The proposed biosensor, illustrated in Figure 1, leverages a multilayered architecture integrated into a SPR configuration based on the Kretschmann prism coupling method (Zheng et al., 2024). At the sensor’s core lies the BK-7 prism, which serves as the substrate and optical foundation. The prism plays a critical role in coupling incident light into the system, enabling the excitation of surface plasmons at the metal-dielectric interface. By directing polarized light at a specific angle (θ), the prism facilitates the resonance condition required for SPR, where a sharp dip in reflected light intensity occurs.

Above the prism is the Ag layer, which forms the metal-dielectric interface crucial for surface plasmon generation. Silver is chosen due to its exceptional plasmonic properties (Naik et al., 2013), including low propagation losses and strong plasmon resonance. This layer amplifies the sensor’s sensitivity to subtle refractive index changes induced by biomolecular interactions. To enhance the propagation of surface plasmons and improve the adhesion between the Ag layer and subsequent functional layers, a Si₃N₄ layer is incorporated as a dielectric spacer (Mudgal et al., 2020). The Si₃N₄ layer is not only a structural component but also a functional one, as its thickness directly influences key optical parameters, such as the FWHM of the resonance curve and the angle shift variation (Tene et al., 2024b). By systematically varying the thickness of the Si₃N₄ layer, the SPR angle is fine-tuned to maximize sensitivity and SNR.

Above the dielectric spacer lies the MoS₂ layer, a 2D material that meaningfully enhances the sensor’s functionality. MoS₂ amplifies the local refractive index changes upon biomolecular binding. Its high refractive index further enhances the SPR signal, while its surface chemistry allows for robust functionalization. In this design, the MoS₂ layer is functionalized with thiol groups, enabling the covalent attachment of ssDNA probes. These probes are designed to specifically hybridize with SARS-CoV-2 RNA, making the biosensor highly selective for viral detection.

The topmost layer, consisting of thiol-tethered ssDNA, serves as the biological recognition element of the sensor. The thiol groups provide a stable attachment to the MoS₂ layer, while the ssDNA sequences offer high specificity for complementary SARS-CoV-2 RNA targets. When the viral RNA binds to the ssDNA, it induces a localized change in the refractive index at the sensor surface. This change is translated into an SPR angle shift (Figure 1), which is detected by the optical system. The specificity of the thiol-tethered ssDNA ensures that the biosensor is not only sensitive but also selective, minimizing false-positive signals from non-target molecules.

A numerical analysis is employed to calculate the reflectance curve using the transfer matrix method (TMM) and Fresnel equations, as described in (Tene et al., 2024c; Varasteanu, 2020). This approach allows for the systematic modeling of light propagation through a multilayer system by considering the optical properties and thickness of each layer. The thickness of each layer is defined along the perpendicular axis (z-axis), which is the direction normal to the plane of the layers. The optical behavior at the interfaces between adjacent layers is governed by boundary conditions that ensure the continuity of the electric and magnetic field components of the electromagnetic wave.

The boundary conditions are applied at the interfaces to connect the electromagnetic fields across each layer. At the first layer, located at Z = Z 1 = 0 , the incident wave interacts with the multilayer structure, and its reflection and transmission coefficients are calculated. Similarly, at the final layer, the wave exits the structure at Z = Z n − 1 , where n represents the total number of layers in the system. These boundary conditions ensure that the physical constraints of the electromagnetic wave, such as the conservation of energy and phase continuity, are satisfied across all interfaces.

The Z-coordinate represents the position along the thickness of the multilayer structure, with Z 1 = 0 marking the start of the first layer (e.g., the BK7 prism) and Z n − 1 corresponding to the end of the last layer (e.g., the ssDNA functionalization layer). The TMM calculates the reflectance by constructing a matrix for each layer based on its refractive index, thickness, and incident angle of the light. These matrices are combined using the boundary conditions to derive the complete reflectance and transmittance of the system. This method enables accurate modeling of the interaction between light and the multilayer structure, providing insights into the sensor’s performance under varying configurations.

Then, the transfer matrix describes the relationship between the tangential components of the electric and magnetic fields as: E 1 H 1 = M E N − 1 H N − 1 where, E 1 , H 1 , E N − 1 , and H N − 1 represent the tangential components of electric and magnetic fields at the first and last layer interfaces, respectively. M is represented by elements M i j as: M = ∏ k = 2 N − 1 M k = M 11 M 12 M 21 M 22

And M k is defined as: M k = cos ⁡ β k − i ⁡ sin ⁡ β k / q k − i q k   sin ⁡ β k cos ⁡ β k here, k is an integer number. Additionally, β k is the phase thickness and q k is the refractive index in each layer: β k = 2 π d k λ 0 ε k − n 1 2 ⁡ sin 2 ⁡ θ

And q k = ε k − n 1 2 ⁡ sin 2 ⁡ θ ε k where, θ is the angle of incidence, λ 0 is the incident wavelength light, n 1 is the refractive index of the prism, d k is the thickness layer, and the local dielectric function ε λ 0 can be adopted as n λ 0 . Hence, the total reflection analysis of the N-layer system is obtained as: R = M 11 + M 12 q N q 1 − M 21 + M 22 q N M 11 + M 12 q N q 1 + M 21 + M 22 q N 2

by using Equation 6 the SPR curve as a function of the angle of incidence is computed. To analyze the performance of the biosensor is necessary to consider the following metrics. The sensitivity of the biosensors ( S ) is defined as the multiplication of the sensitivity to the refractive index change ( S R I ) and the adsorption efficiency of the target analyte ( E ) as: S = S R I ⋅ E

For biosensor optimization, we focus on the sensitivity enhancement ( ∆ S R I ) by optimizing each layer in water and PBS solutions, denoted as: ∆ S R I = S R I P B S − S R I w a t e r / S R I w a t e r

The sensitivity to the refractive index change can be expressed as: S R I = ∆ θ / ∆ n

The parameter ∆ θ represents the angle shift variation and ∆ n is the change in refractive index. The detection accuracy (DA) can be written in terms of ∆ θ and FWHM as: DA = ∆ θ / FWHM

Quality factor (QF) can be expressed in terms of S and FWHM as: QF = S / FWHM

In addition, to compute the FoM, LoD, and SNR, the related equations can be expressed as (Uddin et al., 2021; Hu et al., 2021): FoM = QF / Rmin LoD = ∆ n ∆ θ × 0.005 SNR = ∆ θ F W H M where, Rmin is the resonance minimum from SPR curve and 0.005 is expressed in degree ( 0.005 ° ). All computations in this study were performed using a data sampling density of 50,000 points. This high sampling rate was chosen to ensure statistical accuracy, minimize numerical errors, and provide a robust basis for the analysis.

Supplementary Table S1 illustrates the systematic progression of the systems considered in this work to optimize the SPR biosensor for SARS-CoV-2 detection. It provides a overview of the incremental changes made to the sensor’s architecture, ranging from the baseline configuration (Sys₀) to the fully functional systems incorporating advanced materials and specific biorecognition elements (Sys₉). The baseline system, Sys₀, consists of the simplest architecture: a prism, a silver layer, and water as the sensing medium. This configuration serves as a starting point for computing sensitivity enhancements across more complex systems. In Sys₁, the sensing medium is changed to phosphate-buffered saline (PBS), a biologically relevant environment, to bring the system closer to realistic conditions without altering the structural layers.

From Sys₂ onward, additional layers are incrementally introduced to evaluate their contribution to the biosensor’s functionality. Each subsequent configuration builds upon the previous one by incorporating either Si₃N₄, MoS₂, or thiol-tethered ssDNA. For instance, Sys₂ introduces Si₃N₄ to improve plasmonic performance, while Sys₃ functionalizes this layer with ssDNA to add selectivity for detecting viral RNA. Similarly, Sys₄ and Sys₅ explore the standalone and functionalized use of MoS₂, respectively. More advanced configurations, such as Sys₆, Sys₇, Sys₈, and Sys₉, involve combinations of Si₃N₄ and MoS₂ in various arrangements, with or without functionalization.

