**Understanding the Frying Process of Plant-Based Foods Pretreated with Pulsed Electric Fields Using Frying Models**

**Zihan Xu 1, Sze Ying Leong 1,2, Mohammed Farid 3, Patrick Silcock 1, Phil Bremer <sup>1</sup> and Indrawati Oey 1,2,\***


Received: 11 June 2020; Accepted: 15 July 2020; Published: 17 July 2020

**Abstract:** Deep-fried foods (e.g., French fries, potato/veggie crisps) are popular among consumers. Recently, there has been an increased interest in the application of Pulsed Electric Fields (PEF) technology as a pretreatment of plant-based foods prior to deep-frying to improve quality (e.g., lower browning tendency and oil uptake) and reduce production costs (e.g., better water and energy efficiencies). However, the influence of a PEF pretreatment on the frying process and related chemical reactions for food materials is still not fully understood. PEF treatment of plant tissue causes structural modifications, which are likely to influence heat, mass and momentum transfers, as well as altering the rate of chemical reactions, during the frying process. Detailed insights into the frying process in terms of heat, mass (water and oil) and momentum transfers are outlined, in conjunction with the development of Maillard reaction and starch gelatinisation during frying. These changes occur during frying and consequently will impact on oil uptake, moisture content, colour, texture and the amount of contaminants in the fried foods, as well as the fried oil, and hence, the effects of PEF pretreatment on these quality properties of a variety of fried plant-based foods are summarised. Different mathematical models to potentially describe the influence of PEF on the frying process of plant-based foods and to predict the quality parameters of fried foods produced from PEF-treated plant materials are addressed.

**Keywords:** frying; mathematical model; mass transfer; heat transfer; pulsed electric fields; solid plant foods

#### **1. Introduction**

Pulsed Electric Fields (PEF) technology applies short (μs or ms) and repetitive electric pulses of high voltage to food materials placed between two conducting electrodes, leading to electroporation of cells [1]. When plant tissues are exposed to PEF, the charging process increases the transmembrane potential leading to the breakdown of proteins and the lipid bilayer within the cell membrane. When the transmembrane potential exceeds the range that cells can withstand, the cell membrane is punctured followed by the formation of pores. Pores grow in both size and quantity depending on the intensity of PEF treatment. Critical parameters of PEF processing include the electric field strength (*E*), pulse frequency (*f*), pulse number (*N*), pulse shape and polarity, specific energy input (*W*), pulse width (τ) and duration (*t*) [2]. In recent years, PEF has been recognised as an effective technology to improve

food quality and accelerate heat and mass transfers during food processing, while reducing energy consumption. PEF can be easily integrated into the food industry to assist existing unit operation, such as osmotic dehydration, freeze-drying, frying, freezing, thawing, extraction or clarification [3–5].

Frying (typically at 170 ◦C or above) is one of the oldest unit operations used by food processors; its goal is to produce final products with a crisp texture, an aromatic flavour and a golden-brown colour. There is a large variety of fried plant-based products available in the market, including fried vegetables such as potato, sweet potato, carrot, red beet, taro, celery bulb, squash, pumpkin, green bean and eggplant and fried fruits such as pineapple, apple, banana, peach, grape, guava, jamun and mango [6–9]. Among these, fried potato products (i.e., French fries and potato crisps) are the most widely consumed around the world [10]. Taking the production of French fries as an example, the process line consists of potato washing and sorting, skin peeling, preheating, cutting into fries, blanching, predrying, par frying and finally blast freezing. PEF is recommended to be applied as a pretreatment to the potatoes before the cutting step, potentially replacing the preheating step to reduce energy consumption while achieving equivalent process performance [11,12]. Under suitable operating parameters, PEF treatment will modify the structural and textural properties of potatoes, making them easier/more flexible to cut into fries, thereby reducing "feathering" and at the same time, extending the durability of the cutting blades [11,12]. With respect to the quality of the fried potato products, a lower browning tendency and a crispier texture compared to their non PEF-treated counterparts has been reported in the literature [11–14]. Such advantages offered by PEF in terms of process performance and product improvement make it appealing for the potato industry to adopt this technology.

Frying is a very complex process including simultaneous heat, mass and momentum transfers accompanied by a series of physical and chemical reactions [15,16]. During the frying process, heat is transferred from oil to fried foods leading to mass transfer (e.g., water evaporation and oil uptake). Apart from heat, mass and momentum transfers, chemical constituents (e.g., starch, reducing sugars, amino acids and water) within plant tissue react with each other during frying, and physical reactions (e.g., water evaporation and oil uptake) occur accompanied by structural changes [17]. Frying models, especially those built based on universal physical laws, may provide a better understanding of the frying process and its mechanism [15,18]. Additionally, the incorporation of observational data into kinetic frying models can give predictions of a specified quality of interest (e.g., colour, texture, oil uptake) for fried foods [19].

The purpose of this review is to provide an overview of the different mechanisms and reactions that occur during frying processes and explain how PEF pretreatment of plant materials can result in various quality improvements in the final products. The review will be concluded by discussing the potential use of frying models to explore the effect of PEF on frying to benefit further research and industry application.

#### **2. Frying Process of Food Materials**

Frying is a complex process which involves simultaneous heat, mass and momentum transfers (Figure 1), resulting in the flow of oil and water, phase changes and physicochemical reactions within the raw materials [17]. These reactions account for both the beneficial and the deleterious effects associated with fried foods.

#### *2.1. Heat Transfer*

Frying is an efficient and intensive heat transfer process owing to the high heat transfer coefficients and dynamic conditions within the frying system [20]. During frying, heat transferred from the hot oil to the surface of the cold food materials is driven by *convection* and from the surface to the inner side by *conduction* [18]. Then, water begins to evaporate from the food and creates vapour turbulence because of its rapid evaporation. Heat is transferred by forced convection contributing to the turbulence of the oil around the food [21]. Vapour turbulence enhances the heat transfer rate to its maximum level [22]. When large numbers of vapour bubbles cannot escape from the food material, they form an insulating

layer on the surface limiting heat transfer [23]. During frying, water evaporation dissipates some energy inside the food material, thereby decreasing the available energy for temperature increase. Normally, the change of energy for heat transfer within a nonhomogenous food material can be calculated using the following Equation (1) in Cartesian coordinates [24]:

$$
\rho C\_p \left( \frac{\partial T}{\partial t} + v\_x \frac{\partial T}{\partial x} + v\_y \frac{\partial T}{\partial y} + v\_z \frac{\partial T}{\partial z} \right) = k\_h \left( \frac{\partial^2 T}{\partial^2 x} + \frac{\partial^2 T}{\partial^2 y} + \frac{\partial^2 T}{\partial^2 z} \right) \tag{1}
$$

*T*: temperature, *Cp*: specific heat capacity of the material, *kh*: thermal conductivity of the material, ρ: density, *vi*: fluid velocity in *i*-direction. The right side of Equation (1) represents heat transfer due to conduction (dominates in case of solids) while the left side of the equation represents the unsteady term and convective heat-transfer terms.

In terms of frying, the energy balance equation is as follows [24]:

$$
\rho C\_p \frac{\partial T}{\partial t} = \frac{\partial}{\partial \mathbf{x}} k\_{eff} \frac{\partial T}{\partial \mathbf{x}} + \frac{\partial}{\partial y} k\_{eff} \frac{\partial T}{\partial y} + \frac{\partial}{\partial z} k\_{eff} \frac{\partial T}{\partial z} + \Delta H^{\text{sbl}} w \tag{2}
$$

*ke*ff: effective thermal conductivity, Δ*Hsbl*: latent heat of sublimation, *w*: water content.

There is a lack of information on the influence of PEF on heat transfer of food materials during frying, but current literature has demonstrated that PEF technology can accelerate the heat transfer during drying and osmotic dehydration processes [5]. Since the heat transfer coefficient increases with the rate of moisture transfer [25,26], an increase in tissue porosity and cell electroporation by PEF may increase the hydrodynamic permeability and the moisture transfer rate [27]. Therefore, the heat transfer and efficiency of drying or frying of PEF-treated plant materials is expected to be enhanced.

**Figure 1.** Potential influence of the Pulsed Electric Fields (PEF) pretreatment of potatoes on heat and mass transfer processes during frying. Images of untreated and PEF-treated raw potato tissue from scanning electron cryomicroscopy (cryo-SEM) were obtained from [28].

#### *2.2. Mass Transfer*

During the frying process, mass transfer mainly refers to the evaporation of water from the food material to the oil as well as the uptake of oil by food material [29]. Previous studies have demonstrated that PEF treatment enhances mass transfer in plant-based foods compared to non PEF-treated materials due to its electroporation effect on cell membranes, and subsequently this improves the diffusion of intracellular liquid/cell contents to the outside of the cells. [3]. Ignat et al. [13] have reported that PEF-treated potato cubes had a significantly higher drip loss (10.4%) compared to their non PEF-treated counterpart (2.3%) before frying. This phenomenon occurs due to PEF causing structural changes in the potato tissues which increase drip loss and facilitate the release of sugars which lowers the tendency for the tissues to brown during frying [13]. Moreover, PEF-treated potatoes exhibited a faster rate of water loss compared to non PEF-treated samples after 10 min baking at 100 ◦C [30]. Therefore, PEF pretreatment has the ability to enhance water diffusion from plant materials, which could promote a faster frying efficiency.

#### 2.2.1. Water Transfer

There are four main stages in the frying process which involve heat and mass transfers, namely the initial heating, surface boiling, falling rate and bubble endpoint periods [21,31]. In the *initial heating* stage, raw materials (at cold or ambient temperature) are dropped into the hot oil and are heated up gradually to the boiling point of water. During the *surface boiling* period, water begins to evaporate from the surface of materials along with the formation and release of bubbles leading to a rapid loss of water and the formation of pores at the surface becomes inevitable. A crust also begins to form at the outer surface. In the *falling rate* stage, the humid core region of the food material is heated slowly to the boiling point of water. Meanwhile, crust thickness increases and steam transfer speed decreases during this stage. In the *bubble endpoint* stage, water evaporation slows until bubbles are no longer being released on the surface.

Water transfer in the form of liquid (*Cl*) and vapour (*Cv*) during the unsteady diffusion process at the initial heating stage can be represented by the following equations in Cartesian coordinates [24]:

$$\frac{\partial \mathbf{C}\_{l}}{\partial t} = \frac{\partial}{\partial \mathbf{x}} D\_{i} \frac{\partial \mathbf{C}\_{l}}{\partial \mathbf{x}} + \frac{\partial}{\partial y} D\_{i} \frac{\partial \mathbf{C}\_{l}}{\partial y} + \frac{\partial}{\partial z} D\_{i} \frac{\partial \mathbf{C}\_{l}}{\partial z} - w \tag{3}$$

$$\frac{\partial \mathbb{C}\_{v}}{\partial t} = \frac{\partial}{\partial \mathbf{x}} D\_{i} \frac{\partial \mathbb{C}\_{v}}{\partial \mathbf{x}} + \frac{\partial}{\partial y} D\_{i} \frac{\partial \mathbb{C}\_{v}}{\partial y} + \frac{\partial}{\partial z} D\_{i} \frac{\partial \mathbb{C}\_{v}}{\partial z} + w \tag{4}$$

*Ci*: concentration of liquid (*i* = *l*) or vapour (*i* = *v*), *Di*: diffusivity of *i-*th species in the medium, *w*: water content.

During frying, the food material is transformed into a porous medium consisting of tiny void spaces (or small pores) that are interconnected and filled with fluid (liquid or vapour) [32]. Therefore, the movement of water vapour can no longer be described as diffusion. Darcy's law is considered to be more appropriate in describing the flow of water vapour inside the solid through the porous structure of fried foods [24]. Since plant materials typically contain a high water content (80–95%), Darcy's equation takes into account the significant pressure build up inside the porous food caused by the evaporation of internal water during frying [32]. Moreover, the resistance of the porous structure, which is proportional to the thickness of fried foods, can be integrated into Darcy's law formulating Equation (5) to better describe water loss due to vapour flow through the crust [24].

$$Q\_w = \frac{A \, k \, \Delta P}{\mu \, L} \tag{5}$$

*Qw*: flow rate of water vapour, *A*: cross-sectional area of the fried food, *k*: permeability of the crust layer (related to porosity), Δ*P*: pressure difference/drop over a given distance, μ: viscosity of the water vapour, *L*: thickness of fried foods.

#### 2.2.2. Oil Transfer

There are several mechanisms to explain oil uptake during frying including w*ater replacement*, *capillarity penetration*, the *cooling-phase e*ff*ect* and the *surface-active agent theory*, all of which are associated with water transfer and/or crust formation.

*Water replacement mechanism* explains that oil enters the plant tissues through the voids created by water evaporation, so water loss is considered to be the basis of oil uptake [33]. Gamble et al. [34] found that oil uptake was closely related to the water loss (R2 = 0.989) in potato slices in which the oil absorbed by potato slices was found accumulated in the voids left by the water evaporation. When the water bubbles escaped from food materials, they formed capillary pathways and increased the surface porosity [35].

*Capillarity penetration* describes how the oil moves upwards through narrow pores in fried materials when the adhesive intermolecular forces between the oil and food materials are stronger than the cohesive intermolecular forces in the oil [36]. The pressure difference (Δ*P\** ) between both ends of the capillary pathways mainly drives the capillarity penetration phenomena [37]:

$$
\Delta P^\circ = P\_2 - P\_1 = P\_{atm} - \left(P\_v - \frac{2\sigma\cos\theta}{r} \pm \rho gh \cos\alpha\right) \tag{6}
$$

*Pi*: pressure at the point *i* (*P2:* pressure at the pore surface, *P1:* pressure at the deepest pore point inside the food material, *Patm:* atmospheric pressure, *Pv:* water vapour pressure), θ: contact angle between the oil and the food material, *r*: pore radius, σ: surface tension of the oil, ρ: oil density, *g*: acceleration gravity, *h*: height of the capillary motion, α: angle between the capillary pathway and vertical direction.

Capillary penetration can also be described by the Washburn equation:

$$Q\_{oil} = \frac{\pi r^4 \,\Delta P^\*}{8 \,\mu \, h} \tag{7}$$

*Qoil*: volumetric flow of laminar oil, π: ratio of an oil circumference to its diameter (3.142), *r*: pore radius, μ: oil viscosity.

The penetration of oil over time can be calculated based on the modification of the above two Equations (6) and (7) to yield Equation (8) [37]:

$$\frac{dh}{dt} = \frac{r^2}{8\mu} \left( P\_{atm} - P\_v \right.\\ \left. + \frac{2\,\sigma\,\cos\theta}{r} \pm \rho\,\,\text{g}\,\,h\,\cos\alpha \right) \tag{8}$$

However, the voids or capillary pathways are always filled with water during frying and the inner steam pressure may resist oil penetration. Sometimes, the oil is absorbed after the capillary (food material) is removed from the oil [35,37].

The *mechanism of cooling-phase e*ff*ect* explains oil uptake after the food material has been removed from the oil and this is caused by water vapour condensation and internal pressure reduction during the cooling period. A study by Ufheil and Escher [38] reported that most of the oil in fried potatoes is absorbed into the porous crust after their removal from the oil, implying that oil absorption and water loss are not synchronous and that the cooling-phase effect plays a key role in oil uptake. Oil uptake after removal of the plant tissue from the oil is a balance between the oil drainage and oil adhesion [17]. Adhered oil on the surface of fried material is absorbed due to the "vacuum effect" caused by the condensation of steam during cooling period. When the fried food is removed from the oil, an oil film is formed on the surface and its thickness (*H*) can been calculated by the Landau–Levich–Derjaguin equation [39]:

$$H = 0.944 \frac{\left(\mu \mathcal{U}\right)^{\frac{2}{3}}}{\mathcal{V}^{\frac{1}{6}} (\rho \ g)^{\frac{1}{2}}} \tag{9}$$

μ: oil viscosity, γ: surface tension, *U*: speed of oil removal after frying; ρ: oil density, *g*: acceleration gravity.

Finally, the *surface-active agent theory* has been proposed in addition to the other oil uptake mechanisms. In this theory surface-active agents (e.g., monoglycerides and diglycerides) produced by oil degradation and hydrolytic reactions during frying enhance the interactions between oil and fried food leading to increased oil absorption [35]. Such surface-active agents can increase the foaming tendency of oil and reduce the interfacial tension leading to the increase of surface hydrophobicity [40].

