GPC with ELSD

[me00179]

Hi all,
I have a question about the ELSD. What can I use as a flow marker for this type of detector. I tried a small molecule (tridecanol) but I don't see it in my chromatogram. I'm using it on a breeze system, measuring the Mw of some polymeric materials.
Thanks!

[Kostas Petritis]

Anything non-volatile will do it. Sugars for example are good markers

[me00179]

Do you suggest large molecules?

[Kostas Petritis]

What exactly are you trying to determine?

[Koen Hollebekkers]

I would be very careful to use ELSD for GPC . It’s a detector that’s much more sensitive than RID, but the limited linear range makes it tricky to use this detector for GPC. The area of the slices does not correspond with the concentration of the MW in the polymer. This strongly influences the MW values.

For flow marker you could try ethylene glycol.

[AA]

It is impossible to get accurate MW results using an ELSD. ELSD in a non-liniear technique. You can detect your peaks, make cal curves but you will not get the right MW result for your samples. Stick with your RI.

[mbicking]

As I recall, all of the equations for molecular weight distributions are based on concentration, so results obtained with a "concentration sensitive" detector such as absorbance or refractive index will give appropriate results.

ELSD is a "mass sensitive" detector, and the relative responses will be different across a broadly distributed polymer sample. I don't know if any of the software packages have algorithms to correct for this.

The only situation where you might be able to use ELSD would be if you have a narrow distribution sample, where the only parameter you are measuring is peak MW based on retention time, rather than Mw, Mn, and Mz.

[me00179]

Hi,
I would like to thank you for the response to my question. I’m actually using a GPC/ELSD detector for relatively narrow polymers and copolymers. (PDI~1.2-1.5). Here is a link with some technical bulletins on ELSD use.
http://www.polymerlabs.com/literature/t ... l?dir=ELSD
I don’t really understand the importance of concentration in MWD calculations.

[mbicking]

Upon consulting a GPC expert, I learned that the primary problem with ELSD is the non-linear response, as noted above.

Remember, in GPC you are calibrating by retention time, not area or height, as in most chrom. methods. For a broad MW sample, the software splits up the peak into individual time slices, each representing a certain MW, based on the calibration. The response (height) for that slice represents the amount of material present in the sample at that MW. The software then collects al these slices and calculates the various parameters.

But the main assumption is that the detector response is directly proportional to the amount of material present in the slice. Unfortunately, the ELSD has a non-linear response, so the calculations that result will be in error. For narrow distribution samples, you are really only concerned with the maximum MW, so the answers may be acceptable. However, anytime you try to calculate the usual MWD values (Mn, Mw, Mz), errors will result.

[Kostas Petritis]

You would think that the values can be corrected as you know that you can get linear plots if you plot the log of peak areas vs. log concentration. Maybe you can not really use the commercial GPC software that assume linear detectors but you should be able probably figure something out...

[Koen Hollebekkers]

You would think that the values can be corrected as you know that you can get linear plots if you plot the log of peak areas vs. log concentration. Maybe you can not really use the commercial GPC software that assume linear detectors but you should be able probably figure something out...

This is described in Journal of Chromatography A, 1076 (2005) 51–61. At the moment there is no commercial solution for this. MAybe the new NQAD is an option.

http://www.quant-nqad.com/

[Consumer Products Guy]

True, ELSD is not as linear as most "traditional" detectors. What we do is make sure our standards and samples are relatively close in concentration, and re-make if that's not the case. For example, one could trust quantitation if standard was half or twice concentration in sample.

[mbicking]

CPG:
Yes, this works fine for normal quantitative work, but with GPC the calculations become skewed. The "ends" of the distribution (both large and small molecules) do not give the same relative response, and in some cases are not detected at all. And since these molecules have a siginificant impact on the MWD parameters, the end result is error in the numbers.

According to my contact at Polymer Labs, the only people doing GPC with ELSD are either doing narrow distribution samples, or what he calls "extreme" GPC, where there is really nothing else that works.