J Mater Sci: Mater Electron

Utilization of a triple hexagonal split ring resonator

(SRR) based metamaterial sensor for the improved
detection of fuel adulteration
Yadgar I. Abdulkarim1,* , Şekip Dalgaç2, Fatih Ozkan Alkurt3, Fahmi F. Muhammadsharif4,
Halgurd N. Awl5, Salah Raza Saeed1, Olcay Altıntaş3, Chen Li6, Mehmet Bakır7,
Muharrem Karaaslan3, Mohammad Ameen8, Raghvendra Kumar Chaudhary8, and
Heng Luo6

1
Medical Physics Department, College of Medicals & Applied Science, Charmo University, Chamchamal, Sulaimania 46023, Iraq
2
Department of Electric and Electronics Engineering, Sivas University of Science and Technology, 58050 Sivas, Turkey
3
Department of Electrical and Electronics, Iskenderun Technical University, 31100 Hatay, Turkey
4
Department of Physics, Faculty of Science and Health, Koya University, Koya 44023, Iraq
5
Department of Communication Engineering, Sulaimani Polytechnic University, Sulaimani 46001, Iraq
6
School of Physics and Electronics, Central South University, Changsha 410083, Hunan, China
7
Department of Computer Engineering, Bozok University, 66200 Yozgat, Turkey
8
Department of Electronics Engineering, Indian Institute of Technology (Indian School of Mines), Dhanbad 826004, India

Received: 18 June 2021 ABSTRACT

Accepted: 22 August 2021 In this work, a triple hexagonal split ring resonator incorporated metamaterial
sensor is proposed for the improved detection of fuel adulteration. The pro-
Ó The Author(s), under posed metamaterial structure is made of three resonators, acting to gather the
exclusive licence to Springer effective capacitive and inductive properties whereby any trivial variations in
Science+Business Media, LLC, the dielectric properties of the system can be readily correlated with the shift in
part of Springer Nature 2021 the resonance frequency. The advanced design system and computer simulation
technology software were used to measure the transmission coefficient, fol-
lowed by numerical and experimental investigations. The proposed sensor was
validated on petrol, kerosene and grease samples having different contents and
humidity levels. Results showed a noticeable shift in the resonance frequency of
the samples upon changing their concentration and humidity. The sensor was
able to provide an improved sensitivity and a high-quality factor of 291. The
proposed design can be used for the real time applications in the frequency
range from 1 to 20 GHz.

Address correspondence to E-mail: [Link]@[Link]

[Link]

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 J Mater Sci: Mater Electron

1 Introduction based sensors can be operated in the microwave

frequency range when they are used to detect liquid
The process of fuel adulteration is to add foreign chemicals. Abdulkarim et al. [25] Proposed a MTM-
substances into the standard fuels illegally, by which based sensor incorporating gammadion resonator,
the fuels lose their quality and void to comply with which can be effectively used for the detection of
the standard specifications. Some industries and gas chemical liquids in the frequency range from 8 to12
stations trying to mix kerosene with petrol or diesel GHz. The research groups at Iskenderun Technical
in the process of fuel adulteration. Consequently, University and the University of Adelaide carried out
vehicle engines may get damaged due to the diffi- some investigations on MTM-based sensors for the
culty in burning the adulterated fuels. The most applications of micro-fluid, strain, and rotation
dangerous aspect of the addition of kerosene is that sensing [26–29]. For instance, a MTM sensor based on
the lubricity of the fuel is highly reduced. Therefore, rectangular split ring resonator with a strapline was
the high-pressure fuel pump and nozzles become utilized for the detection of pressure, density and
vulnerable to wear and malfunction. Moreover, when volumetric moisture [30]. Interestingly, MTM sensors
hydrocarbon-based fuels are diluted with an elec- have found themselves in different applications
trolyte to increase the octane number, there would be [31–34]. Bakir et al. [35] proposed a technique to
a large breakdown in the electrical voltage of the detect fuel adulteration of kerosene, ethanol and olive
fuels [1]. Also, running a vehicle with adulterated oil with cotton oil by means of chiral MTMs that
gasoline may lead to frequent blockage of the injector incorporate a single split square and three ring res-
and hence the engine failure. Consequently, onators. Another group at the Henan Normal
researchers used different techniques to identify the University of China presented a MTM-based dual-
fuel adulteration. For instance, Felixe et al.[2] used a band high sensitivity microwave sensor for the fuel
laboratory emission test with the help of fiber grat- adulteration [36]. In the progress of MTM-based
ings to detect the fuel adulteration. Kulkarni and sensor for the detection of fuel adulteration [37–41],
Patrikar [3] performed a simple low-cost fiber optic hexagonal split ring resonator (HSRR) has attained a
coupler with an optoelectronic system for the detec- special attention because of its practical shape and
tion of petrol adulteration, while Bell et al. [4] ease of adaptation for various applications. For
developed a fluorescent paper strip for the detection instance, a monopole antenna loaded with HSRR was
of diesel adulteration using smart phones. Further- suggested for multiband operation [42, 43]. More-
more, Moreira et al. reported an adulteration detec- over, HSRR based MTMs were developed for S and X
tion system, which is based on gas chromatographic band microwave applications [44], terahertz appli-
technique [5]. cation [45], frequency filtration [46] and pH sensing
Despite the efficiency concerns, the implemented [47]. In this work, a novel design of MTM is pro-
methods reported in literature were found to be posed, which is based on three compact hexagonal
expensive or complex [6]. Therefore, the use of arti- split-ring resonators with the aim of improving the
ficial electromagnetic structures, known as metama- detection of fuel adulteration. The work was moti-
terials (MTMs), can be of great importance to develop vated by the high cost and complex techniques pre-
simple and efficient detectors [7]. MTMs are artifi- sented in literature to detect the fuel adulteration.
cially engineered materials capable of providing Hence, a simple and low-cost sensor is developed
special electromagnetic properties such as negative herein with an enhanced detection of the fuel adul-
permittivity (e \ 0), negative permeability(l \ 0) and teration. The novelty of this work is to employ a
a refractive index response that is not commonly seen compact design of three resonators within a single
in the natural materials [8, 9]. Because of exhibiting a sensor structure, which can be efficiently used to
strong electric and magnetic coupling, MTMs have detect the adulterated petrol with kerosene and to
received a considerable attention for designing highly sense the humidity of grease oil. Moreover, the sensor
sensitive detectors [10]. Moreover, MTMs have been can be utilized to measure the thickness, permittivity,
widely used in some important applications such as and permeability of the samples. The proposed low-
perfect absorber [11, 12], cloaking [13, 14], antennas cost sensor showed a high-quality factor, high sen-
[15–17], super-resolution [18, 19], energy harvesting sitivity and high resolution which is viable for the
[20, 21] and sensing [22–24]. It is known that MTM-

