Enhanced aqueous refractive index sensing using an AlSb/BiFeO 3 /AlSb multilayer-integrated surface plasmon resonance platform

1 Introduction

Water is among the most well-known resources needed to sustain life, but it is becoming an extremely threatened aspect due to the immense pollution of the industrial effluents, agricultural runoff, heavy metals and chemical pollutants. Such pollutants even in low concentrations can cause a disruption in the aquatic system, even accumulating in the food web and later leading to extreme threats to human life (Rathod et al., 2024). Therefore, the water resources quality and safety have become the burning question throughout the world. To this, rapid, accurate and sensitive sampling of contaminants is required. The conventional water quality tests, such as chemical analysis, spectroscopic analysis and microbial test are still prevalent but are associated with glaring limitations (Zainurin et al., 2022). They might be tiresome and consuming in terms of time and money and require the services of experienced staff and sophisticated laboratory equipment. More so, these methods are not necessarily sensitive enough to detect low concentrations of harmful substances although these can as well have devastating ecological and health consequences. These inadequacies have created a pressing requirement to establish novel sensing technologies that are not only reliable but can be deployed in practice in field and in real time.

One such alternative has been Surface Plasmon Resonance (SPR) sensors in this regard. The optical phenomenon known as the phenomenon of SPR occurs when polarized light is excited and forms plasmons on the surface on the border of a dielectric and a metal (Zhang et al., 2022). The resonance condition of the probe is highly sensitive to the change in rods of the refractive index of the sensor surface, and thus the SPR sensors are sensitive devices in identifying the biomolecular interaction and environmental pollution. In particular, SPR sensors do not require an additional labeling agent (compared to other common biosensing techniques) in order to operate in real-time and hence can streamline the process of detection and can save on expenses (Anand et al., 2022). However, besides these advantages, there are the performance disadvantages of traditional SPR sensors as sensing water. The resonance signals are most likely diffuse and therefore lack accuracy, sensitive and with low reliability. This is a drawback in large scale water quality monitoring where its accuracy and sensitivity are crucial at trace levels. In this regard, researchers are striving to come up with better designs and new materials to enhance the workings of SPR sensors to sharper resonance curves, higher sensitivity, and long-life operation. In recent years, the progress of SPR sensing techniques has shown that integrating dielectric materials, semiconductor layers and two-dimensional nanomaterials into metal structures can also significantly enhance electromagnetic field confinement, resonance sharpness and refractive index sensitivity (Nayan et al., 2025). Various high-refractive-index dielectric materials in combination with functional nanostructures have been utilized in multilayer SPR designs for applications such as environmental monitoring, chemical sensing, and biomedical diagnostics. It has been demonstrated that multilayer architectures can be strategically crafted to overcome these limitations-for improved sensitivity, accuracy and operational stability of conventional single-metal-based SPR sensors (Fuad et al., 2025; Mahmud et al., 2025). However, the prospects of AlSb/BiFeO₃-based multilayers for aqueous refractive index sensing have mostly not been investigated until now, which is what we mainly motivate here. Along with theoretical studies, multilayer SPR sensors have been practically verified by various experimental works focused on the fabrication of metal–dielectric and semiconductor-enhanced structures using thin-film depositing methods. These experimental explorations have consistently demonstrated enhancements in sensitivity, resonance properties, and sensing stability leading to the validation of a multilayered architecture for refractive index sensing. Motivated by the encouraging agreement between the experimental observations and theoretical predictions reported in literature, the numerical design and optimization presented here provides a strong motivation (Lv et al., 2024; Chochol et al., 2017).

To address these challenges, we have suggested a new multilayer SPR sensor Prism - A g − A l S b − B i F e O 3 − A l S b - Sensing Medium that consists of Prism - A g − A l S b − B i F e O 3 − A l S b - Sensing Medium layers. Each of the materials, namely, silver (A g), is the primary plasmonic material, where the plasmids are most easily excited (Helinski, 2022), and aluminium antimonide (A l S b) is utilized to prevent the oxidation of the silver, and to enhance the degree of confinement of the plasmonic field in addition to bismuth ferrite (B i F e O 3), a multiferroic material exhibiting a strong optical response (Shukla et al., 2025). By a combination of these materials, the proposed multilayer sensor design has the ability to sharpen resonance signal in addition to improving stability and durability of sensor in aqueous condition. By modeling and optimizing and SPR sensor multilayer sensors at 633 nm, we have demonstrated that this is more sensitive, more precise and robust than the conventional designs made with single metal (Fernandes et al., 2024). Such advances can directly be observed in the form of the enhanced ability to detect minor water quality alterations, and the sensor is highly suitable in real-time monitoring of pollutants and contaminants.

