Professional Documents
Culture Documents
G. Lee, F. Hsieh
ABSTRACT. The effects of air temperature and sample thickness on drying kinetics of strawberry fruit leather were
investigated. The mathematical modeling was performed by using three thin‐layer drying models. The independent variables
were sample thickness (1.8, 2.7, and 3.6 mm) and air temperature (50°C, 60°C, 70°C, and 80°C). Drying took place in the
falling rate period. The values of effective moisture diffusivity (Deff ) varied from 2.40 × 10-9 to 12.1 × 10-9 m2 s-1 depending
on drying conditions. The values of activation energy (Ea ) were 35.57, 33.14, and 30.46 kJ mol -1 for 1.8, 2.7, and 3.6 mm
sample thickness, respectively. The two‐term exponential model was found to satisfactorily describe the thin‐layer drying
kinetics of strawberry leather.
Keywords. Activation energy, Diffusivity, Drying, Mathematical modeling, Strawberry leather.
F
ruit leather is a confectionery product manufactured drying is one of the most common unit operations in the food
by dehydrating fruit puree into leathery sheets. It is processing industry. It relies on heating the product with hot
made by drying thin layers of fruit puree in the oven air. The drying behavior is greatly affected by air temperature
or dehydrator to produce a product with the texture and material characteristics. Understanding the drying
of soft leather. These leathery sheets, sometimes referred to kinetics of fruit leather is very important for control of the
as fruit rolls or taffies, are nutritious snacks for backpackers, drying process and quality of the final product. Many studies
campers, and children. Many kinds of fruits, or combinations have been carried out to investigate the effect of some process
of fruits, can be used, such as apricots, apples, grapes, berries, parameters such as drying air temperature and sample
pineapple, oranges, pears, peaches, plums, tomatoes, tropical thickness for various foods (Madamba et al., 1996; Youcef‐
fruits, and others. Fruit leathers especially add value to fruits Ali et al., 2001; Doymaz, 2004), and a few have been reported
that may otherwise not be acceptable for the fresh produce for some fruit leathers (Chan and Cavaletto 1978; Cheman et
market. Strawberries as a typical soft fruit have a high al., 1997; Maskan et al., 2002), but none have been reported
physiological post‐harvest activity with short ripening and for strawberry fruit leather. The objectives of this study were
senescence periods that make marketing of this high‐quality to investigate the effect of process parameters such as drying
fruit a challenge. Thus, one of the most promising methods air temperature and sample thickness on the drying kinetics
for preservation would be to manufacture strawberry fruit of strawberry leather and to find suitable mathematical
leather by drying. models for the drying curves.
Drying is the most commonly employed commercial
technique in the food processing industry. It involves the
vaporization of moisture within the product by heat and its
subsequent removal from the product (Ekechukwu, 1999).
MATERIALS AND METHODS
MATERIALS
Thus, a drying process involves simultaneous heat and mass
Strawberry natural puree was obtained from Sensient
transfer. Many types of dryers are available to the food
Flavors, Inc. (Amboy, Ill.) and pectin (SS200, high methoxy
industry. Drying methods are usually classified into four
pectin of 63% to 67% degree of esterification) was obtained
categories, depending on the heat transfer mechanism
from Danisco USA, Inc. (St. Louis, Mo.). Corn syrup
involved, and they are convection (Azzouz et al., 2002;
(42D.E.) and anhydrous citric acid were purchased from
Dandamrongrak et al., 2002; Krokida et al., 2003),
ADM (Decatur, Ill.).
conduction (Fudym et al., 2003), radiation (Abe and Afzal,
1997; Fu and Lien, 1998; Afzal and Abe, 2000), and
microwave heating (Adu and Otten, 1996). Convection EXPERIMENTAL PROCEDURE
Strawberry fruit leathers were prepared by blending
strawberry puree, corn syrup, pectin, and citric acid in
200:40:2:1 ratios, spreading into thin layers on an aluminum
Submitted for review in October 2007 as manuscript number FPE 7220; weighing dish of 70 mm diameter (Fischer Scientific,
approved for publication by the Food & Process Engineering Institute Pittsburgh, Pa.), and drying in a convection oven
Division of ASABE in August 2008.
