Broadband Dielectric Properties Measurement of

Some Vegetables and Fruits Using Open Ended

Coaxial Probe Technique
Ardhendu Kundu1#, Bhaskar Gupta2
Electronics & Tele-communication Engineering Department, Jadavpur University,
Kolkata - 700 032, India
[email protected], [email protected]
Abstract—Dielectric properties (i.e. permittivity and loss factor)
have been measured for some fresh vegetables and fruits using
open ended coaxial probe technique at two different temperatures
(16ºC and 25ºC). Dielectric properties have been measured for the
first time for some of the fruits and vegetables cited in this work
and other data have been verified. The dielectric properties have
been measured and plotted graphically within frequency range of
200 MHz to 8.5 GHz at 101 sample points. The permittivity and loss
factor have been presented for all the dielectric samples both at
16ºC and 25ºC.
Keywords—Dielectric properties, Fruit, Loss factor, Open
ended coaxial probe technique, Permittivity, and Vegetable.
I. INTRODUCTION
The dielectric constant or permittivity of a material reflects
the extent to which a material interacts with electric fields.
Permittivity is a measure of how an electric field affects, and
is affected by, a dielectric medium. It is determined by the
ability of a medium to polarize in response to an electric field.
Thus, permittivity relates to a material's ability to permit the
electrostatic lines of flux within that material. In real time
microwave applications, the dielectric constant and the
dielectric loss factor are of main concern. Complex relative
permittivity is defined as 􀄰 = 􀄰' – j􀄰''. The real part of dielectric
constant (􀄰') of a material is associated with the energy storage
capability within the material in the presence of an electric
field whereas the imaginary part (􀄰") is associated with the
energy dissipation within the material i.e. the conversion of
electric energy to heat energy in the material. The loss
tangent/dissipation factor is also having an important aspect
from the point of microwave application.
There are several techniques for measuring the dielectric
properties of different samples, among them open ended
coaxial probe method is a popular one for non-destructive
permittivity measurements of several samples. Open ended
probe method for dielectric measurement was pioneered by
Stuchly and Stuchly (1980). The method calculates the
dielectric properties from the phase and amplitude of the
reflected signal at the end of an open-ended coaxial line
inserted or immersed into a sample to be measured. This
technique is extremely useful in measuring the dielectric
properties of liquid, semi-solid and solid samples [1]. Hence,
this technique has been employed in this work to study the
dielectric properties of some new vegetables and fruits.
It is known that vegetables and fruits need to be stored in
storage room and during storage a reasonable portion of that
fruit/vegetable gets destroyed due to pest or bacterial attack
and inappropriate environmental conditions. So, if the
dielectric properties of stored vegetables/fruits are known,
then suitable pest and bacteria controlling methods can be
implemented with microwave dielectric heating at selective
frequencies. Now-a-days, processed coconut water and fruits
are sold in market. Normally, coconut water loses its natural
taste after few hours of extraction from coconut. So knowing
the dielectric properties of coconut water, the natural taste and
quality can be preserved for few days with controlled
microwave exposure. In similar way, food processing can also
be implemented with controlled microwave irradiation. The
correlation of the maturity effect of fruit (apple) with the
change in dielectric properties was discussed by S. O. Nelson
and his coworkers [2]. Besides all these, the SAR evaluation
of fruits and vegetables due to RF emission from cell phone
base station, AM and FM towers is also dependent on the
dielectric properties of those vegetables/fruits as cited by the
authors [3].
S. O. Nelson and his coworkers first reported the dielectric
constant of Wheat with the help of free space measurement
technique in 2002 [4]. In next few years, he reported number
of dielectric parameter studies on different fruits and crops at
various frequency bands in a broad temperature range [5-7].
Next in 2007, Nelson and some other coworkers tried to
correlate the dielectric properties of apple with its maturity
[2]. In 2008, they measured the permittivity and loss in red
winter wheat, orange, watermelon, apple, egg and chicken
meat [8-9]. In this particular paper, the work of dielectric
properties measurement of Agricultural products has been
extended with more vegetables and fruits. The dielectric
properties of apple has been verified with the published work
of Nelson [2] [4-9], whereas the dielectric properties of guava,
coconut water, tomato, brinjal have been measured with
Agilent 85070E Dielectric Probe Kit and Network Analyzer at
two different temperatures (i.e. 16º C and 25º C). An attempt
2014 International Conference on Control, Instrumentation, Energy & Communication(CIEC)
978-1-4799-2044-0/14/$31.00©2014IEEE
480
Authorized licensed use limited to: Notre Dame University - Louaize. Downloaded on
October 14,2021 at 12:38:09 UTC from IEEE Xplore. Restrictions apply.
has been made to note down the difference in permittivity and
loss factor after switching the temperature from 16ºC to 25ºC.
