Innovative Food Science and Emerging Technologies 43 (2017) 1825

Contents lists available at ScienceDirect

Innovative Food Science and Emerging Technologies

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

Driving forces for mass transfer in electrohydrodynamic (EHD) drying MARK

a, a b a a
Alex Martynenko , Tess Astatkie , Nicolas Riaud , Patrick Wells , Tadeusz Kudra
a
Department of Engineering, Faculty of Agriculture, Dalhousie University, Truro, Nova Scotia B2N 5E3, Canada
b
CNRS - Universit de Poitiers, 11 Boulevard Marie et Pierre Curie BP 30179, F86962 Cedex, France

A R T I C L E I N F O A B S T R A C T

Keywords: Electrohydrodynamic (EHD) drying is considered as energy ecient non-thermal technology suitable for de-
Ionic wind watering of heat-sensitive materials, including food products. A factorial experimental design was used to
Corona discharge identify signicant factors aecting mass transfer in EHD drying. The experiment revealed signicant eects of
Electric eld voltage, distance (gap) between electrodes, conguration of discharge electrode, air cross-ow and the char-
Charge distribution
acteristics of material surface on mass transfer. Strong coupling between mass and charge transfer was found for
Evaporation
all voltages, gaps and congurations of discharge electrode. The eect of air cross-ow at 1.0 m/s on mass
Convection
transfer was additive to the eect of ionic wind, consistently increasing total mass transfer by 5.05.1 g/h in all
experimental conditions. These results led to the conclusion that the eect of EHD is convective in nature,
enhancing mass transfer due to ionic wind. The eect of material surface characteristics was attributed to dif-
ferent hydrodynamic conditions of air boundary layer.
Industrial relevance: This research is focused on better understanding of the factors that play a signicant role in
EHD drying, and therefore it is important for industry to facilitate practical applications of EHD in bioprocessing
and food engineering.

1. Introduction Bajgai & Hashinaga, 2001a), higher rehydration ratio

Electrohydrodynamic (EHD) drying is regarded as emerging, non-
thermal dewatering technology, which appears to be a viable alter-
native to conventional thermal drying for certain heat-sensitive mate-
rials, such as high-value bioactive components of fruits and medicinal
plants (polyphenols, avonoids, dietary ber, etc.), living cells (bac-
teria, yeasts and viruses), and non-living substances of biological origin
(blood plasma, serum, hormones, antibiotics, probiotics, nutraceuticals,
etc.) (Alemrajabi, Rezaee, Mirhosseini, & Esehaghbeygi, 2012; Bajgai,
Raghavan, Hashinaga, & Ngadi, 2006; Singh, Orsat, & Raghavan, 2012;
Zhang, Chen, Mujumdar, Zhong, & Sun, 2015). The positive eect of
EHD on mass transfer was also reported for combination with other
drying techniques, such as EHD-assisted hot air drying (Alemrajabi
et al., 2012; Dinani, Havet, Hamdami, & Shahedi, 2014b; Singh et al.,
2017), EHD-assisted vacuum freeze drying (Bai, Yang, & Huang, 2012),
or EHD with auxiliary contact heating (Lai & Wang, 2009). Energy
consumption in EHD drying is much lower than that in hot air drying,
likely because of targeted supply of energy for moisture evaporation
and practically there is no heat lost with exhaust air. The energy-related
issues in EHD drying have been reviewed by Kudra and Martynenko
(2015). The reported benets of EHD compared to hot air drying on
food quality include lesser shrinkage (Alemrajabi et al., 2012;
(Bajgai & Hashinaga, 2001b), preserved color (Alemrajabi et al., 2012;
Bajgai & Hashinaga, 2001a; Esehaghbeygi & Basiry, 2011; Hashinaga,
Bajgai, Isobe, & Barthakur, 1999; Xue, Barthakur, & Alli, 1999), and
higher vitamin C content (Bajgai & Hashinaga, 2001b).
The principle behind EHD drying is the phenomenon of charge
transfer (corona discharge) from a high-voltage electrode. The corona
appears when a high voltage is applied to two electrodes with sub-
stantially dierent radii of curvature, such as a sharp vertical pin or ne
horizontal wire and a at surface, giving respectively point-to-plate or
wire-to-plate congurations (Kulacki, 1982). The discharge electrode
with larger curvature generates ionic wind due to corona discharge,
which impinges the surface of wet material deposited on the collecting
(grounded) electrode (Fig. 1).
EHD drying is the most ecient under conditions of stable glow
discharge from low-energy ions, dragged by electric force (Goldman,
Goldman, & Sigmond, 1985). The space charge in the drift region is the
major factor of drying, determining current density distribution and
mass transfer from the surface of the wet material. The space charge
density in the drift region depends on the gap between electrodes, ex-
pansion of ionization region and thickness of the material under drying.
The limit of ionization region () is determined by electric eld
strength.

