Dynamical in-situ observation of the lyophilization and vacuum-drying processes of a model biopharmaceutical system by an environmental scanning electron microscope
Graphical abstract
Introduction
Many pharmaceuticals and biopharmaceuticals have proved to be unstable in aqueous solutions. Such materials are usually stabilized by drying, with lyophilization (freeze-drying) being the most common method; other approaches to be considered include, for example, vacuum-drying (Mattern et al., 1999, Mattern et al., 1997) and spray-freeze-drying, an unconventional freeze-drying technique that involves the generation of small droplets, flash freezing, and drying (Ishwarya et al., 2015, Wanning et al., 2015). As regards the numerous freeze-dried pharmaceuticals, we can consider, for example, blockbuster protein drugs, including Enbrel and Remicade; small molecular weight antibacterial drugs, exemplified by cephalosporins (such as Convenia); and many viral vaccines (e.g., those employed to treat measles, mumps, and rubella). Vacuum-drying facilitates the production of BOTOX, which has found use in multiple aesthetic and medical applications. A typical lyophilization process comprises 3 steps: freezing, primary drying (ice sublimation), and secondary drying (removal of unfrozen water). Freezing embodies a critical stage in the lyophilization procedure, as both the efficiency of the drying and the quality of the final product are directly influenced by the freezing conditions. In particular, the processes that exploit controlled ice nucleation, where freezing is initiated at a pre-selected temperature (usually slightly below the equilibrium freezing point), could be beneficial due to the shorter production times enabled by the faster ice sublimation (Geidobler and Winter, 2013) or better recovery of the product potency (Fang et al., 2018).
Besides the empirical approaches to freezing and lyophilization, there are numerous attempts, including some based on our research, to designate and separate the freezing-related stresses that might affect the stability and reactivity of frozen and freeze-dried compounds: low temperature (Bhatnagar et al., 2007, Kasper and Friess, 2011), freeze-concentration (Bhatnagar et al., 2007, Heger et al., 2005, Kania et al., 2014, Kasper and Friess, 2011, Krausko et al., 2015, Ondrušková et al., 2018; Takenaka and Bandow, 2007), ice formation (Bhatnagar et al., 2007, Kasper and Friess, 2011), freezing-induced acidity change (Gomez et al., 2001; Govindarajan et al., 2006; Heger et al., 2006, Imrichová et al., 2019, Krausková et al., 2016; Sundaramurthi et al., 2010; Vetráková et al., 2017), and polarity variations (Heger and Klán, 2007). Recently, mathematical simulations were applied to predict the optimal freezing protocol (Arsiccio et al., 2020) and the morphology of freeze-dried products (Capozzi and Pisano, 2018). In order to identify optimal freezing conditions, the morphologies of freeze-dried cakes prepared under diverse freezing processes are commonly evaluated. Methods for evaluating the morphological features of freeze-dried materials include, e.g., determining the specific surface area (Rambhatla et al., 2004), X-ray tomography (Foerst et al., 2019, Pisano et al., 2017), mercury porosimetry (Hottot et al., 2007), optical coherence tomography-based freeze-drying microscopy (Mujat et al., 2012), and scanning electron microscopy (SEM). Of these, the last-named technique embodies probably the most common tool to evaluate the morphologies of freeze-dried cakes: The SEM has been employed, for example, to examine the impact exerted on the cake morphology by controlled ice nucleation via different methods (pressurization-depressurization (Awotwe-Otoo et al., 2013, Iyer et al., 2016, Kawasaki et al., 2018), vacuum-triggered ice nucleation (Oddone et al., 2017), and ice nucleation through partial vacuum followed by its quick release (Izutsu et al., 2014)), to analyze the pores in freeze-dried cakes produced via various freeze-drying cycles (Capozzi et al., 2019, Liu, 2006, Overcashier et al., 1999), to visualize the crystallization of active pharmaceutical ingredients and excipients (Murase et al., 1991, Ogienko et al., 2014, Shalaev et al., 1996), to evaluate the morphologies of vacuum-dried pharmaceuticals (Mattern et al., 1999, Mattern et al., 1997), and to predict the resistance of lyophilized products to vapor flow (Arsiccio et al., 2019, Arsiccio et al., 2017, Grassini et al., 2016, Pisano and Capozzi, 2017).
In such SEM studies, a common underlying assumption is that the morphology (e.g., porosity) of the freeze-dried cake accurately characterizes that of the ice crystals in the frozen sample, if no collapse occurs during the primary and secondary drying. While this assumption is intuitively logical, there remains a lack of research reports on direct comparison of the morphology of a frozen material with that of the freeze-dried cake. Understanding the morphologies of ice crystals is essential for both practical applications, such as developing optimal freezing conditions for the manufacturing of freeze-dried products, and understanding the impact of freezing on the stability of protein molecules.
