Study on different characteristics of TanRai aluminum hydroxide dissolution process in acid and base media were carried out on unactivated and activated samples using 3000rpm centrifugal grinding method under other conditions. The experimental results showed that the centrifugal grinding process did not change the solid phase structure (Gibbsite), and the specific surface of the sample was increased by 1.5 times, average particle size was increased by 10 times, the surface morphology during process of TanRai aluminum hydroxide dissolution were different. In the case of the process of Al(OH)3 dissolution in NaOH media, the corrosion reaction occurs on the all surface of the particle. It goes deep and accesses into the particle. The specific surface might be increased by 10 times in reaction processing, so the grinding did not increase recovery efficiency. In case of dissolution Al(OH)3 in HCl media, the reaction occurs on surface particle so the grinding effects to dissolution recovery that increasing from 2 to 5 times.

ICP-MS method was optimized for the simultaneous analysis of 16 metals (Pb, Cd, As, Hg, Sn, Sb, Co, Ni, Cr, Mn, Mo, Se, Cu, Fe, Zn , Al) in types of medicinal herbs commonly used to produce functional foods. The conditions of samples digestion in a closed system using a microwave and an opened system in Kjeldahl method have been studied in order to flexibly apply different methods of sample preparation in practice. The method was validated with parameters such as calibration curve, limit of detection, limit of quantitation, repeatability RSDr % (0.98 - 19.7%), reproducibility RSDR% (2.72 - 23.5%) and recovery R% (80.3 -109%) meeting the AOAC performance requirements. The method was applied to determine heavy metals in 40 samples of medicinal herbs of apart from the risk of pollution with some toxic heavy metals such as lead and cadmium in medicinal herbs.

In this study, the simultaneous quantitative analysis of samples using multivarite linear regression based on near infrared spectrum has been developed. The model of partial least squares (PLS) matrix with seven value vectors is built with a standard matrix of samples (24 samples containing four analytes together with a total number of excipients) are applied for individual and/or simultaneous determination of guar gum, CMC, carrageenan, xanthan gum in real samples. A matrix of test samples including 10 samples created by fouranalytes to assess the accuracy of the PLS model. The results indicated that the relative error varied from 1.1 to 15%. The optimum experimental conditions were: spectral region: 1700 - 900 cm-1; volume ratio of sample/KBr: 2/98 (w/w); weight per capsule was 15 mg. The method was applied to identify four analytes in 15 material samples and food additive mixture samples.