Mouth model

This unit discusses the use and design of the two mouth simulators, the retronasal aroma simulator (RAS) and the model mouth, that have successfully been verified to produce an effluent with volatile ratios similar to that found in human exhaled breath during eating. Though at a glance the apparatuses seem very different, they produce relatively similar effluents. Of obvious notability is the difference in the size of the reservoir the RAS reservoir is 1 liter and the model mouth reservoir is 70 ml. When determining which apparatus to use, carefully consider concentration needs, absorption characteristics of compounds, and shear resistance of the food. 

The release of aroma compounds in the mouth during eating is primarily determined kinetically, rather than thermodynamically, because of the processes occurring when food is consumed. The model-mouth system was developed to study in vitro-like aroma release and considers the bolus volume, volume of the mouth, temperature, salivation, and mastication (van Ruth et al., 1994). Volatile compounds in the effluent of the model mouth are collected on porous polymers, such as Tenax TA. Alternatively, the effluent can be measured on-line by direct mass spectrometry techniques. The model mouth can be used to study the effects of food composition and structure on aroma release, as well as the influence of oral parameters related to eating behavior. 

Figure G1.7.3 A diagram of the model mouth. A, trap B, ethanol bath (-10°C) C, nitrogen gas source D, water bath (37°C) E, sampling material (in sample flask) F, plunger G, voltage controllers H, motors.

The model mouth was developed as part of the doctoral program of Saskia van Ruth at the Wageningen University in the Netherlands between 1992 and 1995. The hypothesis was that only those volatile compounds released under mouth conditions are relevant for aroma analysis. The volume of the mouth, temperature of the mouth, mastication, and salivation were thought to be critical parameters. Those pa-... 

Figure G1.7.4 A model mouth and human mouth comparison shows a gas chromatogram of volatile compounds released from rehydrated French beans in the model mouth (n = 6 upper chromatogram) and in the mouth of assessors (n = 12, lower chromatogram).

The model mouth and RAS are two examples of mouth simulators. A representative set of mouth simulators are compared in Table Gl.7.1. All account for temperature, breath flow, and mastication. Only the model mouth and the RAS allow for the evaluation of solid foods and have been compared directly to human breath. 

Sample/saliva ratio. The sample/saliva ratio is a critical parameter for volatile release in the model mouth for liquid, semi-solid, and solid foods. With liquid foods, the dilution and the change in lipid and protein concentration have an effect. With more solid types of foods, besides the effects above, salivahas an effect on the dynamics of the release. Saliva, in combination with mastication, affects the rates of mass transfer. Again, a realistic sample/saliva ratio should be chosen (e.g., in the 80 20 to 40 60 range). 

Compounds not detected or detected in lower-than-expected concentrations. First, make sure that the problem is definitely due to a problem with the model mouth. For example, the cause of the problem may be due to the analytical equipment (e.g., gas chromatograph or mass spectrometer), inconsistencies in the food sample, and/or extraction errors. If volatile compounds are not detected or are detected in far lower-than-expected concentrations, there may be a gas leak somewhere in the system. All connections should be checked with a leak detector as described for the RAS. 

The model mouth and the RAS will produce effluents carrying a ratio of volatile compounds that is similar to the ratio of volatiles leaving the human nose when the same food is consumed. The RAS effluent will be -200 times more concentrated than the human breath. The RAS produces a time average representation of the retronasal breath composition. 

Bringing the model mouth to temperature requires -20 min. Food preparation, i.e., cutting and measuring, takes -5 min. Initiating the model mouth takes -1 min, running the model mouth and sampling the effluent usually takes 1 min. Cleaning the model mouth requires 10 min, and may be done concurrently with GC analysis. The time needed for GC and MS analysis is as described for the RAS. 

Provides a description of model mouth and some effects of oral physiological parameters. 

Hansson, A., Giannouli, P, Van Ruth, S. (2003) 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. Journal of Agricultural and Eood Chemistry, 51,4732—4740. 

MS Brauss, B Balders, RST Linforth, S Avison, AJ Taylor. Fat content, baking time, hydration and temperature affect flavour release from biscuits in model-mouth and real systems. Flavour Fragrance J 14(6) 351-357, 1999. 

Representative Sampling of Volatile Flavor Compounds The Model Mouth Combined with Gas Chromatography and Direct Mass Spectrometry... 

SH, static headspace sampling DH, dynamic headspace sampling MM, model mouth sampling GC, gas chromatography APcI-TOFMS, atmospheric pressure chemical ioni2ation-time of flight mass spectrometry PTRMS, proton-transfer-reaction mass spectrometry. 

The model food (10 mL sunflower oil solution, 0.001% v/v for each individual compound) was placed in the sample flask of the latest version of the model mouth [7], The headspace was flushed and the effluent analyzed by GCMS and PTRMS, as described for the dynamic headspace analysis. Headspace concentrations were calculated according to Ref. 6. For real-time PTRMS analysis one specific ion (2-butanone 73 ethyl acetate 89 hexanal 83 2-heptanone 115 ethyl butyrate 117) was monitored for the individual compounds in order to obtain a temporal release profile. Data were corrected for fragmentation. 

Table 3 Volatile Flavor Compounds Sampled Under Static Headspace, Dynamic Headspace, and Model Mouth Conditions Determined by Gas Chromatography and Their Octanol/Water Partition Coefficients (Log P)...

Cumulative release over 1 min. DH, dynamic headspace. Proportions [%] of volatiles sampled. MM, model mouth. 

Sampled under static headspace, dynamic headspace, and model mouth conditions and their octanol/water partition coefficients (log P). Raw data in Fig. 1. 

SH, static headspace DH, dynamic headspace MM, model mouth. 

C [10]), which explains its relatively large proportion under static conditions. According to Voilley and colleagues [10], the compound has a relatively high resistance to mass transfer in oil (27.4) compared to ethyl butyrate (6.7). This may explain the lower proportion under mild dynamic conditions, as well as the beneficial effect of rigorous movements under model mouth conditions. 

The changes in flavor release with time (temporal release) under model mouth conditions were examined by PTRMS (Fig. 1). The maximal intensities of the compounds varied, but the time to reach this intensity did not differ significantly among the compounds. Ethyl acetate showed largest... 

Figure 1 Temporal release profile of five volatfie flavor compounds released from sunflower oil under model mouth conditions determined by proton-transfer-reaction mass spectrometry. (PTRMS). Sample was introduced after 0.5 min.

SM van Ruth, E Boscaini, D Mayr, J Pugh, M Posthumus. Evaluation of three gas chromatography and two direct mass spectrometry techniques for aroma analysis of dried red bell peppers. Int J Mass Spectrom 223-224 55-65, 2003. SM van Ruth, CH O Connor, CM Delahunty. Relationships between temporal release of aroma compounds in the model mouth system and their physicochemical characteristics. Food Chem 71 393-399, 2000. 

SM van Ruth, JP Roozen. Influence of mastication and saliva on aroma release in a model mouth system. Food Chem 71 339-445, 2000. 

---