Negative Absorbance on Agilent DAD 1200 HPLC

[HaruYap]

Hi there. I have a problem with my DAD detector.

I am experiencing negative absorbances during the initial run of my analyses; my total runtime is 60min and I'm having this negative absorbances from injection until about 15-20min before the DAD pick up the positive peaks.

May I know what is the cause of the problem and how am I suppose to correct this?

I have flush the systems and I've also restarted(turn off and on) the DAD (just a try!) and I've also used the autobalance options on the DAD. Any other measures that I can take? If I am not wrong (correct me if I am), the negative absorbance on the initial analyses time might be due to the sample carry over from the previous runs that coincides with the absorbance range of the solvents itself and also the reference wavelength that has been set. If there is such a carryover, peaks will be detected but in the negative region since it corresponds with the solvent peaks as well as from the carryovers, rite?

Anyway, thanks for the attention and I hope that anyone can help me out to verify this situation. Thanks.

[HW Mueller]

Could it be that you are just injecting a sample solvent which absorbs less light than the mobile phase?

[HaruYap]

This problem occurred quite randomly; for a couple of injections, the run is just fine and other subsequent runs will give the same problem again.

My solvent system is water with 0.5% formic acid and organic of ACN with 0.5% formic as well.

My sample solvent is methanol.

Any idea on how to solve this problem? I will try to inject a blank sample solvent of methanol and compared with ACN injection to see if the problem is due to the difference in absorbance of methanol and ACN.

[HW Mueller]

The problem is random? Are you injecting different volumes or are there some diffs in the samples?
Injecting pure ACN should create even a stronger negative peak, inject a solution which is identical to the mobile phase at injection time. Anyway, why does this negative peak bother you?

[Consumer Products Guy]

That's a LONG rin time, especially from someone named Hurry Up !!! (couldn't resist...)

[HaruYap]

Hi Guy; what do you mean by long rin time? Its ok about the Hurry Up..hahaha..

ANyway...Mueller; it bothers me because I dont see this problem from my previous sample runs. Nope, the sample matrices are the same. And this problem occurs even when I'm injecting a blank ie methanol. Do you mean ACN in the gradient my actually causes the negative peaks but why randomly? Randomly in the sense not the sample injection volume or different samples but it occured after a few good runs (good runs = no negative absorbance). Anyway, I think the problem persists now even after using blanks.

Does that mean the DAD is problematic? How do i verified that?

[HW Mueller]

Instead of hurry up it might be good to calm down, thus you could see that I talked of an injection of ACN, not mobile phase ACN, etc. , etc.
It´s understandable that this is random if the negative peak is small and the baseline is not very smooth. The negative peak might just be obliterated, sometimes.

[Peter Apps]

If your detector requires you to set a reference wavelength, the elution of something that absorbs at the reference wavelength will cause a downward deviation in the baseline.

Peter

[HaruYap]

Thanks Peter. Perhaps, I will try to change the reference wavelength and see how does it goes from there.

Just a trivial question (please dont mind); normally, will diode array detector absorbs lower wavelength at earlier times? Meaning, will compounds that absorbs lower wavelength elutes out earlier?

My negative absorbance peaks eluted out at the beginning of the run ie the first 10-15minutes. Maybe its due to the retained compounds in the column towards the end of the run; but anyway, I'll have to verify it.

Thanks everyone for your kind support!!!

[mardexis]

Peter is likely correct about the reference wavelength. The size of the negative peaks will be proportional to their concentration just as they are for the positive peaks... hence they appear randomly (only when those particular components are there)

Take a sample which produced negative peaks and acquire full spectral data for the run. You can then look at the spectra of the negative peaks. You would likely observe that the absorbance is larger at the reference wavelenth than at the monitored wavelength. That would satisfy the intellectual curiousity. Unless the negative peaks are a problem I would just leave the reference wavelenght where it is and be satisfied knowing why you have negative peaks.