Autenticação de origem do sal de cozinha utilizando espectroscopia de absorção atômica em chama e no infravermelho próximo com instrumento portátil

• Ano de defesa: 2025
• Autor(a) principal: Lima, Larissa Rodrigues Zanela
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por; eng
• Instituição de defesa: Universidade Tecnológica Federal do Paraná; Medianeira; Brasil; Programa de Pós-Graduação em Tecnologia de Alimentos; UTFPR
• Palavras-chave em Português: Quimiometria; Sal; Análise de componentes principais; Chemometrics; Salt; Principal components analysis; CNPQ::ENGENHARIAS::ENGENHARIA QUIMICA::TECNOLOGIA QUIMICA::ALIMENTOS; Tecnologia de Alimentos
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/37956

Table salt, composed primarily of sodium chloride (NaCl), is widely used as a seasoning and preservative around the world. The most common types include refined salt, sea salt, Himalayan pink salt, fleur de sel, and light salt. Since its mineral composition varies according to geographic origin, it becomes possible to identify its provenance. This study aimed to authenticate the origin of table salt from four countries using flame atomic absorption spectroscopy (FAAS) and portable near-infrared (NIR) spectroscopy. Chemometric tools were applied to process the data, aiming to reduce dimensionality and facilitate interpretation. Principal Component Analysis (PCA) was used to decompose the original matrix, while the DD-SIMCA method (Data-Driven Soft Independent Modeling by Class Analogy) defined acceptance limits in the new space based on sample distances. The FAAS results revealed high variability in metal content among the countries. Brazilian salt showed higher Na content, while French salt stood out for its elevated levels of K, Zn, Fe, Mg, and Ca. When applying the DD-SIMCA models, Brazilian salt was completely separated from the others, with only five classification errors, confirming that the metal profile can serve as a fingerprint to identify salt origin. In the NIR spectral analysis, origin-related effects were primarily observed in the regions of 1300–1600 nm (second overtone of water) and below 1000 nm (third overtone). DD-SIMCA models using the first derivative of the NIR spectra resulted in 19 classification errors for raw salt, 23 for dried salt, and only 3 for dried and milled salt. These results indicate that particle size and moisture influence the spectra, highlighting bands attributed to free water and hydration water; after drying, correlations improved for Mg, Mn, and Cu contents. Finally, the performance of DD-SIMCA models based on spectra from dried and milled samples was similar to that obtained with FAAS data. Therefore, both FAAS and NIR techniques proved effective in identifying salt origin, with NIR representing a faster and lower-cost alternative, yielding results comparable to those of FAAS.