Vacuum drying characteristics of eggplants (Long Wua, Takahiro Orikasa).pdf

Journal of Food Engineering 83 (2007) 422–429

www.elsevier.com/locate/jfoodeng

Vacuum drying characteristics of eggplants

Long Wu a , Takahiro Orikasa a , Yukiharu Ogawa b , Akio Tagawa a, *

a Graduate School of Science and Technology, Chiba University, 648, Matsudo, Matsudo, Chiba 271-8510, Japan

b Faculty of Horticulture, Chiba University, 648, Matsudo, Matsudo, Chiba 271-8510, Japan

Received 7 February 2007; received in revised form 15 March 2007; accepted 17 March 2007

Available online 28 March 2007

Abstract

The vacuum drying characteristics of eggplant were investigated. Drying experiments were carried out at vacuum chamber pressures

of 2.5, 5 and 10 kPa, and drying temperature ranging from 30 to 50 C. The eﬀects of drying pressure and temperature on the drying rate

and drying shrinkage of the eggplant samples were evaluated. The suitable model for describing the vacuum drying process was chosen

by ﬁtting four commonly used drying models and a suggested polynomial model to the experimental data; the eﬀective moisture diﬀu-

sivity and activation energy were calculated using an inﬁnite series solution of Fick’s diﬀusion equation. The results showed that increas-

ing drying temperature accelerated the vacuum drying process, while drying chamber pressure did not show signiﬁcant eﬀect on the

drying process within the temperature range investigated. Drying shrinkage of the samples was observed to be independent of drying

temperature, but increased notably with an increase in drying chamber pressure. A linear relationship between drying shrinkage ratio

and dry basis moisture content was observed. The goodness of ﬁt tests indicated that the proposed polynomial model gave the best

ﬁt to experimental results among the ﬁve tested drying models. The temperature dependence of the eﬀective moisture diﬀusivity for

the vacuum drying of the eggplant samples was satisfactorily described by an Arrhenius-type relationship.

Keywords: Vacuum drying; Eggplant; Drying characteristics; Drying model

1. Introduction

higher drying rate, lower drying temperature and oxygen

deﬁcient processing environment etc., these characteristics

may help to improve the quality and nutritive value of

the dried products. Presently, vacuum drying has been

applied to dry various food materials, the vacuum drying

kinetics of many fruits and vegetables has been investigated

and the eﬀect of vacuum drying conditions on the drying

process and the qualities of dried products has been evalu-

ated ( Arevalo-Pinedo & Murr, 2006; Arevalo-Pinedo &

Murr, 2007; Bazyma et al., 2006; Cui, Xu, & Sun, 2004;

Jaya & Das, 2003; Methakhup, Chiewchan, & Devahastin,

2005 ).

Eggplant (Solanum melongena var. esculenta)isan

important market vegetable of Asian and Mediterranean

countries and has a very limited shelf life for freshness. In

order to evaluate the practicability of vacuum drying for

improving the quality of dried eggplant, it is necessary to

carry out research on the vacuum drying characteristics

of eggplant fruit. The objectives of this study were to

Drying is one of the most important methods of long-

term food preservation. The removal of moisture from

the food materials prevents the growth and reproduction

of spoilage microorganisms, slows down the action of

enzymes and minimizes many of the moisture mediated

deteriorative reactions. Although drying processing eﬀec-

tively extends the shelf life of agricultural products, loss

of sensory and nutritive qualities is considered inevitable

during traditional drying process due to the undesirable

textural and biochemical changes ( Watson & Harper,

1988 ).

Compared with conventional atmospheric drying, vac-

uum drying has some distinctive characteristics such as

* Corresponding author. Tel./fax: +81 47 308 8847.

