Friday, June 05, 2009

GPC- Quantitation that's qualitative

GPC (gel permeation chromatography a.k.a. SEC (size-exclusion chromatography)) is a fairly common technique for measuring the molecular weight distribution of a polymer. A colleague had some results yesterday and in discussing them, it hit me that despite the technique producing quantitative results, I always look at the results qualitatively.

Here's why: the report we were discussing gave several measure of the molecular weight distribution - Mn, Mw and Mz. The first two have a physical;y intuitive meaning, while the last one doesn't. I've seen other reports that even include Mz + 1, also not a physical quantity. The string of all four of these are simple the first through the fourth moments of the distribution.

So the report had a nice set of numbers for all these moments, calculated out to 3 or 4 significant figures and my colleague was trying to make sense of the numbers. That's when it hit me. I only look at the numbers in a "relative" sense (there are two meanings to that phrase - see below). I seldom care about what the exact number of the molecular weight is, only what the changes are between one sample and the next. Yes, the polydispersity index (Mn/Mw) is a helpful as a measure of the distribution's width, but still, I'm only looking for qualitative changes in the distribution, not exact numbers that I can run calculations with. My colleague seemed a little disappointed with this, as certainly we all take strength from numbers, but to me, the numbers do not add anything.

Back to my double entendre on the "relative" results. The paragraph above gave one meaning, but even the quantitative values are relative in that the instrument is calibrated with polystyrene standards and so the results are always relative to this. (Polystyrene can be easily polymerized anionically, so that the molecular weight distribution is tightly controlled.) There are ways to make the measurements absolute (Wyatt's light scattering technique is prominently advertised), but that is still not something I've ever felt the need for.

In this case, we had samples which we were examining for UV degradation and the associated drop in molecular weight. We found this (qualitatively) so we didn't go any further.

And that's unusual for me. One skill that I learned from my advisor during grad school was to get as much information as possible from each experiment and measurement. It's a skill that I now greatly cherish, although I certainly didn't when I was trying to finish up and he kept wanting me to get more info out of my data! It's a skill that my clients always value, but this is one case where just I don't go any further.

2 comments:

Anonymous said...

I would like to comment that despite the claim that with light scattering absolute molar masses can be obtained, this is rarely really the case, since i) too many parameters need to go into the calculation which are typically assumed and ii)light scattering is stronly discriminating between larger and smaller polymers. Probably most users of GPCs with light scattering detectors are not aware of this, they just believe the stuff companies tell them to buy their equipment. At least that is my impression.

John said...

Definitely good points. I've only been exposed to light scattering in undergraduate classwork. The same goes for neutron scattering, x-ray scattering and any other scattering. Some day I'd like to dive in further, but the need hasn't been there yet.

But I have plenty of experience in being scatter-brained ;)