Corresponding author.
E-mail address: alex.martynenko@dal.ca (A. Martynenko).

http://dx.doi.org/10.1016/j.ifset.2017.07.022
Received 16 February 2017; Received in revised form 26 May 2017; Accepted 9 July 2017
Available online 10 July 2017
1466-8564/ 2017 Elsevier Ltd. All rights reserved.
A. Martynenko et al.
Fig. 1. A typical pin-to-plane corona geometry.
(Adopted from Goldman et al., 1985).
1.1. Existing theories of electrically-induced mass transfer
Mass transfer in EHD drying could be attributed to various physical
phenomena. Some of them, such as ionic wind, impingement of mate-
rial surface, gradient of surface tension, polarization, space charge and
electrocapillarity are briey discussed in the topical literature
(Rounsley, 1985). Early research on EHD considered the ion-drag force
(ionic wind) as a major driving force for mass transfer (Kulacki, 1982;
Robinson, 1961). This force, resulting from exposure of a unit volume
of charged gas to electric eld with strength E, is represented as a
pressure gradient P
P = c E (1)
where c is the space charge density, C/m . 3 slices (Ding, Lu, & Song, 2015), demonstrated exponential decay of
The ionic wind velocity from the electric eld force at the surface of
the collecting electrode can be estimated from the momentum con-
servation law (Robinson, 1961)
d
ue 2
2
= c Edz
0 (2)
where d is the gap between discharge and collecting electrodes (m),
stands for air density (kg/m3), and ue represents ionic wind velocity (m/
s). Taking air density as constant and independent of both the water
vapor density and electric charge density, Barthakur and Al-Kanani
(1989) derived the linear relationship between ionic wind velocity ue
and electric eld strength E
o in water vapor concentration between the sample surface and ambient
ue = E gas (kg/m3) and hm denotes the mass transfer coecient (m/s).
(3)
where o represents dielectric permittivity of vacuum (8.85 pF/m).
This equation considers uniform electric eld at the surface of the
collecting electrode. However, according to Warburg law (Warburg,
1927), discharge from the sharp pin or thin wire causes highly non-
uniform distribution of electric current density, and therefore non-
homogeneous electric eld at the plane surface (see Fig. 1). The
19
Innovative Food Science and Emerging Technologies 43 (2017) 1825
relationship between ionic wind velocity and current density j (A/m2)
at the surface of collecting electrode has been derived by Robinson
(1961) as
jd
ue =
b (4)
with
j = jo cos5 (5)
2
where b is the ionic mobility (m /(s V)), jo is the maximum current
density just underneath of a pin, and is the Warburg angle, de-
termined from the geometry of electrode (see Fig. 1).
Eqs. (4) and (5) predict non-uniform ionic wind velocity over the
surface of the material, which then was conrmed experimentally
(Kawamoto, Yasuda, & Umezu, 2006; Rickard, Dunn-Rankin,
Weinberg, & Carleton, 2006). These results correspond well to early
experimental ndings of Isobe, Barthakur, Yoshino, Okushima, and
Sase (1999), who reported Warburg distribution (Eq. (5)) of mass
transfer, initiated by the pin electrode at the surface of agar gel.
It is important to note that the estimate of ionic wind velocity from
Eq. (4) is about one third of than the one given by Eq. (3). This fact
could be attributed to the averaging the prole of ionic wind velocity
over the surface area. The maximum ionic wind velocity, calculated
from Eq. (3), is usually on the order of several meters per second, which
denitely entails the aerodynamic eect, disturbing the boundary layer
at the material's surface.
The surface action of EHD was conrmed in the experiments with
water evaporation (Kamkari & Alemrajabi, 2010; Li, Li, & Tatsumi,
2000; Wolny & Kaniuk, 1996), wetted solid and perforated glass beads
(Ramachandran & Lai, 2010), sand (Singh et al., 2017), apple slices in
the rst 5 h of drying (Hashinaga et al., 1999), tomato slices in the rst
3 h of drying (Esehaghbeygi & Basiry, 2011), carrots and miscanthus in
the rst 1.4 h of drying (Pogorzelski, Zander, Zander, & Wrotniak,
2013). These experiments demonstrated constant drying rate, which is
typical for convective mass transfer. It corresponds to Type 1 (Dirichlet)
boundary condition, when liquid mass transfer is controlled only by the
gas properties, such as temperature, density, humidity and supercial
velocity.
In contrast, EHD drying of biomaterials, such as rapeseed
(Basiry & Esehaghbeygi, 2010), tomato slices after 3 h of drying
(Esehaghbeygi & Basiry, 2011), kiwi fruits (Dalvand,
Mohtasebi, & Raee, 2013), apple slices (Martynenko & Zheng, 2016),
mushrooms (Dinani, Hamdami, Shahedi, & Havet, 2014a) and carrot
moisture content or falling drying rate, which is typical for the diu-
sion-limited mass transfer. The diusion could be constrained either
because of low internal diusivity of the biomaterial (Ding et al., 2015),
or due to the receding evaporation front with solid-gas interface below
the material surface (Alem-Rajabi & Lai, 2005; Pogorzelski et al., 2013).
This case corresponds to Type 3 boundary condition (Robin), linking
water diusion towards the material surface and convection from the
material surface
dm m
=D = hm (m m)
dt x (6)
2
where D is the water diusivity (m /s) in the material with thickness x
(m), m = m m stands for the concentration gradient or dierence
It was found that mass transfer increases with voltage and decreases
with the gap between electrodes (Lai & Lai, 2002). However, it is still
unclear, whether mass transfer is driven by voltage or by electric eld
strength. Moreover, the eect of electrode conguration on charge
density and mass transfer has never been studied.
The eect of air cross-ow on mass transfer was thoroughly ex-
amined by Lai and Lai (2002) and Lai and Sharma (2005). Whereas
A. Martynenko et al. Innovative Food Science and Emerging Technologies 43 (2017) 1825

