1392

A. Rashidian et al.: Dielectric Characterization of Materials using a Modified Microstrip Ring Resonator Technique

Dielectric Characterization of Materials using a Modified

Microstrip Ring Resonator Technique
Atabak Rashidian, Mohammadreza Tayfeh Aligodarz and David M. Klymyshyn
University of Saskatchewan
Department of Electrical and Computer Engineering
Saskatoon, SK S7N 5A9, Canada
Telecommunications Research Labs of Canada (TRLabs)
111, 116 Research Drive
Saskatoon, SK S7N 3R3, Canada

ABSTRACT
The goal of this study is to present a simple model based on the ring-resonator
technique to measure nondestructively the permittivity and loss tangent of dielectric
materials. The proposed measurement model utilizes a modified ring-resonator
technique in one-layer and two-layer microstrip configurations. This method eliminates
the requirement to metalize the samples and enables characterization of permittivity
and dielectric loss from 2 to 40 GHz. The effects of conductor and radiation losses that
may introduce significant errors in the calculation of the loss tangent, especially at very
high frequencies, are minimized. The measurement precision is evaluated by
comparing the results with those obtained by using two well-known standard
techniques. Uncertainties associated with the proposed model are addressed.
Index Terms Dielectric materials, dielectric measurements, ring resonators,
microstrip, permittivity measurement, dielectric loss tangent measurement.

1 INTRODUCTION
EXTENSIVE research was performed to measure dielectric
parameters and consequently introduce novel functional
materials in the past few years [1- 3]. Dielectric characterization
is a critical part of the qualification of materials for a wide range
of applications. The knowledge of the dielectric constant and loss
tangent is necessary for the design, analysis, and fabrication of
antennas and microwave circuits [4, 5]. By measuring dielectric
properties of agricultural and food products one can make very
useful interpretations regarding the quality of these products [6,
7]. Accurate determination of soil moisture distribution, which is
realized by measuring dielectric constant of soil, plays an
important role in the hydrologic, ecologic, and climatic models
[8]. Because of the variation of tissue permittivity, especially
between malignant and normal tissues, dielectric characterization
is also used in medical applications [9, 10].
Proposed by Troughton to measure dispersion of microstrip
transmission lines [11], microstrip ring resonators have attracted
considerable interest for electrical properties measurements of
dielectrics even at very high frequencies [12, 13]. The advantage
of this technique, which uses the resonance frequency and the Q
factor of the resonator to calculate the permittivity and the loss
tangent of the substrate, is its simplicity and accuracy due to the
absence of end-effects and high Q factor. Numerous analytical
and numerical models are presented in the literature to increase
Manuscript received on 14 October 2011, in final form 15 February 2012.

the accuracy and simplicity of this technique [14, 15]. However,

this technique requires dielectric samples in the form of metalcladded substrates, with the microstrip ring pattern etched or
machined on the metal layer. Because of the low adhesion of
metal to some materials (e.g. typical photoresists), and also the
problems in measuring materials in non-substrate forms (e.g.
powders and liquids), several dielectric characterization methods
based on the modified ring resonator technique are presented in
the literature. For instance the permittivity of liquid solvents is
determined based on the relationship between the resonant
frequency and permittivity [16], which is modeled with a 2nd
order polynomial fit. In another study the microstrip ring
resonator is placed in contact with soil medium to define the soil
dielectric constant from the changes in the resonator parameters
by using a numerical approach based on the method of moments
[8]. The empirical formulation [17], which is validated at 10
GHz, is used [18] to calculate the dielectric constant of a polymer
photoresist. The variational calculation of the line capacitance is
used to compute the effective permittivity of the multilayer
microstrip ring resonator and thus determine the permittivity of
dielectric materials [19].
In this paper, a nondestructive method that can be applied to
bulk material samples of arbitrary thickness by using a
combination of one- and two-layer microstrip ring resonators is
proposed for the first time to measure the dielectric loss tangent.
The closed-form analytical formulations of multilayer microstrip
lines are manipulated and new frequency-corrected permittivity
and loss tangent measurements are performed. Measurements are

1070-9878/12/$25.00 2012 IEEE

IEEE Transactions on Dielectrics and Electrical Insulation

Vol. 19, No. 4; August 2012

1393

performed by using microstrip ring resonators designed for

different fundamental resonant frequencies and implemented on
various substrates. The uncertainties in the permittivities and loss
tangents are calculated based on the possible errors in the
experimental setup and fabrication process. Results are compared
with other measurement methods and good agreements are
achieved. Effects of conductor and radiation losses in calculation
of dielectric loss are discussed.