This innovation to make multilayer this kind of SPR sensor is a good step towards the next-generation of water quality monitoring technology. This approach is also useful in ensuring the safety of the environment and the health of the population as it equips them with quicker, more dependable, and sensitive detection opportunities. Besides, these sensors also may be integrated into portable systems and automated measurements with the use of the flexibility of SPR-based designs, thereby, offering an opportunity to carry out nonstop and decentralized measurements of water quality in industrial settings, agriculture settings, and metropolitan settings (Bk et al., 2025). The main aim of this study is to produce a multilayer SPR sensor with an integrated AlSb / BiFeO₃ / AlSb architecture that achieves high performance for aqueous refractive index sensing applications. The innovation of the proposed design is defined by the synergistic application of high-refractive-index AlSb semiconductor layers with a BiFeO₃ dielectric layer to enhance electromagnetic field confinement, resonance characteristics and consequently sensing performance. Moreover, a systematic optimization based on TMM of prism material and Ag layer thickness is also performed, and the corresponding performance of sensors in sensitivity, QF, SNR, DA were further compared followed by CSF.

2 Design and methodology

This section describes the design considerations and methodology that uses for the performance analysis of the proposed sensor.

2.1 Design considerations

As shown in Figure 1, this work uses a proposed SPR sensor for detection of aqueous samples at a wavelength of 633 nm (Fernandes et al., 2024). The suggested sensor is based on the Kretschmann configuration owing to the Kerr-effect. As in Table 1, to couple the light, we optimize the prism glass by comparing the sensitivity performance of various prisms like BK7, BAK1, BAF10 and SF10 with RIs of 1.515 (Ashrafi and Mohanty, 2022), 1.5704 (Imam et al., 2026), 1.6671 (Yurin et al., 2024) and 1.7231 (Andrade et al., 2025), respectively, at wavelength of 633 nm. In addition, proposed sensor consists of Ag as a metallic layer, which is coated on the surface of prism in order to create the surface plasmon with the RI of 0.0562 + 4.27761. The thickness of metallic layer is optimized with the help of iteration technique and due to that, it is different with respect to the dielectric layer, which is in contact with the metallic layer. A l s b uses over metallic with RI of 3.8638 + 0.003582i and thickness of 0.53 nm in order to shield the metallic layer from its corrosion. The main reason to choose AlSb was for its good optical properties and high refractive index that can improve the optical response of the multilayer SPR structure proposed. In the current work, AlSb is viewed not as a long-term oxidation barrier for silver, but one that instead enhances optical performance. As the proposed study is motivated by theoretical optimization, further experimental validations are needed to assess the long-term chemical stability of ultrathin AlSb layers in aqueous operating conditions using appropriate surface passivation or encapsulation strategies. Thereafter, B i F e O 3 is coated onto A l S b with RI of 2.9680 and thickness of 1 nm, and it is sandwitched between two A l S b layers as shown in Figure 1. Finally, the sensing medium is coated over A l S b, and it has presence of aqueous samples having RI values of 1.33 and 1.35. The optimization was carried out through systematic parametric sweeps using the Transfer Matrix Method (TMM). In order to find the configuration of the configuration with a better sensing performance, four prism materials (BK7, BAK1, BAF10 and SF10) were tested under identical operating conditions. In every step of the optimizations, the thickness of Ag layer was swept from 30 nm to 69 nm with an increment of 3 nm, while all other layer’s parameters were kept constant. The resonance angle, minimum reflectance, full width at half maximum (FWHM), sensitivity, quality factor (QF), signal-to-noise ratio (SNR) and combined sensitivity factor (CSF) were determined for each of the configuration. The best combination was chosen so that maximum sensitivity and minimum reflection were attained at the same time with narrower FWHM with higher QF. Since all candidate configurations within the predefined parameter ranges were systematically evaluated, the reported optimum represents the best-performing configuration within the investigated design space.

TABLE 1. Layer material names, RI and thickness.

It is important to note that the optimized layer thicknesses supplied in Table 1 were calculated using numerical optimization according to the Transfer Matrix Method (TMM), describing a theoretically optimum sensor. In practice, however, the deposition of ultrathin AlSb layers with a nominal thickness close to 0.53 nm is experimentally demanding since these atomic-scale films are vulnerable to island growth, interface roughness and interdiffusion during fabrication. However, with more advanced thin-film deposition techniques like molecular beam epitaxy (MBE), which provides atomic precision in control of thickness, and well-optimized deposition procedures, one can obtain better uniformity and crystal structure quality of the film as well as interface quality. In practice, modest departures from the optimum thickness may be found due to fabrication tolerances and therefore adequate experimental optimization is required for practical application of such a multilayer structure without losing its sensing functions.

2.2 Methodology

In this section, we express the mathematical model to describe the transmitting of p-polarization light through multilayer SPR structure and calculate its reflectance property under resonance circumstance. These derived equations are employed to calculate the resonance angle, reflectance profile, and parameters of this sensor, i.e., sensitivity, quality factor (QF), signal-to-noise ratio (SNR), detection accuracy (DA) and combined sensitivity factor (CSF). The transfer matrix method (TMM) is an exact method widely applied in the modeling of optical multilayer sensors to determine their reflectance (Mackay and Lakhtakia, 2022). When light is incident on the sensor at an angle, maximum energy is absorbed, leading to a minimum reflectance sensed at the output. The theoretical analysis of the multilayer SPR sensor model is performed with one of the most-studied analytical models for calculating the propagation of electromagnetic waves in planar multilayer structures, named TMM (Elsayed et al., 2024). Since the proposed sensor consists of a prism-coupled one-dimensional layered configuration, TMM enables accurate calculation of the optical reflectance, resonance angle, and associated sensing performance with significantly lower computational complexity than full-wave numerical techniques (Luce et al., 2022). In contrast, finite element method (FEM) and finite-difference time-domain (FDTD) approaches are generally more suitable for complex two-dimensional or three-dimensional geometries requiring detailed spatial electric field distributions (Villarim et al., 2023). Therefore, TMM was selected as the most appropriate and computationally efficient method for the design, optimization, and performance evaluation of the proposed multilayer SPR sensor. This particular angle is referred to as the resonance angle and can be defined mathematically as indicated in Equation 1 with the prism, metal, and sensing medium refractive indices represented by, n p , n m and n S, respectively. When analyzing the reflectivity of the sensor, Equations 2, 3 are used, with parameters and variables discussed in Equations 4–8, the transverse refractive index of every layer Equation 4. The characteristic matrix of the overall multilayer sensor irradiated by p-polarized light is provided in Equation 5, and Equation 6 introduces the phase constant of concern with each layer. Furthermore, input angle and wave impedance within the layers are measured using Equations 7, 8; the thickness, permittivity, and permeability of each of these are indicated as μ x, ε x and d x respectively.where θ S P R is the SPR angle, n p is the refractive index of the prism, n m is the refractive index of the metallic (plasmonic) layer, and n d is the refractive index of the dielectric (sensing) medium. The reflectance of the p-polarized incident light is calculated using the reflection coefficient, which determines the resonance characteristics of the proposed SPR sensor.where F p is the reflectance of the p-polarized light, f p is the complex reflection coefficient for p-polarized light, and f p * represents its complex conjugate. The complex reflection coefficient is obtained from the characteristic transfer matrix of the multilayer structure and is expressed as follows.where F 11, F 12, F 21, and F 22 are the elements of the overall characteristic transfer matrix, b 1 is the optical admittance of the incident medium, and b x is the optical admittance of the exit (sensing) medium. The optical admittance and propagation angle for each layer are calculated using the following expressions.where b x is the optical admittance of the x th layer, μ x is the magnetic permeability, ε x is the dielectric permittivity, n p is the refractive index of the prism, and θ x is the propagation angle within the x th layer. The overall transfer matrix of the multilayer SPR structure is obtained by multiplying the characteristic matrices of the individual layers.where F is the overall characteristic transfer matrix of the multilayer structure, and F 11, F 12, F 21, and F 22 denote its individual matrix elements. The phase shift introduced by each layer during electromagnetic wave propagation is calculated using Equation 6.where β x is the phase shift in the x t h layer, λ is the wavelength of the incident light, d x is the thickness of the x t h layer, r x is the refractive index of the x t h layer, and θ x is the propagation angle inside that layer. The propagation angle inside each layer is determined using Snell’s law, as given below.where θ x is the propagation angle in the x th layer, while b x − 1 and b x represent the optical admittances of the x − 1 th and x th layers, respectively. The wave vector component associated with the multilayer structure is determined using Equation 8.where s k is the wave vector component associated with the k t h layer, w x is the angular frequency of the incident light, b k is the optical admittance of the k t h layer, c is the speed of light in free space, λ 633 is the operating wavelength (633 nm), and ε k is the dielectric permittivity of the k t h layer. The following performance metrics are employed to quantitatively evaluate the sensing capability of the proposed SPR sensor. By putting these parameters together, the reflectivity of the six-layer sensor structure is calculated by Equation 9, where the reflection coefficients f 12, f 23, f 34, f 45, f 56 and f 67 and the thickness (d m) and wave vector (w x) components of each layer feature heavily.