#### *2.3. Momentum Transfer*

Momentum transfer is a physical phenomenon that involves convection mechanism between molecules or groups of molecules within the food material [41]. It depends upon the interrelation of the fundamental variables of mass, velocity and time and of changes in the velocity per unit mass [41]. During frying, momentum is transferred by *convection* (vapour leaving the fried materials) or by *molecular forces* (viscous stress or pressure) [42]. Momentum transfer equations are based on the principle that the momentum is conserved in a phase. The momentum balance equation, which contains three velocity components and the *x*-component equation in Cartesian coordinates [15,43] is as follows:

$$\rho \left( \frac{\partial v\_{\mathbf{x}}}{\partial t} + v\_{\mathbf{x}} \frac{\partial v\_{\mathbf{x}}}{\partial \mathbf{x}} + v\_{\mathbf{y}} \frac{\partial v\_{\mathbf{x}}}{\partial \mathbf{y}} + v\_{\mathbf{z}} \frac{\partial v\_{\mathbf{x}}}{\partial \mathbf{z}} \right) = \left( \frac{\partial}{\partial \mathbf{x}} \mu \frac{\partial v\_{\mathbf{x}}}{\partial \mathbf{x}} + \frac{\partial}{\partial y} \mu \frac{\partial v\_{\mathbf{x}}}{\partial y} + \frac{\partial}{\partial z} \mu \frac{\partial v\_{\mathbf{x}}}{\partial z} \right) + \rho \, g\_{\mathbf{x}} \, \beta \left( T - T\_{\text{os}} \right) \tag{10}$$

*vi*: fluid velocity of *i* component, ρ: fluid density, *gx*: acceleration gravity of *x*-component, β: thermal expansion coefficient.

Despite there being no studies directly reporting the influence of PEF pretreatment on momentum transfer in plant materials during frying, Dellarosa et al. [44] have shown that PEF could help to gain momentum to increase mass transfer in plant-based foods during processing, which suggests that the momentum transfer in PEF-treated plants would increase during frying.

#### **3. Changes in the Chemical Constituents of Plant Food Material during Frying**

Plant foods are a good source of carbohydrates (e.g., simple sugars, starch) and proteins. Frying of plant foods at high temperatures (>170 ◦C) initiates a series of complex chemical reactions in these constituents when reacting with oil, and thus generating compounds that can affect the quality of final products, by influencing their flavour, colour, shelf life and nutrient composition [45]. The two chemical changes that this review will focus upon are Maillard reaction and starch gelatinisation.

#### *3.1. Maillard Reaction*

The Maillard reaction is one of the most important chemical reactions that occur during frying because it modifies many quality parameters in the final fried product such as colour, flavour, taste, nutritional value and the level of toxic compounds (e.g., acrylamide) [46]. The Maillard reaction refers to the reaction between an amino group (e.g., amino acids) and a carbonyl group (e.g., reducing sugars). Firstly, a Schiff base is formed and rearranged to Amadori or Heyns products, which then undergo enolisation and are subsequently modified to form reactive α-dicarbonyl compounds, the source of brown pigments production [46,47]. These compounds can react with additional nucleophiles (i.e., guanidines, amines and thiols) and undergo Strecker degradation producing Strecker aldehydes. Furthermore, advanced glycation end products are produced in a series of downstream reactions, and further chemical reactions form a large number of polymerised products, named melanoidins, which result in colour darkening [46]. Temperature, time, reactant type (amino and carbonyl groups) and concentration, water activity and pH are the main factors that influence the Maillard reaction [48]. Most reactants, primarily reducing sugars and amino acids involved in the Maillard reaction are accumulated in vacuoles in the plant's tissues. Since PEF treatment induces a cell electroporation effect on plant tissues which results in leakage, it is not surprising that a considerable amount of the reactants might be released/leached out from the cells, as demonstrated in the potato studies by Janositz et al. [30] and Genovese et al. [49]. An increase in the release of reactant materials from PEF treated potatoes is generally beneficial for high sugar containing potatoes (due to genotype, maturity, and poor storage management), which can exhibit excessive browning when fried. This is because the

overall pool size of the reactants in these potato tissues, after a PEF-pretreatment, that is available to participate in the Maillard reaction and caramelisation during frying can be reduced.

#### *3.2. Starch Swelling and Gelatinisation*

Potato tubers are high in starch, which is a semicrystalline biopolymer. When potatoes are cooked, starch starts to swell and its gelatinisation forms dense and starch-rich areas within the cells, which helps to reduces oil absorption upon frying and resist dehydration and shrinkage, while improving the texture of the cooked products [35]. Starch gelatinisation refers to the collapse of molecular organisation in the starch granule, leading to irreversible changes to its molecular properties resulting in water uptake, swelling, loss of crystallinity, loss of birefringence, unwinding of double helices, starch solubilisation and an increase in viscosity [50]. During frying, starch swelling begins at 60–70 ◦C and the process is completed when the swollen starch granules completely occupy the interior of the cells, which enable them to resist oil penetration [51,52]. Moreover, the extent of starch gelatinisation in the crust and core regions of potato crisps can be very different owing to different rates of heat transfer and water availability [53].

Numerous studies investigating the effect of PEF on starch isolated from different plant sources, have consistently demonstrated that the application of a high intensity PEF treatment (30–50 kV/cm) can alter starch structure and its inherent properties [54–56]. Under the influence of PEF, the crystalline region and the side chains of the amylopectin in the starch can be modified leading to a decrease in the double helix binding force and disruptions to the intragranular molecular arrangement of starch granules. Such changes in the starch promote reactions between water molecules and starch chains, reducing the energy required for starch gelatinisation. Starch gelatinisation temperature and enthalpy of gelatinisation were found to decrease with an increase in PEF intensity [54], suggesting that the starch may become more susceptible to gelatinisation after PEF treatment.

#### **4. An Overview of the Quality Parameters of Fried PEF-Treated Plant Food Materials**

While it is clear that PEF pretreatment of plant materials can influence a range of reactions that occur during frying, including altering the heat and mass transfers, Maillard reaction and starch gelatinisation, all of these processes are expected to influence the quality of the final products. Table 1 summarises how PEF treatment can affect the quality parameters (e.g., colour, moisture content, oil uptake, texture and toxic compound) of plant-based foods, specifically on potato, during frying.

#### *4.1. Colour*

Colour is considered the most important parameter contributing to the visual perception of the quality of foods, and it influences the acceptance and choice of consumers. Apart from Maillard reactions, oil degradation may also affect the colour of fried foods. For example, a high correlation (R2 > 0.9) has been reported between the dark colour of fried tortilla chips and oil degradation time [57]. The polymerisation of triglycerides and the products of triglyceride hydrolysis, such as free fatty acids, monoglycerides and diglycerides, during oil degradation may result in changes in colour [57]. Process variables including raw material properties (i.e., reducing sugar content, amino acids content, protein content and dimensions), frying temperature, time and oil type can also affect the colour of fried food [58].


 **1.** Application of PEF on potato tissues followed by frying process.

**Table**

studies to investigate the quality of food material using a kinetic approach and is not

recommended

 for actual practice.

The effect of PEF treatment on the colour changes of different types of fried plant-based foods has been reported. Specifically considering potatoes, studies by Ignat et al. [13] and Genovese et al. [49] have both reported that the final products produced from PEF-treated potatoes are less brown with a uniform and bright colour after frying. A similar finding was also observed in fried chips produced from PEF-treated sweet potato [19].

#### *4.2. Moisture Content*

When a plant material is placed into hot oil, its surface temperature increases rapidly and surface water rapidly evaporates in the form of water bubbles when its temperature rises to 100 ◦C [21]. As a result, the surface dries quickly and forms a crust. This crust acts as an additional barrier to the escape of water from the inner regions; therefore, the inner region is always moist compared to the outer region [60]. A number of other factors can also affect the moisture content of fried foods such as potato crisps, including frying temperature and time, the size and shape of the product and prefrying procedures such as drying [61].

PEF treatment prior to frying can help to accelerate moisture loss from plant materials during frying based on the results gathered from previous studies [14,30,61]. The effect of fast moisture loss from PEF-treated food material on the hydrolysis of oil can lead to deterioration to frying life of the oil, and hence further research in this aspect is recommended. However, moisture evaporation from fried food can also form a "steam blanket" (or physical barrier) on the surface of the oil and thus prevent the contact between atmospheric oxygen and oil, limiting the oxidation of frying oil [62].

Moreover, a faster reduction in water content has been observed in PEF-treated potato discs/slices during air-drying process followed by frying compared to untreated potato samples [14,59]. Such findings indicate that PEF could be used to enhance/facilitate the overall drying and frying efficiency through accelerating the moisture diffusivity and water loss owing to the cell electroporation effect on plant cell membrane, which results in a greater diffusion of intracellular water out of the cells. However, when most of the surface water within a potato tissue has been evaporated during frying, the influence of PEF on water loss may become insignificant [59].

#### *4.3. Oil Uptake*

Oil is the most important ingredient for frying since it drives heat and mass transfer during frying. As discussed in Section 2.2.2, most of the oil is absorbed by entering pores through which moisture escapes during frying. Factors, such as frying conditions (time, temperature, food-to-oil ratio, repeated frying), oil characteristics (quality and type) and food characteristics (size, shape, surface roughness and porosity) and pretreatments (blanching, edible coating, vacuum drying, PEF, etc.) may influence the amount of oil absorbed [63].

Both Liu et al. [14] and Janositz et al. [30] have reported that the oil uptake for PEF-treated potato slices (1.5–2.5 mm thickness), after frying, was between 34 and 39% lower compared to untreated samples. Therefore, PEF is an effective pretreatment option for solid plant materials in order to reduce oil uptake during frying and may have implications on the management of frying oil and the quality of fried foods. There are several possible reasons to explain this phenomenon. Firstly, PEF treatment appears to increase the porosity of plant tissue leading to a higher vapour pressure and faster vapour movement to the surface, thus limiting the oil absorption [30]. Secondly, a smoother cut surface (less "feathering") of PEF-treated plant material results in a significant reduction in the amount of adhered oil after frying [19]. Thirdly, since PEF pretreatment can hasten starch gelatinisation during frying, this promotes the formation of an impermeable surface layer, which consequently acts to inhibit oil absorption [14]. Another likely phenomenon could be the cell electroporation effect of PEF in conjunction with an increase in the formation of capillaries resulting in an increased tendency for the oil to leach out from the pores/capillaries and thus leave the surface of the fried food.

#### *4.4. Texture*

Fried foods are expected to have a crispy crust and a moist and soft interior [17]. Crispness, porosity and shrinkage are regarded as the main textural indicators of fried foods.

#### 4.4.1. Crispness

Crispness describes a quick fracture under small strain stresses owing to a low water content in the surface layer of a thick fried piece or through a thin slice [17]. It is important to note that the effect of PEF treatment on the crispness of fried foods cannot be assumed since this textural parameter is highly dependent on the par-frying time and temperature, oil content, moisture content, starch content, porosity and roughness of surface [11,61]. Cahayadi et al. [64] reported that fried crisps produced from PEF-treated potatoes were perceived to be crunchier compared to potato crisps from untreated potatoes, and this texture modification of the potato crisps has shown to increase the perceived satiation of an individual and thereby reduce energy intake from snack consumption.

An important factor contributing towards the texture of French fries is their starch content and therefore excessive starch loss during the production of French fries should be avoided [11]. While it has been recognised that PEF treatment can induce the formation of irreversible pores in cell membranes causing leakage of cell contents (e.g., free sugars and amino acids), this electroporation effect on plant tissue may not necessary impact severely on the leakage of polymers such as starch. The retention of starch has been observed in PEF-treated potatoes subjected to industrial scale French fries production [12]. In fact, potato starch granules with an average diameter of 40 μm, which possibly became larger after PEF treatment [55], were not able to pass through the PEF-induced membrane pores that had a maximum size of around 5 μm [65].

#### 4.4.2. Porosity

During frying, some water vapour is unable to move through the food material because of restrictive intercellular diffusion. As a result, superheated vapour distorts the pores and leads to the formation of a porous structure [66]. Normally, intense water evaporation results in the formation of more pores that are larger [66]. There are many factors influencing the porosity of fried foods, including frying conditions (time, temperature, and pretreatments), plant material characteristics (size, shape, component, density) and oil types. For example, it has been reported that the porosity of fried foods increased with increasing oil temperature along with the use of hydrogenated oil [66].

Though the effect of PEF on the porosity of fried foods is yet to be reported in the literature, it is reasonable to expect that PEF treatment may increase the porosity of fried foods according to the observations from PEF-assisted drying process. An increase in surface porosity was found in air-dried products (e.g., apple cubic slabs, apple slices and potato slices) due to a PEF pretreatment [16]. This is likely to occur owing to enhanced heat and mass transfer rates caused by PEF, resulting in an intense water evaporation and thus, an increase in the size and number of pores.

#### 4.4.3. Shrinkage

Shrinkage is caused by water loss, resulting in the reduction of open pores and an increase in the density of fried products [61]. During frying, shrinkage initially occurs at the surface accompanied by the formation of a rigid outer layer. Shrinkage then moves inwards until the final volume is fixed. Shrinkage phenomena are related to the frying time and temperature, shape, size and density of the food materials [61]. For example, it has been reported in fried sweet potatoes that shrinkage is more pronounced with increased frying time and higher frying temperatures [67].

No research has been published studying the influence of PEF on the shrinkage of fried foods, but previous research on PEF-assisted drying process has suggested that PEF pretreatment could lead to product shrinkage. Air-dried products from PEF-treated carrots and red beetroots shrink more

compared to non-PEF treated samples owing to tissue damage caused by the cell electroporation effect [27,68].

#### *4.5. Toxic Compounds*

Toxic compounds that may be produced during frying include acrylamide, hydroxymethylfurfural, furan, ethylcarbamate, heterocyclic amines, polycyclic aromatic hydrocarbons and nitrosamines, all of which are potential health risks to consumers [69]. The presence of acrylamide in fried potato products (typically an average of 400 ng/g, where the upper recommended limit by the European Commission is 1000 ng/g) is unavoidable [70] due to the high concentration of acrylamide precursors, reducing sugars and asparagine in potato tubers [71]. Acrylamide has genotoxic, neurotoxic and carcinogenic risks in animal [72]. Most acrylamide in fried foods are formed by the Maillard reaction between reactive carbonyls and asparagine at a temperature of over 120 ◦C through a series of intermediates [20]. Factors such as frying conditions (temperature, time, pH and pretreatment) and material characteristics (cultivar, chemical constituents, water activity) may influence the acrylamide content in fried plant-based foods [20].

A recent finding by Genovese et al. [49] was that the acrylamide content of potato crisps produced from PEF-treated potatoes was about 30% lower than those for untreated crisps. PEF may considerably reduce, through diffusion, the reactant content available to participate in the Maillard reaction and hence the formation of toxic compounds during frying is inhibited.

#### **5. The Use of Frying Models to Describe the Frying Process of PEF-Treated Plant Materials**

Different frying models have been proposed in the literature (Figure 2). *Physical models* are built based on universal physical laws and the underlying mechanisms behind frying process, so their predictions are usually more precise and based on fundamental physical phenomena. *Observational models* are built based on the fitting of experimental data, so they are also known as data-driven models. *Kinetic models* are built to describe the rates of chemical reactions relevant to some universal physical laws by fitting experimental data into the model. Therefore, kinetic models are classified under both physical and observational models [73]. In this review, selected frying models will be briefly discussed and then evaluated regarding their suitability to be used in describing the influence of PEF treatment on the frying process of plant-based foods.

#### *5.1. Physical Models to Describe Heat, Mass and Momentum Transfers*

Physical frying models require equations that describe changes in mass, heat and momentum transfers, and simultaneously considering the phase changes and physicochemical changes of plant-based foods during frying. They can provide an in-depth understanding of the physical process of frying, and they are usually more precise because the models are built based on the universal laws. Most physical models for frying are macroscopic continuum models of heat, mass and momentum transfers [73]. However, the coordinate systems, suitable equations and boundary conditions may vary between different frying conditions. Because of the complexity of frying process, different types of physical models have been built (Table 3). These can be divided into *simple di*ff*usion-based*, *crust-core moving boundary* and *multiphase porous media* models.

**Figure 2.** Summary of mathematical models suitable to describe frying process and the model building process.

The procedures of building physical models for the frying process are straightforward (Figure 2). The first step is to define the purpose and describe the questions to be solved with this model. Then, some assumptions (about shapes, geometrical dimensions, mass or heat transfer coefficients, material properties and volume changes) are usually made to simplify the complex real-life situation [74]. The governing equations for heat, mass and momentum transfers are the core of theoretical models and the typical equations vary between different types of models (Table 2). *Heat transfer* is usually modelled by conservation of heat equation and Fourier's equation, *mass transfer*is generally modelled by conservation of mass equation and Fick's law of diffusion and *momentum* transfer is always described by conservation of momentum equation and Navier–Stokes equation [74]. Each governing equation has its boundary conditions, which reflect the interaction between the material being fried and surroundings, in order to describe the frying process accurately. Knowledge of boundary conditions occurring during the frying process are required to solve these equations numerically. The commonly used methods to obtain solutions for the boundary conditions include the finite difference, finite element, finite volume, boundary elements, lattice gas cellular automata and lattice Boltzman methods, and many of them rely on the commercial computational software, such as the computational fluid dynamics and computer aided food process engineering. Then input parameters such as density, specific heat capacity, thermal conductivity and permeability are introduced to the model. After building the theoretical model, experimental data from frying process is used to verify the model.


