Measuring the Dielectric Constant of Paper

Using a Parallel Plate Capacitor

Preliminary Communication

Vanja Mandrić Radivojević

Josip Juraj Strossmayer University of Osijek,
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Croatia
[email protected]

Slavko Rupčić
Josip Juraj Strossmayer University of Osijek,
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Croatia
[email protected]

Mario Srnović
Josip Juraj Strossmayer University of Osijek,
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Croatia
[email protected]

Goran Benšić
Mechanical Engineering School of Osijek, Croatia
[email protected]

Abstract – This article is a result of measuring the dielectric constant of a dielectric used in studying the influence of dielectrics on
the antennae reflection coefficients. A paper having a density of 0.797 g/cm3, moisture content of 0% and temperature of 210C, is
used as a dielectric. Although the literature provides a lot of data on the dielectric properties of wood and paper, without direct
measurement of the dielectric constant it is impossible to know its amount for the dielectric used in the defined frequency range.
Dielectric constant measurements are performed in the frequency range from 100 Hz to 100 kHz, while the frequency range of its
impact on the aperture antenna reflection coefficients is up to 2 GHz. The frequency range from 100 KHz to 10 GHz is interpolated
and fitted by using measurements and available literature data and by respecting physical influences and phenomena and
functional changes of the dielectric constant of paper within the given range.

Keywords – dielectric constant, loss factor, parallel plate method, dielectric constant measurement

1. INTRODUCTION dielectric properties of wood was carried out by Skaar

in 1949 [2]. In this paper, we present that the
Paper is made from mechanically treated wood so dielectric constant of wood increases as the moisture
that water can penetrate, soften the fibers and content in- creases, and it decreases with increasing
make them more flexible. This wood processing frequency of the applied field. In 1953, Kroner and
splits and frays the fibers, producing microscopic fibers Pungs [3] deter- mined that the loss tangent depends
that con- tribute to the density and tensile strength of on the moisture content and had a complex form.
the manu- factured paper [2]. The electrical
Further studies (by James in 1975 [4], Gaikwad in
parameters of paper depend on a large number of
1981 [5], Torgovnikov in 1993 [6] and Kempf in 1997
paper parameters: wood types and paper production [7]) dealt with various influences on the dielectric
methods, density, humid- ity, temperature and char- acteristics of the material (moisture, additions,
frequency. Each of these parame- ters can take a large material density, structural directions, frequency,
range of values resulting in a large range and dispersion temperature, etc.) and showed that these variables
of electrical parameters of paper. also have an im- portant influence on the dielectric
The earliest study dealing with the dielectric proper- behavior of wood.
ties of the material was conducted by Debye in Dielectric constant measurement methods cover a
1929, in which he introduces the term polar number of studies and research. We would like to em-
dielectrics [1]. One of the earlier measurements phasize the following: Grove in 2004 [8]
and studies of the (Determining
Volume 9, Number 1, 2018 1
dielectric constants using a parallel plate capacitor);
The real part of the complex dielectric constant   is
Venkatesh in 2005 [9]; Ganchev in 2006 [10] and
a measure of how much energy from an external
Teresh- chemko et al. in 2011. The main objective of
field is stored in a material.
this paper is to use parallel plate methods of
measurement and different models and methods of The imaginary part of the complex dielectric con-
calculating dielectric parameters and dielectric losses stant   is a measure of how dissipative or lossy a
and their applicability to different materials. ma- terial is to an external field. The dielectric loss
angle  is the angle between the real part of the
A large number of studies have been carried to de-
dielectric con- stant and the absolute value of the
termine the dielectric constant of wood and paper,
e.g. physical testing of paper [11] by Mark et al. in dielectric constant, and the loss tangent tan  is
defined as the ratio of the imaginary part to the real
2001 deal- ing with the dielectric constant of papers
and different cellulosic materials. Studies by Sirvio [12] part of the complex dielectric constant (Fig. 1) [14].
and Omari [13] in 2016 provide research and The dielectric constant  is not constant by chang-
measurement of different types of paper for a number ing either the frequency or the temperature. Frequency
of new applications. dependence is important for the considerations and
In this paper, we used the parallel plate method analyses performed for the purpose of this paper.
for measuring the dielectric constant of paper which One of the most important intrinsic properties of
in- volves placing a paper sample between two paper as a dielectric material associated with the
capacitor plates and measuring the resulting fre- quency is its polarization ability. Polarization is a
capacitance and other parameters used an quan- tity which characterizes the material
equivalent capacitor model such as equivalent serial polarization ef- fect under the influence of the
inductance, equivalent serial and parallel resistance, external electric field. The polarization effect is
dissipation factor D=tanδ, the angle between the caused by the change in the spatial arrangement of
current and the voltage θ and ca- pacitor Q factor. electrically charged particles of the wood substance
In addition, linear interpolation of results under the influence of an external electric field that
available from literature is derived for the frequency changes with time and excites an al- ternating electric
range from 100 Hz to 10 GHz. current in the wood (paper) (Fig. 2).
Under this condition, wood (paper) acquires an elec-
2. THE DIELECTRIC CONSTANT (PERMITTIVITY) tric moment [6].
The dielectric constant is equivalent to relative per-
mittivity εr or absolute permittivity ε relative to
permit- tivity of free space ε0. The dielectric constant
is gener- ally a complex number which describes the
interaction of a material with an electric field:

     j    (1  j tan  )   e
j (1)
,
(2)
   r  0 ,
(3)
    tan  ,

where   is a real part of the dielectric constant or

rela- tive dielectric constant; a) b)
  is an imaginary part of the dielectric Fig. 1. a) Vector diagram of the complex dielectric
constant or dielectric loss factor; constant  ; b) Vector schematic diagram of
 is a dielectric loss angle (Fig. 2a); and currents in wood under the action of an external
alternating current field E: J* - total current; Ja
active component
tan  is loss tangent or the of the current; Jrc - reactive component of the current;
dissipation
factor DF. Je - conduction current; Jm - instant displacement
current; Jr - relaxation displacement current; J’ -
The dissipation factor is often used active component of the displacement current; δ -
interchangeably with the term power factor which is
dielectric loss angle; θ - phase displacement angle
approximately equal (Fig.2).
[6].
For a good dielectric      (Fig. 2b):
  energy loss per cycle

2 International Journal of Electrical and Computer Engineering

 The geometric sum of the electric field dielectric multiplied by the electric constant and of
strength vec- tor at a given point of the the polarization vector gives
cos  tan    . (4) the electric displacement (electric induction) vector D
 stored per cycle at this point [11]:
energy

Volume 9, Number 1, 2018 3

 → → → dipole polarization deposit;   - interfacial polarization
D0EP, (5) v
→ deposit;  z - electrolytic polarization deposit; lgf -
wher is electric field strength; loga- rithm of frequency value [6].
→
E
e If the relaxation time τ of dipoles of the dielectric is
is polarization of the material; and
P the same, then the connection between the complex
 0 is the dielectric constant of vacuum. dielectric constant and the frequency is to be found
from the Debye equation:
The connection between the relative dielectric con-
stant and the polarization of the material is as follows: (8)

P  (  1) 0 where   is the limiting high-frequency relative

E→, 
(6)
dielec- tric constant;
where   is the relative dielectric constant of the  s the limiting low-frequency relative dielectric
material.
constant (static);

Unpolarized Polarized ω the angular frequency; and

τ the macroscopic time of relaxation [6].
The polarization process depends on the time t as follows:

P  P 0et /  (9)
Fig. 2. Representation of material polarization
in the presence of an EM field [15].

Summary polarization of wood (and paper as well) P,

includes five kinds of polarization, which take place in
moist heterogeneous dielectrics (Fig. 3):

P  Pe  Pa  Pd  Pv  (7)
Pz .

Electronic polarization Pe arises as a result of the to the ε’ value is as follows: e - electronic polarization
shift of electron orbits relative to the positively deposit;  a - ionic (atomic) polarization deposit;  d -
charged nu- cleus under the influence of an external
electric field.
Ionic (atomic) polarization Pa arises as a result of
an elastic displacement of atoms in the molecules as
well as due to a mutual displacement of charged ions
of op- posite signs in substances with ionic bonds.
Dipole (orientational) polarization Pd consists of
the rotation of dipole molecules in the direction of
an ex- ternal electric field.
Upon application of an electric field to heteroge-
neous dielectrics, free electrons and ions start moving
within each element’s volume, and as a result, the ele-
ment receives a dipole moment -> interfacial polariza-
tion Pv.
The material always has residual electrochemical
or electrolytic polarization. At the instant of imposition
of an electric field, an increase in electrolytic
polarization Pz is observed. The increase of this kind of
polarization is much slower than that of other types of
polarization considered above. Electrolytic
polarization of wood can be characterized by its time
constant, which is ap- proximately equal to 10-4-102 sec.
[6].
The contribution of different kinds of polarization