θ S P R = sin − 1 n m 2 n d 2 n p 2 n m 2 + n d 2 (1)

F p = f p f p * = f p 2 (2)

f p = F 11 + F 12 b x b 1 − F 21 + F 22 b x F 11 + F 12 b x b 1 + F 21 + F 22 b x (3)

b x = μ x ε x 1 / 2 cos ⁡ θ x = ε x ‐ n p ⁡ sin ⁡ θ 2 ε x 2 (4)

F if = ∏ x = 2 N ‐ 1 cos ⁡ β x − i ⁡ sin ⁡ β x − i n x ⁡ sin ⁡ β x cos ⁡ β x if = F 11 F 12 F 21 F 22 (5)

β x = 2 π λ r x ⁡ cos ⁡ θ x s x − s x ‐ 1 = 2 π λ d x ε x − r p ⁡ sin ⁡ θ 2 (6)

θ x = cos 1 − b x ‐ 1 / b x sin ⁡ θ 2 − 1 (7)

s k = w x b k ⁡ cos ⁡ θ k 2 π c / λ 633 ε k 2 (8)

F 123456 = f 123456 2 = f 12 + f 23456 e 2 i w 2 d 2 1 + f 12 f 23456 e 2 i w 2 d 2 2

f 23456 = f 23 + f 3456 e 2 i w 3 d 3 1 + f 23 f 34567 e 2 i w 3 d 3 , f 3456 = f 34 + f 456 e 2 i w 4 d 4 1 + f 34 f 4567 e 2 i w 4 d 4 , f 456 = f 45 + f 567 e 2 i w 5 d 5 1 + f 45 f 56 e 2 i w 5 d 5 , f 12 = ε 2 / w 2 ‐ ε 1 / w 1 ε 2 / w 2 + ε 1 / w 1 , f 23 = ε 3 / w 3 − ε 2 / w 2 ε 3 / w 3 + ε 2 / w 2 , f 34 = ε 4 / w 4 ‐ ε 3 / w 3 ε 4 / w 4 + ε 3 / w 3 , f 45 = ε 5 / w 5 ‐ ε 4 / w 4 ε 5 / w 5 + ε 4 / w 4 , f 56 = ε 6 / w 6 ‐ ε 5 / w 5 ε 6 / w 6 + ε 5 / w 5 (9)

Performance metrics including sensitivity, quality factor (QF), detection accuracy (DA) and sensor resolution factor (CSF) are critical in assessing the sensor. Mathematical descriptions are given in Equations 10–13. The parameters account for change in resonance angle (∇ θ r e s) and change in analyte refractive index change (∇ n), full width at half maximum (FWHM) of the SPR curve, and minimum reflectance, maximum reflectance (Zhang et al., 2024). FWHM characterizes the narrowness of the resonance dip, which directly relates to sensor resolution (Zhang et al., 2024). Collectively, these values give a complete measure of the effectiveness of the sensor and are most important to SPR sensor design for biological and medical diagnostics (Das et al., 2023).

