NIR spectroscopy for the in-line monitoring of a multicomponent formulation during the entire freeze-drying process.

Freeze drying is a complex, time consuming and thus expensive process, hence creating a need for understanding the material behaviour in the process environment and for process optimization. Near-infrared (NIR) spectroscopy offers the opportunity to monitor physicochemical changes of the formulation during freeze-drying. The aim of this work was to examine whether NIR spectroscopy allows in-line monitoring of all components during the entire freeze-drying process of a multi-component pharmaceutical formulation (a solution of fenofibrate and mannitol in a mixture of tertiary-butyl alcohol, and water). To extract useful information of all components in the formulation from the large multivariate data-sets obtained during in-line spectroscopic monitoring, several spectral pre-processing techniques and spectral data analysis techniques such as the mean of selected wavenumbers (Mws), the correlation coefficient (CorrCoef) and principal component analysis (PCA) have been evaluated and compared. To find out whether these chemometric techniques are also able to differentiate between changes in the process settings influencing the freeze-drying process of the formulation, freeze-drying processes were performed at four different conditions. Results demonstrated that in-line measurements using NIR spectroscopy were possible in an icy environment and that a further process understanding could be obtained. Data-analysis revealed the crystallization behaviour of each of the four components. In addition, using the three pre-processing techniques allowed observe the sublimation of the solvents. Mws and CorrCoef have proven to be adequate methods for monitoring the main physicochemical changes of product during the processes; this affirmation was confirmed by observing the outputs of PCA for entire processes.

Real-time determination of critical quality attributes using near-infrared spectroscopy: a contribution for Process Analytical Technology (PAT).

Process Analytical Technology (PAT) is playing a central role in current regulations on pharmaceutical production processes. Proper understanding of all operations and variables connecting the raw materials to end products is one of the keys to ensuring quality of the products and continuous improvement in their production. Near infrared spectroscopy (NIRS) has been successfully used to develop faster and non-invasive quantitative methods for real-time predicting critical quality attributes (CQA) of pharmaceutical granulates (API content, pH, moisture, flowability, angle of repose and particle size). NIR spectra have been acquired from the bin blender after granulation process in a non-classified area without the need of sample withdrawal. The methodology used for data acquisition, calibration modelling and method application in this context is relatively inexpensive and can be easily implemented by most pharmaceutical laboratories. For this purpose, Partial Least-Squares (PLS) algorithm was used to calculate multivariate calibration models, that provided acceptable Root Mean Square Error of Predictions (RMSEP) values (RMSEP(API)=1.0 mg/g; RMSEP(pH)=0.1; RMSEP(Moisture)=0.1%; RMSEP(Flowability)=0.6 g/s; RMSEP(Angle of repose)=1.7° and RMSEP(Particle size)=2.5%) that allowed the application for routine analyses of production batches. The proposed method affords quality assessment of end products and the determination of important parameters with a view to understanding production processes used by the pharmaceutical industry. As shown here, the NIRS technique is a highly suitable tool for Process Analytical Technologies.

Quality by design approach of a pharmaceutical gel manufacturing process, part 1: determination of the design space.

This work was conducted in the framework of a quality by design project involving the production of a pharmaceutical gel. Preliminary work included the identification of the quality target product profiles (QTPPs) from historical values for previously manufactured batches, as well as the critical quality attributes for the process (viscosity and pH), which were used to construct a D-optimal experimental design. The experimental design comprised 13 gel batches, three of which were replicates at the domain center intended to assess the reproducibility of the target process. The viscosity and pH models established exhibited very high linearity and negligible lack of fit (LOF). Thus, R(2) was 0.996 for viscosity and 0.975 for pH, and LOF was 0.53 for the former parameter and 0.84 for the latter. The process proved reproducible at the domain center. Water content and temperature were the most influential factors for viscosity, and water content and acid neutralized fraction were the most influential factors for pH. A desirability function was used to find the best compromise to optimize the QTPPs. The body of information was used to identify and define the design space for the process. A model capable of combining the two response variables into a single one was constructed to facilitate monitoring of the process.

Quality by design approach of a pharmaceutical gel manufacturing process, part 2: near infrared monitoring of composition and physical parameters.