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J Mater Sci: Mater Electron

real-time applications in the frequency range from The sensor layer, with 10 mm of length, was neatly
1to 20 GHz. located behind the backside of the TSHSRR resonator.
Figure 2 shows the unit cell of the proposed
structure with a size of 22.86 9 10.16 9 1.6 (mm3). To
2 Materials and methods achieve the optimal result, parametric studies were
carried out by using genetic algorithm. The other
The schematic view of the proposed MTM-based important dimensions of the proposed sensor, shown
sensor is shown in Fig. 1. The design geometry in Fig. 2a and b, were selected based on our previ-
comprises a triple split-ring resonators, in which ously findings and the use of CST MS optimization
three split ring resonators are deposited onto the technique. To simplify the simulation processes and
front-side and back-side of the substrate to form a to perform the dimensional optimization of the pro-
unique MTM sensor. The full-wave finite integration posed MTM sensor, various boundary conditions
technique (FIT) based on high-frequency electro- were applied with the help of microwave analysis.
magnetic solver in the computer simulation technol- These boundary conditions include perfect electric
ogy microwave studio (CST MWS) was used to carry conductor/perfect magnetic conductor (PEC/PMC),
out the numerical analysis and optimization of the perfect electric conductor (PEC), free space and
sensor geometry. The proposed TSHSRR sensor is periodic pattern. In this design, perfect electrical
composed of three main layers, as shown in Fig. 1b. conductor (PEC) boundary condition was applied in
The resonators are made of copper metal with a the x-/y-directions, while the z-axis was assigned to
thickness of 0.035 mm and conductivity of 5.8 9 107 be opened (add space) alongside the wave propaga-
S/m. The dielectric substrate is flame retardant four tion direction. This is because the side walls of the
(FR-4), which was chosen because of its low-cost and waveguide present a metallic characteristic and that
availability. Its thickness, relative permittivity and the S- parameters are measured by using the WR-90
loss tangent are 1.6 mm, 4.3 and 0.02, respectively. waveguide and a compatible sample holder.

Fig. 1 The proposed metamaterial sensor based on triple hexagonal split-ring resonators: front view (a), back view (b), and perspective
view (c)

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 J Mater Sci: Mater Electron

Fig. 2 The front view (a), perspective view with optimized dimensions (b), and boundary conditions (c) for the designed MTM sensor
with two ports

Figure 2c shows the boundary condition (conducting the proposed design, the capacitance is induced by
wall) with the added two ports at each side of the the split gap and the inductance is generated due to
structure. The simulation process was carried out the metal strips. Hence, the coupling between the
using FIT based on CST MWS. gaps and electric fields leads to generate an electrical
Figure 3a shows the geometry of a hexagonal resonance, while the coupling of magnetic fields with
shaped split ring resonator, while Fig. 3b shows its the loss materials forming a magnetic resonance. The
equivalent RLC circuit. In the equivalent circuit, L1, capacitance of the split gap in the ring can be defined
R1, and Cg represent the inductance, resistance, and by:
gap capacitance of the inner ring, respectively. Simi- A
larly, L2, R2, and Cg is defined for the outer ring C ¼ e0: er ð2Þ
d
resonator. The proposed structure is a common form
of RLC circuit, by which information regarding the where e0 = 8.854 9 10–12 F/m is the permittivity of
resonance frequency of the TSHSRR can be obtained. free space, er denotes the relative permittivity, A is
In Fig. 3a, Ri defines the radius of the inner hexago- the split area and d is the split gap. The total induc-
nal, Ro represents its outer radius, b is the width of tance of the structure can be expressed as follows
the conductor ring, and d is the spacing between the [48]:
8 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi9
two conductor-hexagonal rings. < 2ðd þ g þ hÞ2 ð2w þ g þ hÞ2 þl2 =
The resonance frequency (f 0 ) of the triple hexago- L ¼ 0:01  l0 þ t
:ð2w þ g þ hÞ2 ð d þ g þ hÞ ;
nal split-ring resonators can be expressed by using
the general equation [48]: ð3Þ
1 7
f0 ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffi ð1Þ where l0 ¼ 4p  10 H/m.
2p Leq Ceq The total capacitance of the proposed metamaterial
where Leq and Ceq represent the equivalent induc- can be defined as:
tance and capacitance of the complete structure. In

