摘要
目的采用近红外漫反射光谱技术,以西咪替丁为样本考查定量模型稳定性的影响因素.方法在淀粉辅料中掺入不等量西咪替丁(含量范围为40%～80%),建立校正样品组,采集近红外光谱,运用偏最小二乘法建立数学模型,通过改变激光波长及光圈采集光谱建立预测样品组并进行预测,以考查模型的稳定性.结果校正样品决定系数(R2)=99.77;对照品:波长漂移4cm-1时的决定系数(R2)=99.86;波长漂移8cm-1时的决定系数(R2)=99.76;在poly光圈下的决定系数(R2)=99.81.结论近红外漫反射光谱定量模型相对稳定,为模型转移提供一定的理论基础.
Aim To study the factors influencing the stability of the quantitative model by near-infrared diffuse reflectance spectrometry （NIRDRS）, and then validated by cimetidine. Methods The calibration set was created by adding unequal cimetidine to starch （content range：40%-80% ） and the test set by excursion of wavelength and change of aperture. The NIR spectrum of the samples was collected and a mathematical model was established with partial least squares （PLS） to predict the test set. Results The coefficients of determination （R^2） were 99.77 for the internal crossvalidation set obtained. The coefficients of determination （R^2） were 99.86 for excursion of wavelength 4cm^-1,99.76 for excursion of wavelength 8cm^-1,99.81 for change of aperture. Conclusion The result showed that excursion of wavelength and aperture had some impact on the quantitative result, but little effect on the stability of the quantitative model, which was referential for the transfer of the quantitative model.
出处
《解放军药学学报》
CAS
2005年第4期280-283,共4页
Pharmaceutical Journal of Chinese People's Liberation Army
基金
上海市科学技术发展基金资助项目
No.98441906