On the other hand, Supplementary Table S2 provides the initial parameters used for the SPR biosensor configurations investigated in this study. These parameters show the refractive index and thickness of each material layer, along with references to the experimental or theoretical studies from which these values were derived (Su et al., 2013; Luke et al., 2015; Kumar et al., 2022; Akib et al., 2024). The BK-7 prism, which forms the substrate of the SPR biosensor, has a refractive index of 1.5151. However, its thickness is not reported because, in the numerical modeling approach used here, the prism is treated as an infinite medium for light propagation. This assumption simplifies the simulation process while maintaining the accuracy of the results, as the thickness of the prism does not directly influence the excitation of surface plasmons at the metal-dielectric interface. Similarly, the thickness values for the sensing media, i.e., water and PBS, are not included. In these cases, the sensing medium is considered a semi-infinite layer in the numerical calculations, as its role is to provide the refractive index environment above the biosensor’s active layers. This treatment aligns with the typical approach in SPR modeling (Srivastava et al., 2020).

The Ag layer, with a refractive index of 0.056253 + 4.2760i, has a well-defined thickness of 55 nm. This thickness is optimized to achieve strong surface plasmon resonance and is consistent with values reported in the literature for SPR biosensors. The Si₃N₄ layer, with a refractive index of 2.0394 and a thickness of 5 nm, serves as a dielectric spacer. Its parameters are carefully chosen to enhance the propagation of surface plasmons and improve the adhesion of subsequent layers. The MoS₂ layer has a refractive index of 5.0805 + 1.1723i and a thickness of 0.65 nm, corresponding to a monolayer. These values are derived from theoretical studies that accurately model the optical properties of MoS₂ for plasmonic applications. Lastly, the thiol-tethered ssDNA layer has a refractive index of 1.462 and a thickness of 3.2 nm, which corresponds to a typical monolayer of DNA immobilized on a substrate. As stated, this layer provides the biosensor with its biochemical specificity for detecting SARS-CoV-2 RNA.

The results presented in Figure 2 and Supplementary Table S3 provide a detailed evaluation of the physical and performance metrics of the SPR biosensor configurations, systematically comparing their performance across different architectures. The SPR reflectance curves (Figure 2A) highlight the changes in resonance behavior caused by the incremental addition of functional layers, while the numerical metrics of attenuation, FWHM, and sensitivity enhancement (Supplementary Table S3) quantify the effects of these modifications. This comprehensive analysis is key to identifying the optimal configuration that balances plasmonic interaction, resonance resolution, and sensitivity.

The SPR curves (Figures 2A, B) reveal how each system responds to changes in the angle of incidence. Sys₀, the baseline configuration in water, demonstrates a sharp and narrow resonance with minimal attenuation (0.023%), reflecting its simple and loss-minimized design. However, as additional layers are incorporated in Sys₁ through Sys₉, the resonance depth increases, accompanied by broader curves and higher attenuation values. This trend is particularly pronounced in Sys₄ to Sys₈, where the integration of MoS₂ and ssDNA meaningfully enhances plasmonic interaction. While increased attenuation (peaking at 20.55% in Sys₈) indicates stronger interaction with the functional layers, it must be balanced to prevent excessive broadening of the resonance curve, which could compromise detection resolution.

The broadening of the resonance is quantitatively captured by the FWHM values (Figure 2C), which increase as the system evolves from Sys₀ to Sys₉. The baseline system exhibits the narrowest FWHM (0.88°), indicative of high-resolution performance but limited interaction with the sensing environment. As more layers are added, the FWHM values rise, reaching 2.97° in Sys₈, the widest among the configurations. While broader FWHM can reduce the precision of resonance peak detection, in this context, it reflects the enhanced plasmonic response due to the optimized multilayer structure. This trade-off is critical, as the broadening is accompanied by significant sensitivity gains, underlining the benefits of the multilayer architecture.

The sensitivity enhancement, shown in Figure 2D, provides a direct measure of the biosensor’s ability to detect refractive index changes. Relative to the baseline Sys₀, sensitivity enhancement increases consistently across the configurations, with the highest value observed in Sys₈ (7.50%). This result highlights the effectiveness of the combination of silicon nitride, molybdenum disulfide, and thiol-tethered DNA in optimizing the sensor’s optical and biochemical interactions. The role of each layer in enhancing the refractive index contrast and enabling specific biomolecular binding is evident in the important performance gains of Sys₈ compared to earlier systems. While Sys₉ achieves a similar sensitivity (7.29%), its marginally reduced enhancement compared to Sys₈ highlights the importance of layer arrangement and thickness optimization.

The results in Figure 3 and Supplementary Table S4 describe the impact of varying Ag thickness on the performance of the Sys₈ configuration. The SPR reflectance curves (Figures 3A, B) reveal a clear relationship between Ag thickness and resonance behavior. For thinner layers, such as 40 nm, the resonance dip is shallow, with an attenuation of 3.40%. This low attenuation indicates weaker plasmonic interaction and less effective light coupling. As the Ag thickness increases, the resonance dip becomes deeper, accompanied by a rise in attenuation, which reaches 53.13% for 65 nm. However, thicker layers also introduce substantial optical losses, which reduce the efficiency of the system. At 45 nm, attenuation is minimized at 0.17%, providing an effective balance between plasmonic coupling and energy conservation.

The FWHM values (Figure 3C) provide insight into the sharpness of the resonance peaks. At 40 nm, the FWHM is the widest, measuring 4.56°, indicating poor resolution and diffuse resonance behavior. Increasing the thickness to 45 nm narrows the FWHM to 3.79°, a significant improvement in resolution compared to thinner layers. Beyond 45 nm, the narrowing trend becomes less pronounced, stabilizing around 2.81° at 65 nm. The narrowing of FWHM with increased thickness reflects better optical resolution, though the limited improvements at higher thicknesses do not justify the trade-offs in energy loss and attenuation.

Sensitivity enhancement (Figure 3D) further emphasizes the performance improvements achieved with increasing Ag thickness. At 40 nm, sensitivity enhancement is measured at 0.58%, reflecting the reduced interaction between the plasmonic layer and the sensing medium. This value increases steadily with thickness, peaking at 0.86% for 65 nm. However, the incremental gains between 45 nm (0.67%) and 65 nm do not compensate for the higher attenuation and reduced resolution benefits. At 45 nm, sensitivity enhancement is sufficient for effective detection, while the low attenuation and improved resolution offer a more balanced performance. Hence, the selection of 45 nm as the optimal Ag thickness is supported by its ability to balance all critical performance metrics. The FWHM at 45 nm reflects improved resolution without the unnecessary broadening seen at lower thicknesses. Furthermore, its sensitivity enhancement, while not the highest, is competitive and provides reliable detection without the drawbacks associated with thicker layers.

The results in Figure 4 and Supplementary Table S5 examine the impact of varying the thickness of S₃N₄ on the performance of the Sys₈ configuration. The SPR reflectance curves (Figures 4A,B) demonstrate the effect of S₃N₄ thickness on the resonance behavior. At a thickness of 5 nm, the resonance dip is sharp but shallow, corresponding to an attenuation of 0.17%. This low attenuation indicates limited interaction between the plasmonic layer and the dielectric interface, reducing the system’s ability to effectively respond to refractive index changes. As the thickness increases, the resonance dip becomes more pronounced, peaking at 89.80% for 20 nm. However, such high attenuation introduces excessive optical losses, which can degrade signal response. At 13 nm, attenuation is moderate at 6.99%, providing a balanced interaction that enhances plasmonic coupling without significant energy dissipation.

