Chemometric Methods for Estimating the Strain Hardening Modulus in Polyethylene Resins

Area of Science:

• Polymer Science
• Analytical Chemistry
• Chemometrics

Background:

• Estimating physical properties of polyethylene (PE) copolymers is crucial for material characterization.
• Multidimensional data analysis offers advanced capabilities for complex polymer systems.
• Size exclusion chromatography coupled to Fourier transform infrared spectroscopy (SEC FT-IR) provides rich data for polymer analysis.

Purpose of the Study:

• To explore the feasibility of N-way partial least square (NPLS) for estimating physical properties of 1-butene and 1-hexene PE copolymers.
• To evaluate NPLS performance against traditional methods using simulated multidimensional SEC FT-IR data.
• To assess the accuracy of NPLS in modeling the strain hardening modulus (p>) of bimodal PE structures.

Main Methods:

• Utilized N-way partial least square (NPLS) for direct estimation from multidimensional SEC FT-IR data.
• Generated digital sample sets simulating molecular weight distribution and FT-IR spectra.
• Correlated simulated data to specific Y-block responses using NPLS.
• Compared NPLS results with estimations from various algorithms and response surface methods (RSM).

Main Results:

• NPLS successfully modeled the strain hardening modulus (p>) for bimodal PE-like digital structures.
• NPLS predictions closely mirrored analytical results and expected structural effects.
• The performance of NPLS was comparable to linear response surface methods (RSM).

Conclusions:

• N-way partial least square (NPLS) is a feasible and effective method for estimating physical properties of PE copolymers from SEC FT-IR data.
• NPLS provides accurate predictions for strain hardening modulus, comparable to established methods.
• This study validates NPLS as a powerful tool for analyzing complex polymer structures using multidimensional data.