4 International Journal of Electrical and Computer Engineering

Fig. 3. Vector schematic diagram of dielectric
constant ε’ and loss factor ε” of wood versus
frequency response characteristics with
different kinds of polarization [6].

Fig. 4. Debye relaxation equation

(8) for water at 300C [10].

Volume 9, Number 1, 2018 5

 The resonant frequency fc is identified by a resonant The method uses a frequency range up to 1.8 GHz,
response in ε’ and a peak of maximum absorption in ε’’ has high measurement accuracy and involves a
(fc = 22 GHz for water at 300C in Fig. 4). very simple sample preparation and setup. This
Figure 4 presents a dashed part of the diagram in method is also best for thin flat sheets of samples. The
Fig- ure 3. It is focused on this dielectric constant dielectric constant is derived by knowing the
measure- ment and interpolation. dimensions of the material and plates (width,
length, thickness) and by measuring its capacitance
3. PARALLEL PLATE METHOD - CAPACITOR and dissipation factor.
METHOD
The dielectric constant is calculated as follows:

The measurement methods of dielectric material td  C p (10)

properties depend on physical and electrical param-    S  0 ,
eters of the dielectric material to be measured, the films
fre- quency of interest, and the degree of accuracy
required. The summary of techniques for measuring the * [10]
dielectric properties of materials are presented in
Table 1.
The parallel plate method uses a parallel plate
capaci- tor as a sample holder and auxiliary
parameter being measured, with the material under
the measurement be- tween plates. This method
requires instruments for mea- suring capacitor
impedance (LCR meter) or an impedance analyzer. The
measured capacitance and dissipation fac- tor is then
used to calculate the real and imaginary parts of the
dielectric constant. By using this method, all pa-
rameters of the equivalent model shown in Figure 5
were measured (Ls, Rs, Cp and Rp). In addition,
measurements included measurement of the
dissipation factor D=tanδ, the angle between the
current and the voltage θ and the capacitor Q factor
(Figures 4 to 6).

Table 1. Summary of techniques for measuring the

dielectric properties of materials [16].
Frequency

Measuring
parameter

Character-
range

Method Figure istics and

applications

Accurate, best
Parallel up to 1.8 for low frequen-
εr
plate MHz * cies flat, thin
sheet

Broadband best
Coaxial 100 MHz for lossy mat.,
εr
probe to 50 GHz liquids or semi-
solids

Broadband
best for lossy to
Transmis- 100 MHz
ε r , µr low lossy mat.,
sion line to 60 GHz
machine able
solids

Broadband
Non-contacting
Free 1GHz to
ε r , µr
space 100GHz best for high
temp., large, flat
samples

Accurate, best
for low loss
Resonant 3 GHz to
ε r , µr mat., small
cavity 30 GHz
samples,
substrates, thin

6 International Journal of Electrical and Computer Engineering

 where t is the average thickness of sample (m),
S is guarded electrode’s surface area (m2), 𝜀 stands
for permittivity of free space (8.854 × 10−12 F/m), and
𝐶𝑝 corresponds to the equivalent parallel
capacitance of sample (F) [13].

a)

b)
Fig. 5. a) Parallel plate capacitor with paper as a
dielectric (the diameter of the plates 2a is 20 cm);
b) an equivalent circuit for a capacitor.

In Fig. 5, Ls is equivalent series inductance, Rs is

equiv- alent series resistance, Rp is equivalent
parallel resis- tance and Cp is equivalent parallel
capacitance.

4. EASUREMENT AND MEASUREMENT RESULTS

4.1. Measurement
The dielectric constant and the loss factor were mea-
sured by using the parallel plate method. Dielectric
con- stant and loss factor measurements were
performed as part of antenna aperture reflection
coefficient mea- surements using different materials
as radomes for this antenna [15], [17]. Thus, there was
a need for infor- mation on the dielectric constant of
wood and wood- based materials (paper) that are
analyzed in this paper.
The weight of a sheet of paper used as a sample
was 4.9719 g and the density of 0.797 g/cm3.

