Keywords

9.1 Introduction

Freezing of water is a complex process of ice formation at low temperatures, and is the fundamental principle of bioactive stabilization using freezing and freeze-drying . Freezing and frozen storage prior to freeze-drying is a process of microstructure formation for stability control of frozen and freeze-dried materials. In long-term stabilization, frozen materials need to be stored at conditions where unfrozen solutes remain in a vitrified solid state, while the vitrified solid state of frozen structures is preserved for long-term ambient storage by ice sublimation in the freeze-drying (lyophilization) process. The frozen material shows an expansion corresponding to the volume expansion of the same quantity of water converted to ice during crystallization.

Protection of protein conformation and retention of biological functions subsequent to freezing and thawing or freeze-drying and rehydration is common in the food and pharmaceutical industries. The use of carbohydrates , particularly monosaccharides and disaccharides, as cryoprotectants to retain protein structures against freezing and dehydration damages has been of particular interest (Jorgensen et al. 2009). The phase behavior of water and the state of dissolved substances dominate the thermal properties of cryoprotectant systems. Ice formation always results in separation of a freeze-concentrated unfrozen aqueous cryoprotectant phase and vitrification of the maximally freeze-concentrated unfrozen phase over a cryoprotectant-specific temperature range (Roos and Karel 1991a, b; Luyet and Rasmussen 1968; Franks et al. 1977; Levine and Slade 1986; Pehkonen et al. 2008). The phase and state behavior of cryoprotectant–water systems may be described using state diagrams (Franks et al. 1977; Roos and Karel 1991a, b; Buera et al. 2011). While such diagrams have clarified the physicochemical properties of carbohydrate systems at low temperatures, they lack the information needed to understand carbohydrate–protein interactions in biological, food and pharmaceutical systems as well as intracellular and extracellular ice formation in cellular structures.

Although freezing and freeze-drying are well-established processes for preservation of biological materials, the present chapter highlights physicochemical properties of typical food and related materials affecting structure formation and properties of frozen and freeze-dried materials, with a particular focus on stabilization of bioactives.

9.2 Freezing and the Frozen State

The freezing of food materials is a well-established industry, as freezing stops or significantly delays microbial growth, and delays most biological processes and chemical reactions. Equilibrium freezing (referring here to ice formation and phase and state transitions below 0 °C) properties of eutectic salt solutions such as sodium chloride are well described in general physical chemistry (Atkins and De Paula 2010), but properties of sugar solutions and more complex dispersed structures are often poorly understood. Biological and food materials always contain unfrozen water at low temperatures. The term “unfrozen water ” here refers to water in which interactions with solutes occur, although it is worth noting that the properties of even unfrozen pure water at temperatures within normal limits for supercooling and vitrification, i.e., −40 °C to −130 °C (no man’s land), are poorly understood (Angell 2008, 2014). On the other hand, freezing and frozen storage of biological materials is routinely carried out using liquid nitrogen (liquid-N2, −196 °C) and −80 °C equipment. The use of such temperatures is based on the freezing medium and equipment availability rather than scientific data for optimal material stability.

$$p_{w} = xp_{0}$$
(9.1)

The freezing temperature depression of water by solutes in dilute solutions is well known, and is described by Raoult’s law (Eq. 9.1), which states that the smaller the solute molecules, the larger their effect on water vapor pressure, p w. The equilibrium melting temperature of the last ice crystals during heating of a solution with ice (i.e., freezing temperature), T m, has the requirement of p i = p w (where p i is the vapor pressure of the ice). The T m decreases with solute concentration, which for frozen systems implies that the amount of frozen water (ice) increases with decreasing temperature. Conversely, during frozen storage, temperature fluctuations can result in substantial melting–refreezing phenomena and rapid recrystallization of ice, with a consequent increase in the size of ice crystals. The temperature dependence of ice in frozen solutions is shown in Fig. 9.1. In a binary eutectic system, both the solvent water and the dissolved substance crystallize as solubility is exceeded, causing precipitation of the solute at the eutectic temperature, T e. Eutectic freezing is unusual in complex multicomponent systems, and many biological materials show vitrification of their unfrozen water –solute phase. The glass transition temperature of the maximally freeze-concentrated unfrozen phase is known as T g′. Ice formation in a maximally freeze-concentrated solute phase ceases at T m′ as the solute concentration approaches C g′. A maximally freeze-concentrated material, therefore, shows an onset of the glass transition at T g′ and onset of ice melting at T m′ when measured in heating after freezing to the maximally freeze-concentrated state.

