Determination of the mineral composition of Brazilian rice and evaluation using chemometric techniques Douglas G. da Silva, ab Ieda S. Scarminio, bc Daniela S. Anunciaç~ ao, ab Anderson S. Souza, ab Erik G. P. da Silva d and Sergio L. C. Ferreira * ab The mineral composition of Brazilian rice samples was determined and the data obtained were evaluated using principal component analysis (PCA) and hierarchical cluster analysis (HCA). Nineteen commercial rice samples were analyzed, six being brown, seven parboiled and six white. The elements were determined employing inductively coupled plasma optical emission spectrometry (ICP OES), and the accuracy was confirmed using a certified reference material of rice flour. The PCA showed the formation of three clusters: a group with the brown rice samples, other with the parboiled rice samples and another of the white rice samples. PCA also showed that the brown rice samples are richer in minerals than the white and parboiled rice samples. Calcium, sodium, zinc and manganese in white rice samples are higher than in parboiled rice samples, while the parboiled rice has higher contents of iron, magnesium, strontium, potassium and phosphor. HCA also demonstrated formation of three major groups, confirming the results obtained by PCA. Application of HCA in the subgroups of rice samples clearly showed separation of rice brands and also separation between the raw and cooked samples. The mineral composition in the rice samples analyzed agrees with data reported by other authors. This paper revealed that the mineral compositions for white, brown and parboiled rice are significantly different. 1 Introduction Rice is a grain used to feed more than half human world pop- ulation. Like all plants, the mineral composition of rice depends crucially on the chemical nature of the soil, use of fertilizers and herbicides, irrigation water and other factors. Since this is a food of great importance in the human diet, the nutritional assessment of rice provides data relevant to nutri- tionists and doctors. 1–4 Falahi et al. determined Fe, Zn, Ca, Cu, Pb and Cd in ninety-nine polished white rice samples grown in Iran. 5 Another paper quantied nutrients and toxic elements in rice samples consumed in Thailand. 6 A study characterized using a chemometric technique the rice consumed in Valencia City, Spain. 7 A work evaluated the mineral composition of rice consumed in Pakistan. 8 Antoine et al. determined the content of toxic and essential elements of brown and white rice consumed in Jamaica. 9 Other papers have also been published involving the determination of toxic elements such as mercury 10,11 and cadmium in rice. 12 Principal component analysis (PCA) and hierarchical cluster analysis (HCA) are chemometric techniques of multivariate analysis that allow graphical visualization of analytical data, even when the number of samples and variables is large, examining the presence or absence of natural groupings between samples. 13 PCA reduces the dimensionality of the original dataset, preserving the greatest amount of information. This reduction occurs by means of establishing new orthogonal to each other variables, termed principal components (PCs). 11 HCA can group the samples into classes, based on the similarity of the participants of the same class and differences between members of different classes. The graph obtained is called a dendrogram. 11 PCA and HCA are techniques that complement each other and they have been employed for evaluation of results of data analysis. 14–20 The present paper determined and compared the mineral compositions of brown, parboiled and white rice samples consumed in Salvador City, Brazil. The effect of cooking on the mineral content was also investigated. All the data obtained were evaluated using the multivariate analysis techniques PCA and HCA. 2 Experimental 2.1 Instrumentation The determination of the chemical elements was performed using a Varian model Vista PRO Inductively Coupled Plasma a Instituto de Qu´ ımica, Universidade Federal da Bahia, CEP 40170-290, Salvador, BA, Brazil. E-mail: slcf@ua.br; Fax: +55-71-32355166; Tel: +55-71-32355166 b Instituto Nacional de Ciˆ encia e Tecnologia, INCT, de Energia e Ambiente, Universidade Federal da Bahia, 40170-290 Salvador, BA, Brazil c Departamento de Qu´ ımica, Laborat´ orio de Quimiometria em Ciˆ encias Naturais, Universidade Estadual de Londrina, 86051-980 Londrina, PR, Brazil d Departamento de Ciˆ encias Exatas & Tecnologia, Universidade Estadual Santa Cruz, Ilh´ eus, BA, Brazil Cite this: Anal. Methods, 2013, 5, 998 Received 27th August 2012 Accepted 25th November 2012 DOI: 10.1039/c2ay26158h www.rsc.org/methods 998 | Anal. Methods, 2013, 5, 998–1003 This journal is ª The Royal Society of Chemistry 2013 Analytical Methods PAPER Published on 28 November 2012. Downloaded by UNIVERSIDADE FEDERAL DA BAHIA on 23/02/2015 12:51:03. View Article Online View Journal | View Issue