II. RESEARCH METHODOLOGY
A. Sample Preparation for Dielectric Properties Measurement
Various fresh fruits and vegetables have been collected
from the market. Then they have been kept separately in two
different air conditioned room respectively at 16ºC and 25ºC
temperature for more than 5 hours. In case of coconut water,
two coconuts have been kept separately at 16ºC and 25ºC for 5
hours and then the coconut water has been extracted separately
out of two coconuts for the dielectric properties measurement.
For other fruits and vegetables the extreme outer thin peal has
been removed and cut with the help of knife to get a flat
surface for accurate measurement as shown in Fig. 2.
B. Open Ended Coaxial Probe Technique For Measuring
Dielectric Properties
Open ended coaxial probe method has been used for years
as a non-destructive testing method. Open ended probe
method for dielectric measurement was pioneered by Stuchly
and Stuchly (1980). This particular technique evaluates the
dielectric properties from phase and amplitude of the reflected
signal at the end of a coaxial probe inserted or immersed into
solid, semi-solid or liquid samples for measurement [1]. For
some measurements, it may not be possible to cut out the
sample of a material for measurement. This is especially
important in the case of biological specimens to perform invivo
measurements because the material characteristics may
change. Therefore, with this method the sample can be placed
in close contact with the probe without causing any changes in
the material characteristics. Care must be exercised with this
technique because errors are introduced at very low
frequencies and at very high frequencies, as well as for low
values of dielectric constant and loss factor. Interpretation for
lower-loss materials such as fats and oils must be treated with
caution. The reflection coefficient is measured using a vector
network analyzer. The VNA with a probe system is first
calibrated so that the reflection coefficient measurements are
referenced to the probe aperture plane. The method uses
reference liquids for direct calibration at the open end of the
probe. However, the uncertainties in the measurement are due
to the uncertainties in the characterization of the reference
liquids and the selection of reference liquids as calibration
standard. In the method, all measurements are performed by
placing the standards (a short, an open and a referenced liquid)
at the end of the probe. The referenced liquid is used as a
calibration standard and must be a liquid with “known”
dielectric properties. Water, saline and methanol are usually
selected as the reference liquids. Standard one port full
calibration is then applied. The S-parameters measured on the
MUT can be post-processed to obtain the dielectric parameters
using a program.
C. Dielectric Properties Measurement Set Up
The electrical parameter measurements (like dielectric
constant and loss tangent values with respect to frequency)
required for dielectric properties determination have been
obtained with Agilent 85070E open-ended coaxial line probe
kit and Agilent Network Analyzer at the Microwave lab of
Jadavpur University, Kolkata. Agilent 85070E dielectric probe
kit contains a coaxial probe (shown in the left bottom of Fig. 1)
which can measure dielectric constant from 200 MHz to 20
GHz frequency and can withstand temperature up to 200°C.
Agilent 85070E dielectric probe kit needs to be connected with
Agilent network analyzer with a cable and also must be
interfaced with a computer to install the Agilent 85070E
dielectric probe kit software for computing the dielectric
constant and loss tangent from the reflection coefficient of the
material in contact with the active tip of the probe. As the VNA
present at JU microwave lab can measure up to 8.5 GHz, so the
dielectric properties have been measured in a frequency range
from 200 MHz to 8.5 GHz at 101 sample points on logarithmic
scale. Initially the VNA has been calibrated with an open,
short, and matched load prior to the calibration of the openended
coaxial-line probe with measurements on air, a shortcircuit
block, and distilled water.
Fig. 1. Agilent 85070E Dielectric Probe Kit (courtesy: home.agilent.com)
D. Graphical Representation of Dielectric Properties
Initially the dielectric properties like permittivity and loss
tangent values were measured for each of the fruits and
vegetables, then saved as data file at 101 points with the help
of 85070E software. Next those data files have been converted
to excel files. Then the real part of permittivity and loss
tangent data has been imported in MATLAB for plotting the
graphs.