**Table 2.** Typical equations to model the frying processes.

N/A= data not available.

#### *Foods* **2020** , *9*, 949

The *simple di*ff*usion-based frying model* is the simplest physical model to describe frying. It is built based on the simple heat conduction and moisture diffusion processes while ignoring the oil absorption and water evaporation altogether in the fried material. Rice and Gamble [75] attempted to build a one-dimensional water diffusion model combining both Fick's first law and Arrhenius relationship to predict the moisture loss during the frying process of potato slices and successfully proved that the model made valid predictions regarding the early stage of frying (within the first 180 s). Likewise, Pedreschi et al. [76] applied Fick's law of diffusion with constant and variable effective diffusion coefficients in order to model the water loss during frying. They found that a simple diffusion-based frying model that considered the change of diffusivity coefficient value with frying time was relatively more precise (better fit to experimental data) in predicting the moisture content of potato slices compared to a classic model with a constant effective moisture diffusion coefficient. However, it is important to note that simple diffusion-based frying models are only able to provide a limited understanding of the frying process because the complex pressure driven flow is simplified to effective diffusion and empirical parameters, which influences their accuracy and restricts the application of these models for different plant materials and frying conditions [77].

The *crust-core moving boundary model* is built based on the core and the crust regions formed in the food material during the frying process, taking into account the moving boundary, where the interface between the core and crust regions moves [31]. This model is expected to be more precise than the simple diffusion-based frying model because it considers the diffusional and pressure driven transports as well as the distributed evaporation. For example, a good agreement was found between the experimental data (water content, centre temperature, surface temperature and crust thickness) and predicted values using a one-dimensional moving boundary model that included pressure-driven flow, albeit ignoring diffusion flow in the crust region [78]. Other types of crust-core moving boundary models have been built dealing with different conditions. Acknowledging the temperature difference at different positions of a fried food, Southern et al. [79] developed a moving boundary model using the Fourier's law and energy balance equation to describe heat transfer in both the core and crust regions and significantly improved the theoretical prediction of the experimental temperature-time values in different locations of potato crisps during frying. Moreover, van Koerten et al. [80] built a crust-core moving boundary model based on a Nusselt correlation connecting heat transfer coefficient and water evaporation rate, which was demonstrated to be a simple but effective model for predicting water evaporation and temperature profile in potato cylinders of different diameters (8.5, 10.5 and 14 mm). To allow the models to be applied more widely for different conditions, Farid and Kizilel [42] developed a unified moving boundary model, by defining a parameter which could reflect the extent of mass diffusion relative to thermal diffusion, to predict the temperature and moisture distribution in a food material, which can be applied in any drying and frying processes. The unified model successfully described the temperature and moisture distributions during the frying and air-drying process for thick (25.4 mm) and thin (2–3 mm) potato slices. Thus, crust-core moving boundary model can be a suitable model to describe the frying process regardless of the dimensions and positions of the fried materials. However, the rate of heat transfer during frying process is highly dependent on the food properties, such as the thermal conductivity and water diffusivity and hence, they should be considered in the mathematical model [81]. In addition, analytical solutions for complicated equations and boundary conditions are sometimes unavailable. Application of PEF to plant materials is expected to alter the microstructure and physicochemical properties (e.g., thermal conductivity and water diffusivity), hence influencing the moving boundary of crust-core regions. Therefore, crust-core moving boundary model could be considered in future research to describe the differences of frying process between untreated and PEF-treated plant-based foods in a simple yet precise manner.

*Foods* **2020**, *9*, 949

The *multiphase porous media model* is built based on the simultaneous heat, mass and momentum transfers. Multiphase transport in a porous media can be due to three underlying mechanisms including the molecular diffusion for gases, capillary diffusion for liquids and convection (pressure driven or Darcy flow for liquids or gases) [32]. The multiphase porous media model is considered more realistic, comprehensive and can provide better insight into frying process because it includes the temporal and spatial profiles of temperature and the transport mechanisms of water (liquid and vapour) and air inside the fried materials [82,83]. Researchers have applied the multiphase porous media model for different frying conditions, and hence, are able to explain the frying process from multiple perspectives. For example, Ni and Datta [84] developed a multiphase porous media model to predict the temperature, moisture, oil uptake and crust thickness of potato slices that took into consideration the pressure driven flow for the oil, vapour and air phase in the porous medium. This model shows that there is a pseudo-steady state region in the dry crust and a transient diffusion-like profile in the wet core but it becomes spatially uniform with frying time. Similarly, Halder et al. [77] have also developed a multiphase porous media model for potato frying and postfrying cooling process based on the nonequilibrium equation for evaporation. The estimated heat and mass transfer coefficients accurately reflect the process of different phases including the nonboiling phase and surface boiling and falling rate stages in the boiling phase. As a result, there is a reasonably good agreement between the experimental data and predicted values of quality parameters, such as the oil content, crust thickness and acrylamide content and this model can be applied to describe baking, meat cooking and drying processes with minor modifications. Multiphase porous media model can be applied to vacuum frying, where Warning et al. [85] have developed a model of potato chips by modifying the Darcy's law to account for the Klinkenberg effect. It works well to predict the moisture, temperature, pressure, oil content and acrylamide content during vacuum frying and it implies that the core pressure is approximately 40 kPa higher than the surface pressure of the potato chips. While it is clear that multiphase porous media models are capable of describing the multifaceted physics behind frying process, some of them are difficult to implement in real-life scenario due to the complexity of calculating the evaporation rate. Application of PEF to plant materials is expected to influence the porosity of raw materials. Therefore, multiphase porous media model could be suitable to describe the frying process of PEF-treated plant materials since the model can describe the heat and mass transfers inside a porous material in conjunction with the air flow outside the porous material, and phase changes such as evaporation and condensation [86].


Frying parameters: *T* = frying temperature;

 and *t* = frying time.

*Foods* **2020**, *9*, 949

#### *5.2. Observational Models for Quality Prediction of PEF-Treated Fried Foods*

Observational models are useful in building a relationship between the input and output parameters, especially when the practical situations are too complex to understand and building a physical model becomes unrealistic. When building observational models, assumptions and physical interpretations are not necessary since only experimental data are needed to fit a mathematical equation. Commonly used observational models includes *classical statistical*, *artificial neural network*, *genetic algorithm*, *fuzzy* and *fractal analysis* models [73]. Observational models are particularly useful to help optimise the frying process in order to obtain high-quality fried products. For example, statistical and two-stage fuzzy models have been built to optimize the blanching and frying parameters (e.g., oil temperature, thermal power) resulting in quality improvements of fried foods and efficiency of frying process [92,93]. Artificial neural network models have become popular in recent years and are considered to provide an accurate quality prediction. These models are built by selecting proper network structures and tuning model parameters such as weight, connections and threshold values until the fit to the experimental data is maximised. Mohebbi et al. [91] built a genetic algorithm-artificial neural network model which could predict the moisture and oil content of fried mushroom accurately. However, the predictive results from observational models can be significantly affected when the physical properties or environmental conditions are altered [77]. To study the influence of PEF on some reactions (e.g., Maillard reaction, starch gelatinisation, etc.) and quality parameters (colour, acrylamide content, etc.) of fried foods during frying process, observational models would be suitable because they can predict the results directly without understanding the complex reaction process.

#### *5.3. Kinetic Models for Predicting the Rate of Frying Reactions after PEF Treatment*

The quality of fried foods changes with frying time owing to chemical and physical reactions during the frying process. The changes of food composition and quality parameters are always described by the kinetic rate of frying reactions, which may also influence the heat and mass transfers [81]. The chemical, physical and biochemical changes such as degradation and formation of substances (amino acid, sugar, acrylamide, etc.), texture degradation and starch gelatinisation can be evaluated by kinetic models [94,95]. A clearer understanding of molecular level changes during the frying process can be obtained from kinetic models because they contain characteristic kinetic parameters such as rate constants and activation energies [96]. The frying reactions for PEF-treated plant materials can be influenced by the treatment intensity, and accordingly, alter the kinetic parameters of the chemical, physical and biochemical reactions occurring during frying.

The basis of a kinetic model for predicting quality changes is as follows:

$$\frac{dP}{dt} = \pm k \, P^n \tag{11}$$

*P*: quality parameters, *k*: rate constant, *t*: time, *n*: reaction order.

Normally, reaction orders can be determined from the characteristics of a chemical reaction. The reaction orders may vary under different circumstances due to the complicated nature of food systems. As reported in the literature, the time dependency of chemical reactions and quality changes of texture, colour and nutrient content usually follows a zero or first order of reaction [97,98].

The temperature dependency of the rate constant (*k*) can be modelled using Arrhenius equation and Eyring's absolute reaction rate theory model. The Arrhenius model is very common in the food industry for quality prediction, which can be written as:

$$k = A \exp^{\left(\frac{-\tilde{\mathcal{R}}^{\tilde{\mathcal{A}}}}{kT}\right)} \tag{12}$$

*A*: pre-exponential factor, *Ea*: activation energy, *R*: ideal gas constant (8.3136 J/molK), *T*: absolute temperature (K).

As for the Eyring's absolute reaction rate theory, it is formed based on the transition state theory:

$$A + B \stackrel{k\_1}{\leftrightarrow} A \mathcal{C}^+ \stackrel{k\_2}{\rightarrow} \mathcal{C} \tag{13}$$

*A* and *B*: molecules, *AC*+: activated complex or transition state, *C*: product.

The Eyring equation is as follows:

$$k = \frac{k\_b \ T}{h} \exp^{-\left(\frac{\Delta S^+}{R} - \Delta H^+\right)}\tag{14}$$

<sup>Δ</sup>*S*+: activation entropy, <sup>Δ</sup>*H*+: activation enthalpy *kb*: Boltzmann's constant (1.381 <sup>×</sup> 10−<sup>23</sup> J/K), *<sup>h</sup>*: Planck's constant (6.626 <sup>×</sup> <sup>10</sup>−<sup>34</sup> Js).

The application of Eyring's absolute reaction rate theory model has not been widely used to describe frying process, but it is possible to describe some underlying mechanisms during this process. For example, Moyano and Zúñiga [99] used the enthalpy–entropy compensation approach based on Eyring's absolute reaction rate theory to study the colour kinetics of French fries during frying. The results indicated that the activation entropy decreased during frying because of the limited space for molecular movement after drying.

#### **6. Concluding Remarks and Future Directions**

Frying is a multifaceted process involving the occurrence of heat, mass and momentum transfers that change the physical and chemical states of the food. The application of PEF treatment to plant-based foods is likely to enhance heat, mass and momentum transfers and is an effective emerging technology capable in controlling Maillard reaction and promoting starch gelatinisation, during frying. As a result, numerous quality improvements such as reduction in oil uptake and toxic compound in the final products and improvement in process efficiency such as decreases in energy cost and consumption, have been reported owing to the implementation of PEF technology in the frying industry. However, the underlying mechanisms of how PEF-treated plant materials withstand the frying process and its impact on quality parameters are not yet fully understood. Frying models are widely used to describe the frying process and to predict quality changes but frying models to describe the frying process of PEF-treated plant-based foods have yet to be developed. Building and validation of frying models, either observational, physical or based on kinetics, is of future research interest in order to explain better the underlying mechanisms and influence of PEF pretreatment on fried foods. Moreover, appropriate frying models can then be applied universally across the frying industry to aid the optimisation of PEF processing parameters on a wide range of plant materials in order to achieve the desired quality parameters in the final fried products.

**Author Contributions:** Conceptualisation, I.O.; methodology, Z.X.; investigation, Z.X.; writing—original draft preparation, Z.X.; writing—review and editing, S.Y.L., I.O., M.F., P.B., P.S.; visualisation, Z.X.; supervision, S.Y.L., I.O.; project administration, I.O.; funding acquisition, I.O. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work is supported by the Food Industry Enabling Technologies (FIET) program funded by the New Zealand Ministry of Business, Innovation and Employment (contract MAUX1402).

**Acknowledgments:** Leong and Oey are affiliated to the Riddet Institute, a New Zealand Centre of Research Excellence, funded by the Tertiary Education Commission.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Heat and Mass Transfer Modeling to Predict Temperature Distribution during Potato Frying after Pre-Treatment with Pulsed Electric Field**

**Gohar Gholamibozanjani 1, Sze Ying Leong 2,3, Indrawati Oey 2,3, Phil Bremer 2, Patrick Silcock <sup>2</sup> and Mohammed Farid 1,\***


**Abstract:** Based on unsteady state heat conduction, a mathematical model has been developed to describe the simultaneous heat and moisture transfer during potato frying. For the first time, the equation was solved using both enthalpy and Variable Space Network (VSN) methods, based on a moving interface defined by the boiling temperature of water in a potato disc during frying. Two separate regions of the potato disc namely fried (crust) and unfried (core), were considered as heat transfer domains. A variable boiling temperature of the water in potato discs was required as an input parameter for the model as the water is evaporated during frying, resulting in an increase in the soluble solid concentration of the potato sample. Pulsed electric field (PEF) pretreatment prior to frying had no significant effect on the measured moisture content, thermal conductivity or frying time compared to potatoes that did not receive a PEF pretreatment. However, a PEF pretreatment at 1.1 kV/cm and 56 kJ/kg reduced the temperature variation in the experimentally measured potato center by up to 30%. The proposed heat and moisture transfer model based on unsteady state heat conduction successfully predicted the experimental measurements, especially when the equation was solved using the enthalpy method.

**Keywords:** potato; frying; PEF; variable space network method; enthalpy method; approximate quasi-steady-state analysis; explicit finite difference

#### **1. Introduction**

Deep-oil frying is one of the most common processes used for food preparation, which is estimated as a billion-dollar industry worldwide [1]. Accordingly, monitoring and simulation of the temperature time distribution in food during frying is important to warrant the final quality of the fried food products [2]. Depending on the processing intensity, moisture in foods is removed partially or fully during frying. In fact, a phase change from liquid to vapor happens due to absorbing heat from hot oil during frying. Therefore, in the food frying process, heat and mass transfer phenomena take place simultaneously.

Deep fried potatoes are among the most popular of food products owing to their availability, convenience, price and taste [3]. Several researchers have developed mathematical models to describe the frying process of potatoes [4–6]. These models have been developed to describe heat and mass transfer mechanism for potato frying based on either single or two-phase systems. Based on the assumption that heat and mass transfer processes occur independently, an analytical correlation for heat and moisture transfer coefficients based on different potato geometries was developed by Dincer [7], who used a single-phase model by solving the diffusion equation for both the heat and mass transfer phenomena, without

**Citation:** Gholamibozanjani, G.; Leong, S.Y.; Oey, I.; Bremer, P.; Silcock, P.; Farid, M. Heat and Mass Transfer Modeling to Predict Temperature Distribution during Potato Frying after Pre-Treatment with Pulsed Electric Field. *Foods* **2021**, *10*, 1679. https://doi.org/10.3390/foods 10081679

Academic Editor: Javier Raso Received: 6 June 2021 Accepted: 13 July 2021 Published: 21 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

coupling the two phases together. Costa and Oliveira [8] developed a two-phase model (crust and core) with a time-dependent boundary condition to define the water loss in fried potato discs. In their model system, dynamics were described as a general first-order response to an exponential input, where the rate constants of both phases (crust and core) were highly dependent on the thickness of potato discs while the rate constant of the unfried phase (core) was impacted by the frying temperature. These researchers reported that 63% of the initial moisture content in a potato still resides in the unfried phase.

Farkas et al. [9] developed a two-phase domain model to describe the heat and mass transfers in food frying, including potato frying. In this model it was assumed that water boiling occurs at a moving interface similar to the process of melting and solidification [10]. To avoid complicated and time-intensive models, some researchers have assumed that the process can be described by heat transfer equations only. For example, Farid and Chen [4] employed the unsteady state heat conduction equation and used the finite difference method of Murray and Landis [11] to solve the equation. Their model eliminated the use of excessive parameters such as diffusion coefficient and mass transfer coefficient (in addition to heat transfer parameters) which are difficult to measure analytically. Their work successfully described the frying process but failed to predict the gradual increase of potato center temperature in the final period of frying. To address this issue, Southern et al. [12] defined the center of the potato disc as a region, rather than a point, based on the cross-section area of the thermocouple used. Then, an integration was carried out within that location. This assumption reduced the errors caused by the displacement of the thermocouple from the center of the potato disc. They also considered a variable surface heat transfer coefficient, which was high during boiling and lower during sensible heating. Lastly, they considered the oil penetration effect, which altered the properties of the fried potato discs and thus slowed down the increase in the center temperature of the potato during the final sensible heating period. However, with this approach, the agreement between predicted and measured temperature was poor during this final heating period.

