Dynamic flavor release from sucrose solutions

The initial dynamic flavor release from sucrose solutions was modeled. Modeling was based on the theoretical hydration behavior of sucrose, theoretical physicochemical data of flavor volatiles, and process parameters of a headspace apparatus used for model validation. The rate-limiting factor determining the initial flavor release was the hydration of sucrose, which in turn depends on the molarity of sucrose in the solution and, therefore, on the actual amount of nonbound water. Improved solubility of the more hydrophilic compounds due to their orientation toward the hydration shells of the sugar molecules was considered. The viscosity of nonassociated water forming the microregion for mass transfer of volatiles was considered instead of the bulk solution viscosity. Experimental validation of the model by real-time measurements of dynamic flavor release using foodlike flavor concentrations confirmed the above theory. Increasing sucrose concentrations resulted predominantly in increased flavor release, and bulk solution viscosity showed no effect.     

Flavor release from iota-carrageenan matrices: a kinetic approach

Release of aroma compounds in selected iota-carrageenan systems was studied by static headspace analysis. By varying the sodium chloride content, different rheological behaviors were obtained ranging from solution to gel. From the release curves, mass transfer (h(D)) and partition coefficients (K(ga)) of ethyl butanoate, ethyl hexanoate, and linalool were extracted using a mathematical model based on the penetration theory. This model, previously developed for flavor release from stirred solutions, was found to fit well the data obtained from structured systems (nonstirred conditions) at the beginning and at the end of the release curves: this allowed the determination of h(D) and K(ga). Matrix effects appeared to be dependent on the chemical class of the compounds. For the alcohol, the main effect on both equilibrium partitioning and mass transfer across the interface was ascribed to a salting effect. In the opposite, for esters, iota-carrageenan addition induced an increase of aroma retention and also a slower transfer across the interface. The respective effects of an increasing viscosity of the medium and of the formation of a tridimensionnal network are discussed.     

Partition and release of 21 aroma compounds during storage of a pectin gel system

The increasing popularity of low-fat products increases the need for a better understanding of how flavor release is affected by partial substitution of fat with hydrocolloids. Partitioning and release of aroma compounds from four pectin gels with different compositions were studied with static headspace and with a model mouth. Air/product partition coefficients determine the potential extent of aroma release, and mass transfer determines the rate at which aroma compounds are released to the vapor phase. This study showed that the gel network had large effects on the partition of aroma compounds between the gel and vapor phase. The specific properties of the aroma compounds were also of importance for the air/gel partition. Storage of the four gels showed that one of the weaker gels was influencing the concentration of aroma compounds in the headspace, probably caused by formation of a denser network over time.     

Effect of xanthan on flavor release from thickened viscous food model systems

The influence of xanthan concentration (0, 0.02, 0.1, 0.4, and 0.8% w/w) and bulk viscosity on the release of 20 aroma compounds of different chemical classes (5 aldehydes, 4 esters, 5 ketones, 3 alcohols, and 3 terpenes) was evaluated in xanthan-thickened food model systems having different viscosities. Interactions between flavor compounds and xanthan were assessed by measuring air-liquid partition coefficients, K, of aroma compounds in pure water and in the xanthan solutions by static headspace gas chromatography. Mass transfer of aroma compounds was estimated by dynamic headspace gas chromatography. Notably, limonene and some of the esters and aldehydes exhibited decreased K values in the presence of xanthan, indicating that the release of these volatile aroma compounds was reduced due to interaction with the xanthan matrix. The degree of interaction depended on the physicochemical characteristics of the aroma compounds. A similar tendency was observed at nonequilibrium with the decreases in release rates being most pronounced for limonene, followed by the esters and aldehydes, with no effect for ketones and an apparent "salting out" effect for alcohols. The reduction in flavor release by xanthan was thus dependent on the physicochemical properties of the aroma compounds and was apparently a result of the aroma-xanthan interactions and not influenced by the viscosity of the system itself.     