The authors are Gwi‐Hyun Lee, ASABE Member Engineer, (OV116040, M&L Testing Equipment, Inc., Dundas,
Professor, Division of Agricultural Engineering, Kangwon National Ontario, Canada). The independent variables were product
University, Chuncheon, Korea; and Fu‐Hung Hsieh, ASABE Member thickness (1.8, 2.7, and 3.6 mm) and drying temperature
Engineer, Professor, Department of Biological Engineering, University of (50°C, 60°C, 70°C, and 80°C). Moisture losses were
Missouri, Columbia, Missouri. Corresponding author: Fu‐Hung Hsieh,
Department of Biological Engineering, University of Missouri, Columbia,
recorded by an electronic balance (APX‐1502, Denver
MO 65211; phone: 573‐882‐2444; fax: 573‐882‐1115; e‐mail: hsiehf@ Instrument, Denver, Colo.) with a sensitivity of 0.01 g at
missouri.edu. constant time intervals during drying for the determination of
where A1, A2, k1, and k2 are empirical coefficients. where MRexp,i is the experimental moisture ratio at the ith
A simple model, analogous to Newton's law of cooling, observation, MRcal,i is the calculated moisture ratio at this
also called the Lewis model (Lewis, 1921), is often used in observation, N is the number of observations, and n is the
drying analysis to describe the falling rate period during number of constants. Lower values of x2 and higher values of
thin‐layer drying of products (Temple and van Boxtel, 1999; r2 were chosen as the criteria for best fit. In this study, the
Phoungchandang and Woods, 2000; Panchariya et al., 2002). relationships among the drying air temperature, sample
It assumes that the rate of moisture removal is proportional thickness, and the coefficient of the tested drying models
to the difference between the average moisture content of the were determined.
product being dried and its equilibrium moisture content. It
is mathematically expressed as:
dM RESULTS AND DISCUSSION
= −k (M − M e ) (3) INFLUENCE OF PROCESS VARIABLES
dt
The effects of drying air temperature and sample thickness
The solution of this equation after integrating can be on drying rate of strawberry fruit leather were investigated by
written as: neglecting shrinkage of the product. The drying data
0
0.0 0 20 40 60 80 100
0 100 200 300 400 500 600 Moisture Content (% d.b.)
Drying Time (min)
Figure 1. Drying curves at different drying temperatures for strawberry Figure 3. Drying rate curves at different drying temperatures for
fruit leather of 2.7 mm thickness. strawberry fruit leather of 2.7 mm thickness.
1.0 150
0.8
120
2.7 mm 90 2.7 mm
0.6 3.6 mm
3.6 mm
0.4 60
0.2 30
0.0 0
0 100 200 300 400 500 600 0 20 40 60 80 100
Drying Time (min) Moisture Content (% d.b.)
Figure 2. Drying curves for different sample thicknesses of strawberry Figure 4. Drying rate curves for different sample thicknesses of
fruit leather at 60°C. strawberry fruit leather at 60°C.
involving percentage moisture content (d.b.) were were reduced by about 55.6%, 66.7%, and 70% for sample
transformed to the dimensionless parameter of moisture ratio thicknesses of 1.8, 2.7, and 3.6 mm, respectively, as drying
(MR) and plotted versus drying time (t). Figure 1 shows the temperature was raised from 50°C to 80°C. Similarly,
typical characteristic drying curves for strawberry fruit decreases of 55.6%, 61.5%, 66.7%, and 70.0% in drying time
leather of 2.7 mm thickness during thin‐layer drying process were observed at drying temperatures of 80°C, 70°C, 60°C,
at 50°C, 60°C, 70°C, and 80°C. Increasing the drying and 50°C, respectively, as sample thickness decreased from
temperature significantly increased the drying rate and 3.6 to 1.8 mm. A thinner sample would reduce drying time
reduced the drying time. Similar results were obtained for due to reduced distance of moisture movement and the
sample thicknesses of 1.8 and 3.6 mm. Figure 2 presents the increased surface area exposed for a given volume of
effect of sample thickness on drying rate at 60°C. The product. The pronounced effects of temperature and sample
decrease in sample thickness at constant drying temperature thickness on drying time were similar to previous studies on
significantly reduced the drying time needed to reach garlic slices (Madamba et al., 1996), chicory root slices (Lee
equilibrium moisture content. Similar trends were observed et al., 2004), grape leather (Maskan et al., 2002), and carrot
at drying temperatures of 50°C, 70°C, and 80°C. Based on cubes (Doymaz, 2004).