The use of pulsed electric field (PEF) technology as a pre-treatment in the potato industry, especially for the production of deep-fried potato products (e.g., French fries and potato crisps), has shown promise as a means to: achieve better process performance as the texture of PEF-treated potatoes are more flexible, easier to cut and have a smoother cut surface; reduce energy and water consumption during production as the integration of PEF into existing process line can speed up the production time and reduce energy and water use compared to conventional processing; produce better quality end product as PEF facilitates uniform browning after frying an reduced oil uptake; and minimize food waste, as PEF-treated potatoes have less tissue damage and are less prone to breaking [13,14]. PEF treatment involves applying short (μs or ms in duration) and repetitive electric pulses of high voltage across the potato which is positioned between two conducting electrodes [15]. Electric field strength, pulse frequency, pulse number, pulse shape and polarity, pulse width and duration are important operating parameters of PEF processing [16]. When the electric field strength of the applied pulses exceeds thresholds electrical potential of the cell membrane, existing pores in the membrane can enlarge which affects its microstructure [17]. These changes will also impact heat, mass and momentum transfers, and affect the rate of chemical reactions, moisture content and oil uptake during the subsequent frying process [18]. However, heat and mass transfers during frying of PEF-treated potatoes has not yet been reported.

To accurately model the frying process of potatoes is not trivial, as it is postulated that the water boiling temperature will increase as the boiling interface moves towards the potato center owing to increases in the concentration of solid soluble matters in the potato discs as water is evaporated. To address this, an empirical correlation for water boiling temperature as a function of interface position of fried (crust) and unfried (core) regions was developed in the current study. The unsteady state heat conduction equation was then solved using the enthalpy method, for the first time for potatoes, in addition to the wellknown Variable Space Network (VSN) method previously used by other researchers [19]. The enthalpy method or the effective heat capacity method has previously been widely used to describe melting and solidification of materials [10,20] and unlike the VSN method, it has been applied to two- or three-dimensional geometry samples [21,22]. The enthalpy method was applied to help overcome the computation limitations of the VSN method when an explicit finite difference method is used in the solutions. However, a limitation of the enthalpy method is the assumption that water boils within a very narrow range of temperature. Both the enthalpy method and the VSN method were used in this study to predict temperature-time distribution at different locations of the potato disc. The complete frying time was also measured directly using the method developed by Michael and Farid [23] for flat geometries.

Thus, this study aims to test whether the enthalpy and VSN methods based on a moving interface defined by the boiling temperature of water in potato during frying can describe the temperature-time distribution of PEF-treated potatoes.

#### **2. Materials and Methods**

#### *2.1. Sample Preparation and PEF Treatment*

All potatoes used in this study were harvested together, graded, washed and stored in jute potato bags at 10 ◦C until use. Thirty individual potatoes were randomly allocated to either untreated or PEF treated. Potatoes were PEF-treated using the batch mode of an ELCRACK-HVP 5 (German Institute of Food Technologies, Quakenbrueck, Germany) unit by positioning each potato tuber in the middle of the batch treatment chamber (100 mm length × 80 mm width × 50 mm height, 400 mL capacity) consisting of two parallel stainless-steel electrodes of 5 mm thickness separated by a distance of 80 mm. To standardize the transmission of the electric currents, the whole potato tuber was fully submerged in sodium phosphate buffer (10 mM; pH 7) [24]. Potatoes were treated with PEF at an electric field strength of 1.1 kV/cm, pulse frequency of 50 Hz, and pulse width of 20 μs at varying pulse numbers of 1100 and 3100 that resulted in specific energy inputs at either 54.8–57.7 kJ/kg (thereafter referred as "PEF Low") or 149.2–159.6 kJ/kg ("PEF High"), respectively. During PEF treatment, the pulse shape (square wave bipolar) was monitored on-line using an oscilloscope (Model UT2025C, Uni-Trend Group Ltd., Dongguan, China). The output voltage, output current, pulse energy, pulse number and load resistance was monitored during treatment and the temperature and electrical conductivity of the conducting medium were measured before and after PEF treatment, using a conductivity/temperature meter (CyberScan CON 11, Eutech Instruments, Queenstown, Singapore).

#### *2.2. Moisture Content Measurement*

To determine their moisture content (*w*), samples were cut from the surface and center of each potato (either untreated or PEF-treated) and placed in a convection oven (Binder drying and heating chambers ED115, Tuttlingen, Germany) at 140 ◦C for 4 h. Potato mass was measured over time until a constant mass was achieved and the moisture content was calculated for triplicates independent samples.

#### *2.3. Thermal Conductivity Measurements*

Thermal conductivity was measured using a ThermTest TC-30 (Fredericton, NB, Canada), which works based on transient heat conduction. Fresh or fried potatoes were cut in rectangular shapes (50 mm length × 25 mm width × 5 mm height) and their thermal conductivity was measured as shown in Figure 1. Briefly, the sample was placed on the spring-loaded sensor (Figure 1A,B) and a weight placed on top of it to ensure proper contact (Figure 1C,D). A heat reflectance sensor was used so that the heating element was supported on an insulated backing and surrounded by a guard ring for one-dimensional heat flow. Heat was generated when a current was applied to the sensor coil. Simultaneously, the rate of temperature increase was monitored by the voltage drop in the sensor, which was calibrated to the temperature change. The thermal conductivity of the sample was inversely

proportional to the rate of increase in the temperature. In other words, samples with a lower thermal conductivity showed a steeper rise in temperature. Thermal conductivity was measured in triplicate using independent samples.

**Figure 1.** Stepwise potato sample preparation for the thermal conductivity measurement: (**A**) the empty ThermTest TC-30 machine; (**B**) placing sample in the machine; (**C**) a weight added to the sample; and (**D**) when the weight was on the sample.

#### *2.4. Potato Frying Experiment*

Both the PEF-treated and the untreated potatoes were peeled and cut in circular discs (40 mm diameter and 5 mm thickness). A K-type thermocouple (IEC K 0.5 × 150 mm) was inserted into the center of the potato disc to measure its temperature during frying. The potato disc was placed in a 4 L temperature-controlled deep fryer (Brabantia Deep Fryer BBEK1130, Valkenswaard, The Netherlands, 393 × 280 × 280 mm) as shown in Figure 2. The oil was held at 180 ◦C [25] using a temperature controller. Frying temperatures were recorded over time using PicoLog TC-08 data logger (Cambridgeshire, UK) (Figure 2). The potato preparation including skin removal, cutting and thermocouple insertion in the center of potato discs is shown in Figure 3.

**Figure 2.** Potato frying experiment setup.

**Figure 3.** Stepwise sample preparation for the potato frying experiment: (**A**) intact potato after PEF treatment; (**B**) removal of skin; (**C**) cutting sample in a circular disc form; (**D**,**E**) position of thermocouple inserted into the sample before frying; and (**F**) position of thermocouple in the sample upon completion of frying experiment.

#### *2.5. Measurement of Water Boiling Temperature of Potato*

The water content of a potato ranges between 70% to 80% [26] and during frying most of it is evaporated when the water temperature reaches "boiling" [27]. However, since the soluble matter in a potato such as sugar and potassium become more concentrated during frying, the water boiling temperature in the potato is expected to increase over time. This temperature elevation is expected to continue until a major proportion of free water is removed from potato and the center temperature reaches the targeted oil temperature (180 ◦C). To understand the elevation mechanism of water boiling temperature, potato discs were fried in the same fryer as outlined in Section 2.4. As soon as the center temperature of the potato disc started to rise above the water boiling temperature (i.e., >100 ◦C as monitored using the data logger), the disc was removed immediately from the hot oil. The water content remaining in each partially fried potato disc was determined by placing them in an oven at 140 ◦C to dry, until a constant weight was achieved, which amount was calculated based on the known initial moisture content of potato (Equations (1)–(3)). Equation (3) represents the moisture content of potato in which water boiling point started to deviate from 100 ◦C.

Initial water mass (g) = Initial potato mass before drying (g) − Potato mass after drying in an oven (g) (1)

Water remained after partial frying (g) = Potato mass when frying stopped (g) − Potato mass after drying in an oven (g) (2)

$$\% \text{ Water remaining after partial frying} = \frac{\text{Water remaining after partial frying (g)}}{\text{Initial water mass (g)}} \times 100\% \tag{3}$$

An additional experiment was conducted to estimate the boiling temperature elevation of the water in potatoes. To achieve this, liquid was extracted from almost 2 kg of potatoes, using a juicer (Nutribullet Juicer 800W, Los Angeles, CA, USA), heated and its boiling temperature over time was recorded using a data logger. The change in the mass of the liquid was measured at regular intervals, using an electronic kitchen scale (Wiltshire, NSW, Australia). A correlation was developed to estimate the boiling temperature elevation as a function of potato mass. It should be noted that the data obtained was used to approximate the correlation between the water loss during frying and the boiling temperature elevation since potato liquid does not have the same physicochemical properties as a potato. Furthermore, it was difficult to conduct experiments above 120 ◦C as

the liquid solution became very viscous and starch gelatinization occurred which hindered water evaporation.

#### **3. Mathematical Modeling**

An approximate quasi-steady state was used to estimate the complete frying time of potato and the frying process modeled by VSN and enthalpy methods using MATLAB software (version R2020b, The MathWorks Inc., Natick, MA, USA).

#### *3.1. VSN Method*

In the VSN method, it is assumed that there are two regions of the potato discs separated by an interface namely the "core" and "crust" regions [4]. The physical properties of potato for the two regions of core and crust used in this study are summarized in Table 1. The potato disc sample was assumed to be an infinite slab comprised of a porous solid structure filled with water. The effect of shrinkage and water diffusion in the core region were assumed to be minimal as they were fully saturated with water [4].

**Table 1.** Physical properties of potato discs.


To determine the temperature of the potato discs over time, the heat conduction equation (Equation (4)) in the form of sensible heat was first solved for the whole potato, explicitly using a finite difference method:

$$\frac{\partial T}{\partial t} = \kappa \frac{\partial^2 T}{\partial y^2} \tag{4}$$

where *T* and *t* are temperature and time, respectively; *α* is thermal diffusivity of potato and *y* refers to the position from the surface of potato disc.

As soon as the surface temperature of the potato disc reached its boiling point, a thin layer was assumed to form a crust region needed to start the computation with the presence of two regions as required by the VSN method [4]. Following this, Equation (4) was then solved for both crust and core regions using the following initial and boundary conditions (Equations (5)–(9)):

$$T = 0 \qquad \qquad T = T\_{\bar{l}} \qquad \qquad \text{where } T\_{\bar{l}} = 20 \, ^\circ \text{C} \tag{5}$$

$$y = 0 \quad h \begin{pmatrix} T\_0 \ - \ T\_s \end{pmatrix} = -k\_{cr} \frac{\partial T\_s}{\partial y} \tag{6}$$

$$y = \frac{\delta}{2} \tag{7} \tag{8}$$

$$y = \text{Y}\_{\text{ll}} \qquad \qquad \qquad T\_{\text{co}} = T\_{\text{cr}} = 103 \, ^{\circ}\text{C} \tag{8}$$

$$y = \mathbf{Y}\_n \qquad \qquad \rho\_{co} \, \lambda\_w \, w \, \frac{d\mathbf{Y}\_n}{dt} = k\_{cr} \, \frac{\partial T\_{cr}}{\partial y} - \, k\_{co} \, \frac{\partial T\_{co}}{\partial y} \tag{9}$$

where subscripts "*s*", "*cr*" and "*co*" represent the surface, crust, and core of potato disc, respectively. Subscript "*i*" refers to initial condition of potato and "*o*" refers to oil. *δ*

denotes the full thickness of potato disc. *w* and *λ<sup>w</sup>* define the moisture content, and latent heat of boiling of water, respectively. *Yn* represents the interface position at each timestep, and *h* and *k* are convective heat transfer coefficient and thermal conductivity of the potato, respectively.

The two regions were then divided into equal space increments. To start the computation, a linear temperature distribution was assumed within the initial crust layer. The magnitude of error raised by this simplification was negligible if the assumed crust thickness was small. An extremely small initial crust layer should be avoided; otherwise, computation time would increase substantially to satisfy the stability requirement of the explicit finite difference method. An initial crust layer of between 1% and 2.5% of the potato disc thickness was deemed to be sufficient [4].

As frying proceeds, the crust thickness will increase and hence the space increment width will expand, while the space increment width of core region will decrease. Using the explicit finite difference formulation described by Murray and Landis [11], the nodal equations for both regions are formulated and used to calculate the dimensionless temperature. Equations (10) and (11) represent the nodal discretization of crust and core regions, respectively.

$$
\theta\_{j,n+1} = \theta\_{j,n} + j \left(\theta\_{j+1,n} - \theta\_{j-1,n}\right) \times \frac{(Y\_{n+1} - Y\_n)}{2Y\_n} + \mathfrak{a}\_{\text{cr}} \Delta t \big/ \left(\theta\_{j-1,n} - 2\theta\_{j,n} + \theta\_{j+1,n}\right) Y\_n^2 \tag{10}
$$

$$\theta\_{j,n+1} = \theta\_{j,n} + (2\text{J} - j)(\theta\_{j+1,n} - \theta\_{j-1,n}) \times \frac{\text{Y}\_{n+1} - \text{Y}\_n}{\text{S} - \text{Y}\_n} + a\_{\text{cyl}} \Delta t \big|^2 (\theta\_{j-1,n} - 2\theta\_{j,n} + \theta\_{j+1,n})(\delta - \text{Y}\_n)^2 \tag{11}$$

where *θ* is the difference between temperature and the water boiling temperature (*T* − *Tb)* and subscripts "*j*" and "*n*" are space and time increments, respectively. *J* represents the number of spatial increments in the crust and core regions. The interface position may be calculated from Equation (12).

$$Y\_{n+1} = Y\_n + \left\{\frac{\Delta t}{\rho\_{co} \ w \,\lambda\_w}\right\} \times \left[ \left(\frac{0.5 \, k\_{cr} \, l}{Y\_n}\right) \left(\theta\_{l-1,n} - 4\theta\_{l,n}\right) + \frac{0.5 \, k\_{co} \, l \left(\theta\_{l+3,n} - 4\theta\_{l+2,n}\right)}{\delta - Y\_n}\right] \tag{12}$$

During the two-region period (when both core and crust are present), heat is assumed to transfer from the hot oil to the potato disc surface through a convection boiling process with a coefficient higher than the convection coefficient used for a single region of core or crust, in this case 500 versus 250 W/m2·K (Table 1) [4]. As the interface position approaches the center of the potato disc (i.e., a distance equivalent to 1% to 2.5% of potato disc thickness [4]), the two region computation changes to a single region computation as only the crust is being considered. The sensible heating of the crust region was calculated using Equation (4).

#### *3.2. Enthalpy Method*

In the enthalpy method, water evaporation is assumed to occur over a narrow arbitrary region of boiling (*Tb* ± *ε*) ◦C, where *ε* is a small value (in the order of 0.5 ◦C) representing half the phase change temperature [28]. Since a potato disc was assumed to be an infinite slab [4], the assumption of one-dimensional heat transfer became valid. This assumption meant that Equation (13) could be solved using initial and boundary conditions defined by Equations (5)–(7), using an explicit finite difference method (detailed explanation referred to the study of Gholamibozanjani and Farid [29] for modeling melting and solidification). To guarantee stability of the finite difference method, the condition set by Equation (15) need to be met.

$$
\rho \frac{\partial H}{\partial t} = k \frac{\partial^2 T}{\partial y^2} \tag{13}
$$

where *H* is enthalpy as described by Equation (14).

$$H = \begin{cases} \mathbb{C}\_{\mathcal{O}}T\_{\tau} & T < T\_{b-\varepsilon} \\ \left(T - (T\_{b} - \varepsilon)\right) \times \begin{pmatrix} \frac{\mathbb{C}\_{\mathcal{O}} + \mathbb{C}\_{\mathcal{O}}}{2} + \frac{w \ \lambda\_{w}}{2} \end{pmatrix} + \mathbb{C}\_{\mathcal{O}} \times \left(T\_{b} - \varepsilon\right), & T\_{b} - \varepsilon < T < T\_{b} + \varepsilon \\ \mathbb{C}\_{\mathcal{O}} \ T + w \ \lambda\_{w} + (\mathbb{C}\_{\mathcal{O}} - \mathbb{C}\_{\mathcal{O}}) \times T\_{b}, & T > T\_{b} + \varepsilon \end{cases} \tag{14}$$

$$
\alpha \frac{\Delta t}{\Delta y^2} \le \frac{1}{2} \tag{15}
$$

where subscript "*b*" denotes the water boiling temperature.

