This study aims to validate and develop applications for a novel impedance-based process analytical technology for monitoring the attributes of the product during the entire freeze-drying process (from pre-freezing and annealing to primary and then secondary drying). This measurement approach involves the application of foil electrodes, mounted externally to a conventional glass freeze-drying vial, and coupled to a high-impedance analyser. The location of the electrodes on the outside, rather than the inside of the vial, leads to a description of the technology as a through-vial impedance spectroscopy (TV-IS) technique. The principle observation from this approach is the interfacial-polarization process arising from the composite impedance of the glass wall and product interface. For a conventional glass vial (of wall thickness ~ 1 mm and cross sectional diameter ~ 22 mm) it was shown that the process is manifest within the frequency range 101 to 106 Hz, as a single, broad band peak which spans 2-3 decades of the imaginary part spectrum. Features of the interfacial-relaxation process, characterised by the peak amplitude, C″Peak, and peak frequency, fpeak, of the imaginary capacitance spectra and the equivalent circuit elements that model the impedance spectra (i.e. the solution resistance (R) and solution capacitance (C) were monitored along with the product temperature data during the cycle(s), for a variety of surrogate formulations comprising lactose, sucrose, mannitol or maltodextrin solutions, during the freezing, re-heating, annealing and primary drying stages of freeze drying). It was shown that the parameters, fpeak and R, are strongly coupled to each other and change as a function of the temperature of the solution and its phase state, whereas C″Peak is strongly coupled to the amount of ice that remains during the drying process. Both log fpeak and log R have a linear dependence on the temperature of the solution, provided there was no phase change in the solution. The crystallization process (ice onset, solidification and equilibration to shelf temperature) is characterized well by both log fpeak and log R, whereas the parameter R demonstrates most clearly the formation of eutectic crystallization during freezing. In contrast it was the parameter C which was most sensitive to the detection of the glass transition during re-heating. During primary drying, it was shown that C″peak, is dependent on the amount of ice remaining and therefore provides a convenient assessment of the rate of drying and primary drying end point. The impedance changes during annealing provide a mechanistic basis for the modifications in ice structure which result directly in the observed decrease in primary drying times. The principal observation on annealing of a 10% w/v solution of maltodextrin, was the minimal changes in the glass transition (recorded at ~−16 °C) during the re-heating and cooling step (post-annealing). This result alone appears to indicate that a maximum freeze concentration was achieved during first freezing with no further ice being formed on annealing. The phenomenon of devitrification (and the production of more ice, and hence larger ice crystals) was therefore discounted as the mechanism by which annealing impacts the drying time. Having excluded devitrification from the mechanism of annealing enhanced drying, it was then possible to conclude that the decrease in the electrical resistance (that was observed during the annealing hold time) must necessarily result from the simplified structure of the unfrozen fraction and the improved connectivity of ice crystals that may be the consequence of Ostwald ripening. The application of through vial impedance measurement approach provides a non-invasive, real time monitoring of critical process parameters which subsequently leads to an improved understanding of the mechanisms and effects of different parameters, providing a reliable basis for process optimization, along with improved risk management to ensure optimum quality of the formulation and optimization of the freeze drying process.

Spectroscopy and other optical methods can often be employed with limited or no sample preparation, making them well suited for in situ and in vivo analysis. This dissertation focuses on the use of a near-infrared spectroscopy (NIRS) and polarized light scatter for two such applications: the assessment of cardiovascular disease, and the validation of cleaning processes for pharmaceutical equipment.There is a need for more effective in vivo techniques for assessing intravascular disorders, such as aortic aneurysms and vulnerable atherosclerotic plaques. These, and other cardiovascular disorders, are often associated with structural remodeling of vascular walls. NIRS has previously been demonstrated as an effective technique for the analysis of intact biological samples. In this research, traditional NIRS is used in the analysis of aortic tissue samples from a murine knockout model that develops abdominal aortic aneurysms (AAAs) following infusion of angiotensin II. Effective application of NIRS in vivo, however, requires a departure from traditional instrumental principles. Toward this end, the groundwork for a fiber optic-based catheter system employing a novel optical encoding technique, termed molecular factor computing (MFC), was developed for differentiating cholesterol, collagen and elastin through intervening red blood cell solutions. In MFC, the transmission spectra of chemical compounds are used to collect measurements directly correlated to the desired sample information.Pharmaceutical cleaning validation is another field that can greatly benefit from novel analytical methods. Conventionally cleaning validation is accomplished through surface residue sampling followed by analysis using a traditional analytical method. Drawbacks to this approach include cost, analysis time, and uncertainties associated with the sampling and extraction methods. This research explores the development of in situ cleaning validation methods to eliminate these issues. The use of light scatter and polarization was investigated for the detection and quantification of surface residues. Although effective, the ability to discriminate between residues was not established with these techniques. With that aim in mind, the differentiation of surface residues using NIRS and MFC was also investigated.

