No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
NIR spectroscopy as a process analytical technology (PAT) tool for on-line and real-time monitoring of an extraction process
Abstract
The application of near-infrared (NIR) spectroscopy for on-line monitoring of the extraction process of red paeony root was investigated. For NIR measurements, two fiber optic probes designed to transmit NIR radiation through a 2mm pathlength flow cell were employed to collect spectra in real-time. Partial least squares regression (PLSR) calibration models were developed for the parameters of interest: total solid, paeoniflorin, and benzoic acid. The influences of flow rate and air bubble on the NIR spectra and calibration models were also investigated. The established models were used for on-line and real-time monitoring of extraction process, and a model updating method was proposed for the long-term usage of the developed models. Furthermore, both the moving block of standard deviation (MBSD) and relative concentration changing rate (RCCR) methods were used to identify the end point of extraction process. The results of this particular application of implementing NIR spectroscopy to monitor extraction process are very encouraging. Successful models have been built and applied on-line, which proffers real-time data and instant feedback about the extraction course, and in turn, provides improved control.
... As shown in Figure 2-3, in the process of traditional Chinese medicine extraction and production, online NIR spectroscopy can detect the changes of target components in the extraction in real time, and then ensure the extraction time and extraction end point (Figure 2-4) [18] . In the purification process, the online NIR spectroscopy can detect the concentration change of the target component in the effluent in real time, to control the switch between the mobile phase and the eluent and determine the termination point of the elution process. ...
... Schematic diagram of online near-infrared spectroscopy for monitoring the extraction process of traditional Chinese medicine[18] . ...
... The end point of extraction process was determined based on relative concentration changing rate (RCCR)[18] . ...
In recent decades, with the rapid development of artificial intelligence, data mining, and cloud computing, new chemometric methods have sprung up and become one of the fastest-growing branches in spectroscopic analysis technology, which is also a research hotspot for scholars all around the world. This book mainly discusses the chemometric methods used for spectral analysis, including spectral preprocessing, variable selection, data dimensionality reduction, linear or nonlinear multivariate calibrations, pattern recognition, calibration sample selection, outlier recognition, model update and maintenance, multi-spectral fusion, model transfer, and deep learning algorithms, etc.
... However, in situ NIR spectra are greatly affected by the physical and chemical variations found in large-scale reaction systems [9]. For instance, the variation of process variables such as temperature [10], the presence of two-phase interfaces between liquid and solids [11,12], immiscible liquids and gas bubbles [13], the change in optical properties of the material during reaction [8], as well as changes in the NIR instrumentation (e.g. temporal variation of illumination, changes in light transmission due to fiber optics related issues [13]), need to be addressed on a case-by-case basis. ...
... For instance, the variation of process variables such as temperature [10], the presence of two-phase interfaces between liquid and solids [11,12], immiscible liquids and gas bubbles [13], the change in optical properties of the material during reaction [8], as well as changes in the NIR instrumentation (e.g. temporal variation of illumination, changes in light transmission due to fiber optics related issues [13]), need to be addressed on a case-by-case basis. As a result, transferring the advantages of offline NIR spectroscopy to real time process monitoring remains a challenge for the polyester industry and for similar applications. ...
Recent advances in the latest generation of MEMS (micro-electro-mechanical system) Fabry-Pérot interferometers (FPI) for near infrared (NIR) wavelengths has led to the development of ultra-fast and low cost NIR sensors with potential to be used by the process industry. One of these miniaturised sensors operating from 1350 to 1650 nm, was integrated into a software platform to monitor a multiphase solid-gas-liquid process, for the production of saturated polyester resins. Twelve batches were run in a 2 L reactor mimicking industrial conditions (24 h process, with temperatures ranging from 220 to 240 °C), using an immersion NIR transmission probe. Because of the multiphase nature of the reaction, strong interference produced by process disturbances such as temperature variations and the presence of solid particles and bubbles in the online spectra required robust pre-processing algorithms and a good long-term stability of the probe. These allowed partial least squares (PLS) regression models to be built for the key analytical parameters acid number and viscosity. In parallel, spectra were also used to build an end-point detection model based on principal component analysis (PCA) for multivariate statistical process control (MSPC). The novel MEMS-FPI sensor combined with robust chemometric analysis proved to be a suitable and affordable alternative for online process monitoring, contributing to sustainability in the process industry.
... analysis speed, precise results, less costs, no sample preparation, flexible and easy integration of processes. Therefore, spectroscopy is suitable for on-line measurements [34, [36]. ...
On-line spectral measurement and analysis requires fast data processing and rapid analysis. Most of the published work on on-line measurements and analysis depend on MATLAB as the primary programming language. However, MATLAB which is an interpreted language is unarguably slow and not effective for applications involving fast on-line measurements and analysis. Therefore, the study was aimed at developing a fast and an easy to use on-line spectral measurement and analysis software for agricultural materials in C++ programming. The performance of the developed software was demonstrated on two applications involving agricultural materials; detection of blood spots in eggs, and identification of translucency level in pineapples. A total of 281 eggs and 96 pineapples were used for real-time spectral data collection and analysis with a spectral range of 200-1100 nm and 200-850 nm respectively. The level of translucency was graded on a scale of 0 (no translucency) to 3 (very translucent). Partial least squares-discriminant analysis models for bloody egg detection and pineapple translucency were created using a C++ custom developed software. A sample size of 28 new eggs and 46 new pineapples were used for testing the developed models. The bloody egg classification accuracy achieved for the calibration and validation set was 97% and 93% respectively. The pineapple translucency classification accuracy achieved for the calibration and validation set was 72% and 62% respectively. A total number of 26 new eggs were correctly classified with their respective class. A standard error of 0.293 was achieved for predicting the translucency level of new pineapple samples. The average time for processing results after a spectrum was captured was 6.2 milliseconds. Therefore, the developed software is fast and viable for industrial applications.
... The interaction between light and matter provides information about the chemical composition of the sample. Therefore, VIS-NIRS with chemometric analysis can determine the physio-chemical properties of samples in a non-destructive manner, making it a promising technique for determining pineapple maturity and SSC (Wu et al., 2012). Pathaveerat et al. (2008) used acoustic impulse response nondestructive measurement technique and canonical discriminant function analysis to classify the maturity levels of pineapple fruit achieving an overall accuracy of 77% for three maturity stages (immature, mature, and over-mature). ...
Pineapple is a popular tropical fruit with economic value, and the measurement of its maturity and soluble solids content (SSC) is crucial for quality control and sorting purposes. This study developed a non-destructive and rapid method for internal color-based maturity and SSC in pineapples using online visible and near-infrared spectroscopy (VIS-NIRS). The spectral data for light transmitted through the pineapple sample was measured on the conveyor belt while the machine was moving at a speed of 100 mm/s. A spectrometer with a wavelength range of 200-1100 nm was used during online spectral measurements. The pre-processed spectral data were analyzed using partial least squares regression (PLSR). The internal color-based maturity model achieved a correlation coefficient of double cross-validation (Rv) of 0.97 and a root mean square error of double cross-validation (RMSEV) of 0.034. The SSC model achieved a Rv of 0.88 and a RMSEV of 1.04%. The study demonstrates the potential of online VIS-NIRS as a non-destructive and rapid method for measuring pineapple maturity and SSC. Therefore, this offers a potential for real-time quality monitoring of pineapples at a mass production scale.
... Immersion-probe based NIR has been described for inline monitoring of amino acid concentration during hydrolysis of the traditional Chinese medicine Cornu Bubali using PLS for data processing (61). Similarly, in an other example, NIR was used to monitor total solid, paeoniflorin, and benzoic acid during red peony extraction using a PLS regression model (62). In another case, analysis of the chromatographic purification of crude heparin was described, employing NIR monitoring augmented with PLS to select the variable (signal) representing the heparin content (63). ...
Multivariate analysis (MVA) refers to an assortment of statistical tools developed to handle situations in which more than one variable is involved. MVA is indispensable for data interpretation and for extraction of meaningful data, especially from fast acquisition instruments and spectral imaging techniques. This article reviews trends in the application of MVA to pharmaceutical manufacturing and control. The MVA models most commonly used in drug analysis are compared. The potential of MVA to resolve analytical challenges such as overcoming matrix effects, extracting reliable data from dynamic matrices, clustering data into meaningful groups, removing noise from analytical response, resolving spectral overlaps, and providing simultaneous analysis of multiple components are tackled with examples. Industrial applications of MVA capabilities are described, with special emphasis on process analytical technology (PAT) and how MVA can aid in process understanding and control. A scheme for selecting an MVA model according to the available data and the required information is proposed.
... To confirm four chemical compositions and moisture changes in the process of steaming, about 10 laboratory-scale batches were utilized to establish quantitative models. Wu et al. [29] developed a pragmatic model enabling monitoring of the extraction process online in red peony by NIRS. e establishment of the models was utilized to monitor the online extraction process in real time. ...
The quality of tea leaves (e.g., their color, appearance, and taste) can be directly influenced by the tea production process, which is closely connected with the content of a number of chemical components formed during the production of the tea leaves. However, the production process is now controlled by people's experience, making its quality significantly different. NIRS is a time-saving, cost-saving, and nondestructive method. Therefore, it is necessary to introduce NIRS technology into the quality control of the tea production process. In this study, a quantitative analysis model of caffeine, epigallocatechin-3-gallate (EGCG), and moisture content was established by near-infrared spectroscopy (NIRS) which was united simultaneously with partial least squares (PLSR) for online process monitoring of tea production. The model parameters show that the established model has fine robustness and outstanding measuring accuracy. Then, the feasibility of the established method is verified by the traditional method. Through the verification of the precision of the instrument and the stability of the sample, it is clarified that the model can be further utilized to monitor tea product quality online in a productive process.
1. Introduction
Tea is made of tea buds and leaves of Camellia sinensis. It is known as the second largest drink in the world [1]. Its production process is composed of fixation, withering, rolling, fermentation, polling, drying, etc. The different kinds of tea (green, black, and white tea) are basically different in their production process (Figure 1) [2]. The two most best-selling categories are green (unfermented) tea and black (fully fermented) tea, accounting for 98% of the world’s tea production and consumption approximately [3].
... Additionally, all models of polyclones showed good [27,28]; however, RSEP ≤ 25% was acceptable in some reports [29,30]. These results supported practical applications of the developed PLS models for new coming samples. ...
Antibody titer and viable cell density (VCD) are two important parameters that need to be closely monitored during the process of cell line development and manufacturing of therapeutic antibodies. Typically, determination of each parameter requires 10–100 μL of supernatant sample, which is not suitable for small scale cultivation. In this study, we demonstrated that as low as 2 μL of culture supernatants were sufficient for the analysis using UV-Vis spectrum assisted with partial least squares (PLS) model. The results indicated that the optimal PLS models could be used to predict antibody titer and VCD with the linear relationship between reference values and predicted values at R ² values ranging from 0.8 to > 0.9 in supernatant samples obtained from four different single clones and in polyclones that were cultured in various selection stringencies. Then, the percentage of cell viability and productivity were predicted from a set of samples of polyclones. The results indicated that while all predicted % cell viability were very similar to the actual value at RSEP value of 6.7 and R ² of 0.8908, the predicted productivity from 14 of 18 samples were closed to the reference measurements at RSEP value of 22.4 and R ² of 0.8522. These results indicated that UV-Vis combined with PLS has potential to be used for monitoring antibody titer, VCD, and % cell viability for both online and off-line therapeutic production process.