In this single-phase method, computation is conducted for three periods: First the sensible heating of the potato until its core temperature reaches the lower bound of the evaporating point of water, (*Tb* − *ε*) ◦C. During this sensible heating period, the specific heat capacity and other physical properties of the potato are assumed to be constant in the core region (Table 1). Once the potato temperature reaches (*Tb* − *ε*) ◦C, it is assumed that a two-phase region (liquid and vapor) is formed, and evaporation starts. During this water evaporating period, between (*Tb* − *ε*) ◦C and (*Tb* + *ε*) ◦C, enthalpy rises dramatically. Finally, the sensible heating of the potato crust starts when the potato temperature reaches (*Tb* + *ε*) ◦C.

#### *3.3. Approximate Quasi-Steady State*

Farid et al. [23] developed an analytical method to calculate the approximate time required for complete frying as defined by Equations (16) and (17):

$$\beta = \frac{\left(T\_{\text{o}} - T\_{\text{b}}\right)}{\rho \,\,\lambda\_{\text{w}} \,\, \varepsilon} \tag{16}$$

$$t = \frac{1}{\beta} \left( \frac{\delta}{2 \ h} + \frac{\delta^2}{8 \ k\_{cr}} \right) \tag{17}$$

Equation (17) is derived by ignoring sensible heat, based on the assumption that the temperature reaches the boiling point of water very fast, compared to the time required to complete frying. This sensible heat is considered to be negligible compared to the high latent heat of water vaporization since it accounts for less than 2% of heat absorbed during frying [30].

#### **4. Results and Discussion**

#### *4.1. Experimental Measurements*

4.1.1. Effect of PEF Pre-Treatment on the Moisture Content, Thermal Conductivity and the Temperature-Time Profile of Potatoes during Frying

The average moisture content of the potatoes (considering both the surface and center) in the absence of PEF treatment was 73.5 wt %. As PEF Low- and PEF High treated potatoes a moisture content of 74.4 wt % and 73.3 wt %, respectively, this data shows that that PEF treatment had little effect on the moisture content of the potatoes (Table 2). The results also showed negligible difference for moisture content within the potato from surface to center based on the measurement technique used in this current study.

**Table 2.** Average, minimum and maximum moisture content (wt %) of the surface and center of PEF-treated and untreated potatoes before frying.


\* Data presented as average ± standard deviation from 5 independent potato tubers (*<sup>n</sup>* = 5). † Overall average considering the potato surface and potato center for each pre-treatment applied to the potatoes.

The thermal conductivity of both untreated PEF treated potatoes was estimated before and after frying and the values obtained were used in Equations (4), (10)–(13), (15) and (17) and the boundary conditions were calculated from Equations (6) and (9). The predicted thermal conductivities of untreated and PEF-treated potatoes are shown in Table 3. It was found that thermal conductivity of fried discs from untreated and PEF-treated potato tubers were in the range between 0.10 and 0.11 W/m·K. The average value for the thermal conductivity of fresh potatoes (without frying) was 0.73 W/m·K, with a variation from 0.68 to 0.78 W/m·K. Hence, PEF treatment was concluded not to have a clear impact on thermal conductivity of potato (Table 3). The water content in food is known to be a key factor influencing its thermophysical properties [31,32]. Therefore, negligible difference in the thermal conductivity for untreated and PEF-treated potatoes is expected as they had a similar moisture content (Table 2). Note that the average values of thermal conductivity (0.73 W/m·K for fresh potato and 0.11 W/m·K for fried potato) were used as the basis of modeling in Section 4.2.

**Table 3.** Average, minimum and maximum thermal conductivity (W/m·K) of fresh and fried potatoes.


\* Data presented as average ± standard deviation from 5 independent potato tubers (*n* = 5).

The temperature-time profiles of untreated and PEF-treated potatoes during frying were obtained when the 5 mm-thick potato disc samples were fried in an oil bath at 180 ◦C (as described in Section 2.4). The measured center temperature for both untreated and PEF treated potato discs are depicted in Figure 4, where different experimental measurements are shown in gray dashed lines and the average of all measurements is shown with a black solid line. The average center temperature of the potato disc for the untreated and PEF-treated potatoes shared very similar temperature-time profiles.

As frying proceeded, the center temperature of the potato experienced an initial sharp ascending (sensible heating) period, followed by a leveling off at around 103 ◦C (Figure 4, see red arrows) (evaporation), and a final moderate ascending period (sensible). In the initial period, the water temperature of the potato increased sensibly to its boiling point and then stayed constant at around 103 ◦C for an extended period. All experimental values measured in the current study as well as those reported by other researchers consistently show that the initial boiling temperature of water in potatoes is approximately 103 ◦C [2,4,33,34]. The elevation of the boiling point is believed to occur as a result of nucleation of steam bubbles in the superheated interface separating crust and core [34]. Therefore, the initial boiling temperature of water (*Tb*), in both untreated and PEF-treated potatoes, was set to 103 ◦C, throughout the following modeling section (in Equations (14), (16) and (21)).

**Figure 4.** Experimental measurements of center temperature of potato discs fried at 180 ◦C for (**A**) untreated, or PEF-treated at (**B**) PEF Low (1.1 kV/cm, 54.8–57.7 kJ/kg) or (**C**) PEF High (1.1 kV/cm, 149.2–159.6 kJ/kg), where water in the potato discs typically boiled at 103 ◦C (red arrows) and starts rising after around 550 s (blue arrows). Gray dash lines in (**A**–**C**) represent experimental measurements from 10 independent potatoes. Black solid lines in (**A**–**C**) show the average measurements (*n* = 10 independent potatoes).

As the interface moves close to the center of the potato, the water inside the potato was concentrated with soluble matters such as sugar and potassium, leading to an elevation in water boiling temperature [35,36]. In this study, the elevation of water boiling temperature was observed to occur at around frying time of 550 s (Figure 4, see blue arrows) and then the average center temperature of the potato discs rose steadily after about 550 s.

In the current study, a huge variation in the temperature-time profiles between individual potato discs occurred, probably due to the biological difference between the (untreated) potatoes. Figure 5 illustrates the extent of variation in the center temperature as a function of frying time for untreated and PEF-treated potatoes. During the first 500 s of frying, all the potato disc samples, regardless of pre-treatment, experienced similar temperature-time profiles from the start of the frying until the center temperature of the disc reached the water boiling temperature. This was illustrated by the low standard deviations obtained for the data over the first 500 s of frying where the center temperatures of different potato discs were situated very close to the average value (Figure 5). As the frying lapsed after 550 s, and the center temperature of the disc rose above the water boiling temperature until it reached the oil temperature, a considerable variation in the experimental measurements

between individual potatoes was observed. At this stage, the standard deviations of the experimental temperature measurements between individual potatoes were high, where the center temperature values spread out over a wider range, of 26, 18 and 26 ◦C above the boiling temperature for untreated, PEF Low treated and PEF High treated potatoes, respectively (Figure 5). It is important to note that the standard deviation of the measured center temperature for different potato discs showed that the PEF Low treatment has reduced the experimental variation, by up to 30% compared to untreated potatoes and those pre-treated with PEF High. Potatoes pre-treated with PEF Low had a lower variation in their temperature-time profiles than previously reported in the literature. This result could be because PEF causes loss of cell compartmentalization and allows water to be redistributed across various cell compartments in plant tissue [37], including extracellular spaces [38]. Changes in the water distribution within potato cells would likely influence the water evaporation process from potato discs during frying. In contrast, the high standard deviation in the center temperature measurements of potato discs pre-treated under PEF High may have occurred because the high intensity treatment had severely disrupted the potato cell, ultimately leading to the collapsed of the cell walls [24,39,40].

**Figure 5.** Averaged standard deviation of the measured center temperature for untreated and PEF treated potato discs (*n* = 10 independent potatoes for each treatment).

4.1.2. Increase in the Boiling Temperature of Water in Potato during Frying

The potato discs were removed from the fryer (at 180 ◦C) when the center temperature exceeded 103 ◦C and the percentage of water remaining in them was estimated (Table 4). Approximately 72% to 86 % of the water in potato was removed during the initial frying period.

**Table 4.** Average, minimum and maximum percentage of water remained in partially fried potatoes when the center temperature exceeded 103 ◦C.


\* Data presented as average ± standard deviation from 5 independent potato tubers (n = 5). <sup>a</sup> Calculated based on Equation (1). <sup>b</sup> Calculated based on Equation (2). <sup>c</sup> Calculated based on Equation (3). <sup>d</sup> Percentage of water removed after partial frying = 100% − percentage of water remaining after partial frying.

Following the potato liquid boiling experiment described in Section 2.5, an exponential trend between water boiling temperature elevation and the water loss percentage was found as water was removed due to evaporation (Figure 6). The obtained curve was shifted along the temperature elevation axis by 3 ◦C to compensate for the nucleation of steam bubbles in the superheated interface separating crust and core which occurs during frying (but not in the potato liquid). It was then extrapolated to achieve a 100 wt % water loss, as it was difficult to obtain from the experiment explained in Section 2.5.

**Figure 6.** Elevation in water boiling temperature achieved from boiling the untreated potato liquid ( ) and 3 ◦C temperature correction expected (-) due to the effect of steam bubble nucleation in the superheated interface separating crust and core.

Based on the results reported in Table 4, the experimental measurements of the potato liquid boiling temperature (measured data in Figure 6) and data extrapolation to 100 wt % water losses, Equation (18) was developed. It was assumed that the location of the boiling interface determined the water remaining in the potato disc during frying and hence its boiling point.

$$
\Delta T\_b = 7 \times 10^{-10} \exp(25.455 \times \frac{Y\_m}{\delta/2}) \tag{18}
$$

It has to be noted that Equation (18) is an oversimplification of the frying process of potato discs. During frying, not only water evaporates that causes an increase in the soluble sugar and salt concentration, but the potatoes also experience reactions such as starch gelatinization and protein denaturation [41] which have not been considered in this study. Then, the fixed water boiling temperature (*Tb*) in Equations (10)–(12) and Equations (13) and (14) were replaced with the variable boiling temperature of Equation (18) to be solved via the VSN and enthalpy methods, respectively.

#### *4.2. Mathematical Modeling Results*

In the current study, the frying process of untreated and PEF-treated potatoes was modeled following three methods (VSN, enthalpy and approximate quasi-steady state) as described in Section 3. These models were applied for two conditions, namely: (i) at the constant boiling temperature of water in potato at 103 ◦C (fixed conditions, based on the results in Figure 4) and (ii) when the boiling temperature of water in the potato increases (variable conditions) following Equation (18). Figure 7 shows the application of the VSN method using a fixed and variable water boiling temperature to predict the measured center temperature of untreated and PEF-treated potato discs. In the VSN model, as the interface position moves towards center, the space increment width in the core side becomes very small, requiring a very small time-step that otherwise would create numerical instability. To overcome this computational problem, when the space increment reached a distance about 2% of the potato thickness from the center, the computation was carried out based on a single phase of crust region. As a result, a jump in the center temperature of the potato

disc was observed as shown in Figure 7 (indicated by red and blue arrows when fixed and variable boiling temperatures were used, respectively).

**Figure 7.** Model prediction (colored lines) of measured center temperature for untreated, PEF Low and PEF High treated potato discs (5 mm thickness) fried at 180 ◦C, using VSN method, considering at fixed (those indicated with red arrows) and variable (those indicated with blue arrows) boiling temperatures of the water in the discs.

The modeling results were validated against the experimental measurements of untreated, PEF Low and PEF High-treated potatoes considering the average values of moisture content and thermal conductivity reported in Section 4.1.1 (Tables 2 and 3 respectively).

The measured center temperature of the potato disc was also predicted using the enthalpy method described in Section 3.2. In this method, each location went through three regions namely core, mushy (mixture of core and crust) and crust. Thermal diffusivity for core and crust regions was calculated using Equation (19):

$$\begin{aligned} \alpha\_{cr} &= \frac{k\_{cr}}{\rho\_{cr} \cdot \mathbb{C}\_{cr}}\\ \alpha\_{co} &= \frac{k\_{co}}{\rho\_{co} \cdot \mathbb{C}\_{co}} \end{aligned} \tag{19}$$

In the mushy phase, the density of the crust and core was defined according to Equation (20) [29]. In fact, as frying proceeds, oil penetrates into the potato disc making it difficult to measure its density and hence its thermal diffusivity.

$$
\rho\_{\text{mushy}} = (1 - X\_{n,m})\rho\_{\text{co}} + X\_{n,m}\rho\_{\text{cr}} \tag{20}
$$

where *X* denotes the crust fraction of potato and is calculated from Equation (21).

$$X\_{\rm n,m} = \frac{T\_{\rm n,m} - (T\_b - \varepsilon)}{2\text{ }\varepsilon} \tag{21}$$

where subscript "*mushy*" refers to the phase where a mixture of core and crust exists.

Figure 8 illustrates the model prediction based on the experimental measurements of 5 mm-thick untreated, PEF Low and PEF High treated potato discs fried at 180 ◦C, using the enthalpy method under the conditions of fixed and variable water boiling temperature. Unlike previously developed models [2,4], the current model predicted the experimental temperature-time measurements of potatoes reasonably well, especially at the final stage of frying. Grid and time-step independence analysis was also assured for the enthalpy method following the approach explained in a study conducted by Gholamibozanjani and Farid [29].

**Figure 8.** Model prediction (colored lines) of measured center temperature for untreated, PEF Low and PEF High treated potato discs (5 mm thickness) fried at 180 ◦C, using enthalpy method and the density of the mushy region (Equation (17)), considering fixed and variable boiling temperatures of the water in the discs.

By increasing the space increments and applying smaller mesh sizes during computation, a smoother graph was achieved as shown in Figure 9. In general, increasing the space increments does not contradict the grid independency analysis [29], as the latter verifies the stability of the numerical solution, whereas increasing space increments makes the profile smoother.

Moisture content is an important factor in determining the temperature-time profile of a potato disc. Figure 10 shows the model prediction of the temperature of potatoes based on the minimum, average and maximum measured moisture contents, at both the surface and center of the potatoes, of 69 wt %, 73.8 wt % and 80 wt % for the untreated, PEF Low and PEF High treated potatoes, respectively (Table 2). Along with the effect of other phenomena occurring during the frying process of potato such as gelatinization and PEF pretreatment, the difference in the potato moisture content explains the variation in experimental measurements as shown in Figure 10.

**Figure 9.** Effect of more fine mesh size during the computation of enthalpy method on the smoothness of graph using untreated temperature-time profile as an example.

**Figure 10.** Model prediction (colored lines) of center temperature for untreated, PEF Low and PEF High treated potato discs (5 mm thickness) according to their minimum, average and maximum measured moisture contents (MC) (see Table 2), when fried at 180 ◦C.

The advantage of the enthalpy method compared to the VSN method is that the former does not face any computational problems and it can be applied to multi-dimensional geometry samples, which is not the case with the VSN method. The approximate quasisteady state method (Equation (17)) of Smith and Farid [23] which considers core density and crust thermal conductivity as the basis of calculations, shows that the complete frying

of a 5 mm-thick potato disc happens after 809 s, which is compatible with those obtained from VSN and enthalpy methods based on keeping the water boiling temperature constant.

#### **5. Conclusions**

The developed model, which integrated a variable water boiling temperature into the conduction heat transfer equation and was solved by the enthalpy method, predicted the experimental measurements during potato frying reasonably well. The VSN method suffers from numerical instability when the explicit finite difference is used in its solution. This limitation was eliminated via the use of enthalpy method. The enthalpy and the Variable Space Network (VSN) methods based on a moving interface defined by the boiling temperature of water in potato during frying could adequately describe the temperature time distribution of untreated and PEF-treated potatoes even though PEF treatment did not dramatically change the initial moisture content, thermal conductivity or the total frying time for the potatoes to reach the targeted oil temperature. It was pleasing to note that PEF treatment reduced the variations in the experimental measurements of the potato center temperature time profile by up to 30%, when a low PEF treatment intensity (1.1 kV/cm and 54.8–57.7 kJ/kg) was used. This finding implies that it will narrow the confidence interval between the predicted and measured temperature time distribution. As a consequence, the process uniformity of PEF processing can be more controlled and the changes in quality attributes of potato discs during frying, such as brown color formation and crispiness, can be predicted more accurately.

**Author Contributions:** Conceptualization, I.O., M.F., P.B. and P.S.; methodology, M.F. and I.O.; software, M.F.; validation, M.F.; formal analysis, G.G.; investigation, G.G. and S.Y.L.; resources, M.F. and I.O.; data curation, G.G.; writing—original draft preparation, G.G.; writing—review and editing, S.Y.L., I.O., P.B. and M.F.; visualization, G.G.; supervision, M.F. and I.O.; project administration, M.F. and I.O.; funding acquisition, I.O., M.F., P.B. and P.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the New Zealand Ministry of Business, Innovation and Employment (contract MAUX1402) under the Food Industry Enabling Technologies (FIET) program.