Fig. 9.1
figure 1

Relative quantity of ice forming in biological materials at temperatures corresponding to equilibrium melting temperature, T m; temperatures above the onset of ice melting in a maximally freeze-concentrated state, T m′; and in a maximally freeze-concentrated state at T m′

Full size image

9.3 Glass Transition of Frozen Materials

Glass transition is a reversible change in the state of supercooled noncrystalline materials. The glass transition can be determined from thermodynamic properties such as heat capacity and thermal expansion of supercooled liquids as they are heated and cooled over the glass transition temperature range (Angell 2002; Roos and Drusch 2015). Supercooled states of materials show molecular disorder and random time-dependent characteristics. The glass transition is accompanied by changes in mechanical and dielectric properties as well as other structural relaxations resulting from a change in the molecular mobility of the glass-forming substance. The structural relaxations occurring over the reversible glass transition are referred to as α-relaxation. The α-relaxation is typically studied in heating by monitoring the time and temperature dependence of the molecular response of the material during glass transition . The relaxation time is derived from the time dependence of the α-relaxation. The characteristic relaxation time, τ, for the solid, glassy state is 100 s, corresponding to a viscosity of 1012 Pa s (Angell 2002).

9.4 Carbohydrate Systems

Carbohydrates show a large variation in molecular size as components of biological materials. On the other hand, small sugars are often highly soluble and occur in large quantities in biological materials important to freezing and freeze-drying . Conversely, small sugars are the most common cryoprotectants used for preservation of starter cultures and native protein structures. Sugar systems have been studied in detail for understanding the frozen state phenomena of aqueous solutions of carbohydrates and cryoprotectant systems (Roos 1993; Corti et al. 2010; Roos and Drusch 2015).

Freezing of carbohydrate systems commonly occurs without solute crystallization, as described in Fig. 9.1. Solute crystallization is limited by the high solubility of many carbohydrates and their slow nucleation and crystal growth rates compared to that of ice formation. As a consequence, maximally freeze-concentrated states of carbohydrates are common, and are typically a prerequisite for successful freeze-drying where any liquid water must be part of the maximally freeze-concentrated vitrifying solute phase at temperatures lower than T m′.

State diagrams are often constructed for binary carbohydrate–water systems to describe ice formation as well as freezing and freeze-drying properties of carbohydrate systems. The state diagram of sucrose is shown in Fig. 9.2.

Fig. 9.2
figure 2

State diagram of sucrose showing water content dependence of glass transition temperature, T g; ice melting temperature, T m; onset of glass transition of maximally freeze-concentrated solute phase, T g′; and onset of ice melting in a maximally freeze-concentrated solution, T m′