Graphical abstract
... Extraction process is an essential step during traditional Chinese medicine (TCM) preparation, and water decoction is the most important form for extraction [1,2]. The essence of the extraction process of the active ingredients is a mass transfer process where solute dissolves into solvent. ...
Extraction process is not only a critical manufacturing unit but also the initial process of various extracts and preparations. Taking the most extensive Chinese herbal medicine Danshen (Salvia miltziorrhiza Bge) as an example, salvianolic acid B (Sal B) is its main active pharmaceutical ingredient but lacks accurate characterization of the extraction process. As one of process analytical technologies, near-infrared spectroscopy (NIRS) technology has been widely applied for monitoring pharmaceutical extraction process. In most past studies, water spectral information is often eliminated due to its high absorption. However, this study proposed a method of using water spectrum to understand the whole extraction process and to quickly determine the content of Sal B. Principal component analysis (PCA) was first utilized to investigate the whole extraction process, then the reconstructed spectrum based on PCA was established and analyzed by Aquaphotomics, and finally the partial least squares regression (PLSR) quantitative model of Sal B was established. PCA and Aquaphotomics results showed the whole extraction process could be considered as a dynamic change from structure breaker to structure maker, and the dominance of highly H-bonded water structures increases with the extraction time. Also, the Sal B quantitative model with water spectrum showed higher accuracy and stability than other methods, which parameters (RMSEC, RMSECV, RMSEP, R²c, R²cv, R²p, RPD) were 0.2408 mg/mL, 0.2939 mg/mL, 0.2584 mg/mL, 0.9536, 0.9300, 0.9494, 4.6298, respectively, and the paired t-test showed that Sal B content measured by NIR and HPLC methods had no significant differences (p>0.05). In conclusion, all result indicated that water can be used as a probe to understand the traditional Chinese medicine extraction process with NIRS.
... Near-infrared spectroscopy (NIRs), with the characteristics of fast speed, high e±ciency, low cost and nondestructiveness, has been widely used in many¯elds. [7][8][9][10] Compared with traditional analysis, the cumbersome pretreatments are eliminated. The samples are determined with their original state by NIR method. ...
Coptidis Rhizoma (Chinese: Huanglian) and Phellodendri Chinensis Cortex (Chinese: Huangbo) are widely used Traditional Chinese Medicine, and often used in combination because of their similar pharmacological effects in clinical practice. However, the quality control methods of the two drugs are different and complicated, which is time consuming and laborious in practical application. In this paper, rapid and simultaneous determination of moisture and berberine in Coptidis Rhizoma (CR) and Phellodendri Chinensis Cortex (PC) was realized by near-infrared spectroscopy (NIRs) combined with global models. Competitive adaptive reweighted sampling (CARS) and successive projection algorithm (SPA) method were applied for variable selection. Principal component analysis (PCA) and partial least squares regression method (PLSR) were applied for qualitative and quantitative analysis, respectively. The characteristic variables of berberine showed similarity and consistency in distribution, providing basis for the global models. For moisture content, the global model had relative standard error of prediction set (RSEP) value of 3.04% and 2.53% for CR and PC, respectively. For berberine content, the global model had RSEP value of 5.41% and 3.97% for CR and PC, respectively. These results indicated the global models based on CARS-PLS method achieved satisfactory prediction for moisture and berberine content, improving the determination efficiency. Furthermore, the greater range and larger number of samples enhanced the reliance of the global model. The NIRs combined with global models could be a powerful tool for quality control of CR and PC.
... Computational time: Although this measure may not be relevant for the chosen peak alignment methods used in this study, owing to their fast computations, we included this measure for applicability. Chromatographic and spectroscopic data have been successfully used for bioprocess monitoring [24][25][26], which necessitates fast preprocessing techniques to be on par with bioprocess dynamics [27]. Warping algorithms often have very high computational times [28]. ...
... In this study, the PLS method was thus used to extract valid information from the NIR spectrum and a quantitative calibration model was established. e performance of the established model was evaluated via indicators of standard error of cross validation (SECV), the standard error of calibration (SEC), the primary factor (LV), and the determination coefficient (R) [28,[33][34][35]. ...
This study established an approach to rapidly determine the Tween-80 in traditional Chinese medicine (TCM) injection by using near-infrared (NIR) spectroscopy. Totally 133 standard solutions of Tween-80 were prepared and randomly divided into calibration set and validation set, containing 109 and 24 samples, respectively. Spectral data were preprocessed and then subjected to establish a predictive model using partial least-squares (PLS). The standard error of cross validation (SECV), standard deviation of calibration (SEC), and the determination coefficient ( R ) of the established model were 0.0561, 0.0526, and 0.9986, respectively. The model was successfully applied to determine Tween-80 contents in 25 XBJ Injections and 40 FFSX Injections , and it produced satisfactory quantitative analysis results with average relative deviations 0.60% and 0.16%, respectively, for 25 XBJ Injections and 40 FFSX Injections , and the maximum relative deviations 8.57% and 7.60%, respectively. This work shows that NIR model displayed quite good predictive ability for Tween-80 quantitative analysis, which could potentially be applied to rapid determination of Tween-80 in the production process of TCM injections and other TCM products.
... Accordingly, the near-infrared (NIR) spectroscopy analyzer has been widely used recently [5,6]. The greatest advantage of NIR technology is that it can provide estimation results more rapidly than traditional method [7]. ...
Training sample selection is widely accepted as an important step in developing a near-infrared (NIR) spectroscopic model. For industrial applications, the initial training dataset is usually selected empirically. This process is time-consuming, and updating the structure of the modeling dataset online is difficult. Considering the static structure of the modeling dataset, the performance of the established NIR model could be degraded in the online process. To cope with this issue, an active training sample selection and updating strategy is proposed in this work. The advantage of the proposed approach is that it can select suitable modeling samples automatically according to the process information. Moreover, it can adjust model coefficients in a timely manner and avoid arbitrary updating effectively. The effectiveness of the proposed method is validated by applying the method to an industrial gasoline blending process.
... As a novel quality control technology, near-infrared (NIR) spectroscopy has advantages over other analytical techniques. This is because it is a fast, easy to use, and low-cost method (Wu et al. 2012;Li et al. 2013b). Accordingly, NIR has recently been playing an important role in qualitative analysis. ...
To overcome the numerous disadvantages of existing testing technology, a novel, fast, nondestructive, and quantitative technology for quality evaluation of Chinese eaglewood (CE) based on near-infrared (NIR) technology was proposed in this study. The extractives of CE were qualitatively analyzed to determine the types of volatile compounds using gas chromatography-mass spectroscopy and were quantitatively determined using high performance liquid chromatography (HPLC). Agarotetrol was quantitatively determined by the HPLC analysis. The content was found to range widely from 0.016 to 0.104 mg/g. A quantitative prediction model aimed at quality control was proposed based on the qualitative and quantitative results coupled with a partial least squares regression. The coefficient of correlation and residual predictive deviation of the prediction model were determined to be 0.9697 and 5.77, respectively. The practical tests showed an average error of 0.000327%, which indicated that the method was able to provide a novel, quick, and effective quality evaluation of CE.
... Thus, online vibrational spectroscopy has become an increasingly useful tool for research and process development. 12,13,14 Near-infrared (NIR) and Raman spectroscopy are suitable instrumental techniques in monitoring the formation of a desired product in real time. In comparison to gas chromatography-mass spectrometry (GC-MS) analysis which requires considerable time to provide a result, the NIR and Raman spectroscopic techniques provide an immediate response that considerably shortens analysis time. ...
An efficient, versatile and non-destructive in situ method in reaction monitoring using vibrational spectroscopy is described. A Suzuki cross-coupling reaction was monitored in which the substrate 1-iodo-2-nitrobenzene reacted with the electrophile phenylboronic acid to form the product 2-nitrobiphenyl. To hasten the reaction, palladium(II) acetate and potassium carbonate were added to serve as catalyst and to promote transmetalation, respectively. This reaction was monitored using near-infrared and Raman spectroscopy. The recorded data was subjected to multivariate analysis such as principal component analysis in order to detect spectral changes due to the formation of the product. To confirm the presence of the desired product, offline analyses were performed using gas chromatography-mass spectrometry and nuclear magnetic resonance spectroscopy. The results demonstrate how Raman spectroscopy is able to detect the formation of the product in real time, whereas near-infrared spectroscopy fails to do so.
... The RSEP and RPD values were used to assess the accuracy of the predicted results. If the RSEP value was below 20% and the RPD value was above 3.0, the established model accuracy was acceptable [46]. To validate the applicability, the fifth sample batch (the prediction set) was scanned on-line and predicted using the established models. ...
This paper used near-infrared (NIR) spectroscopy for the on-line quantitative monitoring of water precipitation during Danhong injection. For these NIR measurements, two fiber optic probes designed to transmit NIR radiation through a 2 mm flow cell were used to collect spectra in real-time. Partial least squares regression (PLSR) was developed as the preferred chemometrics quantitative analysis of the critical intermediate qualities: the danshensu (DSS, (R)-3, 4-dihydroxyphenyllactic acid), protocatechuic aldehyde (PA), rosmarinic acid (RA), and salvianolic acid B (SAB) concentrations. Optimized PLSR models were successfully built and used for on-line detecting of the concentrations of DSS, PA, RA, and SAB of water precipitation during Danhong injection. Besides, the information of DSS, PA, RA, and SAB concentrations would be instantly fed back to site technical personnel for control and adjustment timely. The verification experiments determined that the predicted values agreed with the actual homologic value.
... The technology has also been applied to Chinese medicine (CM) in the extraction of an individual ingredients; e.g., Ligusticum chuanxiong (Chuanxiong) [4], Salvia miltiorrhiza (Danshen) [5], Paeonia lactiflora (Shaoyao) [6] and Pueraria lobata Ohwi (Gegen) [7]. However, only a few reports mentioned the application of online NIR analysis for multiple ingredients and APIs of low concentration, e.g., Astragali Radix (Huangqi) [8] and Radix Paeoniae Rubra (Chishao) [9]. ...
Background:
This study aims to analyze the active pharmaceutical ingredients (APIs) of licorice (Radix Glycyrrhizae; gancao), including glycyrrhizic acid, liquiritin, isoliquiritin and total flavonoids, in multi-ingredient and multi-phase extraction by online near-infrared technology with fiber optic probes and chemometric analysis.
