Information Theoretic Analysis of Potentiometric Sensor Array Configurations

Abstract: The simultaneous quantitative determination of multiple ions in complex media can be achieved with the use of potentiometric sensor arrays, or “electronic tongues.” Potentiometric sensor arrays are composed of multiple cross-sensitive sensors whose responses are analyzed with multivariate...

Full description

Saved in:
Bibliographic Details
Main Authors: Sibug-Torres, Sarah May, Enriquez, Erwin
Format: text
Published: Archīum Ateneo 2019
Subjects:
Online Access:https://archium.ateneo.edu/chemistry-faculty-pubs/22
https://ieeexplore.ieee.org/abstract/document/8906332/
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Ateneo De Manila University
Description
Summary:Abstract: The simultaneous quantitative determination of multiple ions in complex media can be achieved with the use of potentiometric sensor arrays, or “electronic tongues.” Potentiometric sensor arrays are composed of multiple cross-sensitive sensors whose responses are analyzed with multivariate chemometric methods to extract the desired quantitative information. One of the most important considerations for designing an array for a given analytical task is the array configuration, or the number and types of sensors incorporated into the array. However, there are currently no theoretical approaches for the a priori design of a potentiometric sensor array, so arrays are currently designed via experimental trial-and-error. In this work, we propose the application of an information theoretic approach as a means to theoretically evaluate the analytical capability of a potentiometric sensor array configuration for a given analytical task. Using a Fisher Information criterion derived for potentiometric sensors, we explored potentiometric sensor array designs with simulated potentiometric sensors, and compared these designs with conventional, empirically-based designs from the literature. Finally, we compute the theoretical performances of various sensor array configurations using the Fisher Information criterion and compare our predictions to the experimental performances from the literature. Our results suggest that Fisher Information can be applied as a theoretical metric to screen for promising array configurations prior to experimental trial-and-error.