Measurement of short chain branching in olefin copolymers using chemometric analysis

A method of determining the short-chain branching distribution in a hydrocarbon sample as a function of its molecular weight distribution, by defining a mathematical relationship between the infrared absorbance of an olefin copolymer test sample in a spectrum of from about 3000 to about 2870 cm−1, and the number of methyl groups per 1000 carbon atoms in the olefin copolymer sample;

• comprising:
• a) providing an olefin copolymer test sample;
• b) isolating at least one fraction from said test sample;
• c) characterizing the molecular weight distribution of said at least one fraction;
• d) determining the infrared absorbance spectrum of said at least one fraction in said spectrum of from about 3000 to about 2870 cm−1;
• and
• e) applying said mathematical relationship to said infrared absorbance spectrum, thereby determining the degree of short-chain branching in said olefin polymer test sample as a function of its molecular weight distribution.

A method for determining the statistical error in the measurement of short-chain branching in an olefin polymer sample as a function of its molecular weight distribution, comprising:

• a) providing multiple replicates of an olefin polymer sample;
• b) isolating from the respective replicate samples at least one fraction having a particular molecular weight distribution;
• c) determining the molecular weight ranges of the replicate sample fractions;
• d) obtaining infrared absorbance spectra of the replicate sample fractions in a wavenumber range useful for determining their degrees of short-chain branching;
• e) determining the degrees of short-chain branching in the replicate sample fractions, by analysis of said infrared spectra;
• f) determining the statistical error in the short-chain branching results in the replicate sample fractions, as a function of molecular weight distribution;
• g) determining the areas of the spectra of the replicate sample fractions, as a function of molecular weight;
• and
• h) finding values of m and b in the following equation that at least approximate the relation between the statistical error and the area of the spectrum for the respective replicate sample fractions:
• E=mA−½+b
• where E is the statistical error for a sample fraction having a particular molecular weight, A is the area of the spectrum of the sample fraction, and m and b are the slope and intercept.