these results, drying times for strawberry leather samples to
reach the moisture content of 12% (w.b.) for safe storage were CALCULATION OF EFFECTIVE MOISTURE DIFFUSIVITY AND
found to vary from 80 to 600 min depending on drying ACTIVATION ENERGY
temperature and sample thickness (table 2). About 55.6%, The drying rate (dM/dt) was calculated from the slopes of
66.7%, and 70% of drying times were reduced for each the moisture content against drying time curves. Figure 3,
sample thickness of 1.8, 2.7, and 3.6 mm as drying showing the variation in drying rate versus moisture content,
temperature was raised from 50°C to 80°C. Drying times presents the drying rate curves at different temperatures. It
was observed that the initial drying rate was high and was
Table 2. Drying times for samples of strawberry fruit followed by a gradual decrease as the product approached the
leather to reach the moisture content of 12% (w.b.).
dried state. As the thickness of sample decreased from 3.6 to
Drying Time (min)
Sample 1.8 mm, drying rates increased, with the drying rate curves
Thickness (mm) 50°C 60°C 70°C 80°C shifted upward for each drying air temperature (fig. 4).
1.8 180 120 100 80 Similarly, a higher drying rate was found at higher
2.7 360 240 180 120 temperature. In this study, the drying of strawberry fruit
3.6 600 360 260 180 leather occurred almost entirely in the falling rate period.
ln MR
being the moisture movement mechanism because of the - 2.0
decrease in moisture available for evaporation with time - 2.5
from the drying surface. The results of this study indicate that
- 3.0
diffusion is the most likely physical mechanism governing
moisture movement in strawberry fruit leather during thin‐ - 3.5
ln MR
A more general form of equation 7 is expressed as follows: - 2.0
M − Me
ln MR = ln = A − Bt (8) - 2.5
Mo − Me - 3.0
where the constant B is related to the effective diffusivity: - 3.5
Deff π 2
- 4.0
B= (9) 0 100 200 300 400 500
4L2 Drying Time (min)
The linear relationships were obtained by plotting the Figure 6. Semi logarithmic plots of drying curves for different sample
moisture ratio (MR) and drying time (t) on a semi‐ thicknesses of strawberry fruit leather at 60°C.
logarithmic scale (fig. 5) at different drying temperatures for
strawberry fruit leather of 2.7 mm thickness in accordance energy (kJ kg-1), R is the gas constant (kJ kmol-1·K-1), and T
with equation 8. Figure 6 shows the drying rate curves plotted is the temperature of air (K). The activation energy for diffusion
for different sample thicknesses at 60°C. In this study, was calculated from the slope of a straight line by plotting
straight lines were satisfactorily fitted to the experimental ln(Deff) versus the reciprocal of the air temperature, 1/T
data, with r2 > 0.97 at all drying temperatures and sample (Madamba et al., 1996; Maskan et al., 2002; Panchariya et al.,
thicknesses investigated. The effective diffusivity was 2002; Doymaz and Pala, 2003). Figure 7 presents the linear
calculated by equation 9 using slopes derived from the linear relationships between ln(Deff) and 1/T in the range of
regression of ln(MR) versus time (t) for different drying temperatures investigated for each sample thickness, indicating
conditions. The values of Deff estimated for different sample Arrhenius dependence. The activation energy calculated from
thicknesses and temperatures are presented in table 3. Both the slope of the straight lines in figure 7 was found to be 35.57
drying air temperature and sample thickness had a (r2 = 0.98), 33.14 (r2 = 0.98), and 30.46 (r2= 1.0) kJ mol-1 for
pronounced influence on the drying rate and, as a result, strawberry fruit leather samples of 1.8, 2.7, and 3.6 mm
greatly affected the value of Deff. The values of Deff for thickness, respectively. The value of activation energy increased
strawberry leather ranged from 2.40 × 10-9 to 12.10 × 10-9 with decreasing sample thickness. It showed the same trend as
m2 s-1 depending on drying conditions. The higher value of the results of Maskan et al. (2002), who reported Ea values of
Deff was obtained at the highest drying air temperatures with 10.3 to 21.7 kJ mol-1 depending on the sample thickness. This
the thickest samples. These values are comparable to result suggests that the thinner samples had significant changes
1.533× 10-9 to 2.885 × 10-9 m2 s-1 for air drying of paprika in diffusivity value compared to the thicker samples due to a
at 60°C (Ramesh et al., 2001), 2.25 × 10-8 to 2.74 × 10-8 m2 smaller variation in temperature during drying. The values of
s-1 for red pepper at 60°C (Doymaz and Pala, 2003), and activation energy in this study were within the general range of
0.776 × 10-9 to 9.335 × 10-9 m2 s-1 for carrot cubes in 15 to 40 kJ mol-1 for various food materials (Rizvi, 1986).
