Journal of Food Engineering 121 (2014) 2431

Contents lists available at ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier.com/locate/jfoodeng

Electrical impedance analysis of potato tissues during drying

Yasumasa Ando a,, Koichi Mizutani b, Naoto Wakatsuki b
a
Graduate School of Systems and Information Engineering, University of Tsukuba, Japan
b
Faculty of Engineering, Information and Systems, University of Tsukuba, Japan

a r t i c l e i n f o a b s t r a c t

Article history: We used electrical impedance spectroscopy (EIS) to explore the changes in the cell physiological status of
Received 3 April 2013 potato tissues during hot air drying at 5080 C. The measured impedance data were analyzed using the
Received in revised form 29 July 2013 modied Hayden model, which is an equivalent circuit model for cellular tissues. From the moisture con-
Accepted 5 August 2013
tent at the initial value to 1.0 (dry basis), focusing on the changes of equivalent parameters of the model,
Available online 14 August 2013
the cell membranes were apparently damaged by drying and heat stresses, and intracellular uid leaked
from the cells. At the moisture content less than 1.0, because of the destruction of cell structure, exper-
Keywords:
imental data could not be tted with the model, and due to the loss of moisture, the impedance magni-
Electrical impedance spectroscopy
Potato tissue
tude increased rapidly as drying proceeded. These results showed the behavior of the cell physiological
Drying status of potato tissues during drying and the potential of EIS as a method for evaluating injuries to bio-
Electrical equivalent circuit logical cells.
Biological cell 2013 Elsevier Ltd. All rights reserved.

1. Introduction
Drying is a common processing for vegetables. The basic objec-
tives when drying vegetables are to assure microbial stability and
extend the shelf-life of the product. Even with the development of
newer drying techniques, most vegetables are still air-dried be-
cause this method of dehydration remains the simplest and most
economical (Mazza, 1983). However, during hot air drying, vegeta-
bles are exposed to heat stress for a long time, causing changes of
color, nutrient content, avor, mechanical properties, etc. Among
these parameters, the physical properties are one of the most
important factors affecting the quality of dried vegetables. The
physical properties of vegetables are closely related to their cellu-
lar status.
For example, many studies have described the effects of cellular
turgor and cell membrane status on physical properties (Virgin,
1955; Falk et al., 1958; Nilsson et al., 1958; Hiller and Jeronimidis,
1996). Blahovec and Lahodov (2011, 2012, 2013) conducted dy-
namic mechanical analyses of potato tissues and found that the
changes of physiological properties at temperatures higher than
70 C were due to the increase in cellular turgor caused by starch
gelatinization. Laza et al. (2001) reported that the storage elastic
modulus and the toughness of potato tissues were decreased by
heating at 60 C due to degradation of the middle lamella, cell wall,
Corresponding author. Address: Graduate School of Systems and Information
Engineering, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8573,
Japan. Tel.: +81 29 853 5468.
E-mail address: ando@aclab.esys.tsukuba.ac.jp (Y. Ando).
and plasmalemma while the starch in the potatoes were partially
gelatinized.
As noted above, many studies have described changes in phys-
ical properties due to the changes in the cellular tissue status
caused by heat and dehydration stress. In the drying of vegetables
as well, physical parameters such as texture are important as a
quality of products. Changes in the texture of vegetables during
drying and rehydration processes have been reported (Lewicki
and Jakubczyk, 2004; Troncoso and Pedreschi, 2007; Cunningham
et al., 2008). These changes are likely to be due to the changes of
cell status caused by heat and drying stresses. Quantitative deter-
minations of cell status and cell damage during drying and rehy-
dration processes will contribute to our understanding of the
physical properties of dried vegetables.
Electrical impedance spectroscopy (EIS) measures the physical
state of materials as a function of frequency. Dielectric analysis
(DEA) is frequently confused with EIS, but although DEA is essen-
tially similar to EIS, DEA measurement are generally conducted
in high frequency areas (100 MHz10 GHz are generally used) to
estimate moisture content and bulk density, etc. (Nelson, 1994,
2005; Mckeown et al., 2012; Jha et al., 2011; Trabelsi and Nelson,
2006; Kandala et al., 1989). The frequency area in which the prop-
erties of cell structures appear is approx. 100 Hz10 MHz, and EIS
is often used in this frequency area.
In the present study therefore, we used EIS to evaluate the sta-
tus of cells. Since EIS is a fast and simple technique, it has been
widely used to estimate the physiological status of various biolog-
ical tissues (Ando et al., 2012; Damez et al., 2007; Yamamoto and
Yamamoto, 1976; Greenham, 1966). EIS generally makes use of
electrical equivalent circuits of materials to characterize the

0260-8774/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jfoodeng.2013.08.008
 Y. Ando et al. / Journal of Food Engineering 121 (2014) 2431
Nomenclature
Cm capacitance of cell membrane
j imaginary unit
M moisture content
p constant phase element exponent
R resistance (real part of the impedance)
b
R approximated values of R
R average value of R
Ri intracellular resistance
Re extracellular resistance
Rm resistance of cell membrane
R2R determination coefcient for R
R2X determination coefcient for X
SD standard deviation
experimental frequency response of impedance. Physical proper-
ties of the materials can be quantied by monitoring the changes
in parameters at the equivalent circuit. Among several equivalent
models proposed to represent biological tissues (Zhang and Willi-
son, 1991; Yamamoto and Yamamoto, 1977), we focus on the Hay-
den model proposed by Hayden et al. (1969). In this model, cellular
structures (i.e., the cell membrane and extra and intracellular uid)
are represented by parameters of the electrical equivalent circuit.
The Hayden model has widely been applied to EIS analyses of
various plant tissues, and its use has provided much useful physi-
ological information about matters such as ripening (Juansah and
Budiastra, 2012), cold injury (Cooley and Evert, 1979) and heat in-
jury (Zhang et al., 1993). Although many EIS studies of plant tissue
have focused attention on the natural physiological properties,
there are few reports about the behavior of electrical impedance
characteristics of vegetables during commonly used food process-
ing techniques such as drying. The use of EIS for investigations of
the cellular physiological status could provide valuable informa-
tion about the physical qualities of the dried vegetables.
Here we applied EIS to potato tissues during hot air drying un-
der several temperature conditions, and we examined the resulting
changes in cellular status. We also investigated the impedance
characteristics during the rehydration of dried samples to under-
stand how the moisture content inuences the impedance charac-
teristics. The objectives of this study were: (1) to analyze the
impedance characteristics of drying potato tissues in order to
clarify the behavior of cells in plant tissues, and (2) to evaluate
the effects of moisture content on the EIS results.
2. Materials and methods
2.1. Sample preparation
Since potato tubers are often used as a model plant tissue
because of their homogeneity, we chose potatoes to use as the
sample. Potatoes (Solanum tuberosum L.) of the variety of
Danshaku-imo were purchased from a local market and stored
at room temperature before the experiment. Danshaku-imo is a
mealy-type potato with a relatively high starch content. The aver-
age moisture content of the fresh potatoes was 4.30 on a dry basis
(g/g).
A schematic diagram of the experimental system used in this
study is shown in Fig. 1. Samples were placed in a thermostatic
chamber (SH-241, ESPEC, Osaka, Japan) and hot-air dried at 50 C
for 12 h, or 60 C, 70 C or 80 C for 10 h (57 samples for each tem-
perature). The relative humidity (RH) was kept under 10%. After
specied drying times, the sample was taken out from the chamber
25
T constant phase element coefcient
X reactance (imaginary part of the impedance)
b
X approximated values of X
X average value of X
Z impedance of sample
ZCPE impedance of constant phase element
h phase angle
x angular frequency
EIS electrical impedance spectroscopy
CPE constant phase element
CNLS complex non-linear least square
DEA dielectric analysis
and weighed. The decrease in the weight was taken as the amount
of water evaporated. Soaking experiments were conducted at 30 C
water temperature for the samples dried at each temperature.
Water absorption was performed by soaking a dried sample in a
beaker containing 200 mL of distilled water. The beaker was placed
in a thermostatically controlled water bath at the soaking temper-
ature. At predetermined soaking time intervals, the sample was
removed from the soaking water, blotted with lter papers to re-
move surface moisture, and weighed. The soaking experiment
was repeated from three to four times and measured data were
averaged. The moisture content M (dry basis (g/g)) of the samples
was determined gravimetrically by drying the samples in the ther-
mostatic chamber at 60 C for 12 h, and then milling and drying
them in an infrared drying moisture meter (EJ-610, A&D, Tokyo,
Japan) at 100 C for 30 min.
2.2. Impedance measurement
The impedance data of the samples were measured using an
impedance analyzer (HP4194A, Hewlett-Packard) with steel needle
electrodes (Fig. 2). The electrodes were connected to the imped-
ance analyzer with coaxial cables by means of the four-terminal
pair conguration (Agilent Technologies, 2013). The impedance
magnitude |Z| and phase angle h of the sample were measured at
predetermined times during drying and water absorption at 81
points (logarithmic frequency intervals) over the frequency range
from 100 Hz to 10 MHz and automatically recorded by a computer
for analysis. The impedance measurement during drying at each
drying temperature was replicated from ve to seven times, and
the soaking experiment was repeated three to four times for the
samples dried at each temperature.
Impedance analizer
PC
GP-IB
Drying chamber
Needle electrodes
Potato tissue sample
(253020 mm)
Fig. 1. The system for impedance measurement.
26 Y. Ando et al. / Journal of Food Engineering 121 (2014) 2431

Moisture content (dry basis (g/g))

4.5
4.0 50C
3.5 60C
70C
3.0
80C Early drying stage

11 mm
2.5
2.0
14 mm
1.5 Late drying stage
Fig. 2. Front and side views of needle electrodes. 1.0
0.5
0
0 200 400 600 800
Drying time (min)

Cm Rm Cm Fig. 4. Changes in moisture content during drying.

