases steadily while no modifications have been made to the system. We work with a gradient (mix: MeOH/water 60/40 --> 100% MeOH --> 60/40) for quantitation of polysorbate 80.
Detector parameters are:
Drift tube temperature: 92°C
Nebulizer heater-cooler set point: 59%
Output full-scale: 2000 mV.
Because of increasing signal we had to continuously reduce the gain. Starting from 800 we are at 65 now.
We did notice that the signal was higher (picks sharper) each time we were changing the column (Waters C-1). But we have got no explanation so far for the observed trends.
Peers, what would be your opinion?
Best from Switzerland,
Anton
Condensation in the drift tube often causes this. You did not say what your flow rate or column dimensions were. If appropriate, you may want to dial down the flow rate, but only after optimizing the ELSD's sample peak S/N ratio using a standard and your existing gradient to make sure you are not biased towards too much signal gain.
There are many other reasons for a steadily increasing signal over time. Poor quality or obstructed exhaust connection (to the ELSD) will also cause excessive noise and changes in baseline over time. Insufficient heat of vaporization (esp with water) causes condensation inside the drift tube resulting in cooling of the system. Usually you will "see" random spikes in the baseline when this happens (from drops of liquid).
Dear Bill, thank you so much for your answer! Especially for technical details which are valuable and highly appreciated!
Flow rate is 0.8 mL/min. Indeed, it could work, but, it is not validated.
We still think that lot-to-lot variability of column is the most probable root cause of the observed phenomenon.
I will let you know how we will fix it. If you are interested of course.
Very best regards,
Anton
You still have not provided us with any HPLC method conditions so can not add anything except that why do you think that column variation is the problem? Are you using a very unusual column type? That is a very unusual and unlikely reason on its own. Again, without detailed information about what your actual HPLC conditions are, no one would be able to provide any specific suggestions. Please note that as an ELSD (or CAD) detector is used, sample is deposited and baked onto the drift tube over time. This changes the output over time and often increases noise too.
You are right:
the column type is quite common: Spherisorb model C1, 4.6x100mm, Waters. No. PSS832612
HPLC and pump parameters:
Eluent A H2O milli-Q
Eluent B MeOH pure
Rinsing of pistons and the injector MeOH/H2O 20 : 80
Volume injected 20 microL
Mode Binary Gradient
Flow 0.80 mL/min
Concentration A 40%
Concentration B 60 %
Run time 20 min
Degasifier on
And we do think that the column can influence because we observe systematically a difference in signal intensity between column in the beginning (freshly installed) - high and its lowering all along the column life time.
Actually, that column would not be one I would choose for this method. In fact, a high quality C18 column (witha higher carbon loading than your C1 with ~ 2%) would be the recommendation for Tween. Column selection is just as important as the selection of the mobile phase. For best results, you want to get both items right as early as possible.
Other comments: Your run time may be too short and your organic not strong enough, esp with that column.No chromatograms were provided so just guessing.
Another area of concern. You may be experiencing a problem of ELSD contamination (a build up of sample inside the ELSD which is very common and not something that the manufacturers like to discuss). If your detector is contaminated, then depending on the running method conditions, both noise and baseline drift tend to occur at ever increasing rates overs time. Signal degradation is the result. You may also be overloading your column with sample. This is also rather common when you are a novice and esp when you have trouble detecting a sample as the common reaction is to keep injecting more (or increasing the concentration). This of course leads to both column fouling and eventually, fouling of the ELSD detector too. Exhaust line obstruction or fouling is another possible problem, as are nebulizer clogging. Lastly, signal will always get worse if the ELSD temp and gas rates are not optimized for the method and sample (this is of course where you start).
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication