Sensitivity enhancement of a SPR biosensor

for urine glucose detection by employing BaTiO3

and BlueP/WSe2
Abror Jawad Momen Sahriar Shoshi Md. Anwar Hossain
Electrical & Electronic Engg. Electrical & Electronic Engg. Electrical & Electronic Engg.
Rajshahi University of Engineering & Rajshahi University of Engineering & Bangladesh University of Business &
Technology Technology Technology
Rajshahi, Bangladesh Rajshahi, Bangladesh Dhaka-1216,Bangladesh
[email protected] [email protected] [email protected]

Abstract—In this study, we present a high-performance since it is label-free, fast, accurate, and real-time [2]. These
surface plasmon resonance (SPR)-based refractive index SPR-based sensors fall into two general configurations: the
biosensor to detect the glucose concentration in the urine Kretschmann configuration and the Otto configuration.
samples. The performance of the suggested sensor was However, the Kretschmann configuration is more
examined and assessed utilizing the commonly used transfer appropriate in numerous cases due to its easily designed
matrix method (TMM). The sensor design utilizes the structure. Here, monochromatic p-polarized light excites the
Kretschmann configuration, comprising of a BK7 glass prism metal interface and surface plasmon wave (SPW) by directly
coated with a coating of Ag (Silver) metallic layer, a dielectric
applying the metal layer to the prism base [3]. During the
layer of BaTiO3 (barium titanate), and a layer of 2D
heterostructure BlueP/WSe2 (blue phosphorous-di-selenide) investigation, we determined the resonance angle (  SPR)
material. The layers' thickness affects the sensor's using the angular modulation approach, and found that the
performance. In order to achieve the lowest reflectance and output of the transmitted light's reflectance decays after
maximum sensitivity, the layer's thickness has also been tuned. passing through the multilayer structure. At the resonance
The optimized thickness of Ag, BaTiO3, and monolayer of condition, the surface plasmons are most excited, minimizing
BlueP/WSe2 has been taken as 52 nm, 11 nm, and 0.34 nm reflectance while maintaining the incoming light's velocity
respectively. The characteristics parameters are also computed and energy. This satisfies the condition ki = kp [4]. Here, ki is
and obtained highest at the glucose concentration of 10 g/dL. the incident wave vector and is defined as.
The maximum sensitivity of 425 deg/RIU with values of
detection accuracy (DA), full width at half maximum (FWHM), 2 n p
ki  sin  i (1)
and quality factor as 0.177305 deg− 1, 5.64 deg, and 75.35 RIU−1 
respectively has been obtained.
Where the variables θi and np denote the incident angle
Keywords—Surface Plasmon Resonance (SPR) sensor, and the refractive index of the incident medium, respectively.
glucose concentration, barium titanate, blue-phosphorus We can represent kp, the propagation constant of the plasmon
tungsten di-selenide, Sensitivity, Transfer Matrix Method (TMM) mode, as follows.
2  m d
kp 
I. INTRODUCTION  m d
. The primary energy source in the human body is (2)
glucose. Every cell in the body uses glucose as fuel, and the where ɛm and ɛd indicate the permittivity constants of the
blood contains it. Insulin, a hormone, stabilizes blood
metaland the dielectric layers, respectively [5]. The Fresnel
glucose levels in our bodies by transporting glucose from the
equations produce the SPR curves, which measure the shift
circulation into the cells for energy use. Diabetes, also
in the resonance curve due to the change in the refractive
known as hyperglycemia, results from disruptions in this
index of the sensing medium. These variations in the SPR
mechanism caused by inadequate hormone synthesis or use.
curve are closely related to the biosensor's sensitivity,
These days, diabetes is a major worldwide illnesss that is
whereas the SPR curve's broadness indicates greater
becoming out of control.If not promptly addressed, elevated
detection accuracy. This change enables the biosensor to
blood sugar levels may harm a patient's kidneys, eyes, brain
respond more quickly. So, the main goal of the study is to
system, and other organs. Inappropriate or excessive therapy
come up with a very sensitive SPR biosensor idea by making
may sometimes cause the blood sugar to drop below normal,
the SPR curve shift more while changing the sample's
which can result in hazardous hypoglycemia episodes. This
refractive index as little as possible.
may result in seizures, unconsciousness, and eventually death
[1]. As a result, it is usually recommended to identify The most crucial and first step in creating a biosensor is
glucose levels early and monitor them continuously. This choosing a metal layer. SPR biosensors use a variety of
work reports on the increased sensitivity of a surface plasmonic materials, each with advantages and
plasmon resonance (SPR)-based biosensor for the detection disadvantages of its own, including gold (Au), silver (Ag),
of glucose levels in urine samples. Medical diagnostics today aluminum (Al), nickel (Ni), copper (Cu), and platinum (Pt)
emplosy SPR, an optical approach that is extremely effective [6]. Silver (Ag) and gold (Au) are considered the best options
among all plasmonic materials due to their appealing phosphorus/WSe2 heterostructure layer improves the
qualities. Ag has outstanding optical qualities, including little performance of the suggested SPR sensor by providing
optical damping, a stronger resonance peak, and no interband enough excitation energy for charge transfer and steady
transmission at the visible light frequency, making it a good analyte collection.
contender. Ag substances have the potential to enhance the
In this work, we have shown the detection of glucose
sensitivity of SPR sensors [7]. This suggested sensor uses Ag
metal as the plasmonic metallic layer on the BK7 prism concentration levels present in urine samples with refractive
substrate. SPR sensors based on Ag metal exhibit a steep index variation using a BK7-Ag-BaTiO3-BlueP/WSe2 SPR
biosensor. We plotted the SPR curves using the transfer
matrix approach for the N-layer model. Numerical
simulations have optimized the metallic layer thicknesses
that offer the best combination of minimal reflection at
resonance and high sensitivity. The suggested sensor, with a
sensitivity of 425 deg/RIU and a quality factor (QF) of
75.35461 RIU-1, can detect glucose concentrations in urine
samples
II. DESIGN AND THEORETICAL MODEL
The multi-layered structure (BK7-Ag-BaTiO3-
BlueP/WSe2-Sensing Medium) of the suggested SPR
biosensor is schematically shown in Figure 1. Use of the
BK7 prism as a first layer has satisfied the need to match the
momentum of light. According to Eq. (3), the BK7 prism's
refractive index on a certain wavelength is 1.5151 [1].