Methods:
High-performance liquid chromatography and ultraviolet spectrophotometry determined the APIs content in different extraction phases by online near-infrared analysis, which included sample set selection by the Kennard-Stone algorithm, optimization of spectral pretreatment methods (i.e., orthogonal signal correction and wavelet denoising spectral correction), and model calibration by the partial least-squares algorithm, moving-window partial least-squares algorithm and synergy interval partial least-squares (SiPLS) algorithm. The relative errors and F values were used to assess the models in different extraction phases.
Results:
The root-mean-square error of correction, root-mean-square error of cross-validation and root-mean-square error of prediction of APIs in the SiPLS model was less than 0.07. The F values of glycyrrhizic acid, liquiritin, isoliquiritin and total flavonoids were 10,765, 32,431, 649 and 6080, respectively, which were larger than 6.90 (P < 0.01).
Conclusion:
The study demonstrated the feasibility of online NIR analysis in the multi-ingredient and multi-phase extraction of APIs from licorice.
... Recently, the application of near-infrared (NIR) spectroscopy has grown rapidly as an efficient online monitoring technique [2], which has been used as an ideal tool for PAT. The growing concentration on NIR is probably a direct result for its advantages of outstanding sensitivity, high speed, low noise, nondestruction, and enabling the analysis of complex samples without the need for pure samples compared to others [3][4][5]. ...
Online near-infrared spectroscopy was used as a process analysis technique in the synthesis of 2-chloropropionate for the first time. Then, the partial least squares regression (PLSR) quantitative model of the product solution concentration was established and optimized. Correlation coefficient (
R
2
) of partial least squares regression (PLSR) calibration model was 0.9944, and the root mean square error of correction (RMSEC) was 0.018105 mol/L. These values of PLSR and RMSEC could prove that the quantitative calibration model had good performance. Moreover, the root mean square error of prediction (RMSEP) of validation set was 0.036429 mol/L. The results were very similar to those of offline gas chromatographic analysis, which could prove the method was valid.
... Currently, the quality control of extraction process relies heavily on stipulated extraction time and high performance liquid chromatography (HPLC). 2,3 Although reliable and relatively accurate, HPLC is time consuming and requires complex sample preparation, limiting its application to rapid detection and process control. Therefore, a fast and accurate method is required to speed up the determination of the intermediate quality attributes to permit quality control during pharmaceutical production. ...
A particle swarm optimization (PSO)-based least square support vector machine (LS-SVM) method was investigated for quantitative analysis of extraction solution of Yangxinshi tablet using near infrared (NIR) spectroscopy. The usable spectral region (5400–6200 cm⁻¹) was identified, then the first derivative spectra smoothed using a Savitzky–Golay filter were employed to establish calibration models. The PSO algorithm was applied to select the LS-SVM hyperparameters (including the regularization and kernel parameters). The calibration models of total flavonoids, puerarin, salvianolic acid B and icariin were established using the optimum hyperparameters of LS-SVM. The performance of LS-SVM models were compared with partial least squares (PLS) regression, feed-forward back-propagation network (BPANN) and support vector machine (SVM). Experimental results showed that both the calibration results and prediction accuracy of the PSO-based LS-SVM method were superior to PLS, BP-ANN and SVM. For PSO-based LS-SVM models, the determination coefficients (R²) for the calibration set were above 0.9881, and the RSEP values were controlled within 5.772%. For the validation set, the RMSEP values were close to RMSEC and less than 0.042, the RSEP values were under 8.778%, which were much lower than the PLS, BP-ANN and SVM models. The PSO-based LS-SVM algorithm employed in this study exhibited excellent calibration performance and prediction accuracy, which has definite practice significance and application value.
Rebaudioside A (RA) is an ideal sweet component extracted from leaves of Stevia rebaudiana Bertoni, and has been widely developed in the food and pharmaceutical fields. Macroporous adsorbent resin is...
Swertia mussotii Franch. (SMF), a traditional Tibetan medicine, which has miraculous effect on treating hepatitis diseases. However, there is no research on its entire production process, and invisible production process has seriously hindered the SMF modern development. In this study, principal component analysis (PCA), subtractive spectroscopy, and two-dimensional correlation spectroscopy (2D-COS) were used to explain changes of characteristic groups in the extraction process. Four main characteristic peaks at 1884 nm, 1944 nm, 2246 nm and 2308 nm were identified to describe the changes of molecular structure information of total active components in SMF extraction process. In addition, multi critical quality attributes (CQAs) models were established by near-infrared spectroscopy (NIRS) combined with the total quantum statistical moment (TQSM). The coefficients of determination (R2eval and R2ival) were both greater than 0.99. The ratios of the standard deviation of validation to the standard error of the prediction (RPDe and RPDi) were greater than five. The quantitative model of AUCT could save time on primary data measurement by not requiring determination of indicator components compared with others. In conclusion, these results demonstrated that it was feasible to understand the SMF extraction process through AUCT and characteristic groups. These could realize the visual digital characterization and quality stability of the SMF extraction process.
The prediction accuracy of calibration models of near-infrared (NIR) spectroscopy typically relies on the morphology and homogeneity of samples. To achieve non-homogeneous tobacco samples for non-destructive and rapid analysis, a...
Rebaudioside A (RA) and Stevioside (STV) are abundant steviol glycosides (SGs) contained in Stevia rebaudiana Bertoni leaves, which exhibit good stability and various pharmacological activities that have been widely developed in the food and pharmaceutical industries. However, the stevia industry still suffers from high consumption, low efficiency, and long-term dependence on the operational experience of workers. The extraction process of Stevia rebaudiana Bertoni leaves one of the fundamental units for the production of SGs, which is crucial for the homogeneous stability of the final product. The applications of near-infrared (NIR) spectroscopy combined with chemometrics in determining the end-point of the extraction process were proposed in this study. Firstly, the quantitative models were established by partial least squares regression (PLSR) and back-propagation artificial neural network (BP-ANN) to rapidly detect the changes of content RA and STV, which have good prediction results. Secondly, the qualitative analysis methods were established by moving block of standard deviation (MBSD), absolute distance of standard deviation (ADSD) and principal component analysis (PCA) to rapidly determine the extraction end-point, which MBSD method was consistent with the high-performance liquid chromatography (HPLC) method. Finally, the variation of the extraction process was revealed by Aquaphotomic to provide a microscopic perspective of water molecules for end-point determination. The HPLC method requires 30 min to determine the content, whereas NIRS requires only 18 s to obtain a spectrum. These results indicate that the extraction end-point of Stevia rebaudiana Bertoni leaves can be determined rapidly and accurately using NIR spectroscopy, which provides a significant reference for other food, medicinal plants, and agricultural products production processes.
Objective
In a chromatographic cycle, the adsorption process is a critical unit operation that has a significant impact on downstream processes and, ultimately, the quality of the final products. The development of a rapid method to determine the endpoints of adsorption processes in a large-scale manufacturing is of substantial importance for herbal medicine (HM) manufacturers.
Methods
In this study, the adsorption of saponins on a macroporous resin column chromatograph, a critical unit operation in Panax notoginseng (Burkill) F.H.Chen injection manufacturing, was considered as an example. The evaluation results of in-line ultraviolet and visible spectra combined with various multivariate analysis methods, including the moving block standard deviation (MBSD), difference between the moving block average and the target spectrum (DMBA-TS), soft-independent modeling of class analogy (SIMCA), and partial least-squares discriminant analysis (PLS-DA), were compared.
Results
MBSD was unsuitable for adsorption processes. The relative standard errors of prediction between the predicted and experimental endpoints were 13.2%, 4.67%, and 5.71% using DMBA-TS, SIMCA, and PLS-DA, respectively.
Conclusions
Among the considered analysis methods, SIMCA and PLS-DA were more effective for endpoint determination. The results of this study provide a more comprehensive overview of the effectiveness of various multivariate analysis methods to facilitate the selection of the most suitable method. This study was also conducive to address the issues of the in-line detection of adsorption endpoints to guide practical HM manufacturing.
Graphical abstract
http://links.lww.com/AHM/A23
Light is an electromagnetic wave that moves in two orthogonal planes of electric and magnetism. The distance between two crests or troughs is seen as the wavelength, denoted by λ. Electromagnetic radiation is a stream of photons propagating through space at high speed, which has the property of both wave and particle.
Concentrating process is an important unit operation during the manufacture of Lonicerae Japonicae Flos (LJF, “Jinyinhua” in Chinese) products. Near infrared (NIR) spectroscopy and real time release testing (RTRT) combined with statistical process control (SPC) methods were presented for on-line monitoring of concentrating and developing quality control strategy for the final LJF concentrates. An automatic NIR detection (AND) device was designed and has proven effective to eliminate the influence of bubble, solid impurity and flow of the concentrates on NIR spectra. Partial least squares (PLS) models were constructed for on-line determination of neochlorogenic acid (NCA), chlorogenic acid (CA), cryptochlorogenic acid (CCA), density and water content, which provided real-time data and exhibited satisfactory predictive abilities. SPC tools, including univariate Shewhart, multivariate Hotelling T² and squared prediction error (SPE), were compared and applied for RTRT. The results of RTRT showed that the releasing criteria for final LJF concentrates were as follows: 1.09 mg/g < NCA < 1.53 mg/g, 10.00 mg/g < CA < 13.44 mg/g, 1.71 mg/g < CCA < 2.43 mg/g, 1.10 g/ml < density < 1.14 g/ml, 64.07% < water content < 72.99% and SPE result < 0.04. Shewhart and SPE charts have proven to be useful in detecting abnormalities and releasing normal batches. This work demonstrated the effectiveness of NIR and RTRT combined with SPC charts in on-line quality control of industrial concentrating process of traditional Chinese medicine.
The yield, quality, and output value of tobacco leaves are strongly affected by the flue-curing conditions. To effectively control the flue-curing and guarantee quality, it is necessary to quantify the chemical composition of tobacco leaves and provide timely feedback during this process. Therefore, the practicability of on-line monitoring of moisture, starch, protein, and soluble sugars for tobacco leaves by near-infrared (NIR) spectroscopy and deep transfer learning was explored. The results showed that the use of an NIR spectrometer equipped with a fiber-optic probe with a deep learning algorithm accurately predicted the content of these components during the curing process. The convolutional neural networks model showed greater potential for on-line monitoring than partial least squares and support vector machines. Furthermore, a network-based deep transfer learning strategy was crafted to include seasonal and temperature variability to accurately predict samples from a new harvest season in a curing barn. The overall studies indicated the efficacy of NIR diffuse reflectance spectroscopy as a rapid and nondestructive method for on-line and simultaneous determination of moisture, starch, protein, and soluble sugars in the flue-curing process to assist in making decisions.
In this work, an analytical model based on near infrared spectroscopy is proposed to predict the main active ingredients of Poria cocos (PC) during the extraction process, acidic polysaccharides. In order to improve the prediction quality, various pretreatment methods and variable selection methods were carefully compared and the partial least squares method was used to relate the predicted values to the reference ones. As a result, a robust and optimal model was obtained, where the coefficient of determination, root mean square error and mean absolute error were 0.9618, 4.05% and 3.26% for calibration, and 0.9609, 4.17% and 3.31% for prediction, respectively. The residual prediction deviation reached 5.15. Such satisfying results clearly showed that the present model had very good prediction ability and thus had great potential for monitoring of the extraction process of PC in the pharmaceutical manufacturing.