III. RESULTS
In this work, dielectric properties measurements of
different fruits and vegetables have been performed with the
Agilent dielectric probe kit 85070E and Agilent VNA at two
different temperatures (16ºC and 25ºC). A PC has also been
used along with this set up to record the permittivity and loss
parameters. After measuring the dielectric properties when the
excel data sheets were ready then they were imported to
MATLAB software for plotting the dielectric curves for each
sample.
􀀹 Real part of permittivity vs. frequency
􀀹 Loss tangent vs. frequency
A. Dielectric Properties Measurement of Tomato
The permittivity measurement set up of a typical medium
size fresh tomato has been shown in Fig. 2. In that figure, the
high temperature dielectric constant measuring probe has been
gently pressed on the inner tissue of tomato so that there is no
2014 International Conference on Control, Instrumentation, Energy & Communication(CIEC)
481
Authorized licensed use limited to: Notre Dame University - Louaize. Downloaded on
October 14,2021 at 12:38:09 UTC from IEEE Xplore. Restrictions apply.
air gap in between them. This set up gives an accurate
measurement of dielectric constant of tomato.
As shown in Fig. 3[A], the real part of permittivity decays
almost exponentially with frequency at both 16ºC and 25ºC
temperature. It seems important to point out that permittivity
curve of tomato is consistently high at 16ºC (green curve) than
at 25ºC (red curve) over frequency range 0.5 GHz to 8.5 GHz.
Similar observations were also reported for some other fruits /
vegetables at GHz frequency scale by S. O. Nelson and his
team [5-6] [9]. The relative permittivity of tomato is around 63
at 200 MHz which falls to 40 at 8.5 GHz, irrespective of
temperatures mentioned above. This implies the energy storage
capability of tomato reduces with frequency. So, SAR due to
high frequency RF radiation would be less than SAR effects at
law frequency with same level of field exposure. The loss
tangent is having minimum value at around 1 GHz to 2 GHz as
shown in Fig. 3[B]. In either side of this frequency band, the
imaginary part of permittivity increases. So, field energy loss /
energy conversion to heat is minimum in the frequency range
of 1 GHz to 2 GHz. There is another important observation that
loss tangent curve of tomato is consistently low at 16ºC (green
curve) than at 25ºC (red curve) over frequency range from 0.2
GHz to 8.5 GHz.
B. Dielectric Properties Measurement of Brinjal
The permittivity measurement set up of brinjal is same as
shown in Fig. 2. The real part of permittivity decays almost
exponentially with frequency as shown in Fig. 4[A]. The
permittivity curves of brinjal also set a similar trend as
dielectric constant (􀄰') is higher at 16ºC (green curve) than at
25ºC (red curve) over the measured frequency range from 0.2
GHz to 8.5 GHz. The relative permittivity of brijal is 36 (25
ºC) at 200 MHz which falls to 26 (25 ºC) at 8.5 GHz. This
implies the energy storage capability of brinjal reduces with
frequency. So, SAR due to high frequency RF radiation would
be less than SAR effects at low frequency with same level of
field exposure. The loss tangent curve is having minima at 1.65
GHz as pointed out by the red dotted curve measured at 25 ºC
cited in Fig. 4[B]. Hence at 25 ºC, field energy loss is
minimum at 1.65GHz. But if the green curve (measured at 16
ºC) is considered in Fig. 4[B], minimum energy loss is
observed at 1 GHz frequency. These observations indicate
energy storage capability and energy loss within vegetable
varies with ambient temperature.
[A]
[B]
Fig. 3. Tomato [A] Real part of permittivity (􀄰') with frequency, [B] Loss
tangent with frequency
[A]
[B]
Fig. 4. Brinjal [A] Real part of permittivity (􀄰') with frequency, [B] Loss
tangent with frequency
Fig. 2. Dielectric Constant measurement set up for tomato with Agilent
85070E Dielectric Probe Kit and VNA
2014 International Conference on Control, Instrumentation, Energy & Communication(CIEC)
482
Authorized licensed use limited to: Notre Dame University - Louaize. Downloaded on
October 14,2021 at 12:38:09 UTC from IEEE Xplore. Restrictions apply.
C. Dielectric Properties Measurement of Apple
The permittivity measurement set up is same as shown in
Fig. 2. The real part of permittivity decays almost in an inverse
manner with frequency at both 16ºC and 25ºC as shown in Fig.