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J Mater Sci: Mater Electron

(a) (b) 3 Results and discussion

Cg
R1 3.1 Metamaterial performance
of the proposed TSHSRR sensor
L1
Figure 4a shows the transmission (S21) and reflection
(S11) coefficients of the proposed TSHSRR that were
obtained from the CST simulation. It was noticed that
the transmission characteristics provided a wideband
response from 8.66 to 11.50 GHz (28.17%). To reveal
the MTM property of the sensor, the real and imag-
inary permeability (l) and permittivity (e) responses
(c) were extracted, as shown in Fig. 4b and c. The neg-
ative permeability and positive permittivity values
were observed from 8.71 to 10.82 GHz, confirming
the mu-negative MTM performance. It was noticed
that after 10.82 GHz the value of permeability grows
higher within a special frequency band, just above
the frequency range with negative permeability,
which makes l [ 1. In Fig. 4b and c, the effective
permeability and permittivity of the proposed MTM-
based sensor were plotted using the following equa-
tions [49]:
l ¼ nz ð5Þ
n
e¼ ð6Þ
Fig. 3 The schematic geometry of the hexagonal split ring z
resonator (a), its equivalent circuit diagram (b), and the
Then, the impedance (z) and refractive index (n) term
transmission spectra achieved from the CST and ADS (c)
were obtained from Eqs. (7 and 8) as follows:
"  # sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2w þ g þ h 2ðd þ g þ h ð1 þ S11 Þ2 S221
C ¼ e0 2
ln t ð4Þ z¼ ð7Þ
2pðd þ hÞ ð a  1Þ ð1  S11 Þ2 S221
In addition, the structure has a sensor layer placed at 1 h  ink0 d 00   ink0 d 0 i
n¼ ln e i ln e ð8Þ
the backside, which is used to hold the samples ðk0 dÞ
under measurement. This sensing layer can be
where S21 and S11 defines the transmission and
defined by the effective capacitance, Cs. In this work,
reflection coefficients of the MTM-based sensor. The
advanced design system (ADS) program has been
exponential term in Eq. (8) is represented by
used to modify the value of these parameters. It can   0
be seen from Fig. 3c that there is a good agreement eink0 d ¼ 1SS21 z1, where ln eink0 d represents the real
11 zþ1
  ink d 00
between the transmission coefficients of the proposed component, ln e 0 denotes the imaginary com-
structure that were obtained by tuning the parame- ponent of the complex number, k0 is the wavenumber
ters and the CST simulation. Before taking the sim- and d represents the maximum length of the unit cell.
ulation and experimental investigations, the dielectric Herein, the complete TSHSRR is considered as a
constant of the petrol, kerosene and grease was single unit cell.
measured by open ended coaxial probe with the
85070E dielectric measurement kit using a vector 3.2 Investigations on the dielectric constant
network analyzer in the frequency range from 8 to and dielectric loss factor
12 GHz.
Initially, the KEYSIGHT brand PNA-L N5234A Net-
work Analyzer and dielectric probe were used to

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 J Mater Sci: Mater Electron

Fig. 4 The transmission and

reflection (a), effective
permeability (b), and effective
permittivity (c) spectra of the
proposed MTM-based sensor

measure the values of dielectric constant and loss determined by dividing the imaginary part of the
tangent of the petrol with kerosene and grease oil dielectric constant (e00 ) by its real part (e0 ), that is tan
with different moisture ratios, as shown in Fig. 5. The d = e00 /e0 .
experimental setup used to measure the electrical Figure 6 shows the measurement setup used to
response of the chemical samples was also utilized to determine the electrical properties of the chemical
extract the real and imaginary parts of the relative liquids. The dielectric probe kit was connected to the
permittivity of the samples with different mixtures. PNA-L N5234A vector network analyzer (VNA) via a
The dielectric loss factor of each sample can be coaxial cable. The important step in this measure-
ment is the calibration of the VNA in the desired
frequency range (8–12 GHz). The measurements
VNA were carried out at room temperature (25 °C). Before
taking the measurements on the selected samples, the
apparatus calibration was carried out. To perform the
calibration, the electrical property of pure water was
given to the VNA while the dielectric was set idle,
and air was measured. The next step was to immerse
the dielectric probe in water followed by the device
Probe kit calibration accordingly. Then, the electrical charac-
teristic of water was measured in order to assure a
correct calibration of the VNA. Finally, the electrical
properties of the selected samples were measured.

Sample
3.3 Electromagnetic properties of the petrol
and kerosene samples

The electrical measurement of the petrol and kero-

Fig. 5 The experimental setup used to measure the electrical sene liquids were performed using a probe kit with
properties of the chemical samples

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J Mater Sci: Mater Electron

Fig. 6 The measured results (a) 2.7 (b)

for the petrol and kerosene 0.35
Petrol
samples: the dielectric constant Petrol

Dielectric Loss Factor

Kerosene
2.6 Kerosene

Dielectric Constant
(a), and dielectric loss factor 0.30
(b) in the frequency range 2.5
0.25
from 8 to 12 GHz
2.4
0.20
2.3
0.15
2.2
0.10
8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0
Frequency (GHz) Frequency (GHz)

