沉香的近红外光谱法鉴别初探

目的建立沉香的近红外光谱快速鉴别法。方法收集全国不同来源的沉香样品53批,经检验后分别采集其近红外光谱,根据样品性质分类,用二阶导数法和因子化法建立鉴别模型A,谱段范围为12000～4000cm-1;用二阶导数法加矢量归一化和因子化法建立鉴别模型B,谱段范围为7600～4000cm-1。结果用36个样品进行验证,通过A和B的组合能很好鉴别出伪劣的沉香。结论近红外光谱法可快速、准确地鉴别出伪劣沉香,可在药品检测车上推广应用。     

阿昔洛韦片通用性近红外定量分析模型的建立

目的 采用近红外漫反射光谱分析技术和化学计量学的方法对阿昔洛韦片进行快速定量分析.方法 以全国不同企业生产的阿昔洛韦片(国家评价抽验品种)为分析对象,用光纤探头测定近红外漫反射光谱,光谱预处理方法为一阶导数和矢量归一化,谱段范围为6684.2～5881.9 cm-1,4755.7～4300.5 cm-1;回归方法采用偏最小二乘法.结果 93个校正集样品经交叉验证建立校正模型,浓度范围为27.57～77.56 mg·mg-1,交叉验证均方根误差(RMSECV)为1.28,决定系数为99.01％;用92个样品进行外部验证,外部验证均方根误差(RMSEP)为1.17,相关系数为99.08％.结论 该法快速、简便,结果准确,满足药品现场快速检测的需要.     

近红外光谱快速鉴别金水宝胶囊的真伪

目的 利用近红外光谱快速鉴别金水宝胶囊的真伪.方法 在12000～4000 cm-1波段范围内对金水宝胶囊进行全谱扫描,并对图谱进行一阶倒数和矢量归一化处理.结果 正品金水宝胶囊与伪品金水宝胶囊的近红外光谱在7317～6927.5 cm-1、6738.5～6476.2 cm-1、5859～5523.4 cm-1范围内存在区别.结论 本方法简单、快速、准确,可应用于金水宝胶囊的日常监督打假.     

运用近红外光谱法对厄贝沙坦片进行一致性检验

目的 应用近红外光谱法对厄贝沙坦片进行一致性检验.方法 使用一阶导数+矢量归一化对某厂家厄贝沙坦片进行预处理,建立一致性检验方法.结果 选择谱段:9000～7500cm-1;6900 ～ 5600cm-1;5000～4250cm-1,预处理方法为一阶导数化+矢量归一化,平滑点数;17,CI:6,使用一致性检验方法可以显...     

烟酰胺片的近红外漫反射光谱快速、无损定量测定

目的建立近红外漫反射光谱法检测烟酰胺片的方法。方法应用傅立叶变换近红外光谱仪和外接积分球以及固体光纤探头,原料粉末采用PbS检测器,素片和空白辅料片采用InGaAs检测器,在12 000～4 000 cm-1,每份样品扫描64次求平均光谱,通过偏最小二乘法建立模型,建模采用波段为9 750.7 cm-1～7 498.2 cm-1。结果交互证实法确定64次扫描最佳主成分数为4,均方误为3.34,预示集平均回收率为99.56%,RSD为2.58%。为确证近红外漫反射光谱法应用的可行性,预示集与参比方法高效液相色谱法测定结果进行比较,采用配对t检验,近红外法与高效液相色谱法比较无显著性差异(P>0.05)。结论显示近红外漫反射光谱法快速,简便,无损,能够用于烟酰胺片的含量测定。     

近红外光谱法用于六味地黄丸粉末混合过程的质量控制研究

目的利用近红外光谱法对六味地黄丸生产粉末混合过程进行质量控制研究。方法利用偏最小二乘、主成分分析-BP神经网络、小波变换-BP神经网络对模拟样品数据进行处理。结果三种方法所得预示集回收率分别为:99.25%、100.0%、102.0%、100.2%、104.6%、99.55%,RSD分别为5.77%、4.79%、6.69%、7.61%、9.92%、6.62%。结论近红外光谱法基本可以满足药品生产过程中粉末混合均匀度测定的要求。     

