Modeling Water Loss During Frying of Potato Slices

Abstract

The effect of blanching and drying on water loss and oil uptake during potato slice frying was studied. Two pretreatments were carried out for the slices: (i) blanching in hot water at 85°C for 3.5 min; (ii) blanching in hot water at 85°C for 3.5 min and then, air-dried until a moisture content of 60 % (wet basis) was obtained. A control treatment, consisting of unblanched potato slices not air-dried was also conducted. After the pretreatments, potato slices were deep-fried in sunflower oil at 120, 150, and 180°C. With respect to the control treatment, blanching provided an increase while blanching-drying a decrease of the oil uptake. Two models, based on Fick's law, were used to describe water loss during frying. The first one is the classic model with an effective moisture diffusion coefficient, which assumed a constant value. The second model considers that the diffusion coefficient varies during the frying process. The variable diffusivity model adjusted the experimental data much better than the constant diffusivity model did. The effective moisture diffusion coefficient, in the variable diffusivity model, increased with frying time and temperature and behaved very similar for both control and blanched slices, while the increase of this coefficient was considerable higher for the blanched-dried slices. The development of a more porous structure with less oil content in the case of blanched-dried slices could explain this fact.

Keywords:

• Frying
• Potato
• Water loss
• Oil uptake
• Diffusivity
• Blanching

INTRODUCTION

Potato chips were invented in 1853 and quickly became very popular. Today they represent 33% of total sales of snack in US market.[1] Retail sales in the US of potato chips are about $6 billion per year.[2] Deep-fat frying is a widespread operation used throughout food processing industries as well in catering operations and home preparation of foods that creates unique textures and flavors in foods. Immersion frying has been defined as the immersion of a food product in an edible oil or fat heated above the boiling point of water,[3] and may, therefore, be considered a dehydration process. Frying temperatures can range from 120–190°C, but the most common temperatures are 170–190°C. These conditions lead to high heat transfer rates, rapid cooking, browning, and texture and flavor development.[4] Development of an oily and dry consistency is a major structural transformation occurring during frying of potato chips due to the high temperature and oil infiltration processes.[5] High heat transfer rates during frying of potato induces major changes in the microstructure that, in turn, determines their final physical and sensorial attributes.[6] The high heat transfer rates are due to the temperature differential between the oil and potato samples and a high convective heat transfer coefficient that is enhanced by the rapid escape of water, in the form of vapor, from the product.[3]

In the potato chip process, raw potatoes are washed, peeled, sorted, and cut into slices. Some potato processing plants use blanching in hot water and drying with warm air to reach a moisture content of 60% (wet basis) prior to frying. Next, potato slices are usually fried in continuous fryers with hot oil (170-190°C), where they remain until the moisture level is less than 2% of the total weight. Finally, potato chips are salted, cooled in ambient air, and packaged. The blanching step improves the color and texture and could reduce the oil uptake by gelatinization of the surface starch.[7] Air dehydration leads to a lower moisture content that also reduces oil absorption.[8]

In recent years, several procedures have been proposed to reduce the amount of absorbed oil in fried potatoes. Fructose has been added to a restructured potato product that resulted in a change of the surface properties, with a reduction of absorbed oil after frying.[9] Soaking of potato strips in NaCl solutions could reduce oil uptake in ∼15% in french fries after frying.[10,11] Predrying of potatoes is a common way to reduce fat uptake in the final fried product.[12,10] Vacuum frying may also be an option for fried potatoes with low oil content and the desired texture and flavor characteristics.[1] Since the properties of the surface of the potatoes are most important for fat uptake, the application of a proper coating is a promising route to reduce oil content.[13] Many options are available to reduce fat uptake by application of coatings or batters. For instance, gellam gum has been used to coat samples so that the resulting film reduced the oil uptake during frying.[14]

In addition to the oil content reduction, knowing how pretreatments, such as blanching or drying, affect the water loss during frying is important. The water loss mechanism during frying is complex, and the transport by molecular diffusion, capillary, and pressure-driven flow should be determined.[15] However, modeling at different levels of complexity has been reported. Moisture content of potato slices has been fitted with frying time and the square root of frying time.[16] A more complex approach tested two models: one controlled by heat transfer and the other one by diffusivity of water in the tissue.[17] A first order kinetic model has been considered in which the rate of moisture loss was proportional to the moisture content.[18,19] In more complex models, the crust and the core have been treated as two regions separated by a moving boundary and included a pressure-driven flow in the crust region.[4] Liquid water transport has also been considered, resulting from convective flow and from capillary flow and vapor by convective flow and diffusion.[15] With different approaches, the Fick's law of diffusion has been extensively used to describe the water loss kinetics during frying.[20,21] Fick's law of diffusion provides a simplified picture of the water loss during frying; however, in order not to oversimplify the transport process, it should be considered that using a variable effective diffusion coefficient would account for the change of physical properties of the material during frying.

