Design and optimisation of pasteurisation conditions for cupuaçu (Theobroma grandiflorum) fruit pulp

Abstract

Cupuaçu (Theobroma grandiflorum) is an Amazonian tropical fruit with a great economic potential. A pasteurisation, by a hot-filling technique, was suggested for the preservation of this fruit pulp at room temperature, and implemented with local communities in Brazil. On account of low activities of enzymes (pectinesterase, peroxidase and polyphenoloxidase) in cupuaçu and/or low temperatures and times required for its complete thermal inactivation, the idea of using enzymes as pasteurisation target for cupuaçu pulp was dropped. A new methodology to establish the pasteurisation criterion, based on Alicyclobacillus acidoterrestris (AAT) spore reduction, was suggested. This bacterium has been detected in several spoiled commercial pasteurised acid fruit juices. The D-value (decimal reduction time) at 91°C and z-value for cupuaçu were, respectively, 4.6 min and 8.9°C (AAT type strain, NCIMB 13137). Based on knowledge from fruit industry in general, literature and storage studies with cupuaçu pulp, one decimal reduction (N/No=0.10) on AAT spores was suggested for this product. The investigation of temperature (T: 85-97ºC), total soluble solids (SS: 5- 60ºBrix or % by weight) and pH (2.5-6.0) influence on D-values, for AAT spore inactivation, was carried out with a response surface methodology and using malt extract broth adjusted to specific T-SS-pH conditions. D-value (between 0.498±0.045 and 94.9±6.7 min) was affected mainly by the temperature, then SS, and, lastly, the pH. A linear decrease in D-value was observed with decreasing SS and pH, whereas a non-linear decrease was noticed with increasing temperature. A second order polynomial was successfully fitted to the data, however, D-values measured in real fruit systems (such as orange, apple and grape juices, blackcurrant concentrates, cupuaçu extract and orange juice drink) were in general higher than those predicted by the malt extract broth model. Care must be taken with model predictions. The kinetics of colour (measured with a colorimeter: total colour difference, TCD*, and lightness retention, L*/Lo*) and sensory changes (‘fresh-notes’ of aroma, FA, and flavour, FF; ‘cooked-notes’ of aroma, CA, and flavour, CF) in cupuaçu pulp were determined from isothermal experiments. At each isothermal experiment TCD*, CA and CF increased and normalised L* , FA and FF decreased with processing time. Colour parameters were described by a power law model, FA and FF follow a first-order model (EaFA=78.24 kJ/mol, EaFF=82.11 kJ/mol), and CA and CF a fractional model (also called first-order reversible) (EaCA=79.67 kJ/mol, EaCF=85.19 kJ/mol). Although ‘cooked-notes’ were linearly correlated with ‘fresh-notes’, the former was a better indicator for quality degradation. A mathematical model was developed to describe the pasteurisation, by hot filling, of fruit purées. The model assumed uniform heating and holding (at the hot-fill temperature), and, after introduced in the container, cooling by conduction. The heat transfer was modelled by a finite difference method. A computer program integrating heat transfer properties, container characteristics and kinetics of AAT spores thermal inactivation, and colour and sensory changes, previously determined, was written. This program is able to make predictions on quality and microbial (safety) parameters for different processing conditions at any container point or in terms of volume average. The relative importance among the process variables (initial product temperature, heating rate, holding temperature and time, container volume and shape, cooling medium type and temperature) on the P (pasteurization value) value and quality was investigated, by performing simulations according to a screening factorial design. The holding temperature (TF) and time affected the P, and the container volume influenced largely the quality. Optimisation was performed for 1 and 100 L container sizes, by generating/simulating equivalent processes (a minimum P91°C=4.6 min was required) and evaluating the final quality. TF≥91°C minimised CA-CF and maximised FA-FF attributes for 1 L container. Concerning the 100 L, the CF and CA development can be minimised with TF≥91°C, but overall, the quality was greatly affected due to long cooling times. A more efficient method to speed up the cooling phase is recommended for industrial size containers. Finally, a storage study for half a year, at two different temperatures (18°C, 38°C), of cupuaçu purée previously pasteurised at 70°C (N/No=0.996) and 90°C (N/No=0.550), was performed. The quality was investigated before and after pasteurisation, and along storage. The pasteurisation was effective for the microbial/enzyme inactivation and stabilisation of the purée. The parameters that most changed during storage were the colour (TCD* increased, and purée got darker), flavour and aroma (some ‘fresh-notes’ were lost) and sugars (conversion of non-reducing into reducing sugars). Within the temperatures studied, the final quality was determined by the storage temperature and not by the pasteurisation temperature (and corresponding time). The total colour difference, TCD*, change during storage was well modelled by first-order reversible kinetics.