S e n s i t i v i t y = ∇ θ r e s / ∇ n ° / R I U (10)

Q F = S e n s i t i v i t y / F W H M R I U − 1 (11)

D A = ∇ θ r e s / F W H M (12)

CSF = Sensitivity × R max − R min / FWHM (13)

3 Results and discussion

The performance evaluation and simulation analysis of the suggested multilayer SPR sensor are presented in this section. To improve sensing performance, the metallic layer thickness, prism material, and dielectric combinations are systematically examined. Using the transfer matrix method (TMM) at a wavelength of 633 nm, reflectance curves and resonance parameters are acquired for various prism configurations and Ag thicknesses (Babu et al., 2023).

3.1 Optimizing Ag’s thickness and prism to enhance the sensor performance

The plasmonic metal layer is important for the coupling efficiency and resonance characteristics of the SPR sensor. In this work, silver was chosen as the plasmonic material due to its lower optical loss and higher conductivity compared with gold (Novaković et al., 2022). In order to investigate the best configuration, the Ag layer thickness is varied from 30 nm to 69 nm in 3 nm steps. R min, full-width half maximum (FWHM), and resonance angle θ r e s are monitored. Figures 2–5 and Tables 2–5 lay out the reflectance behaviour for four prism types—BK7, BAK1, BAF10 and SF10—when the surrounding medium’s refractive index is set to 1.33 or 1.35. The trend is unmistakable: as the silver film thickens the resonance troughs become acutely sharper (Karki et al., 2023). The reflectance (R min) drops betraying a more vigorous coupling of the evanescent wave to the surface plasmons. The measured enhancement in resonance very significant characteristics with increase of Ag thickness can be explained by the coupling of surface plasmons at the metal–dielectric interface with incident electromagnetic waves. However, beyond the optimum thickness, excessive optical absorption reduces plasmon excitation efficiency, leading to broader resonance curves and a decline in sensing performance.

TABLE 2. Measured parameters for Figure 2.

TABLE 3. Measured parameters for Figure 3.

TABLE 4. Measured parameters for Figure 4.

TABLE 5. Measured parameters for Figure 5.

Figure 2 and Table 2 show that with an increase in the thickness of the Ag, the BK7 prism has a strong and sharp resonance dip. The resonance curve is narrowed successively, which means a better angular resolution. The maximum performance is achieved at Ag thickness of 57 nm when the reflectance is at its lowest (0.0081), the FWHM is at its lowest (1.02), and sensitivity is the highest (137.74 ° / R I U). At this thickness the corresponding QF is 134.67 R I U − 1, which confirms great accuracy of detection and signal quality.

In the case of BAK1 prism, as demonstrated in Figure 3 and Table 3, there is an increasing trend in the values with the increase in the Ag thickness, but the sensitivity and QF values are relatively lower than that of BK7. The highest sensitivity of 108.9 ° / R I U occurs at 57 nm Ag thickness and a larger FWHM, which is a disadvantage to the overall factor of quality.

Figures 4a,b and Table 4 present the results of the SPR sensor with the use of the BAF10 prism. As the Ag thickness is increased, the resonance dips become sharper with smaller FWHM and larger angular shift. The optimal response is at 54–57 nm with a sensitivity of 83.28 ° / R I U and quality factor of 92.6 R I U − 1. Nonetheless, the resonance curves in Figure 4 have a broader resonance than that which is received using lower-index prisms. This is because there is greater optical confinement in BAF10 prism, which limits with the sensing medium.

Figures 5a,b shows the reflectance response of the SF10 prism and Table 5 is the list of the corresponding parameters. The angular shift is relatively small though an enhancement in the Ag thickness with increased thickness will enhance coupling with a narrower resonance dip. Table 5 shows a sensitivity of 73.68 ° / R I U at Ag thickness of 54–57 nm (maximum). The quality factor is less than 100 R I U − 1, which represents a worse accuracy in detection. High index prism is prone to excessively confine the field, and that is why it is not applicable in aqueous sensing since its interaction with the plasmon decreases.

