The study of human health and happiness has always been a fascination among researchers. To date, many studies have been carried out to explore human physiology and pathology. In the modern era, the prevalence of diabetes mellitus, often known as “diabetes,” has significantly escalated among humans. From the previous estimation, this disease is projected to reach up to 450 million of the global population by 2030 (Oliver et al., 2009).

A person suffers from diabetes when they have trouble maintaining a healthy blood glucose level throughout a variety of prandial (before and after eating) conditions. Basically, the lack of insulin needed to properly handle glucose production in the body is what leads to diabetes. Usually, this happens because of the destruction of pancreatic beta cells, which are responsible for the production of insulin in the body (Teymourian et al., 2020). Diabetes may be broadly categorised into three distinct subtypes: type 1, type 2, and gestational. Type 1 diabetes occurs when the body’s immune system mistakenly targets and kills insulin-producing pancreatic cells. The afflicted individual will thereafter be unable to produce insulin by their own means. Whereas the majority of people suffer from type 2 diabetes, which arises due to the ineffective use of insulin production in the human body. Later in their pregnancies, some women develop a condition called gestational diabetes, which will usually be cured after the pregnancy in most cases (Hsieh et al., 2004). Both hyperglycaemia (excess glucose concentration) and hypoglycaemia (low glucose concentration) are medical conditions that must be treated as soon as possible to avoid fatal consequences. Common diabetes symptoms include frequent urination, extreme hunger and thirst, rapid weight loss, extreme fatigue, and blurred or distorted eyesight. Kidney disease, stroke, heart disease, nerve damage, and blindness are just some of the complications that can arise from untreated diabetes over time (Heller and Feldman, 2013). As a result of these complications, people with diabetes now have a mortality rate that is 50% greater than that of people without diabetes.

Although there is no permanent cure for diabetes, it can be effectively managed with timely diagnosis of glucose levels and appropriate medication. Injections of insulin or oral glucose-maintaining medications are commonplace in the treatment of diabetes, along with regular checks on sugar levels and correct dosing of medications that may improve the efficacy of these therapies (Heller and Feldman, 2008). In view of this, some advanced and reliable monitoring systems are capable of making the diabetes patient’s life easier. For years, various biological indicators like blood, urine, interstitial fluids, sweat, breath, saliva, and tears have been used to correlate with blood glucose concentration levels to make glucose monitoring simple, easy, more efficient, and less painful.

In this paper, glucose level detection has been studied using a urine sample. Urine is the liquid waste product of human metabolism, consisting of several analytes such as uric acid, urea, and creatinine. It gives a basis for monitoring an overall physical health condition (Zanon et al., 2011). Usually, a urine sample does not show any glucose concentration levels for a person in normal condition. However, for a person with diabetes, glucose concentration levels have been detected and quantified in urine (Li et al., 2015).

Many researchers have reported various techniques in the field of urinary biosensors (Scullion et al., 2013). Recently, metamaterials have emerged as a new type of material that shows potential as a way to improve the performance of the biosensor (Padilla and Averitt, 2021), (Kumar et al., 2022), (Kumar et al., 2020) (Kumar et al., 2021). However, metamaterials with small feature sizes (less than 300 nm) are hard to design. Due to the constraints of current nanolithography techniques, engineered metamaterials do not fall under the meta-regime where the unit-cell dimensions are relatively smaller than the operating wavelength (Noginov and Podolskiy, 2016) (Kaur et al., 2019) (Sortino, 2011). Very Recently, advancements in 2-D photonic crystals have been made. This is based on the hexagonal structure and resonance cavity principle (Aidinis et al., 2020). Furthermore, enhancement has been carried out in the devices, and 2-D hexagonal photonic lattice crystals (Yuan et al., 2018) were also proposed. Glucose levels were detected in all of these devices by measuring the change in the refractive index of the urine sample and then correlating the respective resonance frequencies with the glucose concentration level. These photonic crystal-based techniques can solve many challenges, yet they have a few major downsides, such as the need for pre-sample treatment, biofouling of the delicate crystal structure, and a high fabrication cost.

To overcome these issues, a surface plasmon resonance biosensor has been used for biosensing applications. This highly sensitive phenomenon makes it possible to use it in real-time and label-free biosensing applications where minute changes in refractive index are needed to be measured. SPR biosensors possess various advantages, including ultra-sensitive detection (Homola et al., 1999). They are sensitive even when the concentration of the bio-sample is very low. Extensive efforts have been made to miniaturise SPR biosensors, which are motivated by the idea of a lab-on-chip and its significance in a wide range of applications like detection, biosensing, kinetic and binding experiments, and point-of-care diagnostics (Scarano et al., 2010).

SPR is an optical phenomenon that is caused by the interaction of the p-polarized monochromatic light with the free electrons on the metal-surface interface (Srivastava et al., 2011) (Sharan et al., 2022) (Karmakar et al., 2022) The sensing phenomenon is based on the change in the refractive index of the bio-sample placed at the top layer of the sensor, which results in an angular shift in resonance curves. Based on the light coupling through the prism, these SPR-based structures are of two types. One is the Kretschmann configuration (Kretschmann and Raether, 1968) in which the metal layer is directly in contact with the prism base, and the other is Otto configuration (Otto, 1968) which maintains an air gap between the metal layer and prism base. Moreover, this air gap is filled with any dielectric material having a low dielectric constant. Most of the devices are based on the Kretschmann configuration due to its ease of construction.