forced convection was advantageous in the case of thermal drying be-

cause of increased turbulence at the material-air interface, it was not
benecial for EHD drying. Cross-ow of air, parallel to material surface,
signicantly suppressed the positive eect of ionic wind due to aero-
dynamic eect of convective airow. In contrast, it was found that la-
minar cross-ow facilitated eect of ionic wind on mass transfer
(Lai & Lai, 2002; Sadek, Fax, & Hurwitz, 1972). The positive eect on
EHD drying for 1.0 m/s air velocity has also been reported by
Martynenko and Zheng (2016). Apparently, these ndings indicate the
important role of a boundary layer in the EHD mass transport. Further
study on the eect of material surface on the mass transfer is necessary
for the optimization of EHD-assisted drying.
Despite numerous researches on EHD eect on water evaporation,
the general theory of mass transfer in electrically-charged gas in the
presence of electric eld has not yet been developed. Thus, the objec-
tive of this study was to identify signicant factors of mass transfer and
their interactions to facilitate practical applications of EHD in biopro-
cessing and food engineering.

2. Materials and methods

Fig. 3. Drying chamber as seen through the air exhaust vent.

2.1. Materials
To determine the eect of material on EHD mass transfer, two ca-
pillary-porous materials with distinctively dierent surface character-
istics were used. The rst one was the pack of wet tissue paper
(Cascades Tissue Group, Canada) with 5 mm thickness and smooth at
surface. The second test material was the wet sponge with 5 mm
thickness and rough porous surface. Our postulations related to the
choice of these materials were: (i) free-water surface evaporation; (ii)
negligible moisture gradients within the material, (iii) capillary action,
i.e., insignicant molecular diusion of liquid water within the material
compared to capillary ow, and (iv) negligible shrinkage, which implies
constant thickness of the material and constant gap during drying.
2.2. Experimental apparatus
A lab-scale convective chamber, equipped with multiple pins-to-
plate electrode system, used for EHD drying is shown in Fig. 2.
The experimental apparatus consisted of discharge and collecting
electrodes, computerized system for mass and electric current mea-
surements, controllable speed blower (Fantech, Model K4, Canada), and
universal high voltage DC power supply (Model BAL32-10-V, Universal
Voltronics, USA). The 40 20 20 cm drying chamber was made
from transparent plastic, having two air vents 10.2 cm in diameter on
both sides of the chamber (Fig. 3.).
Air blower provided stabilized airow across the chamber with
supercial velocity of 1.0 m/s, as measured in the drying chamber by
the hot wire anemometer (Digi-Sense, model 20,250, Cole-Palmer,
Canada). The multiple-pin discharge electrode was formed from 1.5 cm
long sharp stainless steel pins 1.35 mm in diameter and 31 cone angle,
attached to the berglass plate 185 125 1.5 mm in the nodes of
the rectangular grid of cells. Two congurations of electrodes with
dierent packing densities were used in this study: one with 143 pins
arranged in 11 13 rows with 1 cm square cells (1 1), and another
with 42 pins arranged in 6 7 rows with 2 cm square cells (2 2).

Fig. 2. Experimental setup: 1 - computer; 2 -

high-voltage power supply; 3 - drying chamber, 4
- multiple-pin discharge electrode, 5 - grounded
plate electrode, 6 - air blower with ow stabi-
lizer.