E-mail address: tagawa@faculty.chiba-u.jp (A. Tagawa).

doi:10.1016/j.jfoodeng.2007.03.030

L. Wu et al. / Journal of Food Engineering 83 (2007) 422–429

423

Nomenclature

A area (m 2 )

D eﬀ eﬀective moisture diﬀusivity (m 2 s 1 )

E a activation energy (kJ kg 1 )

L sample thickness (m)

M moisture content (d.b., decimal)

M 0 initial moisture content (d.b., decimal)

M e equilibrium moisture content (d.b., decimal)

MR moisture ratio

P pressure (kPa)

P 0 mean relative deviation (%)

R gas constant (0.462 kJ kg 1 K 1 )

R 2 coecient of determination

R d drying rate (kg m 2 h 1 )

RMSE root mean square error

t time (h)

T temperature (K)

V volume (m 3 )

W d weight of dry matter (kg)

a,b coecients in drying model

time (s)

v 2

reduced chi-square

investigate the vacuum drying characteristics of the egg-

plant samples, to evaluate the eﬀect of vacuum drying

conditions on the drying process, and to choose a suit-

able drying model for describing the whole drying

process.

2.3. Experimental procedure

The vacuum drying chamber was preheated for 12 h

before the experiments started to obtain stable drying tem-

perature. Drying experiments were conducted in the drying

chamber at temperatures ranging from 30 to 50 C, and

pressures of 2.5, 5, 10 kPa as well as atmospheric pressure,

respectively. One sample was placed on the wire netting

basket and dried in each run, its weight was continuously

recorded at intervals of 5 min using the data acquisition

system throughout the drying process. It was considered

that the sample reached the equilibrium moisture content

(EMC) of drying when the reading of weight remained

the same for 1 h.

Fresh samples were dried under the above-mentioned

conditions for durations ranging from 1 to 15 h individu-

ally to evaluate the drying shrinkage. Approximate volume

and surface area of the dried samples were calculated from

the measured dimensions data. According to preliminary

tests, in which a comparison between the calculated results

and the measured volume of the dried samples using liquid

displacement method ( Maskan, 2001; Orikasa, Tagawa,

Nakamura, & Iimoto, 2005; Zogzas, Maroulis, & Mari-

nos-Kouris, 1994 ) was conducted, the calculated results

were proved to be acceptable, similar results were reported

by Ratti (1994) .

2. Material and methods

2.1. Sample preparation

Fresh eggplants (cultivated in Kochi Prefecture Japan,

cultivar: unknown) were purchased from a local market

and stored at 8 C before experiments started, the storage

time was not more than 12 h in this study. The central part

of each eggplant fruit was cut into a rectangular-shaped

block of 45 25 20 mm for drying treatment. The initial

moisture content of the sample blocks was determined as

94.00% in wet basis (N = 20, standard deviation: 0.52%)

according to the vacuum oven method (i.e., drying at

70 C and 2.5 kPa for 12 h) ( AOAC, 1995 ).

2.2. Experimental setup

A schematic diagram of the experimental vacuum dry-

ing system is shown in Fig. 1 . The system primarily con-

sists of an oil rotary vacuum pump (TSW-300, SATO

VAC, Japan), a vacuum control unit (NVC-2000L, Tokyo

Rikakikai, Japan) to obtain various processing pressures

in the vacuum drying chamber (a glass desiccator) and

a forced convection drying oven (DO600FA, AS ONE,

Japan) to maintain desired drying temperatures. A data

acquisition system composed of a load cell (LTS-50GA,

KYOWA, Japan) which was ﬁxed on a supporting frame,

a wire netting sample holder suspended from the load cell,

an instrumentation ampliﬁer (WGA-710A, KYOWA,

Japan) and a data logger (KEYENCE, NR-1000, Japan)

was used to on-line monitor and record the changes in

sample weight during drying. Hot air drying runs at 30–

50 C and atmospheric pressure were also conducted in

the same glass desiccator with the top lid removed for

comparison.

2.4. Data analysis

The average moisture content of each sample during

drying was calculated from the sample weight recorded

by the data acquisition system (moisture distribution in

the sample was considered to be uniform in this study).