Full size image

9.5 Carbohydrate–Protein Systems

Protection of proteins and peptides against dehydration or freezing damage advances the survival of living organisms as well as the production and stabilization of active enzymes, pharmaceuticals and biological materials. Proteins contain hydration water in cavities and surface pores. Strong hydrogen bonding reduces the chemical potential of water molecules at hydration sites and explains the unfrozen state of hydration water at low temperature (Jansson et al. 2011; Roos and Potes 2015; Swenson and Cerveny 2015). The α-relaxation of hydration water in proteins is coupled with vitrification phenomena in cryoprotection and anhydrobiosis, particularly with the protection of protein conformation when carbohydrate protectants are used. Water molecules in amorphous carbohydrate systems retain high mobility, and they may exchange with water molecules that are hydrogen-bonded to the protein. Depending on the carbohydrate, vitrification of the hydration water and aqueous phase may occur at different temperatures, with consequent effects on protein stability (Roos and Potes 2015). In a protein–carbohydrate system, an exact quantity of water is required to exist as protein hydration water to satisfy a thermodynamic balance with a co-existing carbohydrate phase. At equilibrium, such hydration water has equal water activity a w (equal chemical potential and vapor pressure) to that of water associated with the carbohydrate. Our studies have shown that the sugars co-existing with proteins in a freeze-concentrated solute phase give a T g′ for the unfrozen carbohydrate–unfrozen water phase. Measurement of the T g′ can be used to derive the composition of the unfrozen carbohydrate phase, while the unfrozen protein hydration water can be obtained from the water sorption data (Roos and Potes 2015). Protein hydration water in a maximally freeze-concentrated protein–carbohydrate system showed a quantity that became independent of the initial water content and the carbohydrate content, i.e., maximally freeze-concentrated carbohydrate was plasticized by a measurable unfrozen water fraction. The surrounding protein was maximally freeze-concentrated with a corresponding unfrozen hydration water fraction. A system composed of bovine whey proteins and a glucose–fructose (1:1) carbohydrate phase (GF) showed unfrozen water fractions in the maximally freeze-concentrated state with 0.164 g H2O/g protein and 0.273 g H2O/g GF. The data measured by Roos and Potes (2015) were used to establish a tertiary three-component state diagram shown in Fig. 9.3. The experimental quantification of the unfrozen water in the protein and carbohydrate components of cryoprotectant systems at maximum freeze-concentration as described by Roos and Potes (2015) provided a significant universal method of quantifying (i) protein hydration water and (ii) the unfrozen water in the carbohydrate phase. This is important for the understanding and use of carbohydrate protectants in freezing , lyophilization and dehydration of biological materials, and in the stabilization of bioactive components in formulation and structure engineering of food and pharmaceutical materials. It should be noted that earlier studies of protein hydration water have used a rough estimate of 0.2–0.3 g of unfrozen water/g of protein. Such estimates were made on the basis of unfrozen water measurements using differential scanning calorimetry (DSC) (Doster et al. 1986).

Fig. 9.3
figure 3

Tertiary state diagram for bovine whey proteins (BWP) and glucose:fructose (1:1) as carbohydrates (GF). The glass transition of the GF, T g, is found using differential scanning calorimetry at the temperature dominated by the GF phase. At maximum freeze-concentration, the glass transition, T g′, is that of the carbohydrate, while hydration water on protein shows an α-relaxation at T α′. The water content corresponding to the glass transition of the carbohydrate, T g, and α-relaxation of protein hydration water, T α, can be derived from the component water sorption isotherms

Full size image

Roos and Potes (2015) showed that a separate α-relaxation of the protein hydration water and of the bulk GF were signs that hydration water molecules were primarily hydrogen-bonded to protein molecules, and that hydration water provided a bridge for hydrogen bonding of the protein to the continuous GF phase of the system. Proteins as large molecules were considered as dispersed particles within the carbohydrate phase. At maximum freeze-concentration, a “syrup” with 2.5 mol of water per mol of hexose unit is known to remain unfrozen in sugar solutions (Roos and Karel 1991a, b), while the whey protein isolate (WPI)–GF shows an increased 2.75 mol H2O/mol GF, implying an increase in hydrogen bonding towards the protein hydration interface. At low temperatures and under lyophilization conditions, proteins with carbohydrate cryoprotectants exist with unfrozen hydration water exposed to either a solid (glassy) or liquid (rubbery) unfrozen aqueous carbohydrate phase. A liquid carbohydrate phase causes significant osmotic stress which is not present in the vitrified state of the carbohydrate. According to Roos and Potes (2015), hydrated protein particles within cryoprotectant solutions become encapsulated by the vitrifying carbohydrate phase. Such a mechanism explains the retention of protein conformation in freezing and dehydration by freeze-drying. In other dehydration processes, glass formation around hydrated protein by a protectant carbohydrate can also explain the stabilization of protein conformation and structure.

9.6 The Freeze-Drying Process

Freeze-drying (lyophilization ) involves the freezing of water to ice and establishment of drying conditions where sublimation of ice to water vapor is accelerated and the vapor is condensed as ice on a cold surface. Such conditions can be established by lowering the temperature or using vacuum equipment to lower the pressure of a drying chamber to control the temperature of the ice.