A conventional SEM requires low pressures (<10−2 Pa) to avoid undesirable scattering of the electron beam due to collisions with the atoms and molecules of gases (Reimer, 1998). The low pressure requirement confines this technique to the observation of samples that are stable in vacuum conditions; the majority of water-containing materials cannot be studied with the method. Ice and frozen aqueous solutions are commonly examined via SEMs equipped with a cryo-stage at temperatures below −60 °C (Baker et al., 2007, Baker et al., 2003, Baker and Cullen, 2003; Barnes et al., 2002a, Barnes et al., 2002b, Barnes et al., 2003, Barnes and Wolff, 2004, Blackford et al., 2007, Chen and Baker, 2010, Cross, 1969, Cullen and Baker, 2001, Dominé et al., 2003, Erbe et al., 2003, McCarthy et al., 2013, Obbard et al., 2003, Rango et al., 2000, Rango et al., 1996, Rosenthal et al., 2007, Wergin et al., 2006b, Wergin et al., 2006a, Wergin et al., 1998), as imaging at higher temperatures leads to difficulty maintaining the SEM vacuum due to excessive ice sublimation (Cullen and Baker, 2001). In contrast, temperatures typical of the freezing and primary drying steps of the lyophilization procedure are −50 to −10 °C. Another problem relates to observing dielectric samples when a higher energy electron beam is used: Such samples become negatively charged, and their observation is hindered. This effect is usually eliminated in a sample surface coated with a thin conductive layer (Goldstein et al., 2003). The layer nevertheless interferes with the actual monitoring; in this context, for example, it modifies the chemical composition or masks fine morphological details (Fourie, 1982, Jones, 2012) and might prevent investigation of the processes occurring in the microscopic chamber during the visualization.
To enable direct observation of samples in their native states, the first commercial environmental scanning electron microscope (ESEM) was introduced around 1980 (Danilatos, 1981). Differential pumping enables an ESEM to maintain a high difference between the very low pressure near the electron gun and the relatively high pressure in the specimen chamber. The primary electron beam is formed and focused in a high-vacuum environment and then transferred through pressure-limiting apertures to the specimen chamber, where it travels only a very short distance to reduce the primary-beam scattering in the gas (Danilatos, 1988). Still, a portion of the primary electrons are scattered by the gas, leading to a decrease in the detected signal-to-noise ratio and a lower resolution of the ESEM. Conversely, the gas molecules are ionized due to collisions with the signal electrons emitted from the sample; such procedures then result in efficient signal amplification (Neděla et al., 2018, Neděla et al., 2015) and also compensation of the negative charge on the sample surface. Thus, electrically non-conductive samples do not require conductive coating (Stokes, 2008). Due to the higher pressure (up to thousands of Pa) of the water vapor in the specimen chamber, biological (Vlašínová et al., 2017), polymeric (Bertóková et al., 2015), or other water-containing samples (Navrátilová et al., 2017) can be examined without being dehydrated (Stokes, 2003). In wet samples, controlling the sample temperature and air/vapor pressure in the chamber of the ESEM allows the thermodynamic equilibrium to be established or perturbed to facilitate investigation of the phase transitions (e.g., condensation, evaporation, sublimation, and crystallization) (Donald, 2003, Krausko et al., 2014, Neděla et al., 2020, Schenkmayerová et al., 2014, Tihlaříková et al., 2013, Yang et al., 2017). The ESEM was used for studying the ice nucleation processes (Varanasi et al., 2010, Zimmermann et al., 2008, Zimmermann et al., 2007) and growth of ice crystals (Pedersen et al., 2011), observing metastable liquid water droplets below the pressure of the triple point (Chen et al., 2017), visualizing the brine on the ice surface (Imrichová et al., 2019, Krausko et al., 2014, Vetráková et al., 2019), and examining the sublimation of the frost flowers (Yang et al., 2017) or pure ice (Nair et al., 2018) in the microscope’s chamber. It was shown previously that an ESEM can be used to observe the whole lyophilization process under pharmaceutically relevant conditions (Meredith et al., 1996); this is the only study discussing lyophilization or vacuum-drying performed via an ESEM we are aware of.