**Acknowledgments:** Leong and Oey are affiliated to the Riddet Institute, a New Zealand Center of Research Excellence, funded by the Tertiary Education Commission.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **Abbreviations**

Abbreviations



#### **References**


### *Article* **Kinetics of Colour Development during Frying of Potato Pre-Treated with Pulsed Electric Fields and Blanching: Effect of Cultivar**

**Setya Budi Muhammad Abduh 1,2,3,†, Sze Ying Leong 1,3,†, Chun Zhao 1, Samantha Baldwin 4, David J. Burritt 5, Dominic Agyei <sup>1</sup> and Indrawati Oey 1,3,\***


**Abstract:** The current research aimed to investigate the effect of pulsed electric fields (1 kV/cm; 50 and 150 kJ/kg) followed by blanching (3 min., 100 ◦C) on the colour development of potato slices during frying on a kinetic basis. Four potato cultivars 'Crop77', 'Moonlight', 'Nadine', and 'Russet Burbank' with different content of glucose and amino acids were used. Lightness (*L*\* values from colorimeter measurement) was used as a parameter to assess the colour development during frying. The implementation of PEF and blanching as sequential pre-treatment prior to frying for all potato cultivars was found effective in improving their lightness in the fried products. PEF pre-treatment did not change the kinetics of *L*\* reduction during frying (between 150 and 190 ◦C) which followed first-order reaction kinetics. The estimated reaction rate constant (*k*) and activation energy (*Ea* based on Arrhenius equation) for non-PEF and PEF-treated samples were cultivar dependent. The estimated *Ea* values during the frying of PEF-treated 'Russet Burbank' and 'Crop77' were significantly (*p* < 0.05) lower (up to 30%) than their non-PEF counterparts, indicating that the change in *k* value of *L*\* became less temperature dependence during frying. This kinetic study is valuable to aid the optimisation of frying condition in deep-fried potato industries when PEF technology is implemented.

**Keywords:** kinetic; pulsed electric field; blanching; potato; cultivar; frying; colour; lightness; first order; activation energy; Arrhenius

#### **1. Introduction**

Pulsed electric field (PEF) processing applies a high-voltage electric field in the form of short pulses (in the range of microseconds to milliseconds) across biological cells or foods to develop high transmembrane potential of the cells, leading to pore formation in the cell membrane and hence changes in the cell microstructure [1,2]. In the deep-fried potato industry, PEF processing is conducted as an intermediate step before cutting/slicing, blanching, pre-drying, and frying [3]. Implementation of PEF prior to cutting and blanching has been shown to be effective to soften potato tuber tissues and improve their processing textural quality leading to smoother cutting surface (i.e., reduced feathering) [4,5], reduction in starch loss [4], reduction in oil uptake [4,5] and a more uniform colour development during frying [5], as well as a reduction in acrylamide formation of the final deep-fried potato products [6]. Previous studies have reported the effectiveness of PEF in improving the quality of the final fried potato product. However, these studies conducted PEF

**Citation:** Abduh, S.B.M.; Leong, S.Y.; Zhao, C.; Baldwin, S.; Burritt, D.J.; Agyei, D.; Oey, I. Kinetics of Colour Development during Frying of Potato Pre-Treated with Pulsed Electric Fields and Blanching: Effect of Cultivar. *Foods* **2021**, *10*, 2307. https:// doi.org/10.3390/foods10102307

Academic Editor: Javier Raso Received: 30 August 2021 Accepted: 22 September 2021 Published: 28 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

treatments on a single potato cultivar available at the time of the experiment (e.g., 'Agria', 'Rooster', 'Fontane', and 'Lady Claire') [4–6]. This raises a question as to whether the impact of PEF treatment, when applied at the same process intensity, is consistent across potato cultivars with different physicochemical characteristics in terms of carbohydrate (such as sugar and starch), protein, dry matter, and water contents [7]. So far, no previous study has included the cultivar effect in PEF experimental designs conducted under similar process intensities to assess its impacts on colour development when PEF-treated potatoes are fried. Such study is indispensable for commercial deep-fried potato production since switching or blending potato cultivars with different physicochemical characteristics before entering the cutter/slicer is rather a common practice [8].

Blanching is one of the important unit operations that cannot be omitted during the production of deep-fried potato products because this process results in starch gelatinisation as well as microstructural changes in the potato tissues by weakening the binding between cells (i.e., middle lamella), leading to effective removal of glucose in the potato and hence minimising the development of undesirable dark brown colour in fried potato [9,10] that may lead to acrylamide formation [11]. This thermal process is also necessary to inactivate endogenous polyphenol oxidase enzyme responsible for the enzymatic browning in potatoes [12] and to precondition the tissue structure, enabling texture developments such as hardness and crispiness during frying [13]. Therefore, the adoption of PEF in a commercial potato processing line does not exclude blanching from the entire processing line, except for potato chips/crisps processing line that is using kettle or batch frying technique. Currently, kinetic information on colour development during frying is very limited especially for potatoes that have been subjected to PEF treatment followed by blanching.

The aim of this research was to study the kinetics of colour development during frying of potato slices from different potato cultivars after a sequential pre-treatment of PEF and blanching. In this study, four potato cultivars 'Crop77', 'Moonlight', 'Nadine', and 'Russet Burbank' were selected because they had considerable difference in inherent physicochemical properties and cooking quality or purpose (see Section 2.1). The kinetic rate constant (*k* value) and activation energy (*Ea* value) of colour changes were predicted on the basis of first order reaction to assess the effect of cultivars and PEF pre-treatment applied to the potatoes prior to frying. This thorough kinetic study is a prerequisite to understanding the colour development of potato slices as a function of frying temperature and time. This approach is crucial to aid optimisation of frying conditions after conducting sequential PEF and blanching pre-treatment.

#### **2. Materials and Methods**

#### *2.1. Raw Material and Characterisation of Their Chemical Composition*

Four potato (*Solanum tuberosum* L.) cultivars, namely 'Crop77', 'Moonlight', 'Nadine', and 'Russet Burbank' (50 kg of tubers for each cultivar) were selected in this study. 'Nadine' is a common and readily available cultivar with waxy texture suitable for boiling, salads and braises, mainly sold as a fresh market washed potato in New Zealand [14]. 'Moonlight' is known as an all-rounder or general-purpose potato cultivar, in which the tubers are neither too floury not waxy, mainly grown for the local French fries industry as well as sold as fresh market potatoes [15]. 'Russet Burbank' is mainly used for the French fries industry [16]. 'Crop77' is a breeding line from the Plant and Food Research breeding programme specifically for potato crisping industry [17]. After harvest, the tubers from each cultivar were packed in 10 kg jute sacks and placed in a well-ventilated and temperaturecontrolled dark room at 8.8 ± 0.5 ◦C until use. Prior to processing, the content of total starch, amylose, glucose, total amino acids, and dry matter of the tubers was determined (see Supplementary Materials File S1).

#### *2.2. Pulsed Electric Fields (PEF) Treatment*

For each PEF experiment, potato tubers of similar size from the same cultivar were removed from the temperature-controlled room and acclimatised at room temperature (20 ◦C) overnight, before washing under running tap water to remove adhering soil. The skin of tubers was then peeled manually using a stainless-steel hand peeler. To reduce biological variation between and within tubers for experiment, the tubers were cut transversely into equal halves, separating the stem end from the basal end. Half of the amount of stem ends and basal ends were then allocated for control (later referred to as "Non-PEF") and the remaining amount for PEF treatment (later referred to as "PEF-treated").

The halved tuber was placed in a 400 mL PEF treatment batch chamber (100 mm length, 50 mm depth) consisting of two parallel stainless-steel electrodes (5 mm thickness, 80 mm electrode gap) with the cut surface facing the bottom of the PEF treatment chamber. Subsequently, the chamber was filled with sodium phosphate buffer (10 mM, pH 7.0, electrical conductivity of 1.40 ± 0.01 mS/cm at 13.77 ± 8.72 ◦C, prepared from appropriate mixture of monosodium dihydrogen orthophosphate and disodium hydrogen phosphate from Merck (Darmstadt, Germany)) as an electrical transferring medium, until the potato sample was fully immersed. The total weight of the potato sample and the medium was standardised for every PEF treatment (averaged at 316.97 ± 13.19 g).

Subsequently, potato samples were treated using an ELCRACK® HVP 5 PEF system (German Institute of Food Technologies, Quakenbruck, Germany) at a fixed electric field strength of 1 kV/cm, 20 μs pulse width (square-wave bipolar pulses), 50 Hz pulse frequency at two different levels of specific energy inputs, i.e., approximately 50 kJ/kg (hereafter referred as "PEF Low") and 150 kJ/kg (hereafter referred as "PEF High"). Equation (1) was used to calculate the specific energy input applied during PEF treatment. Considering the dimension of potato used in this study (i.e., halved tuber per PEF treatment to mimic industrial handling of potatoes), these two energy inputs were selected to achieve moderate to severe changes in textural properties based on our preliminary studies. This was consistent with other root vegetable studies on whole/halved potato [18], sweet potato [19] and carrot [20]. While previous studies have employed a much lower energy (<10 kJ/kg) on potato in the form of cubes (10–20 mm3) [5,21], cylinders (20–40 mm diameter, 4–10 mm height) [22–24], discs (9 mm thickness) [25] and slices (2.5 mm thickness) [26] to achieve effective cell disintegration, a higher energy input level is needed when handling whole or halved potato tubers.

$$\text{Specific energy input (kJ/kg)} = \frac{V^2 \cdot (\text{ $n \cdot$ } \text{ $m$ })}{R \cdot W} \tag{1}$$

where *V* is pulse voltage (kV), *n* is pulse number (dimensionless), *m* is pulse width (μs), *R* is pulse resistance (ohm), *W* is the total weight of potato sample and electrical transferring medium in the PEF chamber.

The ELCRACK® HVP 5 PEF system was equipped with built-in measurement sensors that allowed electric field strength (kV/cm), pulse voltage (kV), pulse current (A), pulse power (kW), pulse energy (J), total energy (kJ), pulse resistance (ohm) and pulse number to be monitored in real time during each PEF treatment. The pulse shape and voltage were monitored using a digital oscilloscope (UNI-Trend UTD2042C, Guangzhou, China). The temperature and electrical conductivity of the medium were measured before and after PEF treatment using a temperature/conductivity meter (CyberScan CON 11, Eutech Instruments, Singapore). The average temperature and conductivity increase after "PEF Low" were 6.07 ± 0.94 ◦C and 2%, respectively, while those after "PEF High" treatment were 15.09 ± 1.48 ◦C and 5%, respectively. For each potato cultivar, three independent PEF experiments were performed for "PEF Low" and "PEF High" treatments with each experiment using 8 to 10 potato tubers, where half the amount of stem and basal ends were used for PEF treatment and the remaining half as untreated samples. The total contact time between potato samples and sodium phosphate buffer during PEF treatment (starting from sample immersion in buffer solution in the PEF chamber to the completion of PEF treatment) was averaged at 2 min. For this reason, all the untreated potatoes ("Non-PEF") were immersed in the sodium phosphate buffer for at least 2 min.

#### *2.3. Kinetic Study on the Colour Changes of Non-PEF and PEF-Treated Potato Slices during Frying* 2.3.1. Kinetic Frying Experiment

The potato frying experiment was conducted using an electric fryer (Blue Seal E44E, Birmingham, UK). The fryer was filled with 15 L canola oil (BidFood Smart Choice 19540, Dunedin, New Zealand) and pre-heated for at least 1 h before frying. For each kinetic frying experiment, at least 8 to 10 potato tubers were prepared and then PEF-treated with "PEF Low" and "PEF High" treatments, as described earlier in Section 2.2. After each PEF treatment, all potato tubers were sliced to approximately 1.0 mm thickness using a mandolin (Benriner, Yamaguchi, Japan). Potato slices from different tubers within the same treatment group and cultivar were then pooled together and blanched in boiling distilled water for 3 min over an induction cooker (Micasa MA0239IC, Auckland, New Zealand). After blanching, the excess surface water of the potato slices was reduced using a food dehydrator (Sunbeam DT5600, Auckland, New Zealand) at 75 ◦C for 10 min. Then, the potato slices were immediately used for the kinetic frying experiment.

At each frying temperature, approximately 50 slices of potato (±150 g) were randomly selected and transferred into a stainless-steel frying basket. When the oil reached the targeted temperature as monitored using a digital thermometer (Breville BMP100, Sydney, Australia), the frying basket was immersed in the hot oil. Up to 4 potato slices from 'Russet Burbank', 'Nadine' and 'Moonlight' were removed from the fryer at every time intervals of 60 s, 40 s, and 30 s for frying temperatures of 150, 170, and 190 ◦C, respectively. Our preliminary study showed that the colour of 'Crop77' potatoes changed very minimal when being fried for up to 15 min due to the exceptional low amount of glucose for this cultivar (Table 1). Therefore, the standard protocol for Nadine, Moonlight and Russet Burbank was adapted for 'Crop77' by frying at 170, 180 and 190 ◦C to allow accurate estimation of *k* and *Ea* values. For these reasons, potato slices from 'Crop77' were deep-fried at higher temperatures of 170, 180 and 190 ◦C to better study the changes in their colour as a function of frying time. The temperature-time profile of potato slices during frying was monitored using a K type thermocouple (0.2 mm diameter; Labfacility, South Yorkshire, UK) and recorded using a data logger (Picotech TC-08, Cambridgeshire, UK). As soon as the fried potatoes were removed from the fryer, they were placed on a paper towel to absorb excess surface oil and cooled. The kinetic frying experiment was repeated 3 times for each cultivar, where at least 8 to 10 potato tubers were used for each independent replication.


**Table 1.** Tuber characteristics, dry matter, total starch, amylose, glucose, and amino acid contents of 'Crop77', 'Moonlight', 'Nadine', and 'Russet Burbank' used in this study.

The data are presented as mean ± standard error (*n* = 3–6). Means in the same row not sharing the same letter are significantly different at *p* < 0.05 between cultivars based on one-way ANOVA and Tukey's post hoc test. DW = dry weight. † Originated from the breeding programme of the New Zealand Institute for Plant and Food Research Limited. # 'Crop77' represents the low-glucose control potato in the experimental design of this study. ## 'Nadine' represents the high-glucose control potato in the experimental design of this study.

> The CIE *L*\*, *a*\* and *b*\* colour of the fried samples was measured using a pre-calibrated colourimeter (MiniScanEZ-4500L, Hunterlab, Reston, VA, USA) based upon tristimulus CIE colour combination values, i.e., *L*\* (lightness (100) to darkness (0)), *a*\* (red (+) to green (−)), and *b*\* (blue (+) to yellow (−)) under D65-artificial daylight at 10◦ standard angle.

Due to the uneven surface of the fried potato slices, each sample was crushed finely using a bread roller and transferred into a white porcelain cup for colour measurement. A total of three colour measurements were taken for each fried sample.

#### 2.3.2. Estimation of the Time Dependence of the Colour Change of Potato Slices during Frying

The colour development of potato slices during frying typically followed first-order reaction kinetics (Equation (2)), as reported in other potato cultivars namely 'Panda' based on *b*\* value [27], 'Desiree' based on total colour difference (Δ*E*) [28,29], 'Rosetta' based on *L*\*, *a*\**,* and *b*\* values [30] and 'Russet Burbank' based on *L*\*, *a*\*, *b*\*, and Δ*E* values [31].

In this study, *L*\* (lightness) value was used to describe the colour change of potato slices from all four cultivars, with and without PEF and blanching pre-treatments, during frying as *L*\* was the only parameter that best fitted to the linearised form of the firstorder model (Equation (3)) with R2 close to 0.9. The rate constant to describe changes in the *L*\* value (*k*, in s−1) at each frying temperature was then estimated on the basis of Equation (3) [32].

$$L^\* = L\_0^\* \cdot \mathbf{e}^{-k \cdot t} \tag{2}$$

$$
\ln L^\* = \ln L\_0^\* - \ k \cdot t \tag{3}
$$

where *L*\* is the lightness of potato slices at frying time of *t* (dimensionless), *L*\*0 is the lightness of potato slices at time *t* = 0 s, *k* is the rate constant (s−1) for changes in the *L*\* value during frying, and *t* is frying time (s). The rate constant of lightness change (*k*) at each frying temperature was estimated as the slope from the plot of the natural logarithm of lightness (*L*\*) against frying time (*t*). Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) was used to estimate *k* values (based on Equation (3)) for non-PEF and PEF-treated samples fried at each frying temperature for three independent kinetic frying experiments.

