改性麦糟吸附水中三价砷离子机理及条件优化  

Mechanism and condition optimization of arsenite adsorption from water by modified spent grain

作　　者：陈云嫩[1] 熊昌狮 吴健[2] 梁礼明[2] 聂锦霞[1] 吴速英[1] 

机构地区：[1]江西理工大学资源与环境工程学院,赣州341000 [2]江西理工大学电气工程与自动化学院,赣州341000

出　　处：《环境工程学报》2016年第9期4673-4678,共6页Chinese Journal of Environmental Engineering

基　　金：国家自然科学基金资助项目(51164014);江西省教育厅科技计划项目(GJJ13430)

摘　　要：选取啤酒麦糟作为吸附剂原料,通过聚丙烯酰胺进行改性处理后用于吸附水中亚砷离子。静态吸附条件下考察了p H值、As(Ⅲ)初始浓度、反应温度、吸附剂用量等操作参数对As(Ⅲ)吸附容量的影响。利用扫描电镜及红外光谱等表征了麦糟和改性麦糟的结构特征和物理化学性质。通过BP神经网络方法建立模型,而后用训练好的网络对各参数与As(Ⅲ)吸附容量之间的关系进行仿真,得到的均方误差为0.004 06,表明BP神经网络预测性能较好(R2=0.978 0)。Using spent grain as the raw material for adsorbent, adsorptional experiments of arsenite-contai- ning water onto polyacrylamide-modified spent grain have been performed. The effects of operational parameters including solution pH, initial As （ Ⅲ ） concentration, reaction temperature, and dosage of adsorbent on the ad- sorption capacity of As （ Ⅲ ） were determined under batch experiments. SEM and FTIR were used to characterize the structure and physic-chemical properties of the spent grain before and after modification. The model was based on back-propagation neural network （BPNN）. The relationship between different parameters and the ad- sorption capacity of As （ Ⅲ ） was fitted by the trained model. The test mean square error （MSE） is 0. 004 06 （R2 = 0. 978 0） , which indicated that BPNN performed well.

关 键 词：改性麦糟 三价砷离子 吸附 人工神经网络 

分 类 号：X703[环境科学与工程—环境工程]