Acoustic methods can often be used with limited or no sample preparations making them ideal for rapid process analytical technologies (PATs). This dissertation focuses on the possible use of acoustic resonance spectroscopy as a PAT in the pharmaceutical industry. Current good manufacturing processes (cGMP) need new technologies that have the ability to perform quality assurance testing on all products. ARS is a rapid and non destructive method that has been used to perform qualitative studies but has a major drawback when it comes to quantitative studies. Acoustic methods create highly non linear correlations which usually results in high level computations and chemometrics. Quantification studies including powder contamination levels, hydration amounts and active pharmaceutical ingredient (API) concentrations have been used to test the hypothesis that bootstrap enhanced n-dimensional deformation of space (BENDS) could be used to overcome the highly non linear correlations that occur with acoustic resonance spectroscopy (ARS) eliminating a major drawback with ARS to further promote the device as a possible process analytical technology (PAT) in the pharmaceutical industry. BENDS is an algorithm that has been created to calculate a reduced linear calibration model from highly non linear relationships with ARS spectra. ARS has been shown to correctly identify pharmaceutical tablets and with the incorporation of BENDS, determine the hydration amount of aspirin tablets, D-galactose contamination levels of Dtagatose powders and the D-tagatose concentrations in resveratrol/D-tagatose combinatory tablets.

Modern analytical instrumentation is capable of creating enormous and complex volumes of data. Analysis of large data volumes are complicated by lengthy analysis time and high computational demand. Incorporating real-time analysis methods that are computationally efficient are desirable for modern analytical methods to be fully utilized. The use of modern instrumentation in on-line pharmaceutical process validation, remote sensing, and astrobiology applications requires real-time analysis methods that are computationally efficient. Integrated sensing and processing (ISP) is a method for minimizing the data burden and sensing time of a system. ISP is accomplished through implementation of chemometric calculations in the physics of the spectroscopic sensor itself. In ISP, the measurements collected at the detector are weighted to directly correlate to the sample properties of interest. This method is especially useful for large and complex data sets. In this research, ISP is applied to acoustic resonance spectroscopy, near-infrared hyperspectral imaging and a novel solid state spectral imager. In each application ISP produced a clear advantage over the traditional sensing method. The limitations of ISP must be addressed before it can become widely used. ISP is essentially a pattern recognition algorithm. Problems arise in pattern recognition when the pattern-recognition algorithm encounters a sample unlike any in the original calibration set. This is termed the false sample problem. To address the false sample problem the Bootstrap Error-Adjusted Single-Sample Technique (BEST, a nonparametric classification technique) was investigated. The BEST-ISP method utilizes a hashtable of normalized BEST points along an asymmetric probability density contour to estimate the BEST multidimensional standard deviation of a sample. The on-line application of the BEST method requires significantly less computation than the full algorithm allowing it to be utilized in real time as sample data is obtained. This research tests the hypothesis that a BEST-ISP metric can be used to detect false samples with sensitivity > 90% and specificity > 90% on categorical data.

Innovations and process understanding is the current focus in pharmaceutical industry. The objective of this research was to explore application of high power ultrasound in the slurry crystallisation and application of low power ultrasound (3.5 MHz) as process analytical technology (PAT) tool to understand pharmaceutical processing such as hot melt extrusion. The effect of high power ultrasound (20 kHz) on slurry co-crystallisation of caffeine / maleic acid and carbamazepine / saccharin was investigated. To validate low power ultrasound monitoring technique, it was compared with the other techniques (PAT tools) such as in-line rheology and in-line NIR spectroscopy. In-line rheological measurements were used to understand melt flow behaviour of theophylline / Kollidon VA 64 system in the slit die attached to the hot melt extruder. In-line NIR spectroscopic measurements were carried out for monitoring any molecular interactions occurring during extrusion. Physical mixtures and the processed samples obtained from all experiments were characterised using powder X-ray diffraction, thermogravimetry analysis, differential scanning calorimetry, scanning Electron Microscopy, dielectric spectroscopy and high performance liquid chromatography, rotational rheology, fourier transform infrared spectroscopy and near infrared spectroscopy. The application of high power ultrasound in slurry co-crystallisation of caffeine / maleic acid helped in reducing equilibrium time required for co-crystal formation. During carbamazepine / saccharin co-crystallisation high power ultrasound induced degradation of carbamazepine was negligible. Low power ultrasound can be used as a PAT tool as it was found to be highly sensitive to the changes in processing temperatures and drug concentration.