Interactions between methyl ketones and beta-lactoglobulin: sensory analysis, headspace analysis, and mathematical modeling

Interaction of flavor compounds with proteins is known to have an influence on the release of flavor from food. Hydrophobic interactions were found between beta-lactoglobulin and methyl ketones; the affinity constant increases by increasing the hydrophobic chain. Addition of beta-lactoglobulin (0.5 and 1%) to aroma solutions (12.5, 50, and 100 microL L(-)(1)) of three methyl ketones induces a significant decrease in odor intensity. The chosen methyl ketones were 2-heptanone (K(b) = 330), 2-octanone (K(b) = 950), and 2-nonanone (K(b) = 2440). The release of these flavor compounds (50 microL L(-)(1)) was studied by static headspace in water solution (50 mM NaCl, pH 3) with different concentrations of beta-lactoglobulin (0, 0.5, 1, 2, 3, and 4%). Increasing the concentration of protein increases the retention of volatiles, and this effect is greatest for 2-nonanone, the compound with the highest affinity constant, and lowest for 2-heptanone. A mathematical model previously developed to describe flavor release from aqueous solutions containing flavor-binding polymers (Harrison, M.; Hills, B. P. J. Agric. Food Chem. 1997, 45, 1883-1890) was used to interpret the data. The model assumes that the polymer-flavor interaction is reversible and the rate-limiting step for release is the transfer of volatiles across the macroscopic gas-liquid interface. This model was used to predict the equilibrium partitioning properties and the rate of release of the three methyl ketones. Increasing the affinity constant leads to decreased release rates and a lower final headspace aroma concentration.     

In vitro measurement of volatile release in model lipid emulsions using proton transfer reaction mass spectrometry

The presence of fat in food plays an important role in the way aroma is released during consumption and in the creation of the overall sensory impression. Fat acts as a reservoir for lipophilic volatile compounds and modulates the timing and delivery of aroma compounds in a unique manner. Despite considerable research, reproducible in vitro methods for measuring the effect of fat on volatile release are lacking. An open in vitro cell was used to simulate the open human naso-oropharygeal system and was interfaced with a proton transfer reaction mass spectrometer (PTR-MS) to examine some of the fundamental effects of fat on dynamic volatile release in liquid fat emulsions. Lipid emulsions with various fat contents (0-20%) and droplet sizes (0.25, 0.5, and 5.0 μM) were spiked with flavor volatiles representing a range of lipophilicity (K(o/w) = 1-1380). Preloaded syringes of spiked emulsion were injected into the cell, and temporal changes in release were measured under dynamic conditions. Significant differences in release curves were measured according to the lipid content of emulsions, the vapor pressure, and K(o/w) values of the volatile compounds. With increasing addition of fat, the critical volatile release parameters, maximum concentration (I(max)), time to maximum concentration (T(max)), and the integrated area under the concentration curve (AUC), were affected. The in vitro curves were reproducible and in agreement with theory and correlated with the preswallow phase of in vivo release data. An exponential model was used to calculate changes in mass transfer rates with increased fat addition.     

Experimental and modeling studies showing the effect of lipid type and level on flavor release from milk-based liquid emulsions

The purpose of this work was to study two key parameters of the lipid phase that influence flavor release-lipid level and lipid type-and to relate the results to a mass balance partition coefficient-based mathematical model. Release of 10 volatile compounds from milk-based emulsions at 10, 25, and 50 degrees C was monitored by 1-min headspace sampling with a solid-phase microextraction fiber, followed by GC-MS analysis. As compared to the observations for milk fat, changing to a lipophilic lipid (medium-chain triglycerides, MCT) and adding a monoglyceride-based surfactant did not influence the volatiles release. However, increasing the solid fat content was found to increase the release. At 25 degrees C, and even more so at 10 degrees C, concurrent with an increase in their solid fat content, hydrogenated palm fat emulsions showed increased flavor release over that observed for emulsions made with coconut oil, coconut oil with surfactant, milk fat, and MCT. However, at 50 degrees C, when hydrogenated palm fat emulsions had zero solid fat content, there was no difference in flavor release from that observed for milk fat emulsions. Varying milk fat at nine levels between 0 and 4.5% showed a systematic dependence of the release on the lipid level, dependent on compound lipophilicity. Close correlations were found between the experimental and model predictions with lipid level and percent liquid lipid as variables.     

