Análise dos parâmetros que influenciam a técnica de análise de imagem digital: uma abordagem para quantificação de íons fosfato e ferro (III)

Abstract

Digital image analysis has emerged as an innovative and accessible approach in Analytical Chemistry, enabling the use of smartphones as detectors in quantitative methods. However, the lack of standardization and the scarcity of detailed studies on the parameters influencing image acquisition and processing still limit the reproducibility and reliability of such methodologies. This study aimed to develop and optimize a methodology based on digital image analysis, with emphasis on the systematic evaluation of technical parameters that directly affect analytical performance, such as lighting conditions, capture distance, sensor exposure, and the behavior of RGB color channels. To this end, an analytical platform was constructed using white LEDs and an ABS plastic enclosure to ensure controlled experimental conditions. Phosphate and iron(III) ions were used as model analytes for method development and optimization. The phosphate determination was based on the molybdenum blue reaction, while iron(III) was quantified through the formation of the iron thiocyanate complex. Analytical data were extracted from the RGB channels of images captured by different smartphone models. The systematic evaluation and subsequent optimization of experimental parameters led to significant improvements in method performance. The working range varied from 0.25 to 10.0 mg L⁻¹, with intraday precision ranging from 2.1% to 4.9% and interday precision from 4.5% to 6.8% for phosphate; and intraday from 1.6% to 5.7% and interday from 2.2% to 6.2% for iron(III). For the latter, alternative ranges of 0.1 1.9 mg L 1 and 10.0 80.0 mg L 1 were also proposed. Limits of detection (LOD) were 0.22 mg L 1 for phosphate and 0.18 mg L 1 for iron(III), while limits of quantification (LOQ) were 0.68 and 0.54 mg L 1 , respectively, all mathematically calculated. Using a visual method, LOD and LOQ were 0.1 mg L 1 and 0.3 mg L 1 , respectively, for both ions. Accuracy was evaluated using different approaches: phosphate recoveries ranged from 88.5% to 107%, and iron(III) results showed no significant differences when compared to UV Vis spectrophotometry (F and t tests, 95% confidence). Method reproducibility was confirmed using two additional smartphones as detectors, yielding analytical curves with R² values above 0.99 and similar slopes. Method sustainability was assessed using the AGREE tool, resulting in a “greenness” score of 0.5, and aligned with Sustainable Development Goals (SDGs) 3, 4, 6, 9, 12, and 13. Furthermore, the methodology was applied to air quality monitoring in Catalão GO, evaluating phosphate ion concentrations during autumn and winter. The measured concentrations were within legal limits and did not differ significantly from those obtained by UV Vis spectrophotometry. The results demonstrate that the systematic evaluation and optimization of key parameters are essential to improve the performance of image based methods, ensuring greater reproducibility, reliability, and analytical robustness. In addition to being low cost and feasible, the developed methodology highlights the potential of digital image analysis as a sustainable and accessible alternative with applicability across different analytical contexts. The presented evidence contributes to the development of new image based methods, expanding their use in settings with limited infrastructure.