Food Stability Beyond Water Activity and Glass Transtion: Macro-Micro Region Concept in the State Diagram

Abstract

The water activity concept proposed that a food product is the most stable at its monolayer moisture content. Recently, the limitations of the water activity concept were identified and the glass transition concept was proposed in order to overcome the limitations of water activity. Based on the glass-transition concept, a food is the most stable at and below its glass transition point. Recently it has also become evident that the glass transition concept is not universally valid for stability determination when foods are stored under different conditions. The glass transition concept was used to develop the state diagram by drawing another stability map using freezing curve and glass line. Currently, other components indicating different characteristics are being included in the state diagram. It is being emphasized in the literature to combine the water activity and glass transition concepts. In this paper, an attempt is made to combine these two concepts in the state diagram and to propose a macro-micro region concept for determining the stability of foods.

Keywords:

• Isotherm
• Food stability
• Glass transition
• BET monolayer
• Food spoilage

INTRODUCTION

Humans in the Atone Age (10,000 years ago) stored nuts and seeds for winter and discovered that meat and fish could be preserved by drying in the sun. After the discovery of fire, cooking made food more appetizing and was an aid to preservation. Modern processing developed at the end of the 1700s when the Napoleonic wars raged. As Napoleon pushed forward into Russia, his army was suffering more casualties from scurvy, malnutrition, and starvation. The French government offered 12,000 francs to any one who could develop a method of preserving food. Nicolas Appert took up the challenge. He had a theory that if fresh foods were put in airtight containers and sufficient heat applied, then the food would last longer. Appert packed his foods in bottles, corked them, and submerged them in boiling water, thus preserving them without understanding of bacterial spoilage. After 14 years of experimentation, in 1809 he won the prize and this was given to him by Napoleon himself. A theoretical understanding of the benefits of canning did not come until Louis Pasteur observed the relationships between microorganisms and food spoilage some fifty years later.[1] Eventually canning, drying, and freezing technology evolved. Recently new preservation methods have emerged and been adopted by the food industry. A comprehensive complete coverage of preservation is available in the Handbook of Food Preservation and other references.[2–4]

Determining food stability from a scientific basis rather than an empiricism is a challenge to food scientists and engineers. Microbial death and growth kinetics, deteriorative physical, chemical and biochemical changes during processing and storage depends on many factors, such as water content, food composition, preservatives, pH, and environmental or processing factors (temperature, pressure, electro-energy, gases or vapors etc). In the 1950s, the concept of water activity was proposed to determine the stability of foods, and in the 1980s, significant data on food stability as a function of water activity was published. In order to avoid the limitations of water activity, the glass transition concept was extensively proposed in the 1980s, although this concept initially appeared in the literature in the 1960s. These two concepts provide a strong scientific basis of food stability during drying and freezing. The stability of foods is of paramount interest to both food scientists and engineers; and a better understanding of the factors controlling microbial stability or reaction rates is clearly needed.[2–6] In this paper, the concepts of water activity and glass transition are discussed along with their applications and limitations. In addition, an attempt was made to combine these two concepts in the state diagram by proposing a macro-micro region concept.

WATER ACTIVITY CONCEPT

In the 1950s, scientists began to discover the existence of a relationship between the water contained in a food and its relative tendency to spoil.[7] In the 1980s, Labuza and his group generated significant data on food stability as a function of water activity. They also began to realize that the active water could be much more important to the stability of a food than the total amount of water present. Thus, it is possible to develop generalized rules or limits for the stability of foods using water activity. For example, there is a critical water activity below which no microorganisms can grow, this value is about 0.6. Most of the pathogenic bacteria cannot grow below a water activity of 0.85; whereas most yeasts and molds are more tolerant to reduced water activity but usually no growth occurs below a water activity of about 0.62.

Microbial responses to low water activity are shown in Fig. 1. A well-established response to the temporary loss of turgor pressure after a hyper-osmotic shock (i.e., a reduction of water activity surrounding the cell) is osmoregulation. Exposure of microorganisms to lower water activity causes an instantaneous loss of water, which is accompanied by a decrease in the cytoplasmic volume called plasmolysis. This can also cause lysis. Hypo-osmotic shock generally results in minor changes in cell volume. On the other hand, hyper-osmotic shock causes considerable shrinkage of the cytoplasmic volume. If the osmotic shock is not too severe, after an extended lag phase, the cytoplasmic volume increases as a result of osmotic adjustments made by the cells.[8] After loss of turgor, the microbial cell raises level of the compatible solutes within the cells.[9] Compatible solutes including glycerol, sorbitol, cyclohexaneterol, erythritol, arabitol, or mannitol in fungal cells or amino acids such as proline, aminobutyric acid, and glutamic acid in bacteria. These compatible solutes assist to prevent dehydration, and thus facilitate metabolic activities necessary for growth. This results in an increase in internal osmotic pressure and restores turgor pressure. The type and amount of solute accumulated in cells is not only influenced by water activity, but also other factors in the environment, including those associated with preserving hurdles used to control microbial growth, such as pH and preservatives.[10,11] Besides the accumulation of compatible solutes, changes in the membrane lipid including structural changes in phospholipids and fatty acid were also observed.[12,13] The compatible solutes produced internally are highly soluble, pH neutral, and are usually end product metabolites. They can include sugars from the breakdown of carbohydrates, amino acids from protein degradation, and cations such as K⁺. Examples include betaine, trehalose, glycerol, sucrose, proline, chloline, carnitine, mannitol, glucitol, and ectoine. The cell membrane is selectively permeable to them, allowing the cytoplasmic pool to be determined by the external osmotic pressure.[14,15] The preferred exogenous bacterial-compatible solutes are glycine and betaine, which are found in higher plants, and the amino acid proline.[4] In many foods, however, peptides are more readily available than free amino acids and hence peptides have become an important source of both nutrients and compatible solutes.[16]

Figure 1 Microbial response at low water activity.

A food product is most stable at its monolayer moisture content, which varies with the chemical composition, structure and environmental conditions, such as temperature. However, in many instances the critical limit may also be observed at higher moisture than monolayer moisture content. Figure 2 shows the monolayer and multilayer water on a solid surface.[17] This principle was the main reason why food scientists started to emphasize water activity rather than total water content. Since then, the scientific community has explored the significance of water activity in determining the physical characteristics, processes, shelf life, and sensory properties of foods. It is now used to predict the end point of drying, in process design and control, in ingredient selection, to predict product stability and to make packaging selection. One of the earlier food stability maps based on the water activity concepts considers growth of micro-organisms and different types of bio-chemical reactions.[18,19] The updated food stability map is presented in Fig. 3. In this present map, the trends of microbial growth, bio-chemical reactions and mechanical characteristics are presented in the 3 zones of water activity. In general the rule of water activity concept is: Food products are most stable at its monolayer moisture content or monolayer water activity and unstable above or below monolayer. A water activity map also provides the following rules: below monolayer (zone I):(i) stability can be decreased with decreasing water activity, for example fat oxidation (line ob); (ii) stability can remain constant (line ab); and (iii) stability can increase (line nb). In the adsorbed layer (zone II): (i) stability can increase (line bp); and (ii) stability can decrease (line bc). In multi layer (zone III): stability can increase, decrease, or remain constant.

Figure 2 Concepts of BET mono and multi layers water.

Figure 3 Updated stability diagram based on the water activity concepts.

Recently, the limitations of water activity concept were identified[7,20–24] as: (i) Water activity is defined as at equilibrium, whereas foods may not be in a state of equilibrium, for example low and intermediate moisture foods. (ii) The critical limits of water activity may also be shifted to higher or lower levels by other factors, such as pH, salt, anti-microbial agents, heat treatment, electromagnetic radiation, and temperature. (iii) Nature of the solute used to reduce water activity also plays an important role, for example some solutes are more inhibiting than others even at the same water activity (minimum growth of P. fragi is 0.96 using sodium chloride, while for glycerol water activity is 0.94.[25] (iv) Water activity concept does not indicate the mobility or reactivity of water and the nature of binding to the substrate. It only provides the information on the amount of strongly bound and free water without their precise reactivity. (v) Many physical changes, such as crystallization, caking, stickiness, gelatinization, diffusivity could not be explained based on the water activity alone. However, these limitations do not completely invalidate the concept but rather make it difficult to apply universally. In order to find other alternatives, the glass transition concept was postulated in the literature.

GLASS TRANSITION CONCEPT

Glassy materials have been known for centuries but the glass transition concept was first applied to foods with scientific understanding in the 1980s.[26] A low glass transition means that at room or mouth temperature, the food is soft and relatively plastic, and at higher temperatures it may even flow. In contrast, a food with a high glass transition temperature is hard and brittle at ambient temperature. Early attempts to describe glassy phenomena concluded that glass is a liquid that has lost its ability to flow, thus instead of taking the shape of its container, glass itself can serve as the container for liquids. Food materials are in an amorphous or non-crystalline state below the glass transition temperature and are rigid and brittle. Glasses are not crystalline with a regular structure, but retain the disorder of the liquid state. Physically it is a solid but resembles closely a thermodynamic liquid. Molecular mobility increases 100-fold above glass transition. In kinetic terms, Angell [27] described a glass as any liquid or super-cooled liquid whose viscosity is between 10¹² and 10¹³ Pa s, thus effectively behaving like a solid, which is able to support its own weight against flow due to gravity. To put this viscosity into context, a supercooled liquid with a viscosity of 10¹⁴ Pa s would flow 10⁻¹⁴ m/s in the glassy state, in contrast the flow rate of a typical liquid is in the order of 10 m/s. In other words, a glass is a liquid that flows about 30 μm in a century.[28] This is evidenced by the fact that ancient stained glass windows are thicker at their base due to flow under gravity.[29]

The early papers related to glass transition in food and biological systems appeared in the literature in the 1960s.[30–32] White and Cakebread[31] first highlighted the importance of the glassy state of foods in determining its structural stability. They were perhaps the first food scientists to discuss the importance of the glassy and rubbery states in relation to the collapse of a number of high solid systems. The significant applications of the glass transition concept emerged in food processing in the 1980s, when Levine and Slade [33] and Slade and Levine[34] identified its major merits and wide applications. In the 1990s, Roos, Karel, and other groups generated significant data on the glass transition and components (i.e., characteristic curves) of state diagrams for a number of food ingredients. In 1990s, Chirife, Buera, Bell, Karel, Roos, Labuza, and others started to present data on food stability based on the glass transition and water activity concepts. It has been mentioned in the literature that foods can be considered very stable at the glassy state, since below glass temperature compounds involved in the deterioration reactions take many months or even years to diffuse over molecular distances and approach each other to react.[35] A hypothesis has recently been stated that glass transition greatly influences food stability, as the water in the concentrated phase becomes kinetically immobilized and therefore does not support or participate in reactions. Formation of a glassy state results in a significant arrest of translational molecular motion, and chemical reactions become very slow[36]. The rules of glass-transition concept are: (i) Foods are most stable at and below glass transition; and (ii) the higher the T-T_g or T/T_g (i.e above glass transition), the higher the deterioration or reaction rates. These conditions identify two macro regions (i.e., 3 regions above and one region below) in the state diagram (Fig. 4). Similarly, mechanical and transport properties could also be related with glass transition. It is very interesting to see that this concept has been so widely tested in foods as evident from the literature. In many instances, glass transition concept does not work alone, thus it is now being recommended to use both the water activity and glass transition concepts in assessing process-ability, deterioration, food stability, and shelf-life predictions.[37]

Figure 4 State diagram proposed by Levine and Slade showing 5 macro-regions.

GLASS TRANSITION AND GLASSY STATE

Phase transitions in foods can be divided into two groups: first-order and second-order. At the first-order transition temperature, the physical state of a material changes isothermally from one state to another (e.g., solid to liquid, liquid to gas) by release or absorption of latent heat (e.g., melting, crystallization, condensation, evaporation). Second-order transition occurs (e.g., amorphous state to glassy state) without release or absorption of latent heat.[38] Glass transition is a second-order, time-temperature dependent transition, which is characterized by a discontinuity or change in slope of physical, mechanical, electrical, thermal, and other properties of a material, when plotted as a function of temperature.[38] This transition is most commonly identified as a shift in the thermogram line measured by Differential Scanning Calorimetry (DSC). The process is considered to be a second order thermodynamic transition in which the material undergoes a change in state but not in phase. It is more meaningfully defined as analogue as nature to second-order, since each measurement technique is based on monitoring change in a specific property, and since change or break in properties are achieved within a certain temperature range rather than at a specific temperature.[22] A perfect second-order transition occurs at a specific temperature. During heating, glass transition of food materials does not occur at a fixed point with the change of specific heat. Instead networks soften or transform over quite a large temperature range.[29] However, glycerol produces a step change in heat capacity as a function of temperature at 190 K.[39] Figure 5 shows the ideal second order transition (A) (i.e., shift at a specific temperature) and non-ideal second order transition (i.e., shift over a temperature range); (B) as evident by DSC thermogram.

Figure 5 Second order transition in foods identified by DSC thermogram. A: ideal second order transition, B: non-ideal second order transition.

STATE DIAGRAM AND ITS COMPONENTS

State Diagram

A state diagram is another stability map of different states of a food as a function of water or solids content and temperature.[40] Levine and Slade[33] presented a state diagram of providone N-vinyl pyrrolidone (PVP) illustrating glass line, freezing curve, and intersection of these lines as T_g ′ (in this paper, the author is noted as T_g ′′). The intersection was identified by extrapolation of extending the freezing curve by maintaining a similar curvature. This is most probably the first state diagram in the food science literature. Many of his publications later presented[34] the state diagram of starch by quoting the source reference of van den Berg.[41] Later he presented another state diagram showing glass line, freezing curve, melting line, eutectic point, T_g ′ and vapor line.[42] Starting from the 1990s, Roos, Karel, Kokini, and others presented state diagrams of a number of food components and food products. The main advantages of drawing a map are to help understanding the complex changes when the water content and temperature of foods are changed.[37,43–49] It also assists in identifying stability of foods during storage as well as selecting a suitable condition of temperature and moisture content for processing.[21,50–5] The regions of drying and freezing can be easily visualized in the diagram, and product stability can be assessed based on moisture content and temperature. Figure 6 shows a state diagram indicating the different states as a function of temperature and solute mass fraction (updated from Rahman [22]).

Figure 6 State diagram showing different macro-micro regions (updated from Rahman [22]).

Components of State Diagram

Earlier state diagrams were constructed with only the freezing curve and glass transition line. Recently, attempts have been made to add other structural changes along with the glass line, such as solubility line. Numbers of micro-regions and new terminologies are being included in constructing the state diagram. The presented state diagram shown in Fig. 6 is updated from Rahman.[22,40] In Fig. 6, the freezing line (ABC) and solubility line (BDL) are shown in relation to the glass transition line (EQFS). Point F (X_s ′ and T_g ′) which is lower than T_m ′ (point C) is a characteristic transition (maximal-freeze-concentration condition) in the state diagram, defined as the intersection of the vertical line from (T_m ′) _a to the glass line EQFS.[22] At this maximal-freeze-concentration, all possible freezable water is transformed into ice. The water content at point F or C is considered as the un‐freezable water (i.e., 1- X_s ′). The un-freezable water mass fraction is the amount of water remaining unfrozen even at very low temperature. It includes both un-crystallized free water and bound water attached to the solids matrix. Point Q is defined as T_g ′′ and X_s ′′ as the intersection of the freezing curve to the glass line by maintaining the similar curvature of the freezing curve ABC.

The apparent glass transition (T_g ′′′) _a of the solids matrix in the frozen sample is usually determined by Differential Scanning Calorimetry (DSC) below (T_m ′) _a (Fig. 7A). This is due to the formation of a similar solid matrix associated with un-freezable water and transformation of all free water into ice, although the sample contains different levels of total water before the start of DSC scanning.[52] The values of apparent (T_g ′′′) _a and (T_m ′) _a decreased with increasing solids content and reached to a nearly constant value. The intersections of the lines in Fig. 7B show the T_g ′′′ (point R in Fig. 6) and T_m ′ (point C in Fig. 6). If the sample containing moisture close to the points R and C, these apparent values become similar to T_g ′′′ and T_m ′ as marked in Fig. 7.

Figure 7 Identification of (T_m ′) _a , (T_g ′′′) _a , T_m ′, T_g ′′′, T_g ′′, or T_g ′ in the state diagram.

In the region AGB as shown in Fig. 6, the phases present are ice and solution. Below point B, the first crystallization of solute occurs, transforming the GBCH region to three states: ice, solution and solute crystal. There is no free water (i.e able to form ice) to the right side of point C (T_m ′, end point of freezing with maximal freeze-concentration-condition) and below the very concentrated solution is transformed to the rubber state. The maximal-freeze-concentration condition can be achieved using optimum conditions by slow cooling and/or annealing of the samples so that all freezable water can be transformed into ice. The region HCRI contains ice, rubber, and solute crystal. Point F is the T_g ′, below this point all portion of the rubber state is transformed to the glass state, thus the region KFS contains glass, ice, and solute crystal. The rate of cooling can shift the points B, C, Q and F. More detailed effects of cooling on the shift are discussed by Rahman.[40] The glass line should follow the line EQFS if maximum ideal plasticization occurs in the solid matrix with the addition of water. The line EQF could be shifted upward if water causes different degrees of plasticization (as shown by line EP) in the sample containing un-freezable water. The vertical line from F intersects at point P and the temperature is defined as T_g ^iv . In this situation, fitting the glass transition data to the Gordon-Taylor equation, considering the two ends as T_gs and T_gw , may not be valid. Based on the author's hypothesis of different characteristic temperatures T_g ^iv, T_m ′, T_g ′′′, T_g ′, and T_g ′′, the modified equation could be proposed as:

(1)

where X_w and X_s are the mass fraction of water and solids; T_c and T_gs are the characteristic temperature and glass transition of solids; and k_c is the characteristic of the material, respectively. In the case of the Gordon-Taylor equation, the characteristic temperature is the glass transition of water instead of T_c (i.e., T_g ^iv, T_m ′, T_g ′′′, T_g ′, or T_g ′′).

The region BQEL is important in food processing and preservation, many characteristics, such as crystallization, stickiness, and collapse are phenomena that are observed in this region.[37,53] In case of cereal proteins, Kokini et al.[48] determined an entangled polymer flow region and a reaction zone based on the mechanical characteristics. The line BDL is the melting line which is important when products are exposed to high temperature during processing, for examples frying, baking, roasting, and extrusion cooking. In the case of a multi-component mixture such as food a clear melting is difficult to observe at high temperature due the reactions or interactions between the components. In this case, Rahman.[40] and others have defined melting as the decomposition temperature. Line MDL is the boiling/evaporation line for water from the liquid phase (line MD) and solid matrix with a degree of saturation with water (line DL). It is possible for the melting and evaporation lines to intersect since water evaporation could happen in a saturated matrix before melting of the solid matrix mixed with water.

It was identified that further work is required on the relationships between glass transition, water activity and food stability.[3] It appears that the interrelationships can be very complex, depending on the complexity of the food system and on the type of stability being studied. Recently many papers have presented data on both water activity and glass transition as a function of water content. However the link between them that determines stability has not been identified. Karel et al.[54] attempted to relate water activity and glass transition by plotting equilibrium water content and glass transition as a function of water activity. By drawing a vertical line on the graph, stability criteria could be determined from the moisture isotherm curve and the glass transition line. At any temperature (say 25^oC), stability moisture content determined from the glass transition line was much higher than the stability moisture from the isotherm. Similarly Sablani et al.[55] predicted stability criteria (solids content) from an isotherm model (GAB equation), considering water activity 0.20, and then estimate the solids content from glass model (Gordon-Taylor) considering storage temperature as T_g . They found a surprising gap between the predictions of the stable solids contents using the two different approaches for a number of food products. At present it is a challenge to link them in a meaningful way.

As a first attempt, Rahman[22] plotted the BET-monolayer value as the LNO line in the state diagram shown in Fig. 6. It intersects at point N with the glass line EQFS, which shows that at least in one location (point N), glass and water activity concepts provide the same stability criteria. This also justifies the variability of deteriorations observed by Sablani et al.[55] This approach forms more micro-regions, which could give different degrees of stability in the state diagram. More studies regarding stability need to be done on both the left (above and below glass) and right sides (above and below) of the line LNO. It should be mentioned here that the BET-monolayer could be achieved by mainly removing water from a system (since the isotherm is relatively unaffected by temperature), but glassy state could be achieved by removing water through drying, as well as by decreasing the temperature of the system. A successful combination of water activity and glass transition could develop a more in depth knowledge on stability criteria. In addition, how other factors, such as pH, and preservatives act could be linked with these concepts. At present the scientific community far from developing unified theory.

It is evident from the literature that stability below or above glass transition varies even based on specific cases, indicating that applying only the glass transition temperature for developing the stability rule is not enough. Samples with freezable water are more complex and four characteristic temperatures were defined by Rahman et al.[52] In current paper a total of five temperatures are defined as T_g ^iv  > T_m ′ > T_g ′′′ > T_g ′ > T_g ′′. In addition, the stability below or above T_g ^iv, T_m ′, T_g ′′′, T_g ′, and T_g ′′ should also be explored. There are only a few references available that include all four characteristic temperatures with their moisture content. It is important to know how these temperatures affect the stability of foods. It would be interesting to explore the differences in stability in product within these different ranges.[22]

VALIDITY OF GLASSY STATE

Applications of glassy state concept in food systems are thoroughly reviewed by Kalichevsky-Dong[3] and Rahman.[22] They grouped the applications of glassy state in the processes of controlling diffusion process, structure, crystallization, stickiness, grain damage, pore formation, microbial stability, seed stability, oxidation, non-enzymatic browning, enzymatic reaction, protein denaturation, hydrolysis, and enzyme inactivation. Most of the literature used two criteria to test the validity specifically: (i) whether food is stable if it is in the glassy state or unstable if it is above the glass transition; and/or (ii) whether the change in attributes above glass is related to the (T-T_g ) or T/T_g . It is clear from the literature that all experimental results could not be explained by the above rules, thus it is now important to identify when it fails and why. It is also a challenge to combine other concepts with glass transition.

The water activity concept is based on the binding nature of water molecules in the matrix. When water is bound (i.e., unavailable to take part in reactions) to the solid matrix or non-solvent, then no deterioration reactions could be expected. The glass transition concept is based on the molecular mobility of the reacting components in the matrix, thus diffusion of the reactants through the system to take part in reactions is very slow and stability is achieved. Although combining both the concepts could be a powerful tool for stability determination. A successful combination of water activity and glass transition could open more precise and unified determination of stability criteria. Attempts should be made to provide alternative solutions to combine both concepts. In fact there are other factors which play a role and in many instances it is incorrect to say even combining both would give complete stability. However, other building blocks could be added. Stability could be determined from boundary modeling and then kinetics prediction models.

MACRO-MICRO REGION CONCEPT

BET-monolayer line as LNO in the state diagram (shown in Fig. 6) makes four regions: below BET-monolayer, one above and one below; and above BET-monolayer, one above and another below.[22] This approach forms more micro-regions, which could give different stability in the state diagram and could explain the limitations of each concept. The present hypothesis proposed thirteen micro-regions having highest to lowest stability based on the locations from glass and BET-monolayer lines. For example, the region-1 (relatively non-reacting zone, below the BET-monolayer line and glass line) is the most stable and region-13 (highly reacting zone, far from BET-monolayer line and glass line) is least stable. The stability decreased as the zone number increased. Each micro-region could be studied for specific deterioration separately, considering different characteristics of a medium.

It is clear that the stability based on macro-regions in the state diagram considering below and above BET-monolayer or glass transition would not be enough to determine stability. The literature showed that many systems were stable above glass transition and unstable below glass transition. One of the arguments was proposed that mobility of water could also occur below glass transition. Based on the variability in the stability, micro-regions were identified within the macro-regions. In the state diagram, 3 factors are clearly identified or mapped: (i) how far it is from the BET-monolayer line; (ii) how far it is from the glass transition line; and (iii) how low is the temperature. In addition to the binding nature of water (water activity) and mobility of reactants (glass transition), sample temperature could be identified as a separate factor in the state diagram, considering temperature dependent reaction rates. The reactivity does not increase only based on a function of (T-T_g ) or T/T_g and variation in increasing or decreasing behavior could be observed, for example micro-region 10 is below the BET-monolayer and above the glass transition and could be very reactive. The micro-region 1 is considered most stable since it is below the glass transition and BET-monolayer, and low temperature storage. The most unstable micro-region is region 13 since it is the most reactive. In this case, there is no point in applying the concepts of glass transition or water activity alone and other preservation hurdles must be used. The micro-region could also be sub-divided further to explain the variability of stability, for example regions between 3 to 9 based on T_u, T_m ′, T_g ′′′, T_g ′′, or T_g ′. This could open another dimension to the food stability map with more micro regions. Similarly regions 10 and 11 could be divided based on T_g and T_g ^iv for the samples containing no freezable water. Another important aspect in stability is to study the types of glass formed in foods along with their characteristics, such as fragile/strong glass,[56, 57] degree of plasticization,[57] strength and density of hydrogen bonds [58], variability in molecular mobility and/or relaxation below glass,[59,60] and complex matrix of crystalline/semi-crystalline/amorphous regions. There is a definite need to clarify the validity of each concept by identifying the valid and invalid conditions and systems.

COMBINING MULTICONCEPTS

IFT[6] presented the factors that influence the microbial growth in foods. In this report the multi-factor nature of microbial stability was identified, and how two factors, pH and water activity interact each other were described. The USDA pathogen-modeling program was used to identify critical limits of water activity and pH when both factors are interacting. They also identified that a general model for foods to cover all interactions of atmospheric gases and/or preservative combinations with pH and water activity does not currently exist. In reality the problem of stability determination could not be solved by identifying macro-regions or micro-regions, but it could help in developing scientific, systematic and rational approaches to determine the stability. The next task is to test the stability in each micro region and there may be a possibility to explore more generic rules for stability of individual stability criteria in the micro-region. A knowledge-based approach could be used to identifying how other factors, such as pH, preservatives, and types of solutes affect the stability in each micro-region.

KNOWLEDGE BASE OR DATA MINING

The author believes that developing a knowledge base of the stability for each macro and micro region would be a valued approach for exploring further generic rules for stability. He is confident that a data mining approach and/or a boundary modeling technique could be used to explore and to develop the generic rules in the future. This database, if developed, could be the foundation of new theoretical progress.

CONCLUSION

The state diagram could be used to combine water activity and glass transition concepts by including BET-monolayer and glass lines. In addition different macro- and micro-regions could also be drawn to identify food stability. In this paper 13 micro regions are proposed for determining food stability. A library of knowledge needs to be developed for each macro- and micro-region for developing further generic rules.

ACKNOWLEDGMENTS

This paper was presented in the 18^th International Congress of Chemical and Process Engineering (CHISA 2008), 24–28 August 2008, Prague, Czech Republic as a keynote lecture. The author would like to acknowledge the support of Sultan Qaboos University towards this research in the area of food structure and stability. He is grateful to all members of his research group for their continued support and encouragement. Special thanks to Dr. Ann Mothershaw for checking the clarity of the paper.