CPE

Re
CPE
Re Re Re
Ri Ri Ri In this study, the modied model was applied for the potato tissue
samples. The complex impedance of the modied model is repre-
sented by the equation below.
  
(a) (b) (c) (d) Re 1 xp T 2Ri Re  cosp2 p xp TRi Re Ri
Z  j
fxp TRe Ri g2 2xp TRe Ri  cos p2 p 1
Fig. 3. Equivalent circuit models for biological tissues. (a) Hayden model proposed 
by Hayden et al. (1969), (b) Simplied Hayden model, (c) Modied model, (d) Re- xp TR2e  sin p2 p
CPE parallel circuit model. Cm; capacitance of cell membrane, Rm; resistance of cell   5
membrane, Ri; intracellular resistance and Re; extracellular resistance. fxp TRe Ri g2 2xp TRe Ri  cos p2 p 1
The equivalent circuit parameters were estimated with complex
nonlinear least squares (CNLS) curve-tting (Macdonald, 1992). In
2.3. Models and curve-tting the CNLS tting, the sum of squares of the real and imaginary resid-
uals was minimized. Here, the residual means the difference be-
The resistance R (X) and reactance X (X) were calculated from tween a data value at a given frequency and the corresponding
the following equations: value calculated from the model. To evaluate the goodness of t,
we estimated the determination coefcient for each R and X and
R jZj cos h 1
the standard deviation SD dened as follows:
r
X jZj sin h 2 1Xn b i 2 X i  X b i 2
o
SD Ri  R 6
N i
The relationships between the R and X of the complex impedance
are shown by a ColeCole plot. For biological tissue, the ColeCole P
Ri  Rb i 2
plot was described as a circular arc (Cole, 1932). Model (a): The R2R 1  Pi 7
2
Hayden model (Hayden et al., 1969) for plant tissues shown in
i Ri  R
Fig. 3 takes account of the intracellular resistance Ri, extracellular
resistance Re, and the resistance and capacitance of the cell mem- P b i 2
X i  X
brane, Rm and Cm. We can assume that the membrane resistance R2X 1  Pi 2
8
Rm can be ignored because the value of impedance of Rm is much i X i  X
larger than those of other parameters. Thus the Hayden model b and Xb de-
where N denotes the number of measurement points, R
can be simplied to Model (b): the simplied Hayden model (Wu
note the approximated values of R and X, and R and X denote the
et al., 2008; Zhang and Willison, 1992). This model represents the
average values of R and X.
structure of one cell and describes an exact semicircle as a Cole
The unit of T changes with the values of p. Therefore, the unit of
Cole plot. However, tissues composed of many cells produce a time
T must be xed for the accurate comparison of the capacitive com-
constant distribution, and therefore, the ColeCole plot is described
ponents of the cell membrane. T is converted to apparent C by
as a semiellipse. In order to model this semiellipse, we used a con-
using Eq. (9) (Hsu and Mansfeld, 2001). Note that there is an
stant phase element (CPE) (Zoltowski, 1998) instead of Cm (the
assumption that the relaxation angular frequency (the angular fre-
modied model [Model (c)] shown in Fig. 3). Since the use of a
quency at which the imaginary part of the impedance Z is mini-
CPE can make it easy to t the model to the equivalent circuit accu-
mum) will stay the same.
rately, a CPE has been used in many studies (Itagaki et al., 2002;
Skale et al., 2007; Ricciardi et al., 2009). The impedance of CPE, ZCPE, C Txm p1 9
is calculated by the equation below: where xm is the relaxation angular frequency. xm of the modied
1 model can be obtained from the equation:
Z CPE p 3
jx T 1
1 p  1 p  xm 1
10
cos p  j p sin p 4 TRe Ri p
xTp 2 xT 2
By using Eqs. (9) and (10), we obtain the following:
where j is the imaginary unit, x is the angular frequency, T is a CPE
1 1p
coefcient, and p is a CPE exponent. Here, p is a factor in the range C m T p Re Ri p
11
of 01 that describes the time constant distribution in the system.
 Y. Ando et al. / Journal of Food Engineering 121 (2014) 2431 27

6 ered that the dispersion observed in our study is the b dispersion.

(a) Dispersion area The impedance at the lower frequency range was decreased as the
5
drying proceeded in the early drying stage (Fig. 5a). This result
0 min shows the decrease of extracellular resistance. At the late drying
4
stage, the impedance increased markedly as the drying proceeded,
Z (k)

3 30-min intervals and the frequency areas of dispersion shifted to a low frequency
area compared with those at the early drying stage (Fig. 5b). This
2
result suggests changes in the electrical capacity of the drying
240 min samples.
1
Fig. 7 shows the changes in the ColeCole plots of the samples
0 during drying at 60 C. The right side of each semiellipse describes
1 2 3 4 5 6 7 8
10 10 10 10 10 10 10 10
a low-frequency area, and the left side is a high-frequency area.
Frequency (Hz)
The straight-line part at the low-frequency of the impedance spec-
tra is known as the polarization impedance. This phenomenon was
12
Dispersion area caused by polarization at the electrode surface (Pliquett, 2010; Kal-
10 (b) vy et al., 2011). Since this impedance area does not relate to the
600 min cell structure, the corresponding parts were removed for tting.
8
At the early drying stage, semiellipse of the ColeCole plot shrunk
Z (k)