Qualitative and quantitative detection methods based on near-infrared spectroscopy (NIRs) have been proposed in the process analysis of traditional Chinese medicine in recent years. In this study, rapid monitoring methods were developed for quality control of concentration process of lanqin oral solution (LOS). Partial least squares regression (PLSR) method was adopted to construct quantitative models for epigoitrin, geniposide, baicalin, berberine hydrochloride and density. Simultaneously, the genetic algorithm joint extreme learning machine (GA-ELM) was first applied in qualitative analysis of NIRs to distinguish end point of concentration process. Results of PLSR models were satisfactory with the relative standard error of calibration valued at 3.80%, 3.75%, 3.79%, 11.5% and 1.22% for epigoitrin, geniposide, baicalin, berberine hydrochloride and density respectively, and the residual predictive deviation values were higher than 3. For qualitative analysis, the GA-ELM model obtained 100% prediction accuracy. The PLSR quantitative models and the end point discrimination model constructed by GA-ELM correspond with the requirements of practical applications. The results indicate that NIRs in combination with chemometrics has great potential in improving the efficiency in production.
Introduction:
The on-line analysis of active pharmaceutical ingredients (APIs) during the extraction process in herbal medicine is a challenge. Establishing a reliable and robust model is a critical procedure for the industrial application of on-line near-infrared (NIR) technology.
Objective:
To evaluate the advantages of on-line NIR model development using system optimisation strategy, Glycyrrhiza uralensis Fisch was used as a case. The content of liquiritin and glycyrrhizic acid was monitored during pilot scale extraction process of Glycyrrhiza uralensis Fisch in three batches.
Methods:
High-performance liquid chromatography (HPLC) was used as reference method for content determination of liquiritin and glycyrrhizic acid. The quantitative models of on-line NIR were developed by system optimisation of processing trajectory. For comparison, the models were simultaneously developed by stepwise optimisation. Moreover, the modelling parameters obtained through system optimisation and stepwise optimisation were reused in three batches. Root mean square error of prediction (RMSEP) and residual predictive deviation (RPD) were used to assess the model quality.
Results:
The average values of RMSEP and RPD of systematic model for liquiritin in three batches were 0.0361, 4.1525 (first batch), 0.0348, 4.7286 (second batch) and 0.0311, 4.9686 (third batch), respectively. In addition, the modelling parameters of systematic model for glycyrrhizic acid in three batches were same, and the average values of RMSEP and RPD were 0.0665 and 5.2751, respectively. The predictive performance and robustness of systematic models for the three batches were better than the comparison models.
Conclusion:
The work demonstrated that system optimisation quantitative model of on-line NIR could be used to determine the contents of liquiritin and glycyrrhizic acid during Glycyrrhiza uralensis Fisch extraction process.
Process monitoring is essential for the pharmaceutical industry to control the quality uniformity of final products. In this study, the prompt quantitative methods of geniposide in Lanqin oral solution (LOS) during the process of extraction and concentration were developed by near-infrared spectroscopy (NIRs). Simultaneously, the feasibility of the method for universally monitoring the two processes was investigated. Methods including interval partial least square (iPLS), synergy interval partial least square (siPLS) and competitive adaptive reweighted sampling (CARS) were performed for characteristic variables selection. The single-process models of extraction, concentration and the multi-process model based on the combination of the two processes were constructed and discussed. Results showed that the performance of siPLS models were superior to others. For the extraction, concentration and the combination of these two processes, the Rc values were 0.9486, 0.9746 and 0.9980; while the RSEP values were 10.42 %, 5.41 % and 6.17 %, respectively. Compared with the specific models, the strong accuracy and stability of the multi-process model were also demonstrated. In addition, Pearson correlation analysis and paired t-test were also applied to verify the predictive performance of models. The overall results indicate that NIRs combined with siPLS algorithm is a robust method for predicting the geniposide in the extraction and concentration of LOS. Furthermore, the multi-process model of combination is feasible for the universal supervision of geniposide, which is convenient for industrial production applications.
The volatile organic compounds produced in yeast fermentation are directly related to the degree of fermentation and product quality. This study innovatively proposes a method based on an olfactory visualization sensor system combined with pattern recognition algorithm to ensure the correct discrimination of yeast fermentation stage. First, the olfactory visualization sensor system was developed based on a colorimetric sensor array, which was composed of twelve chemical dyes, including eleven porphyrins or metalloporphyrins and one PH indicator on a C2 reverse silica-gel flat plate. It was employed as an artificial olfactory sensor system to obtain odor information during the process of yeast fermentation. Then, principal component analysis (PCA) was used to reduce the dimension of the data which were obtained from the olfactory visualization sensor system. Finally, three pattern recognition algorithms, which were support vector machine (SVM), extreme learning machine (ELM) and random forest (RF), were used to develop identification models for monitoring of the yeast fermentation stage. Results showed that the optimum SVM model was superior to ELM and RF models with discrimination rate of 100% in the prediction process. The overall results sufficiently demonstrate that olfactory visualization sensor system integrated with appropriate pattern recognition algorithm has a promising potential in in-situ monitoring of yeast fermentation.
In the tide of smart manufacturing in the world, many countries have put forward their own reference frameworks for smart manufacturing system. Based on the framework of smart manufacturing system (SMS) proposed by China, a reference smart manufacturing integration system (SMIS) from multi-level and multi-perspective was proposed. The reference model of SMIS was analyzed by means of system engineering. Based on this model, the overall architecture of SMIS was proposed. Then, the physics subsystem, information integration subsystem, network integration subsystem, data integration subsystem, and visualization integration subsystem for SMIS were proposed through the overall architecture of SMIS. Finally, the above integrated subsystems were collected together to deduce the implementation path of SMIS. Four reference subsystems for SMIS proposed in this paper have achieved good effects in the projects we participated in, which were organized by the state. The research results of this paper can be used as a reference for industry and government to design, set, and carry out SMIS. At the same time, it has a certain reference value for the improvement and supplement of national smart manufacturing system architecture.
In traditional Chinese medicine, an extract intermediate is an extracted and purified intermediary of the production process, and its quality is crucial to the efficacy of the final product. Lumbrokinase (LK), which is found in earthworm extract intermediate, has good fibrinolysis, is an anticoagulant, provides hemorheological improvement and has other pharmacological effects, and it shows significant advantages in the treatment of thromboses. The traditional method for determining lumbrokinase potency is an agarose-fibrin plate assay, but it takes more than a dozen hours to perform and requires expensive culture medium. To develop a new method for rapidly and accurately determining the potency of lumbrokinase, near-infrared (NIR) spectroscopy combined with chemometrics is presented in this study. In addition, the variable selection combination method was investigated to optimize the modeling variables and improve the predictive abilities of the NIR model. After optimization, the parameter values of the correlation coefficient, R²c, R²p, RMSEC and RMSEP, of the partial least squares regression (PLSR) quantitative model were 0.9398, 0.9340, 289.618 U, and 538.313 U, respectively. This study demonstrates that NIR analysis techniques are a potential tool for rapidly determining lumbrokinase potency in earthworm extract intermediate.
A framework of back propagation neural deep learning (BPN-DL) was constructed in this work for Fourier transform near-infrared spectroscopy (FT-NIR) to predict the nutrition components in soil samples. Firstly, the spectra were measured by FT-NIR spectrometer on the full-length range of 10000-4000 cm-1, which resulted in 1512 wavenumbers. The content of organic carbon (OC) was determined by use of the biochemical method of potassium dichromate oxidation. Then, some characteristic wavenumbers were selected as the input variables to the BPN-DL framework based on competitive adaptive reweighted sampling (CARS) algorithm. The back propagation (BP) neural deep learning framework was employed to develop the calibration models for the determination of OC content. With the undergoing computer hard configuration, BPN-DL models were established and pre-set screening for up to 32 hidden layers and 50 nodes. The results were achieved in iteration and parameter identification. The best optimal BP neural deep learning model was constructed by 22 hidden layers and 30 neural nodes, with 91 input wavenumbers selected by CARS. The root mean square error of training was 0.104 and that of testing was 0.279. Another available optimal model was with 19 hidden layers and 46 nodes for 216 characteristic wavenumbers. Finally, the optimal results were compared with the benchmark PCA, PLS and conventional BP network models. The BPN-DL framework showed its excellent performance of prediction and generalization. This study indicated that the FT-NIR spectroscopy technology integrated with appropriate chemometric methods could be utilized to quantitatively determine the OC nutrition content of soil, and BPN-DL reveals its superiority in model training and testing processes.
A novel method was developed for the quality control of Ephedrae herba by near-infrared (NIR) spectroscopy. First, qualitative models established by discriminant analysis and support vector machine were used for the preliminary screening of unqualified samples of E. herba. Then quantitative models of ephedrine and the total alkali (ephedrine and pseudoephedrine) were established by partial least squares regression and particle swarm optimization based least square support vector machine. The contents of test samples were predicted by the established NIR quantitative models. As a result, the accuracies of unqualified identification were 98.9% by discriminant analysis and 100% by support vector machine. The performance of the particle swarm optimization based least square support vector machine models were better than the partial least squares regression models. The correlation coefficients were both more than 0.98 and relative standard errors of calibrations were less than 9% in the calibration sets of particle swarm optimization based least square support vector machine models. As for the test sets, the correlation coefficients were both more than 0.93 and the relative standard errors of prediction were less than 13%, indicating satisfactory predicted results. All of these results demonstrated that NIR spectroscopy may be a powerful tool for the quality control of E. herba.
Quality by Design (QbD) for the pharmaceutical industry includes the design, development and production control of products and production processes from the beginning to the end of the product development phase for ensuring the consistent quality of a pharmaceutical product. The QbD is a systematic scientific approach aimed at meeting the needs of the patient in the desired and targeted quality and aiming to produce the same quality pharmaceutical product in this direction. Process Analytical Technology, which is assessed in that regard, is part of a design quality approach that is used to design, analyze, and control real-time measurements of quality and performance criteria for raw and processed materials to achieve the desired final product. This scientific and systematic approach to pharmaceutical product development, which is also acknowledged and supported by the health authorities, serves to the changing and developing pharmaceutical sector.
To improve the yield of industrial fermentation, this study presented a method based on near infrared spectroscopy to predict the growth process of yeast. The spectral data were measured from fermentation sample by Fourier-transform near-infrared (FT-NIR) spectrometer in the process of yeast culture. Each spectrum was acquired over the range of 10000 to 4000 cm⁻¹. Meanwhile, the optical density (OD) values of fermentation sample were determined with photoelectric turbidity method. A method on the basis of competitive adaptive reweighted sampling (CARS) was used to select characteristic wavelength variables of NIR data, and then extreme learning machine (ELM) algorithm was employed to develop the categorization model about the four growth phases of yeast. The experimental results showed that only 30 characteristic wavelength variables of NIR data were selected by CRAS algorithms, and prediction accuracy of the training set and testing set of the CARS-ELM model was 98.68% and 97.37%, respectively. This study showed that near infrared spectral analysis technique was feasible to predict the growth process of yeast.
