Elsevier

Journal of Food Engineering

Volume 212, November 2017, Pages 154-164
Journal of Food Engineering

A new method developed to characterize the 3D microstructure of frozen apple using X-ray micro-CT

https://doi.org/10.1016/j.jfoodeng.2017.05.028Get rights and content

Highlights

  • First study of the full 3D microstructural changes of apple tissue during freezing.

  • X-ray μCT was applied to study the 3D microstructural changes during freezing.

  • Different freezing methods show significant difference in the 3D microstructure of the frozen apple tissue.

  • A new method to characterize the ice crystals and the 3D microstructure in plant-based products.

Abstract

Non-destructive imaging techniques have become indispensable to improve insights into the microstructural changes occurring in fruit tissue during freezing. Here, apple cortex tissue samples were frozen using different freezing rates: (i) slow freezing (2.0 °C per min.), (ii) intermediate freezing (12.6 °C per min.) and fast freezing (18.5 °C per min.). Temperature-controlled X-ray micro-CT was optimized and then applied to visualize and quantify the 3D microstructure and ice crystal distribution at a pixel resolution of 3.8 μm. Ice crystal size distributions with mean equivalent diameters of about 41 ± 3.5 μm, 55 ± 7.4 μm and 71 ± 9.6 μm were obtained for the fast, intermediate and slow freezing rates, respectively. A new imaging methodology was developed and validated to allow segmentation of the ice crystals in frozen apple tissue using the X-ray attenuation coefficients of the reference model samples, i.e. frozen pure water and concentrated apple juice.

Introduction

Fruit is a highly perishable product and its quality and stability may change considerably during postharvest handling and processing (Canet, 1989). Freezing is one of the most efficient methods used to preserve fruit quality, and retards microbial growth, quality loss as well as slowing chemical reactions. Freezing thus prolongs the storage life of fruit and reduces losses (Zaritzky, 2000).

Water is the most abundant component in most types of fruit. Ice crystal formation during freezing may modify the tissue structure. Fruit poses a particular challenge with respect to freezing. The microstructure of fruit is composed of cells of different sizes and shapes, and interconnected cell walls and intercellular air spaces of different sizes and shapes. During freezing, liquid water crystallizes while the solute concentration in the remaining liquid water increases (Fennema, 1988). The formation of ice crystals will be formed throughout the cellular structure.

Regardless of the different freezing rates employed during freezing, the overall amount of ice formed depends mostly on the freezing temperature and other factors such as the chemical composition. In principle, however, the rate of heat removal defines the final quality of the frozen fruit with respect to ice crystal size, morphology and distribution. A fast freezing rate produces relatively numerous smaller ice crystals that are uniformly distributed (Chassagne-Berces et al., 2009). In food that consists of cellular tissue, smaller ice crystals cause less damage to the tissue structures (Petzold and Aguilera, 2009). On the other hand, a slow freezing rate often generates a limited number of larger ice crystals in the extracellular matrix. The solute concentration increases and decreases the water activity of the unfrozen phase. Water then migrates towards the extracellular ice crystal matrix. Formation of large ice crystals, accompanied by water displacement, involves gradual modifications of the fruit microstructure and produces undesirable effects, including cell damage leading to a loss of cellular material and turgor pressure (Zaritzky, 2000). Furthermore, undesirable changes are more likely to occur during the storage period. If the frozen fruit undergoes temperature fluctuations due to poor refrigeration control, the ice is subjected to melting-diffusion-refreezing cycles leading to crystal growth (Ndoye and Alvarez, 2015). Due to high surface free energy, small crystals are thermodynamically unstable and tend to reduce the energy by melting and re-freezing on the surface of the larger crystals (Alizadeh et al., 2007). In subsequent storage, they are resized and redistributed, leading to further microstructural changes. For frozen fruit, the microstructure is thus strongly linked to ice crystal morphology (size, shape and spatial distribution). The relationship between the microstructure and ice crystal morphology has attracted attention because these factors affect food quality attributes (Zaritzky, 2000, Aguilera, 2005). As a consequence, a more comprehensive insight into the microstructure of frozen fruit is required to elucidate quality changes during freezing and frozen storage.