2 THEORY OF MEASUREMENT
TECHNIQUE
Figure 1 illustrates the two-port microstrip ring resonator
configuration. The ring resonator with mean radius of r and
width of w is printed on a microwave substrate with thickness of
hs. The slab material with thickness of hd is superimposed on the
ring resonator and the differences in Q factor and resonant
frequency before and after slab insertion are used to determine
the dielectric constant and loss tangent of the material.

Figure 1. Two-port microstrip ring resonator structure with a superimposed

slab.

2.1 RELATIVE PERMITTIVITY

The microstrip ring resonator circuit has multiple
resonances that occur when the perimeter of the ring is equal
to an integer number of guide wavelengths. The resonant
frequency, fr,2, of the two-layer resonator is related to the onelayer (i.e. without the slab) resonance, fr,1,

eff , 2

f r ,1

f r ,2

eff ,1

(1)

here eff,1 and eff,2 are the effective permittivity of the one- and
two-layer microstrip ring resonator, and fr,1 and fr,2 can be
extracted from the measured insertion loss of the ring
resonator with and without the slab, as shown in Figure 2.
Based on the analytical formulations of the multilayer
microstrip circuits [20], the relative permittivity of the slab,
r,2, can be found by using the effective permittivity of the
two-layer microstrip configuration and the relative
permittivity of the substrate, r,1,

r ,2

q2
q1 q 2 1

1 q1 2

(2)

eff , 2 q1 r ,1

here q1 and q2 are Wheeler filling factors for the two-layer

geometry, and are functions of geometrical parameters of the
microstrip circuit (i.e. w, hd, and hs).
2.2 DIELECTRIC LOSS TANGENT
The introduction of the slab on the microstrip ring not only
affects the resonance frequency of the circuit, but also changes
the 3-dB bandwidth of the resonance, BW, as shown in Figure
2. If it is assumed that the changes in conductor and radiation
losses due to the superposition of the slab are negligible (this
assumption will be further discussed in Section 4), it can be
concluded that
1
1
1
1
(3)

QD, 2 QD,1 QU , 2 QU ,1

Figure 2. Insertion loss measurement for the 6 GHz ring resonator (r = 5.84
mm, hs = 0.254 mm, w = 0.71 mm) on TLY5 substrate (r = 2.2, tan =
0.0009) before and after insertion of SU-8 slab (hd = 1.73 mm).

where QU,1 and QU,2 are the unloaded Q factors of one- and twolayer configurations, computed using the measured BW [21]
QU

fr
BW

1
1 10

L
20

(4)

QD,1 and QD,2 are the dielectric Q factors of one- and two-layer
configurations, and L is the measured insertion loss in dB of
the ring at resonance. The dielectric Q factor of an N-layer
microstrip configuration, QD,N, is related to the loss tangents
of different layers [22]
N

i, N

tan i

i 1

1
Q D, N

(5)

where tan i is the dielectric loss tangent and pi,N is the

Schneider filling factor for the ith layer. Dielectric loss
tangent of the superimposed slab is found by subtracting the
value obtained from equation (5) for one-layer configuration
from the value obtained from equation (5) for two-layer
configuration. With some manipulations
tan 2

1
p 2, 2

1
1 p1,1 p1, 2

tan 1

p 2, 2
D , 2 Q D ,1

(6)

1394

A. Rashidian et al.: Dielectric Characterization of Materials using a Modified Microstrip Ring Resonator Technique

Table 1. Physical parameters of ring resonators. The relative permittivity and loss tangent of the substrates are r = 6.15, tan = 0.0027 for RT/duroid
6006 and r = 2.2, tan = 0.0009 for TLY-5 with a few percents variation from 2 to 40 GHZ.
Substrate

Ring
Radius (r)

Gap (g)

Microstrip
Width (w)

Fundamental Resonant
Frequency

Material

Thickness (hs)

Copper Cladding

RT/duroid 6006

0.625 mm

17.8 m

11.2 mm

120 m

0.91 mm

2 GHz

TLY-5

0.381 mm

17.5 m

8.66 mm

100 m

1.16 mm

4 GHz

TLY-5

0.254 mm

17.5 m

5.84 mm

70 m

0.71 mm

6 GHz

where p1,1 can be found directly [22] and p1,2 and p2,2 are
derived by manipulating formulations reported elsewhere [20,
22]
p1, 2

r ,1
q1
eff , 2

p 2, 2

r ,2
q 2 1 q1 2
eff ,2 r , 2 1 q1 q 2 q 2 2

(7)

(8)

