Nonsupervised numerical component extraction from pyrolysis mass spectra of complex mixtures.
Nonsupervised numerical component extraction from pyrolysis mass spectra of complex mixtures.
复制标题
从复杂混合物的热解质谱中进行无监督数值成分提取。
DOI：
10.1021/ac00277a009
复制
发表时间：
1984
影响因子：
7.4
通讯作者：
H. Meuzelaar
中科院分区：
文献类型：
--
作者：
W. Windig;H. Meuzelaar
复杂有机混合物的热解质谱数据可以用描述混合物的纯组分及其相对浓度的子模式来表示。所描述的方法涉及因子和判别分析，不需要在数据集中存在纯成分光谱。它以“方差图”的形式表示数据集中的相关性。该程序的应用讨论了各种样品：生物聚合物，褐煤（褐煤）和草叶。最近，Windig等人通过图形旋转描述了热解质谱（Py-MS）数据集的混合分析程序（1-3）。由于复杂混合物的热解质谱不一定是各个组分光谱的精确线性组合，因此选择了这种视觉辅助方法而不是数学程序。此外，通常没有纯生化成分的参考光谱。这严重限制了图书馆搜索系统的适用性，例如GC/MS(4)，或目标旋转方法，例如Malinowski（5,6）开发的Py-MS数据集。混合分析的其他方法，通常基于因子
Pyrolysis mass spectral data of complex organic mixtures can be expressed In subpatterns describing the pure components of the mixtures and their relative concentrations. The ap-proach described Involves factor and discriminant analysis and does not require the presence of pure component spectra In the data set. It Is based on a representation of correlations In a data set In the form of a “variance diagram”. Applications of the procedure are discussed for various sets of sam-ples: biopolymers, lignites (brown coals), and grass leaves.Recently, Windig et al. described a mixture analysis pro-cedure for pyrolysis mass spectrometry (Py-MS) data sets by graphical rotation (1-3). This visually assisted approach was chosen over mathematical procedures because pyrolysis mass spectra of complex mixtures are not necessarily exact linear combinations of the spectra of the individual components. Furthermore, reference spectra of pure biochemical compo-nents are often not available. This severely limits the ap-plicability of library search systems, such as used in GC/MS (4), or of target rotation methods, such as developed by Malinowski (5, 6), for Py-MS data sets. Other methods for mixture analysis, often based on factor