the VNA in the operating frequency from 8 to 35% was found to be about 3.1, 3.8, 6.4 and 6.9,
12 GHz. The real dielectric constant and its imagi- respectively. The humidity of the grease was mea-
nary part were recorded for the samples, while tan sured by a calibrated handheld meter which was
d = e00 /e0 was employed to find the dielectric loss used as a moisture determination tool. Also, the
factor. Figure 6a shows the measured dielectric con- dielectric loss factor for the same percentage at 8 GHz
stant of the petrol and kerosene samples. One can see was estimated to be 0.14, 0.175, 0.225 and 0.23,
from the plots that the dielectric constant of the petrol respectively. It was concluded that the dielectric
and kerosene is about 2.68 and 2.39 at 8 GHz, constant of the liquid samples was highly dependent
respectively. Figure 6b shows the measured loss on the humidity. The increment in the humidity ratio
tangent of the samples at the frequency of 8 GHz. It has led to the increase in the real part of the dielectric
was found that the loss tangent is about 0.19 and 0.14 constant.
for the petrol and kerosene, respectively. However,
when the frequency was increased to 12 GHz, the 3.3.2 Electromagnetic parameters of the petrol
loss tangent for the petrol and kerosene was mea- with kerosene
sured to be 0.30 and 0.225, respectively.
In this section, the electromagnetic parameters (the
3.3.1 Electromagnetic properties of the grease liquid dielectric constant and dielectric loss factor) of vari-
ous percentages of kerosene in petrol were calculated
The second part of this study was devoted to use the and then compared with the modelling results
proposed sensor in the industrial applications such as achieved from Debye’s equation. To calculate the
detecting the quality of grease liquid which is used dielectric properties of the medium, Maxwell–Garnet
for machines lubrication purpose. The dielectric model was used. Generally, the Maxwell–Garnet law
properties of the grease with different humidity [50, 51] can be applied for a dilute composite media.
ratios have been measured by using the dielectric By using this method, one can find the volumetric
probe kit with the Agilent 85070E VNA. The dielec- weight of the mixed liquid. Equation 9 was used to
tric constant and dielectric loss factor were recorded determine the effective permittivity of the mixtures
in the frequency range from 8 to 12 GHz, as shown in for various kerosene concentrations from 10 to 90% in
Fig. 7. After recording the electrical properties of steps of 20%.
each chemical sample, the data files were imported eeff  e1 e2  e1
¼x ð9Þ
into the numerical program, CST MWS suite. Finally, eeff þ 2e1 e2 þ 2e1
the sensing capability of the proposed sensor was
where e1 is the complex permittivity of the petrol, e2
numerically observed by placing each sample in the
is the complex permittivity of the kerosene, eeff is the
sensor layer.
effective complex permittivity of the composite, and x
Figure 7 shows the measured dielectric constant
is the volume fraction occupied by the kerosene [52].
and dielectric loss factor for the grease samples at
The electromagnetic parameters of the samples were
different humidity ratios. The dielectric constant for
the grease with humidity ratio of 0%, 4%, 30% and

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 J Mater Sci: Mater Electron

Fig. 7 The measured (a) 8 (b)0.45

dielectric constant (a), and
0.40 Grease oil humidity 35%

Dielectric Loss Factor

dielectric loss factor (b) for the 7 Grease oil humidity 35%
Grease oil humidity 30%

Dielectric Constant
grease samples with different 0.35
6
humidity percentages in the 0.30
Grease oil humidity 30%
frequency range from 8 to 5
0.25
12 GHz
4 Grease oil humidity 4%
0.20
3 0.15 Grease oil humidity 0%
Grease oil humidity 0% Grease oil humidity 4%
2 0.10
8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0
Frequency (GHz) Frequency (GHz)

extracted using MATLAB programming. The calcu- 90% of kerosene was 2.65, 2.60, 2.54, 2.48 and 2.42,
lated results from the model are shown in Fig. 8. respectively at the resonance frequency of 8.5 GHz.
To verify the calculated results, the electromagnetic The dielectric loss factor for different ratios of petrol-
parameters were theoretically deduced from Debye’s kerosene showed a good agreement between the
model as follows: calculated values and theoretically extracted ones
e1 ðxÞ ¼ 0:3564x2  0:6342x þ 2:039 ð10Þ from the Debye’s equation.
Figure 9 shows the simulated result of the effective
es ðxÞ ¼ 0:1708x þ 0:6159 ð11Þ dielectric constant and resonance frequency shift
  against various concentrations of the kerosene in
sðxÞ ¼ x2  2x þ 2  1011 ð12Þ
petrol. It is worth to notice that by increasing the
where x is the volume fraction of the kerosene in content of the kerosene from 10 to 90%, the effective
petrol. The calculated parameters were inserted into dielectric constant was linearly decreased, whilst the
the Debye’s equation to yield a generalized formula, resonance frequency shift was linearly increased.
which can be used to determine the real and imagi- Concludingly, the proposed triple hexagonal split
nary permittivity of the samples at various volume ring resonator can be used for the quality control
fractions, as follows: assessment of chemical liquids.
e1 ðxÞ  es ðxÞ
e0 ðwÞ ¼ e1 ðxÞ þ ð13Þ 3.4 Numerical investigations
1 þ ½wsðxÞ2
½e1 ðxÞ  es ðxÞws rs Before considering the numerical investigations of
e00 ðwÞ ¼ þ ð14Þ
1 þ ½wsðxÞ2 weo the samples, the sensor layer was left empty. Fig-
ure 10a shows the numerical and measured trans-
The theoretically extracted dielectric constant and mission values with the presence of air in the sensor
dielectric loss factor along with their calculated val- layer. Comparably, two local minima in the simu-
ues are shown in Fig. 8a and b. The dielectric con- lated transmission were found at the resonance
stant for the samples with 10%, 30%, 50%, 70% and

Fig. 8 Comparison of the

calculated and fitted dielectric
constant values (a), and
dielectric loss factor (b) for
different petrol-kerosene
volumes. The solid lines
present the fitted results

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J Mater Sci: Mater Electron

minimum transmission value. The comparison of the

Effective dielectric constant

Frequency (GHz)
8.016 2.65
Effective dielectric constant maximum attainable Q-factor with that reported for
Frequency (GHz)

2.60 other metamaterials-based sensors is given in Table 1.