The freeze-drying process is described in Fig. 9.4. A typical process uses vacuum equipment to lower the pressure of the drying chamber to p < 1 mbar. At such conditions, the ice temperature, T ice, is a direct function of the surrounding pressure. On the other hand, refrigeration equipment is used to lower the temperature of a condensing surface, T c, to below the temperature of the sublimating ice. This allows for a condition where the vapor pressure of the ice, p ice, of the drying material is higher than that found on the condenser surface, p c. Hence, the driving force for sublimation is defined by the vapor pressure difference, ∆p = p ice − p c. Obviously, an increase in ∆p accelerates the freeze-drying process.

Fig. 9.4
figure 4

The principle of freeze-drying (right) and heat transfer and mass transfer to accelerate and control freeze-drying when by conduction (left top) or radiation (left bottom)

Full size image

Solutes in the water phase of biological materials affect the freezing of water, and they each have a specific temperature for maximum freeze-concentration and corresponding onset temperature of ice melting, T m′. A prerequisite for freeze-drying is the solid state for the sublimating ice, and T m′ thus provides a critical temperature for sublimation in freeze-drying. Successful freeze-drying requires that the temperature for sublimation is lower than the T m′ of the water–solutes phase to avoid melting during sublimation. The ice temperature is affected by the chamber pressure and also by the heat and mass transfer within the material undergoing freeze-drying . Heat supplied to the sublimating ice can accelerate freeze-drying, while mass transport inside the drying particle controls the internal pressure of the drying particle (Fig. 9.4). This also causes large differences between freeze-drying processes using radiation and contact heat transfer. As described in Fig. 9.4, during freeze-drying, ice is an excellent heat conductor (heat transfer by conduction through an ice layer on a heated shelf), while an already dried porous layer of a biological material is an efficient insulator for heat transfer (radiation from heated plates and heat conduction to sublimation interface within a drying particle).

Conditions which must be satisfied during a successful freeze-drying process are shown in Fig. 9.5. A preferably external pre-freezing step of a material must ensure that the freezing step produces a desired size of ice crystals and surrounding external wall of unfrozen solids phase thickness (Harnkarnsujarit et al. 2012). Thereafter, the ice temperature must be lowered to well below the T m′, with corresponding ice vapor pressure of p ice < p m′, to ensure that T ice < T m′ applies during loading of the freeze-dryer until proper freeze-drying conditions for maintaining T ice < T m′ have been established. During any condition of T ice > T m′ corresponding to p ice > p m′, a liquid water phase appears and results in viscous flow within the unfrozen water –solutes phase. Such conditions lead to shrinkage and collapse of the structure of the drying material, and often to poor dehydration and loss of quality.

Fig. 9.5
figure 5

Control of a freeze-drying process using information on ice melting properties to maintain ice sublimation at temperatures below the onset temperature of ice melting, T m′

Full size image

9.7 Freezing and Freeze-Drying of Microbes

The survival of starter cultures and preservation of bioactive micro-organisms, particularly lactic acid bacteria, are of fundamental importance in preserving viable probiotic cultures for food use, as well as in the development of dehydrated foods with probiotic activity. Frozen culture concentrates of probiotics exhibit maximal survival in liquid nitrogen (Smittle et al. 1974), while freeze-drying is one of the most successful and convenient methods of preserving microbial cultures (Berny and Hennebert 1991; Carvalho et al. 2003).

The major loss of cell viability due to freezing is attributed mainly to poor freezing or freeze-drying conditions that affect ice crystal formation, high osmotic stress due to high concentrations of internal solutes of cells, leading to membrane damage, and macromolecular denaturation and the removal of water, which affect the properties of many hydrophilic biological macromolecules (Thammavongs et al. 1996). Franks et al. (1991) showed that the activity of proteins was substantially reduced during freeze-drying, and was further reduced by subsequent storage unless the dried material was kept below −20 °C, and ideally at −80 °C. Such finding suggested that stability of protein materials may be achieved at conditions supporting vitrification of protein hydration water and associated low mobility of macromolecular components (Roos and Potes 2015).

