摘要
目的建立拉曼光谱法快速测定紫石英Fluoritum中氟化钙(Ca F2)的含有量。方法采集54批样品的拉曼光谱,EDTA法测定氟化钙含有量,偏最小二乘(PLS)法建立定量模型,联合区间间隔偏最小二乘(si PLS)法进行优化。结果建模谱段为1 550~1 540、1 530~1 520、1 510~1 500、540~535、530~525、475~450、300~250 cm^(-1)。氟化钙含有量为77.76%~99.75%时,预测均方差(RMSEP)为0.029 9,R2为80.29%。模型平均相对偏差为2.66%,平均回收率为100.11%。结论该模型预测性较好,可快速、准确测定紫石英中氟化钙的含有量。
AIM To establish a Raman spectrum method for rapidly detecting the content of calcium fluoride（ CaF2） in Fluoritum. METHODS The Raman spectra of 54 batches of samples were collected. The content of calcium fluoride was detected by EDTA. Partial least squares（ PLS） was used for establishing the quantitative model. Synergy interval partial least squares（ si PLS） was applied to the optimization. RESULTS The modeling spectral bands were 1 550- 1 540 cm（-1）,1 530- 1 520 cm（-1）,1 510- 1 500 cm（-1）,540- 535 cm（-1）,530-525 cm（-1）,475- 450 cm（-1） and 300- 250 cm（-1）. When the contents of calcium fluoride were 77. 76%- 99. 75%,the root mean squared error of prediction（ RMSEP） was 0. 029 9,and R2 was 80. 29%. The average relative deviation of model was 2. 66%,whose average recovery was 100. 11%. CONCLUSION This model shows a good predictability,which can rapidly and accurately detect the content of calcium fluoride in Fluoritum.
出处
《中成药》
CAS
CSCD
北大核心
2016年第7期1560-1564,共5页
Chinese Traditional Patent Medicine
基金
武汉市2012年高新技术产业发展行动计划生物技术与新医药专项(201260523193)