The FWHM values (Figure 4C) offer further insight into the sharpness of the resonance peaks. At 5 nm, the FWHM is narrow, measuring 3.89°, reflecting high resolution but limited plasmonic response. As the thickness increases, the FWHM widens progressively, reaching 14.32° at 20 nm. This widening indicates reduced precision in detecting resonance shifts, which can impact the sensor’s ability to resolve small refractive index changes. A thickness of 13 nm achieves a FWHM of 6.76°, representing an optimal trade-off between maintaining sufficient resolution and accommodating the broader resonance needed for enhanced sensitivity.

Sensitivity enhancement (Figure 4D) highlights the sensor’s performance in detecting refractive index changes. At 5 nm, the enhancement is minimal (0.80%), reflecting limited sensitivity due to the reduced interaction at the interface. Sensitivity increases steadily with thickness, peaking at 17.64% for 15 nm before slightly declining to 16.42% at 20 nm. While thicker layers improve sensitivity, the diminishing returns beyond 13 nm, coupled with the rapid increase in attenuation and FWHM, make thicker configurations less practical for more realistic applications. At 13 nm, the sensitivity enhancement reaches 12.80%, providing a significant improvement without excessive trade-offs in signal response or resolution. The choice of 13 nm reflects a strategic compromise that maximizes the effectiveness of the Sys₈ configuration, supporting its application in precise and reliable biosensing for SARS-CoV-2 detection and beyond.

The results presented in Figure 5 and Supplementary Table S6 assess the effect of varying the number of MoS₂ layers on the performance of the Sys₈ configuration. The SPR reflectance curves (Figures 5A, B) show that for a single layer, the resonance dip is sharp and well-defined, with attenuation at 6.99%. This moderate attenuation reflects efficient plasmonic interaction without introducing significant optical losses. As the number of layers increases, attenuation rises sharply, reaching 89.66% for six layers. This increase in attenuation indicates excessive energy losses that diminish the signal response, reducing its suitability for practical biosensing applications.

The FWHM values (Figure 5C) further highlight the effect of additional layers. For a single layer, the FWHM is narrow at 6.76°, providing precise resolution and clear detection of resonance shifts. However, as the number of layers increases, the FWHM broadens significantly, reaching 16.64° for six layers. This loss of resolution impacts the sensor’s ability to accurately detect small refractive index changes. The narrow resonance peak observed with a single layer supports higher detection precision, making it the most effective option for applications requiring accuracy.

Sensitivity enhancement (Figure 5D) reflects how the number of MoS₂ layers influences the sensor’s ability to detect refractive index changes. In this context, two layers achieve the highest sensitivity enhancement (4.71%), but this comes at the cost of higher attenuation (45.50%) and a wider FWHM (9.61°). Beyond two layers, sensitivity enhancement declines steadily, reaching only 0.44% for six layers. Although two layers show a slight improvement in sensitivity compared to a single layer, the significant trade-offs in attenuation and resolution reduce their overall viability. The single-layer configuration provides sufficient sensitivity (1.19%) while maintaining superior resolution and lower attenuation. Then, these findings demonstrate that a single MoS₂ layer provides the best balance of performance metrics, making it the most effective choice for the Sys₈ configuration.

The results in Figure 6 and Supplementary Table S7 evaluate the effect of varying ssDNA thickness on the performance of the Sys₈ configuration. The SPR reflectance curves (Figures 6A, B) demonstrate the changes in resonance behavior as the ssDNA thickness increases. For a baseline thickness of 3.2 nm, the resonance dip is sharp and well-defined, with an attenuation of 6.99%. This reflects moderate plasmonic interaction and low optical losses. As the thickness increases, the resonance dip deepens significantly, with attenuation peaking at 94.27% for 50 nm. Such high attenuation corresponds to excessive energy losses, reducing signal response and compromising biosensing performance. At 5 nm, attenuation is still low at 8.56%, providing efficient plasmonic coupling without substantial optical losses.

