Integração de sensores de cor e construção de dispositivos analíticos para medidas em solução e em papel

Abstract

This work describes the development and application of analytical devices based on the integration of digital multispectral color sensors (TCS34721, AS7262, AS7341, AS7343, and AS7265x) with microcontrollers, aiming at the construction of intelligent platforms for analytical measurements by reflectance in aqueous solutions and in paper-based analytical devices. Initially, a comparative study was carried out among the TCS34721, AS7262, and AS7341 sensors for the quantification of dyes in aqueous solution. The dyes tartrazine, indigo carmine, ponceau 4R, and sunset yellow were selected because they cover a wide range of the visible spectrum, allowing the evaluation of analytical performance across different spectral regions. The results demonstrated that the AS7341 exhibited superior spectral performance due to its higher multichannel resolution, enabling linear calibration curves, low limits of detection (1 µmol L⁻¹), and adequate precision (RSD = 1.0%). The results obtained with this sensor were comparable to those from a commercial spectrophotometer. Based on these findings, a procedure was developed for the determination of sunset yellow in isotonic beverages, juices, and gelatin, yielding results in agreement with UV–Vis spectrophotometry. Subsequently, the Smart-Sense device was developed, a portable embedded system integrating the AS7341 sensor, electronic control, and a smartphone-based graphical interface, which was successfully applied to the determination of ammonium, iron(II), and nitrite. This system exhibited performance compatible with spectrophotometric methods, in addition to advantages related to portability, low cost, analytical speed, and connectivity. Additionally, the Smart-PAD platform was proposed, consisting of a mechanized system for the reading of paper-based analytical devices. Parameters such as drying time, gain, and integration time were evaluated, with temporal control being critical due to the slow kinetics of the modified Berthelot reaction and the degradation of the reaction product on paper. The platform performs measurements within 8.2 minutes for 18 spots, minimizing the effects of substrate heterogeneity. Its application to ammonium determination showed good linearity (R = 0.997), an RSD of 2.7%, and a detection limit of 6 mg L⁻¹. Overall, the developed devices demonstrate the potential of integrating multispectral sensors with embedded systems as a viable and accessible alternative for chemical analyses, contributing to the democratization of analytical instrumentation and to the advancement of the Internet of Analytical Things (IoAT).