Research Article Vol. 42, No.

1 / January 2025 / Journal of the Optical Society of America B 97

Optimization and characterization of a PCF-based

SPR sensor for enhanced sensitivity and reliability
in diverse chemical and biological applications
Abhijit Kumar Mukhopadhyay, Santu Sarkar,* Subhadeep Mukherjee,
AND N. R. Das
Institute of Radio Physics and Electronics, University of Calcutta, 92 A.P.C. Road, Kolkata, 700009, West Bengal, India
*ssrpe@[Link]

Received 24 July 2024; revised 5 November 2024; accepted 13 November 2024; posted 15 November 2024; published 12 December 2024

This paper presents the meticulous design and characterization of a photonic crystal fiber (PCF)-based surface
plasmon resonance (SPR) sensor, analyzed using finite element method (FEM) simulations. The sensor’s superior
performance is demonstrated by its exceptional sensitivity to minute changes in the refractive index (RI) of various
analytes. The optimized structure reveals a peak resonance at 0.78898 µm with a maximum confinement loss of
546.34 dB/cm. Detailed investigations show resonance wavelength red shifts of up to 0.93% with a 5% increase
in air hole diameter and blue shifts of up to 0.88% with a 5% decrease. Additionally, variations in the plasmonic
gold layer thickness result in resonance shifts of up to 0.76% longer or 0.87% shorter wavelengths. The sensor
achieves remarkable wavelength sensitivity (WS) of up to 13,000 nm/RIU and amplitude sensitivity (AS) of up to
1538.90 RIU−1 , underscoring its high precision in detecting analyte concentration changes. The design’s robust-
ness against fabrication errors, evidenced by minimal variations in resonance characteristics, highlights its practical
reliability. Furthermore, the use of a polynomial regression model with an R2 value near unity accurately approxi-
mates the relationship between resonance wavelength and analyte RI, ensuring precise sensing capabilities without
overfitting. Comparative analysis with existing designs confirms the sensor’s superior performance, rendering it
highly suitable for a wide range of applications, including biosensing of glucose, water contaminated by cholera
germs, mucosa of the human intestine, important components of human blood, and detection of chemicals like
acetone, ethanol, benzene, propanol, glycerol, and expired transformer oil. © 2024 Optica Publishing Group. All rights,
including for text and data mining (TDM), Artificial Intelligence (AI) training, and similar technologies, are reserved.

[Link]

1. INTRODUCTION fields, including biosensing, chemistry, and biomedical research

[3–5]. Some new application areas and designs of PCF-SPR sen-
Photonic crystal fiber (PCF) and surface plasmon resonance
sors that have been explored in recent times include transformer
(SPR) stand as pivotal components in the realm of optical sens-
oil monitoring [6], heavy metal detection [7], merger of spec-
ing, each bearing profound significance. PCF, characterized
troscopy, lasing, and RI sensing [8–11], fusion of internal metal
by its microstructured design featuring an array of air holes,
deposition (IMD) and external metal deposition (EMD)-based
allows for precise control of light guidance and manipulation, approaches [8], and FEM-based laser applications [6,8].
optimizing properties like dispersion and nonlinearity [1]. Small changes in the PCF parameters and the refractive
SPR, on the other hand, is a sensitive optical phenomenon that index of the surrounding analyte modulate the penetration of
occurs at the interface between a metal and a dielectric when light from the core into the cladding and consequently lead to
light is incident at a specific angle, making it a valuable tool for resonance angle shifts.
monitoring changes in refractive index (RI) in real time [2]. The The sensing mechanism in PCF-based SPR sensors is
synergy of these technologies involves incorporating a metallic twofold—internal and external. In the case of internal sens-
layer on the PCF’s surface or within its air holes, enabling SPR ing, the analyte is used to fill up the air holes selectively, whereas
excitation and producing highly sensitive sensors [1,2]. By in the case of external sensing, the analyte is placed on the sur-
managing the interaction between the core mode and surface face of the PCF. In the internal sensing mechanism, since the
plasmon polariton (SPP) mode at the metal–dielectric interface injected analyte directly affects the initial refractive index dis-
within the PCF, these sensors can detect minute alterations in tribution of the fiber, it has the potential to exhibit better levels
the surrounding environment, finding applications in diverse of sensitivity. Numerous sensor designs employing internal