Moisture ratio (MR) of the sample was determined by

the following equation:

MR ¼ ð M M e Þ

ð M 0 M e Þ

ð 1 Þ

where M is the average moisture content of a sample at any

time of drying, M 0 and M e stand for the initial and equilib-

rium moisture content, respectively.

424

L. Wu et al. / Journal of Food Engineering 83 (2007) 422–429

Fig. 1. Schematic diagram of vacuum drying system: 1. vacuum pump, 2. cold trap, 3. vacuum control unit, 4. forced convection drying oven, 5. glass

desiccator, 6. load cell, 7. supporting frame, 8. wire netting sample holder, 9. instrumentation ampliﬁer, and 10. data logger.

Drying curves (MR vs. time) were plotted and ﬁtted by

four empirical drying models (i.e., proposed polynomial

model, exponential model, Page’s model and logarithmic

model), and a theoretical model based on the Fick’s diﬀu-

sion law; model coecients were calculated using Origin-

Pro 7.5 software (OriginLab Corp.). The goodness of ﬁt

was evaluated by the coecient of determination (R 2 ),

the root mean square error (RMSE), the reduced chi-

square (v 2 ) and mean relative deviation modulus (P 0 )

deﬁned by the following equation

conditions were plotted. From Fig. 2 , the drying time

needed to reach the EMC was shortened notably with an

increase in drying temperature due to a larger driving force

for heat and mass transfer at higher drying temperature.

Fig. 3 showed that drying chamber pressure ranging from

2.5 to 10 kPa did not aﬀect the drying process as strongly

as drying temperature did. For the present vacuum drying

conditions, the eﬀect of drying chamber pressure on the dry-

ing process was not signiﬁcant. According to the reports of

Arevalo-Pinedo and Murr (2006) and Arevalo-Pinedo and

Murr (2007) for carrot and pumpkin, Cui et al. (2004) for

carrot, Giri and Prasad (2007) for mushroom, and Metha-

khup et al. (2005) for Indian gooseberry, drying pressure

had a certain eﬀect on the drying process, the drying time

was reduced by decreasing drying pressure. The diﬀerentia-

tion between the results of this study and the literature

could be attributed to the diﬀerent processing conditions

as well as diﬀerent degrees of boiling point elevation caused

by various plasma concentrations of the tested materials.

P 0 ¼ 100

N X

Y exp ; i Y pre ; i

ð 2 Þ

Y exp ; i

i ¼ 1

where Y exp,i is the experimental result of the investigated

variable, Y pre,i is the predicted value from various mathe-

matical models, N is the number of observations ( Chen &

Morey, 1989; Jena & Das, 2007; Madamba, Driscoll, &

Buckle, 1996; Sacilik & Elicin, 2006 ).

The best model describing the vacuum drying process of

the eggplant samples was chosen as the one with the high-

est R 2 and the least RMSE, v 2 and P 0 .

Comparisons between means were performed in SPSS

12.0 software (SPSS Inc.) using Duncan’s multiple range

tests at a signiﬁcance level of 0.05.

3.2. Drying shrinkage

Drying shrinkage aﬀects not only the product quality

but also the drying process and rehydration capability of

the dried food materials ( Karathanos, Anglea, & Karel,

1993; Maskan, 2001; Mcminn & Magee, 1997a, 1997b ).

Ratti (1994) indicated that changes in the dimensions of

dried sample were independent of drying conditions but

dependent on the geometric shape and type of foodstuﬀ.

Souma, Tagawa, and Iimoto (2004) reported that the hot

air drying shrinkage of eggplant is very remarkable and

the reduction in sample volume was larger than the volume

of removed water due to its high porosity, similar tendency

were also observed in this study, as shown in Fig. 4 .

