Exploring the purity of chitin from crustacean sources using deep eutectic solvents: A machine learning approach

Area of Science:

• Biomaterials Science
• Polymer Chemistry
• Machine Learning Applications

Background:

• Chitin, a natural polymer, is abundant in crustacean shells and valuable for biomedical applications.
• Current chitin recovery methods face challenges in achieving high purity and yield.
• Deep eutectic solvents offer an eco-friendly approach to chitin extraction.

Purpose of the Study:

• To develop a novel and robust method for investigating chitin purity from crustacean shell waste.
• To identify key influencing parameters for obtaining pure chitin using deep eutectic solvents.
• To apply machine learning for modeling and predicting chitin purity.

Main Methods:

• Utilized deep eutectic solvents for chitin extraction from crustacean shells.
• Employed machine learning algorithms to model chitin purity based on experimental data.
• Selected input variables to predict chitin purity as the output.

Main Results:

• The XGBoost machine learning model demonstrated the highest predictive accuracy (0.95).
• XGBoost achieved minimal Root Mean Square Error (RMSE) and Mean Absolute Error (MAE).
• Identified optimal input variables for pure chitin production with reduced processing time.

Conclusions:

• Machine learning provides an effective solution for complex chitin purity analysis.
• This approach offers a cost-effective and time-saving method for chitin extraction.
• Validates the utility of machine learning in optimizing natural polymer recovery.