Headspace GC and sensory analysis characterization of the influence of different milk additives on the flavor release of coffee beverages

Previous investigations of coffee flavor have been confined to the analysis of the aroma substances. These investigations showed that about 30 volatile compounds were substantially responsible for the coffee flavor. The aim of this study was to investigate the influence of different milk additives and one coffee whitener on the release of flavor impact compounds from coffee beverages. For the investigation of these effects an external static headspace technique was developed. With this technique the most potent odorants of the coffee beverage were determined. Analyses were performed by gas chromatography/olfactometry, flame ionization detection, and mass spectrometric detection. In addition, sensory studies of the odor profiles were performed. Milk and vegetable products as additives for coffee beverages affected the release of aroma substances in the brew through their lipid, protein, and carbohydrate components. All beverages with an additive showed reduced, but typical, odor profiles for each additive.     

In vivo flavor release from gelatin-sucrose gels containing droplets of flavor compounds

Gelatin-sucrose gels containing the same amount of flavor compounds present as either suspended droplets or homogenously distributed in the gel (dissolved) were eaten, and the in vivo flavor release was studied using atmospheric pressure chemical ionization-mass spectrometry. The maximum intensity of release was higher from all droplet-containing samples as compared with the dissolved sample (by a factor of 4-2500-fold). When the flavor was dispersed as a greater number of smaller droplets rather than one 1 microL droplet, the intensity of in vivo release was slightly lower. The release of 16 of the flavor compounds varied in their Log P (range 0.26-4.83) and vapor pressure (Log vapor pressure ranged from -1.09 to 1.99). The differences in release for flavors present as either droplets or dissolved in the gel matrix were strongly influenced by both of these factors. This suggested a different mechanism for flavor release from droplets as compared to the classical partition mechanism established for dissolved flavors.     

Volatile release from aqueous solutions under dynamic headspace dilution conditions.

Static equilibrium was established between the gas phase (headspace) and an unstirred aqueous phase in a sealed vessel. The headspace was then diluted with air to mimic the situation when a container of food is opened and the volatiles are diluted by the surrounding air. Because this first volatile signal can influence overall flavor perception, the parameters controlling volatile release under these conditions are of interest. A mechanistic model was developed and validated experimentally. Release of compounds depended on the air-water partition coefficient (K(aw)) and the mass transport in both phases. For compounds with K(aw) values <10(-)(3), K(aw) was the factor determining release rate. When K(aw) was >10(-)(3), mass transport in the gas phase became significant and the Reynolds number played a role. Because release from packaged foods occurs at low Reynolds numbers, whereas most experiments are conducted at medium to high Reynolds numbers, the experimentally defined profile may not reflect the real situation.     

Verification of a mouth simulator by in vivo measurements

The volatile content of the effluent from the retronasal aroma simulator (RAS) was compared with that of human breath using mass spectroscopy (MS-Nose). The ratios of volatile compounds from the RAS were closely related to those from the panelists' breath with the correlation coefficients ranging from 0.97 to 0.99 from model food systems. A greater sensitivity using the RAS was achieved because higher concentrations of volatiles in the MS-Nose were produced from the RAS than from the breath. In analyzing the effects on volatility of RAS parameters including airflow rate, temperature, saliva ratio, and blending speed, airflow rate had the greatest effect. The correlation coefficients for the real food systems studied ranged from 0.83 to 0.99. The RAS gives a good approximation of time-averaged flavor release in the mouth as defined by breath-by-breath measurements.     

The influence of gel strength on aroma release from pectin gels in a model mouth and in vivo, monitored with proton-transfer-reaction mass spectrometry

The course of events from taking a food into the mouth to the perception of the food's flavor involves many steps, from dilution with saliva, mastication, and transportation of the compounds to the olfactory epithelium to transformation into signals that go to the brain. In addition, there are also the effects of the food's structure and properties. In this study, a proton-transfer-reaction mass spectrometer (PTR-MS) was used to investigate how four pectin-containing systems with different structures and strengths affected the release of aroma compounds in a model mouth and in the nose of an assessor. Both the model mouth and the in-nose measurements showed that the strength and structure of pectin-containing systems are important with regard to the quantity of aroma compounds that are released. Mastication and saliva were also shown to have a large influence on how much of the aroma compound is released from the mouth to the nose.     

