SEC/GPC RI/LS discrepancies

[acho]

I am determining molecular weights with polymers containing acetoxy substituent groups and have difficulty obtaining consistent molecular weight values. The RI data is produced using polystyrene standards and Mark-Houwink parameters and LS data is produced using an independently measured dn/dc value (the GPC columns, RI, and LS detectors are all inline). The LS molecular weight data is higher than the RI data, and additionally, the discrepancy between the two detectors increases with higher molecular weight samples. Could this be due to column adsorption or some error in calibration? If anyone has suggestions please let me know.

[tom jupille]

What about conformation variations/differences?

[acho]

We believe our polymers are linear, however, we aren't completely sure what the cause of the discrepancy is. We don't believe we have long chain branching since the weight and number average molecular weights both increase. The RI and LS readings are close at lower molecular weights (<100k) but diverge as the weights increase. Basically, the RI readings are pretty constant while the LS readings increase rapidly towards a plateau value. I am by no means a chromatography expert, it could be that our polymers undergo some conformational change we are unaware about (hydrogen bonding, adsorption onto columns?). If you have suggestions/ideas please let me know.

[Uwe Neue]

Please describe the nature of your polymer, the columns that you are using, the temperature and the solvent system that you are running. This would help in troubleshooting if there is any adsorption possible, and what one can do about it.

[acho]

Our polymer is poly(4-acetoxystyrene), it has an acetoxy substituent group exposed in the para position of the phenyl ring, which could possibly be a cause of adsorption of hydrogen bonding. We are running at room temperature in THF using Waters Styragel columns.

[Uwe Neue]

I do not expect to see any adsorption of your polymer in THF on a DVB-type column. I also do not see a reason for self-interaction of this polymer in this solvent. On the other hand, the experimental proof that none of these things are an issue would be to run your system at elevated temperature. If nothing changes, and your problems remain the same, I am completely on the wrong track.

[HW Mueller]

Isn´t there a controversy about these calculations and/or are they not yielding different theoretical radii? Sorry that I can not do better than this, I looked at this a few years ago, couldn´t get the necessary monetary resources to get into this. Right now there is no time to review the material. But Wyatt and Viscotek have a lot of (opposing?) info on this. Hopefully someone can fill in here.

[acho]

Do you have any references regarding any sort of similar situation? We are in the dark as to what could be causing this- we don't know if this is due is a conformational issue with our polymer or some discrepency with our detectors.

[HW Mueller]

Had a little time to look at my old material. I only saw Mark-Houwink used in connection with viscosity, not RI (refractive index). RI gives what the "experts" call secondary information only (like UV, mol. weights [mw] based on retention time) and dn/dc. Viscosity yields primary info: Rh (hydrodynamic ratio) and mw. MALS also allows "primary info.", namely, calc. of Rg (radius of gyration) and mw. Dynamic LS gives Rh and mw (primary info.). The controversy seems be about which LS (MALS, LALS, RALS) works with what mw.

Now, if the different methods would not produce diverging data one could not use them to get at conformational differences, branching, etc.

Refs: As mentioned above + books like Freifelder, Physical Biochemistry, Freeman and Co, NY.