Optimisation processing of green kiwifruit-blackcurrant leather using response surface methodology

Abstract

The purpose of this study was to efficiently produce a green kiwifruit-blackcurrant fruit leather with acceptable colour and taste characteristics. The effects of different levels of sugar, blackcurrant purée and green kiwi fruit purée and a small addition of pectin were analysed using response surface methodology to identify a range of optimal ingredient levels. An advanced software programme, Design-Expert which used a Box-Behnken method was used to design the experiment. Sixteen different combinations suggested by the programme were investigated, using constant drying conditions, to determine the final moisture contents, water activity, colour, texture and ascorbic acid contents of the green kiwifruit-blackcurrant fruit leather produced. The outputs of the data-derived analysis from this programme were then analysed using multiple regression and ANOVA analysis was used to identify the significance of the developed models. The data were then plotted using three dimensional surface plots so that the interaction of the different parameters could be observed. This analysis showed that the optimum combination for the manufacture of kiwifruit-blackcurrant fruit leather was 81.02% kiwifruit purée, 8.95% blackcurrant purée, 10% sugar and 0.03% pectin. Central composite analysis was then used to investigate the effect of drying conditions; temperature, time and sample thickness on this optimum ingredient combination. Eighteen different combinations of drying time, temperature and sample thickness suggested by the programme were then applied to the optimised fruit leather mixture and the outcomes analysed for final moisture contents, water activity, colour, texture and ascorbic acid contents of this green kiwifruit-blackcurrant fruit leather dried at each set of conditions. The data were then plotted using three dimensional surface plots so that the interaction of the three drying parameters could be observed. The optimum drying conditions for a product with the highest desirability as defined by the programme, were 14.73 hours of drying time at 67.32°C with a sample thickness of 8.00 mm. Under these conditions the predicted responses were moisture content of 33.72 g/100 g DM; water activity, 0.67; L*, 28.25; a*, 6.89; b*, 1.08; chroma, 6.57; puncturing force, 0.19 N/mm and ascorbic acid content 164.51 mg/100 g DM. Sensory analysis showed that the optimised recipe was appreciated by almost all panellists and this was confirmed by their overall liking score for this recipe. The flavour and colour of the fruit leather were the most important characteristics appreciated by panellists. The very dark colour of fruit leather made from blackcurrant alone was the most disliked feature. The optimised recipe derived from the data generated by response surface methodology using a mixture of kiwifruit and blackcurrant purée and processed using optimum conditions produced the most acceptable colour and texture and received the highest overall liking score. The textural property, which was critical for customer acceptance in the taste test was determined by measuring the force needed to puncture the fruit leather. The puncturing force of fruit leather increased with increasing pectin content, increasing drying temperature and decreasing sample thickness. Overall, the fruit leather made from a combination of kiwifruit and blackcurrant purées using the most efficient combination of materials and processing conditions was identified by response surface methodology methods. The optimised fruit leather was soft and tasty and had the highest overall liking of all the fruit leathers tested in a sensory evaluation experiment.