The principal objective of this work was to i) compare the oil uptake during frying between control potato chips and those pretreated (blanched and blanched-dried); ii) apply Fick's law of diffusion, with constant and with variable effective diffusion coefficients, in order to model the water loss kinetics during frying at three temperatures for: (a) control potato slices, (b) blanched potato slices, and (c) blanched-dried potato slices.

MATERIALS AND METHODS

Materials

Potatoes (variety Panda, ∼20 % of dry solids) and sunflower oil (Chef^TM, Coprona, Chile) were the raw materials. Potatoes that had been stored in a dark room at 8°C and 95% of relative humidity were washed and peeled. Slices (thickness of 2.5 mm) were then cut from the pith of the parenchymatous region of potato tubers using an electric slicing machine (Berkel, model EAS65). A circular cutting mold was used to provide chips with a diameter of 37 mm.

Pretreatments

Slices were rinsed immediately after cutting for 1 min in distilled water to eliminate some starch material adhering to the surface prior to frying. The following treatments were used: (i) Control or unblanched slices without predrying; (ii) Slices blanched in hot water at 85°C for 3.5 min; (iii) Slices blanched in hot water at 85°C for 3.5 min and then dried in a convection oven (Memmert, ULM 500, USA) at a dry bulb temperature of 60±1°C and an air velocity of 1±0.1 m/s until reaching a moisture content of 60±1 % (wet basis).

Frying conditions

Ten slices per sample time of each pretreatment were deep fried in 8 liters of hot oil contained in an electrical fryer (Beckers, Model F1-C, Italy) at each of the three temperatures tested: 120°C, 150°C and 180°C, with two replicates at each temperature. Potato-to-oil weight ratio was maintained as low as possible (∼0.0035) in order to keep the temperature of frying constant (±1°C). Slices were fried at different time intervals until a final moisture content of ∼1.8 % (wet basis) was reached. Previously, the corresponding minimum frying times to reach moisture content of ∼1.8% and the sampling intervals for each frying temperature were determined experimentally. For each selected sampling time, the fried slices were drained over a wire screen for 5 minutes and allowed to cool to room temperature before oil and water content analysis were done. The oil was preheated for 1 hr prior to frying, and discarded after 6 hr of use.[22]

Analysis

Mean moisture content of potato chips was measured by drying the samples in a convection oven at 105°C until constant weight. The oil content was determined by a simple and rapid method that consists of an initial extraction with a mixture of 1:2:0.8 (v/v/v) in chloroform, methanol, and water. This mixture was then adjusted to 2:2:1.8 (v/v/v) to continue the oil extraction. In this way, the chloroform layer contained the purified oil.[23]

Modeling Water Loss

Two models were used to describe water loss during frying. In the first, Fick's law of diffusion was used, which for one-dimensional transport is given by:[24]

(1)

where l= half thickness of the slab (m); m_t =moisture content (g water/g dry solid); m₀= initial moisture content (g water/g dry solid); m_e= equilibrium moisture content (g water/g dry solid); t= frying time (s) and D_eff effective moisture diffusion coefficient (m²/s). A constant value during the frying process was assumed. It is reasonable to assume that moisture content is negligible when equilibrium is reached in the frying process, so m_e= 0 in equation (1).[14,20,25–27]

The second model considers that physical properties of the potato chip vary with changes in the temperature and moisture content during the frying process; in this way, the effective moisture diffusion coefficient is no longer a constant value but rather a function of time. Simplified methods, such as the method of slopes or numerical methods, have been used in order to obtain values of variable effective diffusivities from air drying curves.[28] In this work, an extension of the analytical model of variable diffusivities proposed for spherical geometry to study air drying of grapes,[29] and water loss during frying of potato strips[10] was used. In the case of slab geometry, this model results in the following equation:

(2)

where D_eff is postulated as a function of time as:

(3)

in which D₀ is the effective diffusivity at time = 0, and b is a dimensionless parameter.

Since the diameter is considerable larger than the width in potato slices, it was assumed that the mass transfer takes place only in the axial direction (unidirectional).

The parameters of Eqs. (1) and (2) were obtained by non-linear regressions, using as objective function by the minimization of the relative deviations.