0740-3224/25/010097-08 Journal © 2025 Optica Publishing Group

 98 Vol. 42, No. 1 / January 2025 / Journal of the Optical Society of America B Research Article

sensing mechanisms grace the annals of literature, each offering an AS of 70 RIU−1 , and a resolution of 2 × 10−5 RIU. This
distinct contributions to the domain of refractive index sensing design has the drawback of a very low AS, in addition to having
[12,13]. For instance, in Ref. [12], a sensor design making use a narrow sensing range. Titanium nitride may introduce higher
of a gold coating, which demonstrated a maximum wavelength optical losses compared to noble metals, impacting the intensity
and amplitude sensitivity of 11,000 nm/RIU and 1420 RIU−1 , of the SPR signal. The foregoing discussion shows that many
respectively, was reported. The paper [13] proposed the design of the existing works on PCF-SPR sensors have certain gaps
of a sensor that made use of gold-coated hollow fiber core as that need addressing. Hence, in this paper, a novel design of
the microfluidic analyte channel. However, the challenges a PCF-based SPR sensor is proposed, which does not require
associated with these internal sensing mechanisms prompted the cutting of any kind of grooves or removal of portions of
the quest for alternatives. Thus, many researchers have utilized the PCF, thereby making the sensor’s fabrication process easy.
the external sensing mechanism to design PCF-SPR sensors. In On top of this, the plasmonic material used is gold, which is
Ref. [14], a novel design of a PCF-SPR sensor making use of a chemically very stable. The sensor proposed here can detect
u-grooved selective coating for extremely sensitive remote and analytes over a wide RI range of 1.30–1.39, thus making it
biosensing applications was developed. This sensor achieved a suitable for sensing multiple analytes. In terms of performance,
maximum amplitude sensitivity (AS) of 1189 RIU−1 , highest the sensor boasts the highest AS of 1538.90 RIU−1 , maxi-
wavelength sensitivity (WS) of 12,500 nm/RIU, and maximum mum WS of 13,000 nm/RIU, best wavelength resolution of
resolution of 8 × 10−6 RIU while detecting analytes in the 7.69 × 10−6 RIU, and highest FOM of 400 RIU−1 .
RI range of 1.29–1.40. However, the creation of the u-groove The rest of the paper is organized as follows. In Section 2, the
with a selective coating of a gold plasmonic layer thereon is a design of the sensor and its associated theoretical background
challenge from a fabrication point of view. The complications are discussed. Section 3 investigates the sensor performance by
involved in grinding to achieve the deep polishing depth of optimizing different design parameters using simulations based
the u-groove make this type of sensor design unfit for mass on the finite element method (FEM). Section 4 concludes the
production. The design of an SPR-based PCF sensor with two paper.
open channels for detecting analytes in the refractive index
range of 1.33–1.40 was introduced in Ref. [15]. This sensor
2. SENSOR DESIGN AND THEORETICAL
exhibited the highest wavelength and amplitude sensitivities of
BACKGROUND
7000 nm/RIU and 593.61 RIU−1 , respectively, and a resolu-
tion of 1.43 × 10−5 RIU. This type of design suffers from the The PCF-SPR sensor is conceptualized as a dual hexagonal ring
limitation that it is difficult to fabricate the open channels and structure. In order to improve the coupling between the core
selectively coat those with metal, and its figure of merit (FOM) mode and the SPP mode along the Y axis, air holes in the rings
of 94.97 RIU−1 is quite low. Side polishing as a technique to along the Y axis are removed. Additionally, a column of air holes
obtain better performance was utilized in the design reported in is introduced centrally inside the core along the Y axis to create a
Ref. [16], which achieved a maximum WS of 12,500 nm/RIU dual-core structure with strong coupling. Figure 1 represents the
and a maximum wavelength resolution of 8 × 10−6 RIU in the X–Y cross section of the design.
RI detection range of 1.35–1.39. Besides having a small sensing The dimensions of the different structure components are
range, this sensor’s fabrication requires precise removal of a fixed as follows. The fiber radius is set at 3.5 µm. A plasmonic
specific portion of the PCF, which raises the complexity level of gold layer of thickness 40 nm is deposited on the fiber. The
the manufacturing process. Furthermore, the extra polishing width of the analyte region is taken as 0.98 µm. The width of
layer on the flat surface adds to the cost of the fabrication proc- the perfectly matched layer (PML) is set at 3.5 µm. The pitch
ess. A dual-core structure based on a hexagonal arrangement size is kept at 2 µm. The diameter of the air holes is related to the
of circular air holes was designed in Ref. [17], wherein it was pitch through different scale factors, k1 to k6 . The air hole sizes
reported that the sensor showed peak wavelength and amplitude of the structure are given in Table 1. The values of plasmonic
sensitivities of 10,700 nm/RIU and 1770 RIU−1 , respectively, layer thickness, pitch size, and air hole diameters mentioned
and a wavelength resolution of 9.34 × 10−6 RIU within the above are the optimized values for the sensor structure proposed.
narrow sensing range of 1.39–1.40. As is evident, this sensor The amplitude interrogation method is chosen to optimize the
design is unsuitable for detecting analytes over a wide RI range, structure since it is less complex and cost-effective compared to
thus restricting its usage only to a very niche segment of analytes. the wavelength interrogation method. The hexagonal arrange-
A sensor design employing air holes in two rings for sensing RI ment of air holes is chosen as it limits light scattering through
in the range of 1.33–1.40 was reported in Ref. [18]. The sensor’s
performance was characterized by its highest WS, highest AS,
and maximum resolution of 12,000 nm/RIU, 1656 RIU−1 ,
and 8.33 × 10−6 RIU, respectively. The sensor achieved its best
performance when silver was used as the plasmonic material.
However, silver can easily get oxidized in humid conditions,
which affects its sensing capabilities. The narrative of sensor
designs continued to unfold with the exploration of titanium
nitride-coated PCF in the quest for refined sensing, as in Ref.
[19]. The resulting sensor demonstrated a sensing RI range of
1.385–1.40, boasting a maximum WS of 10,000 nm/RIU, Fig. 1. XY cross-section of the proposed sensor.
 Research Article Vol. 42, No. 1 / January 2025 / Journal of the Optical Society of America B 99

