利用中红外光声光谱法对动物粪便进行特征分析。

Characterization of animal manures using mid-infrared photoacoustic spectroscopy.

机构信息

The State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, East Beijing Road 71, Nanjing 210008, China.

出版信息

Bioresour Technol. 2010 Aug;101(15):6273-7. doi: 10.1016/j.biortech.2010.03.010. Epub 2010 Mar 24.

DOI:10.1016/j.biortech.2010.03.010

PMID:20338754

Abstract

The feasibility of using Fourier transform mid-infrared photoacoustic spectroscopy (FTIR-PAS) for rapid characterization of animal manures was investigated. Animal manure samples were collected from various places in China, and probabilistic neural networks (PNN) and partial least squares (PLS) were initially applied in the qualitative and quantitative analysis of animal manures, respectively. The animal manures exhibited distinctive bands, specifically around 2900-3700 cm(-1), 1200-1800 cm(-1) and 500-1100 cm(-1). There were numerous differences in the spectra of different animal manures, and manures were successful identified by PNN model; organic matter contents in animal manure were well predicted by PLS model, and the calibration coefficient (R(2)), validation error and RPD (ratio of standard deviation to predicted error) were 0.93, 2.38% and 2.58%, respectively, suggesting the potential application of FTIR-PAS for the fast characterization of animal manures.

摘要

利用傅里叶变换中红外光声光谱（FTIR-PAS）快速表征动物粪便的可行性进行了研究。采集了来自中国各地的动物粪便样本，初步应用概率神经网络（PNN）和偏最小二乘法（PLS）对动物粪便进行定性和定量分析。动物粪便显示出独特的谱带，特别是在 2900-3700 cm(-1)、1200-1800 cm(-1) 和 500-1100 cm(-1) 左右。不同动物粪便的光谱存在许多差异，PNN 模型成功地对其进行了识别；PLS 模型很好地预测了动物粪便中的有机物含量，其校准系数（R(2)）、验证误差和 RPD（标准偏差与预测误差的比值）分别为 0.93、2.38%和 2.58%，表明 FTIR-PAS 在动物粪便的快速特征化方面具有潜在的应用。