The influence of volatile organic compound release, texture and microstructure on the perception of apple flavour

Aroma and texture are known to influence the overall perception of apple flavour and are subconsciously evaluated by consumers during consumption. Aroma is derived from the volatile organic compounds (VOCs) released from the apple both before (orthonasal) and during consumption (retronasal). The degree to which the VOC impacts upon the flavour profile is affected by the mechanical breakdown of the apple flesh during mastication. The characteristics of apple texture, microstructure and breakdown are cultivar specific. This thesis aimed to understand the cultivar specific differences of apple texture, microstructure and VOC release in relation to their role in overall flavour perception and to explore the correlations between sensory derived attributes and instrumental derived parameters. In vitro (orthonasal) and in vivo (retronasal) measurements of VOC release were carried out using either a proton transfer reaction quadrupole mass spectrometer (PTR-QUAD-MS) or a proton transfer reaction time of flight mass spectrometer (PTR-ToF-MS). Both instrumental techniques separated cultivars based on their in vitro headspace volatile organic compound (VOC) profiles of either intact or cut apple portions. However, PTR-ToF-MS was found to be superior to PTR-QUAD-MS due to its higher sensitivity. The in vivo nosespace method involving the measurements of dynamic VOC release from the nasal cavity, during swallowing as panellists consumed apple portions, resulted in improved data and temporal resolution, as well as an increase in the number of detectable nosespace VOCs when the PTR-ToF-MS was used. Ester related VOCs were predominantly detected in the in vivo nosespace measurements followed by carbonyl and alcohol related compounds. Measured nosespace parameters showed potential for use in understanding differences in the mechanical breakdown of apple flesh during mastication; with for example juicy cultivars inducing a faster time to first swallow (Tswal). Panellists also displayed longer consumption times (Tcon) when consuming firm samples, as opposed to softer cultivars. In a preliminary study using the PTR-QUAD-MS, the time to maximum VOC intensity (Tmax) was found to be shorter for soft apple flesh owing to their faster breakdown during mastication and their resulting increase in surface area. However, the PTR-ToF-MS study which used more cultivars suggested that Tmax could also be shorter for firm juicy cultivars, owing to the occurrence of more swallowing events during mastication as juice was released. The feasibility of being able to differentiate between cultivars using solely instrumental techniques was investigated by comparing instrumentally derived results to those gained from descriptive sensory analysis. Sensory odour/ flavour data was compared to VOC profiles obtained from the PTR-ToF-MS, while sensory texture was compared to mechanical and acoustic textural parameters. Both analytical and sensory techniques characterised the cultivars similarly, indicating the potential of solely instrumental techniques to separate apple cultivars based on their VOC profiles and textural properties. Different VOC groups (terpenes; esters and their associated fragments; butanoate esters) were associated to specific sensory attributes. For example, butanoate esters (m/z 103, 117, 131) were positively correlated to lemon flavour and odour whereas hexyl acetate (m/z 145) was positively correlated to overall apple odour signifying its importance towards base apple flavour. Physico-chemical properties such as titratable acidity (TA) were positively correlated to sour taste which could be considered in prediction models. Sweet taste was negatively correlated to TA but was not correlated to soluble solids content (SSC). However, a positive correlation between sweet taste and fruity ester related compounds (m/z 43, 61, 145) was observed indicating a possible aroma-taste interaction. Instrumentally derived texture parameters were well correlated to sensory descriptions. However, the former also provided additional information that could differentiate between cultivars that were firm but not crunchy. As the behaviour of apple flesh breakdown during mastication was shown to affect perceived texture, X-ray micro-computer tomographic (μ-CT) scanning was used to investigate microstructural differences that could affect apple texture. Morphological properties of apple parenchyma such as connectivity and porosity were associated with VOC release, whereas anisotropy was associated with firm texture. Porosity was negatively correlated to firm texture however, this was cultivar specific owing to the fact that porosity is a summation of all intercellular spaces within a volume; it does not take into account the variation in size and distribution which has been shown to affect texture. The use of an ethylene inhibitor, 1-methylcyclopropene (1-MCP) retained the apple texture at a cost of lower VOC emission. However, its efficacy favoured cultivars of high ethylene concentration (Golden Delicious) as opposed to low ethylene emitting cultivars (Fuji). Therefore, it is important to understand the effects of 1-MCP on specific cultivars. This study demonstrated the viable use of PTR-MS and texture measurements to understand the interrelationships between texture and aroma release during consumption. These techniques were shown to characterise apple cultivars similarly to sensory measurements. Moreover, μ-CT scanning proved to be a powerful tool to visualise the role of microstructure in mechanical texture. Overall, these analyses showed that apple cultivars are distinctly different and suggest that such information could be used to compile a database of phenotypes, such as the cultivar specific VOC and texture characteristics. This database could be used as a reference to create high quality apples through careful selection of specific traits in order to create customised apples targeted towards consumer specific needs