6 10-min intervals
4 520 min
2
0 and heat stress. Therefore, in our study, we suspect that these
1 2 3 4 5 6 7 8
10 10 10 10 10 10 10 10
Frequency (Hz)
Fig. 5. Temporal changes in the impedance spectra of the potato tissue during
drying at 60 C plotted with respect to frequency. (a) Impedance characteristics of
the sample dried for 0240 min measured at 30-min intervals. (b) Impedance
characteristics of the sample dried for 520600 min measured at 10-min intervals.
In this study, the apparent Cm was obtained by assigning the values
of each parameter in the modied model to Eq. (11) and dened as
the cell membrane capacitance.
3. Results and discussion
3.1. Behavior of the impedance characteristics during drying
Fig. 4 shows the changes in moisture content during drying at
50 C, 60 C, 70 C and 80 C. At each drying temperature, the mois-
ture content decreased as the drying time proceeded. Here, we de-
ne the period from the moisture content at the initial value to 1.0
(dry basis) as the early drying stage, and the period of moisture
content below 1.0 (dry basis) as the late drying stage. The imped-
ance of the samples plotted with respect to frequency is shown in
Fig. 5. The impedance characteristics of the samples dried for
0240 min measured at 30-min intervals and those dried for
520600 min measured at 10-min intervals are shown. Focusing
on the fresh sample (0 min drying) in Fig. 5a, it is seen that the
impedance declined very markedly as the frequency increased
between 104 and 106 Hz. The phenomenon in which the decrease
of impedance depends on the increase of frequency is called
dispersion.
A pattern diagram of the cells and the ow of the electric cur-
rent are shown in Fig. 6. In the low-frequency area, because of
the high electrical capacity of cell membranes, the electrical cur-
rent owed only through extracellular uid, which has relatively
high resistance. However, in the high-frequency area, the imped-
ance decreases immensely because the current is able to ow
through intracellular uid, which has relatively low resistance.
This phenomenon resulting from cell structures in biological tissue
is known as b dispersion (Pethig and Kell, 1987), and we consid-
as the drying proceeded (Fig. 7a). It was reported that a decrease in
the impedance of plant tissues occurred during freezing (Repo
et al., 1994; Wu et al., 2008; Zhang and Willison, 1992) and heating
(Zhang et al., 1993; Halder et al., 2011) due to the injury of cell
membrane. During the drying, the samples were exposed to drying
stresses led to injury of the cell membranes of the samples and re-
sulted in shrinkage of the ColeCole plot.
However, at the late drying stage, semiellipse was greatly en-
larged as the drying proceeded (Fig. 7b). Stamm (1929) reported
that at moisture content below 0.3 (dry basis), for wood tissues,
a linear relationship exists between the logarithm of the electrical
resistance and the moisture content, and this relationship, how-
ever, fails to hold in higher moisture content. Because dried vege-
tables have porous structure as dried wood, we suspect that the
same tendency existed in our study, therefore, the impedance
magnitude increased rapidly. We thus considered that the mois-
ture content is a dominant factor of the behavior of impedance
characteristics at the late drying stage.
To quantify the changes of impedance characteristics shown in
Figs. 5 and 7, we conducted an equivalent circuit analysis. In order
to test the validity of the model tting, we tted each model to
experimental values of impedance and compared the tting accu-
racies. Fig. 8 shows the approximation results of the models and
the experimental results of fresh potato tissue. As shown in Table 1,
the modied model has higher accuracy of t for both R and X com-
pared to the simplied Hayden model. By using the CPE, the phase
angles of Cm were changed exibly, and thus the modied model
could describe the impedance characteristics of inhomogeneous
tissues of the samples. From the above results, the validity of the
modied model was conrmed.
Fig. 9 shows the approximation results of the modied model
and the Re-CPE parallel circuit model shown in Fig. 3d for the
experimental impedance of dried potato tissue at the moisture
content 0.21. The goodness of t indexes are shown in Table 2.
The modied model has slightly high accuracy of t because it
has more variables than the Re-CPE parallel circuit model. How-
ever, it is reasonable to assume that Ri in the modied model
was vanished; that is, the Re-CPE parallel circuit model is the suit-
able model for dried samples. This model means that the samples
have a quite simple electrical structure which has both resistive
and dielectric behavior. The impedance data of samples at the late
drying stage (moisture content below 1.0) show the same ten-
dency. In this period, therefore, we performed an equivalent circuit
analysis by applying the Re-CPE parallel circuit model to the exper-
imental data. The complex impedance for the Re-CPE parallel
circuit model is obtained by substituting zero into Ri in Eq. (5)
and is represented as follows:
28 Y. Ando et al. / Journal of Food Engineering 121 (2014) 2431

Intracellular fluid

Extracellular fluid

Cell membrane

Low-frequency current High-frequency current

Fig. 6. Pattern diagram of the cells and ow of the electric current.

8
3 (a) 0 min Approximations
Experimental values 6
(I) Measured R
30-min intervals High freq. Measured X
2 4 Approximated R
-X (k)

R, X (k)
240 min Approximated X
Low freq. 2
1 0

-2
0
0 1 2 3 4 5 6 -4
2 4 6 8
10 10 10 10
R (k)
Frequency (Hz)
60
8
(b) 600 min Approximations
Experimental values 6 (II) Measured R
Measured X
40 10-min intervals High freq.
4 Approximated R
-X (k)

R, X (k)

Approximated X
520 min Low freq. 2
20
0
4
-2
0
0 20 40 60 80 100 -4
2 4 6 8
10 10 10 10
R (k)
Frequency (Hz)
Fig. 7. Temporal changes in the ColeCole plot of the potato tissue during drying at
60 C. (a) ColeCole plot of the sample dried for 0240 min measured at 30-min Fig. 8. Approximation results of three models and experimental results of fresh
intervals. (b) ColeCole plot of the sample dried for 520600 min measured at potato tissue. (I) Simplied Hayden model (Model (b)) and (II) modied model
10 min intervals. Solid lines in (a) represent approximations given by the modied (Model (c)).
model, and those in (b) represent approximations given by the Re-CPE parallel
circuit model.
Table 1
Goodness of t indexes of the simplied Hayden model and the modied model tted

x p
TR2e
 cos p2 p Re to impedance data of fresh potato tissue.
Z  j
xp TRe 2 2xp TRe  cos p2 p 1 Model R2R R2X SD (X)

xp TR2e  sin p2 p Simplied Hayden model 0.6278 0.4411 1129.80
  12 Modied model 0.9998 0.9995 28.00
xp TRe 2 2xp TRe  cos p2 p 1

The parameters of the circuit were estimated with CNLS in the same
manner as described above.
Fig. 10 shows the relationship between moisture content and
each equivalent circuit parameter during drying. The modied
model (Model (c) in Fig. 3) was used to analyze the impedance
spectra of the samples at the early drying stage, and the Re-CPE
parallel circuit model (Model (d) in Fig. 3) was used to analyze that
at the late drying stage. The solid lines in Fig. 7 represent the
approximated values. The experimental data showed a good t
with the approximations. The values of R2R of each sample were
within the range from 0.956 to 0.999, and the values of R2X were
within the range from 0.946 to 0.999. In the early drying stage,
the ratio of the standard deviation and average values (variation
coefcient) for each plot were smaller than 0.636 for Cm, 0.712
for Re, 0.664 for Ri, and those in the late drying stage were smaller
than 2.441 for Cm, 2.298 for Re. The variation coefcients in the late
drying stage were larger than those of in the early drying stage due
to the marked changes of the parameters in the late drying stage.
From the moisture content of the initial value to approx. 3.5,
each parameter was kept constant as shown in Fig. 10. The changes
of central temperature of the samples during drying are shown in
Fig. 11. We suspect that from the moisture content of the initial
value to 3.5, the cell membrane was not injured because the
sample was kept relatively low temperature (below 45 C). At the
moisture content below 3.5, the capacitance of the cell membrane
 Y. Ando et al. / Journal of Food Engineering 121 (2014) 2431 29

400 80
(I) 50C
70

Central temperature (C)

60C
Measured R
70C
R, X (k)

200 Measured X 60
80C
Approximated R
Approximated X 50
0
40

30
-200
2 4 6 8
10 10 10 10 20
Frequency (Hz) 0 1 2 3 4 5
Moisture content (dry basis (g/g))
400
(II) Fig. 11. Changes of central temperature of potato tissues during drying.

Measured R
200
R, X (k)

Measured X
Approximated R 40
Approximated X
0
30

Re / Ri
-200
2 4 6 8 20
10 10 10 10
Frequency (Hz) 50C
60C
Fig. 9. Approximation results of the models and experimental results of dried 10
potato tissue at moisture content of 0.21. (I) Modied model (Model (c)) and (II) Re-
70C
CPE parallel circuit model (Model (d)). 80C
0
0 1 2 3 4 5
Table 2 Moisture content (dry basis (g/g))
Goodness of t indexes of the modied model and the Re-CPE parallel circuit model
tted to impedance data of dried potato tissue at moisture content of 0.21. Fig. 12. Relationship between moisture content and Re/Ri during drying at 50 C,
60 C, 70 C and 80 C.
Model R2R R2X SD (X)