Volume 9, Number 1, 2018 7

 The measured value of capacitance may be poses a problem. This problem appears when the di-
thought of as two capacitors Cp and Cd in series; one is
electric constant and the loss factor in a given frequen-
a perfect parallel capacitor filled with the dielectric cy range or at specific temperature are not available.
material Cd and the other capacitor is a perfect
In this paper, equation (8) available from literature
parallel capacitor filled with air Cair: does not help much because of frequency
1 1 1 1 t t discrepancy be- tween the experimental and
theoretical determina-
   tion and the lack of measured (precise) values of the
Cair  S    0 S  
d air
Cmeas Cp Cd (11)
interconstants of the particular observed sample (  
0 is the limiting high frequency relative dielectric
constant,
where 𝑡 is the thickness of the dielectric material (m),  s is the limiting low frequency relative dielectric con-
𝑡air is the dthickness of unwanted air gaps (m) (Fig. 6), S stant (static); τ is the macroscopic time of relaxation
is
the capacitor surface area (m2), is the permittivity of time constants, appearing in this equation [5]).
𝜀 0

free space (8.854 × 10 F/m), and ε’ is the permittivity

−12

of the dielectric material: 4.2. Results from available literature


  The analysis performed in this section is based
  on the measurements of the dielectric constant and
1  (12) loss factors of wood and wood-based materials
td  (paper)
1 tair  S in [4], [6] and [13]. [4] deals with measured values of
  0 the dielectric constant and loss tangent of white oak,
C S

 p 0  dielectric.
The capacitor filled with air (Cair) is an equivalent ca-
pacitor of all air gaps between capacitor plates (Fig. Measurement equipment in these measurements
6). The influence of unwanted air gaps on the total consists of a circular shape capacitor of 8 cm in diam-
capac- ity as well as on the dielectric constant and the eter and LRC meter UNI-T, type UT612.
loss fac- tor is increased with an increase in paper Determining the dielectric constant value of paper at
layers and the thickness of the dielectric (Figures 7 and specific temperature and frequency range sometimes
8).

a)

b)
Fig. 6. Capacitance measurement (a) thin
dielectric
– single sheet paper dielectric, (b) thick multilayer

8 International Journal of Electrical and Computer Engineering

Douglas-fir, and four commercial hardboards at fre-
quencies from 20 Hertz (Hz) to 50 Megahertz
(MHz), moisture conditions from oven dry to
complete satu- ration, temperatures from -20°C
to +90°C, and for the natural wood, with the
electric field aligned with three principal
structural orientations. Oven-dry wood is ob-
tained by using the specimens originally
conditioned and tested in equilibrium with 30
percent relative hu- midity. These specimens
were dried in a vacuum oven at 60°C. Table 2
presents the real part and loss tangent of the
dielectric constant as a function of frequency for
tangential oak of density 0.74 g/cm3 and
temperature ranging between 5 and 250C.

Real part of dielec- Tg δ

Frequency tric constant
50C - 250C 5 C - 250C
0

100Hz 2.50 - 2.60 0.050 - 0.0076

1kHZ 2.30 - 2.30 0.033 - 0.021

10kHZ 2.20 - 2.20 0.035 - 0.033

100kHz 2.10 - 2.20 0.038 - 0.040

1MHz 2.20 - 2.20 0.047 - 0.046

10MHz 2.10 - 2.20 0.052 - 0.054

Table 2. Real part and loss tangent of the

dielectric constant of density 0.74 g/cm3 and
temperature
5-250C [4].

[6] deals with dielectric parameters of wood

and wood based materials.
From this reference, for the purpose of
comparing measured values and, more
importantly, for interpo- lation purposes,
measured data of the dielectric con- stant and
the loss factor for oven-dry paper of density
0.80 g/cm3 and temperature 20-250C with vector E
ori- ented perpendicularly to the sheet surface,
were used.