近红外一致性检验模型快速鉴别苯妥英钠片

目的：利用近红外光谱法快速鉴别苯妥英钠片的真伪。方法以全国不同企业生产的苯妥英钠片为分析对象,用光纤探头测定近红外反射光谱,光谱预处理方法为一阶导数和矢量归一化法,谱段范围为9000～7500cm -1、6900～5600cm -1和5000～4250cm -1。结果模型能准确鉴别苯妥英钠片并区分不同的生产厂家。结论所建立的方法可以为快速鉴别不同厂家生产的苯妥英钠片提供参考。     

近红外光谱法快速检验熊胆粉

目的:建立熊胆粉的近红外光谱快速检验方法。方法:收集全国不同来源的熊胆粉样品20批,经检验后分别采集其近红外光谱,以二阶导数加矢量归一化法和因子化法建立鉴别模型,谱段范围为7550～4 000cm^-1。以一阶导数加矢量归一化法建立含量模型,谱段范围为12000～4250cm^-1。结果:鉴别模型鉴别10批样品准确率达到100%,含量模型测定20批样品偏差不超过3%。结论:近红外光谱法可用于熊胆粉的快速检验,适合在药品检测车上推广应用。     

近红外光谱对血浆中格列齐特浓度的检测

目的建立一种高效无损检测格列齐特血药浓度的方法 .方法采用近红外光谱法检测,近红外光谱的检测范围为12500~4000cm-1,分辨率4cm-1,扫描次数为64次.结果偏最小二乘回归在6200~5400cm-1范围内进行分析,检测浓度范围在0.05~2.15μg·mL-1,RMSECV和RMSEP分别为0.153和0.467.结论用近红外光谱对格列齐特血药浓度的测定是可行的.     

近红外光谱法测定腰痛宁胶囊中士的宁、马钱子碱和麻黄碱

目的通过考察近红外光谱法对腰痛宁胶囊粉末中士的宁、马钱子碱和麻黄碱的检测,探索一种腰痛宁胶囊生产过程中的快速检测方法。方法采集不同批号腰痛宁胶囊药粉进行近红外光谱检测,以偏最小二乘法建立校正模型,通过该模型与测得的近红外光谱对腰痛宁胶囊中主要成分进行检测。结果经过预处理后的近红外光谱数据与样品中的士的宁、马钱子碱、麻黄碱数据呈现明显的线性关系,各指标的相关系数分别为0.924、0.944、0.874。预测均方差(RMSEP)分别为0.03、0.06、0.02。结论近红外光谱法能快速检测腰痛宁胶囊粉末中士的宁、马钱子碱,可以指导生产过程中的质量控制。     

A Process Analytical Technology approach to near-infrared process control of pharmaceutical powder blending. Part III: Quantitative near-infrared calibration for prediction of blend homogeneity and characterization of powder mixing kinetics

The Process Analytical Technology (PAT) initiative, undertaken by the Food and Drug Administration (FDA), paves the way for improvement of drug manufacturing through real-time measurements that allow better process understanding. This study is the third and final Part in a series of studies that represent an integrated approach for real-time blend uniformity assessment using near-infrared (NIR) technology. In this study, the development of a quantitative NIR model for prediction of blending end point is presented. Process signature was built into NIR calibration models by using blend samples that were collected from actual blend experiments under different processing conditions. Evaluation of various calibration algorithms including principal component regression (PCR), partial least squares (PLS), and multi-term linear regression (MLR) was performed. It was found that linear regression, using a single wavelength, yielded optimum calibration and prediction results. The blending profiles predicted by the NIR quantitative model correlated well to those determined by the UV reference analytical method. Characterization of intra-shell versus inter-shell powder mixing kinetics and its implication in sensor positioning was also performed and will be discussed.     