Modied model 0.9993 0.9925 3719.27

Re-CPE parallel circuit model 0.9992
decreased as the moisture content decreased. Water content plays
an important role to maintain the structural and functional integ-
rity of biological membranes (Crowe and Crowe, 1982). Therefore,
we suspect that the decrease in cell membrane capacitance was
caused by the structural changes in the cell membrane which
occurred in the dehydration process.
For example, the decrease of the area of the cell membrane
caused by plasmolysis during drying could be related to the
decrease in the capacitance of the cell membrane. Lew (1996)
investigated the interplay between cell turgor pressure and the
electrical characteristics of the cell membrane, using growing
0.9914 4014.53 Arabidopsis root hairs. He also found that electrical characteristics
of the cell membrane changed with the ion efux across the mem-
brane caused by the osmotic changes of the extracellular uid.
Thus, the movement of the ions due to plasmolysis during drying
could have caused the changes in cell membrane capacitance. In
the moisture content less than 2.0, the decrease in capacitance
was more pronounced at high drying temperature. Zhang and
Willison (1992) reported that a reduction in the capacitance of
the cell membrane arises after heat injury of plant cells. We thus
considered that heat damage of the cell membrane occurred in
the latter term of the drying.
The Re and Ri take low values at the higher drying temperatures,
indicating that these parameters had negative temperature depen-
dence. The electrical conductivity of the electrolyte solution is due

5
5 10 400

4
Cm Re Ri 50C
300 60C
4
10 70C
3
C (nF)

Re ()

Ri ()