The tests on all prisms showed that BK7 had the greatest angular sensitivity and a resonance curve with the narrowest width (Ashrafi and Mohanty, 2022), whereas SF10 had a relatively high intensity of reflectance and low sensitivity at the same time due to its high refractive index (Andrade et al., 2025). The configuration that involved BK7 and Ag had its optimal thickness of Ag set to 55 nm, where the minimum reflectance got down to 0.0081, FWHM was 1.02 ° and sensitivity was 137.74 ° / R I U. Thus, the quality factor (QF) was 134.67 R I U − 1. The moderate refractive index of BK7 (1.5151) ensures efficient coupling of incident light into surface plasmons while maintaining strong field penetration into the sensing region. Consequently, the BK7–Ag configuration is selected as the optimized base structure for further enhancement using dielectric and functional layers in the proposed multilayer SPR sensor (Akjouj and Mir, 2022).

3.2 Proposed structure impact

To enhance the performance of the sensor, the multilayer structure Prism-A g - A l S b - B i F e O 3 - A l S b - Sensing Medium was considered. The combination of AlSb and B i F e O 3 enhances the trapping of surface plasmons at the dielectric and metal interface while reducing damping effects. This resulted in not only sharper reflectance dips but also larger angular shifts for small refractive index variations in the analyte. Figure 6 and Table 6 show the reflectance responses for different structural configurations S1 to S4. Structure S1, comprising BK7, A g, A l S b, B i F e O 3, A l S b, and SM, exhibited the maximum angular shift Δθ = 2.43 ° and the highest sensitivity of 121.72 ° / R I U, thus outperforming configurations S2, S3, and S4. For this design, the FWHM was 0.84 °, giving a QF value of 144.95 R I U − 1 and an SNR of 2.9.

TABLE 6. Measured parameters for Figure 6.

Figure 7 illustrates the reflectance curves for different numbers of first A l S b layers at sensing medium refractive indices of 1.33 and 1.35. With more layers, a resonance dip is observed to move towards higher incident angles which points to improvement in the interaction between plasmids and the analyte. Table 7 indicates that the sensitivity increases with the number of layers as 137.74 ° / R I U to 176.175 ° / R I U. Beyond seven layers, there is a drastic improvement. The highest angular movement of 7.3673^o is achieved at 9 layers. This gives the higher sensitivity of 368.36 ° / R I U. Therefore, adding the first layers of AlSb increases sensor performance considerably.

TABLE 7. Performance analysis of 1^st A l S b layer variation.

The reflectance response to different numbers of second A l S b layers with the same change in refractive index is shown in Figure 8. The resonance angle moves consistently like the increase in the layer number to higher values. This verifies enhanced optical confinement and better coupling of plasmids. Table 8 shows that the sensitivity also becomes higher as the number of layers increase gradually; it is 137.7 ° / R I U to 233.835 ° / R I U. There is a significant enhancement at higher levels. At a ten-layers deep, the maximum angular shift of 7.3674^o is obtained. As a result, the maximum sensitivity of 368.37 ° / R I U is achieved. This shows that the optimization of the second A l S b layer is as well a result in ultra-high sensitivity performance.

TABLE 8. Performance analysis of 2^nd A l S b layer variation.

The One Dimensional (1D) plot used to depict the Electric Field Intensity distribution versus the distance between the prism and metal interface (z) at the proposed multilayer SPR sensor structure is shown in Figure 9. The X-axis is used to indicate the distance in nano meters (nm) and the Y-axis indicates Electric field intensity (E x 2). It is also clearly observed that the maximum intensity of the electric field is at the metal–dielectric interface where surface plasmids are highly excited. The field intensity gradually reduces exponentially with distance beyond the interface, which validates the evanescent character of the plasmonic wave. The optimized structure resulted in the calculation of the penetration depth at 332.04 nm, which is high and it shows that the field extends considerably into the aqueous analyte region. Thus, the 1D field distribution of Figure 9 confirms the enhanced sensitivity of the sensor and high coupling of plasmon to the analyte of the proposed design.

