Standard Reference Material 2881 - National Institute of Standards ...

9. Experimental Results9.1 Calibration of Spectra—Fitting of the DataWe will now describe experimental tests of the various relationships derived inSection 6 culminating in Type A and Type B uncertainties for the MMD. For a fixedmass of matrix, as the amount of polymer increases so should the mass spectral signalfrom the polymer. This has been studied by Goldschmidt and Guttman for PS about3000 u 23 . They show a relatively large range of linearity. We have repeated the samework for three polystyrenes in the range from 6000 u to 12 000 u. An example of thosedata is shown in Figure 2. The problem with this data is not lack of linearity but poorrepeatability. The data is essentially linear but getting a fairly accurate slope seems to beparticularly difficult. Extensive experience in this lab showed that fairly goodrepeatability using the electrospray sample preparation technique could be obtained.However, at times, even while taking great precautions, repeatability would be lost. Theuncontrolled variable could not be ascertained.Although it was necessary to show linearity in the concentration versus signalintensity relationship for single analytes, this was not sufficient to demonstrate that theproper MMD was necessarily derived from the MMS. To do this a mixture of twopolymers ─one of which is held at a constant polymer to matrix ratio and another that isvaried needed to be measured. In this way the mixture can be used to control thepolydispersity. We start with polymers having two different end groups but with similarmolecular mass distributions, specifically OctylA and ButylA. OctylA was held at afixed concentration of 25 mg polymer/ 1 g matrix while ButylA was varied from0 mg polymer/ 1 g matrix to 50 mg polymer/ 1 g matrix. In other cases where we hadOctyl end grouped polymer of 6000 u and 12000 u the butyl was held fixed at about thesame matrix to polymer ratio as the Octyl A and the Octyl ended PS was varied. This isshown in Figure 4 and an example MALDI spectrum is shown in Figure 3. The fact thatthe slope is very close to one is indicative of the negligible effect of the different endgroup on the ablation or charging of the polymer. To check this in more detail at a fixedtotal polymer to matrix mass ratio, the octyl to butyl ratio was varied and the slope of the27