2.3 FREQUENCY DEPENDENCIES OF MICROSTRIP

RESONATOR
To calculate frequency dependent relative permittivity of the
slab, r,2(f), a general method of frequency improvement in the
form of [23]
r eff f 0
(9)
eff f r
1 P f
is used for the effective permittivity of the one-layer
microstrip circuit in equation (1). An accurate and r
independent formulation for the frequency correction term,
P(f), is that of Edwards and Owens [24].
To calculate the frequency dependent dielectric loss tangent
of the slab, the frequency dependent filling factors, pi,N(f),
should be derived. For the one-layer configuration substituting
equation (9) into the formulation of p1,1 results in
2 P f 1 G r ,1

(10)
p1,1 f
2 P f 2 eff ,1 f
hS

where G 1 10 w . For the two-layer structure, an

analytical frequency dependent effective permittivity is
required to define Schneider filling factors. For this purpose,
the effective permittivity formulation of the two-layer
structure is first used to calculate the quasi TEM expression.
This value is then considered as the effective permittivity of a
one-layer microstrip structure, and the relative permittivity of
the new one-layer configuration, r, is extracted as a function
of r,1 and r,2
1 q1 2
2
G 1
(11)
'r

r ,1q1 r , 2

r , 2 1 q1 q 2 q 2
1 G
2
Substituting equation (11) and the quasi TEM effective
permittivity of the two-layer structure into equation (9), the

frequency dependent effective permittivity of the two layer

configuration is derived. After some manipulation

1 q1 2
(12)
eff , 2 f A r ,1q1 r , 2
B
r , 2 1 q1 q 2 q 2

1 G P f . This
2 P f 1 G
where A
and B
1 P f 1 G
1 P f 1 G
equation is incorporated to derive frequency dependent filling
factors
p1,2 f Aq1

r ,1

(13)

eff ,2 f

q 2 1 q1 2
p 2, 2 f A
1 q q q 2
1
2
2
r ,2

r ,2

eff , 2 f

(14)

3 EXPERIMENTAL RESULTS
The two-layer microstrip ring resonator technique was used
to measure the dielectric properties of materials in this section.
The experimental results are compared with other
measurement methods in some cases to evaluate the accuracy
of the proposed method. Since the ring resonator parameters,
such as substrate thickness and permittivity, line width, ring
diameter, and coupling gaps may affect the accuracy of the
results [14], microstrip ring resonators of varying diameters
were designed and fabricated on different substrates. Table 1
shows the physical parameters of different rings. In all designs
the line width to ring radius ratio (w/r) of the ring resonator
structure is kept smaller than 0.2 to avoid excitation of
possible higher-order modes [21]. The first ten resonances of
the large rings (2 GHz and 4 GHz designs) and the first six
resonances of the small ring (6 GHz design) are within the
desired frequency range of measurement (up to 40 GHz), have
clear and sharp peaks (Q>100), and are suitable for
calculations to determine the electrical properties of the slab
materials. The slabs are superimposed on the rings and the
whole circuit is mechanically stabilized with small plastic
clips installed electrically far from the ring resonator. Figure 3
shows the experimental setup and the ring resonator circuit.
An Agilent 8722ES Vector Network Analyzer was used to
measure the insertion loss of the two-port ring resonators from
1 to 40 GHz. The universal substrate test fixture WK-3001-G
from Inter-continental Microwave was directly used to

IEEE Transactions on Dielectrics and Electrical Insulation

Vol. 19, No. 4; August 2012

1395

Relative Permittivity

3.6

3.4

3.2

(a)
3

10

15

20
25
Frequency (GHz)

30

35

40

30

35

40

(a)
0.05

(b)
Figure 3. (a) Microstrip ring resonators with different designs and fabricated
on different substrates; (b) experimental setup when transparent PMMA is
superimposed on the 2 GHz microstrip ring resonator.

Dielectric Loss Tangent

0.04

0.03

0.02

0.01

measure the circuits without additional connectors. Two

standard SOLT (Short-Open-Load-Through) calibration
procedures, one from 1 GHz to 20 GHz, and the other
from 20 GHz to 40 GHz were performed to de-embed the
loading effects of the test fixture and cables using the
Agilent 85056 Calibration Kit. The full 1601 point sweep
was used, giving a frequency resolution of 11.87 MHz for
the first and 12.49 MHz for the second band. The
measured insertion loss of the 6 GHz ring with and
without a slab is shown in Figure 2 as a sample. The sharp
resonances are clear and the effect of the slab in lowering
the amplitude and frequency of the resonances and
widening the 3-dB bandwidth is noticeable. Postprocessing (Spline interpolation) was performed on the
measured data with MATLAB to further enhance the
frequency resolution to less than 1 MHz. The extrema of
the interpolated curves before and after insertion of the
slabs are extracted as resonant frequencies. For each
resonant frequency, nearest left and right frequencies in
which the curve is 3 dB lower are then extracted and the
difference between them is calculated as the 3 dB
bandwidth for that particular resonant frequency. The Q
factors are then computed by using equation (4). Finally,
the relative permittivity and loss tangent of the slabs are
calculated at the resonant frequencies by using equation
(2) and equation (6), respectively.

