A predictive model for flavor partitioning and protein-flavor interactions in fat-free dairy protein solutions

Ombeline Viry, Remko Boom, Shane Avison, Mirela Pascu, Igor Bodnár*

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

8 Citations (Scopus)

Abstract

Flavor perception is directly related to the concentration of aroma compounds in the headspace above a food matrix before and during consumption. With the knowledge of flavor partition coefficients, the distribution of aroma compounds within the food matrix and towards the headspace can be calculated. In this study static headspace measurements and modelling are combined to predict flavor partitioning of a wide range of flavor compounds above fat-free dairy protein mixture solutions. AFFIRM® (based on Atmospheric Pressure Chemical Ionization-Mass Spectrometry) was used to measure the static headspace concentrations of 9 flavor compounds (3 esters, 3 aldehydes and 3 alcohols) above protein solutions with different concentrations and ratios of sodium caseinate and whey protein isolate. Proteins had a small pushing out effect, leading to increased release of hydrophilic flavor compounds. This effect was negligible for more hydrophobic compounds, where clear retention was observed. An increased total protein concentration and higher whey to casein ratio increased the retention for all flavor compounds. Within the same chemical class, the retention increased with chain length. The experimental data was interpreted with a model describing flavor partitioning in protein solutions (Harrison & Hills, 1997), thereby enabling to extract protein-flavor binding constants. A clear power law was found between the protein-flavor binding constant and log P (octanol-water partition coefficient). Assuming solely non-specific hydrophobic interactions gave satisfying partitioning predictions for the esters and alcohols. For aldehydes specific chemical interactions with proteins turned out to be significant. This rendered a binding constant for whey protein that is 5 times higher than for caseinate in case of esters and alcohols, and 3 times higher in case of aldehydes. The model can accurately predict equilibrium flavor partitioning in dairy protein mixtures with only the knowledge of the octanol-water partition coefficients of the flavor compounds, and the composition of the protein mixture.
Original languageEnglish
Pages (from-to)52-58
JournalFood Research International
Volume109
DOIs
Publication statusPublished - 1 Jul 2018

Keywords

  • APCI-MS
  • Flavor
  • Hydrophobicity
  • Partitioning
  • Sodium caseinate
  • Whey protein isolate

Fingerprint Dive into the research topics of 'A predictive model for flavor partitioning and protein-flavor interactions in fat-free dairy protein solutions'. Together they form a unique fingerprint.

  • Cite this