The FWHM values (Figure 6C) reflect how the resonance sharpness evolves with increasing ssDNA thickness. At 3.2 nm, the FWHM is 6.76°, representing high resolution and precise detection of resonance shifts. However, as the thickness increases, the FWHM broadens, reaching 19.11° at 50 nm. This loss of resolution reduces the sensor’s ability to detect subtle refractive index changes. At 5 nm, the FWHM increases slightly to 6.92°, maintaining a strong balance between resolution and enhanced plasmonic interaction.

Sensitivity enhancement (Figure 6D) shows a steady increase with thicker ssDNA layers, peaking at 6.90% for 20 nm. At 5 nm, the sensitivity enhancement is 2.06%, a noticeable improvement over the baseline thickness of 3.2 nm. Beyond 5 nm, the higher sensitivity values are counterbalanced by the rapid rise in attenuation and broadening of the FWHM. This trade-off limits the practical applicability of thicker ssDNA layers despite their higher sensitivity. Indeed, at a thickness of 5 nm, the ssDNA layer provides a balance that minimizes energy losses, preserves adequate resolution, and ensures reliable sensitivity enhancement. This configuration avoids the excessive attenuation and reduced resolution associated with thicker layers while still offering improved biosensing capabilities.

To emphasize, Supplementary Table S8 presents the optimized parameters of Sys₈ following the detailed analysis of each layer’s thickness. These parameters, which include refractive index and thickness values, represent the final configuration of the SPR biosensor for effective detection of SARS-CoV-2 in PBS. The results also include the refractive index changes corresponding to different virus concentrations, based on experimental observations reported by Kumar et al. (2022), who observed a linear relationship between virus concentration and refractive index.

Hence, the optimized configuration of Sys₈ consists of a BK7 prism as the substrate, which has a refractive index of 1.5151. This serves as the base for the plasmonic structure. The Ag layer, with a thickness of 45 nm, was selected for its superior balance of attenuation, FWHM, and sensitivity enhancement. The S₃N₄ layer has an optimized thickness of 13 nm, contributing to enhanced plasmon propagation and improved resonance behavior. The MoS₂ layer has a thickness of 0.65 nm, corresponding to a monolayer. Its refractive index is 2.7611 + 1.6987i, which ensures efficient light-matter interaction, further enhancing the system’s sensitivity. The ssDNA layer, which could be functionalized for specific binding to SARS-CoV-2 RNA, has a thickness of 5 nm and a refractive index of 1.462.

The PBS medium, with a refractive index of 1.334, serves as the surrounding environment for detecting the target analyte. We point out that the refractive index values of SARS-CoV-2 in PBS range from 1.340 for a virus concentration of 150 mM to 1.355 for 525 mM. However, the concentrations reported in the experimental study (Kumar et al., 2022), ranging from 150 to 525 mM, are not representative of realistic physiological conditions, where virus concentrations are typically much lower. To address this limitation and enable the realistic testing of the proposed Sys₈ biosensor, we extended the linear relationship to extrapolate the refractive index of SARS-CoV-2 in PBS for concentrations ranging from 0.01 mM to 150 mM (Supplementary Table S9). This extrapolation allows us to simulate and evaluate the biosensor’s performance under more plausible scenarios, providing a deeper understanding of its sensitivity and detection capability in practical applications.

We now move to the focus of the present work, say, the performance of the optimized Sys₈ biosensor was evaluated across SARS-CoV-2 concentrations ranging from 0.01 to 150 mM in PBS, as shown in Figure 7 and Supplementary Table S9. At baseline (PBS@0.0 mM), the SPR reflectance curve shows a shallow resonance dip, reflecting the absence of analyte interaction (Figure 7A). As virus concentration increases, the resonance dip deepens progressively, with noticeable angular shifts at higher concentrations (50–150 mM). While the shifts at lower concentrations (0.01–0.1 mM) are minimal, the biosensor begins to demonstrate a measurable response starting at 1.0 mM, where sensitivity enhancement reaches 0.01% (discussed below). This indicates the biosensor’s capability to detect changes in refractive index at detectable levels while still being limited in resolving extremely dilute samples.