Cryoprotectant molecules used in freezing and freeze-drying are often small sugars such as glucose and sucrose, while lactose may be also present depending on the fermentation media. Skim milk and/or combinations of sucrose, trehalose and dextran are most commonly used as cryoprotectants (Hubalek 2003). Sugars exhibit typical glass-forming properties of carbohydrates , and they have a significant impact on freezing, freeze-drying and storage characteristics. Such properties were reported for common sugars by Roos (1993), and are useful in the selection of cryoprotectant systems and freezing and freeze-drying conditions. Reconstituted skim milk solids are among the most useful suspension media for the freezing or freeze-drying of starter cultures due to their cryoprotective effects on cells (Abraham et al. 1990), and supplementing skim milk solids with different cryoprotectants may increase the cryoprotective effect (King and Lin 1995). A large number of compounds have been assessed for use as cryoprotectants, including skim milk solids, glycerol, mannitol, sorbitol, trehalose, sucrose, maltose, lactose, fructose, glucose, betaine, monosodium glutamate, honey, and amino acids and their salts.

A typical protectant should provide cryoprotection of probiotic bacteria cells during freezing and freeze-drying processes and ease of rehydration after long-term storage (Costa et al. 2000). Protectants present in a suspending medium have two main modes of action in preserving the viability of freeze-dried cells. Firstly, a glass-forming carbohydrate will provide mechanical support during storage. Secondly, the living cells are biochemically protected against damage during the drying process and subsequent storage (Berny and Hennebert 1991). Stabilization by small carbohydrate molecules has also been explained using the “water replacement hypothesis”, which refers to their capacity to form hydrogen bonds with components of biological structures in the absence of water (Crowe et al. 2001). The ability of skim milk to protect freeze-dried microorganisms is thought to be due to its capacity to stabilize the cell membrane constituents and to create a porous structure in the freeze-dried materials (Selmer-Olsen et al. 1999).

Interestingly, sugar composition within microbial cells cannot be assumed to be the same as that of a cryoprotectant medium. Sugar metabolism involves sugar uptake into the cell and breakdown into simpler products that are easily metabolized for energy. In lactic acid bacteria, there are two systems for the transport and metabolism of lactose. The two methods for transporting lactose into the cells are as follows (Walstra et al. 1999):

  1. 1.

    A phosphoenol pyruvate-dependent phosphotransferase system (PEP/PTS): As lactose is transported into the cell, it is transformed to lactose-P. Once inside the cell, phospho-β-galactosidase (P-β-gal) hydrolyzes the lactose-P to glucose and galactose-6-P. Glucose is further converted to glucose-P, and both phosphated sugars are further metabolized.

  2. 2.

    An ATP-dependent permease system: Lactose is transported into the cell as described above, and hydrolyzed by β-galactosidase or lactase into glucose and galactose. Glucose is converted to glucose 6-P. In galactose-fermenting bacteria, galactose is converted to glucose-6-P by the Leloir pathway.

Galactose is not metabolized if enzymes of the Leloir pathway are not present (i.e., Streptococcus thermophilus and Lactobacillus delbrueckii ssp. bulgaricus). Therefore, galactose present in such probiotic cells is either secreted or converted to (exo)polysaccharides (Walstra et al. 1999). Obviously, fermentation conditions and cell composition at the time of freezing may have an important effect on cell viability in both freezing and freeze-drying , as intracellular sugar composition may show large variations. Fonseca et al. (2000) found that the larger the surface area of the cell, the greater the membrane damage, owing to extracellular ice crystal formation during freezing. Water content as such is also an important factor affecting the viability of bacteria during freeze-drying and subsequent storage (Zayed and Roos 2004).

Crowe et al. (1984) reported that the role of trehalose in the stabilization of dry biological membranes was by hydrogen bonding to the polar head group of the phospholipid membrane. In addition to possible hydrogen bonding to cell membranes, trehalose and sucrose as glass formers play a role in protecting cell membranes from fusion and breakage (Crowe et al. 1984). Such carbohydrates also stabilize proteins against denaturation during freezing and freeze-drying (Crowe et al. 1988; Franks et al. 1991). Zayed and Roos (2004) demonstrated that the capacity of lactic acid bacteria (LAB) to metabolize trehalose did not seem to confer greater resistance following use as a protectant during the freeze-drying process. Pehkonen et al. (2008) also showed that maximum ice formation during pre-freezing to satisfy the condition T ice < T m′ was more important than the use of lactose, trehalose or their mixture as cryoprotectant for Lactobacillus rhamnosus GG. Post-freeze-drying conditions require maintenance of the glassy state of cryoprotectants and possible cell constituents which typically require storage in hermetic packages in the absence of water, at almost “zero” water content and relative humidity of approximately 0 % (Pehkonen et al. 2008). Pehkonen et al. (2008) reported slightly improved stability for L. rhamnosus GG when freeze-dried with trehalose or lactose–trehalose compared to lactose. Such differences in storage stability with various cryoprotectants may depend primarily on the intracellular sugar composition, which may be strongly affected by cell activity in the cryoprotectant system under pre-freezing conditions (Pehkonen et al. 2008).

