Modeling the Time to Fail of Peach Nectars Formulated by Hurdle Technology

Abstract

The use of regression with life-data is helpful to observe whether one or more factors affect the failure time (spoilage) of a product, obtaining a model that predicts the time to fail (TTF). TTF models link kinetic (lag time) and probabilistic (growth /no-growth prediction) models for selected formulation/storage conditions. Our objective was to assess the individual and combined effects of pH, a_w, and the incorporation of potassium sorbate (KS) or sodium benzoate (BNa) at selected concentrations on the microbial stability of peach nectar during storage at 25°C, in order to model and predict TTF. Peach nectars were formulated with 40% fruit pulp and the necessary sucrose syrup and citric acid to attain a_w 0.96, 0.97, or 0.98 and pH 3.0, 3.5, or 4.0; while 0, 500, or 1000ppm of KS or BNa were added. Nectars were stored for 180 days in glass jars at 25°C, and periodically analyzed (standard plate as well as yeast and mould counts). The experimental design and analyses were replicated three times. Storage times that revealed microbial populations higher than 10⁴ CFU/mL and signs of spoilage were registered to model TTF by survival analysis. From the 54 combinations tested, 9 formulations (without antimicrobials) exhibited early spoilage (<5 days). For the combinations formulated with 500ppm of BNa, spoilage was detected after 30 days; much longer spoilage times were observed for nectars with 1000ppm of KS or BNa. In general, KS was more effective than BNa in delaying spoilage when 1000ppm were added. TTF models included individual and interaction effects of the evaluated factors and revealed good agreement among experimental and predicted data (R²>0.90). Survival analysis through TTF models can be used to predict spoilage time under specific factor combinations or to select factor levels for a specific shelf-life of peach nectars.