End point determination of blending process for trimebutine tablets using principle component analysis (PCA) and partial least squares (PLS) regression

This study showed near Infrared (NIR) and Raman spectroscopy with a multivariate calibration approach were very effective to determine blend uniformity end-point. A set of 36 trimebutine samples containing magnesium stearate, stearic acid, colloidal silicon oxide, talc as excipients (0.9%??1.8%) was acquired from six positions during blending processing with U-type blender from 0 to 30 min. Principle component analysis (PCA) with NIR and Raman spectral data was used to confirm the end-point of blending. After 30 min, the scores of principle component (PC) 1 and principle component (PC) 2 for samples moved into one point, which clearly indicated the mixture of sample became homogenous. In addition, NIR and Raman spectroscopy has been applied to the quantitative analysis of 20 trimebutine samples containing 2??40% in mixture granules, which divided into a calibration set of 15 samples and a prediction set of 5 samples for NIR spectral data. The standard error of calibration (SEC) and standard error of prediction (SEP) are 0.15% and 0.13%, respectively using NIR while SEC and SEP of 0.95% and 0.91% are obtained using Raman spectroscopy. The results showed the NIR and Raman spectroscopy with a multivariate calibration such as PCA and PLS provide the possibility of real time monitoring of homogeneity and content uniformity during blending process.     

近红外光谱法快速分析罗红霉素片的含量

目的：建立测定罗红霉素片含量的近红外光谱(NIR)快速分析方法。方法：以全国不同企业生产的罗红霉素片为分析对象．用光纤探头测定近红外漫反射光谱；定量模型的预处理方法为一阶导数与矢量归一化，波长范围9048．7—6098cm^-1，回归方法为最小二乘法。定性鉴别方法为马氏距离与限定值相比较。结果定量模型的浓度范围为19．5％-70．8％，相对标准偏差小于6％：定性鉴别方法可将罗红霉素片与其他抗生素片剂及安慰剂相区别。结论：该法快速、简便，具有一定的专属性，可用于药品的快速检验。     

近红外光谱法快速测定异烟肼片

目的研究近红外光谱法在异烟肼片快速测定中的应用。方法应用偏最小二乘法建立计算模型,通过方差分析法选择计算波长,主成分分析法选择验证集和训练集,交互验证法选择适当的计算因子数。结果应用所建立的偏最小二乘法模型,对9份异炯肼片测定异烟肼含量,与HPLC法相比,所测结果相对误差≤±0．8％,方法准确可靠。结论可将近红外光谱法应用于异烟肼的快速测定,在异烟肼生产中的过程控制和快速质量检测上有较大应用前景。     

头孢拉定胶囊剂通用性近红外定量分析模型的建立

目的:利用近红外漫反射光谱(NIRDRS)分析技术和化学计量学的方法对头孢拉定胶囊剂进行尤损、快速定量分析。方法:以全国不同企业生产的头孢托定胶囊为分析对象,用光纤探头测定近红外漫反射光谱,光谱预处理方法为一阶导数和矢量归一化,谱段范围为9751～7498.4 cm~(-1),回归方法为偏最小二乘(PLS)法。结果:100个样品经内部交叉验证建立预测模型,浓度范同为58.9%～94.5%,交叉验证均方差(RMSECV)为1.70%,相关系数为0.9693。用46个样品进行外部验证,外部验证均方差(RMSEP)为2.24%,平均相对偏差为2.4%。结论:该方法快速、简便,在胶囊壳外面采集光谱即可,结果准确,可用于药品的现场快速分析。     

Mass-balanced blend uniformity analysis of pharmaceutical powders by at-line near-infrared spectroscopy with a fiber-optic probe

A near-infrared (NIR) fiber-optic probe was used at-line to study the blending dynamics of pharmaceutical powder blends on a mass-balanced basis. The probe with a flexible fiber-optic cable can be inserted directly into a 10-L bin blender to get quantitative readings. NIR calibration models were developed and validated for the probe based on a designed 50-sample calibration set. A model formulation containing acetaminophen, mannitol, Avicel, magnesium stearate and AcDiSol was used in the study. The blending study was conducted at 18 rpm for 20 min. NIR probe scans were performed at 1 min intervals and five different locations in the bin. Thief samples were collected and later analyzed by a bench-top NIR instrument to confirm the results from the probe. Complete blending profiles were constructed and compared based on the assay results from both instruments.     

