Morphological Signatures of Salt Crystals under Controlled Humidity Using Advanced Image Analysis

Area of Science:

• Materials Science
• Chemical Crystallization
• Data Analysis

Background:

• Crystallization processes are influenced by environmental factors.
• Understanding salt crystallization morphology is crucial for various applications.
• Relative humidity (RH) is a key environmental parameter affecting crystal formation.

Purpose of the Study:

• To investigate the effect of controlled relative humidity (RH) on the crystallization patterns of sodium chloride (NaCl) and ammonium chloride (NH4Cl).
• To analyze the morphological and textural features of salt deposits using advanced imaging and statistical methods.
• To evaluate the potential of image analysis and machine learning for salt identification based on crystallization patterns.

Main Methods:

• Design and fabrication of a humidity control chamber.
• Controlled crystallization experiments of NaCl and NH4Cl on glass slides at varying RH levels.
• High-resolution imaging and MATLAB-based analysis for feature extraction.
• Principal Component Analysis (PCA) for pattern recognition.
• Development and testing of deep learning neural network models for salt classification.

Main Results:

• Relative humidity significantly affected salt drying times and crystal morphologies.
• Ammonium chloride (NH4Cl) formed complex dendritic structures that increased in complexity with higher humidity.
• Sodium chloride (NaCl) formed cubic/hopper crystals, with size and aggregation varying based on humidity.
• PCA revealed distinct, humidity-specific crystallization patterns for each salt.
• Deep learning models accurately predicted salt types from crystal morphologies (>97% accuracy).

Conclusions:

• Controlled relative humidity systematically alters salt crystallization dynamics and resulting morphologies.
• Advanced image analysis techniques can precisely quantify these morphological signatures.
• Machine learning, particularly deep learning, shows high efficacy in identifying salts based on their crystallization patterns, demonstrating the potential for automated analysis and quality control.