Volume 9, Number 1, 2018 9

The frequency range of the dielectric constant and the Table 4. Real and imaginary parts of the dielectric
loss factor is from 100 Hz to 10 GHz. The term oven-dry constant as a function of frequency for agava
wood (or paper) implies wood dried to a constant americana paper with a hand sheet characteristic:
mass in the air at a temperature equal to + 103 ±2 0C hand sheet weight: 1.23 g, fines content: 46.55%,
[6]. average fiber length: 0.52 mm, porosity: 7637 mL/
Using the data from the above reference, the min, roughness: 10.93 𝜇m, crystallinity rate: 44.05 %
inter- polation of the dielectric constant value and and young modulus: 715.50 MPa [13].
the loss factor for the frequency range above the
Real part of Imaginary part of
measurement range was performed. Frequency
dielectric constant dielectric constant
Table 3 presents measured values of the dielectric 100Hz 2.77 0.99
constant and the loss factor for oven-dry paper of den-
sity 0.80 g/cm3 and temperature 20-250C with vector E 1kHZ 2.50 0.33
oriented perpendicularly to the sheet surface based on
data from [6]. 10kHZ 2.20 0.15
From [13], which deals with the dielectric
constant of paper made of pulp of several materials 100kHz 2.00 0.09
based on natural products, the values of the dielectric
constant and the factor of losses for agave paper were 1MHz 1.80 0.06
used. The agave (agave americana) hand sheets were
prepared according to a standard papermaking 10MHz 0.80 0.01
method (TAPPI T205 sp-954). This paper sheet is
heterogeneous (cel- lulose, lignin, fines, and other
impurities related to ex- traction processes of fibers The dielectric measurements were carried out at
and the origin of the plant) and hydrophilic, with room temperature with an impedance meter (Agile-
rough surfaces and consequently a lack of flatness, nt4294A) using a cell (Agilent 16451B) allowing
necessary for a good contact with the electrodes [13]. direct measurement. The frequency range was
between 100Hz and 10MHz. The samples were dried
in an oven at 1050C to remove residual water [13]. The
Table 3. The dielectric constant and the loss
dielectric constant is calculated by using equation (10).
factor of oven-dry paper as a function of
frequency with density 0.80 g/cm3 and
4.3. Measurement results and
temperature 20-250C with vector E oriented
interpolatione
perpendicularly to the sheet surface [6].
The standard SMART LINE OFFICE A4, 80 g/m2,
ε’ Tgδ ε’’= ε’·tgδ MONDI paper was used as a sample paper for
Frequency measurement. There are no information on dielectric
20 C - 25 C
0 0
20 C - 25 C
0 0
200C - 250C
properties for this paper. Since there is an extremely
large number of differ- ent types of paper with
100 1.80 – 3.30 0.0030 – 0.012 0.0054 – 0.0396
different parameters, it was not possible to find
information about the paper used. Mea- surements
1kHz 1.80 – 3.30 0.0030 – 0.015 0.0054 – 0.0495
were made with the available equipment and they
provided the measured data. Measurements were
10kHz 1.80 – 3.20 0.0044 – 0.021 0.0079 – 0.0672 performed as part of measurements done for the an-
tenna aperture reflection coefficient using paper as a
100kHz 1.80 – 3.10 0.0067 – 0.031 0.01206 – 0.0961 ra- dome for this antenna. The frequency range of the
used aperture antenna is significantly greater than the
1MHz 1.80 – 3.00 0.010 – 0.043 0.0180 – 0.1290 scope of measurement and for that range of the
frequency band the interpolation of the data from
10MHz 1.80 – 2.80 0.014 – 0.053 0.0252 – 0.1484 literature com- plies with the measured data.
Wood is a raw material for paper building. An
100MHz 1.70 – 2.60 0.013 – 0.046 0.0221 – 0.1196 oak sample was used in [4] to determine the dielectric
con- stant and loss factors. The dielectric constant and
1GHz 1.60 – 2.20 0.011 – 0.032 0.0176 – 0.0704 the loss factor of paper used in these measurements
and oak were compared [4] (with similar parameters,
10GHz 1.60 – 2.10 0.011 – 0.030 – 0.0630 i.e. hu- midity, density and temperature), as illustrated
in Fig. 9. It is evident that the dielectric properties of
these two materials are very similar.
A variation of the dielectric constant and loss In the second part, the results obtained by using
factors with frequency for different papers are measurements of the dielectric constant and the loss
represented in Table 4. factor of paper are compared with experimental values

1 International Journal of Electrical and Computer Engineering

 from [6] and [13] and presented in Fig. 10 and Fig. 11.