This paper presents dynamical in-situ observation of the drying step in the lyophilization of a model protein BSA, both pure and mixed with sucrose, inside the chamber of an ESEM. The solutions were subjected to freezing under various conditions (spontaneous freezing, controlled ice nucleation, and spray freezing), whereas the drying conditions were kept constant. While the first two freezing methods are commonly used in conventional freeze-drying, the third one embodies the emerging technology of spray-freeze-drying (Ishwarya et al., 2015). The effect of the freezing method on the progress of the lyophilization and the structure of the resulting lyophile was thoroughly examined (in this article, the terms structure and morphology are used interchangeably). Unexpectedly, and contrary to the underlying assumptions proposed in previous SEM studies of freeze-dried materials, the lyophile morphology can be significantly altered during the ice sublimation, even at temperatures well below the collapse temperature (Tc). Therefore, studying a final freeze-dried material might not be enough to yield relevant data on the morphology of the ice crystals formed during the freezing step. This observation could have a far-reaching impact on both the design of the freezing conditions for various freeze-dried products and the fundamental understanding of the protein/ice interaction and destabilization of the protein molecules by ice. Finally, the vacuum-drying process was also monitored, revealing remarkable differences in the sample morphology.
Section snippets
Materials
The solutions of pure BSA and BSA mixed with sucrose were prepared by dissolving powdered BSA (fraction V, >98%, ~66 000 g mol−1, Carl Roth) and sucrose (Penta) in MilliQ water. A constant total solid content of 50 mg g−1 was maintained in all of the solutions. The concentrations of the BSA and sucrose in the mixtures were [40/10], [30/20], and [10/40] mg g−1, respectively.
Instrumentation
The lyophilization procedures were performed with two ESEMs: a non-commercial AQUASEM II (a Tescan VEGA SEM customized at
Effect of freezing method: Pure BSA
In the pure BSA, the lyophilization process strongly depended on the applied freezing method, as described below and summarized in Table 1. All the frozen samples were solid, without a visible layer of liquid on the surfaces or a tendency to flow during the whole lyophilization cycle. No collapse of the samples or formation of bubbles were observed.
Comparing lyophilization techniques: An ESEM versus a common freeze-dryer
Spontaneous freezing and controlled ice nucleation are procedures widely used in conventional freeze-drying. The main difference between the samples used in our study and typical ones (frozen in a usual container) consists in their volumes. By extension, the spray-freezing method finds application in the emerging technology of spray-freeze-drying (Ishwarya et al., 2015), where the size of the samples markedly resembles that of our spray-frozen ones. Concerning the size of a sample, the ESEM
Conclusion
In this study, a 50 mg g−1 BSA solution and BSA/sucrose mixtures with the protein/sugar concentrations of [40/10], [30/20], and [10/40] mg g−1 were freeze-dried in an ESEM, with the ice sublimation performed either below the Tc (the pure BSA and BSA/sucrose [40/10] and [30/20] mixtures) or above it (BSA/sucrose [10/40]). The course of the lyophilization was monitored in a step-by-step manner, and the morphological changes of the samples during the drying phase of the lyophilization process were
CRediT authorship contribution statement
Ľubica Vetráková: Investigation, Formal analysis, Validation, Writing - original draft, Visualization. Vilém Neděla: Resources, Writing - review & editing. Jiří Runštuk: Investigation. Eva Tihlaříková: Investigation. Dominik Heger: Validation, Writing - review & editing. Evgenyi Shalaev: Conceptualization, Validation, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgement
The research was supported by Czech Science Foundation via projects GA 19-08239S and GA 19-03909S.
The authors’ contributions
Ľ.V. conducted the experiments, analyzed the data, and wrote the manuscript; J.R.and E.T. operated the ESEM; E.S. developed the idea of the project; and V.N., D.H., and E.S. contributed to the discussions and design of the final paper.
References (112)
- et al.
Considerations on protein stability during freezing and its impact on the freeze-drying cycle: a design space approach
J. Pharm. Sci.
(2020) - et al.
Impact of controlled ice nucleation on process performance and quality attributes of a lyophilized monoclonal antibody
Int. J. Pharm.
(2013) - et al.
Study of the individual contributions of ice formation and freeze-concentration on isothermal stability of lactate dehydrogenase during freezing
J. Pharm. Sci.
(2008) - et al.
Looking inside the ‘black box’: freezing engineering to ensure the quality of freeze-dried biopharmaceuticals
Eur. J. Pharm. Biopharm.
(2018) Foundations of environmental scanning electron microscopy
Adv. Electron. Electron Phys.
(1988)- et al.
Effect of freezing on lyophilization process performance and drug product cake appearance
J. Pharm. Sci.
(2016) - et al.
Effect of controlled ice nucleation on stability of lactate dehydrogenase during freeze-drying
J. Pharm. Sci.
(2018) - et al.
Controlled ice nucleation in the field of freeze-drying: Fundamentals and technology review
Eur. J. Pharm. Biopharm.
(2013) - et al.
Impact of freeze-drying on ionization of sulfonephthalein probe molecules in trehalose-citrate systems
J. Pharm. Sci.
(2006) - et al.
Frequency domain image analysis for the characterization of porous products
Meas. J. Int. Meas. Confed.
(2016)