Abstract:

Objective This research develops a comprehensive model for enhancing poly(ethylene terephthalate) (PET) production via liquid-state polymerization. By focusing on falling film flow and process simulation, the model aims to optimize the design of falling film reactors and refine the polymerization process, thereby improving the overall efficiency of industrial PET fiber production.

Method A mathematical model for continuous polymerization was developed, focusing on a six-reactor system including five reactors and an additional falling film reactor, based on a specific industrial setup. This model, calibrated with industrial data, defines the input parameters for the falling film reactor. It analyzes changes in key quality metrics like molecular weight and end-group content and explores the influences of temperature and pressure on these metrics within the reactor.

Results In this study, a mathematical model was constructed for a falling film liquid-phase polymerization reactor, tailored to simulate and analyze the distribution of component concentrations and polymer molecular weights within the reactor under varying operational conditions. The model demonstrated high accuracy, particularly under reaction conditions of 287 ℃ and 100 Pa, successfully yielding a product with an intrinsic viscosity (η_intr) value of 1.004 5 dL/g. The η_intr value at the product outlet showed only a 0.05% deviation, demonstrating the model's precision and its ability to reflect the reactor's production process accurately. The model meticulously details the axial distribution of components, showing a decrease in acid and hydroxyl end groups (E_a) and hydroxyl end group (E_g) along the reactor's length, while vinyl end groups (E_v) and degree of polymerization (DPN) increase almost linearly. Notably, the molecular flow rates of small molecules, ethylene glycol quality flow (Q_EG) and water quality flow (Q_w), exhibit a nonlinear response along the reactor length, with an initially high evaporation rate that gradually slows as the reaction proceeds and the DPN increases. Furthermore, the model assesses how molecular weight (M_WN) fluctuates with temperature changes, especially in the range of 270-300 ℃. The M_WN initially rises with the temperature increase, then declines, particularly when the temperature exceeds 290 ℃, leading to a sharp increase in E_v concentration. At 300 ℃, the concentration of E_a surges dramatically to 40 mmol/kg. Concurrently, the contents of diethylene glycol (DEG) and DEG end groups (E_DEG) decrease under these conditions. Under controlled conditions of 270-300 ℃ and 0.01-1.0 kPa, an increase in vacuum level results in a rise in M_WN and a reduction in the production of E_a, E_DEG, and DEG. This meticulous process optimization was evidenced by the enhanced M_WN value of approximately 39 000 g/mol at the outlet, underscoring the model's effectiveness in optimizing PET production.

Conclusion This study established a mathematical model for a polyester falling film liquid-state polymerization reactor, integrating a liquid phase plug flow model, a fully mixed gas phase model, and coupling PET reaction kinetics with gas-liquid mass transfer and high-viscosity fluid dynamics to simulate the polycondensation process. This model reflects the concentration of components and molecular weight distribution inside the vertical falling film reactor, indicative of the degree and progression of high-viscosity molten polymerization, typically challenging to measure online in industrial production. Exploring the relationship between component and molecular weight distribution along the reactor axis could lead to the development of direct spinning processes for polyester products with different intrinsic viscosities. The model also analyzes the impact of reaction temperature and pressure on the molecular weight and end-group content during polymer melt polycondensation, suggesting that controlling reactor temperature (285-290 ℃) and pressure (0.1 kPa) with an input melt viscosity of 0.63 dL/g and carboxyl end group of 31 mmol/kg can produce high-viscosity polyester products with an η_intr value about 1.0 dL/g. Sensitivity analysis of the model to determine optimal operating parameters offers high industrial application value.

Key words: poly(ethylene terephthalate), liquid-state polymerization, reactor, polymerization, process simulation, polymerization process