To improve the yield of industrial fermentation, herein, we report a method based on Fourier-transform near-infrared spectroscopy (FT-NIR) to predict the growth of yeast. First, the spectra were obtained using an FT-NIR spectrometer during the process of yeast cultivation. Each spectrum was acquired over the range from 10000 to 4000 cm⁻¹, which resulted in spectra with 1557 variables. Moreover, the optical density (OD) value of each fermentation sample was determined via photoelectric turbidity method. Then, using a method based on competitive adaptive reweighted sampling (CARS), characteristic wavelength variables were selected from the preprocessed spectral data. Gaussian mixture regression (GMR) algorithm was employed to develop the prediction model for the determination of OD. The results of the model based on GMR were achieved as follows: only 13 characteristic wavelength variables were selected by CRAS, the coefficient of determination Rp² was 0.98842, and the root mean square error of prediction (RMSEP) was 0.07262 in the validation set. Finally, compared to kernel partial least squares regression (KPLS), support vector machine (SVM), and extreme learning machine (ELM) models, GMR model showed excellent performance for prediction and generalization. This study demonstrated that FT-NIR spectroscopy analysis technology integrated with appropriate chemometric approaches could be utilized to monitor the growth process of yeast, and GMR revealed its superiority in model calibration.
A direct synthesis process of clevidipine butyrate was used for testing the real-time detection performance of Raman spectroscopy in this study. The decrease of the reactant (Chloromethyl butyrate) and the increase of the product (Clevidipine butyrate) were used as indexes to evaluate the synthesis process using a 785-nm Raman spectrometer. Calibration models were built for the quantification of chloromethyl butyrate and clevidipine butyrate. The issue of linear regression distortions was overcome by setting solvent (acetonitrile) as internal standard. The validation results indicated that the Raman spectroscopy method was reproducible and accurate for monitoring the synthesis process. The real-time data were achieved to evaluate the esterification reaction of clevidipine butyrate under different conditions.
In this work, near infrared (NIR) spectroscopy was used in combination with chemometrics to determine the epimedin A, epimedin B, epimedin C, icariin, and moisture contents of Herba Epimedii. The variable selection method genetic algorithm (GA) and regression tool support vector machine (SVM) were used to improve the model performance. Four different calibration models, namely Full-PLS, GA-PLS, Full-SVM, and GA-SVM, were established, and their performances in terms of prediction accuracy and model robustness were systemically studied and compared. In conclusion, the performances of the models based on the efficient variables selected through GA were better than those based on full spectra, and the nonlinear models were superior over the linear models. In addition, the GA-SVM model demonstrated the optimal performance in predicting five quality parameters (viz. epimedin A, epimedin B, epimedin C, icariin, and moisture). For GA-SVM, the determination coefficient (Rp(2)), root-mean-square error (RMSEP), and residual predictive deviation (RPD) for the prediction set were 0.9015, 0.0268%, and 2.20 for epimedin A; 0.9089, 0.0656%, and 3.08 for epimedin B; 0.9056, 0.1787%, and 3.18 for epimedin C; 0.8192, 0.0657%, and 2.26 for icariin; and 0.9367, 0.2062%, and 4.12 for moisture, correspondingly. Results indicated that NIR spectroscopy coupled with GA-SVM calibration can be used as a reliable alternative strategy to measure the epimedin A, epimedin B, epimedin C, icariin, and moisture contents of Herba Epimedii because this technique is fast, economic, and nondestructive compared with traditional chemical methods.
Discriminating the maturity levels of tobacco leaf with in-situ measurement can effectively reduce loss rate and quality decline due to misjudgment of the maturity levels of tobacco leaf. In the meantime, the regular way we use to determine the maturity levels of tobacco, which is depend on tobacco leaf age and judgment of tobacco grower, lacks of objectivity. So this paper proposed a method to identify maturity levels of tobacco leaf by using spectral feature parameters combined with the method of support vector machine (SVM). In this paper, a total of 351 tobacco leaf samples collected in 5 maturity levels including immature (MD, unripe (M2), mature (M3), ripe (M4), and mellow (M5) determined by experts were scanned by field spectroscope(ASD FieldSpec3) with in-situ measurement for getting their reflectance spectrum. Through spectral analysis we found that the spectrum of tobacco leaf with different levels of maturity can be distinguished in visible band but not easily be distinguished in near-infrared band, so we use the tobacco leaf spectrum in visible band as the sensitive bands to analyze and model. To find the most suitable input variables for modeling, we use continuous spectrum (350 similar to 780 nm), feature band (496 similar to 719 nm) and spectral feature parameters (the reflectance of green peak, location of green peak, first order differential value of red-edge and blue edge, red-edge and blue-edge area, location of red-edge and blue-edge) in visible region as the input variables, and using these three kinds of input variables in the method of SVM to establish a discriminant model for identifying maturity levels of tobacco leaf. The result shows that, the model using spectral feature parameters gains the accuracy rate of 98. 85%. While the accuracy rates of other two models were 90. 80% and 93. 10%, respectively. The conclusion was drawn that using spectral feature parameters in visible spectrum as the input variables in SVM can improve the model performance. It is feasible to use this method to identify maturity level of tobacco leaf with in-situ measurement.
As a rapid and accurate technique, near infrared (NIR) spectroscopy has been widely used in process monitoring to improve efficiency and ensure quality consistency. In the present work, the application of NIR spectroscopy for on-line quantitative monitoring of the concentration process of Wangbi tablets was investigated. Partial least squares regression (PLSR) models based on the full-range spectra or the key wavelengths selected by the competitive adaptive reweighted sampling (CARS) method were established. The accuracy of the CARS-PLSR method was greater than that of full-spectrum PLSR for three quality parameters (soluble solid contents, paeoniflorin and icariin). For the prediction set samples using the CARS-PLSR models the coefficients of determination (r2) and root mean square error [RMSEP (%)] were 0.97 and 1.29 for soluble solid contents, 0.93 and 0.014 for paeoniflorin and 0.85 and 0.009 for icariin, respectively. Overall the results indicated that NIR spectroscopy coupled with CARS-PLSR calibration is a reliable and non-destructive alternative method for on-line monitoring of the concentration process of Wangbi tablets on an industrial scale.
Variable selection is widely accepted as an important step in quantitative analysis of near-infrared (NIR) spectroscopy. However, the variables preselected based on the calibration set might not be representative of the effective variables in future prediction process due to the large variability among soil sample sets. In this work, variable-updating methods (i.e., update both the model coefficients and effective variables in the prediction process) have been applied to support the robustness of the calibration model when it used to predict heterogeneous samples. Partial least squares regression (PLSR), recursive PLSR (RPLSR), and three variable-updating methods, namely variable importance in the projection combined with PLSR (VIP-PLSR), VIP-RPLSR, and uninformative variable elimination combined with PLSR (UVE-PLSR) were used to construct calibration models for the prediction of soil nitrogen (N) and organic carbon (OC) based on NIR spectroscopy. The entire data set was split into calibration set and prediction set according to soil type. The model VIP-RPLSR achieved the optimal performance for soil N and OC. The values of residual prediction deviation (RPD) were 2.9 and 2.8 for N and OC respectively. The results indicated that VIP-RPLSR was able to learn the information from the latest samples by adapting both model coefficients and effective variables at every sample interval. The proposed method VIP-RPLSR has the advantages of wider applicability and better performance for NIR prediction of soil N and OC in comparison with PLSR, RPLSR, VIP-PLSR and UVE-PLSR modeling techniques.
Soil available phosphorus (P) and available potassium (K) don't possess direct spectral response in the near infrared (NIR) region. They are predictable because of their correlation with spectrally active constituents (organic matter, carbonates, clays, water, etc.). Such correlation may of course differ between the soil sample sets. Therefore, the NIR calibration models with fixed structure are difficult to achieve good prediction performances for soil P and K. In this work, the method of recursive partial least squares (RPLS), which is able to update the model coefficients recursively during the prediction process, has been applied to improve the predictive abilities of calibration models. This work compared the performance of partial least squares regression (PLS), locally weighted PLS (LW-PLS), moving window LW-PLS (LW-PLS2) and RPLS for the measurement of soil P and K. The entire data set of 194 soil samples was split into calibration set and prediction set based on soil types. The calibration set was composed of 120 Anthrosols samples, while the prediction set included 29 Ferralsols samples, 23 Anthrosols samples and 22 Primarosols samples. The best prediction results were obtained by the RPLS model. The coefficient of determination (R2) and residual prediction deviation (RPD) were respectively 0.61, 0.76 and 1.60, 2.05 for soil P and K. The results indicate that RPLS is able to learn the information from the latest modeling sample by recursively updating the model coefficients. The proposed method RPLS has the advantages of wider applicability and better performance for NIR prediction of soil P and K compared with other methods in this work.
Ginkgo leaves are widely utilised in Chinese herbal medicines and functional food additives. However, the quality of ginkgo leaves fluctuates obviously due to the variety of geographical environments and climate conditions. Real time release testing (RTRT) combined with near infrared (NIR) spectroscopy was used to improve the quality control of ginkgo leaves. The RTRT of ginkgo leaves was achieved by qualitative and quantitative analysis using NIR spectroscopy and acceptable releasing criteria. Partial least squares regression models were developed for quantitative analysis of flavonol glycoside (FG), moisture and extract contents in ginkgo leaves. The coefficients of determination for leave-one-out cross-validation in calibration were 0.93, 0.92 and 0.89 for FG, moisture and extract contents, respectively, and relative standard errors of prediction were 9.01%, 6.67% and 3.22%, respectively. A discriminant analysis model was developed for qualitative analysis of ginkgo leaves. The Mahalanobis distance values were used as the qualitative releasing criteria of RTRT based on the discriminant analysis. In addition, FG content >= 0.7%, moisture content <= 12% and extract content >= 25% were used as the quantitative releasing criteria of RTRT according to the Chinese Pharmacopoeia. The accuracy of RTRT for ginkgo leaves was 86.7% according to qualitative and quantitative analyses based on NIR spectra. The results obtained in this work demonstrated that RTRT combined with NIR spectroscopy is a powerful tool for the quality control of ginkgo leaves.