20
A. Martynenko et al. Innovative Food Science and Emerging Technologies 43 (2017) 1825

The discharge electrode was connected to the positive pole of the high
voltage source providing DC current up to 32 kV and 10 A. The col-
lecting electrode made of aluminum plate 12.5 20 0.15 cm was
centered below discharge electrode and connected to the ground of the
high voltage power supply. The gap between discharge and grounded
electrodes was adjustable in the range from 2.0 to 4.0 cm. Both voltage
and current were displayed on the control panel of the power supply,
and continuously recorded on the computer through a digital interface
NI USB-6210 (National Instruments, USA). Air velocity in experiments
with forced convection was maintained constant at 1.0 0.1 m/s.

2.3. Drying experiments

The eect of EHD on the water mass transfer was evaluated from
drying experiments of wet material with capillary action at room
temperature of 20 1 C and air humidity in 5570% range. The
sample of porous wet material with high capillarity was formed from
either 5 mm sponge 10 16 cm or a pack of 40 dry tissue paper
10 16 cm, which were wetted with tap water to 50.0 1.0 g and
placed centrally on the surface of the collecting electrode. The surface
area of wet sample was entirely exposed to ionic wind produced by the
array of multi-pin electrode. Drying experiments were carried out for
dierent combinations of voltages (14 and 19 kV DC +), congurations
of discharge electrode (1 1) and (2 2), gaps (3.5 and 4.0 cm), and
air velocities (0 and 1.0 m/s). In this study, both current and mass
transfer were considered as output (response) variables. Each experi-
ment was carried out at constant voltage in the range of currents from 0
to 300 A. Mass reduction of the wet material was measured by top-
loaded digital balance HCB1002 (Adam Equipment, Danbury, CT, USA)
and continuously recorded through a USB interface to the computer.
Drying rate g/(sm2) was calculated from time series of mass mea-
surements with respect to specic area of evaporation. Mass transfer to
gaseous phase was considered equal to drying rate from wet material.
The duration of each drying experiment was set to 1 h.

2.4. Statistical analysis Fig. 4. EHD drying kinetics of wet paper tissue (A) and sponge (B) for (1 1) electrode:
circles-14 kV, triangles-19 kV, lled-mass, emptycurrent. Gap 40 mm, no airow.

Unreplicated 25 factorial design was used to determine the eect of

operational variables (voltage, gap, electrode conguration, air velo-
city, and type of the material) on two output (response) variables
(drying rate and current). Since the full factorial model of unreplicated
designs cannot be analyzed using Analysis of Variance (ANOVA) be-
cause of the zero degrees of freedom for the error, a two-phase analysis
was conducted. In the rst phase, normal probability plot of the eects
along with the Lenth test was used to identify active eects, and then in
the second phase, ANOVA of a reduced model (reduced by moving the
inactive eects to the error) was used to re-conrm the results
(Astatkie, Joseph, & Martin, 2006). Normal distribution and constant
variance of the error terms assumptions were veried by examining the
residuals of the second phase ANOVA model as described in
Montgomery (2013). The analysis was completed using Minitab soft-
ware (Minitab 17, Minitab Inc., State College, PA, USA).
3. Results and discussion
3.1. Drying kinetics
Drying kinetics of the wet paper tissue and sponge measured within
1 h at dierent EHD settings is presented in Fig. 4.
This linear drying kinetics, observed in both cases (Fig. 4A and
Fig. 4B), is specic for the materials with capillary action. It is dierent
from drying of biomaterials with diusion-limited mass transfer, like
apples (Martynenko & Zheng, 2016) or mushrooms (Dinani, Havet,
et al., 2014). The constant drying rate, observed in this experiment
(Fig. 4) validated the rst postulation about free water evaporation
from the material surface. It also proved the second postulation about
negligible moisture gradients inside of the material, at least within the
rst hour of drying. Constant drying rate, observed for EHD drying
implies constant water content on the surface of the material and
constant coecient of mass transfer hm. It means that the third postu-
lation about insignicant eect of diusion on the mass transfer was
correct. It is important to note that electric current in all experiments
was also constant for the entire period of drying, which indicated sta-
tionary drying conditions.
It was also noticed that electric eld directly aected weight mea-
surements. In our experiment, immediate decrease of weight by 22.5 g
was recorded in the presence of an electric eld (Fig. 4A). Generally,
this eect was proportional to the strength of electric eld between
electrodes and was fully reversible when voltage was turned o. It did
not change when wet material was removed from the gap, which im-
plies possible eect of Coulomb force between discharge and collecting
electrodes.
The eect of EHD and convective airow (separately and in com-
bination of both) on the mass transfer is clearly illustrated in Fig. 5.
Constant drying rate of capillary-porous materials during long
period of time was quite benecial for quantifying eects of multiple
factors on mass transfer in EHD drying. From Fig. 5 it can be concluded
that the mass transfer depends on the electric eld strength and air
cross-ow. Mass transfer due to the cross-ow was 1.2 and 5.0 g/h for
free and forced convection, respectively. Mass transfer due to EHD was
3.9 g/h for free convection and 9.0 g/h for forced convection.
Current set of experiments completed within the chosen range of
voltages, gaps and electrode congurations clearly demonstrated strong
correlation between mass transfer rate and electric current. Apparently,

