Model system flavor studies

The effects of seven processing variables which, on the basis of previous literature reports (7, 27) and our chemical studies on model systems, might affect the irradiation flavor intensity or consumer acceptance of irradiated steaks were evaluated in statistically designed and analyzed experiments. Steaks were then prepared using the processing conditions shown to be optimum by these experiments and evaluated for consumer acceptance and storage stability. 

Use of Model Systems. From time to time it may be desirable to utilize a model or artificial citrus juice system as an aid in studying certain flavor attributes. A big advantage, of course, is the resultant standard "juice" being completely reproducible at any time. However, the big disadvantage is that no matter how well a citrus product is simulated in a model juice system, many taste panelists apparently cannot feel really comfortable when evaluating a purely artificial product. 

Heterocyclic compounds are primarily formed through non-enzymatic browning reactions. Recent studies of deep-fat fried food flavors led to the identification of pyrazines, pyridines, thiazole, oxazoles and cyclic polysulfides which had long-chain alkyl substitutions on the heterocyclic ring. The involvement of lipid or lipid decomposition products in the formation of these compounds could account for the long-chain alkyl substitutions. Model systems were used to study the participation of lipids in the formation of pyrazines, pyridines, thiophenes and cyclic polysulfides. 

In addition to simple model systems, more complex systems which are closer to actual foodstuffs have been used to investigate the formation of flavor chemicals in the Maillard reaction. Sixty-three volatile chemicals were isolated and identified from starch heated with glycine (4). When beef fat was used as a carbonyl compound precursor in a Maillard model system with glycine, 143 volatile chemicals were identified (6). These included fifteen n-alkanes, twelve n-alkenes, thirteen n-aldehydes, thirteen 2-ketones, twelve n-alcohols, and eleven n-alkylcyclohexanes. Recently, the effect of lipids and carbohydrates on the thermal generation of volatiles from commercial zein was studied (7). 

The composition of the volatile fraction of bread depends on the bread ingredients, the conditions of dough fermentation and the baking process. This fraction contributes significantly to the desirable flavors of the crust and the crumb. For this reason, the volatile fraction of different bread types has been studied by several authors. Within the more than 280 compounds that have been identified in the volatile fraction of wheat bread, only a relative small number are responsible for the different notes in the aroma profiles of the crust and the crumb. These compounds can be considered as character impact compounds. Approaches to find out the relevant aroma compounds in bread flavors using model systems and the odor unit concept are emphasized in this review. A new technique denominated "aroma extract dilution analysis" was developed based on the odor unit concept and GC-effluent sniffing. It allows the assessment of the relative importance of the aroma compounds of an extract. The application of this technique to extracts of the crust of both wheat and rye breads and to the crumb of wheat bread is discussed. 

In initial work, L- a-phosphatidylcholine (lecithin) from egg yolk was selected as the phospholipid, and later studies compared other phospholipids and lipid extracts from meat. As the study originated from investigations of cooked meat flavor, the model system reactions were carried out in aqueous solution buffered with phosphate at an initial pH of 5.7 and concentrations of the reactants were selected to approximate their relative compositions in mammalian muscle. The reactions were carried out under pressure... 

In summary, model studies are very efficient for the identification and structure elucidation of important flavor components. Most of the compounds reported here have not been identified in meat and have not yet been reported as constituents of food volatiles. Nevertheless, there are good reasons to believe that minute traces of these sulfur-containing components are present in roasted and/or cooked meat volatiles because our model system was based solely on naturally occurring precursors. We believe that only minute trace amounts of these types of components need to be present in natural products to be of prime significance due to their extremely low odor threshold values. 

Concentrations of thermally generated meat flavor components are diminished by protein adsorption when soy extenders are added to fresh meat products before heating. The amounts of individual alkyl pyrazines, thermally generated by heating beef diffusate, decreased linearly as the amount of whole soy, soy 7S or soy 11S proteins were increased in a model system. Similar recoveries were obtained when pyrazines were mixed with soy either as chemical standards or from diffusate. Stoichiometry and energetics of interaction were determined for methyl pyrazine congeners with soy proteins at 120° and 145°C. Results of this study suggest that flavorants can be added in readily determined amounts to compensate for losses due to adsorption in meat-soy products. 

This paper reviews the interactions between aroma compounds and other components of a wine matrix colloids, fining agents and ethanol. Studies are carried out with model systems and instrumental methods to investigate flavor-matrix interactions. 

The knowledge of the composition of volatile compounds in food has greatly increased during the past decade. Many studies continue to report the identity and the concentration of volatile compounds in food matrices. However concentration alone appears insufScient to explain flavor properties of food. The lack of our knowledge concerning the influence of non-volatile constituents of food on the perception of aroma has to be filled by studies such as those presented in this paper. Data on interactions between aroma and matrix in wine are scarce compared with other food matrices studied. Flavor-matrix interactions in wine have generally been obtained in model systems and with instrumental experiments. However it is possible to develop some hypotheses on the possible sensory contribution of some non-volatile compounds of wine on overall aroma. 