80C
200
2 50C 50C
3
10
60C 60C 100
1 70C 70C
80C 80C
2
0 10 0
0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5
Moisture content (dry basis (g/g)) Moisture content (dry basis (g/g)) Moisture content (dry basis (g/g))
Fig. 10. Relationship between moisture content and equivalent circuit parameters during drying at 50 C, 60 C, 70 C and 80 C. Cm: capacitance of cell membrane, Ri:
intracellular resistance and Re: extracellular resistance.
30 Y. Ando et al. / Journal of Food Engineering 121 (2014) 2431
-1
10
C Re
-2
10
C (nF)
Re ()
-3
10
-4
10
0 1 2 3
Moisture content (dry basis (g/g))
Fig. 13. Relationship between moisture content and parameters in the Re-CPE parallel circuit model during water absorption process at 30 C.
to the ion conductivity. If the received energy becomes larger, the
ions become active and the electrical conductivity of the electro-
lyte solution is increased. That is to say, the resistance of electro-
lyte solution is decreased as temperature increases. The
intracellular resistance Ri values were increasing and the extracel-
lular resistance Re values were decreasing at the moisture content
from 3.5 to 1.0. In normal cells, intracellular uid with a relatively
high concentration of electrolytes and extracellular uid with a
low concentration are divided by the cell membrane. The present
results can be explained by assuming that the cell membrane of
the drying sample was injured by drying and heat stresses and that
intracellular uid leaked from the inside of the cells to the outside.
The result that the values of Re and Ri were approaching each other
as the drying progressed support this assumption.
From the above results, the ratio of the values of ReRi could be
an index of the healthiness of the cells. The relationship between
the moisture content and the ratio of ReRi at each drying temper-
ature is shown in Fig. 12. The ratio of Re and Ri takes a value of
greater than 1, and the value is smaller as the injuries of the cells
become greater. The initial value of the ratio of ReRi was approx.
3235 and decreased as the drying progressed. The higher drying
temperatures caused faster changes in the values, indicating that
the heat stress had strongly inuenced the cell membranes.
In the period of moisture content below 1.0 in Fig. 10, the values
of Cm had been quite small and the values of Re were sharply in-
creased. These results can be explained as follows: the cell mem-
brane was completely disrupted and the marked change in Re in
the late drying stage is mainly attributable to the loss of the mois-
ture content, as mentioned above. This view is also supported by
the nding that our experimental data in the late drying stage do
not t with the modied model.
3.2. Behavior of the impedance characteristics during water absorption
process
From these experiments, we hypothesized that the dominant
factor of the behavior of impedance characteristics in the early dry-
ing stage were (1) injury of the cell membrane, and (2) leakage of
electrolytes from the intracellular uid to the extracellular uid. In
addition, in the late drying stage (after the disruption of cell mem-
brane), the changes in the impedance characteristics were caused
by a decrease in conductivity associated with moisture loss. To test
this hypothesis, we investigated the behavior of impedance charac-
teristics during water absorption using dried samples obtained
from the drying experiment. In this process, we found that the
Re-CPE parallel circuit model was the most suitable model as an
equivalent circuit model, which indicates that the cell structures
of the samples were completely collapsed due to drying.
6
10
The sample dried at 50C
The sample dried at 60C
5
10 The sample dried at 70C
The sample dried at 80C
4
10
3
10
2
10
0 1 2 3
Moisture content (dry basis (g/g))
Fig. 13 shows the relationship between moisture content and
equivalent circuit parameters during the water absorption process
obtained by applying the Re-CPE parallel circuit model to the
experimental data. The variation coefcient values of C for each
plot were within the range from 0.004 to 1.180, and those of Re
were within the range from 0.011 to 1.850. In the period of mois-
ture content below 1.0, the Re values were greatly decreased from
the start of water absorption, and the relationship between mois-
ture content and Re was almost consistent with that seen during
the drying. At the higher moisture content, in contrast, there was
a small change in Re; that was, the change in Re was almost inde-
pendent of the moisture content in the higher moisture content
area.
The C values were slightly increased between approx. 103 and
2
10 (nF) with the increase of moisture content. We suspect that
this result was caused by absorption of water, which has a high
dielectric constant compared to solid materials. The changes of C
during the water absorption were quite small compared to those
during drying. This result suggests that the changes in C during
drying were independent of the moisture content changes and
originated from structural changes in the cell membrane. These re-
sults support that the behavior of the impedance in the early dry-
ing stage was based on the changes in electrical properties of the
cells regardless of the moisture content.
4. Conclusions
We investigated the impedance characteristics of drying potato
tissues to evaluate the changes in the cell physiological status dur-
ing drying. At the early drying stage (from the initial moisture con-
tent to 1.0 [dry basis]), the modied Hayden model was found to
be used to describe the impedance frequency characteristics. From
the equivalent circuit analysis, we conclude that the cell mem-
branes were damaged by drying and heat stresses and that intra-
cellular uid leaked out of the cells. At the late drying stage, the
cell membrane was completely disrupted and due to the loss of
moisture, the impedance magnitude increased rapidly as the dry-
ing proceeded. We also investigated impedance characteristics
during the water absorption process of the dried samples to under-
stand how the moisture content affects the impedance of the sam-
ples. From the results of the equivalent circuit analysis using the
Re-CPE parallel circuit model, we conclude that the changes in
the equivalent circuit parameters at the early drying stage were
due to membrane injury, and not to changes of moisture content.
The results obtained showed the behavior of the cell physiological
status of potato tissues during drying. In addition, the potential of
EIS as a method for evaluating the healthiness of biological cells
was demonstrated.
 Y. Ando et al. / Journal of Food Engineering 121 (2014) 2431
Acknowledgement
This work was supported by the JSPS KAKENHI [Grant-in-Aid
for JSPS Fellows, 24 1232].
References
Agilent Technologies. Agilent Impedance Measurement Handbook, fourth ed.
<http://cp.literature.agilent.com/litweb/pdf/5950-3000.pdf> (accessed
01.29.13)
Ando, Y., Mizutani, K., Wakatsuki, N., 2012. Electrical impedance analysis of potato
tissues during drying process. In: 2012 IFT Annual Meeting & Food Expo
Technical Program Book of Abstracts, 077128, 7576.