21
A. Martynenko et al.
Fig. 5. Drying kinetics of wet paper tissue: 1 - free (natural) convection; 2 - forced con-
vection 1.0 m/s; 3, 4 - EHD at 14 kV, gap 4.0 cm and (1 1) electrode, without (3) and
with (4) airow 1.0 m/s.
higher ionic wind resulted in higher electric current and, as a con-
sequence higher mass transfer rate. To study the relationships between
mass transfer, ionic wind and forced airow under dierent experi-
mental settings, the multifactorial experiment was carried out.
3.2. Eects of voltage, gap, electrode conguration, airow and type of the
material on mass transfer
To study the eect of the selected factors, the experiment was car-
ried out according to an unreplicated ve-factor, each with two levels
(25) factorial design that has 32 runs, but conducted in a random order
(Astatkie et al., 2006). The levels of each factor have carefully been
chosen based on preliminary research on current-voltage characteristics
(Table 1).
Preliminary analysis of the eect of these factors indicated that both
the current and drying rate for wet paper tissue were much higher than
the ones for the sponge. This eect of material on the mass transfer and
current was more pronounced for (1 1) than for (2 2) electrode.
The positive eect of air cross-ow on mass transfer was consistent for
all experimental settings. Among all main and interaction eects, the
eects shown in Table 2 were signicant.
From Table 2 it follows that all factors, including the two-way in-
teractive eect between electrode and material, notably aect mass
transfer. Signicant eect of the interaction between electrode and
material (E M) indicates that electrode geometry should be carefully
chosen with respect to material under drying. However, the
Table 1
Levels of the factors in 25 factorial design.
Factor Level
1 1
Voltage (kV) 14
Gap (cm) 3.5
Electrode conguration (cm) (1 1)
Air velocity (m/s) 0 1.0
Material Tissue paper
Innovative Food Science and Emerging Technologies 43 (2017) 1825
Table 2
Results of statistical analysis of reduced ANOVA model for drying rate. The signicant
main eects and interaction eect that require further multiple means comparison are
shown in bold.
Eect Symbol p-Value
Voltage V 0.001
Gap G 0.001
Electrode E 0.001
Air A 0.001
Material M 0.001
Electrode material E M 0.003
insignicance of the other interaction eects that are not included in
the ANOVA model conrm that the dierences between the two levels
of Voltage, Gap and Air shown in Fig. 6 are valid regardless of the
settings of the other factors. The relative eect of input factors on
drying rate (DR) could be seen in the tted regression model (Eq. 7)
expressed in coded units.
DR (g h) = 9.912 + 2.688V 1.038G + 1.356E + 2.763A1.30M
+ 0.731E M (7)
The coecients of Eq. 7 reect the most signicant positive eect of
air cross-ow, applied voltage and larger spacing between pins, while
eect of the gap and material roughness on drying rate was negative.
The highest drying rate, obtained with V = 19 (+1), G = 3.5 ( 1),
E = 2 (+1), A = 1 (+1) for the wet paper tissue is DRmax = 18.325 g/
h.
The plots for main and interaction eects on drying rate are pre-
sented in Fig. 6.
From Fig. 6A and C, it follows that voltage and airow are the two
most signicant factors, facilitating mass transfer. Eect of the gap on
the mass transfer was negative, which corresponds well with previous
research (Kudra & Martynenko, 2015; Lai & Lai, 2002). From Fig. 6D, it
follows that drying of wet paper tissue was more ecient than that of
sponge, which could be related to the smooth surface structure which
does not disturb so much the air boundary layer. The dierence in
drying rate was higher for (1 1) electrode and lower for (2 2 cm)
electrode. Also, the electrode with larger spacing between pins
(2 2 cm) positively aected mass transfer. Therefore, from our ex-
perimental results it follows that larger spacing between pins is pre-
ferable for EHD drying. However, optimization of electrode geometry
requires further experimental investigation.
Statistical analysis of the eect of the factors on current demon-
strated signicant eect of only two factors: voltage and type of the
material. The tted regression model using coded units is given in Eq. 8:
I ( A) = 135 + 84.6V 44.4M (8)
According to Fig. 7, the highest current would be obtained for wet
tissue paper ( 1) dried under high voltage (+1). At this setting, and
using Eq. 8, the highest current achievable was estimated to be
263.98 A. In contrast, the lowest current (94.8 A) was obtained for
sponge (+ 1) dried under low voltage (1).
The regression equation also shows the dominating eect of voltage,
which corresponds well to previously reported results (Goldman et al.,
1985; Kulacki, 1982; Robinson, 1961; Singh et al., 2012). Signicant
eect of the material type on charge transfer has never been reported.
However, we postulate that it could be attributed to the dierences in
surface properties of the two materials used in this study. Rough surface
texture of the sponge could inuence charge transfer through the in-
terfacial boundary layer. This hypothesis, however, requires careful
19 experimental verication.
4.0 The informative plot for signicant eects of material and voltage
(2 2)
on the current that have been obtained in this study is shown in Fig. 7.
Sponge It is important to note that for our experimental design, electric
current was independent of airow and electrode geometry parameters
22
A. Martynenko et al. Innovative Food Science and Emerging Technologies 43 (2017) 1825

