• +41 (0)21 729 12 34
  • [javascript protected email address]

Guidelines for GPC in Eurolignin

Guidelines for GPC in Eurolignin

The "guidelines" which I put last year on the ILI Internet Forum have been updated (see below). They have to be further discussed.

Remarks from Polymer Standard Services (PSS) - Mr. Kilz

Items 4:

I assume the peak area of the UV signal is meant and not the "surface", but this is a minor point. importantly, the proposed methods (which have been published, but proven unreliable) will not give a quantitative measure of detected amount of the sample:

  • detector calibration curve will give linear response but incorrect slope (normally the absorbed fraction (not amount) is constant, therefore not visible in this kind of test
  • the problem is that solvent and injection peaks are measured with the sample response; since there is no separation the peak area are wrong and can depend strongly on analytical and injection conditions (repeatability)
  • this method is suitable if the reference sample has identical chemical and physical properties as the sample under investigation I would delete methods a) and b) and add another one which is not depending on any preconditions:
  • the UV absorption coefficient is determined with the sample under investigation in an UV/VIS spectrometer. The UV GPC detector is calibrated by injecting a reference sample of accurately known absorption coefficient and concentration (the reference sample should have no adsorption in the GPC phase system). I suggest a narrow PS reference standard. The same procedure can be applied to RI detectors. In that case the specific refractive increment (dn/dc) is measured and used for detector calibration.

These methods rely on the following equation:

  • A = K(Instrument) * k(sample) * concentration(sample)
  • k(sample) is dn/dc for RI detection and adsorption coefficient for UV/VIS"

Items 6:

I agree to using toluene as an internal standard. However, the resolution of the system cannot be checked with toluene since the resolution is molar mass dependent. Therefore I recomment the use of toluene only for the determination of the plate number and asymetry. I would determine the resolution for each sample and specify a minimum number for the resolution, Rsp, or even better calculate the separation efficiency, T, as specified in the GPC standards ISO 13885-1 and DIN 55672-1. This ensures that there is enough effective separation (and not only potential resolution):

T = (V(Mx) - V(10Mx) [in ml]) / A(col) [in cm²] > 6,0

parameters used:

  • V(Mx) elution volume for the molar mass Mx;
  • V(10Mx) elution volume for die 10 times the molar mass of Mx
  • VA(col) column cross section area (A= ƒ d² / 4)

select Mx so that the peak maximum of the investigated sample is in the middle (center) between V(10Mx) and V(Mx)."

ONE VERY IMPORTANT POINT for the guidelines is that viscometry should be done in all cases (even if there are some technical problems, see item 11): in my opinion the systematic utilisation of that signal (even if noisy and no "smoth curves") is the most important new element in the recent work that will allow to better understand what is the nature of problematic peaks and tails.