Determination of amylose in Iranian rice by multivariate calibration of the surface plasmon resonance spectra of silver nanoparticles

Abstract

A simple and non-separative analytical method for selective determination of amylose in Iranian rice has been developed. It was based on the reduction of silver ions by amylose and production of Ag nanoparticles, which exhibit surface plasmon resonance (SPR) spectra in the ultraviolet/visible region. The formation of Ag nanoparticles in the presence of amylose was monitored by transmission electron microscopy (TEM) and dynamic light scattering (DLS). The experimental conditions were optimized to obtain the highest yield for nanoparticle formation. Partial least square (PLS) regression as an efficient multivariate spectral calibration method was employed to make a connection between the SPR spectra of the generated Ag nanoparticles and the amylose content (AC) of the rice starch. The number of PLS latent variables was optimized by leave-one-out cross-validation utilizing prediction residual error sum of square (PRESS). The proposed model exhibited a high ability for prediction of amylose concentration in both standard starch samples and real rice samples prepared from different regions of Iran. The relative errors of prediction were almost lower than ±5% for different real samples and the detection limit was 3.23 weight percent of amylose in rice. In comparison to the reference method (Juliano method), the proposed method is simpler and does not need tedious sample preprocessing steps.