Volume 9, Number 1, 2018 1

1
 Figure 10 shows measured values of the dielectric the data from Table 3 for temperature 21.3°C.
constant and loss factors together with literature data
[6] for paper of the same density, i.e. 0.8 g/m3.
Mea- surement of dielectric parameters is performed
in the frequency range from 102 to 105 Hz, while
the data from literature cover the full range of
frequencies (102
- 1010 Hz). The data from 6 were interpolated for the
ap- proximate temperature at which measurements
were performed (21.3oC) and compared with the
measure- ments. For the purpose of comparing data to
the extent not covered by the measurements,
measured data are fitted using the exponential
function (equation (13)).
In this diagram, it can be seen that measured (and
fitted) data of the real part of the dielectric constant,
using the parallel plate method, are very close to
the values mentioned in the references. It is
therefore as- sumed that for the non-available
frequency band the extrapolation of known values
can be interpolated while maintaining the same
conditions and param- eters of the materials used.
The same procedure is performed for the data
shown in Fig. 11, only by using the interpolated data
of the previous analysis. The reason is that there is no
data in- terpolated frequency band in [13].
What is still remarkable is that measured data on the
dielectric constant and the loss factor of the measured
paper and paper derived from the agave match very
well, but there is a visible discrepancy in the curve
passing through measured data and showing an
expo- nential tendency to “saturation” and
maintaining a con- stant value in the frequency range
from 105 to 1010 Hz.

Table 5. Measured values of equivalent capacitor

circuit parameters as a function of frequency for a
paper dielectric sheet of density 0.797 g/ cm3 (for
a 5 g paper sheet) and temperature 210C.
Paper
thickness 0.1 1 10
[mm]

f Cp Cp Cp
D D D
[kHz] [pF] [pF] [pF]

0.1 7932.0 0.214 911.0 0.219 166.0 0.392

1 6631.0 0.095 764.6 0.094 114.5 0.205

10 6125.0 0.041 707.2 0.042 96.4 0.087

100 5878.0 0.023 678.8 0.023 88.2 0.049

Table 6 presents linear interpolated values of the di-

electric constant and the loss factor of oven-dry paper
of density 0.80 g/cm3 and temperature 21.30C with
vector E oriented perpendicularly to the sheet surface
based on data from [6]. The interpolation is derived for
1 International Journal of Electrical and Computer Engineering
 Table 6. Interpolated values of the dielectric dielectric constant increases with increasing of the
constant and the loss factor as a function of dielectric thickness, i.e. an increase in unwanted air
frequency for oven-dry paper of density 0.80 gaps as well
g/ cm3 and temperature 21.30C with vector E
oriented perpendicularly to the sheet surface
[6].

Frequency ε’ tgδ ε’’= ε’·tgδ

100Hz 2.19 0.00534 0.01169

1kHz 2.10 0.01860 0.03906

10kHz 2,06 0.00872 0.01799

100kHz 2.14 0.006084 0.01302

1MHz 2.11 0.008806 0.01858

10MHz 2.06 0.011718 0.02414

100MHz 1.93 0.011181 0.02158

1GHz 1.76 0.009352 0.01646

10GHz 1.73 0.009139 0.01594

Table 7. Interpolated values of the dielectric

constant and the loss factor as a function of
frequency of tangential oak, the dielectric
constant and loss tangent of density 0.74
g/cm3 and temperature 210C [4].

Frequency ε’ tgδ ε’’= ε’·tgδ

100Hz 2.58 0.18318 0.47260

1kHz 2.30 0.04370 0.10051

10kHz 2,20 0.07260 0.15972

100kHz 2.18 0.08720 0.19009

1MHz 2.20 0.10120 0.22264

10MHz 2.18 0.11554 0.25188

Table 7 presents linear interpolated values of

the di- electric constant and the loss factor of
tangential oak, the dielectric constant and loss
tangent of density 0.74 g/cm3 and temperature
210C.
The parallel plate method is suitable for thin
dielec- trics and deviations and errors due to
unwanted air gaps are the smallest in the
dielectric constant and loss factors for dielectric
thickness of 0.1 mm.
Therefore, if the values of the real and the
imaginary part of the dielectric constant of the
tiny dielectric (0.1 mm) are taken as a reference,
the deviations and errors of these parameters at
two thicknesses (1 and 10 mm) can be determined.
These errors are listed in Table 8.
It can be seen that the relative error of the
Volume 9, Number 1, 2018 1
3
as the dispersion of the electric field outside the space
between the capacitor plates.