Fig. 1. Schematic diagram of the proposed SPR model 1.039612122 1.010469452 0.2317923442
nBK7  1  2  2
 0.00600069867  103.560653  0.0200179144
2
drop in the SPR curve.
(3)
However, oxidation is still an issue when employing
silver metal. We might place a protective layer over the Ag where λ indicates the wavelength of the incident light,
layer for optimal performance. This protective layer serves as which is633 nm. The Ag layer's refractive index as a
a line of defense against problems with corrosion and function of radiation wavelength is expressed as follows [6]:
oxidation that might impair the SPR sensor's functionality. 1
   c  2
Recent studies have shown that when used as a protective nmetal      1  2
  (  i ) 
layer, a sheet of barium titanate (BaTiO3), which has  p c 
exceptional dielectric qualities including low loss and a high (4)
index, may provide a significant resonance angle shift with a
The values of the plasmon wavelength(λp) and the
little change in the analyte index [8]. Due to their exceptional
collision wavelength (λc) for Ag is λp = 1.4541 × 10-7 m and
optical and electrical properties, two-dimensional
λc = 1.7614 × 10-5, respectively [6], and λ indicates the
nanomaterials (2D) such as graphene (G), PtSe2, Ti3C2Tx-
incident light wavelength. With an 11 nm thickness, BaTiO3
Mxene, blue phosphorus (BlueP), black phosphorus (BlackP),
is the third layer under consideration. BaTiO3's refractive
hexagonal boron nitride (hBN), and transition metal
index has been determined to be 2.4042 [8]. The fourth layer
dichalcogenides (TMDc) have gained a lot of attention in the
(BlueP/WSe2) has a calculated refractive index of 2.68 +
last twenty years. Optoelectronic devices, energy storage
0.220i [10]. Each BlueP/WSe2 layer has a thickness of 0.34
devices, optical sensors, solar cells, supercapacitors, fuel
* L nm, where L is the layer number. We have employed a
cells, and other devices utilize these materials. Currently, the
2-layered (L = 2) BlueP/WSe2 layer in our suggested
TMDc family includes approximately forty compounds, also
biosensor. The Ag metal layer has a thickness of 52 nm,
known as MX2. Chalcogen materials such as sulfur (S),
while the BaTiO3 layer has an 11 nm thickness, which
tellurium (Te), and selenium (Se) use X, while transition
provides the best sensitivity. The fifth layer has been used as
metals such as tungsten (W), molybdenum (Mo), tantalum
a sensing medium and consists of biosample.Here, we have
(Ta), and niobium (Nb) use M. By interacting with metal
taken urine samples with glucose concentrations of 0–15
sheets, these two-dimensional nanomaterials improve
mg/dL, 0.625 gm/dL, 1.25 gm/dL, 2.5 gm/dL, 5 gm/dL, and
molecular interactions for biomolecule sensing. However,
10 gm/dL with the corresponding refractive indices of 1.335,
research is moving away from traditional 2D nanomaterials.
1.336, 1.337, 1.338, 1.341, and 1.347 [10].
The heterostructure is the newest class of 2D nanomaterials.
It essentially combines transition metal dichalcogenides The structure's reflectivity may be determined by using
(MX2) with traditional 2D nanomaterials. Examples include the transfer matrix approach for the N-layer system. The
graphene-MoS2, BlueP/MoS2, BlueP/MoSe2, BlueP/WS2, tangential fields of the starting limit, Z = Z1 = 0, and the final
and BlueP/WSe2, among others. Heterostructure 2D limit, Z=ZN-1, are correlated. Equation 5 showed the
nanomaterials exhibit a higher energy bandgap, work relationship between the tangential fields at the starting and
function, charge transfer, mobility, and optical absorption in final limits [7]:
comparison to typical 2D nanomaterials [9]. We use the
U1  U N 1 
BlueP/WSe2 heterostructure has a higher optical absorption   M 
coefficient, energy bandgap, work function, charge transfer, V1  VN 1 
and mobility than traditional 2D nanomaterials, we are using (5)
it in our suggested SPR sensor. As a result, the blue
Whither U1 illustrates the tangential elements of electric (11)
fields and tangential elements of the magnetic field can be
assimilated with V1 at the limiting surface. In addition, the A high value of sensitivity is required for good
performance of an SPR sensor Sensitivity (S) can be
analogous bounds owing to the limit of the Nth layer are UN-1
and VN-1. The characteristics matrix of this compound expressed as the ratio of change in the resonance angle (∆θsp )
with respect to the change in the refractive index of a
architecture is the Mij. In this way, for the sake of plane-
polarized light, the conditions are defined by equation 6 to 7 biosample (∆n) as noted in equation 11 [8]:
[7]:  sp
S (deg/RIU)
 