9.8 Intracellular Sugar Composition

Although there is limited knowledge regarding the intracellular sugar composition of living cells during fermentation and freezing, we analyzed sugar utilization by selected probiotic bacteria to evaluate the effects of cryoprotectants on their intracellular sugar composition for freezing and freeze-drying applications. L. acidophilus was able to ferment several sugars, while L. casei fermented lactose, trehalose, glucose and galactose. L. bulgaricus grew successfully in maltose, lactose and glucose. Lactobacillus viridescens showed positive results for maltose and glucose, whereas Lactobacillus sakei gave positive results for such sugars except lactose (Table 9.1).

Table 9.1 Sugars fermented by Lactobacillus species
Full size table

We further studied the sugar composition of Lactobacillus acidophilus and Lactobacillus casei cells grown with skim milk and various sugars containing growth media. Probiotic cultures were washed and sonicated to remove extracellular sugars and release cell contents, respectively. Sonicated L. acidophilus cells showed little or no maltose uptake, and glucose concentrations remained low. These data showed that upon entering the cell, maltose was immediately hydrolyzed into two glucose moieties and utilized by L. acidophilus. A significant accumulation of maltose within cells was found for L. casei.

Galactose , glucose and lactose content in both L. acidophilus and L. casei varied depending on the length of fermentation, as shown in Fig. 9.6. Lactose was readily accumulated over time in the cells of both strains. Lactose uptake more than quadrupled from 5 min to 1 h after inoculation of L. casei into 15 % (w/w) lactose solution, and the intracellular lactose concentration reached 10 g/100 g of solids at 48 h. In addition to the high uptake of lactose, its hydrolysis to glucose and galactose in L. casei cells was confirmed by high concentrations of intracellular monosaccharides, especially glucose. The glucose concentration increased steadily with time to a little over 6/100 g of solids, and the galactose reached a concentration of 1.7/100 g of solids at 48 h (Fig. 9.6).

Fig. 9.6
figure 6

Intracellular galactose (A), glucose (B) and lactose (C) content in L. acidophilus and L. casei cells during growth in 15 % (w/w) lactose solution

Full size image

Lactose uptake by the L. acidophilus cells was similar to that of the L. casei cells, but the accumulated concentration was not as high, with a maximum concentration of only 6.8/100 g of solution accumulating after 48 h. The concentration of glucose within the L. acidophilus varied.

Our study showed greater disaccharide utilization by L. acidophilus than by L. casei while variations in intracellular composition implied significant differences in intracellular freezing and freeze-drying properties. Sonication of the cells at different time points showed that lactose was accumulated almost immediately by L. acidophilus, and that its concentration was proportional to the length of time of immersion in the media. However, L. casei showed uptake of a larger concentration of lactose after 48 h than L. acidophilus, and also showed a much greater increase in monosaccharide formation. The other disaccharides showed a similar pattern, with the exception of maltose. The maltose concentration within the cells was quite low and remained constant up to 48 h; however, accumulation within the L. casei cells after 7 days suggested that maltose uptake was directly proportional to utilization needs, and only accumulated over time in a saturated environment.

9.9 Conclusions

Freezing and freeze-drying are important technologies for the stabilization of bioactive materials used in the food industry as ingredients or by consumers in foods and supplements. The effects of freezing have been studied in biological materials, but little information is available on the properties of water and ice in complex biological materials at low temperatures. A better understanding of complex carbohydrate–protein–lipid systems in freezing and freeze-drying is needed in order to improve the stability of bioactive materials during freezing, thawing, drying and reconstitution.