8.012
Noteworthy, the quality factor (Q- Factor) of the
2.55
8.008 y = 2.665+0.29*x
proposed MTM-based sensor was calculated to be
R2 = 0.9973 2.50 about 291, which is superior to those reported in the
8.004 literature (see Table 1). Moreover, an experimental
2.45
bandwidth of 350 MHz was obtained for the pro-
8.000 2.40 posed sensor which is better than those reported by
0 20 40 60 80 100 other researchers [23, 35]. Finally, the measured
Petrol concentration with Kerosene dielectric properties of the petrol and kerosene
showed almost similar electrical properties to those
Fig. 9 Simulated result of the effective dielectric constant and
reported in literature, validating the measured
resonance frequency shift versus kerosene content (% volume)
values.
frequencies of 8.5 GHz and 10.6 GHz, which are The simulated electric field distribution for the first
corresponding to - 18 dB and - 24 dB, respectively. and second resonance frequency has been obtained
The experimental results were also recorded in the by the numerical CST MWS at the desired operating
same condition of the simulated ones. The experi- frequencies, as shown in Fig. 10b and c. The electric
mental transmission curve presented three minima field mostly concentrated at the upper part of the
of - 28 dB, - 22 dB and - 22.5 dB at the corre- triple hexagonal resonator at the resonance frequency
sponding resonance frequency of 8.75 GHz, 9.98 GHz of 8.5 GHz. When the frequency was increased to
and 10.6 GHz. Overall, a good agreement between 10.6 GHz, which is the optimum value of the second
the numerical and experimental results was resonance frequency, a noticeable localization was
observed. The small difference between the simu- observed for the electric field at the left-hand and
lated and measured results might be due to the cal- middle hexagonal resonator. However, a small part
ibration errors in the VNA, fabrication tolerance, and of the electric field was concentrated around the
errors which may arise at the laboratory during the rectangular ring resonator.
testing process of using cables, connectors, and The parametric study and genetic algorithm
waveguide. Nevertheless, this deviation between the approach were used to determine the required
simulation and experimental transmission minima dimensions of the triple hexagonal SRR and overall
does not affect the sensing efficiency. This is because proposed structure. This approach was a built-in
the sensing mechanism is interpreted in terms of the function inside the CST software, which can be used
shift in the resonance frequency. In MTM-based to obtain the optimum result. The effect of the
f dimension of the triple hexagonal SRR on the reso-
sensors, the Q -factor is determined by Df0 , where Df is
nance frequency of the structure was determined by
the bandwidth at ? 3 dB with respect to the parametric study, as shown in Fig. 11. The effect of

Fig. 10 The simulated and

measured results of the
transmission value with empty
sensor layer (a) along with the
electric field distribution at 8.5
(b), and 10.6 GHz (c),
respectively

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 J Mater Sci: Mater Electron

Table 1 Comparison of the achieved dielectric constant, frequency range and maximum Q-factor with those reported in literature

References Dielectric constant Frequency (GHz) Quality factor Bandwidth (MHz)

[23] Not available 2–3 230 70

[36] 2.20 for petrol 8–12 400 270
[35] 2.41 for petrol & 2.26 for kerosene 2.24–2.34 114 40
[34] 2.43 for gasoline & 2.07 for diesel 10–11 205 100
[37] 80 for water & 9 for methanol 1.7–2.1 130 120
This work 2.68 for petrol & 2.39 for kerosene 8–12 291 350

Fig. 11 The shift in resonance

frequency due to the changes
in: thickness of the sensor
layer (a), permittivity (b), type
of the resonator layer (c), and
permeability (d)

thickness on the resonance frequency was investi- was increased. The main factor for this change in the
gated, as shown in Fig. 11a. The resonance frequency resonance frequency is that the inductive effect exists
seems to be slightly dependent on the thickness at in the sensor layer.
8.5 GHz as there is a small change in the resonance Further investigations were made on the impact of
frequency shift. The same topology was applied for the resonator material type on the transmission
the second resonance frequency at 10.6 GHz. The spectrum, as shown Fig. 11c. Noticeably, with the
second parametric study is the dielectric constant presence of copper metal, the resonator showed
characterization. In this part, the sensor layer was highest transmission activity. Meanwhile, the reso-
fixed at 10 mm, while the permittivity of the sensor nance frequency shift was seen to be trivial in the low
layer was increased from 2 to 3 in steps of 0.5. It was frequency range compared to that of the high fre-
observed from Fig. 11b that the optimum permittivity quency one. As such, the copper metal can be the best
value of the proposed structure is 1, where the choice in addition to its availability and low-cost.
transmission value presented the highest peak in the Figure 11d shows the effect of permeability change
low frequency range. The second resonance fre- from 1 to 3 in steps of 0.5. The optimum value was
quency was shifted backward when the permittivity found to be at 1 due to the presence of a slight change