In order to investigate the drying shrinkage of the sam-

ples, drying experiments with diﬀerent drying durations

3. Results and discussion

3.1. Eﬀect of vacuum drying conditions on drying process

After vacuum drying, the moisture content of the egg-

plant samples was reduced from a initial value of 15.67 to

less than 0.2 kg water/kg dry matter. The eﬀects of drying

temperature and pressure on the vacuum drying process

are shown in Figs. 2 and 3 a,b, in which drying curves (mois-

ture content in dry basis vs. time) under diﬀerent drying

L. Wu et al. / Journal of Food Engineering 83 (2007) 422–429

425

P = 2.5 kPa

1.0

50 ° C Hotair drying

50 ° C 2.5 kPa Vacuum drying

30 ° C

40 ° C

50 ° C

0.8

0.6

0.4

Drying time (h)

0.2

Fig. 2. Changes in average moisture content of samples during vacuum

drying at 2.5 kPa and 30–50 C.

0.0

0.2

0.4

0.6

0.8

were carried out under various drying conditions, the vol-

ume and surface area of the dried samples were calculated.

Fig. 5 presents the relationship between the volume shrink-

age ratio (V/V 0 ) and the moisture content of the samples

dried at 2.5 kPa and various drying temperatures, the

results indicated that drying temperature had insigniﬁcant

eﬀect on the drying shrinkage of the eggplant samples for

the investigated temperature range. Fig. 6 shows the eﬀect

of drying chamber pressure on the drying shrinkage of the

samples at 50 C, it can be seen that the shrinkage became

more severe at higher vacuum chamber pressures. This

phenomenon could be explained as follows: when water

is removed from the material during drying, a pressure

unbalance is generated between the interior of the dried

material and the external environment, and induces the

contracting stresses that lead to drying shrinkage. The dry-

ing shrinkage of eggplant is particularly severe because of

the collapse of the unconsolidated porous structure of egg-

(V 0 -V)/V 0

Fig. 4. Comparison between volumetric shrinkage of eggplant and volume

of removed water during drying at 50 C and atmospheric pressure.

plant tissue during drying. In contrast with atmospheric

drying, the pressure unbalance during vacuum drying is

substantially reduced due to the reduction in air pressure,

consequently the drying shrinkage could be inhibited.

Until now, many theoretical and empirical models for

describing drying shrinkage have been proposed ( Mayor

& Sereno, 2004 ). Among them, linear equation:

V 0 ¼ aM þ b

ð 3 Þ

where V is the volume of a sample at any time of drying

(m 3 ), M is the average moisture content of the sample at

P = 2.5 kPa

0.8

T = 30 ° C

0.6

2.5 kPa

5 kPa

10 kpa

30 ° C

40 ° C

50 ° C

0.4

0.2

Moisture content (d.b. decimal)

Fig. 5. Volume shrinkage ratio (V/V 0 ) as a function of moisture content

of samples at 2.5 kPa and diﬀerent drying temperatures.

Drying time (h)

T = 50 ° C

2.5 kPa

5 kPa

10 kPa

0.8

0.6

2.5 kPa

5 kPa

10 kPa

Atmospheric pressure

0.4

0.2

0.4

0.6

0.8

Drying time (h)

M/M 0

Fig. 3. Changes in average moisture content of samples during vacuum

drying at diﬀerent drying chamber pressures: (a) 30 C and (b) 50 C.

Fig. 6. Volume shrinkage ratio (V/V 0 ) as a function of moisture content

of samples at 50 C and diﬀerent drying chamber pressures.

426

L. Wu et al. / Journal of Food Engineering 83 (2007) 422–429

the same time, V 0 is the sample’s initial volume (2.25

10 5 m 3 (initial moisture content: 15.67) in this study),

has been successfully used for describing the drying shrink-

age of a wide range of foodstuﬀs under various drying con-

ditions ( Baini & Langrish, 2007; Lozano, Rotstein, &

Urbicain, 1980; Lozano, Rotstein, & Urbicain, 1983; Ratti,

1994; Suzuki, Kubota, Tsutomu, & Hosaka, 1976; Zogzas

et al., 1994 ).