The attenuation values (Figure 7B) highlight the sensor’s performance in preserving signal response. At concentrations below 10 mM, attenuation remains relatively low, ranging between 8.56% and 8.82%, which supports minimal optical losses and efficient plasmonic coupling. However, for higher concentrations such as 50 mM and 150 mM, attenuation increases substantially, reaching 14.06% at 150 mM. This rise in attenuation reflects stronger plasmonic interaction with the analyte but introduces optical losses that may impact precision at elevated analyte concentrations.

The FWHM (Figure 7C) remains consistent across lower concentrations, ranging between 6.97° and 6.99° up to 10 mM. This stability highlights the sensor’s ability to maintain resolution for detecting subtle refractive index changes. At higher concentrations, such as 150 mM, the FWHM broadens slightly to 7.23°, indicating a gradual decline in resolution. While the broadening is not drastic, it points to the inherent trade-off between sensitivity and resolution as analyte concentrations increase.

The sensitivity enhancement (Figure 7D) further reinforces the biosensor’s behavior at varying concentrations. At the lowest concentrations (0.01 and 0.1 mM), sensitivity enhancement remains negligible. The biosensor begins to exhibit measurable sensitivity at 1.0 mM, with an enhancement of 0.01%. As the virus concentration increases, the sensitivity improves significantly, reaching 1.92% at 150 mM. This trend highlights the biosensor’s effectiveness in detecting moderate to high concentrations of the virus, although the minimal enhancement at very low concentrations indicates limitations in resolving trace amounts of the target analyte.

The Sys₈ biosensor offers several strengths, including low attenuation and consistent FWHM at lower concentrations, which ensure efficient plasmonic coupling and high resolution. The measurable sensitivity response beginning at 1.0 mM and its significant improvement at higher concentrations confirm its utility for detecting moderate to elevated virus levels. However, the limited sensitivity enhancement at concentrations below 1.0 mM and the increased attenuation at higher concentrations highlight areas for potential refinement.

Here, the performance evaluation of the optimized Sys₈ biosensor focuses on angular shift (Δθ), sensitivity to refractive index changes (S), detection accuracy (DA), and quality factor (QF). These metrics, presented in Figure 8 and Supplementary Table S10, provide a detailed understanding of how the biosensor responds to varying concentrations of SARS-CoV-2, ranging from 0.01 to 150 mM. The angular shift, shown in Figure 8A, measures the resonance angle change caused by variations in the refractive index due to the presence of the virus. At very low concentrations (0.01 and 0.1 mM), the Δθ is negligible, indicating limited interaction between the sensor and the analyte at these levels. However, starting at 1.0 mM, the biosensor begins to exhibit a measurable angular shift, with Δθ reaching 0.01°. As the concentration increases to 150 mM, the angular shift rises significantly to 1.57°. This increasing trend demonstrates the biosensor’s capability to effectively detect higher concentrations, although its limited response at lower concentrations highlights the challenge of detecting trace amounts of the virus.

The sensitivity, quantified in Figure 8B, reflects the angular shift per refractive index unit ( ° /RIU) and highlights the biosensor’s responsiveness to refractive index changes. For concentrations below 1.0 mM, sensitivity remains at 0.0°/RIU due to the negligible Δθ values. At 1.0 mM, sensitivity increases sharply to 200°/RIU, demonstrating the biosensor’s capacity to respond to detectable refractive index variations. As the concentration rises to 150 mM, sensitivity improves further, reaching 261.33°/RIU. These results emphasize the biosensor’s strong performance for moderate to high concentrations while also underlining its reduced effectiveness for detecting very low concentrations where refractive index variations are minimal.