5[A]. It is clear from the green (16ºC) and red dotted (25ºC)
curves that relative permittivity of apple decreases as the
ambient temperature increases over the frequency range from
0.5 GHz to 8.5 GHz. The observation is similar to the
observations reported by S. O. Nelson and his coworkers which
show that the set up gives an accurate measurement of
dielectric constant [5-6] [9]. The relative permittivity of apple
is around 70 at 200 MHz which falls approximately to 50 at 8.5
GHz within the temperature range from 16ºC to 25ºC. This
implies energy storage capability of apple reduces with
frequency. So, SAR due to high frequency RF radiation is
expected to be less than SAR at low frequency with same level
of RF exposure. On other side, loss tangent curve (red dotted)
at 25ºC is having a minima at frequency band of 0.6 GHz to 1
GHz as shown in Fig. 5[B]. In either side of this frequency
band, the imaginary part of permittivity/ loss increases. So,
field energy loss / energy conversion to heat is minimum over
the frequency band from 0.6 GHz to 1 GHz at 25ºC ambient
temperature. Next the green colored loss tangent curve
(measured at 16 ºC) is considered, which shows a lower energy
loss than the red dotted curve (measured at 25ºC) over the
measured frequency scale. Surprisingly, almost no energy loss
is observed within apple in GSM 900 MHz band at 16ºC
temperature. These observations indicate energy loss within
apple decreases as the ambient temperature goes down.
[A]
[B]
Fig. 5. Apple [A] Real part of permittivity (􀄰') with frequency, [B] Loss tangent
with frequency
D. Dielectric Properties Measurement Of Guava
The permittivity measurement set up for guava is also same
as shown in Fig. 2. The real part of permittivity decays almost
in an inverse manner with frequency as shown in Fig. 6[A]. It
is the first observation in this study where relative permittivity
goes up as temperature rises from 16ºC to 25ºC in GHz
frequency scale. The relative permittivity of guava is 76 at 200
MHz which falls to the scale of 50 at 8.5 GHz within the
temperature range from 16ºC to 25ºC. This implies the energy
storage capability of guava reduces with frequency. The loss
tangent curves (both at 16ºC and 25ºC temperature) are having
minima at around GSM 900 MHz frequency band as shown in
Fig. 6[B]. Hence, energy loss / energy conversion to heat is
minimum in the frequency band around 1GHz. As pointed out
by the loss tangent curves, loss within guava slightly decreases
as the temperature goes up from 16ºC to 25ºC which also goes
against the observations reported by Nelson and his team [5-6].
[A]
[B]
Fig. 6. Guava [A] Real part of permittivity (􀄰') with frequency, [B] Loss tangent
with frequency
E. Dielectric Properties Measurement of Coconut Water
The permittivity measurement set up of coconut water
(liquid sample) is same as that shown in Fig. 1. In this case, the
permittivity sensor probe needs to be inserted totally within the
liquid without any air bubble on the surface of the dielectric
sensor. This set up gives an accurate measurement of dielectric
constant of coconut water.
The real part of permittivity decays almost in an inverse
manner with frequency as shown in Fig. 7[A]. No such
significant variation in relative permittivity is observed with
2014 International Conference on Control, Instrumentation, Energy & Communication(CIEC)
483
Authorized licensed use limited to: Notre Dame University - Louaize. Downloaded on
October 14,2021 at 12:38:09 UTC from IEEE Xplore. Restrictions apply.
the change in ambient temperature from 16ºC to 25ºC. The
relative permittivity of coconut water is around 80 at 200 MHz
which falls to 62 at 8.5 GHz. This implies the energy storage
capability of coconut water reduces with frequency. So, SAR
due to high frequency RF radiation would be less than SAR
effects at law frequency with same level of field exposure.
The loss tangent curve is having minimum value at around 2
GHz at 25ºC (1.3 GHz at 16ºC) as shown in Fig. 7[B]. Hence,
energy loss is minimum in the frequency around 2 GHz at
25ºC (1.3 GHz at 16ºC). Lower loss factor is observed at 16ºC
compared to 25ºC from 0.2 GHz to 4 GHz frequency band.
[A]
[B]
Fig. 7. Coconut water [A] Real part of permittivity (􀄰') with frequency, [B]
Loss tangent with frequency
IV. CONCLUSION
In this work, complex dielectric constant measurement of 2
vegetables and 3 fruits has been reported. Among those
vegetables / fruits, only the dielectric constant measurement of
apple has been repeated and verified with the earlier result
reported by S. O. Nelson to justify the accuracy of the
dielectric properties measurement work reported in this paper
[2] [6] [8-9]. As seen in case of coconut water, the relative
permittivity is having maximum value w. r. t all other samples
over the measured frequency band. In case of other vegetables
and fruits, the real part of permittivity is reasonably high due
to high moisture content with exception in case of brinjal
where relative permittivity is lower due to less moisture and
more air in its body. Hence, above obtained results establish
the fact that higher permittivity is expected in vegetables/fruits
with higher moisture content. If the loss tangent plot is
observed for different vegetables, fruits and coconut water,
one important observation is that the loss is minimum at
around 1 GHz to 2GHz band for all the measured samples.