Recent studies on salmon flavors revealed that a single compound appears to be responsible for the characterizing cooked salmon flavor (39). The cooked salmon flavor compound was found to have an extremely low threshold, and was Initially detected only by odor assessment of a fraction eluting at Ig of 9.6-9.7 on a Carbowax 20M packed column when headspace volatiles were analyzed from canned salmon meat. Accelerated oxidation of salmon oil did not yield salmon-llke aromas before the development of fishy oxidized aromas. However, when salmon oil was coated onto Cellte supports, and allowed to oxidize at room temperature, a distinct salmon-loaf-llke aroma developed within 24 h after Initiation of oxidation. A variety of supports were evaluated In model systems with salmon oil for their ability to produce the salmon aroma compound. Odor assessments of the oxidizing systems Table II Indicated that a range of odors developed from salmon-loaf-llke to oxidized fishy aromas, and only the Cellte system provided the aroma. 

Ames, J.M. and Hincelin, O., Novel sulfur compounds from heated thiamine and xylose/thi-amine model systems, in Progress in Flavor Precursor Studies, Schreier, P. and Winterhalter, P., Eds., Allured Publishing Corp., Carol Stream, IL, 1992, p. 379. 

One challenge cing the flavor industry today deals witii the stability of process/reaction flavors to heat, UV-light, oxidation, binding to food mafrices, diffusion and interactions with the enviroiunent in v ch they find application. Seeventer et al. (75) studied the stability of thiols formed from model system ribose/cystein, and reported that 2-methyl-3-furandiiol, 2-furfiirylthiol, 2-mercapto-3-butanone and furaneol decreased during storage. In brewed coffee. 

The Maillard reaction between reducing carbohydrates and amines is among the most important flavor generating reactions in thermally processed foods (5). Thus, it might be expected that in foods treated with HHP, but at low temperatures, some of the typical aroma compounds might not be formed. Only two studies about the influence of HHP on the formation of volatiles in Maillard model systems are currently available (6, 7). Bristow and Isaacs (d) reported that at 100°C, the formation of volatiles from xylose/lysine was generally suppressed when HHP was applied. Hill et al. (7) confirmed this observation for a glucose/lysine system. However, it has to be pointed out that the samples analyzed were not reacted in a buffered system and, also, the reaction time of the pressure-treated and untreated sample were not identical. 

The purpose of the present study was to identify those volatile compounds which significantly contribute to the seasoning-like note of fenugreek using the approach of sensory directed chemical analysts. Gas chromatography in combination with olfactometry and mass spectrometry have been used as key steps of this approach (18,19). The formation of flavor impact compound(s) was studied in model systems using the quantification technique Isotope Dilution Assay (20, 21) The mechanistic study was based on a hypothetical pathway proposed for the formation of sotolone via thermally induced oxidative deamination of HIL (10). 

Tertiary butylhydroquinone (TBHQ), a commonly-employed synthetic antioxidant in foods, was also included in the studies as a model phenolic compound, and it had the same effect on depletion of methanethiol as caraway seed fiavonoids (Table ni). These results indicated that TBHQ could be used also for suppressing unpleasant flavors in foods that are caused by methanethiol and its oxidative derivatives. Because the methanethiol-depleting effect of caraway seed fiavonoids and TBHQ increased with increasing oxygen tensions and pH values in the model systems... 

The influence of dimethyl disulfide on flavor quality of modified atmosphere stored broccoli florets was further demonstrated by data shown in Table VI which revealed that aroma assessments of undesirable sulfurous aromas in broccoli florets more closely corresponded with concentrations of dimethyl disulfide than with methanethiol. In this experiment broccoli florets were treated by dipping in various solutions containing either ascorbic acid (500 ppm), sodium hy oxide (0.01 M), or phosphoric acid (0.1 M). Broccoli florets dipp in distilled water were used as the control samples. Samples held in sealed Curlon 850 pouches were analyzed after storage for 4 days at 10°C. Ascorbic acid was included in this study because it acts as a reducing agent for methanethiol under anaerobic conditions in model systems (25). 

Strecker aldehydes are quantitatively the major products of the Maillard reaction. In addition to their intrinsic flavor, they are very reactive and participate in numerous reactions that make additional contributions to flavor development in foods. There is a lack of information on the reaction kinetics of these Strecker aldehydes as well as other flavor compounds. Thus a kinetic study on the formation of methional and two secondary products (dimethyl disulfide and 2-acetylthiophene) from the reaction of amino acids (0.075 mole) and glucose (0.5 mole) in aqueous model systems was conducted. Systems were heated at temperatures from 75 to 115°C at times from 5 min. to 7.5 h and pH s of 6, 7, and 8. Kinetic data are presented and discussed. 

---