Development of a chemometric approach to improve the accuracy of copolymer sequence information obtained from pyrolysis gas-chromatography mass-spectrometry data

Number-average copolymer sequence length information can be obtained by pyrolysis-gas chromatography (Py-GC) by comparing the ratios of formed oligomers (i.e. dimers and trimers). The formation constants of the oligomers and their detection efficiency are not constant for all fragments, however. This can lead to unrepre-sentative peak ratios in the chromatogram. In these cases, calibration with an external method (e.g. NMR) is required. In this work, we introduce an algorithm that improves the copolymer sequence accuracy yielded from chromatograms with unrepresentative peak areas. The algorithm even functions in cases where oligomer data is missing as the rate of formation of certain oligomers is too low to detect them. One Py-GC measurement and one NMR measurement are required to train the developed algorithm for the determination of average monomer reactivity ratios and relative pyrolysis constants. Afterwards, Py-GC measurements of copolymers containing the same monomers, albeit with different compositions, can be corrected using the previously estimated constants. The algorithm was tested on various styrene-acrylate copolymers, yielding more accurate sequence information, even when limited oligomer information was available.

Automated summary

Information about the number-average copolymer sequence length can be obtained through pyrolysis-gas chromatography (Py-GC) by comparing the ratios of formed oligomers. However, the formation constants of the oligomers and their detection efficiency are not constant, leading to unreliable peak ratios in the chromatogram. To address this, an algorithm is introduced to improve the accuracy of copolymer sequence from chromatograms with unrepresentative peak areas. The algorithm even works when oligomer data is missing.