GPC determines the average molecular weight, you have synthesized new polymer it's all synthetic process depended on physical state and composition of reaction mass. Now you should try in different conditions.

If you have access to it, an accurate technique could be Static Light Scattering. You have to measure various polymer concentration solutions at various angles.

generally when charged polymers are put in to solutions they show poly-electrolyte effect because of charge and exhibit different hydrodynamic volume. when analyzing charged polymers you should first neutralize the charge by adding salt, in your case solvent DMF you can use Lithium chloride and check the molecular weight.

GPC is calibrated with polymer which is not charged. Therefore the value you got obviously erratic. You try the way Suresh has suggested. Of course, the solubility must be checked. 

See if you can tag with one fluorescent molecule per polymer chain. You can estimate Mn to a greater level of accuracy. In general try End group analysis.

The most important thing to realize is that GPC is primarily a chromatographic effect based on molecular size in solution, not molecular weight. . Secondary effects-like ionic repulsive effects and adsorption can cause samples to elute at very different elution volumes than the size exclusion effect.

Have you tried mass spectrometry?

Probably its molecular wt is more than 40K so I have not tried 

How about gel electrophoresis? It's not going to tell you an accurate mass but it will give you an approximate size. 

MALLS Multiangle light láser scattering may be adequate I think. Or SAXS

Could you be specific about the columns and standards you are using in the GPC, and  your copolymerization chemistry?

You could be getting interactions between the charged polymer and the column, causing your sample to have a longer elution time, this usually leads to tailing or fronting in your peaks so you could check that, too, One way to solve this, is adding lithium chloride to your eluent and making sure you are selecting the right columns. Remember as it was already mentioned SEC, when using calibration standards, is a relative method based on molecular size in solution not molecular weight and it assumes that the sample and column do not interact.

Dear Arvind,

From the description description of your problem it seems that youactually have a depolymerization occuring.

If it has a high enough molecular weight, you may want to consider finding the molecular weight distribution of your polymer in the molten state from rheological mesurements instead of using GPC/SEC.

Kind regards

Guy Havet, PhD