Afkhami Abbas, Abbasi-Tarighat Maryam, Khanmohammadi Hamid
Faculty of Chemistry, Bu-Ali Sina University, Hamadan, Iran.
Talanta. 2009 Jan 15;77(3):995-1001. doi: 10.1016/j.talanta.2008.07.065. Epub 2008 Aug 28.
New complexes of Co(2+), Ni(2+), Cu(2+) and Zn(2+) with a recently synthesized Schiff base derived from 3,6-bis((aminoethyl)thio)pyridazine were applied for their simultaneous determination with artificial neural networks. The analytical data show the ratio of metal to ligand in all metal complexes is 1:1. The absorption spectra were evaluated with respect to Schiff base concentration, pH and time of the color formation reactions. It was found that at pH 10.0 and 60min after mixing, the complexation reactions are completed and the colored complexes exhibited absorption bands in the wavelength range 300-500nm. Spectral data was reduced using principal component analysis and subjected to artificial neural networks. The data obtained from synthetic mixtures of four metal ions were processed by principal component-feed forward neural networks (PCFFNNs) and principal component-radial basis function networks (PCRBFNs). Performances of the proposed methods were tested with regard to root mean square errors of prediction (RMSEP%), using synthetic solutions. Under the working conditions, the proposed methods were successfully applied to simultaneous determination of Co(2+), Ni(2+), Cu(2+) and Zn(2+) in different vegetable, foodstuff and pharmaceutical product samples.
将钴(II)、镍(II)、铜(II)和锌(II)与最近合成的源自3,6 - 双((氨基乙基)硫代)哒嗪的席夫碱形成的新配合物用于通过人工神经网络进行同时测定。分析数据表明,所有金属配合物中金属与配体的比例均为1:1。针对席夫碱浓度、pH值和显色反应时间对吸收光谱进行了评估。结果发现,在pH 10.0且混合60分钟后,络合反应完成,有色配合物在300 - 500nm波长范围内呈现吸收带。使用主成分分析对光谱数据进行降维处理,并将其应用于人工神经网络。从四种金属离子的合成混合物中获得的数据由主成分 - 前馈神经网络(PCFFNNs)和主成分 - 径向基函数网络(PCRBFNs)进行处理。使用合成溶液,针对预测均方根误差(RMSEP%)对所提出方法的性能进行了测试。在工作条件下,所提出的方法成功应用于同时测定不同蔬菜、食品和药品样品中的钴(II)、镍(II)、铜(II)和锌(II)。