Chromatography Forum

I am new to GPC and I am currently using an Agilent Infinity II1260 GPC/SEC system (OpenLab CDS-11 Version 2.6) with a refractive index detector to measure the molecular weight of my standards and samples. I have questions about calibration and triple analysis.
We are conducting molecular weight control in accordance with the United States Pharmacopoeia monograph. We used Dextran standards ranging from 5000 to 410000 Da, a refractive index detector, and two columns: 7.8-mm × 30-cm; packing L39 with pore sizes of 1000 and 120 Å, respectively.
My calibration curve, based on retention time by GPC/SEC, is perfect with an R2 = 0.9999 when using my standards, which already have known Mw values. However, when we calculate the Mw of our samples, the results do not meet the requirements and are almost twice as high. We have noticed that when we change the Mark-Houwink constants (used for the viscometric detector), the results change significantly. But I know that we don`t need this parameters in our calculation because we don`t have viscosity.
Could it be possible that the detector was incorrectly selected during the program installation, as the change in Mark-Houwink constants critically affects the result? What could we be doing wrong?"

Are your samples also the SAME type of "Dextran" as your standards are? If they are, then it is reasonable to achieve similar Mw within the same ranges that you calibrated across. If however, your samples and standards are different compounds with different hydrodynamic volumes, then it would be unreasonable to assume that your data would match your expectations.

My samples are the same type of dextrans as standards.
What about Mark-Houwink parameters? When we must use it?

The M-H constants are dependant on the solvent, the Mwt of the dextran and temperature. They are critical in getting an accurate determination of the Mwt.
The triple detector is probably using default values that are far removed from your application. Variations in the alpha don't have a big effect but the small changes in the kappa can easily give you Mwts that are twice what they should be.
The link below shows a table of M-H values for dextran.
https://www.semanticscholar.org/paper/D ... e546df08c8

If you are new to GPC Wikipedia is also a good place to learn some more.