Quantitative Post-Processing Characterization Techniques for Freeze-Dried Products

Sandra Leidholdt/Getty Images

Visual appearance, a primary evaluation method for freeze-dried products, is subjective and likely to vary between laboratories and over time. This article discusses how visual methods can be quantified and presents an overview of mechanical methods of characterizing and quantifying the properties of freeze-dried materials.

A well-dried product is one that will reconstitute into a viable product. It must have been frozen at a rate that will produce a desirable structure, and then kept at a low enough temperature during drying to maintain that structure without collapse.

The structure of the dried product reflects how successful the drying process has been; if there has been a collapse, then parts of the product will have been less thoroughly dried. Defects in the structure and, therefore, in the drying of a product will have an impact on how well it will rehydrate, whether any bioactivity will be preserved, and how long and under what conditions it can be stored. Even if the product has been dried well, the structure of the dried product still has a bearing on how easily it will rehydrate.

Assessing the visual appearance

CLICK IMAGE TO ENLARGE Figure 1: (A) Catastrophic failure is visually evident by the collapse of the freeze-dried product. (B) There can still be differences between products that are not easy to relate to specific defects; the product is uniform and adhered to the vial (left); uniform but not adhered to the vial (middle); has a peaked cake (right).

The primary method of assessing the quality of a freeze-dried product is by visual inspection. Catastrophic failure is visually evident by a collapse of the product (Figure 1A). Other product defects are not readily apparent but can still be determined by an assessment of appearance. For example, a partial collapse may have occurred during drying, or a suboptimal crystal structure may have been achieved during freezing so that the product will not reconstitute evenly. Some of these events may result in a change of appearance (Figure 1B), but it is not easy to relate changes in appearance to specific defects.

Aspects of visual appearance that can be assessed include shrinkage, collapse, skin formation, color, contaminants, and consistency. In order to analyze changes, it can be helpful to give a score to each of these attributes. The scores can then be analyzed and referred to during cycle development, scale up, or equipment changes. However, the process remains subjective and the consistency of scoring is likely to vary by day, analyst, or laboratory.

Microscopic features reveal product structure

CLICK IMAGE TO ENLARGE Figure 2: Scanning electron microspcope (SEM) images of two samples of freeze-dried mannitol reveal the differences in structure resulting from fast and slow cooling.

A microscopic visual assessment using scanning electron microscopy (SEM) or tunneling electron microscopy (TEM) reveals more information about product structure. SEM and TEM are valuable methods for looking at the microstructure to reveal possible micro-collapse and to determine the porosity of the product, which has a direct bearing on rehydration. The SEM images shown in Figure 2 reveal that fast cooling creates a structure with smaller crystals and therefore lower porosity than slow cooling.

This information can be used during cycle development for determining the impact of different methods on product structure. For example, SEM clearly shows the difference in structure caused by adding an annealing phase to the freezing of mannitol (Figure 3). It is also an excellent way of determining the quality of an individual sample. However, SEM still requires a visual assessment of the images, which results in variation due to subjectivity.

It is possible to develop a discrete classification with terms that could be applied to different regions in the same cake, but assessing the microscopic features remains a primarily qualitative method.

Figure 3: Scanning electron microscopy (SEM) images of two freeze-dried mannitol–sucrose mixtures. The mixture on the left has been annealed and reveals a rougher structure. The mixture on the right was not annealed and shows a smoother, amorphous structure.

Quantitative characterization methods

The challenge is to find a quantitative method that relates reliably to the product structure. The ideal would be a mechanical measurement that relates consistently to pore size, structure, and any presence of micro-collapse. The following mechanical tests can be applied alongside a visual and microscopic analysis:

• Gas adsorption to determine the specific surface area and the mean pore diameter

• X-ray diffraction to check for different polymorphic forms

• X-ray microcomputed tomography scanning to measure the porosity and heterogeneity

• Pressure testing to assess the mechanical properties.

Gas adsorption methods
The specific surface area and the mean pore diameter of a material can be determined by measuring the rate of adsorption and desorption (evaporation) of nitrogen gas to and from the material’s surface at low temperatures and under varying degrees of pressure.

The rate at which gas is adsorbed depends on the surface energy of the material, which itself depends on the surface structure. A rough surface has a larger surface area than a smooth surface. There also will be more surface atoms that are incompletely bound, which will increase the adsorption rate of a gas. The Brunauer-Emmett-Teller (BET) equation allows the determination of specific surface area from the gas adsorption rate.

The overall porosity of the material can be related to the difference between the adsorption and the desorption rate of the nitrogen. The Barrett-Joyner-Halenda (BJH)/Kelvin equations can be applied to the data to determine the pore-size distribution.

The following example shows how this method can be applied to samples of freeze-dried mannitol. Nitrogen adsorption and desorption isotherms were measured at -195.8 °C using an ASAP Tristar 3000 (Micromeritics Instrument Corp.) volumetric adsorption system for a fast-cooled mannitol solution, a slow-cooled mannitol solution, and liquid nitrogen (LN₂)-quenched mannitol.