Table 1. Diameter of Different Air Holes

Description Denoted by Relation with Pitch Value (µm) Scale Factors Involved
Central hole d1 k1 × pitch 0.3200 k1 = 0.16
Holes immediately above and below d2 k3 × k2 × k1 × pitch 0.2304 k1 = 0.16, k2 = 1.80, k3 = 0.40
the central hole
Holes at the outer extremes of the d3 k3 × k4 × k1 × pitch 0.1856 k1 = 0.16, k3 = 0.40, k4 = 1.45
central column
Larger holes in the inner hexagon d4 k2 × k1 × pitch 0.5760 k1 = 0.16, k2 = 1.80
Smaller holes in the inner hexagon d5 k5 × k1 × pitch 0.1920 k1 = 0.16, k5 = 0.60
Larger holes in the outer hexagon d6 k4 × k1 × pitch 0.4640 k1 = 0.16, k4 = 1.45
Smaller holes in the outer hexagon d7 k6 × k1 × pitch 0.2336 k1 = 0.16, k6 = 0.73

the cladding compared to a rectangular arrangement, thereby The AS is measured at a single operating wavelength. So,
increasing the coupling between the core-guided mode and to carry out this measurement, no variation of wavelength is
the SPP mode [20,21]. Gold is chosen as the metal to create necessary, thus making its measurement easier compared to WS.
a coating of the plasmonic layer on the outer surface of the The formula for calculating the AS is as follows:
cladding because of its chemical inertness, large number of sur-
1 ∂α(λ, n a )
face free electrons, ease of deposition over silica, greater shift of S A (λ) = − RIU−1 ,


× (5)
wavelength, and less loss in the visible and near-infrared region α(λ, n a ) ∂n a
[3,5,22–24]. where α(λ, n a ) is the confinement loss at a particular λ and n a ,
The background material of the fiber is chosen to be fused and ∂α(λ, n a ) is the difference in confinement loss values at a
silica. Its dispersion relation is given by the Sellmeier equation particular λ for a small change ∂n a in the refractive index of the
[24]: analyte.
An important metric to characterize the performance of the
s
B1 λ2 B2 λ2 B3 λ2
n(λ) = 1 + 2 + 2 + 2 , (1) sensor is the FOM, which is defined as follows [24]:
λ − C1 λ − C2 λ − C3
Sλ
RIU−1 ,


where λ is the free-space wavelength (µm). The Sellmeier coeffi- FOM = (6)
FWHM
cients B1 , B2 , B3 , C 1 , C 2 , and C 3 are as provided in Ref. [24].
The dielectric constant of gold can be calculated using the where Sλ is the wavelength sensitivity and FWHM is the full
Drude–Lorentz equation [24]: width at half maximum.
For any sensor, its resolution measures the precision with
ω2D 1ε2L which the sensor can detect an input change. For this case, the
εgold = ε∞ − − 2 , (2) sensor resolution can be measured using the following formula
ω(ω + j γ D ) ω − 2L + j 0 L ω
[25,26]:
where εgold is the dielectric constant of gold, ε∞ is the dielectric
constant at high frequencies, ω D is the plasma frequency, ω is 1n a 1λmin
R= (RIU), (7)
the angular frequency, γ D is the decaying frequency, and 1ε is 1λpeak
the weighting factor. The values of these parameters are available
where R represents the sensor’s wavelength resolution, 1n a is
in Ref. [24].
the variation in the refractive index of the analyte, 1λmin is the
An important indicator of the performance of PCF-SPR
minimum spectral resolution, and 1λpeak is the difference in
sensors is the confinement loss, which is expressed as follows
resonance peak shifts.
[25,26]:
 
2π dB
αloss = 8.686 × Im(n eff ) × 10 4
, (3) 3. RESULTS AND DISCUSSION
λ cm
The sensor design proposed herein was simulated based on the
where λ denotes the wavelength in µm and Im(n eff ) represents FEM to demonstrate and analyze the functionality of the design.
the imaginary part of the effective refractive index. Figure 2 illustrates the dispersion relationship between
The sensitivity of the sensor can be expressed in terms of WS the core-guided mode and the SPP mode at the analyte RI of
and AS [25,26]. The WS is measured as the change in the reso- 1.38 for the sensor structure described in Section 2. Since the
nance wavelength with a change in the refractive index of the tar- y-polarized mode shows a larger evanescent field, stronger inter-
get analyte. The formula for calculating the WS is as follows: action with the analytes, and higher loss depth as compared to
1λpeak  nm  the x-polarized mode, in this work, only the y-polarized mode
Sλ = (4) has been considered.
1n a RIU
The effective mode index of the core-guided mode and SPP
where 1λpeak is the shift in the wavelength peak in nm for the mode shows a steady variation with changes in wavelength, with
corresponding change 1n a in the refractive index of the target the intersection point being the resonance condition. Under this
analyte in RIU. condition, maximum optical energy leaks from the core to the
 100 Vol. 42, No. 1 / January 2025 / Journal of the Optical Society of America B Research Article

Table 3. R2 Values for Linear and Polynomial Fits

R2 Values for R2 Values for
Linear Fit Polynomial Fit Reference
0.9647 0.9978 [25]
0.93135 0.9876 [27]
0.95756 0.9909 [27]
0.95014 0.9768 [27]

Fig. 2. Dispersion relationship of the core-guided mode, SPP

mode, and loss spectra.