M
M ij   kN21M k   11
M12 

ij  M 21 M 22  (12)
n

(6) The FWHM is de_ned as the difference between two

 cos  k  (i sin  k ) / qk  predictor variable points for which the response variable is
With , M k    equal to half of its maximum value. QF is the ratio of angular
  iqk sin  k cos  k  sensitivity to FWHM [8] and is expressed as,
(7)
S (RIU-1)
1/ 2 QF 
1/ 2    n 2 sin 2   FWHM
  k 1 1
Where, qk   k  cos k    (13)

 k  k
Detection Accuracy is the inverse to the FWHM of the
(8) reflectance intensity curve and can be denoted as [8],
2 2d k  1/ 2
2 2 1
k  nk cos k ( z k  z k 1)    k  n1 sin 1  (9) DA  (deg-1)
    FWHM
Here, dk illustrates the density of the ledges that are (14)
thought towards the z-axis. Likewise, this can be resolved
that a dielectric constant and refractive index are found in the
kth layer, which could be shown through εk and դk III. RESULTS AND DISCUSSION
independently. The reflection coefficient(rp) for plane- This work utilizes a transfer matrix method (TMM)
polarized incident light is expressed by equation 10 [9]: implemented using the "Matlab" program for the detection of
(� +� � )� −(� +� � ) glucose concentration in a urine sample with changes in the
�� = (�11+�12��)�1+(�21+�22��) refractive indices. The light has been directed towards the
11 12 � 1 21 22 �
(10) prism BK-7, and the angular interrogation technique has
been used. The intensity of reflection has been measured at
At last, the reflection intensity for plane-polarized light is every angle of incidence
given by equation 11 [9]:
Initially, we investigate how the thickness of the Ag layer
2 affects the minimum reflectivity and sensitivity. This
R p  rp analysis is conducted for various thicknesses of the barium