3.3 Fabrication feasibility steps

Figure 10 explains the fabrication steps of the proposed multilayer SPR sensor. First, the BK7 prism with refractive index (RI) of 1.5151 is cleaned to remove dust and unwanted particles. Then, a thin layer of silver (Ag) is deposited on the prism surface, which helps in generating surface plasmons (Angadala et al., 2025). After that, a very thin A l S b layer is coated to protect the silver layer and improve performance. Next, a B i F e O 3 layer is added to increase the interaction between light and the sensing region. Another A l S b layer is deposited on top to form a stable sandwich structure. All thin-film layers are assumed to be deposited with precise thickness control using established thin-film deposition techniques to ensure reproducibility and optimal optical performance. Finally, the sensing medium with RI values of 1.33–1.35 is placed above the structure to detect changes in the sample. This completes the fabrication of the proposed SPR sensor.

Although the present work is based on numerical simulations using the Transfer Matrix Method (TMM), the proposed multilayer SPR sensor has been designed with practical fabrication feasibility in mind. The Ag plasmonic layer can be deposited using magnetron sputtering or thermal evaporation, whereas the BiFeO₃ thin film can be fabricated using pulsed laser deposition (PLD). The AlSb layers may be realized using molecular beam epitaxy (MBE) or metal–organic chemical vapor deposition (MOCVD), which are well-established techniques for high-quality semiconductor thin-film growth. After fabrication, the experimental characterization of angular interrogation SPR sensor is performed by prism coupling to verify predicted resonance characteristics, sensitivity, quality factor and detection limit. In practical implementation, this sensing performance may be compromised by fabrication tolerances, such as layer thickness variations, fabrication-related surface roughness and interfacial defects that arise during thin-film deposition. These contribute to small shifts in the resonance angle, as well as broadening of the resonance dip resulting in minimal degradation of sensitivity and detection accuracy. Moreover, oxidation of the Ag layer under long operation time may weaken its plasmonic performance. In the intended structure, the AlSb layers will give partial protection of the Ag film at the same time expanding electromagnetic field confinement. In addition, the high chemical stability of the BiFeO₃ layer renders higher durability to this multilayer architecture, resulting in its suitability for long-term aqueous sensing. Future experiments will focus on how the practical performance of the sensor changes with variations in fabrication tolerances and environmental stability. The sensing parameters are critical parameters for evaluating SPR performance. The optimized sensor achieved a maximum sensitivity of 368.37 ° / R I U, QF of 129.08 R I U − 1, SNR of 2.581 and DA ≈ 0.35 deg − 1, showing a marked improvement over conventional single-layer structure. The smaller FWHM and the higher QF indicate sharper resonance peaks, hence yielding an improved angular resolution while precisely detecting small refractive index changes in aqueous samples. In addition, the A l S b − B i F e O 3 − A l S b sandwich structure provides more stability and robustness under environmental conditions, which is important for long-term sensing. The comparing tests conducted with prisms of BK7, BAK1, BAF10, and SF10 against other configurations gave one confirmation that BK7 is the best prism for sensing in aqueous environmental conditions because of its medium-high refractive index of 1.5151 which guarantees good light coupling and intense evanescent field interaction with the medium of the substance to be analyzed (Sassi et al., 2025).

