Comparative Chemometric Analysis for Classification of Acids and Bases via a Colorimetric Sensor Array.

Area of Science:

• Analytical Chemistry
• Sensor Technology
• Chemometrics

Background:

• Colorimetric sensor arrays offer a simple method for analyte identification and quantification.
• Multidimensional data from these arrays require chemometric analysis for effective interpretation.

Purpose of the Study:

• To evaluate chemometric methods for classifying and quantifying acid and basic analytes using an 8-sensor colorimetric array.
• To determine the optimal chemometric approaches for analyte identification and concentration determination.

Main Methods:

• Utilized an 8-sensor colorimetric array to analyze acid and basic samples (0.5 - 10 M).
• Applied Principal Component Analysis (PCA), Hierarchical Cluster Analysis (HCA), and Linear Discriminant Analysis (LDA) for data visualization.
• Employed LDA, K-Nearest Neighbors (KNN), and Hierarchical Quantization Index (HQI) for analyte identification and concentration analysis.

Main Results:

• PCA, HCA, and LDA effectively separated clusters for different acid/base analytes and concentrations.
• PCA identified sensors with the highest analyte identification efficacy.
• HQI and KNN achieved 100% correct analyte identification; LDA achieved 95% accuracy.
• Solvent and image effects did not necessitate control for the chemometric methods used.

Conclusions:

• Colorimetric sensor arrays coupled with chemometric analysis are effective for analyte identification and quantification.
• Machine learning algorithms (HQI, KNN, LDA) demonstrate high accuracy in classifying analytes and their concentrations.
• The developed method is robust, requiring no specific control for solvent or image variations.