2.3.3. Estimation of the Temperature Dependence of Rate Constant for Changes in *L*\* Value during Frying

The temperature dependence of the rate constant *k* for changes in *L*\* value during frying was estimated using Arrhenius equation (Equation (4)), which can be linearised using a logarithmic transformation, leading to Equation (5) [33].

$$k = A \cdot \exp\left(\frac{-E\_a}{R \cdot T}\right) \tag{4}$$

$$
\ln k = \ln A - \frac{E\_a}{R \cdot T} \tag{5}
$$

where *k* is the rate constant (s−1) for changes in *L*\* value at a specific frying temperature as estimated from Equation (3), *A* is a pre-exponential factor with the same dimension as that of *k*, *Ea* is the activation energy (kJ/mol−1), *R* is the universal gas constant (8.314 J·mol−1·K−1), and *<sup>T</sup>* is the actual frying temperature (K). The kinetic parameter *Ea* was estimated using a linear regression analysis by plotting the natural logarithm of the rate constant (*k*) of lightness (*L*\*) change versus the reciprocal of the absolute temperature (1/*T*) using Microsoft Excel (Microsoft Corporation, Redmond, WA, USA).

#### *2.4. Statistical Analysis*

The statistical significance of difference in the chemical content and colour parameters of potato samples between cultivars or between treatments was calculated using Student's *t*-test for single comparison or using an analysis of variance (ANOVA) for multiple comparisons, followed by post hoc Tukey's Honestly Significant Difference (HSD) test. Pearson's correlation coefficient (*r*) was used to determine the linear correlation between chemical content in potato tuber and the colour parameters of fried samples, followed by evaluation of statistical significance of the linear correlation with two-tailed probability values. The criterion employed for a statistical significance of the difference was *p* < 0.05. All statistical

analysis was performed using IBM SPSS Statistics version 25 (IBM Corporation, New York, NY, USA).

#### **3. Results and Discussion**

#### *3.1. Comparison on the Chemical Composition of Four Different Potato Cultivars*

Table 1 summarises the dry matter, total starch, amylose, glucose and total amino acid contents of the four potato cultivars used in this study. Dry matter of potato tubers varied significantly (*p* < 0.05) between the four cultivars, with both 'Crop77' and 'Russet Burbank' had the highest dry matter (>24%) and the lowest in 'Nadine' (<15%). Dry matter is known to be associated with the texture of potato when fried, where potatoes rich in dry matter usually resulted in a crispier texture after frying compared to those produced using potato cultivars with a lower dry matter [34]. Starch is a major component in potato, and it was found that most potato cultivars used in this study had similar total starch content (>708 mg/g DW), except for 'Moonlight' with the lowest total starch content, averaged at 587.34 mg/g DW. Moreover, it was found that amylose made up at least 24% of the total starch content in 'Moonlight' and 'Nadine', while amylose only constituted at least 10% of the total starch content for 'Crop77' and 'Russet Burbank'. Previous studies have shown that differences in the amylose content between potato cultivars is not uncommon and can influence the functional properties of starch, cooking quality and its end use for food application [35,36].

Glucose is an important precursor involved in the colour development of potato during frying, i.e., through caramelisation and the Maillard reaction [30]. When comparing the glucose content between the potato cultivars, 'Nadine' displayed the highest levels of glucose (104.64 mg/g DW), followed by 'Moonlight' and 'Russet Burbank' sharing similar range between 21 and 22 mg/g DW, and it was interesting to detect a very low amount of glucose in 'Crop77' (0.75 mg/g DW) (Table 1). Therefore, it is expected that the high glucose content in 'Nadine' will lead to excessive colour browning in the potato slices when fried. According to Bartlett, et al. [37], potato cultivar with total reducing sugars of greater than 50 mg/g is generally considered as tubers with a high acrylamide risk in any deep-fried potato products manufacturing line (e.g., French fries and potato crisp). Another key precursor involved in Maillard reaction to allow reaction with the carbonyl group of glucose is amino acid [30]. In this study, it was found that the total amino acid contents of 'Nadine', 'Russet Burbank' and 'Moonlight' were similar and were significantly (*p* < 0.05) higher than that of 'Crop77'.

Clearly, 'Crop77' potatoes were characterised by low glucose content, low in amino acids but high in dry matter and total starch. 'Russet Burbank' and 'Moonlight' potatoes shared a few similarities in the chemical contents, where both cultivars had moderate levels of glucose and amino acids. However, the starch content in 'Russet Burbank' was higher than in 'Moonlight'. High glucose and amino acids contents, in conjunction with low dry matter are the chemical features for 'Nadine' cultivar. With this in mind, 'Crop77' and 'Nadine' represent the low- and high-glucose control potatoes in this study, respectively. It is, however, important to note that the use of 'Nadine' tubers for deep-fried potato processing line is typically unfavourable due to its exceptionally high glucose content and low dry matter.

#### *3.2. Colour Evaluation of Fried Potato Slices Produced from Non-PEF and PEF-Treated Potatoes*

To justify the effectiveness of PEF in controlling the colour development of potato slices during frying, the inclusion of low- and high-glucose control potatoes (i.e., using 'Crop77' and 'Nadine' respectively in this study) is considered for the first time in the literature. The colour characteristics of potato slices fried under the same frying condition i.e., at 180 ◦C for 3 min, without PEF nor blanching pre-treatments were initially assessed. In agreement with the high glucose content found in 'Nadine', it is not unexpected that the resulting fried samples from this cultivar displayed the lowest *L*\* (reduced lightness) and *b*\* (reduced yellowness) (Table 2) compared to other potato cultivars, when all fried at

180 ◦C for 3 min. On the contrary, potato slices from 'Crop77', with the glucose content averaged at least 140-fold lower than 'Nadine', exhibited the highest *L*\* (lightness was at least 2-fold higher than fried 'Nadine'), lowest *a*\* (redness was at least 1.3-fold lower than fried 'Nadine'), and the highest *b*\* (yellowness was at least 2-fold higher than fried 'Nadine') when fried (Table 2). The average *a*\* values for fried potato slices from 'Moonlight' and 'Russet Burbank' were rather similar (Table 2) and significantly redder compared to fried 'Nadine' and 'Crop77'. While both 'Russet Burbank' and 'Moonlight' shared similar glucose content (Table 1), 'Russet Burbank' fried samples were found to demonstrate a higher average *L*\* and lower *b*\* value compared to fried samples from 'Moonlight'. Based on the Pearson's correlation analysis, the difference in the colour characteristics of fried potato slices (Table 2) from four cultivars, especially *L*\* (*r* = −0.9893) and *b*\* (*r* = −0.9810), appeared to correlate significantly (*p* < 0.05) with the level of glucose (i.e., one of the important precursors involved in the Maillard reaction browning during frying) found in the tubers (Table 1). The *a*\* value of fried potato slices was found to be positively correlated with the total amino acids content in the tuber instead (*r* = 0.8769, *p* < 0.05). Coincidentally, previous studies have consistently evidenced that both *L*\* and *b*\* values of fried potatoes are highly associated with the glucose content of potatoes [7], while the colour parameter *a*\* correlates better with amino acids and acrylamide concentration [38].

**Table 2.** The effect of PEF pre-treatment alone (without blanching) on the *L*\*, *a*\* and *b*\* values of fried potato slices (180 ◦C for 3 min) from 'Crop77', 'Moonlight', 'Nadine', and 'Russet Burbank'.


The data are presented as mean ± standard deviation (*n* = 15 measurements, 3 measurements from 5 independent potato tubers per treatment). Means in the same row for each cultivar not sharing the same lowercase letter in subscript are significantly different at *p* < 0.05 between untreated and PEF-treated potatoes based on one-way ANOVA and Tukey's post hoc test. Means in the same column for each colour parameters not sharing the same uppercase letter in superscript are significantly different at *p* < 0.05 between cultivars of the same treatment based on one-way ANOVA and Tukey's post hoc test.

3.2.1. The Effect of PEF Pre-Treatment Alone (without Blanching) on the Colour Characteristics of Fried Potato Slices

Based on Table 2, the colour characteristics of fried potatoes was found to be highly dependent on the initial glucose content of the tubers and the intensity of PEF treatment. Fried 'Nadine' slices produced from PEF Low treatment exhibited a higher lightness (*L*\*) and reduced redness (*a*\*) compared to fried samples from untreated 'Nadine'. Moreover, *a*\* was the only colour parameter for fried 'Russet Burbank' slices, produced from PEF Low treatment, found to be reduced significantly (*p* < 0.05) when compared to untreated

counterpart. Therefore, the results indicate that application of PEF Low pre-treatment on the tubers prior to frying is particularly effective for high glucose-containing 'Nadine' to limit the degree of Maillard browning upon frying. The cell electroporation effect of PEF on promoting the leakage of chemical constituents [39] from potato tuber is likely the reason to minimise the level of precursors (e.g., sugars) involved in Maillard reaction. With respect to the 'Crop77' and 'Moonlight', the application of PEF Low treatment on the tubers did not influence both the lightness and redness of fried potato slices. Therefore, it can be concluded that an application of PEF Low pre-treatment on potatoes was promising in controlling the colour development during frying when applied on a high glucose-containing potato cultivar such as 'Nadine' and posed almost negligible effect when applied on a low glucosecontaining potato cultivar such as 'Crop77'. This is because the glucose content in 'Nadine' potatoes was reduced by 74% with PEF Low pre-treatment while no apparent changes in the glucose content for 'Crop77' potatoes before and after PEF treatment. PEF treatment tends to cause changes to the microstructure of plant material and a recent study by Zhang et al. [40] showed that potato cells can be severely disrupted by PEF in which larger pores can then be formed, accelerating the mass transfer process such as leakage of cell constituents after PEF.

Similar to PEF Low treatment, the application of PEF at higher energy (i.e., increased from 50 to 150 kJ/kg at a fixed electric field strength of 1 kV/cm) did not cause significant changes in the colour parameters of the fried 'Crop77' slices. The reason is due to the exceptional low glucose content for this potato cultivar as no considerable in glucose reduction was observed after PEF. If leaching of glucose from the 'Crop77' potato slices did occur due to PEF, the amount of glucose remained in the tissue could be too low to influence the colour characteristics of fried potato slices. On the contrary, it was found that PEF High treatment significantly (*p* < 0.05) improved the colour parameters for both 'Russet Burbank' (*a*\* reduced from 5.6 to 4.2) and 'Nadine' (*L*\* increased from 17 to 20 and *a*\* reduced from 4.5 to 3.5) fried samples to that of their untreated counterparts. However, it is important to note the extent of improvement in lightness and reduction of redness of fried samples for these two cultivars by PEF High pre-treatment were similar as when PEF Low treatment was applied to these tubers prior to frying. This result strongly suggested that application of PEF at an electric field strength of 1 kV/cm and energy ~50 kJ/kg on potatoes was adequate in controlling the degree of colour changes upon frying.

3.2.2. The Effect of Sequential PEF and Blanching Pre-Treatment on the Colour Characteristics of Fried Potato Slices

When the potato slices were blanched and then fried immediately at 180 ◦C for 3 min, fried samples from all non-PEF potatoes, except for 'Crop77', exhibited a significant (*p* < 0.05) improvement in lightness (*L*\*) compared to their unblanched counterparts (Table 3). The extent of *L*\* increase was the greatest for fried 'Nadine', followed by 'Moonlight' and 'Russet Burbank'. The fact that fried 'Nadine' exhibited the greatest increase in *L*\* indicates that blanching is a vital unit operation to effectively remove sugars from the potato of inherently high sugar load to reduce the excessive colour development of fried potato slices associated with Maillard reaction. In this study, blanching was able to cause up to 66% reduction in glucose in 'Nadine'. Blanching pre-treatment also led to a significant (*p* < 0.05) reduction in redness (*a*\*) of fried samples for all potato cultivars, except 'Nadine'. With respect to the yellowness (*b*\*) of the fried potato slices, the impact of blanching treatment was not consistent between cultivars. Compared to their unblanched counterparts, *b*\* reduced significantly for blanched 'Crop77', increased significantly for both 'Russet Burbank' and 'Nadine', and no changes in *b*\* was observed for fried samples between unblanched and blanched 'Moonlight'. Despite this, the *b*\* value of fried potato slices from all 4 cultivars appeared to fall within the same range (between 12.88 and 18.03), indicating that the variation in *b*\* of fried samples between cultivars was minimised with blanching pre-treatment and such observation was not found in other colour parameters of *L*\* and *a*\* of fried potato samples.


**Table 3.** The effect of sequential PEF and blanching pre-treatments on the *L*\*, *a*\* and *b*\* values of fried potato slices (180 ◦C for 3 min) from 'Crop77', 'Moonlight', 'Nadine', and 'Russet Burbank'.

The data are presented as mean ± standard deviation (*n* = 15 measurements, 3 measurements from 5 independent potato tubers per treatment). Means in the same row for each cultivar not sharing the same lowercase letter in subscript are significantly different at *p* < 0.05 between untreated and PEF-treated potatoes based on one-way ANOVA and Tukey's post hoc test. Means in the same column for each colour parameters not sharing the same uppercase letter in superscript are significantly different at *p* < 0.05 between cultivars of the same treatment based on one-way ANOVA and Tukey's post hoc test. \* indicates significant difference (*p* < 0.05) in the colour parameters, based on a Student's *t*-test, due to the blanching treatment when compared to the untreated and PEF-treated potato samples without subjected to a blanching treatment (Table 2).

Moreover, the results from this study showed that the lightness of fried 'Russet Burbank' slices with blanching pre-treatment overlapped with 'Crop77' and 'Moonlight', which was not exhibited in their unblanched counterparts. With respect to the redness of fried potato slices, unblanched 'Moonlight' and 'Russet Burbank' upon frying exhibited the highest *a*\* among all four cultivars (Table 2) and the subsequent application of blanching step reduced their redness significantly (Table 3). Clearly, blanching can reduce the colour development of fried potato slices and possibly minimise variation in colour across different potato cultivars of varying levels of glucose due to the effective removal of inherent sugars prior to frying. In fact, recent studies by Bartlett, et al. [37] and Zhang, et al. [41] consistently reported that extending the blanching duration further provoked changes in the microstructure of potato tissues, consequently facilitating the release of sugars and amino acids involved in Maillard reaction, and thus maximising the reduction of acrylamide of potato slices after frying.

The sequential PEF and blanching treatment on the colour characteristics of fried potato slices is presented in Table 3. When PEF Low was applied to the potatoes, followed by blanching, the *L*\* of resulting fried samples improved significantly (*p* < 0.05) for those produced from 'Crop 77 and 'Russet Burbank'. Such observation was not found in those fried samples when the potatoes were treated with PEF Low alone without subjected to the subsequent blanching step (Table 2). In particular, the subsequent blanching step after PEF Low treatment further increased the lightness of fried potato slices of low glucosecontaining 'Crop 77 . Furthermore, the redness of fried samples reduced significantly (*p* < 0.05) for all cultivars after the sequential PEF Low and blanching treatment compared to their non-PEF counterparts, while the yellowness of the resulting fried samples for all potatoes were similar to their non-PEF counterpart. Among all the cultivars, the sequential PEF Low and blanching treatment was found effective in causing between

22 (least in 'Crop77') and 74% (the greatest in 'Nadine') reduction in glucose. In this context, the effective removal of glucose from potato tissues could be attributed by the cell electroporation effect after PEF combined with the depolymerisation and solubilisation of pectin during blanching step [42].

When the sequential PEF High and blanching treatment applied to the potatoes, only fried 'Russet Burbank' samples experienced a significant (*p* < 0.05) increase in *L*\* compared to non-PEF counterparts. However, the lightness of fried 'Russet Burbank' was similar when the tubers were treated with PEF Low and PEF High, suggesting that the application of sequential PEF Low and blanching treatment was adequate to improve the lightness of potato slices upon frying. A similar result was observed with respect to the redness and yellowness of fried samples. The glucose result shows that the glucose amount remaining in 'Russet Burbank' was similar when pre-treated with either PEF Low or PEF High treatment. However, the sequential PEF High and blanching treatment on 'Moonlight' appeared effective in reducing the yellowness of the resulting fried samples (Table 3). The reason behind could be due to the ability of sequential PEF High and blanching treatment in causing up to 80% reduction in glucose, which has not been observed in any other cultivars.

Overall, it is interesting to observe that the colour characteristics (especially *L*\* and *a*\*) of fried potato slices for all cultivars were impacted more significantly due to sequential PEF and blanching treatments compared to when PEF treatment alone (without blanching step) was applied. Therefore, findings from this study revealed a positive synergistic effect of the sequential PEF and blanching treatment on potato, when tested across four potato cultivars with varying content of inherent glucose and amino acids, in improving the colour characteristics of fried potato slices. The result also showed that PEF treatment could be possibly applied at a lower intensity by combining with blanching step during the production of fried potato slices to achieve improved colour characteristics. For example, fried 'Russet Burbank' and low glucose-containing 'Crop77' from any PEF-treated tubers shared similar lightness. Due to consistent and significant improvements in the colour characteristics of fried potato slices observed across four cultivars, the effect of PEF followed by blanching treatments on the kinetics of their colour changes during frying was modelled and is further discussed in the next section (Section 3.3). The inclusion of combined PEF as a pre-treatment with a subsequent blanching step is considered to be a more effective and strategic acrylamide mitigation measure when producing deep-fried potato products [43].

#### *3.3. Kinetic Study on the Colour Changes of Sequential PEF and Blanching Pre-Treated Potatoes during Frying*

The Maillard browning reaction is a complex reaction, and colour development is considered the final phase in this reaction [32]. The first-order kinetic model has been applied on the initial, intermediate, and final phases of the Maillard reaction [33]. Therefore, in this study the change in colour parameters *L*\*, *a*\* and *b*\* of pre-blanched potato slices for all four cultivars were modelled as a function of frying time and temperature. Compared to other colour parameters, the changes in *L*\* value followed a first-order reaction, and the measured data fitted well to the first order reaction equation (Equation (3)) with *R*<sup>2</sup> close to 0.9. Therefore, the *L*\* parameter was used in this study to describe the colour development of potato slices during frying.

#### 3.3.1. Time and Temperature Dependences of *L\** Value Change for Non-PEF (Blanching Only) Pre-Treated Potatoes during Frying

Representative plots of the natural logarithm of *L*\* (ln *L*\*) against frying time, for fried potato slices from the four cultivars and the fitting of the first-order kinetic model to the experimental lightness data, are illustrated in Figure 1. Clearly, changes in the lightness for all investigated cultivars during frying obeyed first order reaction kinetics (Equation (3)). The rate constant (*k*) describing the changes of *L*\* at each frying temperature was estimated using linear regression analysis (Table 4). The estimated rate constant *k* increased with increasing frying temperature, indicating the rate of changes in lightness of potato slices typically accelerated when they are fried at higher temperatures. At the same frying

temperature (i.e., 170 and 190 ◦C), there was an apparent cultivar effect on the estimated *k* values in *L*\*, where the fried potato slices from 'Nadine' were found to darken (reduction in *<sup>L</sup>*\*) the fastest (1.93–2.96 × <sup>10</sup>−<sup>3</sup> <sup>s</sup><sup>−</sup>1), followed by 'Russet Burbank' (1.63–2.87 × <sup>10</sup>−<sup>3</sup> <sup>s</sup><sup>−</sup>1) and 'Moonlight' (1.27–1.87 × <sup>10</sup>−<sup>3</sup> <sup>s</sup><sup>−</sup>1), and the slowest in 'Crop77' (0.24–0.83 × <sup>10</sup>−<sup>3</sup> <sup>s</sup><sup>−</sup>1). Moreover, findings from this kinetic study revealed that the estimated *k* value in *L*\* for high glucose-containing 'Nadine', together with 'Russet Burbank' and 'Moonlight', fried at 150 ◦C are likely to fall into a similar rate as low glucose-containing 'Crop77', when fried at 190 ◦C (Table 4).

**Figure 1.** An example of fitting the first-order kinetic model to the experimental lightness data (natural logarithm of *L*\* values against frying time) at different frying temperatures for 'Crop77', 'Moonlight', 'Nadine' and 'Russet Burbank' subjected to blanching treatment only. Symbols represent the experimental lightness data (average measurements from three independent frying experiments per frying time point where each experiment involved a batch of 8–10 potatoes) and the predicted lightness values from the first-order reaction (Equation (3)) are represented by the continuous dots. The *L*\* of potato slices prior to frying was averaged at 57.15 ± 1.33, 57.09 ± 1.48, 54.60 ± 2.17 and 52.76 ± 1.49 respectively for 'Crop77', 'Moonlight', 'Nadine' and 'Russet Burbank'.


**Table 4.** Estimated kinetic parameters of time dependence *k* and temperature dependence *Ea* of changes in lightness (*L*\*) of fried potato slices produced from cultivars 'Crop77', 'Moonlight', 'Nadine' and 'Russet Burbank' without and with PEF pre-treatment followed by blanching.

\* The rate of changes in *L*\* (*k*) is presented as average of estimated kinetic parameter ± standard deviation of the estimates from three independent frying kinetic experiments (where each experiment involved a batch of 8–10 potatoes). Estimated *k* values within the same cultivar and frying temperature not sharing the same lowercase letter in subscript are significantly different at *p* < 0.05 between non-PEF and PEF-treated potatoes based on one-way ANOVA and Tukey's post hoc test. \*\* Activation energy (*Ea*) is presented as average of estimated kinetic parameter ± standard deviation of the estimates from three independent frying kinetic experiments (where each experiment involved a batch of 8–10 potatoes). Estimated *Ea* values within the same cultivar not sharing the same lowercase letter in subscript are significantly different at *p* < 0.05 between non-PEF and PEF-treated potatoes based on one-way ANOVA and Tukey's post hoc test. Estimated *Ea* values within the same treatment group not sharing the same uppercase case letter in superscript are significantly different at *p* < 0.05 between cultivars based on one-way ANOVA and Tukey's post hoc test.

> To describe the temperature dependence of *k* values for the four potato cultivars, the Arrhenius equation (Equation (5)) was used. The result of fitting the Arrhenius equation to the natural logarithm of *k* (ln *k*) describing the change in *L*\* for fried potatoes and the reciprocal of absolute temperature (1/T) is depicted (Figure 2). It was found that the estimated *Ea* values varied significantly (*p* < 0.05) according to the potato cultivar (Table 4),

ranging from 31.73 kJ/mol (lowest in 'Moonlight') to 105.53 kJ/mol (highest in 'Crop77'). Moreover, the estimated *Ea* values for the changes in lightness during frying of 'Nadine' potato slices was significantly (*p* < 0.05) lower than 'Russet Burbank'. The estimated *Ea* value obtained in this study for fried 'Russet Burbank' (45.37 kJ/mol) was slightly higher than previous study reported at 43.2 kJ/mol [31]. Potato cultivar displaying a lower estimated *E*<sup>a</sup> values typically indicate a lower temperature dependency of the *k* values with respect to the lightness of fried samples, and vice versa for those potato cultivars displaying a higher estimated *E*a values. Overall, the *k* values of colour parameter *L*\* during the frying of 'Moonlight', 'Nadine' and 'Russet Burbank' potato slices appeared to be less temperature sensitive (or more temperature stable) than that of low glucose-containing 'Crop77' potatoes.

**Figure 2.** Arrhenius plot of the natural logarithm of *k* values for changes in *L*\* during frying against the reciprocal of absolute frying temperature for 'Crop77', 'Moonlight', 'Nadine' and 'Russet Burbank' subjected to blanching pre-treatment only. Symbols represent the estimated *k* values for *L*\* from first-order reaction (Equation (3)) and the standard error bars represent the standard deviation (average from three independent frying experiments where each experiment involved a batch of 8–10 potatoes).

3.3.2. Time Dependency of *L\** Value for PEF-Treated Potatoes during Frying

Table 4 summarises the rate constant (*k*) describing the reduction of *L*\* increased with frying temperature, estimated for PEF-treated potatoes of different cultivars. Interestingly, when either PEF Low or PEF High were applied to the 'Russet Burbank' potatoes, a significantly (*p* < 0.05) lower *k* value for lightness in the fried potato slices at 190 ◦C was observed in comparison to their non-PEF (only subjected to blanching) counterparts. This observation strongly suggests that PEF-treated 'Russet Burbank' experienced a slower reduction in the *L*\* colour parameter during frying at high temperature of 190 ◦C. It could be that the cell electroporation effect of PEF treatment, followed by the subsequent blanching step, modified the structural integrity of potato cells [42] and possibly altered the levels of precursors (e.g., reducing sugars and amino acids) responsible for Maillard browning reactions in the potato slices prior to frying, thus delaying the colour darkening (i.e., *L*\* reduction) in the fried potato slices. Nevertheless, fried potato slices produced from the other three potato cultivars after sequential PEF and blanching treatments did not exhibit a similar trend in rate of colour change in *L*\* to that of 'Russet Burbank' at any frying temperature. In other words, the estimated *k* values for changes in *L*\* at each frying

temperature for potato slices from low glucose-containing 'Crop77', 'Moonlight' and high glucose-containing 'Nadine' were not affected at a statistical significant level (*p* > 0.05) by any of the PEF pre-treatments applied (Table 4). This result implies that the effectiveness of sequential PEF and blanching treatments in reducing the rate of lightness during potatoes frying could be highly dependent on the cultivar. Apart from the differences in the availability of precursors for the colour development in Maillard reaction, variation in the solid content of potato has been reported to influence their tissue conductivity, thus impacting the efficacy of PEF treatment [22].

#### 3.3.3. Temperature Dependency of *k* for *L\** Value for PEF-Treated Potatoes during Frying

Table 4 shows the estimated activation energy (*Ea*) describing the changes of *L*\* for each potato cultivar during frying due to sequential PEF and blanching treatments. Regardless of the intensity of PEF pre-treatment applied, the fried samples from 'Crop77' consistently demonstrated the highest estimated *Ea* value with respect to the changes in *L*\* during frying. On the contrary, fried samples from any pre-treated 'Moonlight' consistently demonstrated the lowest estimated *Ea* value.

The estimated *Ea* values for the rate of *L*\* changes of fried potato slices were significantly (*p* < 0.05) lower in all PEF-treated 'Russet Burbank' compared with the non-PEF samples (Table 4). A similar finding was found for fried samples from PEF-treated low glucose-containing 'Crop77' but not for 'Moonlight' and high glucose-containing 'Nadine'. The *k* values describing the changes of colour parameter *L*\* during the frying of 'Russet Burbank' and 'Crop77' potato slices appeared to be significantly (*p* < 0.05) more temperature stable (i.e., reduction of estimated *Ea*) because of the application of sequential PEF and blanching treatment. This is contrary to the opposite behaviour exhibited by PEF-treated sweet potato during frying [19], where the estimated *Ea* value describing the temperature dependency of the rate of changes of *L*\* increased for PEF-treated sweet potato, suggesting the resulting slices become more temperature sensitive towards browning/darkening when fried. The discrepancy between the sweet potato result and the findings from the present study on potato tubers could be due to the tissue type difference between the two vegetable matrices [44], and also attributed to the fact that a blanching step was applied to all potatoes in the present study prior to frying.

Another interesting finding from this study is that the estimated *Ea* values for 'Crop77' seem to demonstrate a PEF processing intensity specific effect, where a lower *Ea* value was estimated for fried samples from 'Crop77' pre-treated at PEF Low compared to those treated at PEF High. This finding was not observed for 'Russet Burbank' as the *Ea* values for fried potato slices from PEF Low-treated were significantly (*p* < 0.05) higher than PEF High-treated 'Russet Burbank' potato slices. Overall, potato slices from PEF-treated 'Russet Burbank' (especially at PEF High) and low glucose-containing 'Crop77' (especially at PEF Low) are expected to experience a small change in *L*\* parameter (i.e., slower browning) during frying over a wider temperature range (i.e., more temperature stable).

The availability of glucose and amino acids on the surface of potato slice, as a reactant for the formation of melanoidin, as part of the Maillard reaction that led to browning of potato slices during frying, is highly associated with the diffusion characteristic of potato slices, the mass transfer of frying oil penetrating the potato slice, and moisture leaving the potato tissue during frying [45,46]. Pore formation at the cell membrane of potato tissue due to PEF treatment has been reported to cause changes in the diffusion characteristics of potatoes [22,23,39,47], thus affecting the removal of sugar and the subsequent heat and mass transfer processes during frying [48,49]. Moreover, the inclusion of a blanching step after PEF treatment, prior to frying, is another effective way of removing excess sugar from the potatoes by diffusion process due to severe disruption of potato tissues of the thermal effect of blanching in weakening the binding between cells (i.e., middle lamella) [10,42]. The aforementioned are likely the reasons that the estimated kinetic parameters underpinning colour development during potato slices frying was significantly altered for 'Crop77' and 'Russet Burbank' when both were subjected to sequential PEF and blanching treatments

(Table 4). However, it is not possible to explain why similar results were not observed for 'Moonlight' and high glucose-containing 'Nadine'.

#### **4. Conclusions**

This study provides an important insight with respect to the colour development of fried potato slices when potato tubers with significant inherent differences in the content of precursors involved in Maillard reaction browning, namely glucose and amino acids, were subjected to sequential PEF and blanching treatments prior to frying. PEF pre-treatment did not change the kinetics of changes in *L*\* values during frying for any of the four cultivars, namely 'Crop77', 'Moonlight', 'Nadine' and 'Russet Burbank', which followed first-order reaction kinetics. While frying of potato slices of low glucose-containing 'Crop77' exhibited the highest *Ea* value for changes in *L*\* among all cultivars, it was found that the estimated *Ea* value decreased significantly (i.e., more temperature stable), by at least 18% when the potatoes were pre-treated with PEF Low prior to frying. With respect to high glucose-containing 'Nadine', this cultivar is generally not suitable for commercial deepfried potato process lines, but the findings from this study reveal that sequential PEF and blanching treatment is very effective at reducing the *L*\* of 'Nadine' slices, with reductions of up to 38%. In addition, the frying kinetic result of non-PEF and PEF-treated 'Nadine' showed that the rate of changes in *L*\* was less temperature sensitive than equivalently pre-treated low glucose-containing 'Crop77'. In this study, it was rather surprising to find that potato slices from moderate glucose-containing 'Moonlight' exhibited the lowest estimated value of *Ea* among all cultivars for any PEF pre-treatment applied to the tubers prior to frying, indicating the *k* value for changes in *L*\* of the potato slices for this cultivar is less temperature sensitive with increasing frying temperature. The sequential PEF and blanching treatments appeared to benefit 'Russet Burbank' the most in which changes in *L*\* of the potato slices became more temperature stable (i.e., up to 30% reduction in *Ea* value with PEF High treatment) over a wide range of frying temperature when higher intensity of PEF was applied to the tuber prior to frying. Clearly, this research provides new evidence that colour development of potato slices during frying can be modulated with PEF pretreatment on tubers in conjunction with smart selection of potato cultivar. The reduction of estimated *Ea* value due to sequential PEF and blanching treatments was prominent, particularly when frying potato slices from 'Crop77' and 'Russet Burbank', which may bring advantages to the process control of deep-fried potato industries, especially when the temperature distribution inside the fryer is not uniform during processing. It is expected that the results of the present study would be helpful in predicting the impact of sequential PEF and blanching treatments on the colour development of potato tubers from a wide range of physicochemical properties.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2304-815 8/10/10/2307/s1, File S1: Determination of the content of total starch, amylose, glucose, total amino acids, and dry matter of the tubers.

**Author Contributions:** Conceptualisation, S.B.M.A., S.Y.L., S.B., D.J.B. and I.O.; methodology, S.B.M.A., S.Y.L., C.Z. and I.O.; software, S.B.M.A. and C.Z.; validation, S.B.M.A. and C.Z.; formal analysis, S.B.M.A., S.Y.L. and C.Z.; investigation, S.B.M.A., S.Y.L. and C.Z.; resources, S.B., D.J.B. and I.O.; data curation, S.B.M.A., S.Y.L. and C.Z.; writing—original draft preparation, S.B.M.A. and S.Y.L.; writing—review and editing, S.B., D.J.B., D.A. and I.O.; visualisation, S.B.M.A.; supervision, S.Y.L., S.B., D.A. and I.O.; project administration, S.B., D.J.B. and I.O.; funding acquisition, S.B., D.J.B. and I.O. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by New Zealand Ministry of Business, Innovation and Employment (contract MAUX1402) for the Food Industry Enabling Technologies (FIET) programme. This study is also supported by the Riddet Institute, a New Zealand Centre of Research Excellence, funded by the Tertiary Education Commission.

**Data Availability Statement:** The datasets generated for this study are available on request to the corresponding author.

**Acknowledgments:** Abduh acknowledges the Indonesia Endowment Fund for Education (LPDP) for supporting him with a PhD scholarship. The authors would also like to thank Alex McDonald Limited for kindly providing tubers of the cultivars 'Nadine', 'Moonlight' and 'Russet Burbank'. Technical staff from the Department of Botany (Pamela Cornes and Toni Renalson) and Department of Food Science (Ian Ross) are thanked for providing support during the transport and storage of potatoes.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. The New Zealand Institute for Plant and Food Research Limited is a government-owned research institute and provided potato tubers from international public domain and domestically bred varieties.

#### **References**


*Article*