10

15

20
25
Frequency (GHz)

(b)
Figure 4. (a) Dielectric constant and (b) loss tangent vs. frequency of SU-8.
The error bars are calculated based on uncertainties in resonant frequency,
Q-factor, microstrip width, substrate thickness, and substrate permittivity.

Figures 4 and 5 show the dielectric constant and dielectric loss

tangent of two typical photoresist materials (SU-8 and
Polymethylmethacrylate (PMMA)) from 2 to 40 GHz. The results
of the different ring designs are in good agreement. A worst case
variation of about 5% is observed for the measured permittivity
points at the common frequency points over the entire frequency
range. For both samples, the loss tangent significantly increases
with frequency and the dielectric constant is stable with frequency,
with a slight linear drop as the test frequency increases 20-fold. The
results for SU-8 and PMMA measurements are consistent with
reported results [25, 26], respectively. These two materials are
popular dielectric materials since they can be directly structured by
using lithographic techniques.
The results of other measurement methods provided by the
manufacturer [27] are compared with the proposed technique

1396

A. Rashidian et al.: Dielectric Characterization of Materials using a Modified Microstrip Ring Resonator Technique

Long Stripline Method (Linear Fit)

Cavity Resonator Method (Linear Fit)
Two-layer Ring Resonator Method (Linear Fit)
Two-layer Ring Resonator Method (Measured)

2.6

Relative Permittivity

2.8

Relative Permittivity

2.7

2.6

2.4

2.2

2.5

2.4

10

15

20
25
Frequency (GHz)

30

35

10

20
Frequency (GHz)

30

40

(a)

40

(a)
0.020

Dielectric Loss Tangent

0.010

Dielectric Loss tangent

0.015

0.010

Long Stripline Method (Linear Fit)

Cavity Resonator Method (Linear Fit)
Two-layer Ring Resonator Method (Linear Fit)
Two-layer Ring Resonator Method (Measured)

0.008

0.006

0.004

0.002

0.005
0
0

10

20

Frequency (GHz)

30

40

(b)
5

10

15

20
25
Frequency (GHz)

30

35

40

(b)
Figure 5. (a) Dielectric constant and (b) loss tangent vs. frequency of
PMMA. The error bars are calculated based on uncertainties in resonant
frequency, Q-factor, microstrip width, substrate thickness, and substrate
permittivity.

for RT/duroid 5880 and RO3006 in Figures 6 and 7. The

trends are linear least squares fitted lines, and the actual
measured data are shown by markers. In general, the two-layer
ring resonator plot gives a good approximation of the values
found by the standard long stripline and cavity resonator
methods. The relative permittivity achieved by the ring
resonator technique varies within about 5% while average
values of 2.17 and 6.11 (very close to 2.2 and 6.15 for the
other methods) are obtained over the entire frequency for these
two low loss materials (Figures 6a and 7a). The loss tangent
trends of all methods for RT/duroid 5880 (Figure 6b)
demonstrate a slight increase over the frequency band.
However, the trends of the proposed and the long stripline
methods for RO3006 (Figure 7b) show a slight decrease over
the frequency band while the cavity resonator trend is

Figure 6. Comparison of other measurement techniques with the proposed

two-layer ring resonator method. (a) Relative permittivity of RT/duroid
5880; (b) dielectric loss tangent of RT/duroid 5880. The error bars are
calculated based on uncertainties in resonant frequency, Q-factor, microstrip
width, substrate thickness, and substrate permittivity.

approximately constant. The loss tangents achieved by the ring

resonator agree well with other measurement techniques. The
fairly high scattering in the actual measured loss tangent
points of the proposed method is comparable with those of the
standard methods, shown in Figure 7b.
The uncertainties in the permittivities are calculated based on
the highest possible errors in resonant frequency (2 MHz, two
times the resolution), microstrip width (8 m, fabrication
resolution), substrate thickness and substrate permittivity (7%
and 1-2%, respectively, based on manufacturers data), and
arrived at by means of variational calculations with the
MATLAB code. The errors are calculated for all measured
points, and the overall cumulative errors for SU-8 and PMMA are
2-4% and for RT/duroid 5880 and RO3006 are 4-6%. To
maintain the readability, the error bars are shown in Figures
4 and 5 only for three different frequencies at the beginning
(8 GHz), middle (20 GHz), and end (36 GHz) of the band, each