Fig. 2. Minimum reflectance and sensitivity versus silver layer thickness of the structure prism-Ag-BaTiO3-BlueP/WSe2-sensing medium for different
thicknesses of barium titanate (a) d3 = 0 nm, (b) d3 = 7 nm, (c) d3 = 9 nm and (d) d3 = 11 nm. (λ=633 nm, number of BlueP/WSe2 layers=2 and glucose
concentration, C=10 gl−1).
Urine
Sensitivity FWHM QF
glucose RI Rmin
(0/RIU) (deg) (RIU-1)
levels
(0–.15)
1.335 ref 0.0267 4.16 ref
mg∕dL
.625
1.336 310 0.0312 4.25 72.94118
g∕dL
1.25
1.337 320 0.0365 4.34 73.73272
g∕dL
2.5
1.338 330 0.0428 4.45 74.1573
g∕dL
5 g∕dL 1.341 356.67 0.0708 4.79 74.46138
10
1.347 425 0.2367 5.64 75.35461
g∕dL

Fig. 3. Minimum reflectance and sensitivity versus the number of BlueP/WSe2 layer thickness of the structure prism-Ag-BaTiO3-BlueP/WSe2-sensing
medium for different thicknesses of barium titanate (a) d3 = 0 nm, (b) d3 = 7 nm, (c) d3 = 9 nm and (d) d3 = 11 nm. (λ = 633 nm, thickness of silver(Ag)
metal d2 = 52 nm and C=10 gl−1).
titanate (BatiO3) layer while keeping the wavelength constant theoretical connection, derived from the angle scan, between
at λ = 633 nm and using two layers of BlueP/WSe2. Angle the number of BlueP/WSe2 layers and minimum reflectivity.
scans (each minimum reflectance corresponds to a different In this instance, the thicknesses of the barium titanate layers
incidence angle) were used to establish theoretical are similarly assumed to be 0 nm (Figure 3(a)), 7 nm (Figure
correlations between minimum reflectivity and different 3(b)), 9 nm (Figure 3(c)), and 11 nm (Figure 3(d)). The
thicknesses (0 nm, 7 nm, 9 nm, and 11 nm) of barium findings show that for the number of BlueP/WSe2 layers L =
titanate layers. Each minimum reflectance corresponds to a 3, the value of minimum reflectance is 8.08 × 10-06 when the
certain incidence angle, as shown in figures 2(a)–(d). It is thickness of the barium titanate layer is maintained at zero
clear that the thickness of the Ag layer has a significant (figure 3(a)). In the other panels of the figure (figures 3(b)–
impact on the lowest reflectance. The results show that, (d)), the minimum reflectivity changes as the number of
provided the barium titanate layer thickness remains constant,
the minimum reflectivity progressively drops toward zero as
the Ag layer thickness grows and subsequently rises as the
Ag layer thickness increases further. The values of minimum
reflectance are 2×10-03, 7.35×10-04, 6.5×10-03, and 9.76×10-04
for barium titanate layer thicknesses of 0 nm (figure 2(a)), 7
nm (figure 2(b)), 9 nm (figure 2(c)) and 11 nm (figure 2(d)),
respectively. When comparing the four panels of figure 2, it
can be observed that the reflectance profiles show a
minimum reflectivity at Ag metal layer thickness of 52 nm,
with barium titanate layer thicknesses of 7 nm (figure 2(b)).
The suggested SPR-based sensor's angular sensing capability
is compared by examining the sensitivity in relation to the
thickness of the Ag layer at glucose concentration, C=10 gl-1 Fig. 4. Reflectivity curve to analyze the variation in incidence