La mise en place de la démarche PAT (Process Analytical Technology), initiée par la FDA (Food and Drug Administration) s’est développée au cours de ces dernières années, au sein de l’industrie pharmaceutique. Grâce à des contrôles en continu au coeur des procédés de fabrication, elle permet une meilleure compréhension et une maîtrise de la formulation et du procédé, afin d’assurer la qualité finale des médicaments.A travers ce travail, nous avons mis en place un suivi en temps réel, par spectroscopie proche infrarouge, d’une opération d’enrobage suite à l’intégration d’une sonde à l’intérieur d’une turbine d’enrobage. La quantité d’enrobage, déterminée par une simple et rapide pesée mais néanmoins soumise à la variabilité de la masse des comprimés nus, ainsi que l’épaisseur du film, obtenue avec précision par imagerie térahertz ont servi de valeurs de référence pour calibrer l’information spectrale. Dans les deux cas, ces deux attributs qualité critiques ont été prédits avec de faibles erreurs de prédiction, qui se sont révélées être similaires. Par ailleurs, la prédiction en temps réel des propriétés de dissolution de comprimés prêts à être libérés, à partir de spectres acquis in-line, a permis de déterminer l’arrêt optimal de l’opération d’enrobage.Suite à un enrobage réalisé à partir d’une dispersion aqueuse de polymère une étape supplémentaire de traitement thermique ou curing est généralement nécessaire afin de stabiliser le film d’enrobage. Un travail de caractérisation mené à partir de techniques innovantes a permis d’apporter un nouvel éclairage sur la compréhension des phénomènes impliqués dans la formation du film au cours du curing. La caractérisation approfondie de la structure d’enrobage de comprimés soumis à un curing en turbine (conditions dynamiques) a mis en évidence la diminution de la porosité, couplée à l’évaporation de l’eau et à une meilleure organisation des chaînes de polymère au cours du curing. L’étude de comprimés soumis à un curing de référence en étuve durant 24 h (conditions statiques) a confirmé l’obtention d’un film stable après 4 h de curing dynamique. De nouveaux phénomènes, indépendants du curing, liés à la cristallisation et à la migration de l’alcool cétylique, couplée à la migration du lauryl sulfate de sodium, au sein de la couche d’enrobage ont été détectés au cours de la conservation des comprimés enrobés. / Implementation of PAT (Process Analytical Technology) approach has recently been promoted by the FDA (Food and Drug Administration) within the pharmaceutical industry. A desired goal of the PAT framework is to enhance understanding and control of the manufacturing process through timely measurements, during processing, to ensure final product quality. Real-time monitoring of a coating operation was performed from in-line Near Infrared (NIR) measurements inside a pan coater. Mass of coating materials, determined by simple and fast weighing but depending on core tablet weight uniformity, and film coating thickness, obtained from accurate and non-destructive Terahertz Pulsed Imaging (TPI) measurements, were used as reference values to calibrate NIR spectral information. In both cases, these two critical quality attributes were predicted with low predictive errors, which were found to be similar. In addition, real-time predictions of drug release from cured tablets were carried out by in-line NIR measurements. The coating operation was successfully stopped when desired dissolution criteria were achieved. A post-coating thermal treatment, known as curing, is generally required to stabilize film coating from aqueous polymer dispersion. Innovative techniques were jointly used to elucidate the underlying mechanisms of film formation along the curing process. This study provided a new insight into the tablet coating structure, highlighting a reduced internal coating porosity, a decrease in water content and showing a better structural rearrangement of polymer chains, with dynamic curing. All investigated techniques confirmed that a stabilized state was reached after a 4 h dynamic curing in comparison with a reference curing carried out in an oven for 24 h. Interestingly even prior to curing, new findings were pointed out, during coated tablets storage, related to the crystallisation and the upward migration of cetyl alcohol, coupled to the downward migration of sodium lauryl sulfate within the coating layer.

Software sensors are a potent tool to improve biotechnological real time process monitoring and control. In the current project, algorithms for six partly novel, software sensors were established and tested in a microbial reactor system. Eight batch and two fed-batch runs were carried out with a recombinant Escherichia coli to investigate the suitability of the different software sensor models in diverse cultivation stages. Special respect was given to effects on the sensors after recombinant protein expression was initiated by addition of an inducer molecule. It was an objective to figure out influences of excessive recombinant protein expression on the software sensor signals. Two of the developed algorithms calculated the biomass on-line and estimated furthermore, the specific growth rate by integration of the biomass changes with the time. The principle of the first was the application of a near infrared probe to obtain on-line readings of the optical density. The other algorithm was founded on the titration of ammonia as only available nitrogen source. The other two sensors analyzed for the specific consumption of glucose and the specific production of acetate and are predicted on an in-line HPLC system. The results showed that all software sensors worked as expected and are rather powerful to estimate important state parameters in real time. In some stages, restrictions may occur due to different limitation affects in the models or the physiology of the culture. However, the results were very convincing and suggested the development of further and more advanced software sensor models in the future.