Fig. 4. Polynomial fit showing the relationship between resonance

Fig. 3. Distribution of y-polarized electric fields of the (a) core- wavelength and analyte refractive index.
guided mode, (b) SPP mode, and (c) resonance mode at the analyte RI
of 1.38.
A comparative listing of R2 values for linear and polynomial
fits in some previously reported works is mentioned in Table 3.
As a result, this paper utilizes a polynomial regression model to
Table 2. Wavelength Sensitivity for Different Analyte approximate the relationship between sensor output and input
RIs as follows:
Analyte Res. Peak Res. Peak Shift WS
RI Wavelength (µm) (nm) (nm/RIU) y = 50.379x 2 − 132.17x + 87.264, (8)
1.30 0.57 where y represents the resonance wavelength and x stands for
1.31 0.58 10 1000 the analyte refractive index. The R2 value of 0.973 is very close to
1.32 0.60 20 2000 unity, which indicates that the chosen model fits the data well. It
1.33 0.61 10 1000 also suggests that the changes in the peak resonance wavelength
1.34 0.63 20 2000 have a strong association with changes in the analyte RI, prob-
1.35 0.65 20 2000 ably the most important requirement of a RI sensing device.
1.36 0.69 40 4000 Quadratic regression has been used to model the relationship
1.37 0.73 40 4000
between the resonance wavelength and analyte refractive index.
1.38 0.79 60 6000
Higher-degree polynomials, though they have higher R2 values,
1.39 0.92 130 13,000
have not been chosen to avoid overfitting. Figure 4 shows the
relevant plot.
metal surface, resulting in a sharp loss peak. For the above plot, The confinement loss curve of the core mode changes
the confinement loss jumps to 483.64 dB/cm at the resonance with the changing refractive index of the analyte, as shown in
wavelength of 0.79 µm. Fig. 5(a). With an increase in the analyte refractive index, the
Figure 3 illustrates the distribution of y-polarized electric evanescent field of the core-guided light more strongly overlaps
fields for the core-guided, SPP, and resonant modes at the with the plasmonic field at the interface. This deepens the inter-
analyte RI of 1.38. In the core-guided mode, light is confined action between the evanescent field and the plasmonic field,
within the fiber core, whereas in the SPP mode, SPPs are formed which leads to greater absorption of light energy by the plasmon
in the metal-coated sensing region. In the resonance mode, waves and reduced confinement within the core of the fiber.
maximum optical energy leaks from the core-guided mode to Consequently, there is an upsurge in confinement loss, resulting
the SPP mode due to phase matching. in a decrease in the intensity of transmitted light through the
Table 2 illustrates the WS of the proposed sensor across vari- fiber core.
ous refractive indices of the analyte. As the analyte RI increases, The maximum wavelength resolution of the sensor is calcu-
the resonance peak wavelength experiences a red shift. This lated to be 7.69 × 10−6 RIU, which indicates the capability of
phenomenon occurs because the wavevector of the SPPs at the the sensor in detecting extremely minute RI changes of the order
metal–analyte interface increases. Consequently, to uphold of 10−6 .
phase matching, the light within the PCF must resonate at a Apart from the wavelength interrogation method, the
longer wavelength. The results indicate a nonlinear relationship amplitude interrogation method was also used to study the per-
between the resonance peak wavelength and the analyte RI. formance of the proposed sensor. The AS is shown in Fig. 5(b),
 Research Article Vol. 42, No. 1 / January 2025 / Journal of the Optical Society of America B 101

Fig. 5. (a) Confinement loss versus wavelength and (b) amplitude Fig. 6. (a) Dependence of the confinement loss curve on the air hole
sensitivity versus wavelength. diameter and (b) amplitude sensitivity variation with the air hole diam-
eter.

