Moderate and high intensity pulsed electric fields: Effect on microbial inactivation, shelf life and quality of fruit juices

Pulsed Electric Field (PEF) processing has gained a lot of interest the last decades as mild processing technology as alternative to thermal pasteurisation, and is suitable for preservation of liquid food products such as fruit juices. PEF conditions typically applied at industrial scale for pasteurisation are high intensity pulsed electric fields aiming for minimal heat load, with an electric field strength (E) in the range of 15 − 20 kV/cm and pulse width (τ) between 2 − 20 μs. Alternatively, moderate intensity pulsed electric fields with an electric field strength in the order of E = 0.5 − 5.0 kV/cm are used as a pre-step to disintegrate plant cells or electroporate micro-organisms for mass transfer in food and biotechnological, but not as an alternative preservation technology. The work described in this thesis investigated how moderate and high intensity PEF processing conditions, product matrix and  characteristics of target species affect microbial inactivation, shelf life, enzyme activity and quality of fruit juices. This was performed by a systematic evaluation of the impact of most important process parameters in PEF processing, including electric field strength, pulse width and temperature, in a continuous-flow configuration. Microbial inactivation after PEF treatment was compared to that of an equivalent thermal process to distinguish the electroporation and thermal effects responsible for inactivation.

In Chapter 1, an overview of the current status in the scientific field on fruit juice processing, pulsed electric field processing and kinetic modelling is provided, knowledge gaps were identified and the aim and research questions following from that are described.

In Chapter 2, the impact of a high intensity PEF process (E = 20 kV/cm and τ = 2 μs) was studied on the inactivation of Escherichia coli, Salmonella Panama, Listeria monocytogenes and Saccharomyces cerevisiae in apple, orange and watermelon juice. Kinetic data showed that for identical process and matrix conditions, the yeast S. cerevisiae was the most sensitive micro-organism, followed by S. Panama and E. coli, which displayed comparable inactivation kinetics. L. monocytogenes was most resistant towards the treatment conditions tested. A synergistic effect between temperature and electric pulses was observed at inlet temperatures above 35 °C, hence less electrical energy for inactivation was required at higher temperatures. The different juice matrices resulted in a different degree of inactivation, predominantly determined by pH, where more acidic conditions led to more inactivation.

In Chapter 3, the effects of high intensity PEF processing conditions (E = 13.5 − 24.0 kV/cm) and storage temperature on the outgrowth of surviving yeast and mould populations naturally present in a fresh fruit smoothie were assessed over time. Results showed that untreated smoothie was predominantly spoiled by the outgrowth of yeasts, typically after 8 days (stored at 4 or 7 °C), while initial number of moulds present in the smoothie declined during storage. PEF inactivated most yeasts present in the smoothie, thereby providing outgrowth opportunities for the moulds, which were visually observed after 14 days (stored at 7°C) or 18 days (stored at 4°C). The intensity of the electric field strength both affected the log10 reduction of yeasts and the lag-time, the period the cell required to grow out. A similar effect of electric field strength on the degree of inactivation has been observed for moulds, although electric field strength did not influence the period to visual mould growth.

Chapter 4 and 5 focussed on the development of a model to be used to fit and predict non-linear log-time inactivation of a thermal process at a holding time comparable to a PEF process. In Chapter 4, a rational thermodynamic model based on Gaussian distribution and Eyrings rate constant has been developed that can accurately model non-linear log-time inactivation kinetics of enzymes and micro-organisms exposed to a thermal and/or chemical (acid) treatment. This so-called Gauss-Eyring model is a bivariate log-normal distribution with temperature and time as independent variables. Model parameters standard activation enthalpy and entropy are directly related to reference temperature Tr and Z-value, commonly used in kinetic analysis in food microbiology. An essential feature of the kinetic model is that its parameters are treated as stochastic variables, owing to the underlying physics, based on the Lumry-Eyring model for unfolding of proteins using transition state theory. The performance of the model was evaluated using published data on enzyme inactivation and microbial inactivation including a wide range of temperatures and pH.