Volatile flavor analysis and sensory evaluation of custard desserts varying in type and concentration of carboxymethyl cellulose

The influence of type and concentration of carboxymethyl cellulose (CMC) on flavor and textural properties of custard desserts was examined. A synthetic strawberry flavor mixture was used to flavor the custards; it comprised 15 volatile flavor compounds. The viscosity of the custards was determined using rheometric measurements. Static headspace gas chromatography and in-nose proton transfer reaction-mass spectrometry analyses were conducted to determine the custards' volatile flavor properties. Perceived odor, flavor, and textural properties were assessed in sensory analysis experiments using magnitude estimation against a fixed modulus. Both type and concentration of CMC altered the viscosity of the custards. Softer custards had higher static headspace flavor concentrations. On the contrary, firmer custards demonstrated higher in-nose flavor concentrations. In sensory analysis, firmer custards showed higher thickness and lower sweetness intensities than their low-viscosity counterparts. The thickness perception corresponded to the viscosity of the custards. Removal of sucrose from the custards affected sweetness intensity only and not the intensity of other attributes. Therefore, the influence of the viscosity of the custards on the release of sweet-tasting components is held responsible for the effect on perceived sweetness intensity. Odor intensities were generally higher for the low-viscosity custard, whereas fruity flavor intensities were higher for the firmer custards. Odor intensities correlated with static headspace concentrations and flavor intensities related reasonably well with in-nose concentrations. Opening and closing of the nasal cavity is regarded as an important factor determining the discrepancy between static and in-nose measurements.     

Use of fiber interface direct mass spectrometry for the determination of volatile flavor release from model food systems.

Described in this paper is a fiber interface direct headspace mass spectrometric system for the real-time measurement of flavor release. The system was optimized for the detection of the garlic aroma volatile, diallyl disulfide, from water. Parameters investigated included interface temperature, flow rate through the fiber, flow rate through the sample vessel, and sample stir rate. The delay time for detection of sample after introduction into the sample vessel was determined as 43 s. The system proved to be reliable and robust with no loss in sensitivity or contamination of the mass spectrometer over a 6 month period. The technique was applied to a homologous series of aliphatic alcohols from C(2) to C(7). Results showed that as polarity decreased with increasing chain length the release of volatile into the headspace was faster and gave a higher maximum intensity. Release of the garlic aroma volatile from different commercial mayonnaise products clearly showed a decrease in the release of diallyl disulfide as fat content increased. These results demonstrate the potential of using this technique as a tool for understanding the complex interactions that occur between flavor compounds and the bulk food matrix.     

Flavor release and perception in hard candy: influence of flavor compound-compound interactions

The influence of flavor compound-compound interactions on flavor release properties and flavor perception in hard candy was investigated. Hard candies made with two different modes of binary flavor delivery, (1) L-menthol and 1,8-cineole added as a mixture and (2) L-menthol and 1,8-cineole added separate from one another, were analyzed via breath analysis and sensory time-intensity testing. Single-flavor candy containing only L-menthol or 1,8-cineole was also investigated via breath analysis for comparison. The release rates of both L-menthol and 1,8-cineole in the breath were more rapid and at a higher concentration when the compounds were added to hard candy separate from one another in comparison to their addition as a mixture (conventional protocol). Additionally, the time-intensity study indicated a significantly increased flavor intensity (measured as overall cooling) for hard candy made with separate addition of these flavor compounds. In conclusion, the flavor properties of hard candy can be controlled, at least in part, by flavor compound-compound interactions and may be altered by the method of flavor delivery.     

