In the food industry, high sensory quality and stability of products are crucial factors for consumer satisfaction and
market shares. Sensory quality is normally being evaluated by two major approaches: instrumental (volatile and nonvolatile
compounds) approach and sensory approach by trained or consumer panels. Sensory evaluation is a primary
measurement for providing immediate information of human perception on the products. Instrumental methods give
objective analysis of compounds that potentially contribute to food flavour. These two kinds of analysis, basically, give
different types of information about food flavour but correlate to each other. The exploration of relationships between
sensory and instrumental data is one important aspect for fundamental understanding of the mechanisms of sensory
perception. This thesis has investigated the importance and limitation of aroma analysis and multivariate data analysis
for linking volatile compounds to sensory analysis. Based on the method applied in each study, different outcomes and
correlations can be extracted.
The first part of the project included two studies focusing mainly on aroma analysis, where analysis of aroma and
sensory data was done separately. Study 1 has investigated the application of volatile analysis and difference from
control test in winemaking practices as an example of a situation where the relationship cannot be estimated by
multivariate techniques. Volatile compound analysis revealed that some oxidative compounds increased with the
absence of SO2 in wine. Different from control test supported that lack of SO2 in wines during storage significantly
affected flavour attributes related to oxidation. Study 2 revealed the different temperature applied during cheese
ripening significantly affected the level of some compounds, especially acids, aldehydes, esters and ketones. The higher
flavour intensity of high temperature maturated cheeses, compared with the less mature cheeses was explained by the
high levels of many of the volatile compounds.
The second part of the project has investigated the gross or calculated relationships (Studies 3, 4) by using multivariate
statistical techniques to explore the associations between the two datasets. Study 3 and 4 addressed these approaches
with descriptive sensory analysis on specific product examples. Descriptive sensory analysis has demonstrated its
efficiency in relation to aroma and sensory methods in understanding the pre-fermentation treatment on sensory quality
of wine (Study 3). In Study 4, the RATA method was used to provide the intensity of significant sensory descriptors
that discriminate the significant differences between chocolate samples.
Part three step by step moves closer towards the direct link between volatiles and sensory sensation. In Study 5, GC-O
results showed that esters were the main odour active compounds which contribute to fruity notes of Solaris wine. An
optimized APCI-MS/MS method in monitoring in nose (in vivo) aroma release for select odour active compounds, and
the impact of the prior consumption of two types of cheeses (low fat, high fat) was evaluated. The compound with the
least hydrophobicity (3-methyl-1-butanol) showed the greatest in vivo persistence and the more hydrophobic
compounds (3-methylbutyl acetate, ethyl hexanoate, and ethyl octanoate) were shown to interact with the lipid fraction
of the cheese as measured by retronasal aroma release, resulting in suppression of their release in nose. The
consumption of cheese with wine impacts not only the total delivery of aroma to the nose but also changed the relative
abundance of each compounds resulting in an altered aroma profile.
In summary, this PhD study provides insights in how aroma compounds are linked to sensory quality of a product,
which serves as a starting point to practical applications and further studies. For instance, it is demonstrated how SO2
influence the aroma compounds and consequently change the sensory quality in wine which provide the useful
information of wine quality management to winemakers to as well as knowledge on the behaviour of wine oxidation.
Additional, studies focused on understanding the development of volatiles during accelerated cheese ripening; detecting
changes of volatiles as a result of production errors in chocolate production, and monitoring aroma release provide
useful information to food sensory quality control.
For each application, both techniques were faced with challenges that need to be handled in different ways. Due to the
complex nature of flavour, some sensory-aroma relationships are still weak or difficult to interpret. The developments
of instrumental and sensory analysis, improved statistical approaches are expected to give an enhanced and valid
measurement of food flavour.