In Chapter 5, this Gauss-Eyring model was used to fit inactivation data of E. coli, L. monocytogenes, Lactobacillus plantarum, Salmonella Senftenberg and S. cerevisiae in orange juice. Thermal inactivation data was collected by exposing capillary tubes with target organisms to different temperature time combinations using a water bath to obtain inactivation kinetics, either via isothermal (a fixed temperature with varying holding time) or isotime (a fixed holding time with varying temperature) series. The model fitted well to the inactivation data of individual cultures. Variability between the different cultures of the five tested micro-organisms was observed. Therefore, an average value of the parameters of the individual cultures was used to predict inactivation as a function of temperature for a chosen (short) holding time.

In Chapter 6, a systematic evaluation of the individual effects of electric field strength and pulse width in combination with heat on the inactivation of E. coli, L. monocytogenes, S. Senftenberg, L. plantarum and S. cerevisiae in orange juice was carried out. A wide range of conditions has been tested, including both moderate intensity as well as high intensity PEF. Both electric field strength and pulse width were shown to be important for microbial inactivation. Inactivation kinetics of the tested conditions were compared to an equivalent thermal reference process, based on parameter estimates obtained in Chapter 5. A nonthermal pulse effect was observed for three specific sets of conditions in addition to the thermal effect responsible for inactivation. A non-thermal pulse effect was found for high intensity PEF treatment at E = 15 or 20 kV/cm and τ = 2 μs, but also at moderate intensity PEF condition of E = 2.7 kV/cm and τ = 1000 μs. The effectivity of this moderate intensity PEF condition was evaluated for E. coli and L. monocytogenes in watermelon juice and coconut water, varying in pH and conductivity. Interestingly, this moderate intensity PEF condition showed the same effectivity for all matrices in the pH range of 3.8 to 6.0, while high intensity PEF conditions at E = 20 kV/cm did show a strong dependence on product pH for microbial inactivation (Chapter 2). This suggests that a different mechanism is responsible for inactivation at moderate intensity conditions compared to high intensity conditions. Speculations on the mechanism responsible for inactivation are made in Chapter 8.

In Chapter 7, the impact of moderate intensity PEF (E = 0.9 − 2.7 kV/cm) and long pulse width (τ = 1000 μs) at variable maximum temperatures was evaluated on quality attributes of freshly squeezed orange juice, and compared to the impact of two thermal processes using either mild or severe pasteurisation conditions. No differences for pH and soluble solids were found after application of any treatment, and only small differences were observed for colour and vitamin C content after PEF and thermal treatment, mainly for the conditions applied at higher temperature. A large processing effect was measured in the enzyme activity of pectin methylesterase (PME), responsible for undesired cloud instability. Reduction of the remaining enzyme activity depended on the maximum applied temperature, and levels below the critical value to obtain shelf stable juices were found, showing that it is possible to select moderate intensity PEF conditions for adequate pasteurisation of fruit juices, both with respect to microorganisms and enzymes. The impact of processing on volatile flavour compounds was moderate when compared to untreated, although some deviations between moderate intensity PEF treated and thermally processed orange juice were found for individual compounds, with a better retention of the flavour compounds after application of moderate intensity PEF.

In Chapter 8, the main results of this thesis were discussed and concluding remarks and recommendations were presented. In conclusion, this thesis provided novel insight in the use of pulsed electric field processing as a mild pasteurisation process with improved quality as alternative to thermal pasteurisation of fruit juice. The work described in this thesis resulted in more insight in the individual effects of electric field strength and pulse width, and presented an effective combination of long pulse width and moderate intensity electric field strength as alternative PEF condition to the currently used high intensity electric field strength conditions for preservation. In addition, these moderate intensity/long pulse duration conditions are promising for industrial application, as they showed to be less sensitive for differences in the characteristics of the micro-organisms than high intensity PEF conditions and they are effective in both high-acid as well as low-acid products.