Hydrocolloid clustering based on their rheological properties

AbstractIn this study, we proposed an objective classification of seven commercial hydrocolloids and four novel hydrocolloids. Total of 74 rheological parameters was generated by steady (flow behavior, hysteresis loop, single shear decay, in‐shear structural recovery experiments), dynamic (strain sweep and frequency sweep tests), and transient (creep/recovery and stress relaxation) shear measurements. Subsequently, the parameters were classified into seven categories with more than 60% similarity indexes in each group using agglomerative hierarchical clustering based on those properties related to the number of linkage, strength of linkage, distance of linkage, rupture and flow, rate of destruction, the extent of destruction, and the state of destructured samples in the absence of flow field. Fuzzy c‐means classifier used to extract patterns for each class. Our results correspond to four different classes; κ‐carrageenan and agar gum were categorized in the first class, high methoxyl pectin, xanthan, sage seed gum and basil seed gum in the second class, alginate gum and Balangu‐Shirazi seed gum in the third class, and guar gum, cress seed gum and carboxymethyl cellulose in the fourth class. Using this classification technique, complete rheological patterns can be extracted for each class. This classification provides a map for other researchers to rationally design the best test type which could describe adequately different properties of materials and avoid experiments with a similar type of parameters. The main reason for the frequent use of hydrocolloids in various industries is their ability to modify the rheology. A lot of works have been done to study the rheological behavior of many hydrocolloids in model and food systems. As there is still demand for new sources of hydrocolloids with more specific functionality in foods, probing the similarities among commercial and emerging hydrocolloids could help us to rationally design structural features in different formulations, besides gives insight into the structure–function relationship between them. This object could be attained by clustering, a part of the pattern recognition theory. Contrary to the traditional clustering methods, in which the membership of a product is exclusive for only a class, in constraint clustering by fuzzy logic methods, a partial membership can be shared by two or more classes. In this way, using the fuzzy logic clustering method, we clustered a number of commercial and novel hydrocolloids based on the steady, transient, and dynamic shear rheological properties and found a specific pattern among them.Practical applicationsThe main reason for the frequent use of hydrocolloids in various industries is their ability to modify the rheology. A lot of works have been done to study the rheological behavior of many hydrocolloids in model and food systems. As there is still demand for new sources of hydrocolloids with more specific functionality in foods, probing the similarities among commercial and emerging hydrocolloids could help us to rationally design structural features in different formulations, besides gives insight into the structure–function relationship between them. This object could be attained by clustering, a part of the pattern recognition theory. Contrary to the traditional clustering methods, in which the membership of a product is exclusive for only a class, in constraint clustering by fuzzy logic methods, a partial membership can be shared by two or more classes. In this way, using the fuzzy logic clustering method, we clustered a number of commercial and novel hydrocolloids based on the steady, transient, and dynamic shear rheological properties and found a specific pattern among them.