Table 8. Relative errors of the real and the

imaginary part of the dielectric constant with
the dielectric parameters of smallest thickness
dielectric (0.1 mm) as a reference.
Freq. 1kHz 10kHz 100 kHz
Ima
t Real Real Ima Real Ima
d Err Err Err Err Err
mm % Err
% % % %
%

0.1 -1.,49 -62.93 -22.81 -85.21 -25.96 -92.043

1 -16.16 -64.01 -22.56 -85.15 -25.61 -92.1871

10 -30.99 -63.91 -42.04 -87.14 -46.90 - Fig. 9. Measured, linear interpolated and fitted
Fig. 7. A plot of the real part of the dielectric dielectric constant (real and imaginary part) as a
function of frequency for 0.1 mm paper sheet and
measured dielectric constant of oak (tangential)
[4].

The number and size of unwanted air gaps are too

large for a large number of paper sheets to cause
sig- nificant errors when using the parallel plate
method. Therefore, a single sheet of paper was
taken as rele- vant, and the results of the
comparison with the mea- sured values in the
literature confirm this claim. The measured values of
the dielectric constant and the loss factor when
compared with the measurements in the literature
show good value matching.

constant versus the dielectric thickness for paper

sheet with density 0.797 g/cm3 at 210 C calculated
using equations (10) and (3) from measured
values presented in Table 5.

Fig. 10. Measured, linear interpolated and fitted

dielectric constant (real and imaginary part) as a
function of frequency for 0.1 mm paper sheet and
measured dielectric constant of the oven-dry sheet
Fig. 8. A plot of the imaginary part of the dielectric
paper [6].
constant versus the dielectric thickness for paper
sheet with density 0.797 g/cm3 at 210 C calculated
Measured data as well as measured data from
the literature are fitted and inserted in Figures 9 to 11.
Ex- ponential fitting of results was performed according
to the formula:
using equations (10) and (3) from measured y0
values presented in Table 5. y  A exp(x / t) 

1 International Journal of Electrical and Computer Engineering

 (13)

Volume 9, Number 1, 2018 1

5
 entire range of frequencies of interest (from 100 Hz to
10 GHz). The obtained values are compared with linear
interpolated and exponential fitted values of different
paper of the same density, temperature and initial val-
ues of the dielectric parameters at the frequency range
from 100 Hz to 10 GHz from the available literature.
This comparison showed relatively small deviations
of the real part of the dielectric constant of up to
17.5%. Bigger deviations of the imaginary part of the
dielec- tric constant are the result of very low
measured values as well as the imprecision of the
measured equipment and the inadequacy of the
exponential fitting in that frequency bandwidth. The
measured values of the di- electric constant and the
loss factor correspond well to the measurements in
the literature. The data thus obtained are necessary
Fig. 11. easured, linear interpolated and fitted for the simulation and mea- surement of the
dielectric constant (real and imaginary part) as a reflection coefficient of the antenna aperture with a
function of frequency for 0.1 mm paper sheet and radom made of materials of similar di- electric
measured dielectric constant of the papers with fibers characteristics such as paper and paper-based
from agave [13]. materials. Important applications of paper could be
shown as dielectric materials in the construction of
Maximum deviations from the fitted result of mea- radomes. This will be further elaborated in our future
surement and measurement results according to the work related to this paper.
literature at the frequency of interest (2 GHz) are pre-
sented in Table 9.
6. REFERENCES
Table 9. Relative errors of the real and the [1] P. Debye, “Polar molecules”, Chemical Catalog
imaginary part of the dielectric constant of fitted Company, 1929.
measured results and measured results from
literature [2] C. Skaar, “The dielectric properties of wood at
Real part error Imag. part error sev- eral radio-frequencies”, Technical Bulletin, Vol.
Dielectric constant
% %
69, 1949.
Tangential oak paper [4] -5.86 +79.85
[3] K. Kroner, L. Pungs,“Dielectric loss factor in
Dry-oven paper [6] -17.24 -49.02 natural wood at high frequencies”, Holyforsch.,
Vol. 7, No. 1, 1953, pp. 12-18.
Agava paper [13] - +11.28
[4] W. L. James, “Dielectric properties of wood and
The deviations of the real part of the dielectric con- hard board: variation with temperature
stant do not exceed 18%, while the deviations of the moisture and grain”, U.S. Department of
imaginary part are significantly higher. This can be ex- Agriculture Forest Service, Forest Products
plained by the very low values of the imaginary part
Laboratory Madison, Wis- consin, USA, 1975.
and hence the greater probability of measurement
and equipment inaccuracy. Furthermore, it is highly [5] M. G. Gaikwad, Correlation of positronium life-
likely that the exponential fitting model is not
times and dielectric constants in organic
appropriate for an area of about 2 GH for the
imaginary component of the dielectric constant. liquids, Savitribai Phule Pune University, Pune,
India, PhD Thesis, 1981.

