Lecture
Flavour Analysis. Past Present and Future
- at -
- ICM Saal 4b
- Type: Lecture
Lecture description
Before the advent of modern chemical analytical techniques, flavour research was limited to studying and mixing raw materials such as vanilla, mint and spices to make new flavours. Once Gas Chromatography (GC) was invented, flavour researchers were able to separate the chemical compounds in a particular flavour and identify them. In the 12 years after GC was invented, around 4000 publications involving flavour analysis by GC were published, indicating a significant step change in flavour research. To prepare samples for GC, flavour researchers developed a variety of methods to isolate flavour and also modified the GC operation to make analysis quicker.
Knowing the key components of a flavour (e.g., strawberry) helped the flavour industry to formulate flavours for different applications. Mixing flavour chemicals (esters, acids, pyrazines etc.) to mimic a natural flavour gave a standardised flavour to food products and the flavour quality did not depend on botanical issues such as maturity, breed variety and the weather.
Over the years since GC was invented, thousands of publications have recorded the volatile aroma profiles of fresh fruits and vegetables. The formation of cooked flavours through the Maillard reaction was studied using GC analysis as well as some sophisticated techniques, such as isotope labelling, to define the complex chemical pathways that occur when an amino acid reacts with a sugar molecule.
More modern techniques have also been adopted by academic and industry researchers. Direct-MS was developed in the late 1990s to measure aroma release during eating, 2D-GC and chiral chromatography are essential tools to separate coeluting compounds, which is important, because some chiral pairs have different aromas. The non targeted approach has also found uses in the analysis of complex and variable flavours, for example pet foods, where different raw materials from different origins, make the flavour profile extremely complex and challenging for the analyst.
Flavour analysis now faces new challenges. The AI tools currently available have reignited a long-standing question in flavour research, namely “Is it possible to predict the flavour of a mixture simply from its chemical composition”. To address this challenge, high quality data is necessary but, unfortunately, the data that is currently available is recorded in scattered in different places and in different formats, which means that combing individual studies into a master flavour data base is a complex task. Examples and potential ways to overcome some of these data issues will be presented.
Knowing the key components of a flavour (e.g., strawberry) helped the flavour industry to formulate flavours for different applications. Mixing flavour chemicals (esters, acids, pyrazines etc.) to mimic a natural flavour gave a standardised flavour to food products and the flavour quality did not depend on botanical issues such as maturity, breed variety and the weather.
Over the years since GC was invented, thousands of publications have recorded the volatile aroma profiles of fresh fruits and vegetables. The formation of cooked flavours through the Maillard reaction was studied using GC analysis as well as some sophisticated techniques, such as isotope labelling, to define the complex chemical pathways that occur when an amino acid reacts with a sugar molecule.
More modern techniques have also been adopted by academic and industry researchers. Direct-MS was developed in the late 1990s to measure aroma release during eating, 2D-GC and chiral chromatography are essential tools to separate coeluting compounds, which is important, because some chiral pairs have different aromas. The non targeted approach has also found uses in the analysis of complex and variable flavours, for example pet foods, where different raw materials from different origins, make the flavour profile extremely complex and challenging for the analyst.
Flavour analysis now faces new challenges. The AI tools currently available have reignited a long-standing question in flavour research, namely “Is it possible to predict the flavour of a mixture simply from its chemical composition”. To address this challenge, high quality data is necessary but, unfortunately, the data that is currently available is recorded in scattered in different places and in different formats, which means that combing individual studies into a master flavour data base is a complex task. Examples and potential ways to overcome some of these data issues will be presented.