Eq. (3) was ﬁtted to the experimental data under diﬀer-

ent drying conditions, goodness of ﬁt of the equation was

evaluated by R 2 and P 0 . The statistical results indicated that

under present conditions, linear model was adequate to

predict the drying shrinkage of the eggplant samples, the

R 2 of the linear regression reached about 0.99, and P 0

was less than 7%, as shown in Table 1 . The results also

proved that the drying shrinkage caused by vacuum drying

was obviously less than that caused by atmospheric drying

at the same drying temperature.

Theoretically, the surface area of the sample during dry-

ing can be predicted by the following equation ( Orikasa

et al., 2005; Pabis, 1999; Pabis & Jaros, 2002; Suzuki

et al., 1976 ):

0.006

30 ° C 2.5 kPa Experimental

0.005

0.004

0.003

y = 0.0001x + 0.0027

R 2 = 0.9802

0.002

0.001

P' = 8%

Moisture content (d.b. decimal)

Fig. 7. Variation in surface area with respect to moisture content of

samples during drying at 2.5 kPa and 30 C.

3.3. Rate of vacuum drying

According to Toei (1975) , drying rate (R d ) is deﬁned as

R d ¼ W d

ð 6 Þ

where R d is the drying rate (kg m 2 h 1 ), W d is the weight

of dry matter of the sample (kg), A is the drying area of the

sample (m 2 ), M is the volume-averaged moisture content, t

is the drying time (h).

Substituting Eq. (5) into Eq. (6) gives

A 0 ¼ k

3 l

¼ k ð aM þ b Þ 3 l

V 0

ð 4 Þ

where A is the surface area of a sample at any time (m 2 ), A 0

is initial surface area of the sample. Due to the complexity

of drying process of food materials, Eq. (4) lost its accuracy

in some cases. According to the results reported by

Nakamura, Tagawa, Orikasa, and Iimoto (2005) Orikasa,

Tagawa, Soma, Iimoto, and Ogawa (2005) , the relationship

between surface area and moisture content in dry basis of

the sample could be approximately expressed by a linear

equation:

A ¼ a 0 M þ b 0

R d ¼

a 0 M þ b 0 dM

W d

ð 7 Þ

ð 5 Þ

Eq. (5) was ﬁtted to the experimental data of surface area

and moisture content, the model coecients and the

indexes of goodness of ﬁt (R 2 and P 0 ) were calculated.

The results showed that the linear equation was adequate

to describe the changes in surface area with respect to the

moisture content of the samples during drying. The exper-

imental data at 30 C and 2.5 kPa and the results of linear

regression are shown in Fig. 7 .

The drying rate of the samples under various drying condi-

tions was calculated using Eq. (7) and plotted with respect

to the moisture content in dry basis. Fig. 8 shows the

changes in drying rate as a function of moisture content

at 2.5 kPa and various drying temperatures, similar trends

were observed at other drying chamber pressures. From the

ﬁgure, the drying temperature aﬀected the drying rate

signiﬁcantly, drying at higher temperature was apparently

faster than that at lower temperatures. The results also

indicated that the drying rates of the samples decreased

with decreasing moisture content throughout the drying

processes, that is to say, vacuum drying of the eggplant

samples under the investigated drying conditions took

place in the falling rate period only.

0.8

P= 2.5 kPa

Table 1

Results of linear regression for modelling drying shrinkage with respect to

moisture content

Drying condition

0.6

30 ºC 40 ºC

50 ºC

Linear equation

coecient

R 2

P 0 (%)

0.4

Vacuum drying (2.5 kPa, 50 C) a = 0.6124,

b = 0.3961

0.9864 3.3819

0.2

Vacuum drying (5 kPa, 50 C) a = 0.6773,

b = 0.3657

0.9929 3.0343

Vacuum drying (10 kPa, 50 C) a = 0.7238,

b = 0.307

0.9919 3.5381

Moisture content (d.b. decimal)

Hot air drying (atmospheric

pressure, 50 C)

a = 0.8438,

b = 0.2471

0.9897 6.0831

Fig. 8. variation in drying rate with respect to moisture content of samples

at 2.5 kPa and diﬀerent drying temperatures.

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: