Freeze drying is used to improve long-term stability of labile drugs. It comprises three different stages: freezing, primary drying, and secondary drying. Due to low values of pressure and temperature required in these stages, freeze drying is a cost-intensive process. Various authors have recently shown that the manipulation of the temperature of nucleation can provide benefits in terms of process optimization, and various technologies have been proposed to this purpose, such as electrofreezing, ultrasound, depressurization method, and ice fog. However, due to various problems of these technologies, such as the requirement of additional equipment and difficulties in scale-up, in this study the vacuum induced nucleation method was taken into consideration for its ease of application. Unfortunately, this method, as proposed in literature, produces defects in the cake structure because of blow-up of the frozen product and flake formation on the product surface, which are detrimental for the elegance of the final product and rejected by the pharmaceutical industry. For these reasons, a refined control technique has been here investigated and validated, showing that an accurate control of vacuum conditions can give a drastic reduction in cycle time along with an elegant product. The vacuum induced nucleation method consists in reducing the pressure within the drying chamber for a short time during the freezing stage. This pressure reduction produces the partial evaporation of water, which causes a reduction in product temperature and promotes the nucleation of ice. In this work I have developed a new method, still based on pressure reduction, that solved some of the problems of the original technique. In particular, once the desired value of pressure (which is product dependent) was reached, the drying chamber was isolated from the condenser for about 1 min. Because of that isolation, the pressure inside the drying chamber increased because of vapor accumulation, thus avoiding any esthetic defects to the product. A detailed investigation about the impact of the settings of the new control method on product and process performances was performed. To this purpose, various solutions containing mannitol, sucrose, lactose, dextran, glycine, polyvinylpyrrolidone, trehalose, and cellobiose were used. The results obtained showed that the method led to a high quality of the final product. Furthermore, showed that the induction of nucleation at higher temperature promoted the formation of very large crystals (see Fig. 1) with a consequent drastic reduction in cycle duration (due to lower values of product resistance to vapor flow, Rp). In order to promote the industrial scale-up of the method, a monitoring system based on thin-film technology, that allows, besides the monitoring of the product temperature, the detection of the end of the nucleation event was developed. Once the method was developed, a detailed investigation on its impact on both product (i.e., structure and homogeneity) and process (i.e., primary and secondary drying time) was carried out. The impact of the method on within-batch (inter-vial) and within-vial (intra-vial) variability was investigated. Comparing the onset-offset of pressure ratio curves of various cycles, it was found that the use of controlled nucleation (instead of conventional shelf-ramped freezing) dramatically increased the inter-vial homogeneity (by about 50%). This result was also confirmed by comparing the residual moisture values at the end of drying, which showed lower variance for controlled nucleation with respect to uncontrolled. A study, focused on mannitol-based formulations, showed that both nucleation temperature and filling volume had an impact on pore size and its distribution along the lyophilized cake. This within-vial variability was quantified by analyzing the average pore dimension along the cake (by Image-J program). The homogeneity of the product was also studied in terms of polymorph formation. To this regard, X-ray diffraction and off-line Raman were used to identify the mannitol polymorphs present within the product, which was found to vary with the freezing protocol and the temperature of secondary drying. Off-line Raman spectroscopy was also used to evaluate the polymorph distribution along the cake, showing that the spectra collected at the top of each sample often contained slightly higher amounts of α and δ mannitol in comparison with the spectra recorded in the other regions of the same cake. Finally, the Raman tool was also used for the in-line monitoring of the product characteristics, increasing the knowledge related to polymorph formation during freeze drying. To this purpose, it was found that the hydrate form of mannitol (which can damage the product if released during product storage) was formed during the entire cycle (freezing and drying stages), but was transformed into anhydrous polymorph immediately after the vacuum within the chamber was released at the end of the cycle. The primary drying stage of the cycles was performed using in-line monitoring (DPE+ algorithm) and control (LyoDriver) to further optimize the freeze-drying cycle. The use of these tools showed that, to obtain the optimum freeze-drying cycle, it was necessary to work on both freezing and primary drying stages. The mass-transfer resistance values obtained from each cycle (by DPE+ algorithm estimation) were used to calculate the design space and simulate the process. The cycle simulation highlighted the reduction of the cycle duration using the control method proposed (in comparison with an uncontrolled stochastic cycles) and the influence of its operating conditions on freeze-drying cycle optimization. It was, finally, shown that the primary drying duration could be decreased up to about 55% in the case of controlled nucleation. A study focused on the moisture content of the product during secondary drying showed that the controlled nucleation led to higher values of residual moisture content in comparison with conventional freezing. This comparison was done for different secondary drying temperatures. The estimated kinetic constant of desorption was always higher in the case of uncontrolled cycles with respect to that for the controlled cycles, thus resulting in an increase of time needed for desorption when controlled nucleation was used. This was an expected result as control of nucleation leads to higher porosity and, thus, to a lower specific surface area available for the desorption of residual moisture. Although higher values of residual moisture content led to an additional time needed in the case of controlled nucleation to complete secondary drying stage, it was found that the total time to carry out the entire cycle was still smaller in the case of controlled nucleation in comparison with uncontrolled, thus confirming the advantage in using this control technique. Finally, the design space built for the secondary drying stage gave additional information about the impact of some operating conditions of the process on secondary drying time, thus giving a great contribution in the optimization of this step. The difference in water content between conventional and controlled freezing protocol was also investigated by in-line NIR spectroscopy analysis. This technology was able to discriminate between bound and surface water in freeze dried products. Differences in bound and surface water were detected only in the case of different temperature during secondary drying independently from the freezing protocol used.

Vacuum Induced Nucleation as a method for freeze drying optimization / Oddone, Irene. - (2016).