(A) Main effect of Voltage (B) Main effect of Gap (C) Main effect of Air
Me an Drying Rate (g/h)

12.0 12.0 12.0

10.5 10.5 10.5

9.0 9.0 9.0

7.5 7.5 7.5

6.0 6.0 6.0

14 19 3.5 4.0 0 1
Voltage Gap Air

(D) Interaction effect of Electrode and Material

Me an Drying Rate (g/h)

12.0

10.5

9.0

7.5 Material
Towel
Sponge
6.0
1 2
Ele ctrode
Fig. 6. Main eects (A), (B), and (C), and interaction eect (D) of input factors on drying rate.

(spacing between pins and the gap between electrodes). This fact in- ionic wind. Considering the evident relationship between ionic wind
dicates that current is mostly limited by spatial charge in the drift re- and current (Eq. 4), drying rate versus square root of total current was
gion and interfacial charge of a boundary layer. It also suggests the plotted, which represents ionic wind for the particular geometry
primary role of ionic wind as a determinant of mass transfer from the (Fig. 9).
surface of the material. Although space charge and ionic wind were not From Fig. 9 it follows that linear relationship between drying rate
measured in our study, the relationship between measurable variables, and square root of total current is valid for all voltages gap, cong-
i.e., current and mass transfer could be depicted in Fig. 8. uration of electrode and material under drying. This relationship was
found almost linear in the range of currents from 25 to 150 A, but
3.3. Relationship between mass transfer and current trended to saturation at higher currents. High current also increases
variability in the drying rate, which resulted in larger scatter around the
The initial assumption that mass transfer is driven by ionic wind trend line. From the other side, drying rate is not sensitive to the cur-
was veried by developing the relationship between drying rate and rent below 25 A, being mostly dependent on the convective airow.

(A)Main effect of Voltage (B) Main effect of Material

225 225

200 200

175 175
Mean Current

150 150

125 125

100 100

75 75

50 50
14 19 Sponge Towel
Voltage Material
Fig. 7. Main eects of voltage and material on the current.