temperature range of 50°C to 70°C (Doymaz, 2004).
The effect of temperature on effective diffusivity is Table 3. Values of effective diffusivity (Deff) for drying
generally described by an Arrhenius‐type equation: of strawberry fruit leathers (r2 values in parentheses).
Effective Diffusivity (× 10‐9)
Sample
⎛ E ⎞ (m2 s‐1)
Deff = Do exp ⎢ − a ⎟ (10) Thickness
⎝ RT ⎠ (mm) 50°C 60°C 70°C 80°C
1.8 2.40 (0.94) 3.09 (0.94) 4.68 (0.95) 7.32 (0.97)
where Do is a constant equivalent to the effective diffusivity 2.7 3.77 (0.96) 5.36 (0.96) 6.91 (0.97) 11.10 (0.99)
at infinite high temperature (m2 s-1), Ea is the activation 3.6 4.57 (0.97) 6.24 (0.96) 8.38 (0.97) 12.10 (0.98)
Table 4. Results of non‐linear regression obtained from different thin‐layer drying models.
Sample
T Thickness
Model (°C) (mm) Model Constants r2 χ2
Lewis 50 1.8 k = 0.0137 0.9558 2.8896 × 10‐3
model 2.7 k = 0.0080 0.9534 2.9731 × 10‐3
3.6 k = 0.0049 0.9540 2.6914 × 10‐3
60 1.8 k = 0.0197 0.9701 2.2066 × 10‐3
2.7 k = 0.0110 0.9706 2.1767 × 10‐3
3.6 k = 0.0068 0.9714 1.9158 × 10‐3
70 1.8 k = 0.0238 0.9757 1.7097 × 10‐3
2.7 k = 0.0141 0.9778 1.5459 × 10‐3
3.6 k = 0.0089 0.9782 1.4425 × 10‐3
80 1.8 k = 0.0328 0.9935 7.9269 × 10‐4
2.7 k = 0.0191 0.9935 5.4204 × 10‐4
3.6 k = 0.0120 0.9903 7.3472 × 10‐4
Page 50 1.8 ko = 0.0929, m = 0.5785 0.9863 1.4004 × 10‐3
model 2.7 ko = 0.0494, m = 0.6315 0.9917 7.4201 × 10‐4
3.6 ko = 0.0326, m = 0.6555 0.9939 4.2218 × 10‐4
60 1.8 ko = 0.1212, m = 0.5705 0.9893 1.4735 × 10‐3
2.7 ko = 0.0563, m = 0.6508 0.9912 8.6502 × 10‐4
3.6 ko = 0.0356, m = 0.6840 0.9942 6.4287 × 10‐4
70 1.8 ko = 0.1003, m = 0.6443 0.9924 8.6003 × 10‐4
2.7 ko = 0.0577, m = 0.6836 0.9928 6.9469 × 10‐4
3.6 ko = 0.0376, m = 0.7112 0.9951 4.4546 × 10‐4
80 1.8 ko = 0.0918, m = 0.7287 0.9941 5.6536 × 10‐4
2.7 ko = 0.0470, m = 0.7856 0.9946 4.1130 × 10‐4
3.6 ko = 0.0367, m = 0.7642 0.9961 3.7068 × 10‐4