Vacuum Induced Nucleation as a method for freeze drying optimization

ODDONE, IRENE
2016

Abstract

Freeze drying is used to improve long-term stability of labile drugs. It comprises three different stages: freezing, primary drying, and secondary drying. Due to low values of pressure and temperature required in these stages, freeze drying is a cost-intensive process. Various authors have recently shown that the manipulation of the temperature of nucleation can provide benefits in terms of process optimization, and various technologies have been proposed to this purpose, such as electrofreezing, ultrasound, depressurization method, and ice fog. However, due to various problems of these technologies, such as the requirement of additional equipment and difficulties in scale-up, in this study the vacuum induced nucleation method was taken into consideration for its ease of application. Unfortunately, this method, as proposed in literature, produces defects in the cake structure because of blow-up of the frozen product and flake formation on the product surface, which are detrimental for the elegance of the final product and rejected by the pharmaceutical industry. For these reasons, a refined control technique has been here investigated and validated, showing that an accurate control of vacuum conditions can give a drastic reduction in cycle time along with an elegant product. The vacuum induced nucleation method consists in reducing the pressure within the drying chamber for a short time during the freezing stage. This pressure reduction produces the partial evaporation of water, which causes a reduction in product temperature and promotes the nucleation of ice. In this work I have developed a new method, still based on pressure reduction, that solved some of the problems of the original technique. In particular, once the desired value of pressure (which is product dependent) was reached, the drying chamber was isolated from the condenser for about 1 min. Because of that isolation, the pressure inside the drying chamber increased because of vapor accumulation, thus avoiding any esthetic defects to the product. A detailed investigation about the impact of the settings of the new control method on product and process performances was performed. To this purpose, various solutions containing mannitol, sucrose, lactose, dextran, glycine, polyvinylpyrrolidone, trehalose, and cellobiose were used. The results obtained showed that the method led to a high quality of the final product. Furthermore, showed that the induction of nucleation at higher temperature promoted the formation of very large crystals (see Fig. 1) with a consequent drastic reduction in cycle duration (due to lower values of product resistance to vapor flow, Rp). In order to promote the industrial scale-up of the method, a monitoring system based on thin-film technology, that allows, besides the monitoring of the product temperature, the detection of the end of the nucleation event was developed. Once the method was developed, a detailed investigation on its impact on both product (i.e., structure and homogeneity) and process (i.e., primary and secondary drying time) was carried out. The impact of the method on within-batch (inter-vial) and within-vial (intra-vial) variability was investigated. Comparing the onset-offset of pressure ratio curves of various cycles, it was found that the use of controlled nucleation (instead of conventional shelf-ramped freezing) dramatically increased the inter-vial homogeneity (by about 50%). This result was also confirmed by comparing the residual moisture values at the end of drying, which showed lower variance for controlled nucleation with respect to uncontrolled. A study, focused on mannitol-based formulations, showed that both nucleation temperature and filling volume had an impact on pore size and its distribution along the lyophilized cake. This within-vial variability was quantified by analyzing the average pore dimension along the cake (by Image-J program). The homogeneity of the product was also studied in terms of polymorph formation. To this regard, X-ray diffraction and off-line Raman were used to identify the mannitol polymorphs present within the product, which was found to vary with the freezing protocol and the temperature of secondary drying. Off-line Raman spectroscopy was also used to evaluate the polymorph distribution along the cake, showing that the spectra collected at the top of each sample often contained slightly higher amounts of α and δ mannitol in comparison with the spectra recorded in the other regions of the same cake. Finally, the Raman tool was also used for the in-line monitoring of the product characteristics, increasing the knowledge related to polymorph formation during freeze drying. To this purpose, it was found that the hydrate form of mannitol (which can damage the product if released during product storage) was formed during the entire cycle (freezing and drying stages), but was transformed into anhydrous polymorph immediately after the vacuum within the chamber was released at the end of the cycle. The primary drying stage of the cycles was performed using in-line monitoring (DPE+ algorithm) and control (LyoDriver) to further optimize the freeze-drying cycle. The use of these tools showed that, to obtain the optimum freeze-drying cycle, it was necessary to work on both freezing and primary drying stages. The mass-transfer resistance values obtained from each cycle (by DPE+ algorithm estimation) were used to calculate the design space and simulate the process. The cycle simulation highlighted the reduction of the cycle duration using the control method proposed (in comparison with an uncontrolled stochastic cycles) and the influence of its operating conditions on freeze-drying cycle optimization. It was, finally, shown that the primary drying duration could be decreased up to about 55% in the case of controlled nucleation. A study focused on the moisture content of the product during secondary drying showed that the controlled nucleation led to higher values of residual moisture content in comparison with conventional freezing. This comparison was done for different secondary drying temperatures. The estimated kinetic constant of desorption was always higher in the case of uncontrolled cycles with respect to that for the controlled cycles, thus resulting in an increase of time needed for desorption when controlled nucleation was used. This was an expected result as control of nucleation leads to higher porosity and, thus, to a lower specific surface area available for the desorption of residual moisture. Although higher values of residual moisture content led to an additional time needed in the case of controlled nucleation to complete secondary drying stage, it was found that the total time to carry out the entire cycle was still smaller in the case of controlled nucleation in comparison with uncontrolled, thus confirming the advantage in using this control technique. Finally, the design space built for the secondary drying stage gave additional information about the impact of some operating conditions of the process on secondary drying time, thus giving a great contribution in the optimization of this step. The difference in water content between conventional and controlled freezing protocol was also investigated by in-line NIR spectroscopy analysis. This technology was able to discriminate between bound and surface water in freeze dried products. Differences in bound and surface water were detected only in the case of different temperature during secondary drying independently from the freezing protocol used.
2016
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2640463
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo