Model for heat and mass transfer in freeze-drying of pellets.

Lyophilizing frozen pellets, and especially spray freeze-drying, have been receiving growing interest. To design efficient and safe freeze-drying cycles, local temperature and moisture content in the product bed have to be known, but both are difficult to measure in the industry. Mathematical modeling of heat and mass transfer helps to determine local freeze-drying conditions and predict effects of operation policy, and equipment and recipe changes on drying time and product quality. Representative pellets situated at different positions in the product slab were considered. One-dimensional transfer in the slab and radial transfer in the pellets were assumed. Coupled heat and vapor transfer equations between the temperature-controlled shelf, the product bulk, the sublimation front inside the pellets, and the chamber were established and solved numerically. The model was validated based on bulk temperature measurement performed at two different locations in the product slab and on partial vapor pressure measurement in the freeze-drying chamber. Fair agreement between measured and calculated values was found. In contrast, a previously developed model for compact product layer was found inadequate in describing freeze-drying of pellets. The developed model represents a good starting basis for studying freeze-drying of pellets. It has to be further improved and validated for a variety of product types and freeze-drying conditions (shelf temperature, total chamber pressure, pellet size, slab thickness, etc.). It could be used to develop freeze-drying cycles based on product quality criteria such as local moisture content and glass transition temperature.

Effect of controlled ice nucleation on primary drying stage and protein recovery in vials cooled in a modified freeze-dryer.

The freezing step influences lyophilization efficiency and protein stability. The main objective of this work was to investigate the impact on the primary drying stage of an ultrasound controlled ice nucleation technology, compared with usual freezing protocols. Lyophilization cycles involving different freezing protocols (applying a constant shelf cooling rate of 1 degrees C/min or 0.2 degrees C/min, putting vials on a precooled shelf, and controlling nucleation by ultrasounds or by addition of a nucleating agent) were performed in a prototype freeze-dryer. Three protective media including sucrose or maltodextrin and differing by their thermal properties and their ability to preserve a model protein (catalase) were used. The visual aspect of the lyophilized cake, residual water content, and enzymatic activity recovery of catalase were assessed after each lyophilization cycle and after 1 month of storage of the lyophilized product at 4 degrees C and 25 degrees C. The freezing protocols allowing increasing nucleation temperature (precooled shelf and controlled nucleation by using ultrasounds or a nucleating agent) induced a faster sublimation step and higher sublimation rate homogeneity. Whatever the composition of the protective medium, applying the ultrasound technology made it possible to decrease the sublimation time by 14%, compared with the freezing method involving a constant shelf cooling rate of 1 degrees C/min. Concerning the enzyme activity recovery, the impact of the freezing protocol was observed only for the protective medium involving maltodextrin, a less effective protective agent than sucrose. Higher activity recovery results were obtained after storage when the ultrasound technology or the precooled shelf method was applied. Controlling ice nucleation during the freezing step of the lyophilization process improved the homogeneity of the sublimation rates, which will, in turn, reduce the intervial heterogeneity. The freeze-dryer prototype including the system of controlled nucleation by ultrasounds appears to be a promising tool in accelerating sublimation and improving intrabatch homogeneity.

Effect of ethanol, temperature, and gas flow rate on volatile release from aqueous solutions under dynamic headspace dilution conditions.

On the basis of a mechanistic model, the overall and liquid mass transfer coefficients of aroma compounds were estimated during aroma release when an inert gas diluted the static headspace over simple ethanol/water solutions (ethanol concentration = 120 mL x L(-1)). Studied for a range of 17 compounds, they were both increased in the ethanol/water solution compared to the water solution, showing a better mass transfer due to the presence of ethanol, additively to partition coefficient variation. Thermal imaging results showed differences in convection of the two systems (water and ethanol/water) arguing for ethanol convection enhancement inside the liquid. The effect of ethanol in the solution on mass transfer coefficients at different temperatures was minor. On the contrary, at different headspace dilution rates, the effect of ethanol in the solution helped to maintain the volatile headspace concentration close to equilibrium concentration, when the headspace was replenished 1-3 times per minute.

Mechanistic mathematical model for in vivo aroma release during eating of semiliquid foods.

The paper describes a mechanistic mathematical model for aroma release in the oropharynx to the nasal cavity during food consumption. The model is based on the physiology of the swallowing process and is validated with atmospheric pressure chemical ionization coupled with mass spectrometry measurements of aroma concentration in the nasal cavity of subjects eating flavored yogurt. The study is conducted on 3 aroma compounds representative for strawberry flavor (ethyl acetate, ethyl butanoate, and ethyl hexanoate) and 3 panelists. The model provides reasonably accurate time predictions of the relative aroma concentration in the nasal cavity and is able to simulate successive swallowing events as well as imperfect velopharyngeal closure. The most influent parameters are found to be the amount of the residual product in the pharynx and its contact area with the air flux, the volume of the nasal cavity, the equilibrium air/product partition coefficient of the volatile compound, the breath airflow rate, as well as the mass transfer coefficient of the aroma compound in the product, and the amount of product in the mouth. This work constitutes a first step toward computer-aided product formulation by allowing calculation of retronasal aroma intensity as a function of transfer and volatility properties of aroma compounds in food matrices and anatomophysiological characteristics of consumers.

Effect of product temperature during primary drying on the long-term stability of lyophilized proteins.

Our objective was to investigate the effect of performing primary drying at product temperatures below and above Tg' (glass transition temperature of the freeze-concentrated phase) on the long-term stability of lyophilized proteins. Two protective media differing in the nature of the bulking agent used (amorphous or crystalline) were selected. Several lyophilization cycles were performed by using various combinations of shelf temperature and chamber pressure to obtain different values of product temperature during primary drying. The antigenic activity of the proteins was measured after lyophilization and after 6 months of storage at 4 degrees C and 25 degrees C. After 6 months of storage and regardless of the protective medium, the losses of antigenic activity of both toxins increased from 0% when primary drying was performed at a product temperature lower than Tg' and to 25% when the product temperature was higher than Tg'. The use of partially crystalline systems makes it possible to withstand high primary drying temperatures (above Tg'). However, the shelf life of lyophilized proteins may be decreased when the amorphous phase including the protein and the stabilizing molecule changes to the viscous state.

Effect of sucrose and maltodextrin on the physical properties and survival of air-dried Lactobacillus bulgaricus: an in situ fourier transform infrared spectroscopy study.

The effect of sucrose, maltodextrin and skim milk on survival of L. bulgaricus after drying was studied. Survival could be improved from 0.01% for cells that were dried in the absence of protectants to 7.8% for cells dried in a mixture of sucrose and maltodextrin. Fourier transform infrared spectroscopy (FTIR) was used to study the effect of the protectants on the overall protein secondary structure and thermophysical properties of the dried cells. Sucrose, maltodextrin and skim milk were found to have minor effects on the membrane phase behavior and the overall protein secondary structure of the dried cells. FTIR was also used to show that the air-dried cell/protectant solutions formed a glassy state at ambient temperature. 1-Palmitoyl 2-oleoyl phosphatidyl choline (POPC) was used in order to determine if sucrose and maltodextrin have the ability to interact with phospholipids during drying. In addition, the glass transition temperature and strength of hydrogen bonds in the glassy state were studied using this model system. Studies using poly-L-lysine were done in order to determine if sucrose and maltodextrin are able to stabilize protein structure during drying. As expected, sucrose depressed the membrane phase transition temperature (Tm) of POPC in the dried state and prevented conformational changes of poly-L-lysine during drying. Maltodextrin, however, did not depress the Tm of dried POPC and was less effective in preventing conformational changes of poly-L-lysine during drying. We suggest that when cells are dried in the presence of sucrose and maltodextrin, sucrose functions by directly interacting with biomolecules, whereas maltodextrin functions as an osmotically inactive bulking compound causing spacing of the cells and strengthening of the glassy matrix.

Physical characterisation of formulations for the development of two stable freeze-dried proteins during both dried and liquid storage.

The development of stable freeze-dried proteins requires maintaining the physical and biological integrity of the protein as well as increasing the efficiency of the manufacturing process. Our objective was to study the effects of various excipients on both the physical characterisation and the dried and liquid stability of two proteins. Thermo-physical properties of 13 formulations were determined using both differential scanning calorimetry and freeze-drying microscopy. The antigenic activity was evaluated immediately after freeze-drying and after subsequent storage in both dried and liquid state. From the comparison between glass transition (T'g) and collapse (T coll) temperatures, we concluded that the collapse temperature was a more relevant parameter than T'g for freeze-drying cycle development and optimisation. One crystalline formulation composed of 4% mannitol and 1% of sucrose protected efficiently both proteins during subsequent storage in dried state (6 months at 25 degrees C) and in liquid state (3 months at 4 degrees C after rehydration). However, the freeze-drying behaviour of this crystalline formulation remained difficult to predict and control. On the other hand, two amorphous formulations composed of 4% of maltodextrin and 0.02% of Tween 80, or 5% of BSA preserved antigenic activity during storage in dried state. The glassy character of these formulations as well as their high collapse temperature values (-9 and -12 degrees C, respectively) should allow simplification and shortening of freeze-drying process.

1H nuclear magnetic resonance relaxometry study of water state in milk protein mixtures.

1H NMR signal was used to characterize highly hydrated milk protein dispersions (3-20% dry matter) with various micellar casein concentrations (3-15%), whey protein concentrations (0-3%), lactose concentrations (0-7.5%), CaCl(2) concentrations (0-2 mM), and pH (6.2-6.6). The results showed the predominant effect of micellar casein concentration on water state and were consistent with the three-site relaxation model in the absence of lactose. The relaxation rates observed for these dispersions were explained by the free water relaxation rate, the hydration water relaxation rate, and the exchangeable proton relaxation rate. Hydration water was found to be mainly influenced by casein micelle concentration and structure. The variations in hydration with pH were consistent with those observed for classical measurement of voluminosity observed at this range of pH. The effects of lactose and whey protein content are discussed.

Composition rather than viscosity modifies the aroma compound retention of flavored stirred yogurt.

The influence of thickening agents (modified starch/pectin mixture 0 and 7 g/L) and mechanical treatment (low, medium, and high) on the retention of esters (pentyl acetate and ethyl pentanoate), aldehydes (hexanal and (E)-2-hexenal), and a lactone (gamma-octalactone) in low-fat flavored stirred yogurts were investigated under equilibrium conditions. In the range studied, the thickening agent and mechanical treatment had little influence on aroma compound retention compared to the decreasing effect of increasing dairy protein concentration on aldehyde retention and the "salting out" effect of carbohydrates on esters. Moreover, experiments in dynamic mode (study of the release of hexanal when yogurts were heated) showed, in the conditions studied, that heat and mass transfer coefficients were not influenced by any of the studied factors (thickening agents and mechanical treatment). These results under static and dynamic conditions are not related to the significant decreasing effect of thickening agents on apple sensory scores associated with hexanal, observed in a previous sensory study. Thus, this sensory effect of thickening agents may be due to sensory interactions between perceptions rather than physicochemical interactions.

Collapse temperature of freeze-dried Lactobacillus bulgaricus suspensions and protective media.

Optimization of the freeze-drying process needs to characterize the physical state of frozen and dried products. A protocol to measure the collapse temperature of complex biological media such as concentrated lactic acid bacteria using freeze-drying microscopy was first elaborated. Afterward, aqueous solutions of one or several components as well as concentrated lactic acid bacterial suspensions were analyzed in order to study how the structure of these materials is degraded during freeze-drying. A similar behavior toward collapse was observed for all aqueous solutions, which was characterized by two temperatures: the "microcollapse" temperature (T(microc), beginning of a local loss of structure) and the "collapse" temperature (T(c), beginning of an overall loss of structure). For aqueous solutions, these two temperatures were close, differing by less than 3 degrees C. Nevertheless, when lactic acid bacteria were added to aqueous solutions, the collapse temperatures increased. Moreover, the interval between microcollapse and collapse temperatures became larger. Lactic acid bacterial cells gave a kind of "robustness" to the freeze-dried product. Finally, comparing glass transition, measured by differential scanning calorimetry (DSC) and collapse temperature for aqueous solutions with noncrystallizable solutes, showed that these values belonged to the same temperature range (differing by less than 5 degrees C). As suggested in the literature, the glass transition temperature can thus be used as a first approximation of the collapse temperature of these media. However, for lactic acid bacterial suspensions, because the difference between collapse and glass transition temperatures was about 10 degrees C, this approximation was not justified. An elegant physical appearance of the dried cakes and an acceptable acidification activity recovery were obtained, when applying operating conditions during freeze-drying in vials that allowed the product temperature to be maintained during primary drying at a level lower than the collapse temperature of lactic acid bacterial suspensions. Consequently, the collapse temperature T(c) was proposed as the maximal product temperature preserving the structure from macroscopic collapse and an acceptable biological activity of cells.

Collapse temperature of bacterial suspensions: the effect of cell type and concentration.

The characterisation of the physical state of frozen and freeze dried biological products delivers powerful information for freeze-drying process optimisation. The influence of lactic acid bacterial cell size, shape and concentration on collapse temperature of concentrated bacterial suspensions was investigated. Lactobacillus bulgaricus (long rods), and Streptococcus thermophilus (small spherical cells) were used as cellular models for this study. Whatever the strain, when lactic acid bacterial cells were added to protective solutions, the collapse temperature increased, thus allowing the use of higher sublimation temperatures during primary drying than expected from the protective medium alone. Moreover, the higher the cell concentration, the greater the effect, linear relationships existing between the collapse temperatures and the total dried matter. Cells of both strains gave a kind of robustness to the freeze-dried product, but the increase observed in collapse temperature was considerably higher (3 - 5 degree C) for L. bulgaricus compared to S. thermophilus. This result was ascribed to the different size and shape of the strains.