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J Mater Sci: Mater Electron

in the resonance frequency at 10.5 GHz. The perme-

ability increase caused a reverse change in the reso-
nance frequency at 10.6 GHz, which counted for
500 MHz. Therefore, it was concluded that the effect
of parametric study at the higher frequency is more
pronounced than that at the low frequency.
To understand the sensing mechanism of the pro-
posed structure and the resonance behavior, the
surface current and magnetic field distribution were
investigated at the two resonance frequencies of
8.5 GHz and 10.6 GHz, as shown in Figs. 12, 13. The
simulated surface current distribution at the first and
second resonance frequency are shown in Fig. 12a
and b, respectively. One can see from Fig. 12a that the
maximum surface current is concentrated close to the
middle and right-hand side of the hexagonal res- Fig. 13 Simulated color map of the magnetic field distribution for
onator. Also, there exists a small current flow around the proposed design at 8.5 GHz (a), 10.6 GHz (b) resonance
the right-hand side of the rectangular ring resonator. frequency
Comparably, taking the surface current at 10.6 GHz,
as shown in Fig. 12b, one can observe that the current field in the direction of propagation, Ez = 0, and the
is largely distributed around the rectangular ring magnetic components exist in the same direction of
resonator in the parallel and antiparallel direction. propagations, Hz = 0.
The parallel currents are responsible to the genera- Figure 13 shows the numerical investigation per-
tion of electric field, while the antiparallel current formed on the magnetic field distribution for the
acting upon controlling the magnetic response. The proposed triple hexagonal split-ring resonators at
measurements and simulations in this study were two different resonance frequencies of 8.5 GHz and
carried out in the transverse electric TE10 mode, 10.6 GHz, respectively. The simulation result was
which means there are no components of the electric recorded when the sensor layer was empty. Fig-
ure 13a shows that a weak magnetic field is estab-
lished at the half middle and right-hand side of the
hexagonal shape resonator. When the frequency was
increased to reach the second resonance frequency at
10.6 GHz, the magnetic field was localized around
the rectangular ring resonator. Noticeably, no mag-
netic field was located at the hexagonal shaped res-
onator, as shown in Fig. 13b. Concludingly, the
physical mechanism of the two-resonance frequency
observed at 8.5 GHz and 10.6 GHz are different. The
resonance at 10.6 GHz was observed when the
hexagonal shaped resonator and magnetic field
component of the incident field were coupled. The
resonance phenomenon at 8.5 GHz is resulted from
the electrical coupling of the electron mobility at the
middle and right-hand side of the hexagonal shaped
resonator with the electric field component of the
incident field.

Fig. 12 Simulation of the surface current distribution of the

proposed structure at two different resonance frequencies:
8.5 GHz (a), and 10.6 GHz (b)

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 J Mater Sci: Mater Electron

3.5 Experimental validations important to distinguish the good quality of petrol

from kerosene. The numerical and experimental
To verify the simulation results, the sensor with triple studies were carried out when the sensor layer of the
hexagonal split-ring resonators was fabricated using proposed structure was filled with air, petrol and
LPKF ProtoMAT E33 prototyping computerized kerosene. The results of these investigations are
numerical control (CNC) machine. The photograph shown in Fig. 15a and b, respectively. The fuel
of the proposed structure is depicted in Fig. [Link] adulteration study was realized by simulating the
dimensions of the fabricated structure were chosen to transmission value of the petrol and kerosene sam-
be similar to that of the numerical design. The res- ples followed by the experimental tests using VNA
onator was made of copper placed on front and when the sensor layer was filled with the samples. As
backside of the substrate. The experimental investi- it can be seen from the results, the value of trans-
gations were taken by using Agilent PNA-L series mission in the simulated and test results are very
(VNA), coaxial cable, two waveguide, WR90 and close, showing that the proposed design has been
sample holder in the frequency range from 8 to accurately fabricated and effectively tested. Note-
12 GHz. In this experiment, the depth of the sample worthy, the proposed structure can easily distinguish
holder was 10 mm, while its face was tied by a cap- the liquid samples placed in the sample holder.
tion band, and it was placed at the front side of the Furthermore, fuel adulteration sensing was studied
waveguide. The backside of the sample holder was by using the theoretical model of Maxwell–Garnett
coated by a thin copper band, as shown in Fig. 14- [51]. The effective permittivity of kerosene ratio in the
[Link] experimental set up is shown in Fig. 14c. Before petrol from 10 to 90% in steps of 20% was calculated,
taking measurements, the VNA was calibrated via a and then the data file was imported into the CST
special calibration kit with open circuit, short circuit software as a new material. In this way, the value of
and 50 Ohm load connector for the two ports inside transmission was recorded for each concentration, as
the VNA in the frequency range from 8 to 12 GHz. shown in Fig. 16a. To support the numerical results,
The numerical and measured results for the fuel experimental investigation was made for various
adulteration and sensing application are shown kerosene ratios, as shown in Fig. 16b. There have
Figs. 15, 16 and 17. In the first course of investiga- been some fluctuations in the resulted curves due to
tions, fuel adulteration was carried out, which is very the manufacturing process. Even though the electri-
cal properties of these liquids are close to each other,
Sample holder
the designed structure can easily determine the per-
centage ratio of the mixed chemical liquids. The
sensing application of the proposed design can be
generalized to other liquid materials. The measure-
(a) Fabricated design (b) ment errors can be originated from the calibration
and measurement equipment. Comparably, the res-
VNA
onance frequency shift in Figs. 15 and 16 are close
enough to each other.
Validation of the sensing capability of the proposed
structure was also made on the grease sample with
humidity of 0%, 4%, 30% and 35%. The grease oil can
Coaxial cable be used as a sealant to minimize the leakage, to
prevent contamination, and to hold solid lubricants
in the suspensions. The numerical and experimental
Port 1 Port 2 studies were performed with different values of
(c) humidity, as shown in Fig. 17a and b. It was noticed
WR90
that the numerical and experimental results are in a
Fig. 14 Photos of the fabricated triple hexagonal split-ring good agreement. It can be concluded that the pro-
resonators type metamaterial (a), the waveguide and sample posed structure is highly sensitive against different
holder with manufactured sample (b), and experimental setup types of liquids, which can be effectively utilized as a
showing the proposed structure under test (c)

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J Mater Sci: Mater Electron

Fig. 15 Comparison of the

simulated (a), and measured
results (b) for the petrol and
kerosene samples

Fig. 16 Comparison of
simulated (a) and measured
(b) results for different ratios
of kerosene in petrol

Fig. 17 Verification of the

sensing process from the
simulated (a) and measured
results (b)

viable detector for sensing application of the chemical resonance frequency shift and the ratio of kerosene in
materials. petrol from 10 to 90%. Furthermore, the resonance
frequency shift was interestingly used to interpret the
humidity level of the grease liquids. It was concluded
4 Conclusions that with the presence of two resonance frequencies
at 8.5 GHz and 10.6 GHz, the proposed sensor
In this work, the design and fabrication of a low-cost showed a higher sensitivity and quality factor com-
MTM-based sensor incorporating triple hexagonal pared to those reported in literature. The simulation
split ring resonators was successfully employed for and experimental results were found to be in a good
the improved detection of fuel adulteration in the agreement. The designed structure can be used for
frequency range from 8 to 12 GHz. From the inves- various important sensing applications in addition to
tigations, a linear response was achieved between the the detection of fuel adulteration.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

 J Mater Sci: Mater Electron

Acknowledgements 6. P.K. Kanaujia, B.P. Vempatapu, Trends Anal. Chem. 92, 1–11
(2017)
The authors would like to thank the Central South 7. L. Deng, Y.I. Abdulkarim, M. Karaaslan, Ş Dalgaç, R.H.
University and Iskenderun Technical University for Mahmud, F.O. Alkurt, F.F. Muhammadsharif, H.N. Awl, S.
the technical supports. Huang, H. Luo, Electronics 9(825), 1–17 (2020)
8. D.R. Smith, W.J. Padilla, D. Vier, S.C. Nemat-Nasser, S.
Schultz, Phys. Rev. Lett. 84(18), 4184 (2000)
Author contributions 9. R.A. Shelby, Science 292(5514), 77–79 (2001)
10. L. Deng, Y.I. Abdulkarim, O. Altıntaş, E. Ünal, M. Karaaslan,
YIA and MK conceived the idea; YIA and HNA Phys. E: Low-dimens. Syst. Nanostruct. 114, 113593 (2019)
performed the simulations; ŞD and FOA performed 11. F. Dincer, K. Muharrem, C. Sabah, J. Electromagn. Waves
the experiment; YIA wrote the manuscript, OA, CL, Appl. 29(18), 2492–2500 (2015)
FFM, HL, MA, MB, RKC and SRS revised the 12. M. Mao, Y. Liang, R. Liang, L. Zhao, N. Xu, J. Guo, F. Wang,
manuscript; MK and HL supervise this work. All H. Meng, H. Liu, Z. Wei, Nanomaterials 9(8), 1101 (2019)
authors have read and agreed to the published ver- 13. D. Ramaccia, D.L. Sounas, A. Alu, F. Bilotti, A. Toscano,
sion of the manuscript. IEEE Antennas Wirel. Propag. Lett. 17(11), 1968–1972
(2018)
14. S. Guo, Y. Zhang, Y. Ge, S. Zhang, H. Zeng, H. Zhang, Adv.
Funding Mater. 31(39), 1902352 (2019)
15. K. Sun, S. Han, J.H. Choi, J.K. Lee, IEEE Antennas Wirel.
This research was funding by the National Key
Propag. Lett. 17(7), 1162–1165 (2018)
Research and Development Program of China (Grant
16. R. Junyi, S. Gong, W. Jiang, IEEE Antennas Wirel. Propag.
No. 2017YFA0204600), the National Natural Science
Lett. 17(1), 102–105 (2017)
Foundation of China (Grant No. 51802352), Central
17. J.D. Binion, E. Lier, T.H. Hand, Z.H. Jiang, D.H. Werner,
South University (Grant No. 2018zzts355) and
Nat. Commun. 10, 108 (2019)
Teaching reform for postgraduate students of Central
18. L. Aiping, X. Zhou, J. Acoust. Soc. Am. 132, 2800 (2016)
South University (Grant No. 2019JG085).
19. G. Yuan, K.S. Rogers, E.T.F. Rogers, N.I. Zheludev, Phys.
Rev. Appl. 11(6), 064016 (2019)
20. M. Bağmancı, M. Karaaslan, E. Unal, M. Özaktürk, O.
Data availability
Akgol, F. Karadağ, A. Bhadauria, M. Bakır, Int. J. RF
The data that support the findings of this study are Microw. Comput. Aided Eng. 29(7), e21597 (2019)
available from the corresponding author upon rea- 21. T. Tan, Z. Yan, H. Zou, K. Ma, F. Liu, L. Zhao, Z. Peng, W.
sonable request. Zhang, Appl. Energy 254, 113717 (2019)
22. M.T. Islam, A. Hoque, A.F. Almutairi, N. Amin, Sensors
Declarations 19(1), 169 (2019)
23. S. Ogawa, Y. Takagawa, M. Kimata, Materials 12(19), 3157
Conflict of interest The authors declare that there is (2019)
no conflict of interest. 24. Z. Yi, C. Liang, X. Chen, Z. Zhou, Y. Tang, X. Ye, Y. Yi, J.
Wang, P. Wu, Micromachines 10(7), 443 (2019)
25. Y.I. Abdulkarim, L. Deng, M. Karaaslan, E. Unal, Chem.
References Phys. Lett. 732, 136655 (2019)
1. Global Marketing, Diesel Fuels Technical Review (Chevron 26. O. Altintas, M. Aksoy, O. Akgol, E. Unal, M. Karaaslan, C.
Corporation, San Ramon, CA, 2007) Sabah, J. Electrochem. Soc. 164(12), B567–B573 (2017)
2. P.A. Udaykiran, V.J. Felix, K. Ganesan, Indian J. Sci. Tech- 27. M. Bakir, J. Electrochem. Soc. 164(9), 8488 (2017)
nol. 8(S2), 90–95 (2015) 28. M. Bakir, M. Karaaslan, O. Akgol, O. Altintas, E. Unal, C.
3. S. Patrikar, S. Kulkarni, AIP Conf. Proc. 2136(1), 050004 Sabah, Optik 168, 741–746 (2018)
(2019) 29. W. Withayachumnankul, K. Jaruwongrungsee, A. Tuantra-
4. R. Gotor, J. Bell, K. Rurack, J. Vis. Exp. 141, e58019 (2018) nont, C. Fumeaux, D. Abbott, Sens. Actuators A Phys. 189,
5. L.S. Moreira, L.A. d’Avila, D.A. Azevedo, Chromatographia 233 (2013)
58(7–8), 501–505 (2003)

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

J Mater Sci: Mater Electron

30. O. Akgol, E. Unal, M. Bağmancı, M. Karaaslan, U.K. Sevim, 42. M.K.A Rahim, M. Aminu-Baba, F. Zubir, M. Fairus, M.
M. Öztürk, A. Bhadauria, J. Electron. Mater. 48(4), Yusoff, N. Asmawati Samsuri, O. Ayop, K.I. Jahun, in IEEE
2469–2481 (2019) Asia-Pacific Conference on Antennas and Propagation
31. O. Altintas, M. Aksoy, E. Unal, M. Karaaslan, J. Elec- (APCAP), Auckland, New Zealand (2018).
trochem. Soc. 166(6), B482–B488 (2019) 43. R.S. Daniel, R. Pandeeswari, S. Raghavan, Int. J. RF and
32. M. Bakir, M. Karaaslan, F. Karadag, S. Dalgac, E. Unal, O. Microw. Comput. Aided Eng. 28(2), e21401 (2018)
Akgol, Appl. Comput. Electromagn. Soc. J. 34(5), 799–806 44. M. Samsuzzaman, M.S. Islam, G.K. Beng, N. Misran, N.
(2019) Amin, IEEE Access 28, 68239–68253 (2020)
33. M.A. Tümkaya, E. Ünal, C. Sabah, Int. J. Mod. Phys. B 45. W. Jianbo, C. Xinyi, C. Guibo, S. Guancheng, C. Jian, L. Jun,
33(24), 1950276 (2019) Infrared Laser Eng. 41, 622 (2012)
34. C. Qiu, J. Wu, R. Zhu, L. Shen, B. Zheng, Opt. Commun. 46. T. Saktioto, R.F. Syahputra, S. Punthawanunt, J. Ali, P.
451, 226–230 (2019) Yupapin, Microw. Opt. Technol. Lett. 59(6), 1337–1340
35. M. Bakır, Ş Dalgaç, M. Karaaslan, F. Karadağ, O. Akgol, E. (2017)
Unal, T. Depçi, C. Sabah, J. Electrochem. Soc. 166(12), 47. M.T. Islam, F.B. Ashraf, T. Alam, N. Misran, K.B. Mat,
B1044–B1052 (2019) Sensor 18(9), 2959 (2018)
36. W. Liu, M. Wang, Y. Niu, L. Xu, J. Electrochem. Soc. 48. M.R. Islam, M.T. Islam, M. Moniruzzaman, M. Samsuzza-
166(15), B1452–B1460 (2019) man, H. Arshad, Sci. Rep. 11, 1–22 (2021)
37. V. Rawat, V. Nadkarni, S.N. Kale, S. Hingane, S. Wani, C. 49. X. Chen et al., Phys. Rev. 70, 016608 (2004)
Rajguru C, in Proceedings of the 2nd International Sympo- 50. J.C. Maxwell-Garnett, Philos. Trans. R. Soc. Lond. A 203,
sium on Physics and Technology of Sensors (ISPTS), vol. 257 385–420 (1904)
(1EEE, 2015). 51. J.M. Garnett, Philos. Trans. R. Soc. Lond. Ser. A
38. V. Rawat, V. Nadkarni, S.N. Kale, Def. Sci. J. 66, 421 (2016) 205(387–401), 237–288 (1906)
39. M.A. Tümkaya, F. Dinçer, M. Karaaslan, C. Sabah, J. Elec- 52. J.H. Santeen, D. Polder, Phys. E: Low-dimens. Syst. Nanos-
tron. Mater. 46, 4955 (2017) truct. 12(5), 257–271 (1946)
40. W. Xian-Ju, L. Xin-Fang, Chin. Phys. Lett 26(5), 056601
(2009) Publisher’s Note Springer Nature remains neutral with
41. A. Singh, S.K. Sharma, Int. J. Res. Eng. Technol. 03(10), 144 regard to jurisdictional claims in published maps and
(2014) institutional affiliations.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

onlineservice@[Link]