Skip to main content
ARS Home » Research » Publications at this Location » Publication #149897
Title: ANALYSIS OF FLAVORS USING A MASS SPECTRAL BASED CHEMICAL SENSOR

Author
item KINTON, VANESSA - GERSTEL, INC.
item Goodner, Kevin

Submitted to: Eastern Analytical Symposium
Publication Type: Abstract Only
Publication Acceptance Date: 5/30/2003
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Development of flavor formulations can be a challenging task. Individual components of different flavor formulations can be varied in composition and quantity depending on the final product desired. For example, if a new formula needs to be developed that matches a certain target, a series of trial formulations can be developed and a method that compares samples is needed. A technique widely used in flavor analysis is GC/MS. The output, a TIC, can be challenging to compare between many samples. This technique can also be time consuming with sample analysis between 30-60min/sample. In order to simplify the comparison and decrease analysis times, a mass-spectral-based-chemical-sensor was used in this study. The chemical sensor consists of an x, y, z robot autosampler that is capable of sampling headspace, SPME or liquids. The autosampler is directly coupled to a mass spectrometer without a GC which results in MS analysis times of 1-3min/sample. Multivariate data analysis is then applied to the samples mass spectral fingerprint and comparison of sample profiles can be easily done with exploratory data analysis (PCA or HCA). Different flavor formulations were easily compared against a target sample by visual inspection of the fingerprint projections into a 3D-PC space. These results were validated by running the same samples with a GC/MS instrument. The GC/MS data was also analyzed in the same manner as the MS-only data and compared. Overall, the fast and easy interpretation of output from the chemical sensor was found to be an advantage over the traditional GC/MS analysis.