with change in Refractive indices 1.341 to 1.347 for different angle for detecting urine glucose in a bio-sample using a specific
barium titanate layer thicknesses and plotted in figures 2(a)– structure (Prism-Ag-BaTiO 3-BlueP/WSe2-Bio-sample).
(d). It was shown that the angular sensitivity changes and
reaches its maximum as the thickness of the Ag layer BlueP/WSe2 layers rises. However, the minimum reflectivity
is seen when the number of BlueP/WSe2 layers is 2, 1, and 1
increases. The barium titanate layer thickness impacts the
for barium titanate layer thicknesses of 5 nm, 9 nm, and 11
optimum Ag layer thickness, where the sensitivity is at its
nm, respectively. When comparing the four panels of figure
highest. At a number of BlueP/WSe2 layers L = 2, the
3, it can be seen that the minimal reflectivity drops shifts
thickness of silver d2 = 52 nm, and barium titanate d3 = 11
towards a lower number of BlueP/WSe2 layers as the
nm, the greatest sensitivity of 425 deg RIU−1 can be achieved.
thickness of the barium titanate layer increases. The barium
TABLE I. PERFORMANCE PARAMETERS COMPUTATION OF PROPOSED titanate and BlueP/WSe2 layers function as absorption
SPR SENSOR mediums. The energy received by these layers is transferred
to the metal layer (Ag) in order to induce a strong surface
plasmon resonance (SPR) excitation. Decreasing the
The influence of the number of BlueP/WSe2 layers on the thickness of barium titanate layers necessitates an increased
sensitivity and minimum reflectivity is investigated for a number of BlueP/WSe2 layers to achieve full energy transfer.
given thickness of Ag (52 nm), radiation wavelength (633 To compare the angular sensing performance of the proposed
nm), and barium titanate layer. Figures 3(a)–(d) show the SPR sensor, the change of the sensitivity with the number of
BlueP/WSe2 layers at a glucose concentration of C=10 gl-1 degrees, following the same pattern. Figure 5(a) displays the
for different thicknesses of the barium titanate layer is sensitivity(blue curve) and quality factor(red curve) as they
calculated and plotted in figures 3(a)–(d). The research has relate to glucose concentration. In figure 5(b), we can
found that the angular sensitivity increases as the number of observe the relationship between the minimum reflectance
BlueP/WSe2 layers increases. The maximum angular (blue curve) and the DA (red curve) as the glucose
sensitivity values are 130 deg RIU−1 (figure 3(a)), 208.33 concentration varies. From figure 5(b), it is evident that the
deg RIU−1 (figure 3(b)), and 290 deg RIU−1 (figure 3(c)). FWHM slightly increases as the glucose concentration
These values are achieved when the number of BlueP/WSe2 increases. The angle increases from 4.16 deg at a
layers is L=4, and the barium titanate layer thicknesses are 0 concentration of glucose 0-15 mg/l to 5.64 deg at a
nm, 7 nm, and 9 nm, respectively. In figures 3(d), the concentration of 10 g/l.
sensitivity decreases after reaching its maximum of 425 deg
RIU-1 at number of BlueP/WSe2 layers L = 2. TABLE II. COMPARING THE SENSITIVITY OF THE CURRENT WORK
WITH THE MOST RECENT WORKS.

