摘要
数据平滑和扫描步长是两个重要求导条件。本工作用五点三次多项式平滑技术对扫描数据进行平滑处理以消除随机噪音,但平滑次数增大会导致信号失真从而使导数光谱的分辨能力降低。扫描步长减小能提高导数光谱的分辨能力,但同时导致导数光谱灵敏度减小。导数光谱中干扰物等效浓度IEC的大小是上述因素的综合反映。本文以IEC为考察指标,讨论了平滑次数和扫描步长的优化。
Raw data obtained in scans include random noise which hampers the calculation of derivatives. In the present paper, a polynomial smoothing technique including a fivepoint cubic polynomial function was used for smoothing raw data. The technique is effective for the elimination of random noise. However, the data smoothing may also bring signal distortions and therefore decrease the resolving power of numerical derivative spec trometry. Step size in scans is another important condition. With a smaller step size, a better sparation of the derivative signals of an analyte from interfering signals can be achieved. But, meanwhile, the derivative sensitivity of an analyte decreases with step size. In order to achieve the best performance of numerical derivative spectrometry, the number of data-smoothings and step size should be optimized with interferent equivalent concentration in derivative spectra as the criterion.
出处
《分析化学》
SCIE
EI
CAS
CSCD
北大核心
1991年第9期993-996,共4页
Chinese Journal of Analytical Chemistry
关键词
ICP-AES
导数光谱法
数据平滑
Inductively coupled plasma
Atomic emission spectrometry,Spectral interferences Derivative spectrometry
