Dear all,

I'm working with a ELSD detector and a Diol column using a gradient elution.
Solvent A: CHCl3/MeOH/NH4OH (80/19/1)
Solvent B: CHCl3/MeOH/NH4OH (60/39/1)
I start with 100% A after 5min I have 70% A and after 12min 30%A.
From 18min I run with 100%B.

MY PROBLEM: I have an unknown peak (probably from the gradient system) which starts at 7min and finishes at 12min.

Has somebody an explanation and/or an idea to prevent or to reduce this peak???

Many thanks in advance for your help

Carol

Unlike a UV detector, an evaporative light-scattering detector (ELSD) is rather resistant to 'ghost' peaks or gradient upsets. Typically, response in the ELSD is strongly related (though not linearly) to the mass of the substances being eluted from the column. There are no refractive index problems, or difficulties with elution of small quantities of high-absorbance components.

In my experience, when you get noise on an ELSD, it looks like noise (very ragged) and is due to droplets of incompletely evaporated HPLC mobile phase passing the laser. Usually changing the conditions (MP flow, gas flow, tube temp.) will eliminate the noise (though such changes can also compromise detection of your analytes).

With gradient elution into an ELSD, noise can occur during one part of the chromatogram and not during others. Such noise can be highly consistent run-to-run, or quite inconsistent. In either case, it indicates that a change in ELSD condiditons is in order.

If your "peak" is a clean-looking peak, then it probably represents the elution of some real material.

You say it's "probably" from the gradient system. How sure are you of this? Does this peak occur in a blank run? In every one of several blank runs? I have often seen (by UV detection) evidence of some contaminent in the mobile phase accumulating on the column during isocratic pumping and then eluting during a gradient. In such cases, the first blank injection, after prolonged isocratic pumping, showed a much larger unknown peak than the subsequent blank injections.

Other questions:
What is the pH of your MP? Is it within the acceptable range for your diol column?

Do you get the peak when you run the gradient without the column?

5 min peak??? what is the peak height?

Thanks for your answers

First to Anonymous:
No! I don't get the "unknown peak" when I run the gradient without the column.

When I'm running (monitoring the baseline) with 100%A and after 4-5min
I change immediately to 100% B and I observe this peak (with a big intensity).
But with this run, the peak is narrower (just 1min) than for the normal gradient system.
And if I change again to 100% A and after 3-4 min again to 100% B, the peak occurs again.

In a blank run (and even after several blank runs) the peak is always present.

So maybe this peak comes from some contaminent or It comes the material from the diol column.

Thanks for your help

Carol

P.S. the pH from my MP is 10,4 but I'm not sure.
Question: How can I measure the pH in an organic solvent?

To john B

Well the peak increases slowly during 4 min and then in 1min decreases.
The max. height is 300mV. This is comparable with some other peaks.

If some of you are interested I can send per email a chromatogram in pdf.

Carol

Carol,

You're right. I wasn't thinking. You can't measure pH in an organic solvent. However, I'm wondering whether all that ammonia might be harming your column. I have no experience with mobile phases like the one you're using, so can't say. Is your diol column silica-based?

You never said whether the peak was "ragged" or smooth. That makes a difference in interpreting what's going on.

I'm having trouble rationalizing the symptoms. You seem to be saying that any time you change between the two mobile phases, you get this strange peak. Since any way you look at it, one of those mobile phases must be stronger than the other with respect to your column and any real analyte, this symptom does not make sense if we hypothesize the accumulation of some material from the mobile phase on the column.

In fact, it doesn't make much sense for any hypothesis except noise, and that would show a ragged peak (in my experience, anyway). However, if it were noise, it should occur regardless of the presence of the column, unless the column has some secondary effect, like changing the flow rate.

Let me create a working hypothesis: Your "peak" is due to noise (and is probably not smooth, but ragged). The noise is due to incomplete evaporation of the mobile phase at some range of compositions intermediate to your MP A and MP B. This peak will occur regardless of the kind of column you install. It will occur with no column in the system (providing you provide sufficient back-pressure to your pump that the pump functions properly). This noise peak can be eliminated by changing the ELSD conditions: an increase in temperature will probably work; an increase in nebulizer flow may work; a decrease in mobile phase flow to the detector will probably work (but changes your HPLC conditions).

I realize that hypotheses contradicts one or two observations you've made, but I can't come up with anything else right now. How badly does it contradict your observations?

By the way, what ARE your conditions? MP flow rate? drift tube temp? nebulizer gas and flow rate?

Has this always been a problem? I have never run a gradient with ELSD (I am sure it is possible).... Since you don't see any peaks when running isocratic perhaps the change in solvents (refractive index) is causing this response.

Try developing a single solvent mixture that you can run isocratic.

Carol,

If you want, you may email me the chromatogram at the address hyperlinked to my name. I suggest you delete any proprietary information, since for all I know our companies may be competitors. Please answer the questions above as I might not be able to help without those answers.

Carol,
My guess is the high pH is destroying your column releasing the coating and silica to make the peak.

Measuring pH in a non-aqueous system is tricky. Do you have a non-aqueous system? Have you added ammonia gas or ammonia dissolved in water? For chemical reactions a little water tends to make the solution behave like a water solution. For example, 1 N sodium propoxide in propyl alcolcol is much more basic than 1 N sodium hydroxide, but one percent water in the propyl alcohol behaves more like sodium hydroxide. You can use a bridge to measure the pH but pH indicators might do your job easier.
Jim

Dear all,

I have solved my problem.
I changed only the supplier of the chloroform and the unknown peak is disappeared.
Thank you very much for your help!
Carol

Does the new supply of chloroform use a different stabilizer (or none)?