NGHIÊN CỨU XÁC ĐỊNH ĐỒNG THỜI CÁC DẠNG ASEN VÔ CƠ VÀ HỮU CƠ BẰNG PHƯƠNG PHÁP PHỔ HẤP THỤ NGUYÊN TỬ SỬ DỤNG KỸ THUẬT HIĐRUA HÓA (HVG- AAS) KẾT HỢP VỚI THUẬT TOÁN HỒI QUI ĐA BIẾN I. KHẢO SÁT TÍNH CỘNG TÍNH VÀ XÂY DỰNG MÔ HÌNH HỒI QUI ĐA BIẾN TUYẾN TÍNH

Tóm tắt

Speciation in analytical chemistry have a difficulty in separating each specie to determinate them, by using multivariate regression analysis we can simultaneous determinate them without separating. In this study, investigation linear and setup multivariate calibration were base on results in study conditions for the reduction of arsenic species. The results measuring absorption of eight blank solutions in every acide media HCl 6M, 1M and tactric/tactrate buffer (pH 2, 3, 4) were used to calculate LOD (limit of detection) and LOQ (limit of quantitation) values, with As(III) the LOD is 0,1ppb and the LOQ is 0,34ppb; for As(V) 0,6ppb- 2ppb, for MMA 0,15ppb-0,5ppb, for DMA 0,1ppb-0,34ppb. In experiments investigation linearing indicated that each species arsenic have investigated highly with each other, the linear range for As(III) 0,2-10ppb, As(V) 1-40ppb, MMA 0,5-15ppb, DMA 0,5-30ppb. In another experiment, the absorptions of 40 solutions contained four species arsenic in linear range were measured at five acide media and used to calculating coefficients of multivariate calibration by inverse least squares (ILS) method and principal component regression (PCR) method. With PCR, two initial principal component (PC) contains 99,8% variance, are chosen to change origin data and built multivariate regession model. All funtions were programmed and run by Matlab software.