(4)

where V_obs is the observed or experimental value, V_pred is the predicted values, N the number of data points, and RMS is the root mean square error.

Statistical analysis

Analyses of variance were carried out when required using Statgraphic Statistical Package (Statistical Graphics Corporation, Version 4, Rockville, USA), which included multiple range tests (P>0.05) for separation of least square means.

RESULTS AND DISCUSSION

Table 1 shows the final oil content for control and pretreated slices with final moisture content of ∼1.8 % (wet basis). For all the frying temperatures, the blanched-dried slices present the lowest oil content (P>0.05): for temperatures of 120 and 150°C, it was found that a decrease of around 20% with respect to control slices, while at 180°C the decrease was only 6%. On the other hand, oil content in blanched slices was always higher than control slices. Predrying decreases the initial moisture content of potato pieces and is an efficient way to reduce fat uptake in the final fried product.[8,10,12] Contrary to expectation, blanched slices absorbed more oil than control slices; blanching has been reported as a pretreatment that could improve the color and texture of the chips and reduce their oil uptake.[7] In agreement with our results, other research has found that blanching for high temperatures and short times (e.g. 97°C, 2 min) before frying of potato strips resulted in higher oil content than in control strips. Consumers find this result unacceptable in the product.[30] The above blanching conditions are close to that used in this work (85°C for 3.5 min). On the other hand, some authors reported that low temperature blanching (e.g. 55-70°C) before frying activates pectinesterase enzyme (PME), and the resulting reactions decrease porosity and, hence, reduce oil uptake.[31] Not only the frying temperature but also the pretreatment have a significant effect (P>0.05) on the final oil content of the potato chips.In the three pretreatments studied, the lower the frying temperature the higher the oil absorption.

Table 1 Final oil content in potato slices (g/g dried solid)*.

	Frying temperature
	120°Χ	150°Χ	180°Χ
Control slices	0.3938 ± 0.010	0.3653 ± 0.016	0.3026 ± 0.006
Blanched slices	0.4148 ± 0.014	0.3841 ± 0.015	0.3642 ± 0.014
Blanched & dried slices	0.3118 ± 0.010	0.3013 ± 0.010	0.2849 ± 0.016
*Mean value ±standard deviation.

At the frying temperature of 180°C, the fit of moisture content with time data in Eqs. (1) and (2) are shown in Fig. 1 for control slices. The variable diffusivity model adjusted the experimental data better than the constant diffusivity model did (%RMS < 10% as an average). Tables 2 and 3 show the parameters obtained for both models. As can be seen, the values obtained for D_eff are similar to those reported during frying of potato slices,[20] discs of dough made from a biscuit mix,[14] and potato strips;[10] however, in all cases the RMSs for constant diffusity model were very high, indicating that this model is inadequate to represent the water loss of potato chips during frying. The effect on water loss in a thin and changing potato structure will be better represented by a diffusivity coefficient whose value also changes with time. Activation energies for constant diffusion coefficients calculated from data, shown in Fig. 2, were 4.5, 4.7, and 6 kJ/mol for control, blanched, and blanched and dried potato chips, respectively.

Figure 1 Observed and predicted water loss of control slices during frying at 180°Χ.

Table 2 Effective diffusivity coefficient and RMS values for constant diffusivity model.

		Temperature
		120°Χ	150°Χ	180°Χ
Control	Deff (m^2/s)	3.86 ± 0.05 × 10^−9	9.86 ± 0.57 × 10^−9	1.72 ± 0.02 × 10^−8
	RMS (%)	46.07	34.82	38.28
Blanched	Deff (m^2/s)	3.55 ± 0.06 × 10^−9	9.76 ± 0.47 × 10^−9	1.71 ± 0.01 × 10^−8
	RMS (%)	21.65	35.80	28.82
Blanched and dried	Deff (m^2/s)	4.53 ± 0.06 × 10^−9	1.28 ± 0.07 × 10^−8	3.48 ± 0.05 × 10^−8
	RMS (%)	18.89	35.02	41.94

Table 3 Parameters and RMS values for variable diffusivity model.