<p>Software sensors are a potent tool to improve biotechnological real time process monitoring and control. In the current project, algorithms for six partly novel, software sensors were established and tested in a microbial reactor system. Eight batch and two fed-batch runs were carried out with a recombinant <em>Escherichia coli</em> to investigate the suitability of the different software sensor models in diverse cultivation stages. Special respect was given to effects on the sensors after recombinant protein expression was initiated by addition of an inducer molecule. It was an objective to figure out influences of excessive recombinant protein expression on the software sensor signals.</p><p>Two of the developed algorithms calculated the biomass on-line and estimated furthermore, the specific growth rate by integration of the biomass changes with the time. The principle of the first was the application of a near infrared probe to obtain on-line readings of the optical density. The other algorithm was founded on the titration of ammonia as only available nitrogen source. The other two sensors analyzed for the specific consumption of glucose and the specific production of acetate and are predicted on an in-line HPLC system.</p><p>The results showed that all software sensors worked as expected and are rather powerful to estimate important state parameters in real time. In some stages, restrictions may occur due to different limitation affects in the models or the physiology of the culture. However, the results were very convincing and suggested the development of further and more advanced software sensor models in the future.</p>

Many areas of science now generate huge volumes of data that present visualization, modeling, and interpretation challenges. Methods for effectively representing the original data in a reduced coordinate space are therefore receiving much attention. The purpose of this research is to test the hypothesis that molecular computing of vectors for transformation matrices enables spectra to be represented in any arbitrary coordinate system. New coordinate systems are selected to reduce the dimensionality of the spectral hyperspace and simplify the mechanical/electrical/computational construction of a spectrometer. A novel integrated sensing and processing system, termed Molecular Factor Computing (MFC) based near infrared (NIR) spectrometer, is proposed in this dissertation. In an MFC -based NIR spectrometer, spectral features are encoded by the transmission spectrum of MFC filters which effectively compute the calibration function or the discriminant functions by weighing the signals received from a broad wavelength band. Compared with the conventional spectrometers, the novel NIR analyzer proposed in this work is orders of magnitude faster and more rugged than traditional spectroscopy instruments without sacrificing the accuracy that makes it an ideal analytical tool for process analysis. Two different MFC filter-generating algorithms are developed and tested for searching a near-infrared spectral library to select molecular filters for MFC-based spectroscopy. One using genetic algorithms coupled with predictive modeling methods to select MFC filters from a spectral library for quantitative prediction is firstly described. The second filter-generating algorithm designed to select MFC filters for qualitative classification purpose is then presented. The concept of molecular factor computing (MFC)-based predictive spectroscopy is demonstrated with quantitative analysis of ethanol-in-water mixtures in a MFC-based prototype instrument.

The U.S. Food and Drug Administration has accepted the guidelines put forth by the International Conference on Harmonization (ICH-Q8) that allow for operational flexibility within a validated design space. These Quality by Design initiatives have allowed drug manufacturers to incorporate more rigorous scientific controls into their production streams. <br>Fully automated control systems can incorporate information about a process back into the system to adjust process variables to consistently hit product quality targets (feedback control), or monitor variability in raw materials or intermediate products to adjust downstream manufacturing operations (feedforward control). These controls enable increased process understanding, continuous process and product improvement, assurance of product quality, and the possibility of real-time release. Control systems require significant planning and an initial investment, but the improved product quality and manufacturing efficiency provide ample incentive for the expense. <br>The fluid bed granulation and drying unit operation was an excellent case study for control systems implementation because it is a complex unit operation with dynamic powder movement, high energy input, solid-liquid-gas interactions, and difficulty with scale-up development. Traditionally, fluid bed control systems have either used first principle calculations to control the internal process environment or purely empirical methods that incorporate online process measurements with process models. This dissertation was predicated on the development of a novel hybrid control system that combines the two traditional approaches. <br>The hybrid controls reduced the number of input factors for the creation of efficient experimental designs, reduced the variability between batches, enabled control of the drying process for a sensitive active pharmaceutical ingredient, rendered preconditioned air systems unnecessary, and facilitated the collection of data for the development of process models and the rigorous calculation of design spaces. Significant variably in the inlet airstream was able to be mitigated using feedforward controls, while process analytical technology provided immediate feedback about the process for strict control of process inputs. Tolerance surfaces provided the ideal tool for determining design spaces that assured the reduction of manufacturing risk among all future batches, and the information gained using small scale experimentation was leveraged to provide efficient scale-up, making these control systems feasible for consistent use. / Mylan School of Pharmacy and the Graduate School of Pharmaceutical Sciences / Pharmaceutics / PhD / Dissertation