We applied the principles of quality by design to the production process of a pharmaceutical gel by using the near infrared spectroscopy (NIRS) technique in combination with multivariate chemometric tools. For this purpose, we constructed a D-optimal experimental design having normal operational condition (NOC) batches as central point. The primary aim here was to develop an expeditious NIRS method for determining the composition of a pharmaceutical gel and assess the temporal changes in major physical factors affecting the quality of the product (specifically, viscosity and pH). Gel components were quantified by using partial least squares (PLS) calibration models of the PLS1 type. The study was completed by using the batch statistical process control method to compare product batches included in the experimental design with NOC batches. Similarities and differences between the two types of batches were identified by using control charts for residuals (Q-statistic) and Hotteling's T2 (D-statistic). The ensuing models, which were subject to errors less than 5%, allowed the gel production process to be effectively monitored. As shown in this work, the NIRS technique is a highly suitable tool for process analytical technology.

Nir-chemical imaging study of acetylsalicylic acid in commercial tablets.

Near Infrared Chemical Imaging (NIR-CI) is demonstrating an increasing interest in pharmaceutical research since it meets the challenging analytical needs of pharmaceutical quality and may serve as a versatile adjunct to conventional NIR spectroscopy in many fields. The direct analysis of samples by using hyperspectral imaging techniques, which provide a NIR spectrum in each pixel of the image, generates a big amount of information from one sample. Focusing the interest in pharmaceutical research, several chemometric algorithms are demonstrating their usefulness extracting the relevant information (i.e. quantitative determination of the component in one sample) in tablets with only one sample and without damaging it. In this work, a quantitative method to analyze different commercial Acetylsalicylic acid tablets is proposed by using Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS) method to the hyperspectral image and without any previous calibration model. For this purpose, a large concentration range of active pharmaceutical ingredient (ASA, Acetylsalicylic acid in this work), between 82% and 12%, was covered depending on the manufacturer. MCR-ALS allowed obtaining a concentration maps for acetylsalicylic acid and therefore, consequent analysis of the ASA distribution in the tablet was developed by using the histograms of the distribution of concentration. Results certified the good distribution of ASA despite the different origins of the tablets. Moreover, the obtained values of concentration showed a very good concordance with the nominal value of ASA. As a matter of fact, the quality of the results demonstrated the useful of encompassing NIR-CI techniques with MCR-ALS and, consequently, the well development on the production of Acetylsalicylic acid tablets.

Determination of low analyte concentrations by near-infrared spectroscopy: effect of spectral pretreatments and estimation of multivariate detection limits.

Near infrared spectroscopy (NIRS) was used in combination with partial least squares (PLS) calibration to determine low concentrated analytes. The effect of the orthogonal signal correction (OSC) and net analyte signal (NAS) pretreatments on the models obtained at concentrations of analyte near its detection limit was studied. Both pretreatments were found to accurately resolve the analyte signal and allow the construction of PLS models from a reduced number of factors; however, they provided no substantial advantage in terms of %RSE for the prediction samples. Multiple methodologies for the estimation of detection limits could be found in the bibliography. Nevertheless, detection limits were determined by a multivariate method based on the sample-specific standard error for PLS regression, and compared with the univariate method endorsed by ISO 11483. The two methods gave similar results, both being effective for the intended purpose of estimating detection limits for PLS models. Although OSC and NAS allow isolating the analyte signal from the matrix signal, they provide no substantial improvement in terms of detection limits. The proposed method was used to the determine 2-ethylhexanol at concentrations from 20 to 1600 ppm in an industrial ester. The detection limit obtained, round 100 ppm, testifies to the ability of NIR spectroscopy to detect low concentrated analytes.

Analytical monitoring of alcoholic fermentation using NIR spectroscopy.

Alcoholic fermentation under Saccharomyces cerevisiae yeasts is governed largely by glucose uptake, biomass formation, ethanol and glycerin production, and acidification. In this work, PLS calibration models were developed with a view to determining these analytical parameters from near infrared spectra and analytical data provided by the corresponding reference methods. The models were applied to a set of samples obtained from various fermentation processes. The glucose, ethanol, and biomass values predicted by the models exhibited a high correlation with those provided by the reference method.