Table 4. Highest Amplitude Sensitivity at Different RI

Changes
Analyte RI Change Highest AS (RIU−1 ) Wavelength (µm)
1.30–1.31 77.07 0.60
1.31–1.32 101.79 0.61
1.32–1.33 138.80 0.62
1.33–1.34 189.54 0.64 Fig. 7. (a) Variation of the confinement loss curve with plasmonic
1.34–1.35 273.15 0.66 layer thickness and (b) amplitude sensitivity variation with plasmonic
layer thickness.
1.35–1.36 408.14 0.69
1.36–1.37 672.34 0.73
1.37–1.38 1538.90 0.79 of silica in the cladding. Initially, this might seem to promote
higher confinement loss due to reduced confinement efficiency.
Table 5. Air Hole Diameters for Different Values of
However, at a sufficiently small diameter, the fiber’s structure
Scale Factors may begin to resemble that of a more homogeneous medium
rather than a highly structured one. This uniformity can lead to
k1 d1 d2 d3 d4 d5 d6 d7 a more stable guiding condition and better confinement of light,
0.14 0.28 0.2016 0.1624 0.504 0.168 0.406 0.2044 resulting in reduced confinement losses.
0.15 0.30 0.2160 0.1740 0.540 0.180 0.435 0.219 The thickness of the plasmonic layer is a critical parameter
0.16 0.32 0.2304 0.1856 0.576 0.192 0.464 0.2336 in a PCF-based SPR sensor that influences its performance,
particularly the confinement loss at resonance conditions.
with the analyte RI being varied over the range of 1.30–1.38. Figure 7(a) shows the variation of confinement loss with the
Table 4 shows the AS of the proposed sensor over different ana- wavelength at plasmonic gold layer thicknesses of 30, 40, and
lyte RI transitions. The maximum refractive index resolution of 50 nm for an analyte RI of 1.38. The corresponding AS plots for
the proposed sensor obtained using the amplitude interrogation analyte RI changes from 1.37 to 1.38 are shown in Fig. 7(b). The
method is 6.50 × 10−6 RIU. During the calculation of both figures clearly indicate that the optimum plasmonic layer thick-
wavelength and amplitude resolutions of the sensor, practi- ness for maximum SPR effect is 40 nm. When the plasmonic
cal noise influence has not been considered, and it has been layer is thinner than this optimum thickness, the interaction
assumed that the sensor could detect a minimum 1% change of between the evanescent field of the guided mode and the sur-
the transmission intensity. face plasmons weakens, leading to a drop in the confinement
The study then examines the impact of the air hole diameter loss. Since the evanescent field decays exponentially from the
and the thickness of the plasmonic layer on the confinement loss fiber core into the cladding, it has a limited penetration depth.
of the sensor. For this purpose, the scale factor k1 , which affects Therefore, increasing the plasmonic layer thickness beyond
the diameters of all the air holes, is varied keeping all other scale the optimal value results in higher intrinsic metal losses that
factors fixed. Table 5 lists the hole diameters for different values do not contribute to the SPR effect, leading to a decrease in
of k1 . confinement loss.
Figure 6(a) shows the variation of confinement loss with The overall performance of the proposed sensor structure
wavelength at different air hole diameters for the analyte RI of can be expressed with the FOM parameter. This parameter
1.38. Figure 6(b) shows the variation of the AS of the proposed indicates the detection accuracy of a sensor, and it implies the
sensor with wavelength at different air hole diameters for a output proficiency of the sensor. The proposed design has a
change in analyte RI from 1.37 to 1.38. maximum FOM of 400 RIU−1 , which is an improvement over
The investigation reveals that both increasing and decreas- those reported in Refs. [1,3,14,15,24].
ing the diameter of the air holes can reduce confinement loss It is common to encounter deviations ranging from approxi-
under resonance conditions. Increasing the diameter of the mately ±2% to ±5% from the optimal design parameters dur-
air holes enhances the refractive index contrast between the ing the fabrication process of the sensor [28]. The proposed sen-
air-filled cladding and the silica core. This higher contrast can sor design has been simulated at analyte RI of 1.38 to test for fab-
lead to stronger confinement of the guided mode within the rication tolerance, with the results provided in Tables 6 and 7.
core, reducing the likelihood of light leaking into the cladding. Upon simulating up to five decimal points, the optimized
Conversely, decreasing the diameter of the air holes reduces the structure demonstrates a peak resonance at 0.78898 µm,
refractive index contrast by effectively increasing the amount accompanied by a loss of 546.34 dB/cm. The peak resonance
 102 Vol. 42, No. 1 / January 2025 / Journal of the Optical Society of America B Research Article

Table 6. Results for Fabrication Tolerance of the Air Table 7. Results for Fabrication Tolerance of Gold
Hole Diameter Thickness
Change (%) Res. Peak Wavelength (µm) Res. Peak Loss (dB/cm) Change (%) Res. Peak Wavelength (µm) Res. Peak Loss (dB/cm)
+5 0.79632 525.65 +5 0.79501 441.65
+4 0.79483 529.59 +4 0.79381 496.65
+3 0.79335 533.91 +3 0.79260 571.74
+2 0.79189 537.97 +2 0.79143 563.31
+1 0.79045 542.02 +1 0.79023 555.04
−1 0.78756 549.78 −1 0.78767 537.63
−2 0.78616 554.09 −2 0.78633 528.12
−3 0.78478 557.75 −3 0.78494 518.82
−4 0.78338 561.49 −4 0.78358 509.05
−5 0.78201 565.19 −5 0.78213 499.40