Objective: To establish the quantitative models for analyzing the content of critical quality indicators in the purification process of Gardenia jasminoides intermediate in Reduning Injection using near-infrared (NIR) spectroscopy. Methods: The contents of shanzhiside, geniposidic acid, deacetyl asperulosidic acid methyl ester, genipin-1-β-D-gentiobioside, geniposide, chlorogenic acid, and total acid were determined by the reference method and NIR spectra were acquired. After removing the outliers, selecting the optimal spectral preprocessing method and selecting the best spectral wavelength, partial least squares (PLS) and the least squares support vector machines (LS-SVM) were used to build the models for predicting the contents of the above quality indicators in 18 unknown samples. Results: For shanzhiside, geniposidic acid, deacetyl asperulosidic acid methyl ester, genipin-1-β-D-gentiobioside, geniposide, chlorogenic acid, and total acid, the relative standard errors of prediction (RSEP) was lower than 3% for PLS models and LS-SVM models, indicating both methods could exhibit the satisfactory fitting results and predictive abilities. However, the LS-SVM models of shanzhiside and total acid showed lower predictive errors than PLS models. For geniposidic acid, deacetyl asperulosidic acid methyl ester, genipin-1-β-D-gentiobioside, geniposide, and chlorogenic acid, both models have the closer predictive errors. Conclusion: S-SVM shows better predictive performance than PLS. The established NIR quantitative models can be used for rapidly measuring the content of critical quality indicators in the purification process of G. jasminoides intermediate in Reduning Injection. ©, 2015, Editorial Office of Chinese Traditional and. All right reserved.
As the objective of simulation study, the samples of gardenia fructus extract were collected every 5 min while boiling for 60 min. Each sample was filtrated by different sizes of sieve meshes, and then scanned by the near infrared spectra. Using the HPLC determined value for reference, quantitative calibration models of geniposide were established by the partial least square method. Based on on-line near infrared spectral technology, the effects of by-pass preprocess system and filter mesh on the accuracy of calibration model were further studied. By the analysis of paired t test, the filtrated sample was compared with original extract, which illustrated that the filtrated sample and the original extract weren't significant different, so the sample after filtration could be representative of the original gardenia fructus extract for the data analysis. In addition, after the sample was filtrated by a sieve with meshes size of 0.049 mm, quantitative prediction model, which was built by raw spectrum pretreatment, could reach the ideal prediction result (Root Mean Square Error of Cross Validation (RMSECV) was 0.1962, Root Mean Square Error of Prediction (RMSEP) was 0.1867, Bias was 0.0256). It was recommended that proper size of sieve mesh could be selected by the simulation study, and then the feasibility of sample pretreatment system could be verified on-line, finally the stable preprocess system could be established. These steps could improve the performance of prediction models dramatically.
The preparation and electrochemical characterization of a carbon paste electrode modified with layered birnessite-type manganese oxide for use as a sodium sensor is described. The effects of powder synthesis process (sol-gel and redox precipitation) for birnessite on the electrochemical activity of the sensor was investigated by cyclic voltammetry. The carbon paste electrode modified with birnessite-type manganese oxide that was synthesized by the sol-gel method showed a best electrochemical for sodium ions. The detection is based on the measurement of anodic current generated by oxidation of Mn(III) to Mn(IV) at the surface of the electrode and consequently the sodium ions extraction into the birnessite structure. The best voltammetric response was obtained for an electrode composition of 15% (w/w) birnessite oxide in the paste, a TRIS buffer solution of pH 8.0 and a scan rate of 50 mV s(-1). A sensitive linear voltammetric response for sodium ions was obtained in the concentration range of 7.89x10(-5) to 3.49x10(-4) mol L(-1) with a slope of 37.5 microA L mmol(-1) and a detection limit (3sigma/slope) of 3.43x10(-5) mol L(-1) using cyclic voltammetry. Under the working conditions, the proposed method was successfully applied to determination of sodium ions in urine samples.
Fourier transform infrared (FTIR) spectroscopy has been used by chemists as a powerful tool to characterize inorganic and organic compounds. In this study, we examine the potential of FTIR microspectroscopy for early evaluation of therapy efficiency. For this purpose, we examine the effect of acyclovir (a known antiherpetic drug) on the development of herpes simplex virus type 1 (HSV-1) infection in cell culture. Also, we examine spectral changes in lymphocytes obtained from leukemia patients after appropriate chemotherapy treatment. Our results show early and significant spectral indicators for successful infection of Vero cells with HSV-1. Treatment of these infected cells with increasing doses of acyclovir reduces clearly the spectral changes caused by the infection in a correlation with inhibiting the development of the cytopathic effect (CPE) induced by this virus. Also significant and consistent spectral differences between lymphocytes from human leukemia patients compared to that from healthy persons are obtained. Treatment of these leukemia patients with appropriate drugs reduces significantly these spectral differences in a correlation with the improvement of the patient's clinical situation. It seems that FTIR spectroscopy can be used as an effective tool for early evaluation of the efficiency of drugs.
The application of near infrared spectroscopy in bioprocessing has been limited by its dependence on calibrations derived from single bioreactor at a given time. Here, we propose a multiplexed calibration technique which allows calibrations to be built from multiple bioreactors run in parallel. This gives the flexibility to monitor multiple vessels and facilitates calibration model transfer between bioreactors. Models have been developed for the two key analytes: glucose and lactate using Chinese hamster ovary (CHO) cell lines and using analyte specific information obtained from the feasibility studies. We observe slight model degradation for the multiplexed models in comparison to the conventional (single probe) models, decrease in r(2) values from 89.4% to 88% for glucose whereas for lactate from 92% to 91.8% and a simultaneous increase in the number of factors as the model incorporates the inter-probe variability, nevertheless the models were fit for purpose. The results of this particular application of implementing multiplexed-NIRS to monitor multiple bioreactor vessels are very encouraging, as successful models have been built on-line and validated externally, which proffers the prospect of reducing timelines in monitoring the vessels considerably, and in turn, providing improved control.
Water content and granule size are recognized as critical process and product quality parameters during drying. The purpose of this study was to enlighten the granule behavior during fluid bed drying by monitoring the major events i.e. changes in water content and granule size.
NIR spectra collected during drying and water content of sampled granules were correlated by principal component analysis (PCA) and partial least squares regression (PLSR). NIR spectra of dried granules were correlated to median granule size in a second PCA and PLSR.
The NIR water model discriminates between various stages in fluid-bed drying. The water content can be continuously predicted with errors comparable to the reference method. The four PLS factors of the granule size model are related to primary particle size of lactose, median granule size exceeding primary particle size and amorphous content of granules. The small prediction errors enable size discrimination between fines and granules.
For product quality reasons, discrimination between drying stages and end-point monitoring is highly important. Together with the possibilities to determine median granule size and to distinguish fines this approach provides a tool to design an optimal drying process.
Objective: To research and develop a model of the representative active component's content by NIR spectroscopy, so as to realize the on-line quality control of extracting process for multiple herbal medicine system in product scale. Methods: The on-line NIR detection of extracting process was used to obtain the NIR spectrum, HPLC detection of the extracts was carried out to determine the content of danshensu, and PLS method was used to establish the relationship between the information of NIR and HPLC. Results: The optimum NIR wavelength range was 9 715-7 082 cm-1, R=0.9594, RMSEC= 0.0494, the average relative error was 7.2%. Conclusion: NIR Technique could be used in the on-line quality control of Compound Danshen's extracting process.
The conductive polyaniline doped with instu decorating nano-silica was continue prepared by pulsed laser ablation, the polyaniline/nano-silica hybrid film was characterized by four probe stander conductivity test, TEM, UV-Vis, TG, XRD, XPS. discussing the thermal and conductivity of hybrid films was affect by different content of nano-silica. Get the conclusion that:nano silica prepared by laser ablation have a small size, not aggregating, and well disperseing in the hybrid. there is strong interaction between nano-silica and polyaniline, which have destroyed the order of polyaniline molecular chain and decreased the concentration and the mobility rate of carrier. Decreasing the conductivity;but get a more well oxidation resistance.
OBJECTIVE: To Study the relationship between on-line NIR spectrum information and off-line HPLC information of the product during water extracting process of Salvia miltiorrhiza Bunge. METHOD: On-line NIR detection of water extracting process of Salvia miltiorrhiza Bunge was simulated in the laboratory to obtain the NIR spectrum information, HPLC detection of the extracts was carried out to determine the contents of salvianolic acid B and tanshinone II A as the representative active components. PLS method was used to establish the relationship between the information of NIR and HPLC. RESULTS: The optimum NIR wavelength range for establishing the calibration model was 1300-1 600 nm and 2 200-2 400 nm. For tanshinone II A, r2 = 0.942 7, SEC = 0.917 7, the largest absolute predication error was 1.40%; For salvianolic acid B, r 2 = 0.914 3, SEC = 1.121 2, the largest absolute predication error was 3.08%. CONCLUSION: NIR technique can be used in the on-line detection and quality control of the exacting process of TCM, and can provide analytic results that satisfy precision requirement of industrial production.
In attempting to analyze, on digital computers, data from basically continuous physical experiments, numerical methods of performing familiar operations must be developed. The operations of differentiation and filtering are especially important both as an end in themselves, and as a prelude to further treatment of the data. Numerical counterparts of analog devices that perform these operations, such as RC filters, are often considered. However, the method of least squares may be used without additional computational complexity and with considerable improvement in the information obtained. The least squares calculations may be carried out in the computer by convolution of the data points with properly chosen sets of integers. These sets of integers and their normalizing factors are described and their use is illustrated in spectroscopic applications. The computer programs required are relatively simple. Two examples are presented as subroutines in the FORTRAN language.
Near-infrared (NIR) reflection spectroscopy was used for in-line monitoring of the conversion and the thickness of thin UV-cured acrylate coatings applied to polymer foils. Quantitative analysis of the spectroscopic data was performed either with the aid of PLS-based chemometric models or by band integration according to the Beer–Lambert law. Unintended changes of the thickness of the coating, e.g. caused by variation of the web speed, were found to preclude the correct analysis of the conversion by chemometric methods. In order to correct the conversion data for such changes, NIR spectra were recorded before and after UV irradiation. The conversion was determined from the ratio of the band integrals of the overtone of the acrylic double bond at 1620nm. It was shown that quantitative conversion data with high precision were achieved in this way. The method was used for in-line monitoring of the conversion in clear and pigmented coatings, which were applied to OPP foil by roll coating at line speeds up to 120mmin−1.
Near-infrared (NIR) spectroscopy has been utilized to demonstrate its feasibility for the measurement of major components in the acetic acid process. In order to simulate the acetic acid process, synthetic mixtures were prepared from five different components: acetic acid, methyl acetate, methyl iodide, water, and potassium iodide. Partial least squares (PLS) regression was utilized to differentiate the spectral characteristics as well as to quantify each component for the mixtures. The spectral features of acetic acid, methyl acetate, methyl iodide, and water are noticeably different with each other over the entire NIR region. The quantity of iodide ion, which does not absorb NIR radiation, was determined using the wavelength shift and intensity change of water absorption band caused by the change of iodide ion concentration. The PLS calibration results of the five components show good correlation with reference data. They also demonstrate the technical feasibility of NIR spectroscopy for monitoring important components in the acetic acid process.
Quantitative spectrometric analysis of mixture components is featured for systems with low spectral selectivity, namely, in the ultraviolet, visible, and infrared spectral range. Limitations imposed by data reduction schemes based on ordinary multiple regression are shown to be overcome by means of partial least-squares analysis in latent variables. The influences of variables such as noise, band separation band intensity ratios, number of wavelengths, number of components, number of calibration mixtures, time drift, or deviations from Beer's law on the analytical result has been evaluated under a wide range of conditions providing a basis to search for new systems applicable to spectrophotometric multicomponent analysis. The practical utility of the method is demonstrated for simultaneous analysis of copper, nickel, cobalt, iron, and palladium down to 2 X 10â»â¶ M concentrations by use of their diethyldithiocarbamate chelate complexes with relative errors less than 6%. 26 references, 4 figures, 6 tables.