Sr. Sensitivity
Structure Year References
No. (o/RIU)
BK7-Ag-BaTiO3-BlueP- Proposed
1 WSe2-Bio- sample 425 2024
work
BK7/Al/BaTiO3/Ni/WS2/s
2 330.5 2023 [8]
ensing medium
BK7/Au/ZnO/Ag/BP/sensi
3 ng medium 289 2022 [6]
CaF2 Prism/ZnO-Au-
4 BlueP-MoS2/Sensing 235 2021 [9]
medium
Prism/Ag/PtSe2/WS2/Bio-
5 194 2020 [7]
sample

Consequently, the provided numerical analysis offers a

effective solution for accurately detecting glucose levels in
urine samples. In table II, a comparison is made between the
proposed structure and similar work that has already been
reported. It can be seen that the proposed biosensor has a
notable amount of sensitivity S= 425 deg/RIU, QF =105.38
RIU-1 and DA=0.177305 deg-1 compared to the other
biosensors that have been reported.
IV. CONCLUSIONS
A biosensor based on SPR has been suggested for
Fig. 5. (a) Sensitivity and quality Factor versus glucose concentration measuring the glucose concentration in human urine samples.
(b) Minimum reflectance andFWHMversus glucose concentration of
the proposed structure.
The Kretschmann configuration is used for detection,
utilizing the attenuated total reflection principle. The
We have performed an analysis of the effectiveness of performance parameters, including sensitivity, FWHM, QF,
the suggested sensor in identifying glucose levels in human and DA, have been calculated to demonstrate the
urine at the optimized value of Ag-layer and BaTiO3-layer effectiveness of the proposed structure. The measured values
thicknesses. The refractive index of human urine ranges from for sensitivity, FWHM, QF, and DA of the suggested SPR
1.33 to 1.35 when the glucose concentration varies from 0 glucose detection sensor are 425 deg/RIU, 5.64 deg,
g/dl to 10 g/dl. Figure 4 displays the SPR response curves of 75.35461 RIU−1, and 0.1777305 deg−1. These improved
the proposed sensor for the six concentrations of glucose in performance parameters create opportunities for this
urine. It can be seen that the greater value of glucose proposed design in sensing-based applications.
concentration results in a more significant change in the
resonance angle. In table I, we present the refractive index
for each concentration, minimum reflectance (Rmin), full
REFERENCES
width at half maximum (FWHM), sensitivity (S), and quality
factor (QF) to demonstrate the performance of the proposed [1] A. Yadav, A. Kumar, P. Sharan, and M. Mishra, "Highly sensitive
bimetallic-metal nitride SPR biosensor for urine glucose detection,"
sensor. With a RI change of 0.001, a maximum sensitivity of IEEE Transactions on NanoBioscience, vol. 22, no. 4, pp. 897-903,
330 deg∕RIU has been determined, while a minimum 2023.
reflectance of 0.0267 has been observed. Furthermore, when [2] H. V. Hsieh, Z. A. Pfeiffer, T. J. Amiss, D. B. Sherman, and J. B.
considering a 0.006 RI change, the sensitivity is found to be Pitner, "Direct detection of glucose by surface plasmon resonance
a remarkable 425 deg/RIU (for 10 g/L), with the lowest with bacterial glucose/galactose-binding protein," Biosensors and
reflectance recorded at 0.2367. When the glucose Bioelectronics, vol. 19, no. 7, pp. 653-660, 2004.
concentration varies from 0 − 15 mg∕dL to 2.5 g∕dL, there is a [3] E. Kretschmann and H. Raether, "Radiative decay of non radiative
surface plasmons excited by light," Zeitschrift für Naturforschung A,
small change in the angle of incidence from 81.45 degrees to vol. 23, no. 12, pp. 2135-2136, 1968.
82.41 degrees for a 0.001 RI variation. For a RI change of
0.006, the angle of incidence shifted from 83.48 to 86.03
[4] Y. A. Akimov and W. Koh, "Resonant and nonresonant plasmonic [10] L. Han, H. Ding, N. N. A. Landry, M. Hua, and T. Huang, "Highly
nanoparticle enhancement for thin-film silicon solar cells," sensitive SPR sensor based on Ag-ITO-BlueP/TMDCs-graphene
Nanotechnology, vol. 21, no. 23, p. 235201, 2010. heterostructure," Plasmonics, vol. 15, pp. 1489-1498, 2020.
[5] S. Hayashi and T. Okamoto, "Plasmonics: visit the past to know the
future," Journal of Physics D: Applied Physics, vol. 45, no. 43, p.
433001, 2012.
[6] A. Yadav, A. Kumar, and P. Sharan, "Sensitivity enhancement of a
plasmonic biosensor for urine glucose detection by employing black
phosphorous," JOSA B, vol. 39, no. 1, pp. 200-206, 2022.
[7] M. M. Rahman, M. M. Rana, M. S. Rahman, M. Anower, M. A.
Mollah, and A. K. Paul, "Sensitivity enhancement of SPR biosensors
employing heterostructure of PtSe2 and 2D materials," Optical
Materials, vol. 107, p. 110123, 2020.
[8] M. Uwais, A. Bijalwan, and V. Rastogi, "SPR-Based refractive index
sensor using BaTiO3 and WS2 with Al-Ni bimetal for glucose
sensing," Physica Scripta, vol. 98, no. 10, p. 105515, 2023.
[9] S. Singh, A. K. Sharma, P. Lohia, and D. Dwivedi, "Theoretical
analysis of sensitivity enhancement of surface plasmon resonance
biosensor with zinc oxide and blue phosphorus/MoS2
heterostructure," Optik, vol. 244, p. 167618, 2021.