wavelength of 0.78898 µm is critical, as it marks the point of

Table 8. Sensor’s Performance Comparison
maximum coupling between the core-guided mode and surface
plasmon mode on the metal layer, maximizing the sensor’s Max.
sensitivity to refractive index changes. At this wavelength, even Sensing RI Max. WS Max. AS Resolution
slight variations in the surrounding medium’s refractive index Range (nm/RIU) (RIU−1 ) (RIU) Reference
cause measurable shifts, enabling precise analyte detection. This 1.20–1.34 10,900 N.A. 9.17 × 10−6
[1]
resonance wavelength, positioned in the near-infrared (NIR) 1.32–1.41 9000 1540 1.11 × 10−5 [2]
spectrum, aligns the sensor for optimal performance in sens- 1.33–1.44 11,200 505.04 8.92 × 10−6 [3]
ing weak biological and chemical signals. The peak resonance 1.23–1.29 13,000 333.8 7.69 × 10−6 [4]
wavelength exhibits a maximum red shift of 0.93% when the 1.33–1.37 4000 320 2.5 × 10−5 [5]
diameter of the air holes is changed by 5%, while it shows a 1.29–1.40 12,500 1189 8.0 × 10−6 [14]
maximum blue shift of 0.88% for a variation of −5%. These 1.32–1.40 7000 593.61 1.43 × 10−5 [15]
shifts occur due to slight increases in the air hole diameter, lower- 1.35–1.39 12,500 N.A. 8.0 × 10−6 [16]
1.39–1.40 10,700 1770 9.34 × 10−6 [17]
ing the effective refractive index of the fiber’s guided modes and
1.33–1.40 12,000 1656 8.33 × 10−6 [18]
resulting in a red shift toward longer wavelengths. Conversely, a
1.385–1.40 10,000 70 2 × 10−5 [19]
slight decrease results in a higher effective refractive index, lead- 1.33–1.35 10,000 1019.94 9.8 × 10−6 [22]
ing to a blue shift toward shorter wavelengths. Fabrication errors 1.33–1.36 2200 266 3.75 × 10−5 [23]
in depositing the gold layer cause the peak resonance wavelength 1.45–1.48 8000 700 1.25 × 10−5 [24]
to shift up to 0.76% longer for a 5% increase in gold layer thick- 1.33–1.38 4600 420.4 2.17 × 10−5 [25]
ness, and up to 0.87% shorter for a 5% decrease. Resonance 1.33–1.37 4200 300 2.38 × 10−5 [26]
requires alignment between the wavevector of light within the 1.33–1.39 6000 573.83 N.A. [29]
fiber and the plasmon waves on the gold surface. As the effective 1.33–1.40 10,300 2245 N.A. [30]
refractive index of the plasmon mode increases with a thicker 1.26–1.38 5626.86 N.A. 1.78 × 10−5 [31]
gold layer, the resonance condition (phase matching) shifts to 1.20–1.33 12,000 N.A. 8.3 × 10−6 [32]
longer wavelengths, causing a red shift in the peak resonance 1.30–1.39 13,000 1538.90 7.69 × 10−6 Proposed
wavelength. Similarly, with a decreased effective refractive index
of the SPP mode due to a thinner gold layer, the phase matching The high sensitivity of the PCF-SPR sensor is crucial in
condition for resonance shifts to shorter wavelengths, causing a detecting small changes in analyte concentrations, benefiting
blue shift in the peak resonance wavelength. The shift from the medical and chemical applications. In medical diagnostics, it
peak resonance wavelength of 0.78898 µm lies within 0.01 µm, enables early detection of low-concentration biomarkers, aiding
which is the minimum shift demonstrated by the sensor with in timely intervention and improved outcomes, especially for
RI change, as can be seen from Table 2. The design proposed chronic conditions like diabetes. In chemical applications, it
herein, thus, displays good tolerance to fabrication errors and enables the detection of trace contaminants in water, food,
is robust against errors of up to at least ±5%. This robustness and pollutants, safeguarding public health. This sensor can
supports scalability, facilitating mass production for industrial be used to detect a wide array of analytes whose refractive
and medical applications. Additionally, its tolerance to initial indices lie within the sensor’s measurable range of 1.30–1.39.
fabrication errors suggests long-term stability and reliability Notable among them are water contaminated by Vibrio cholerae
under real-world conditions, including exposure to temperature (RI = 1.365) [33], 10% and 20% glucose solutions in water
and physical stresses. The minor variations in peak resonance (RI = 1.347 and 1.3635, respectively), mucosa of the human
wavelength that creep in at higher levels of fabrication errors can intestine (RI = 1.329 to 1.338), acetone (RI = 1.36), ethanol
be taken care of easily by recalibration due to the high R2 value (RI = 1.361), human liver cells (RI = 1.369) [34], impor-
of this design. The results obtained show that this design stands tant components of human blood like WBCs (RI = 1.36),
out in comparison to many of the existing designs, as is evident hemoglobin (RI = 1.38) and plasma (RI = 1.35) [35], car-
from Table 8. cinomic basal cells (RI = 1.38) [36], chemicals like benzene
 Research Article Vol. 42, No. 1 / January 2025 / Journal of the Optical Society of America B 103