IEEE Transactions on Dielectrics and Electrical Insulation

1397

Long Stripline Method (Measured)

Cavity Resonator Method (Measured)
Two-layer Ring Resonator Method (Measured)
Long Stripline Method (Linear Fit)
Cavity Resonator Method (Linear Fit)
Two-layer Ring Resonator Method (Linear Fit)

Relative Permittivity

Vol. 19, No. 4; August 2012

5
10

20

30
Frequency (GHz)

40

50

Figure 8. Comparison of sum of radiation and conductor losses before and

after slab insertion on the 6 GHz microstrip ring resonator. RT/Duroid 5880
and RO 3006 were superimposed as dielectric slabs.

(a)

Dielectric Loss Tangent

0.008

higher absolute loss of the SU8 composites, resulting in a

wider 3-dB bandwidth that is less sensitive to variations.
Again, to maintain the readability, the error bars are
shown in Figures 4 and 5 for three different frequencies,
each for a different ring. The error bars are shown for all
measured points in Figures 6 and 7.

Long Stripline Method (Measured)

Cavity Resonator Method (Measured)
Two-layer Ring Resonator Method (Measured)
Long Stripline Method (Linear Fit)
Cavity Resonator Method (Linear Fit)
Two-layer Ring Resonator Method (Linear Fit)

0.006

4
0.004

0.002

0
0

10

20
30
Frequency (GHz)

40

50

(b)

Figure 7. Comparison of other measurement techniques with the proposed

two-layer ring resonator method. (a) Relative permittivity of RO 3006; (b)
dielectric loss tangent of RO 3006. The error bars are calculated based on
uncertainties in resonant frequency, Q-factor, microstrip width, substrate
thickness, and substrate permittivity.

QC

for a different ring. The error bars are shown for all
measured points in Figures 6 and 7.
The errors for loss tangents are calculated based
on the uncertainties in the above mentioned parameters,
uncertainties in Q-factor measurements (5%), and also the
calculated uncertainty in slab permittivity (2-6%). The
errors are calculated for all measured points, and the
overall cumulative errors were on the order of 10% for all
samples. For instance, the minimum and maximum loss
tangent at 36 GHz are 0.034 and 0.042 for SU-8
(equivalent to 19% variation), 0.013 and 0.018 for PMMA
(equivalent to 28% variation), 0.0018 and 0.0033 for
RT/duroid 5880 (equivalent to 45% variation), and 0.0007
and 0.0013 for RO3006 (equivalent to 46% variation).
The errors are less for SU-8 than for PMMA and the other
two low loss materials, which can be attributed to the

DISCUSSIONS AND CONCLUSIONS

A particular advantage of the proposed dielectric properties

measurement method is the removal of conductor and
radiation loss calculations. In other words, it is supposed that
the sum of the conductor and radiation loss is similar before
and after slab insertion. The sum of conductor and radiation
losses for one-layer and two-layer microstrip resonators can be
calculated as follows at the resonant frequencies:

1
QR

1
QU

1
QD

(15)

To verify this assumption the unloaded (QU) and dielectric

(QD) quality factors of the one-layer (without slab) and twolayer (with slab) ring resonators with samples of known
permittivities and loss tangents up to 40 GHz (provided by
Rogers and Taconic corporations) are extracted from insertion
loss measurements and Schneider formulation for multilayer
microstrip configurations, equation (5).
Figure 8 compares the cumulative conductor and radiation
losses of the one-layer microstrip ring resonator with the twolayer microstrip ring resonator for the 6 GHz design (TLY 5
substrate material) when RT/duroid 5880 (r = 2.2 0.02, tan
e = 0.0006-0.0011) and RO 3006 (r = 6.15 0.15, tan e =
0.002-0.003) are superimposed as dielectric slabs. An
excellent agreement is achieved among the three curves,
confirming the accuracy of the assumption. The total
conductor and radiation loss increases with increasing
frequency and a maximum of 1.910-3 deviation occurs at 36
GHz between the ring resonator without slab and with the RO
3006 slab. It is worth mentioning that the samples with higher
permittivities will result in more deviation as shown in Fig. 8,

