Kaolin–Polyvinyl Alcohol–Potato Starch Composite Films for Environmentally Friendly Packaging: Optimization and Characterization

Abstract

This research paper introduces an innovative methodology to produce biodegradable composite films by combining kaolin, polyvinyl alcohol (PVA), and potato starch (PS) using a solvent casting technique. The novelty of this study resides in the identification and implementation of optimal synthesis conditions, which were achieved by utilizing the Response Surface Methodology—Central Composite Design. The study defines starch, polyvinyl alcohol (PVA), and kaolin as independent variables and examines their influence on important mechanical qualities, water absorption capacity, moisture content, and degradability as primary outcomes. The study establishes the ideal parameters as 5.5 weight percent Kaolin, 2.5 g of starch, and 3.5 g of PVA. These settings yield notable outcomes, including a tensile strength of 26.5 MPa, an elongation at break of 96%, a water absorption capacity of 21%, a moisture content of 3%, and a remarkable degradability of 48%. The study emphasizes that the augmentation of kaolin content has a substantial impact on many properties, including degradability, tensile strength, and elongation at break. Simultaneously, it leads to a reduction in the water absorption capacity and moisture content. The study’s novelty is reinforced by conducting an additional examination on the ideal composite film, which includes investigations using FTIR, TGA, and SEM-EDX techniques. The consistency between the predicted and experimental results is noteworthy, as it provides further validation for the prediction accuracy of Design Expert software’s quadratic equations. These equations effectively capture the complex interactions that exist between process parameters and selected responses. This study presents novel opportunities for the extensive utilization of PVA/PS composite films, including kaolin in various packaging scenarios, thereby significantly advancing sustainable packaging alternatives. The statistical analysis provides strong evidence supporting the relevance of the models, hence increasing our level of trust in the software’s prediction skills. This conclusion is based on a 95% confidence level and p-values that are below a threshold of 0.05.