Multi-layered analyses using directed partitioning to identify and discriminate between biogenic amines

Area of Science:

• Analytical Chemistry
• Polymer Science
• Chemometrics

Background:

• Conjugated polymers offer potential for chemical sensing applications.
• Accurate classification of volatile and non-volatile amines is crucial in biological and environmental monitoring.
• Traditional multivariate statistical methods can be limited in complex mixture analysis.

Purpose of the Study:

• To develop and validate a multi-layered statistical analysis for enhanced classification of non-volatile amines using a single conjugated polymer sensor.
• To improve upon the classification accuracy and provide concentration information compared to single-layer analysis.

Main Methods:

• Utilized a single, cross-reactive conjugated polymer sensor for amine detection.
• Employed a multi-layered statistical approach, including directed partitioning and group-ungroup analysis.
• Applied linear discriminant analysis (LDA) for initial classification and principal component analysis (PCA) for concentration determination.

Main Results:

• Achieved 98% accuracy in classifying amines into general classes (aromatic, aliphatic, polyamines) in the first analysis layer.
• Reached near 99% accuracy for aliphatic and aromatic amines and approximately 90% for polyamines in the second layer.
• Determined analyte concentrations within approximately 10% accuracy in the third layer.

Conclusions:

• The multi-layered statistical analysis significantly enhances the classification accuracy of biogenic amines compared to traditional methods.
• This approach enables both qualitative identification and quantitative determination of amines using a single polymer sensor.
• The developed method demonstrates a powerful strategy for complex chemical mixture analysis in relevant biological contexts.