1398

A. Rashidian et al.: Dielectric Characterization of Materials using a Modified Microstrip Ring Resonator Technique

and consequently less precise results will be achieved. Other

experiments show that by increasing the substrate permittivity
(e.g. 2 GHz ring resonator design with RT/duroid 6006 as
substrate) the sum of the conductor and radiation losses
decreases to about one third of that of the 6 GHz ring and
becomes less important in dielectric loss calculation using this
method. Nevertheless, higher permittivity substrates will reduce
the coupling to the sample being measured and thus reduce the
sensitivity of the measurement of the permittivity of the sample.
It is to be noted that the results of the proposed technique
should be more accurate for higher dielectric loss materials
(i.e. SU-8 and PMMA) where the dielectric loss is dominant
and the difference in conductor and radiation loss before and
after slab insertion is negligible compared to the dielectric
loss. However, care should be taken for very lossy material
slabs (e.g. usually with loss tangent more than 0.1), which
decrease the Q-factor significantly, destroy clear and sharp
peaks, and make the measurements almost impossible.
A major source of the error in this method is the setup
resolution. In fact, more than half of the errors for both
permittivity and loss measurements come from the frequency
resolution of the setup (i.e. uncertainties in resonant frequency
and Q-factor measurements). Subsequently, the overall error
can be significantly reduced by using a higher resolution
measurement, e.g. by dividing the desired frequency range
into more than two sub-bands.
The main advantage of the proposed technique is in its
simplicity for use with a wide variety of prepared samples in a
nondestructive way. By fabricating only one ring resonator,
one can measure electrical properties of as many materials as
desired. Using a few ring resonators, the electrical properties
of the samples can be measured at numerous points over a
broad frequency range. This can also be used to increase the
reliability of the results when the measured points for different
rings coincide in frequency. The method is not specific to
sample thickness and is capable of measuring test slabs with
any thickness and dimension as long as the sample completely
covers the superimposed ring. There is no need for metalizing
the samples under test to measure their electrical properties.
This is especially important with specific materials that are
difficult to metalize, for instance because of low adhesion
(e.g. various polymer materials). Complex and inaccurate
equations to calculate the conductor and radiation losses are
avoided in the proposed method, since they do not introduce
any significant error. They appear almost identical before and
after slab insertion on the microstrip ring resonator circuits, as
shown in Figure 8, and therefore cancel each other out. The
conductor loss has a great impact on stripline and direct
microstrip methods, especially at higher microwave
frequencies. In the direct microstripline case, the radiation loss
must be calculated as well, which is not only significant at
high frequencies, but also difficult to calculate theoretically.
The proposed technique can also be used to measure materials
in non-substrate forms such as powders and liquids.

REFERENCES
[1]

[2]
[3]

[4]

[5]
[6]

[7]
[8]
[9]

[10]

[11]
[12]

[13]
[14]
[15]
[16]

[17]

[18]
[19]
[20]

ACKNOWLEDGMENT

[21]

The authors would like to thank Rogers and Taconic Companies

for providing measurement data and materials for this project.

[22]