This observation implies, dielectric loss is minimum at that
frequency band and RF field at that frequency band provides
an optimum angular velocity for the polar particles in those
samples due to which the friction becomes nominal to reduce
the heat generation within the vegetables and fruits. The
permittivity and loss parameter measurements of the above
mentioned fruits (except Apple), vegetables and coconut water
have been done for the first time and those measured data can
contribute to the standard list of permittivity and loss tangent.
It is observed from the results that as the temperature
rise from 16ºC to 25ºC, the relative permittivity goes down and
loss tangent goes up over measured frequency band from 1
GHz to 8.5 GHz with only exception in case of guava. These
observations are similar to the observations reported by S. O.
Nelson [5-6] [9]. It is expected that maturity level of those two
guavas were not exactly same, which resulted in an exceptional
temperature dependence dielectric and loss curves for guava.
Measured permittivity and loss parameters of several
agricultural products can be useful in employing controlled
environment RF dielectric heating for pest control at selective
frequencies without any harm to the stored vegetables and
fruits. Besides, huge cumulative RF power is radiated from
mobile phone towers to the nearby plants, crops and fruits that
might result in harmful physiological changes within the cells
/ tissues of fruits and vegetables. Hence, Specific Absorption
Rate (SAR) should be evaluated in those fruits and vegetables
to find out the real hazards in their growth. The measured
permittivity and loss tangent data may be used in SAR
evaluation of the above mentioned vegetables / fruits due to
the RF radiation coming from mobile phone towers.
REFERENCES
[1] M.S. Venkatesh and G.S.V. Raghavan, “An overview of dielectric
properties measuring techniques,” Canadian Biosystems Engineering,
Volume 47, 2005, pp. 7.15-7.30.
[2] Wen-chuan Guo, Stuart O. Nelson b, Samir Trabelsi, and Stanley J. Kays,
“10–1800-MHz dielectric properties of fresh apples during storage,”
Elsevier, Journal of Food Engineering 83 (2007), pp. 562–569.
[3] Kundu A. and Gupta B., “SAR Evaluation of Apple as per FCC RF
Exposure Guideline,” Recent Development in Electrical, Electronics &
Engineering Physics (RDE3P-2013), MCKVIE, W.B. (India), ISBN:
978-81-8424-877-7, pp. 152-156, 26-27 October 2013.
[4] A.W. Kraszewski, S. Trabelsi and S.O.Nelson, “Broadband Microwave
Wheat Permittivity Measurement in Free Space,” Journel of Microwave
Power & Electromagnetic Energy, Vol. 37 No. 1, 2002.
[5] Nelson, S.O., "Measuring dielectric properties of fresh fruits and
vegetables," Antennas and Propagation Society International Symposium,
2003. IEEE, vol.4, pp.46-49, 22-27 June 2003.
[6] Nelson, S.O., "Dielectric Spectroscopy of Fresh Fruits and Vegetables,"
Instrumentation and Measurement Technology Conference (IMTC),
Proceedings of the IEEE, vol.1, pp.360,364, 16-19 May 2005.
[7] Nelson, S.O.; Trabelsi, S.; Kays, S.J., "Correlating Dielectric Properties
of Melons with Quality," Antennas and Propagation Society International
Symposium 2006, IEEE, pp.4849-4852, 9-14 July 2006.
[8] S.O. Nelson, “Dielectric properties of agricultural products and some
applications,” RES. AGR. ENG., 54, 2008 (2): pp. 104–112.
[9] S. O. Nelson and S. Trabelsi, “Dielectric Spectroscopy Measurements on
Fruit, Meat, and Grain,” Transactions of the ASABE,American Society of
Agricultural and Biological Engineers ISSN 0001-2351, Vol. 51(5): pp.
1829-1834.
2014 International Conference on Control, Instrumentation, Energy & Communication(CIEC)
484
Authorized licensed use limited to: Notre Dame University - Louaize. Downloaded on
October 14,2021 at 12:38:09 UTC from IEEE Xplore. Restrictions apply.