5. CONCLUSION [6] G. I. Torgovnikov, “Dielectric Properties of Wood

In this paper, the parallel plate method is used to and Wood-Based Materials”, 1st Ed., Springer
mea- sure capacitor parameters with paper as a Ver- lag, 1993.
dielectric.
[7] U. W. Kempf, “Microsphere containing circuit
Based on the measurement, in the frequency
range from 100 Hz to 100 kHz, the values of the real board paper”, US6042936 patent FIberMark, 1997.
and the imaginary part of the dielectric constant of
[8] T. T. Grove, M. F. Masters, R. E. Miers,
paper are calculated. Furthermore, the fitting of
measured data was made by using the exponential “Determining dielectric constants using a
function for the parallel plate capaci-
1 International Journal of Electrical and Computer Engineering
 tor”, American Association of Physics Teachers, [14] G. L. Johnson, Solid State Tesla Coil, http://hot-
Vol. 73, No. 1, 2005, pp. 52-56. streamer.deanostoybox.com/TeslaCoils/OtherPa-
[9] M. S. Venkatesh and G.S.V. Raghavan,“An pers/GaryJohnson/tcchap1.pdf (accessed: 2018)
overview of dielectric properties measuring [15] G. J. Meyer, Polyurethane Foam: Dielectric Mate-
techniques”, Canadian Biosystems Engineering, rials for Use in Radomes and Other
Vol. 47, 2005, pp. 15-30. Applications, General Plastics Manufacturing
[10] S. Ganchev, “Materials Measurement Techniques Company, https://
& Applications“, Americas TM Contact Center, www.generalplastics.com/wp-content/up-
Agi- lent Technologies, USA, 2006. loads/2016/12/White-Paper-PU-Foam-Dielectric-
Materials-for-Use-in-Radomes-and-Other_Appli-
[11] R. E. Mark, C. Habeger, J. Borch, M. B. Lyne,
cations.pdf (accessed: 2018).
“Hand- book of Physical Testing of Paper:
Volume 1 and 2”, 2ndEd., CRC Press, 2001. [16] O. V. Tereshchenko , F. J. K. Buesink, F. B. J.
Leferink, “An overview of the techniques for
[12] P. Sirvio, “Electrical and Dielectric Properties of measuring the dielectric properties of materials“,
Un- coated and Coated Wood-Free Paper for Proceedings of the 2011 XXXth URSI General
Electro- phozography”, Faculty of Science and Assembly and Scien- tific Symposium, Istanbul,
Engineering Ĺbo Akademi University Abo, Turku, Turkey, 13-20 August 2011, pp. 1-4.
Finland, PHD Thesis, 2016.
[17] S. Rupcic, Z .Sipus, V. Mandric, “Input
[13] H. El Omari, A. Zyane, A. Belfkira, M. Taourirte, admittance of Circular Waveguide Opening
F. Brouillette, “Dielectric Properties of Paper Mounted on Dielectric-Covered Grounded
Made from Pulps Loaded with Ferroelectric Spherical Surface”, Proceedings of the
Particles”, Journal of Nanomaterials , Vol. 2016, International Symposium EL- MAR, Zadar,
10 pages. Croatia, 28-30 September 2012, pp. 317-320.

Volume 9, Number 1, 2018 1

7