The detection accuracy, as shown in Figure 8C, is a measure of the precision of the biosensor in detecting angular shifts. For virus concentrations below 1.0 mM, the DA remains negligible due to the absence of measurable angular shifts. At 1.0 mM, the DA improves to 0.001, increasing progressively to 0.217 at 150 mM. This trend confirms the biosensor’s reliability in detecting moderate to high analyte levels while indicating the need for enhancements to improve accuracy at very low concentrations.

The quality factor, displayed in Figure 8D, evaluates the efficiency of the biosensor by combining sensitivity and resolution. For concentrations below 1.0 mM, the QF is negligible, consistent with the zero angular shifts and sensitivity observed at these levels. At 1.0 mM, the QF reaches 28.68 RIU⁻¹, and it increases steadily with higher concentrations, peaking at 36.16 RIU⁻¹ for 150 mM. The consistent improvement in QF with rising concentrations demonstrates the biosensor’s robustness for detecting substantial analyte levels, where the combination of high sensitivity and reasonable resolution enhances its total effectiveness.

The results collectively show that the Sys₈ biosensor performs well for moderate to elevated virus concentrations, demonstrating measurable angular shifts, high sensitivity, improved detection accuracy, and strong quality factors. However, its limited response at very low concentrations (0.01–0.1 mM) highlights a challenge in detecting trace amounts of the analyte, which is critical for early diagnostic applications. Addressing this limitation would significantly enhance the sensor’s applicability for realistic scenarios where low analyte concentrations are common.

The performance of the optimized Sys₈ biosensor is further analyzed through the metrics of Figure of Merit (FoM), Limit of Detection (LoD), and Signal-to-Noise Ratio (SNR), as presented in Figure 9 and Supplementary Table S11. The FoM, shown in Figure 9A, captures the efficiency of the biosensor by combining sensitivity and resolution into a single metric. For the lowest concentrations (0.01 and 0.1 mM), the FoM remains at 0.0 RIU⁻¹, reflecting the negligible angular shifts and sensitivity observed at these levels. The FoM rises significantly at 1.0 mM, reaching 333.97 RIU⁻¹, marking a clear improvement in the biosensor’s ability to detect measurable refractive index changes. The performance peaks at 10 mM, where the FoM attains its maximum value of 405.50 RIU⁻¹, indicating the optimal balance between sensitivity and resolution. However, at higher concentrations, such as 150 mM, the FoM decreases to 257.23 RIU⁻¹, suggesting that optical losses and broader resonance peaks at elevated analyte levels begin to affect the sensor’s efficiency.

The LoD, reported in Figure 9B, highlights the biosensor’s ability to detect minimal analyte concentrations. At 1.0 mM, the LoD is 2.50 × 10⁻⁵, demonstrating the biosensor’s capability to detect low concentrations with reasonable precision. As the virus concentration increases, the LoD improves steadily, reaching 1.91 × 10⁻⁵ at 150 mM. This consistent trend reflects the sensor’s capacity to detect higher analyte levels with increasing accuracy. However, for concentrations below 1.0 mM, the LoD remains unresponsive, which accentuates a challenge in achieving reliable detection at trace analyte levels critical for early-stage diagnostics.

The SNR, presented in Figure 9C, evaluates the biosensor’s signal reliability relative to background noise. At very low concentrations (0.01 and 0.1 mM), the SNR is negligible (0.0) due to zero angular shifts and weak sensitivity responses. The SNR begins to rise at 1.0 mM, reaching 0.001, and increases progressively with higher concentrations. At 50 mM, the SNR improves to 0.072, and it peaks at 0.217 for 150 mM. This progression demonstrates the biosensor’s ability to produce a more distinguishable and reliable signal as analyte levels increase, but it also highlights its limited performance at very low concentrations where the signal remains indistinguishable from noise.

These results collectively indicate that the Sys₈ biosensor performs remarkably well at moderate to high SARS-CoV-2 concentrations. The high FoM at 10 mM highlights the biosensor’s efficiency in balancing sensitivity and resolution, while the steady improvements in LoD and SNR with increasing concentrations validate its reliability for detecting clinically relevant analyte levels.