Blahovec, J., Lahodov, M., 2013. Storage induced changes of potato properties as
detected by DMA. LWT-Food Science and Technology 50, 444450.
Blahovec, J., Lahodov, M., 2012. DMA thermal analysis of different parts of potato
tuber. Food Chemistry 133, 11011106.
Blahovec, J., Lahodov, M., 2011. DMA peaks in potato cork tissue of different
mealiness. Journal of Food Engineering 103, 273278.
Cole, K.S., 1932. Electric phase angle of cell membranes. The Journal of General
Physiology 15 (6), 641649.
Cooley, D.R., Evert, D.R., 1979. Normalized electrical impedance evaluates cold
injury to stem sections of Delicious apple. Journal of the American Society for
Horticultural Science 104, 561563.
Crowe, J.H., Crowe, L.M., 1982. Induction of anhydrobiosis: membrane changes
during drying. Cryobiology 19, 317328.
Cunningham, S.E., Mcminn, W.A.M., Magee, T.R.A., Richardson, P.S., 2008. Effect of
processing conditions on the water absorption and texture kinetics of potato.
Journal of Food Engineering 84, 214223.
Damez, J.L., Clerjon, S., Abouelkaram, S., Lepetit, J., 2007. Dielectric behavior of beef
meat in the 11500 kHz range: simulation with the Fricke/ColeCole model.
Meat Science 77, 512519.
Falk, S., Hertz, H.C., Virgin, H.I., 1958. On the relation between turgor pressure and
tissue rigidity. I. Physiologia Plantarum 11, 802817.
Greenham, C.G., 1966. Bruise and pressure injury in apple fruits. Journal of
Experimental Botany 17, 404409.
Halder, A., Datta, A.K., Spanswick, R.M., 2011. Water transport in cellular tissues
during thermal processing. AIChE Journal 57, 25742588.
Hayden, R.I., Moyse, C.A., Calder, F.W., Crawford, D.P., Fensom, D.S., 1969. Electrical
impedance studies on potato and alfalfa tissue. Journal of Experimental Botany
20, 177200.
Hiller, S., Jeronimidis, G., 1996. Fracture in potato tuber parenchyma. Journal of
Materials Science 31, 27792796.
Hsu, C.H., Mansfeld, F., 2001. Concerning the conversion of the constant phase
element parameter Y0 into a capacitance. Corrosion 57, 747748.
Itagaki, M., Taya, A., Watanabe, K., Noda, K., 2002. Deviations of capacitive and
inductive loops in the electrochemical impedance of a dissolving iron electrode.
Analytical Sciences 18, 641644.
Jha, S.N., Narsaiah, K., Basediya, A.L., Sharma, R., Jaiswal, P., Kumar, R., Bhardwaj, R.,
2011. Measurement techniques and application of electrical properties for
nondestructive quality evaluation of foodsa review. Journal of Food Science
Technology 48 (4), 387411.
Juansah, J., Budiastra, I.W., 2012. Electrical behavior of Garut citrus fruits during
ripening: Changes in resistance and capacitance models of internal fruits.
International Journal of Engineering and Technology 12 (04).
Kalvy, H., Johnsen, G.K., Martinsen, .G., Grimnes, S., 2011. New method for
separation of electrode polarization impedance from measured tissue
impedance. The Open Biomedical Engineering Journal 5, 813.
Kandala, C.V.K., Nelson, S.O., Lawrence, K.C., 1989. Non-destructive electrical
measurement of moisture content in single kernels of corn. Journal of
Agricultural Engineering Research 44, 125132.
31
Laza, M., Scanlon, M.G., Mazza, G., 2001. The effect of tuber pre-heating temperature
and storage time on the mechanical properties of potatoes. Food Research
International 34, 659667.
Lew, R.R., 1996. Pressure regulation of the electrical properties of growing
Arabidopsis thaliana L. root hairs. Plant Physiology 112, 10891100.
Lewicki, P.P., Jakubczyk, E., 2004. Effect of hot air temperature on mechanical
properties of dried apples. Journal of Food Engineering 64, 307314.
Macdonald, J.R., 1992. Impedance spectroscopy. Annals of Biomedical Engineering
20, 289305.
Mazza, G., 1983. Dehydration of carrots. Effects of pre-drying treatments on
moisture transport and product quality. Journal of Food Technology 18, 113
123.
McKeown, M.S., Trabelsi, S., Tollner, E.W., Nelson, S.O., 2012. Dielectric spectroscopy
measurements for moisture prediction in Vidalia onions. Journal of Food
Engineering 111, 505510.
Nilsson, S.B., Hertz, C.H., Falk, S., 1958. On the relation between turgor pressure and
tissue rigidity. II. Physiologia Plantarum 11, 818837.
Nelson, S.O., 1994. Measurement of microwave dielectric properties of particulate
materials. Journal of Food Engineering 21, 365384.
Nelson, S.O., 2005. Dielectric spectroscopy in agriculture. Journal of Non-Crystalline
Solids 351, 29402944.
Pethig, R., Kell, D.B., 1987. The passive electrical properties of biological systems:
their signicance in physiology, biophysics and biotechnology. Physics in
Medicine and Biology 32, 933970.
Pliquett, U., 2010. Bioimpedance: a review for food processing. Food Engineering
Review 2, 7494.
Repo, T., Zhang, M.I.N., Ryypp, A., Vapaavuori, E., Sutinen, S., 1994. Effects of freeze-
thaw injury on parameters of distributed electrical circuits of stems and needles
of Scots pine seedlings at different stages of acclimation. Journal of
Experimental Botany 45, 823833.
Ricciardi, S., Ruiz-Morales, J.C., Nuez, P., 2009. Origin and quantitative analysis of
the constant phase element of a platinum SOFC cathode using the state-space
model. Solid State Ionics 180, 10831090.
Skale, S., Dolecek, V., Slemnik, M., 2007. Substitution of the constant phase element
by Warburg impedance for protective coatings. Corrosion Science 49, 1045
1055.
Stamm, A.J., 1929. The ber-saturation point of wood as obtained from electrical
conductivity measurements. Industrial and Engineering Chemistry, Analytical
Edition 1 (2), 9497.
Trabelsi, S., Nelson, S.O., 2006. Nondestructive sensing of bulk density and moisture
content in shelled peanuts from microwave permittivity measurements. Food
Control 17, 304311.
Troncoso, E., Pedreschi, F., 2007. Modeling of textural changes during drying of
potato slices. Journal of Food Engineering 82, 577584.
Virgin, H.I., 1955. A new method for the determination of the turgor of plant tissues.
Physiologia Plantarum 8, 954962.
Wu, L., Ogawa, Y., Tagawa, A., 2008. Electrical impedance spectroscopy analysis of
eggplant pulp and effects of drying and freezing-thawing treatments on its
impedance characteristics. Journal of Food Engineering 87, 274280.
Yamamoto, T., Yamamoto, Y., 1976. Electrical properties of the epidermal stratum.
Medical and Biological Engineering 14, 151158.
Yamamoto, T., Yamamoto, Y., 1977. Analysis for the change of skin impedance.
Medical & Biological Engineering & Computing 15, 219227.
Zhang, M.I.N., Willison, J.H.M., 1991. Electrical impedance analysis in plant tissues:
a double shell model. Journal of Experimental Botany 42, 14651475.
Zhang, M.I.N., Willison, J.H.M., 1992. Electrical impedance analysis in plant tissues:
the effect of freeze-thaw injury on the electrical properties of potato tuber and
carrot root tissues. Canadian Journal of Plant Science 72, 545553.
Zhang, M.I.N., Willison, J.H.M., Cox, M.A., Hall, S.A., 1993. Measurement of heat
injury in plant tissue by using electrical impedance analysis. Canadian Journal
of Botany 71, 16051611.
Zoltowski, P., 1998. On the electrical capacitance of interfaces exhibiting constant
phase element behavior. Journal of Electroanalytical Chemistry 443, 149154.