(RI = 1.366) and propanol (RI = 1.378) [37], glycerol concen- 11. A. K. Shakya and S. Singh, “Development of a generalized Fourier
trations up to 35% (RI = 1.3330 to 1.3766) [38], and expired transform model for distinct household oil samples by performing
spectroscopy analysis,” Results Opt. 10, 100355 (2023).
transformer oil (RI = 1.340) [39]. 12. A. A. Rifat, F. Haider, R. Ahmed, et al., “Highly sensitive selectively
coated photonic crystal fiber-based plasmonic sensor,” Opt. Lett. 43,
891–894 (2018).
4. CONCLUSION 13. L. Duan, X. Yang, Y. Lu, et al., “Hollow-fiber-based surface plasmon
resonance sensor with large refractive index detection range and
A new design of a wide RI range sensing PCF-SPR sensor is
high linearity,” Appl. Opt. 56, 9907–9912 (2017).
proposed and analyzed in this paper. The sensor was studied 14. A. M. T. Hoque, K. F. Al-Tabatabaie, M. E. Ali, et al., “U-grooved selec-
in the visible and near-infrared wavelength range over a refrac- tively coated and highly sensitive PCF-SPR sensor for broad range
tive index range of 1.30–1.39. The designed sensor exhibited analyte RI detection,” IEEE Access 11, 74486–74499 (2023).
15. N. Islam, M. F. H. Arif, M. A. Yousuf, et al., “Highly sensitive open
the highest WS of 13,000 nm/RIU, wavelength resolution of
channel based PCF-SPR sensor for analyte refractive index sensing,”
7.69 × 10−6 RIU, AS of 1538.90 RIU−1 , amplitude resolution Results Phys. 46, 106266 (2023).
of 6.50 × 10−6 RIU, and FOM of 400 RIU−1 . The sensor 16. Y. Wang, S. Li, M. Wang, et al., “Refractive index sensing and filter-
design reported herein has the advantages of an outer analyte ing characteristics of side-polished and gold-coated photonic crystal
fiber with a offset core,” Opt. Laser Technol. 136, 106759 (2021).
channel, outer sensing method, high wavelength and amplitude
17. A. Shafkat, “Analysis of a gold coated plasmonic sensor based on a
sensitivities, excellent resolution, and a very good FOM. Thus, duplex core photonic crystal fiber,” Sens. Bio-Sens. Res. 28, 100324
the sensor proposed in this paper is suitable for being used in (2020).
sensing a wide range of chemical and biological analytes with a 18. M. A. Mahfuz, M. A. Mollah, M. R. Momota, et al., “Highly sensitive
high degree of accuracy and precision. photonic crystal fiber plasmonic biosensor: design and analysis,”
Opt. Mater. 90, 315–321 (2019).
Acknowledgment. The authors acknowledge technical support from 19. V. Kaur and S. Singh, “Design of titanium nitride coated PCF-SPR
the I-STEM (Indian Science, Technology, and Engineering facilities Map) sensor for liquid sensing applications,” Opt. Fiber Technol. 48,
159–164 (2019).
program, funded by the Office of the Principal Scientific Adviser to the Govt.
20. T. Wu, Y. Shao, Y. Wang, et al., “Surface plasmon resonance bio-
of India. The authors would like to thank all their colleagues in their research
sensor based on gold-coated side-polished hexagonal structure
group for their help and useful suggestions.
photonic crystal fiber,” Opt. Express 25, 20313–20322 (2017).
21. S. Singh and Y. K. Prajapati, “Highly sensitive refractive index sensor
Disclosures. The authors declare no conflicts of interest.
based on D-shaped PCF with gold-graphene layers on the polished
surface,” Appl. Phys. A 125, 437 (2019).
Data availability. Data underlying the results presented in this paper are
22. M. F. O. Hameed, Y. K. A. Alrayk, and S. S. A. Obayya, “Self-
not publicly available at this time but may be obtained from the authors upon
calibration highly sensitive photonic crystal fiber biosensor,” IEEE
reasonable request.
Photonics J. 8, 6802912 (2016).
23. M. R. Hasan, S. Akter, A. Rifat, et al., “A highly sensitive gold-
coated photonic crystal fiber biosensor based on surface plasmon
REFERENCES resonance,” Photonics 4, 18 (2017).
1. H. Wang, J. Zhao, X. Yi, et al., “A low refractive index microstructured 24. M. N. Sakib, M. B. Hossain, K. F. Al-tabatabaie, et al., “High perform-
fiber sensor with wide range detection,” Plasmonics 19, 327–334 ance dual core D-shape PCF-SPR sensor modeling employing gold
(2024). coat,” Results Phys. 15, 102788 (2019).
25. M. R. Hasan, S. Akter, A. A. Rifat, et al., “Spiral photonic crystal fiber-
2. S. M. M. Monazah, F. Emami, and M. R. Salehi, “Design and analysis