Figure 4: The adsorption and desorption of nitrogen (LN2) under pressure by different samples of mannitol.

Isotherms were similar for each method (Figure 4), but the hysteresis between the adsorption and the desorption curves was wider for the LN₂ quench-cooled mannitol. This result indicates that the energy needed for evaporation from the pores was distinctly different from the energy associated with condensation within it, which implies that desorption was inhibited due to constriction, thereby suggesting that this sample had a smaller pore size (Figure 5).

X-ray diffraction to determine polymorphic forms

Figure 5: Determining the characteristics of mannitol lyophilized with different cooling methods; (A) The Brunauer-Emmett-Teller (BET) adsorption analysis to determine the surface area of the cake; (B) The Barrett-Joyner-Halenda (BJH) adsorption analysis to determine the mean pore diameter

X-ray diffraction is a well-established method of determining crystal structure. When used on lyophilized materials, the diffraction spectrum can reveal the different polymorphic forms resulting from different cooling methods.

Figure 6 shows the different spectra obtained from samples of dried mannitol that were frozen using different cooling rates. The spike at a diffraction angle of 9.5°2Φ for the slow-cooled solution is characteristic of the δ polymorphic form./

CLICK IMAGE TO ENLARGE Figure 6: X-ray diffraction (XRD) spectra of mannitol lyophilized with different freezing rates. The spectra reveal the different polymorphic form in the sample cooled at 0.5°C per minute (bottom).

Micro-CT scanning for porosity and heterogeneity
Micro-computed tomography (CT) scanning is a noninvasive method of visualizing the internal micro-structure of a three-dimensional (3D) object. Cross-sectional X-rays are taken at small intervals across the product. CT software takes these cross sections and builds them into a 3D-visualization of the internal structure.

Computational methods have been developed for quantifying the 3D-images, allowing measurement of porosity, pore-size distribution, pore connectivity, and particle size throughout the sample matrix. The nondestructive and penetrating nature of the analysis allows observation of the structure throughout the entire sample, without risk of deforming the material by cutting or other methods to access the lower parts of the product that could invalidate the results.

Pressure testing to determine cake strength
Work is underway between Imperial College London and Biopharma Technology Ltd to develop a miniature load cell to measure stress and strain in a lyophilized cake while it is still in the vial. Measuring the stress (σ) and strain (ε) gives the elasticity, Young’s modulus (E) (see Equation 1):

The following example shows how this method works in practice. The aim was to compare the strength of different cakes through compression testing. The samples were tested in the vial using a Stable Microsystems Texture analyzer and Lloyds EZ50 column test stand (Figure 7). Three different samples of mannitol were tested: fast-cooled, slow-cooled, and LN₂-quench cooled. The pressure was applied to the cakes at a speed of 1 mm/sec, tested in increments of 0.01 mm/sec to a depth of 3 mm, starting at 1 g of force.

Figure 7: A miniature load cell is used to apply pressure to the freeze-dried cake while it is still in the vial.

The results showed that the fast-cooled lyophilized mannitol was the least flexible (i.e., had the greatest resistance to deformation when force was applied) and that the slow-cooled lyophilized mannitol was the least resistant to deformation (Figure 8). This result is likely to be due to morphology; the slow-cooled mannitol has larger pores, and therefore a weaker structure. An SEM analysis of the samples revealed small holes that are visible in the LN₂ and slow-cooled samples; holes are the weakest points in the cakes where they can crack or break easily.

The clear differential shown between the Young’s modulus for these mannitol samples merits further work on how this method can be developed to provide a quantitative assessment of the mechanical properties of freeze-dried materials, either to compare different formulations or batches of the same product./

Conclusion

Figure 8: The results of the pressure test can be used to determine Young’s modulus for different samples. The difference in results can be related to the different structure of the samples.

Product-appearance assessment can be subjective. Microscopic analysis is a valuable tool for determining product structure, but it still requires a subjective visual analysis that is likely to vary across operators, sites, and time.

Gas adsorption, X-ray diffraction, micro-CT, and pressure testing are quantifiable methods of measuring aspects of product structure. These techniques can be used to improve the consistency of freeze-drying processes across different sites and over time. They can also improve the understanding of the process or formulation changes that are required to achieve high quality freeze-dried products.

ALL FIGURES ARE COURTESY OF BIOPHARMA TECHNOLOGY LTD.

About the Author
Katriona Scoffin is a freelance writer with extensive experience in the life sciences. She works from Cambridge, UK. Contact her at articles@scoffin.com.

Article DetailsBioPharm International
Vol. 28, No. 12
Pages: 51–55

Citation: When referring to this article, please cite it as K. Scoffin, "Quantitative Post-Processing Characterization Techniques for Freeze-Dried Products," BioPharm International 28 (12) 2015.