Chromatography Forum

I am looking for some advice here. We currently have an application in our laboratory where we use an Evaporative Light Scattering Detector. As is often the case with ELSD we see some variability from run to run.

I am wondering if anyone has experimented with the use of an internal standard to address this problem. I guess the key question becomes: will there be significant variability in the course of a single run. If so then using an internal standard may not help - and might even work against you (or at least it would be critical for the internal standard retention time to be close to that of the analyte).

Any experience people have to share on this issue would be much appreciated.

An "easy" way to check, if your existing chromatograms have more than one peak, is to look at the ratio of areas of two peaks (preferably of comparable size). In other words, pretend that one of your existing peaks is an IS.

If the reproducibility of the ratios is better than than of the areas, then an internal standard will help.

I tried your suggestion Tom.

Surprisingly, even for two peaks that are only separated by about 2 minutes, the area counts do not "track together" (meaning when the area of one peak increases the area of the other peak does the same, and vica versa).

This suggests that the variability one gets from an ELSD detector changes not just from run to run but from second to second!! It suggests that use of an internal standard would not be well advised with ELSD in general.

I'm curious to hear - have other people also found this to be the case.

Adam,

I would suggest lowering temperature of ELSD and injecting sample without column several times and measuring peak area for each injection. May be some of your problems come from injection or trapping of some of he compounds somewhere (column, lines, injector).
We are using sodium hydroxide as internal standard for ELSD. We measure peak area for sodium every time we are injecting something. SHould stay the same from injection to injection (same volume of injection and same concentration of sodium)

This is quite of a surpise to me. For columns that are fully equilibrated, run to run reproducibility of non-volatile analytes is pretty good (assuming that you do not inject around the LOD).

Well...I'm still curious to hear other opinions on this.

The fundamental issue is that the nebulization process is not under control as much as we would like. That is why ELSD has higher RSDs than other detectors. It is for the same type of reason that mass spec has higher variability than something like UV!

It suggests that use of internal standards is not well advised for these techniques.

If no one has feedback to share on ELSD, has anyone had a mass spec situation where you got better results without an internal standard.

Thanks

In the case of mass spec you can use 13C isotopes of the same compound that will have the same ionization efficiency and retention time as the analyte so I would also be suprised if anybody reports that they get better results without an internal standard when compared to isotopically labeled internal standard methods