based dual-polarized surface plasmon resonance biosensor,” IEEE
of nano gold coated refractive index sensor with asymmetric multiple
Sens. J. 18, 133–140 (2018).
holes,” Plasmonics 18, 931–940 (2023).
26. M. R. Momota and M. R. Hasan, “Hollow-core silver coated photonic
3. Q. M. Kamrunnahar, J. R. Mou, and M. Momtaj, “Dual-core gold
crystal fiber plasmonic sensor,” Opt. Mater. 76, 287–294 (2018).
coated photonic crystal fiber plasmonic sensor: design and
27. Z. Fan, S. Chu, X. Zhang, et al., “Two kinds of liquid crystal filled PCFs
analysis,” Results Phys. 18, 103319 (2020).
temperature and RI sensors based on SPR,” IEEE Sens. J. 23, 5766–
4. C. Liu, L. Yang, X. Lu, et al., “Mid-infrared surface plasmon reso-
5772 (2023).
nance sensor based on photonic crystal fibers,” Opt. Express 25,
28. M. R. Hasan, M. I. Hasan, and M. S. Anower, “Tellurite glass defect-
14227–14237 (2017).
core spiral photonic crystal fiber with low loss and large negative flat-
5. A. A. Rifat, G. A. Mahdiraji, Y. A. Sua, et al., “Surface plasmon
tened dispersion over S + C + L + U wavelength bands,” Appl. Opt.
resonance photonic crystal fiber biosensor: a practical sensing 54, 9456–9461 (2015).
approach,” IEEE Photonics Technol. Lett. 27, 1628–1631 (2015). 29. N. Islam, M. M. U. Masum, M. F. H. Arif, et al., “Enhanced sensitivity
6. A. K. Shakya and S. Singh, “Novel merger of spectroscopy and of open channel SPR-based PCF sensor employing plasmonic mate-
refractive index sensing for modelling hyper sensitive hexa-slotted rials for analyte sensing,” Plasmonics 17, 2075–2087 (2022).
plasmonic sensor for transformer oil monitoring in near-infrared 30. H. Fang, C. Wei, H. Yang, et al., “D-shaped photonic crystal fiber
region,” Opt. Quant. Electron. 55, 764 (2023). plasmonic sensor based on silver-titanium dioxide composite
7. A. K. Shakya and S. Singh, “Designing of a novel PCF biosensor hav- micro-grating,” Plasmonics 16, 2049–2059 (2021).
ing octagonal core and based on SPR for chemical and heavy metal 31. H. Pan, F. Pan, A. Zhanga, et al., “Wide refractive index detection
sensing,” in 12th International Conference on Cloud Computing, range surface plasmon resonance sensor based on D-shaped
Data Science & Engineering (Confluence), Noida, India, 2022, photonic crystal fiber,” Opt. Quant. Electron. 54, 393 (2022).
pp. 171–175. 32. J. Wang, L. Pei, L. Wu, et al., “A polarization-independent SPR sensor
8. A. K. Shakya and S. Singh, “State of the art alliance of refractive index based on photonic crystal fiber for low RI detection,” Plasmonics 15,
sensing and spectroscopy techniques for household oils analysis,” 327–333 (2020).
Plasmonics 18, 2347–2364 (2023). 33. E. Haque, A. A. Noman, and F. Ahmed, “Numerical investigation of
9. A. K. Shakya, F. Cheng, and T. Carmon, “Ultracoherent emission by photonic crystal fiber-based biosensor for pathogens detection in
orthogonal lasers,” Proc. SPIE 13012, 130120F (2024). water,” IEEE Access 10, 88885–88893 (2022).
10. A. K. Shakya and S. Singh, “Performance analysis of a developed 34. M. R. Islam, M. M. I. Khan, A. J. Yeasir, et al., “Design and analysis of
optical sensing setup based on the Beer-Lambert law,” Plasmonics a highly sensitive SPR based PCF biosensor with double step dual
19, 447–455 (2024). peak shift sensitivity,” Heliyon 9, e18782 (2023).
 104 Vol. 42, No. 1 / January 2025 / Journal of the Optical Society of America B Research Article

35. S. Jain, K. Choudhary, and S. Kumar, “Photonic crystal fiber-based 38. Y. Wang, Q. Huang, W. Zhu, et al., “Novel optical fiber SPR temper-
SPR sensor for broad range of refractive index sensing applications,” ature sensor based on MMF-PCF-MMF structure and gold-PDMS
Opt. Fiber Technol. 73, 103030 (2022). film,” Opt. Express 26, 1910–1917 (2018).
36. A. Ramola, A. Marwaha, and S. Singh, “Design and investigation of 39. A. K. Shakya, A. Ramola, S. Singh, et al., “Design of an ultra-sensitive
a dedicated PCF SPR biosensor for CANCER exposure employing bimetallic anisotropic PCF SPR biosensor for liquid analytes
external sensing,” Appl. Phys. A 127, 643 (2021). sensing,” Opt. Express 30, 9233–9255 (2022).
37. A. A. Rifat, K. Ahmed, S. Asaduzzaman, et al., “Development of pho-
tonic crystal fiber-based gas/chemical sensors,” in Computational
Photonic Sensors, M. Hameed and S. Obayya, eds. (Springer, 2019).