Quality Control of PVC granulate

Significance of the topic

PVC is a widely used polymer known for its chemical resistance, mechanical stability and flame retardant properties. The molecular weight of PVC directly affects its processing behavior and final material properties. Rapid and reliable monitoring of molecular weight during production is essential to ensure consistent product quality and to optimize manufacturing parameters.

Objectives and study overview

This study aims to demonstrate a fast alternative to size exclusion chromatography for determining PVC molecular weight. A calibration model based on near-infrared spectroscopy (NIRS) is developed and validated to quantify average polymer chain length in seconds.

Methodology and instrumentation

The study used 33 PVC granulate samples covering molecular weights from 113 000 to 192 000 g/mol. Spectra were recorded on a DS2500 Solid Analyzer in diffuse reflectance mode. Vision Air Complete software handled data acquisition, preprocessing and model building. Cross-validation was applied to assess prediction performance.

Main results and discussion

A robust calibration model was obtained with a coefficient of determination (R2) of 0.988 between NIR predictions and reference SEC values. The standard error of calibration was 3 640 g/mol and the standard error of cross-validation was 5 375 g/mol. The NIRS method showed an average deviation of 1.5 percent from the reference method, enabling molecular weight estimation in under one minute without sample preparation.

Benefits and practical applications

• Rapid analysis in seconds enables real-time process monitoring and immediate feedback for production control
• No chemical reagents or sample preparation reduces operating costs and environmental impact
• Robust instrument design allows use in harsh industrial environments
• Simple operation supports decentralized quality control in both laboratories and production lines

Future trends and opportunities

Integration of NIR sensors into process chromatography lines and full automation can further enhance continuous quality assurance. Advances in chemometric algorithms and machine learning will improve prediction accuracy and broaden the range of measurable polymer properties. Connectivity with Industry 4.0 systems and digital twins can enable predictive maintenance and process optimization.

Conclusion

Near-infrared spectroscopy on the DS2500 Solid Analyzer provides a fast, reliable and maintenance-free method for PVC molecular weight determination. The high correlation with reference SEC values and low prediction error make NIRS an attractive tool for routine quality control in polymer production.

References

AN-NIR-098 Quality Control of PVC granulate, Metrohm AG