3.4 Performance comparative study

Table 9 presents the comparative analysis of recently reported SPR sensor structures along with the proposed design. The 2024 structure BK7 – Au – P t S e 2 – BlueP/ W S 2 achieved a sensitivity of 234 RIU⁻¹ with a high QF of 390 RIU⁻¹ (Basak et al., 2024), while the 2023 BK7 / Cu / Pt / BP design reported 309 RIU⁻¹ sensitivity and a QF of 69.39 RIU⁻¹ (Maheswari et al., 2023). Another 2023 configuration, Prism / Ag / B a T i O 3 / Ag / G / SM, showed 220 RIU⁻¹ sensitivity with a QF of 101.85 RIU⁻¹ (Taya et al., 2023), and the 2022 BK7 – Ag₁ – MXene – Ag₂ – ZnO – Graphene structure demonstrated 161 RIU⁻¹ sensitivity with a QF of 67.93 RIU⁻¹ (Karki et al., 2022). Unlike the above research studies, the current study presents a Prism ag A l S b − B i F e O 3 − A l S b Sensing Medium multilayer structure and a much higher sensitivity of 368.37 ° / R I U. The two layers of AlSb are used to increase the confining of the evanescent fields, whereas the B i F e O 3 layer has a high dielectric interaction with the analyte, which leads to a better plasmon-analyte coupling. Therefore, the proposed sensor performs better than the earlier reported sensors in terms of sensitivity, and it proves to be appropriate in high-precision and real-time aqueous environmental sensing. The competitive data indicates notable performance improvement of our suggested AlSb / BiFeO₃ / AlSb multilayer architecture compared to previously reported structures for an SPR sensor. The improved sensing performance largely comes from the synergistic effect of AlSb semiconductor and BiFeO₃ dielectric layers, which lead to better confinement of electromagnetic fields at metal – dielectric interface and stronger interaction between evanescent field and the analyte medium. Moreover, the systematic optimization of prism material and thickness of Ag layer results in sharper resonance dip with higher quality factor. These results corroborate the unique multilayer structure of the presented design and establish its feasibility in aqueous refractive index sensing applications.

TABLE 9. Performance comparison table.

3.5 Limitations, simulation setup and future directions

The current work is theoretical, and relies on numerical simulations using the Transfer Matrix Method (TMM), with no experimental validation at this time. The angular interrogation simulations were performed with operating wavelength fixed at 633 nm and prism materials of BK7, BAK1, BAF10 & SF10. While the Ag layer thickness was generally varied within 30–69 nm by a step of 3 nm, thus consistent with previous work, AlSb and BiFeO₃ layer thicknesses were preserved at their optimized states in each run. Sensing performance was evaluated for aqueous refractive indices in the range of 1.33–1.35, with sensitivity, quality factor (QF), signal-to-noise ratio (SNR), detection accuracy (DA) and combined sensitivity factor (CSF) being the main performance metrics. To minimize the possibility of Volmer–Weber Island growth and interfacial diffusion, careful optimization of substrate temperature, deposition rate, and interface engineering would be required during thin-film fabrication. Such process optimization can improve film continuity and interface quality, although experimental optimization remains essential for realizing the theoretically optimized layer thicknesses. Despite the potentially impressive numerical performance of the proposed sensor, several practical issues remain. The numerical model employed in the present work does not account for any effects arising from fabrication tolerances, material imperfections, interface roughness, oxidation due to environmental exposure conditions acting over time or temperature sensitivities. In future work, we will focus on experimental fabrication and characterization of the proposed multilayer SPR sensor; analyzing tolerance, stability under practical operating conditions, and extension of the proposed platform to selective biochemical as well as environmental sensing applications.

4 Conclusion

Researchers put together a multilayer SPR sensor. It features a structure like Prism, A g, A l S b, B i F e O 3, A l S b, Sensing Medium. The goal was environmental sensing in water-based samples. Simulations using the TMM method at 633 nm wavelength showed improvements in sensitivity, quality factor, and detection accuracy. These gains beat out traditional single-layer and double-layer setups. For the specific BK, A g, A l S b, B i F e O 3, A l S b, SM arrangement, sensitivity hit around 368.37 ° / R I U at its peak. The quality factor came in at 129.08 R I U − 1. Detection accuracy sat near 0.35 deg − 1. All this points to a sensor with strong precision and dependability. Layers of A l S b and B i F e O 3 helped trap the field more effectively. They also lowered optical losses and strengthened plasmonic coupling. The BK7 prism handled light entry smoothly in water environments. That layered setup improves stability in everyday scenarios. It stands up way better to oxidation too. Things like that make the design fit well for continuous checks on water pollutants or contaminants. In the end, this multilayer SPR sensor feels like a reliable option. It runs without any labels and delivers on-the-spot optical readout. A configuration this way supports detailed evaluation and ongoing monitoring of water quality.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Xujin Yuan, Beijing Institute of Technology, China

Reviewed by: Biswajit Brahma, McKesson, United States

Mahamudul Hassan Fuad, Dhaka International University, Bangladesh