23
A. Martynenko et al.
Fig. 9. Relationship between drying rate and square root of electric current in the range
of voltages from 14 to 19 kV, gaps of 3.5 and 4 cm, electrodes (1 1) and (2 2).
Squares indicate experiments with wet paper tissue, circles - wet sponge. Trend lines show
signicant dierence between clusters of data with forced airow at 1.0 m/s (top) and
without airow (bottom).
It is worth noting that all data from the multifactorial experiment
fall into two distinctively dierent clusters: with and without con-
vective airow. The trend lines for these clusters are almost parallel,
which manifests that the eect of airow at 1.0 m/s on mass transfer is
additive, consistently enhancing total mass transfer by 5.05.1 g/h in
all experimental settings. These values coincide with mass transfer due
to convective airow (5.0 g/h), calculated without EHD. From this
observation it could be concluded that the eect of EHD is convective in
nature, creating additional mass transfer due to ionic wind. Hence, in
the range of the experimental design, this additive eect could be ex-
pressed by the following simple equation:
dm
= DREHD + DRC
dt
where DREHD is the EHD-induced mass transfer (g/h), and DRC stands
for convective mass transfer (g/h).
Eqs. (6) and (9) could be combined into Eq. (10):
dm
= (hEHD + hm)(m m)
dt
This equation implies that mass transfer in EHD-convective drying is
governed by the air cross-ow and ionic wind. This hypothesis, how-
ever, needs additional experimental verication for dierent airow
velocities and EHD settings.
4. Conclusions
1. There are two major driving forces for mass transfer in EHD drying:
ionic wind and forced convective airow, which are additive in
principle. EHD-induced mass transfer is convective in nature,
Innovative Food Science and Emerging Technologies 43 (2017) 1825
Fig. 8. Hypothesised driving forces for mass
transfer in EHD drying.
depending on the voltage, gap, electrode geometry, cross-ow air
velocity and material under drying.
2. Total current in EHD drying is aected by voltage and material
properties, being independent of electrode geometry and airow
rate.
3. Signicant eect of the characteristics of the material surface on
charge and mass transfer could be attributed to convective boundary
conditions at the solid-gas interface.
It should be noted that the simple linear regression in the current
research was used only for the preliminary screening of factors playing
signicant role in EHD drying. Our study demonstrated signicance of
voltage, gap, electrode conguration, air velocity, and the material on
the EHD-assisted mass transfer. However, since these results were ob-
tained from xed eects (as opposed to random eects) using ANOVA,
the predictive power of the model is limited and conclusions apply only
to the two levels, used in the experiment. The next step to explore the
major driving forces for EHD drying will require further experiments to
analyze the eects of uncontrolled factors, such as humidity, atmo-
spheric pressure, electrode surface conditions, etc., that would reveal
more general inference of random factors, not accounted in the current
experiment. Another promising approach is the method of response
surface optimization (Montgomery, 2013) to determine optimum set-
tings of voltage, gap, electrode and air velocity, which is important for
scaling-up of the EHD dryer towards industrial unit.
Acknowledgments
This research was supported by NSERC Discovery grant RGPIN
355740-08 and InnovaCorp ESCF (Nova Scotia, Canada) fund awarded
to Dr. Alex Martynenko.
References
Alem-Rajabi, A., & Lai, F. C. (2005). EHD-enhanced drying of partially wetted glass beads.
Drying Technology, 23(3), 597609.
Alemrajabi, A. A., Rezaee, F., Mirhosseini, M., & Esehaghbeygi, A. (2012). Comparative
evaluation of the eects of electrohydrodynamic, oven, and ambient air on carrot
(9) cylindrical slices during drying process. Drying Technology, 30(1), 8896.
Astatkie, T., Joseph, A. A., & Martin, R. C. (2006). A two-level un-replicated factorial
experiment to determine the eect of organic and inorganic fertilizers on dry matter
yield of permanent pasture. Renewable Agriculture and Food Systems, 21(2), 106113.
Bai, Y., Yang, Y., & Huang, Q. (2012). Combined electrohydrodynamic (EHD) and va-
cuum freeze drying of sea cucumber. Drying Technology, 30(10), 10511055.
Bajgai, T. R., & Hashinaga, F. (2001a). Drying of spinach with a high electric eld. Drying
Technology, 19(9), 23312341.
(10) Bajgai, T. R., & Hashinaga, F. (2001b). High electric eld drying of Japanese radish.
Drying Technology, 19(9), 22912302.
Bajgai, T. R., Raghavan, G. S. V., Hashinaga, F., & Ngadi, M. O. (2006).
Electrohydrodynamic drying A concise overview. Drying Technology, 7(7), 905910.
Barthakur, N. N., & Al-Kanani, T. (1989). Impact of air ions of both polarity on eva-
poration of certain organic and inorganic liquids. International Journal of
Biometeorology, 33, 136141.
Basiry, M., & Esehaghbeygi, A. (2010). Electrohydrodynamic (EHD) drying of rapeseed