The microstructure of frozen food has been widely imaged using optical microscopy, either directly by light microscopy (Caillet et al., 2003) or indirectly, after freeze-drying the food materials by means of scanning electron microscopy (SEM) (Russell et al., 1999, Chassagne-Berces et al., 2009). 3D imaging techniques produced without substantial sample preparation have become available (Mousavi et al., 2005, Mousavi et al., 2007, Verboven et al., 2008, Ho et al., 2013, Cantre et al., 2014, Herremans et al., 2015). Non-invasive and non-destructive techniques have been developed to describe the microstructure of foods and enable the generation of 3D structures, such as nuclear magnetic resonance imaging (MRI) and X-ray micro-computed tomography (X-ray micro-CT). The freezing process has been investigated by monitoring the 3D structure of ice crystals in a sucrose solution using MRI at an isotropic voxel size of about 78 μm (Mahdjoub et al., 2006). However, this spatial resolution is limited and larger than the typical microstructure of fruit (Cantre et al., 2014, Herremans et al., 2015). X-ray micro-CT (μCT) generates a series of X-ray radiograph shadows of the material acquired from different angles. The radiographies represent the attenuation of the X-rays that depends on the energy of the X-rays and material properties such as density and thickness. 3D images are then reconstructed from the radiographies. X-ray μCT has been proven to offer a powerful tool when used to investigate the fruit network structure (Cantre et al., 2014, Herremans et al., 2015). Moreover, its use to visualize the microstructure of frozen foods has been applied to freeze-dried foodstuffs (Mousavi et al., 2005, Mousavi et al., 2007). The samples were subjected to a freeze-drying process to lyophilize frozen water and leave holes representing ice crystals. The freeze-drying process, however, may have changed the structure investigated through shrinkage (Voda et al., 2012), resulting in inconclusive results. In frozen product, Pinzer et al. (2012) showed the applicability of X-ray μCT to investigate the 3D microstructure of ice cream with the help of a contrast agent. Therefore, the freeze-drying process can be avoided when the X-ray attenuation coefficients of the food components are sufficiently different.

The objective of this work was to develop and validate a methodology to visualize, characterize and quantify the tissue microstructure and ice crystals in plant-based products using a high-resolution X-ray μCT system. To this end, apple cortex tissue was considered as a model system with a microstructure consisting of cells and intercellular spaces of different sizes and shapes (Herremans et al., 2015). Five replicate apple samples were subjected to three different freezing protocols (corresponding to slow, intermediate and fast freezing rates) after which the microstructure and ice crystals were measured using X-ray μCT at a steady state temperature of −20 °C. An image analysis protocol was developed to segment frozen from liquid water, using reference samples of pure ice and sugar solutions. The resulting images were used to quantify and compare the pore spaces, liquid and ice phases among the different treatments.

Section snippets

Apple sample and preparation

‘Jonagold’ apples (Malus × domestica Borkh.) were harvested from an orchard in Rotselaar (Belgium). Apples were picked within their optimum picking window as determined by the Flanders Centre of Postharvest Technology (VCBT), Belgium, and immediately transported to the lab where the apples were stored under controlled atmosphere at 1 °C for 6 months. Prior to sample preparation, the apples were transferred to a room at 18 °C and stored overnight to equilibrate. Cylindrical samples of Jonagold

Ice crystal validation procedure

Fig. 2a shows the μCT slice of the frozen apple sorbet. The dark regions correspond to bubbles, the intermediate grey regions denote the ice crystals, and the light grey pixels represent the unfrozen phase. Fig. 2b shows the microscopic image of the frozen apple sorbet. The dark regions represent the bubbles and the roundish objects with dark boundaries are the ice crystals. Fig. 2c shows the cumulative probability density functions of the ice crystal size in four replicates acquired using the

Effect of freezing rate on the 3D microstructure

One of the major concerns of the frozen food industry is to implement a controlled cold chain to guarantee an optimal sensory quality, and thus the economic viability of the sector. In frozen plant-based products such as fruit, changes in the ice crystal size distribution in the cellular tissue during freezing, in subsequent storage and transport will affect their structural integrity and consumer perception. Our observations of ice crystal size distributions in ‘Jonagold’ apple as affected by

Conclusion

In this work, a non-destructive imaging technique was successfully developed to visualize, characterize and quantify the 3D microstructure and the ice crystal size distribution of apple cortex tissue during freezing under different conditions. The three different freezing protocols that were implemented showed a significant effect on the microstructure of the frozen apple tissue and produced different ice crystal size distributions with mean equivalent diameters of about 41 ± 3.5 μm,

Acknowledgements

The authors wish to acknowledge the financial support from DIM ASTREA, proposed by the Regional Council of Ile-de-France, France (Ast 140054), which enabled this research work to be conducted. Victor Vicent is a doctoral fellow under the joint doctoral programme organized by Irstea (France) and KU Leuven (Belgium).

References (26)

  • A. Voda et al.

    The impact of freeze-drying on microstructure and rehydration properties of carrot

    Food Res. Int.

    (2012)
  • N.E. Zaritzky

    Factors affecting the stability of frozen foods

  • AOAC International

    Official Methods of Analysis

    (2000)
  • Cited by (0)

    View full text