Effect of beta-cyclodextrin on aroma release and flavor perception

Binding and release of volatile compounds to and from beta-cyclodextrin were measured in model aqueous systems using static equilibrium headspace and dynamic headspace dilution. Beta-cyclodextrin decreased the static equilibrium headspace for some volatiles (e.g., ethyl octanoate and decanone) due to binding, but dilution studies demonstrated that binding was readily reversible. Dynamic release of hydrophobic volatile compounds was similar to that observed from emulsions. When beta-cyclodextrin was added to fat free yogurt, the release of a commercial lemon flavoring was modified and was similar to release from a regular fat yogurt. Sensory difference testing confirmed the release results. The data demonstrate that beta-cyclodextrin can be used to modify flavor delivery in both model and real systems; the effects in the latter are sensorially significant.     

Flavor release and perception in hard candy: influence of flavor compound-flavor solvent interactions

The release kinetics of l-menthol dissolved in propylene glycol (PG), Miglyol, or 1,8-cineole (two common odorless flavor solvents differing in polarity and a hydrophobic flavor compound) were monitored from a model aqueous system via atmospheric pressure chemical ionization mass spectrometry (APCI-MS). Breath analysis was also conducted via APCI-MS to monitor release of l-menthol from hard candy that used PG and Miglyol for l-menthol incorporation. The quantities of l-menthol released when dissolved in PG or Miglyol from the model aqueous system were found to be similar and overall significantly greater in comparison to when dissolved in 1,8-cineole. Analogous results were reported by the breath analysis of hard candy. The release kinetics of l-menthol from PG or Miglyol versus from 1,8-cineole were notably more rapid and higher in quantity. Results from the sensory time-intensity study also indicated that there was no perceived difference in the overall cooling intensity between the two flavor solvent delivery systems (PG and Miglyol).     

Proton transfer reaction mass spectrometry and time intensity perceptual measurement of flavor release from lipid emulsions using trained human subjects

The effect of the fat component of liquid emulsions on dynamic "in-nose" flavor release was examined using a panel of trained human subjects (n = 6), proton transfer reaction mass spectrometry (PTR-MS), and time intensity (TI) sensory evaluation. A rigorous breathing and consumption protocol was developed, which synchronized subjects' breathing cycles and also the timing of sample introduction. Temporal changes in volatile release were measured in exhaled nostril breath by real-time PTR-MS. Corresponding changes in the perceived odor intensity could also be simultaneously measured using a push button TI device. The method facilitated accurate examination of both "preswallow" and "postswallow" phases of volatile release and perception. Volatile flavor compounds spanning a range of octanol/water partition coefficient (K(o/w)) values (1-1380) were spiked into water (0% fat) or lipid emulsions with various fat contents (2, 5, 10, and 20% fat). Replicate samples for each fat level were consumed according to the consumption protocol by six subjects. Statistical comparisons were made at the individual level and across the group for the effects of changes in the food matrix, such as fat content, on both pre- and postswallow volatile release. Significant group differences in volatile release parameters including area under the concentration curve (AUC) and maximum concentration (I(max)) were measured according to the lipid content of emulsions and volatile K(o/w). In a second experiment, using single compounds (2-heptanone, ethyl butanoate, and ethyl hexanoate), significant decreases in both in-nose volatile release and corresponding perceived odor intensities were measured with increasing fat addition. Overall, the effect of fat on in vivo release conformed to theory; fat had little effect on compounds with low K(o/w) values, but increased for volatiles with higher lipophilicity. In addition, significant pre- and postswallow differences were observed in AUC and I(max), as a result of changing fat levels. In the absence of fat, more than half of the total amount of volatile was released in the preswallow phase. As the content of fat was increased in the emulsion systems, the ratio of volatile released postswallow increased compared to preswallow. These data may provide new insights into why low-fat and high-fat foods are perceived differently.     

Altering the fat content affects flavor release in a model yogurt system.

Flavored yogurts differing in fat content were eaten, and the release of flavor volatiles was measured by monitoring the volatile composition of air from the nose in real time by atmospheric pressure ionization mass spectrometry. Low-fat yogurts (0.2%) were found to release volatiles more quickly and at higher intensity but with less persistence than yogurts containing fat at 3.5 and 10% fat. Yogurts with increasing fat content had higher viscosity and lower relative particle size. Lipophilic compounds were more affected by fat for maximum volatile intensity, but not time-to-maximum intensity or persistence. Sensory assessment of the yogurts found significant differences in intensity and speed of onset of flavor, but not overall length of perception.