K. Ishimoto, E. Kanegae, Y. Ohki, T. Tanaka, Y. Sekiguchi, Y. Murata, and C.

C. Reddy, Superiority of dielectric properties of LDPE/MgO nanocomposites
over microcomposites, IEEE Trans. Dielectr. Electr. Insulat., Vol. 16, pp.
1735-1742, 2009.
S. Singha and M. J. Thomas, Permittivity and tan delta characteristics of epoxy
nanocomposites in the frequency range of 1 MHz1 GHz, IEEE Trans.
Dielectr. Electr. Insulat., Vol. 15, pp. 2-11, 2008.
X. Huang, L. Xie, Z. Hu, and P. Jiang, Influence of BaTiO3 nanoparticles
on dielectric, thermophysical and mechanical properties of ethylene-vinyl
acetate elastomer/BaTiO3 microcomposites, IEEE Trans. Dielectr. Electr.
Insulat., Vol. 18, pp. 375-383, 2011.
J. M. Heinola and J. P. Strom, Evaluation of dielectric properties of printed
wiring board materials by using a microstrip-ring and strip-line ring
resonator methods, IEEE Electr. Insul. Mag., Vol. 23, No. 3, pp. 23-29,
2007.
F. Declercq, H. Rogier, and C. Hertleer, Permittivity and loss tangent
characterization for garment antennas based on a new matrix-pencil two-line
method, IEEE Trans. Antennas Propag., Vol. 56, pp. 2548- 2554, 2008.
M. P. Abegaonkar, R. N. Karekar, Y. K. Cho, and R. C. Aiyer, Study of nonlinear variation of dielectric constant with moisture for mug grains using
microstrip ring resonator, Microwave and Optical Technology Letters, Vol. 36,
pp. 221-224, 2003.
S. O. Nelson, Agricultural applications of dielectric measurements, IEEE
Trans. Dielectr. Electr. Insulat., Vol. 13. pp. 688-702, 2006.
K. Sarabandi and E. S. Li, Microstrip ring resonator for soil moisture
measurements, IEEE Trans. Geoscience and Remote Sensing, Vol. 35, pp.
1223-1231, 1997.
R. Ortega-Palacios, L. Leija, A. Vera, and M. F. J. Cepeda, Measurement of
breast-tumor phantom dielectric properties for microwave breast cancer
treatment evaluation, 7th Intl. Conf. Electr. Eng., Computing Sci., Automatic
Control, pp. 216-219, 2010.
I. S. Tsakanyan, V. K. Ivanov, and Y. I. Malyshenko, Complex dielectric
permittivity of animals organs and tissues measurements in microwave band
for the needs of modern veterinary medicine, 4th Intl. Sympos. Physics and
Engineering of Microwave, Millimeter and Submillimeter Wave (MSMW),
Ukraine, pp. 945, 2001.
P. Troughton, Measurement technique in microstrip, Electron. Lett., Vol. 5,
pp. 25-26, 1969.
D. C. Thompson, O. Tantot, H. Jallageas, G. E. Ponchak, M. M. Tentzeris, and
J. Papapolymerou, Characterization of liquid crystal polymer (LCP) material
and transmission lines on LCP substrates from 30 to 110 GHz, IEEE Trans.
Microw. Theory Tech., Vol. 52, pp. 1343-1352, 2004.
G. Zou, H. Gronqvist, J. P. Starski, and J. Liu, Characterization of liquid crystal
polymer for high frequency system-in-a-package applications, IEEE Trans.
Advanced Packaging, Vol. 25, pp. 503-508, 2002.
J. Heinola and K. Tolsa, Dielectric characterization of printed wiring board
materials using ring resonator techniques: a comparison of calculation models,
IEEE Trans. Dielectr. Electr. Insulat., Vol. 13, pp. 717-726, 2006.
X. Fang, D. Linton, C. Walker, and B. Collins, Dielectric constant
characterization using a numerical method for the microstrip ring resonator,
Microwave and Optical Technology Letters, Vol. 41, No. 1, pp. 14-17, 2004.
K. Saeed, A. C. Guyette, I. C. Hunter, and R. D. Pollard, Microstrip resonator
technique for measuring dielectric permittivity of liquid solvents and for
solution sensing, IEEE Intl. Microw. Theory Tech. Sympos. (MTT-S),
Honolulu, USA, pp. 1185-1188, 2007.
R. G. Sumesh Sofin and R. C. Aiyer, Measurement of dielectric constant using
a microwave microstrip ring resonator (MMRR) at 10 GHz irrespective of the
type of overlay, Microwave and Optical Technology Letters, Vol. 47, No. 1,
pp. 11-14, 2005.
J. Dewdney and J. Wang, Characterization the microwave properties of SU-8
based on microstrip ring resonator, Wireless and Microwave Technology
Conf., pp. 1-5, 2009.
P. A. Bernard and J. M. Gautray, Measurement of dielectric constant using a
microstrip ring resonator, IEEE Trans. Microw. Theory Tech., Vol. 39, pp.
592-595, 1991.
J. Svacina Analysis of multilayer microstrip lines by a conformal mapping
method, IEEE Trans. Microw. Theory Tech., Vol. 40, pp. 769-772, 1992.
K. Chang and L. H. Hsieh, Microwave Ring Circuits and Related Structures,
Hoboken, New Jersey, USA, John Wiley & Sons, 2004.
M. V. Schneider, Dielectric loss in integrated microwave circuits, The Bell
System Technical J., pp. 2325-2332, 1969.

IEEE Transactions on Dielectrics and Electrical Insulation

Vol. 19, No. 4; August 2012

[23] W. J. Getsinger, "Microstrip dispersion model", IEEE Trans. Microw. Theory

Tech., Vol.21, No.1, pp. 34- 39, 1973.
[24] T. C. Edwards, and R. P. Owens, 2-18 GHz dispersion measurements on 10100 ohm microstrip lines on sapphire, IEEE Trans. Microw. Theory Tech.,
Vol. 24, pp. 505-513, 1976.
[25] F. D. Mbairi and H. Hesselbom, High Frequency Design and Characterization
of SU-8 Based Conductor Backed Coplanar Waveguide Transmission Lines,
Intl. Sympos. Advanced Packaging Material: Processes, Properties, and
Interfaces, pp. 243-248, 2005.
[26] J. Baker-Jarvis, R. G. Geyer, J. H. Grosvenor, M. D. Janezic, C. A. Jones, B.
Riddle, and C. M. Weil, Dielectric characterization of low-loss materials - a
comparison of techniques, IEEE Trans. Dielectr. Electr. Insulat., Vol. 5, pp.
571-577, 1998.
[27] A. Horn, Dielectric constant and loss of selected grades of Rogers high
frequency circuit substrates from 1-50 GHz, Technical Report 5788, Rogers
Corporation, USA, 2003.