(Brassica napus L.). Journal of Electrostatics, 68, 360363.
Dalvand, M. J., Mohtasebi, S. S., & Raee, S. (2013). Eect of needle number on drying
rate of kiwi fruit in EHD drying process. Agricultural Sciences, 4(1), 15.
Dinani, S. T., Hamdami, N., Shahedi, M., & Havet, M. (2014a). Mathematical modelling of
hot air/electrohydrodynamic (EHD) drying kinetics of mushroom slices. Energy
Conversion and Management, 86, 7080.
Dinani, S. T., Havet, M., Hamdami, M., & Shahedi, M. (2014b). Drying of mushroom slices
24
A. Martynenko et al.
using hot air combined with an electrohydrodynamic (EHD) drying system. Drying
Technology, 32(5), 597605.
Ding, C., Lu, J., & Song, Z. (2015). Electrohydrodynamic drying of carrot slices. PloS One,
10(4), 112.
Esehaghbeygi, A., & Basiry, M. (2011). Electrohydrodynamic (EHD) drying of tomato
slices (Lycopersicon esculentum). Journal of Food Engineering, 104(4), 628631.
Goldman, M., Goldman, A., & Sigmond, R. S. (1985). The corona discharge, its properties
and specic uses. Pure and Applied Chemistry, 57(9), 13531362.
Hashinaga, F., Bajgai, T. R., Isobe, S., & Barthakur, N. N. (1999). Electrohydrodynamic
(EHD) drying of apple slices. Drying Technology, 17(3), 479495.
Isobe, S., Barthakur, N., Yoshino, T., Okushima, L., & Sase, S. (1999).
Electrohydrodynamic drying characteristics of agar gel. Food Science and Technology
Research, 5(2), 132136.
Kamkari, B., & Alemrajabi, A. A. (2010). Investigation of electrohydrodynamically-en-
hanced convective heat and mass transfer from water surface. Heat Transfer
Engineering, 31(2), 138146.
Kawamoto, H., Yasuda, H., & Umezu, S. (2006). Flow distribution and pressure of air due
to ionic wind in pin-to-plate corona discharge system. Journal of Electrostatics, 64(6),
400407.
Kudra, T., & Martynenko, A. (2015). Energy aspects in electrohydrodynamic drying.
Drying Technology, 33(13), 15341540.
Kulacki, F. A. (1982). Electrohydrodynamic enhancement of convective heat and mass
transfer. In A. S. Mujumdar, & R. A. Mashelkar (Eds.), Advances in transport processes
(pp. 105147). New York: Wiley Eastern Ltd.
Lai, F. C., & Lai, K. W. (2002). EHD-enhanced drying with wire electrode. Drying
Technology, 20(7), 13931405.
Lai, F. C., & Sharma, R. K. (2005). EHD-enhanced drying with the multiple needle elec-
trode. Journal of Electrostatics, 63(34), 223237.
Lai, F. C., & Wang, C. C. (2009). EHD-enhanced water evaporation from partially wetted
glass beads with auxiliary heating from below. Drying Technology, 27(11),
11991204.
Li, L., Li, F., & Tatsumi, E. (2000). Eects of high voltage electrostatic eld on evaporation
of distilled water and okara drying. Biosystem Studies, 3, 4352.
25
Innovative Food Science and Emerging Technologies 43 (2017) 1825
Martynenko, A., & Zheng, W. (2016). Electrohydrodynamic drying of apple slices: Energy
and quality aspects. Journal of Food Engineering, 168(1), 216222.
Montgomery, D. C. (2013). Design and analysis of experiments (8th Ed). New York: Wiley.
Pogorzelski, M., Zander, Z., Zander, L., & Wrotniak, M. (2013). Drying kinetics of plant
material using the electrohydrodynamic (EHD) method. Chemical Engineering and
Equipment, 44(6), 552553 (in Polish).
Ramachandran, M. R., & Lai, F. C. (2010). Eects of porosity on the performance of EHD-
enhanced drying. Drying Technology, 28(12), 14771483.
Rickard, M., Dunn-Rankin, D., Weinberg, F., & Carleton, F. (2006). Maximizing ion-driven
gas ows. Journal of Electrostatics, 64(6), 368376.
Robinson, M. (1961). Movement of air in the electric wind of the corona discharge.
Transactions AIEE, 80, 143150.
Rounsley, R. R. (1985). Drying in the presence of electrostatic eld. 1984 Engineering
Conference Proceedings. Tappi Journal, 108112.
Sadek, S. E., Fax, R. G., & Hurwitz, M. (1972). The inuence of electric elds on con-
vective heat and mass transfer from a horizontal surface under forced convection.
Journal of Heat Transfer, 94(2), 144148.
Singh, A., Orsat, V., & Raghavan, G. S. V. (2012). A comprehensive review on electro-
hydrodynamic drying and high-voltage electric eld in the context of food and bio-
processing. Drying Technology, 30(16), 18121820.
Singh, A., Vanga, S. K. K., Nair, G. R., Gariepy, Y., Orsat, V., & Raghavan, G. S. V. (2017).
Electrohydrodynamic drying of sand. Drying Technology, 35(3), 312322.
Warburg, E. (1927). About silent discharge in gases. In H. Geiger, K. School, & W.
Westphal (Eds.), Handbuch der Physik (pp. 149170). Berlin-Heidelberg: Springer (in
German).
Wolny, A., & Kaniuk, R. (1996). The eect of electric eld on heat and mass transfer.
Drying Technology, 14(2), 195216.
Xue, X., Barthakur, N. N., & Alli, I. (1999). Electrohydrodynamically-dried whey protein:
An electrophoretic and dierential calorimetric analysis. Drying Technology, 17(3),
467478.
Zhang, M., Chen, H., Mujumdar, A. S., Zhong, Q., & Sun, J. (2015). Recent developments
in high-quality drying with energy-saving characteristic for fresh foods. Drying
Technology, 33(13), 15901600.