		Temperature
		120°Χ	150°Χ	180°Χ
Control	D0 (m^2/s)	1.29 ± 0.20 × 10^−10	4.17 ± 1.13 × 10^−9	7.14 ± 1.06 × 10^−9
	b (−)	4.51 ± 1.07	3.16 ± 1.35	3.07 ± 0.83
	RMS (%)	8.96	9.36	7.83
Blanched	D0 (m^2/s)	2.05 ± 0.67 × 10^−10	3.89 ± 0.67 × 10^−10	7.77 ± 2.42 × 10^−9
	b (−)	1.82 ± 1.28	3.33 ± 0.88	2.75 ± 1.56
	RMS (%)	12.70	8.58	4.60
Blanched and dried	D0 (m^2/s)	2.67 ± 0.54 × 10^−10	5.55 ± 0.46 × 10^−10	1.13 ± 0.10 × 10^−9
	b (−)	1.66 ± 0.82	3.05 ± 0.64	4.69 ± 0.60
	RMS (%)	7.62	11.52	9.76

Figure 2 Arrhenius constant diffusivity plot for potato chips.

Experimental data and the fit to variable diffusivity model of water loss vs. frying time for control slices and blanched-dried slices are shown in Fig. 3A and Fig. 3B, respectively. Water loss increases with the frying temperature and time for both cases as expected. However, water loss is faster for each of the three frying temperatures in the case of blanched-dried slices (Fig. 3B). The variable diffusivity model fitted the experimental data properly since the average %RMS value for all the studied cases (3 frying temperatures and three pretreatments) was less than around 10% (Table 3). The value for D₀ increased with the frying temperature for the 3 cases studied. It seems reasonable that the effect on water loss in a changing potato microstructure will be better represented by a diffusivity coefficient whose value changes with frying time. Fig. 4 shows that, at a given temperature, D_eff increases as the sample further loses moisture and also gains oil, probably because of the development of a more porous microstructure with frying time. The effect of the increase in temperature of water vapor during frying is higher than the resistant effect of the formed crust. As a result, D_eff increases as drying progresses during frying. This fact suggests that the water leaves the potato tissue easily without a pronounced resistance of the external crust progressively formed that could eventually make D_eff diminish with time, as was found earlier in potato strips fried at 180°C.[10]

Figure 3 Observed and predicted water loss during the frying of (A) control slices; (B) blanched and dried slices. Circles, triangles, and squares represent experimental data for the frying temperatures of 120°C, 150°C, and 180°C, respectively. Solid lines represent the predicted values by the variable diffusivity model (Eq. (2)).

Figure 4 Effective moisture diffusivity during frying of control potato slices at 120°C, 150°C, and 180°C (Eq. (3)).

For the same frying temperature (e.g., 150°C), D_eff behaves very similar in time for both control and blanched slices (Fig. 5). This fact was repeated also at 120°C and 180°C (results not shown). However, the increase in D_eff with frying time at the same frying temperature (e.g., 150°C) for blanched-dried slices was considerably higher than for control and blanched slices. Blanched-dried potatoes have around 20% less oil. The effect of oil content on the moisture diffusion coefficient during air drying has been studied for some biomaterials such as avocado and oil-based model systems,[33] and pepperoni.[34] In all cases, it was found that the moisture diffusion decreases as the oil increases. This fact was attributed to the hydrophobic character of fat that imposes a resistance to the flow of water. However, during frying, most of the oil (70-80%) is absorbed at the end when the potato piece is removed from the fryer, so the flow resistance imposed by the oil to the flow of water during frying is negligible.[35,36,37]

Figure 5 Effective moisture diffusivity during frying at 150°C of potato slices with different treatments (Eq. (3)).

D_eff increases as the potato slices dry during frying (Fig. 6). As the moisture content diminishes in potato slices, a more porous microstucture is formed. This makes the diffussion water easier. For the same moisture content, the higher the frying temperature the higher the D _eff . When potato slices were fried at 150°C, D_eff increases as the moisture content decreases during frying, independent of the pretreatment received by the potato slices (Fig. 7). The same trend was observed for the frying temperatures of 120°C and 180°C (results not shown). D_eff depends principally on the moisture content of the potato slices during frying.

Figure 6 Effective moisture diffusivity vs. moisture content of control potato slices fried at 120°C, 150°C, and 180°C.

Figure 7 Effective moisture diffusivity vs. moisture content during frying at 150°C of potato slices with different treatments.

CONCLUSION

Blanched and dried potato slices absorbed less oil than control and blanched slices after frying. As the frying temperature increased, the potato slices absorbed less oil. The variable diffusivity model fit the experimental moisture data better than the constant diffusivity model. Effective moisture diffusivity increased with the frying time and temperature, and depended principally on the moisture content of the potato slices during frying. In the case of potato slices, the effect of an increase in temperature in the food during frying (which increases D_eff) was more important than the increase in crust resistance (which decreases D_eff).

ACKNOWLEDGMENTS

The authors acknowledge financial support from FONDECYT Project N° 1030411.