Statistical analysis of the physico-chemical characterization of dairy and non-dairy food powders - Institut Agro Rennes-Angers Accéder directement au contenu
Poster De Conférence Année : 2012

Statistical analysis of the physico-chemical characterization of dairy and non-dairy food powders

Résumé

Food powders are generally characterised by their dry matter content. However, better understanding and further use would result from taking into account biochemical properties (protein, carbohydrate, lipid and mineral composition), as well as microbiological and physical properties (e.g., density, interstitial and occluded air, particle size, solubility, dispersibility, wettability, flowability, floodability, hygroscopicity, etc.). Indeed, all of these properties, which are discussed in this paper, determine: - the preservation/storage conditions of powders, in relation to their sensitivity to different types of damage (non-enzymatic browning or Maillard reactions, oxidation, microbial growth, etc.), - the properties of powders in use (flowability, solubility, rehydration, etc.) - and, more generally, powder quality from functional, nutritional and sensory points of view. We provide a comprehensive analysis to analyse the correlations between the properties of 10 commercial dairy powders and 15 commercial non-dairy food powders obtained by spray drying. Principle component analysis (PCA) was carried out to provide an overall representation of the experimental set up. Almost 45 % of the total variability was represented by the projection of the experimental set up on the map defined by the first two principle components (PC) of the PCA. This projection was used to group the following: - 1st group (bottom right quartile): the total nitrogen content, the water content of the monolayer and the total water content. This correlation was expected because proteins are constituents that bind primarily to water at low aw. - 2nd group (top left quartile): dispersibility criteria, dry matter and total dry matter, negatively correlated with the first group; - 3rd group (bottom left quartile): density criteria (bulk, tapped, true) - 4th group (top right quartile): total and occluded air content (partial, total) negatively correlated with the third group. The experimental set was subsequently projected onto the map determined by PC1 and PC2. It is of note that the hierarchical cluster analysis performed on the PCA data grouped powders into 6 distinct classes: powders high in fat in class , simple carbohydrates in class and more complex carbohydrates in class . The analysis identified two classes for high protein powders: class contained micellar casein and gelatin, which were thus distinguished from other protein powders (class ). Finally, class contained powders of various origins and inhomogeneous powders. To conclude, it is important to note that the correlations identified in this study take into account the variability of often very divergent processing methods and storage conditions. This emphasizes the scope of correlations of variables and classes of samples identified by the PCA.
Fichier non déposé

Dates et versions

hal-01209381 , version 1 (02-10-2015)

Identifiants

  • HAL Id : hal-01209381 , version 1
  • PRODINRA : 175167

Citer

Pierre Schuck, Anne Dolivet, Valérie Lechevalier-Datin, Romain Jeantet. Statistical analysis of the physico-chemical characterization of dairy and non-dairy food powders. 5. International Symposium on Spray Dried Dairy Products, Jun 2012, Saint Malo, France. , 2012. ⟨hal-01209381⟩
37 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More