Root mean square groupsizes computed from Mahalanobis distances were investigated as a means of determining, before the calibration step is executed, the proper sample preparation to use for performing near-infrared reflectance analysis. This multivariate mathematical technique appears to be a viable and useful method of predetermining which of several sample preparation methods to use. The advantage of this technique is that it can be applied to optical data alone, without the need to measure reference laboratory values and perform calibration calculations before determining how the samples should be prepared.
IntroductionQuantitative AnalysisQualitative AnalysisSignal ProcessingNew DevelopmentsReferences
A reversed phase high performance liquid chromatography method was established for the simultaneous determination of eight
major constituents, namely gallic acid, paeoniflorin sulfonate, catechin, albiflorin, paeoniflorin, benzoic acid, pentagalloylglucose
and benzoylpaeoniflorin in red and white peony roots, the two commonly used traditional Chinese medicinal herbs. The optimal
conditions of separation and detection were achieved on a C18 analytical column with a gradient mobile phase consisting of acetonitrile and 0.015% phosphoric acid at the flow rate of
1.0 mL min−1 and detection wavelength set at 230 nm. All calibration curves showed good linear regression (r>0.9995) within test ranges.
This method provided good reproducibility with overall intra-and inter-day precision of less than 5% and 4% and good accuracy
with recovery of more than 93%, respectively. The method was successfully applied to determine 71 samples of red and white
peony roots collected from different areas. The results indicated that the contents of eight compounds varied significantly
among the samples determined, which mainly resulted from processing procedure and habitat variation. The roots of Paeonia veitchii Lynch. contained a much higher amount of gallic acid and pentagalloylglucose than that of Paeonia lactiflora Pall., which could be used to distinguish the two similar species.
The particle size distribution of a solid product can be crucial parameter considering its application to different kinds of processes. The influence of particle size on near infrared (NIR) spectra has been used to develop effective alternative methods to traditional ones in order to determine this parameter. In this work, we used the chemometrical techniques partial least squares 2 (PLS2) and artificial neural networks (ANNs) to simultaneously predict several variables to the rapid construction of particle size distribution curves. The PLS2 algorithm relies on linear relations between variables, while the ANN technique can model non-linear systems.Samples were passed through sieves of different sieve opening in order to separate several size fractions that were used to construct two types of particle size distribution curves. The samples were recorded by NIR and their spectra were used with PLS2 and ANN to develop two calibration models for each. The correlation coefficients and relative standard errors of prediction (RSEP) have been used to assess the goodness of fit and accuracy of the results.The four calibration models studied provided statistically identical results based on RSEP values. Therefore, the combined use of NIR spectroscopy and PLS2 or ANN calibration models allows determining the particle size distributions accurately. The results obtained by ANN or PLS2 are statistically similar.
Multivariate calibration models are of critical importance to many analytical measurements, particularly for spectroscopic data. Generally, considerable effort is placed into constructing a robust model since it is meant to be used for extended periods of time. A problem arises, though, when the samples to be predicted are measured on a different instrument or under differing environmental factors from those used to build the model. The changes in spectral variations between the two conditions may make the model invalid for prediction in the new system. Various standardization and preprocessing methods have been developed to enable a calibration model to be effectively transferred between two systems, thus eliminating the need for a full recalibration. This paper presents an overview of the different methods used for calibration transfer and a critical assessment of their validity and applicability. The focus is on methods for transfer of near-infrared (NIR) spectra.
Bayberry plays an important role in the nutrition and is a very important fruit-product. It has a high economic and officinal value. In this study, glucose, fructose and sucrose in bayberry juice were detected and quantified using near-infrared (NIR) spectroscopy. The HPLC method was assumed to provide the reference value of the analyte for calibration. Partial least-squares regression (PLSR) was used to construct calibration models with different pre-processing methods. The number of PLS factors was optimised. The results show PLS models are good for predicting glucose, fructose and sucrose concentrations in bayberry juice, and their prediction accuracy can be improved by using derivative process with the exception of the glucose. The best models were mostly given by the second derivative processed spectra, especially for sucrose with the determination coefficient, R2 of 0.9931. This demonstrates the potential of NIR spectroscopy to quickly detect these components simultaneously in bayberry juice with the reference method of HPLC.
The frequent non-linearity of the calibration models used in infrared reflectance spectroscopy (NIRSS) is the main source of large errors in analyte determinations with this technique. Non-linearity in this type of system arises from factors such as the multiplicative effect of differences in particle size among samples or an intrinsically non-linear absorbance–concentration relationship resulting from interactions between components, hydrogen bonding, etc. In this work, calibration methods including partial least-squares (PLS) regression, linear quadratic PLS (LQ-PLS), quadratic PLS (QPLS) and artificial neural networks (ANNs) were used in conjunction with the NIRRS technique to determine the moisture content of acrylic fibres, the wide variability in linear density of which results in differential multiplicative effects among samples. Based on the results, PC-ANN is the best choice for the intended application. However, the joint use of an effective spectral pretreatment and computational methods such as PLS and LQ-PLS, the optimization of which is much less labour-intensive, provides comparable results. Standard normal variate (SNV) was found to be the best of the spectral pretreatments compared with a view to reducing the non-linearity introduced by scattering. The subsequent application of PLS provides accurate results with linear systems (absorption band at 1450 nm). A non-linear calibration model must be applied instead, however, if the system concerned is intrinsically non-linear. Under these conditions, the three methods tested for this purpose (LQ-PLS, QPLS and ANN) provide comparable results.
A novel approach of dynamic graphic software watermark is proposed. In this scheme, many fake watermarks created through encoding multi-constant, which's structure is similar to the only true one's, are introduced to enhance the stealth and anti-attack ability of dynamic graph watermark and keep software from sabotage. Meanwhile, a detailed analysis in theory is made in terms of the principle, feasibility and merits of this approach. A series of tests are made on the basis of the prototype system, analyzing the two sub-systems separately about their validity, robustness and performance overload caused by watermark embedding. Moreover, bit-rate of the three graph watermark structures in the system is analyzed in theory.
The application of near-infrared (NIR) spectroscopy for in-line monitoring of extraction process of scutellarein from Erigeron breviscapus (vant.) Hand-Mazz was investigated. For NIR measurements, two fiber optic probes designed to transmit NIR radiation through a 2 mm pathlength flow cell were utilized to collect spectra in real-time. High performance liquid chromatography (HPLC) was used as a reference method to determine scutellarein in extract solution. Partial least squares regression (PLSR) calibration model of Savitzky-Golay smoothing NIR spectra in the 5450-10,000 cm(-1) region gave satisfactory predictive results for scutellarein. The results showed that the correlation coefficients of calibration and cross validation were 0.9967 and 0.9811, respectively, and the root mean square error of calibration and cross validation were 0.044 and 0.105, respectively. Furthermore, both the moving block standard deviation (MBSD) method and conformity test were used to identify the end point of extraction process, providing real-time data and instant feedback about the extraction course. The results obtained in this study indicated that the NIR spectroscopy technique provides an efficient and environmentally friendly approach for fast determination of scutellarein and end point control of extraction process.
A method for rapid quantitative analysis of four kinds of Tanreqing injection intermediates was developed based on Fourier transform near infrared (FT-NIR) spectroscopy and partial least squares (PLS) algorithm. The NIR spectra of 120 samples were collected in transflective mode. The concentrations of chlorogenic acid, caffeic acid, luteoloside, baicalin, ursodesoxycholic acid (UDCA), and chenodeoxycholic acid (CDCA) were determined with the HPLC-DAD/ELSD as reference method. In the PLS calibration, the NIR spectra were pretreated with different methods and the number of PLS factors used in the model calibration was optimized by leave-one-out cross-validation. The performance of the final PLS models was evaluated according to the root mean square error of calibration (RMSEC), root mean square error of cross-validation (RMSECV), root mean square error of prediction (RMSEP), BIAS, standard error of prediction (SEP), and correlation coefficients (R). The R values in the prediction sets were all higher than 0.93, and the SEPs for the 6 compounds are 1.18, 6.02, 2.71, 155, 126, 30.0mg/l, respectively. The established models were used for the liquid preparation process analysis of Tanreqing injection in three batches, and a model updating method was proposed for the long-term usage of the established models. This work demonstrated that NIR spectroscopy is more rapid and convenient than the conventional methods to analyze the intermediates of Tanreqing injection, and the presented method is helpful to the implementation of process analytical technology (PAT) in pharmaceutical industry of Chinese Medicines Injections.
A rapid method for simultaneous determination of main phenolic acids in Radix Salvia Miltrorrhiza extract solutions was developed using Fourier transform near infrared spectroscopy in transflective mode and multivariate calibration and HPLC-UV as the reference method. Partial least squares (PLS) algorithm was conducted on the calibration of regression models. The multiplicative scatter correction, Norris derivative and second derivative were adopted for the spectral pre-processing, and the number of PLS factors were optimized by leave-one-out cross-validation. The performance of the final model was evaluated according to root mean square error of prediction (RMSEP) and correlation coefficient (R). The R values achieved in the prediction set were above 0.93. The developed models were used for analysis of unknown samples and routine monitoring with satisfactory results. This work demonstrated that NIR spectroscopy combined with PLS algorithm could be used for the rapid determination of the main phenolic acids of Salvia Miltrorrhiza extract solutions.
A licensed pharmaceutical process is required to be executed within the validated ranges throughout the lifetime of product manufacturing. Changes to the process, especially for processes involving biological products, usually require the manufacturer to demonstrate that the safety and efficacy of the product remains unchanged by new or additional clinical testing. Recent changes in the regulations for pharmaceutical processing allow broader ranges of process settings to be submitted for regulatory approval, the so-called process design space, which means that a manufacturer can optimize his process within the submitted ranges after the product has entered the market, which allows flexible processes. In this article, the applicability of this concept of the process design space is investigated for the cultivation process step for a vaccine against whooping cough disease. An experimental design (DoE) is applied to investigate the ranges of critical process parameters that still result in a product that meets specifications. The on-line process data, including near infrared spectroscopy, are used to build a descriptive model of the processes used in the experimental design. Finally, the data of all processes are integrated in a multivariate batch monitoring model that represents the investigated process design space. This article demonstrates how the general principles of PAT and process design space can be applied for an undefined biological product such as a whole cell vaccine. The approach chosen for model development described here, allows on line monitoring and control of cultivation batches in order to assure in real time that a process is running within the process design space.
To explore a method on-line determining the contents of tanshinone II A in returning extraction process of Salvia miltiorrhiza by AOTF-Near infrared spectroscopy.
