Determinação de cálcio em petróleo utilizando colorimetria com imagens digitais

• Ano de defesa: 2022
• Autor(a) principal: Baumann, Luiza
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Química; UFSM; Programa de Pós-Graduação em Química; Centro de Ciências Naturais e Exatas
• Palavras-chave em Português: Petróleo; Cálcio; Imagens digitais; Calibração multivariada; PhotoMetrix; Crude oil; Calcium; Digital images; Multivariate calibration; CNPQ::CIENCIAS EXATAS E DA TERRA::QUIMICA
• Link de acesso: http://repositorio.ufsm.br/handle/1/24583

In this work, a method was developed for the determination of Ca in petroleum using colorimetry with digital images with the PhotoMetrix application. For this purpose, a portable system was used based on the capture of digital images by an endoscopic camera and the processing and treatment of the images was done using the PhotoMetrix PRO application. The colorimetric analysis was based on complexation reactions with Ca using calcein, murexide and calcon. The pH parameters and concentration of each complexing agent necessary for the formation of stable complexes with Ca were evaluated. A three variable Doehlert matrix (DOE), quadratic model, was also used to evaluate the influences of the total volume of solution (1 to 2 mL), camera focal length (2 to 20 mm) and light intensity (30 at 180 lx) for digital image acquisition. Finally, the regions of interest for capturing digital images (32x32, 64x64 and 128x128 pixels) were optimized and the possible interference caused by the presence of magnesium was evaluated. After the capture of the images, the data were processed in the PhotoMetrix PRO application using the multivariate calibration model by the partial least squares (PLS) method. ASTM D4807 was evaluated as a sample preparation method for further determination of Ca in petroleum samples. Therefore, the saline species contained in the crude oil were isolated in a 0.45 μm membrane after filtration, as recommended by the ASTM D4807 standard. The proposed method with optimized conditions was applied to 5 crude oil samples with Ca concentrations ranging from 5 to approximately 900 mg kg-1, to evaluate the predictive capacity of the model. The results of the proposed method were compared with the values obtained by the solubilization of the petroleum samples with determination by ICP OES and also with the values obtained by microwave-assisted wet digestion using single reaction chamber (MAWD-SRC) with determination by ICP OES, as reference methods. Determination coefficients greater than 0.997 were obtained for the calibration models built for the three complexing agents, using 5 factors. RMSECV values of 0.26, 0.35 and 0.32 mg L-1 were obtained for the regression models for calcein, murexide and calcon, respectively. No significant differences were observed between the proposed method and the reference methods (ANOVA, 95% confidence interval). The detection and quantification limits were 0.62 and 2.05 mg kg-1, 0.12 and 0.40 mg kg-1 and 0.15 and 0.51 mg kg-1, for the complexing agents calcein, murexide and calcon, respectively. No significant interference of magnesium was observed in petroleum samples used in this work. Finally, the colorimetric method by digital images with PhotoMetrix is considered a viable alternative for the determination of Ca crude oil samples, due to its possibility of miniaturization, low cost, operational ease and speed in obtaining and processing digital images.