Atabak Rashidian (S08-M11) was born in 1977 in

Khorramabad, Iran. He received the M.Sc. degree
from Tarbiat Modares University, Tehran, in 2004,
and the Ph.D. degree from the University of
Saskatchewan, Saskatoon, Canada in 2011, both in
electrical engineering. In 2000, Dr. Rashidian joined
Telecommunication Company of Iran (TCI) as a
Microwave Engineer, working on fixed wireless
microwave communication systems. From 2001 to
2006, he was with Iran Telecom Research Center
(ITRC), where he was involved in several research projects to adapt the legacy
network to the next generation network. As part of his Ph.D. degree research, he
worked in the Institute of Microstructure Technology (IMT) at the German National
Laboratory Forschungszentrum Karlsruhe (FZK) as a Guest Scientist in 2009,
optimizing microfabrication process technologies for antenna and microwave
applications. After his Ph.D. degree was completed, he was a Postdoctoral Fellow at
the University of Saskatchewan for several months. Since 2011, he is a research staff
member in the Antenna Lab at the University of Manitoba. He was a recipient of
TRLabs Scholarship from 2007 to 2011. He was the author of a highlighted paper in
Journal of Micromechanics and Microengineering (JMM) in 2010. Dr. Rashidian is
the recipient of the IEEE Antennas and Propagation Edward E. Altshuler Prize Paper
Award in 2012. He has published 25 papers in international journals and conferences.
His research interests include dielectric structures for antennas and microwave
components, microfabrication processes and techniques for RF applications, advanced
microwave materials including microcomposites and nanocomposites, and microwave
measurements.

1399

Mohammadreza Tayfeh Aligodarz (S09) was

born in 1977 in Esfahan, Iran. He received the
M.Sc. degree from Tarbiat Modares University,
Tehran, Iran, in telecommunication engineering
in 2004. In 2004, he joined G.E. Pardis,
Siemens
partner
in
enterprise
telecommunications, first as an engineer and
then as the Chief Technical Officer. Since 2009,
he is with the Department of Electrical and
Computer
Engineering,
University
of
Saskatchewan, where he pursues his Ph.D. degree studies in
microfabrication of microwave and millimetre-wave devices with XRay lithography. Mr. Aligodarz is a recipient of University of
Saskatchewan Ph.D. Graduate Scholarship and TRLab Fellowship, and
was granted by Iran Telecom Research Center during his M.Sc. His
research interests include the areas of design of millimeter-wave
antennas and microwave devices suitable for X-ray lithography
fabrication, novel microwave materials for microfabrication processes,
fractal and non-Euclidian radiating structures, meta-materials, and
microwave measurement.
David M. Klymyshyn (S95-M98) was born in
Yorkton, Canada in 1968. He received the B.E.,
M.Sc., and Ph.D. degrees in electrical
engineering
from
the
University
of
Saskatchewan, Canada in 1990, 1995, and 1998,
respectively. From 1990 to 1994, Dr.
Klymyshyn was with SED Systems (Canada),
and in 1995 was with WaveCom Electronics
(now Vecima Networks - Canada), where he was
involved in the design of various microwave
communication systems. Currently he is a
Professor of Electrical and Computer Engineering at the University of
Saskatchewan, teaching undergraduate and graduate microwave
courses.
He is also a Research Scientist with Telecommunication
Research Labs (TRLabs - Canada), and also has international research
collaboration with the Institute for Microstructure Technology, KIT
(Germany), where he is a Visiting Scientist. His research interests
include linear and nonlinear microwave devices and circuits. His group
was first in Canada to demonstrate fabrication of high-aspect-ratio RF
MEMS coupling, resonant, and reactive structures fabricated using deep
X-ray lithography (XRL). He is co-Principal Investigator at Canadas
deep XRL facility, SyLMAND (Synchrotron Laboratory for
Micro/Nano Devices). Dr. Klymyshyn has 4 patents in microwave
devices and circuits, has supervised more than 40 graduate students and
researchers, and has published over 70 papers.