Firstly, the sample was collected in the extraction process of Salvia miltiorrhiza. Then HPLC was used as a reference method to determine the contents of tanshinone II A in the sample. Multivariate calibration models based on PLS1 algorithm were developed to correlate the spectra and the corresponding values determined by the reference methods.
RMSECV of the models for tanshinone II A was 0.0092. The correlation coefficients of the calibration models was 0.9918. External validation with external validation samples proved that the relative deviation was 5.74%.
AOTF-NIRS can be used in the determination of tanshinone II A in returning extraction process of Salvia miltiorrhiza. The method is rapid, accurate and non-destructive, and can be applied for process analysis and quality control of Chinese medicine manufacturing process.
The objective of this study was to develop an integrated process monitoring approach for evaluating powder blending process kinetics and determining blending process end-point. A mixture design was created to include 26 powder formulations consisting of ibuprofen as the model drug and three excipients (HPMC, MCC, and Eudragit L100-55). The mixer was stopped at various time points to enable near-infrared spectroscopy scan of the powder mixture for obtaining the time course of the blending process. The evaluation of the blending process kinetics and process end-point was studied through three quantitative approaches: (1) Spectra linear superposition method; (2) Characteristic peak method; (3) Moving block standard deviation method. It was found that the powder blending experienced an initial rapid process to reach a quasi- end point within the first few minutes. Afterwards, a demixing occurred. Then, a real blending end-point was reached as characterized by an inflection point. ANOVA shows that the compositions of ibuprofen and MCC are the most statistically significant variables that impact the time required to reach the blending end-point. This highlighted the critical importance of developing quantitative chemometric approaches to extract critical process information and generate essential process knowledge to enable real-time release of the blending process.
In present work, we investigated the feasibility of universal calibration models for moisture content determination of a much complicated products system of powder injections to simulate the process of building universal models for drug preparations with same INN (International Nonproprietary Name) from diverse formulations and sources. We also extended the applicability of universal model by model updating and calibration transfer. Firstly, a moisture content quantitative model for ceftriaxone sodium for injection was developed, the results show that calibration model established for products of some manufacturers is also available for the products of others. Then, we further constructed a multiplex calibration model for seven cephalosporins for injection ranging from 0.40 to 9.90%, yielding RMSECV and RMSEP of 0.283 and 0.261, respectively. However, this multiplex model could not predict samples of another cephalosporin (ceftezole sodium) and one penicillins (penicillin G procaine) for injection accurately. With regard to such limits and the extension of universal models, two solutions are proposed: model updating (MU) and calibration transfer. Overall, model updating is a robust method for the analytical problem under consideration. When timely model updating is impractical, piecewise direct standardization (PDS) algorithm is more desirable and applied to transfer calibration model between different powder injections. Both two solutions have proven to be effective to extend the applicability of original universal models for the new products emerging.
Obtaining completely uniform distribution of the active principle and excipients in a pharmaceutical preparation is essential with a view to ensuring correct dosage. Uniformity in pharmaceutical formulations has usually been controlled by collecting samples at different stages of the process in order to determine the active principle using a chromatographic or UV-Visible spectroscopic method. In this work, near infrared reflectance spectroscopy (NIRS) was used to monitor blending in order to ensure uniformity in a mixture consisting of three typical pharmaceutical excipients and one active principle. To this end, a method for calculating the Mean Square of Differences between two consecutive spectra was developed with a view to expeditiously identifying the time mixture homogeneity was attained. The performance of the proposed method was compared with that of two others routinely employed to monitor blending by use of the NIRS technique and the results were found to be quite consistent.
Quantification analysis with near-infrared (NIR) spectroscopy typically requires utilizing chemometric techniques, such as partial least squares (PLS) method, to achieve the desired selectivity. This article points out a major limitation of these statistical-based calibration methods. The limitation is that the techniques suffer from the potential for chance correlation. In this article, the impact of chance correlation on the robustness of PLS model was illustrated via a pharmaceutical application with NIR to the content uniformity determination of tablets. The procedure involves evaluating the PLS models generated with two sets of calibration tablets incorporated with distinct degree of concentration correlation between the active pharmaceutical ingredient (API) and excipients. The selectivity and robustness of the two models were examined by using a series of data sets associated with placebo tablets and tablets incorporated with variations from excipient content, hardness and particle size. The result clearly revealed that the strong correlation observed in the PLS model created by the correlated design was not solely based on the API information, and there was an intrinsic difference in the variances described by the two calibration models. Diagnostic tools that enable the characterization of the chemical selectivity of the calibration model were also proposed for pharmaceutical quantitative analysis.
The feasibility of rapid analysis for oligosaccharides, including isomaltose, isomaltotriose, maltose, and panose, in Chinese rice wine by Fourier transform near-infrared (FT-NIR) spectroscopy together with partial least-squares regression (PLSR) was studied in this work. Forty samples of five brewing years (1996, 1998, 2001, 2003, and 2005) were analyzed by NIR transmission spectroscopy with seven optical path lengths (0.5, 1, 1.5, 2, 2.5, 3, and 5 mm) between 800 and 2500 nm. Calibration models were established by PLSR with full cross-validation and using high-performance anion-exchange chromatography coupled with pulsed amperometric detection as a reference method. The optimal models were obtained through wavelength selection, in which the correlation coefficients of calibration (r(cal)) for the four sugars were 0.911, 0.938, 0.925, and 0.966, and the root-mean-square errors of calibrations were 0.157, 0.147, 0.358, and 0.355 g/L, respectively. The validation accuracy of the four models, with correlation coefficients of cross-validation (r(cv)) being 0.718, 0.793, 0.681, and 0.873, were not very satisfactory. This might be due to the low concentrations of the four sugars in Chinese rice wine and the influence of some components having structures similar to those of the four sugars. The results obtained in this study indicated that the NIR spectroscopy technique offers screening capability for isomaltose, isomaltotriose, maltose, and panose in Chinese rice wine. Further studies with a larger Chinese rice wine sample should be done to improve the specificity, prediction accuracy, and robustness of the models.
The application of NIR in-line to monitor and control fermentation processes was investigated. Determination of biomass, glucose, and lactic and acetic acids during fermentations of Staphylococcus xylosus ES13 was performed by an interactance fiber optic probe immersed into the culture broth and connected to a NIR instrument. Partial least squares regression (PLSR) calibration models of second derivative NIR spectra in the 700-1800 nm region gave satisfactory predictive models for all parameters of interest: biomass, glucose, and lactic and acetic acids. Batch, repeated batch, and continuous fermentations were monitored and automatically controlled by interfacing the NIR to the bioreactor control unit. The high frequency of data collection permitted an accurate study of the kinetics, supplying lots of data that describe the cultural broth composition and strengthen statistical analysis. Comparison of spectra collected throughout fermentation runs of S. xylosus ES13, Lactobacillus fermentum ES15, and Streptococcus thermophylus ES17 demonstrated the successful extension of a unique calibration model, developed for S. xylosus ES13, to other strains that were differently shaped but growing in the same medium and fermentation conditions. NIR in-line was so versatile as to measure several biochemical parameters of different bacteria by means of slightly adapted models, avoiding a separate calibration for each strain.
FT-Raman spectroscopy (in combination with a fibre optic probe) was evaluated as an in-line tool to monitor a blending process of diltiazem hydrochloride pellets and paraffinic wax beads. The mean square of differences (MSD) between two consecutive spectra was used to identify the time required to obtain a homogeneous mixture. A traditional end-sampling thief probe was used to collect samples, followed by HPLC analysis to verify the Raman data. Large variations were seen in the FT-Raman spectra logged during the initial minutes of the blending process using a binary mixture (ratio: 50/50, w/w) of diltiazem pellets and paraffinic wax beads (particle size: 800-1200 microm). The MSD-profiles showed that a homogeneous mixture was obtained after about 15 min blending. HPLC analysis confirmed these observations. The Raman data showed that the mixing kinetics depended on the particle size of the material and on the mixing speed. The results of this study proved that FT-Raman spectroscopy can be successfully implemented as an in-line monitoring tool for blending processes.
High-speed counter-current chromatography was successfully used for the first time for the preparative separation and purification of paeoniflorin from the Chinese medicinal plant Paeonia lactiflora Pall. using a two-phase solvent system composed of n-butanol-ethyl acetate-water (1:4:5, v/v) in a single run. From 160 mg of the crude sample containing 22.0% paeoniflorin, 33.2 mg of paeoniflorin was yielded at 98.2% purity as determined by HPLC analysis. The recovery of paeoniflorin was 94.3%.
Near-infrared (NIR) spectroscopy and imaging are fast and nondestructive analytical techniques that provide chemical and physical information of virtually any matrix. In combination with multivariate data analysis these two methods open many interesting perspectives for both qualitative and quantitative analysis. This review focuses on recent pharmaceutical NIR applications and covers (1) basic principles of NIR techniques including chemometric data processing, (2) regulatory issues, (3) raw material identification and qualification, (4) direct analysis of intact solid dosage forms, and (5) process monitoring and process control.
This paper developed a rapid method using near infrared spectroscopy (NIRS) to differentiate two species of cortex phellodendri (CP), cortex phellodendri chinensis (PCS) and cortex phellodendri amurensis (PAR), and to predict quantitatively the content of berberine and total alkaloid content in all cortex phellodendri samples. Three alkaloids, berberine, jatrorrhizine and palmatine were analyzed simultaneously with a Thermo ODS Hypersil column by gradient elution with a new mobile phase under high-performance liquid chromatography-diode array detection (HPLC-DAD). Berberine content determined by HPLC-DAD was exploited as a critical parameter for successful discrimination between them. Multiplicative scatter correction (MSC), second derivative and Savitsky-Golay (S.G.) were utilized together to correct the scattering effect and eliminate the baseline shift in all near infrared diffuse reflectance spectra as well as to enhance spectral features in order to give a better correlation with the results obtained by HPLC-DAD. With the use of principal component analysis (PCA), samples datasets were separated successfully into two different clusters corresponding to two species. Furthermore, a partial least squares (PLS) regression method was built on the correlation model. The results showed that the correlation coefficients of the prediction models were R=0.996 for the berberine and R=0.994 for total alkaloid content. The influences of water absorption bands present in the NIR spectra on the models were also investigated in order to explore the practicability of NIRS in routine use. The outcome showed that NIRS possibly acts as routine screening in the quality control of Chinese herbal medicine.
An improved method based on an ensemble of Monte Carlo uninformative variable elimination (EMCUVE) is presented for wavelength selection in multivariate calibration of spectral data. The proposed algorithm introduces Monte Carlo (MC) strategy to uninformative variable elimination-PLS (UVE-PLS) instead of leave-one-out strategy for estimating the contributions of each wavelength variable in the PLS model. In EMCUVE wavelength variables are evaluated by different Monte Carlo uninformative variable elimination (MCUVE) models. Moreover, a fusion of MCUVE and the vote rule can obtain an improvement over the original uninformative variable elimination method. Results obtained from simulated data and real data sets demonstrate that EMCUVE can properly carry out wavelength selection in the course of data analysis and improve predictive ability for multivariate calibration model.