近红外光谱法鉴别普伐他汀钠片及其原料药晶型的一致性研究

目的 建立全覆盖抽样的普伐他汀钠片的近红外光谱法一致性检验模型,考察制剂工艺的差别和原料药晶型的差异,通过稳健、准确、代表性强的近红外光谱一致性模型实现普伐他汀钠片的快速检验和筛查。方法 对评价性抽验抽取的5个企业中的4个共65批样品建立普伐他汀钠片近红外一致性检验模型,并对4个厂家的原料药的近红外光谱图进行比较。结果 建立了4个厂家普伐他汀钠片剂的近红外一致性模型,预测成功率均为100%;4种原料药和1种无定型粉末的近红外光谱图显示不同晶型光谱图具有差异。结论 近红外光谱法能够用于快速鉴别质量工艺稳定的普伐他汀钠片产品,对制剂工艺进行考察,并能够区分不同晶型的原料药。     

Application of process analytical technology in tablet process development using NIR spectroscopy: blend uniformity, content uniformity and coating thickness measurements

Near-infrared (NIR) spectroscopy was employed as a process analytical technique in three steps of tabletting process: to monitor the blend homogeneity, evaluate the content uniformity of tablets and determine the tablets coating thickness.

A diode-array spectrometer mounted on a lab blender (SP15 NIR lab blender) was used to monitor blend uniformity using a calibration-free model with drug concentration ranging from 2.98 to 9.25% (w/w). The method developed accurately depicted the changes in concentration of the drug during blending and the positive effect of a delumping step in the production process. Blend homogeneity was reached within 2 min of the blending step post-delumping, with relative standard deviation (R.S.D.) values varying from 1.0 to 2.5% depending on the drug concentration of the blend.

A Fourier-transform spectrometer (Bruker MPA) was used to analyze content uniformity and coating thickness with calibration based models. Prediction of a validation set with tablets compacted at pressures not present in the calibration set yielded an root mean square error of cross validation (RMSEP) of 1.94%; prediction of tablets compacted at pressures present in the calibration set yielded a RMSEP of 1.48%. Performance of the model was influenced by several physical tablet properties, which could be reduced by spectral pre-processing.

A model based on reflectance spectra predicted coating thickness and its variation more accurately than the model based on transmission spectra. Inter-tablet coating variation was predicted with NIR and compared to reference thickness measurements. Both methods gave comparable results. Initial inter-tablet variation of tablets sampled in-process during coating was high, but stabilized after 30 min into the process.     

近红外漫反射光谱法快速测定头孢氨苄胶囊含量

目的：利用近红外漫反射光谱（NIRDRS）分析技术对头孢氨苄胶囊进行快速定量分析。方法：以22个企业的98批头孢氨苄胶囊为分析对象,用光纤探头测定近红外漫反射光谱;定量模型的预处理方法为一阶导数与矢量归一化,波长范围11995．6～6098．1cm,回归方法为偏最小二乘法（PLS）。结果：80个样品经内部交叉验证建立预测模型,浓度范围为42．46％～92．32％,内部交叉验证均方差（RMSECV）为1．53％,相关系数为0．9871。用18个样品进行外部验证,外部验证均方差（RMSEP）为1．18％,平均回收率为100．1％,RSD为1．79％。结论：该法快速、准确、简便、无损,可用于药品的快速检验。