This work is based on the Kretschmann configuration for the simulation of the proposed biosensor. The angular modulation method has been employed for the calculation of the resonance angle (θ _spr ). At resonance, the electrons in the metal surface absorb the incident photons, due to which the intensity of reflected light drops. As the light travels across the layers, the reflected intensity decreases.

Due to the utmost excitation of the surface plasmons, reflectance intensity is reduced to a minimum (Hayashi and Okamoto, 2012) by conserving the momentum and energy of the incident light at resonance conditions, which satisfies the equation k i = k p (Akimov and Koh, 2010),Where k i = ω c n 1 ⁡ sin ⁡ θ i , represents the incident wave vector with θ i and n 1 as the incident angle and refractive index of the incident medium, respectively.

And, k p is the propagation constant of the plasmon mode, and this is represented as k p = ω c ε m e t a l ε d i e l e c t r i c ε m e t a l + ε d i e l e c t r i c (1)

Further solving the Fresnel equations allows the reflectance intensity as a function of incident angle to be obtained. Moreover, the dip of the reflectance curve at the resonance shifts upon replacing the bio-sample with different refractive indices (Otto, 1968). This shift in dip is directly related to the biosensor’s sensitivity.

For the design of SPR based biosensor, a thin film of plasmonic metals such as Gold (Au), Silver (Ag), Nickle (Ni), Platinum (Pt), Aluminium (Al), and Copper (Cu) (Moznuzzaman et al., 2020) is deposited on a dielectric substrate such as a prism made up of BK-7 glass, and further, it is covered with a sensing medium consisting of a bio-sample. Due to the poor adsorption property of the metals, they do not show many effects in the detection of bio-samples, and further, they are more susceptible to oxidation and corrosion. As a result, they must be coated with a chemically inert material in addition to meeting the requirement of high adsorption against the bio-sample in order to improve sensing performance parameters. Hence, the 2-D nanomaterials (Liu et al., 2015) such as graphene, Black Phosphorus (BP), and transition metal dichalcogenides (TMDCs) such as Molybdenum disulfide (MoS₂), Molybdenum diselenide (MoSe₂), Tungsten disulfide (WS₂) and Tungsten diselenide (WSe2) have gained great attention for the last 2 decades to enhance the sensitivity of the SPR biosensor. Because of its unique optical and electrical characteristics, it has recently been a research focus for biosensing applications. Additionally, they possess much better thermal and mechanical conductivity and a large value of elasticity, which makes them best suited for enhancing the SPR performance parameters (Singh et al., 2020).

In this study, two SPR structures have been proposed based on a typical Kretschmann configuration. In Structure-I, Ag has been used as a plasmonic metal, whereas Au has been used in Structure-II. Both of these metals have low complex refractive index values, indicating a higher capacity for absorption. Further, the effect of the silicon Layer has been simulated and analysed over both structures that simultaneously enhances the resonance effect and detection sensitivity as compared to conventional biosensors. (Green, 2008). The bio-sample consists of urine glucose samples of 0–15 mg/dL, .625 gm/dL, 1.25 gm/dL, 2.5 gm/dL, 5 gm/dL, and 10 gm/dL with the corresponding refractive indices of 1.335, 1.336, 1.337, 1.338, 1.341, and 1.347, respectively (Sani and Khosroabadi, 2020). Now, we have demonstrated the incorporation of 2-D nanomaterials over the Si layer in both Structure-I and Structure-II. After theoretical simulation and analysis of structure-I with a monolayer of WS₂, the sensitivity increases up to 1.3 times that of the sensor using Ag-Si layer and 1.7 times that of the conventional biosensor (BK7/Ag/Bio sample). Using a monolayer of BP sensitivity, on the other hand, improves sensitivity up to 1.3 times that of the Au-Si layer-based sensor and 2.2 times that of the conventional (BK7/Au/bio sample) biosensor in Structure-II. Additionally, Structure-II shows a sensitivity 1.4 times greater than that of Structure-I, which resulted in the attainment of a sensitivity of 273.4 °/RIU. We expect the proposed structure has the potential to detect glucose concentration levels with a quick response in terms of resonance angle shift in reflectance vs incident angle curves.

This paper is structured into the following parts: A brief introduction to diabetes and the SPR sensing phenomenon has been explained in Section 1. Section 2 describes the design methodology for the device structure. Section 3 explains the theory and theoretical modelling of the proposed biosensor. Section 4 illustrates the findings, followed by the analysis and discussion of those findings and a comparison of the results with existing work. The overall conclusion has been reached in Section 5, and references have been added later on.

Figure 1 shows the schematic diagram of the proposed multilayer biosensor in a typical Kretschmann configuration. The structure is comprised of five layers, which are in the following sequence: Prism-Ag/Au-Si-2D-Bio-sample. The thickness of all layers is in the nanometre (nm) range, and the operating wavelength has been considered as 633 nm to get the best performance parameters of the proposed biosensor. Here, the prism, which is fabricated from BK-7 glass, has been taken as the first layer. The prism’s refractive index was calculated using Eq. 2 (Johnson and Christy, 1972) for an operating wavelength of 633 nm. n = 1 + 1.03961212   λ 2 λ 2 − 0.00600069867 + 1.01046945 λ 2 λ 2 − 103.560653 + 0.231792344 λ 2 λ 2 − 0.0200179144 1 / 2 (2)

The base of the prism has been coated with a thin layer of Ag (a single layer of thickness d1 = 56 nm), which is later covered by a thin layer of Silicon (Si) with a thickness of d2 = 3 nm. The prism’s base has now been coated with a thin layer of Au (a single layer of thickness d1 = 50 nm), which is then covered by a thin layer of Si of thickness d2 = 3 nm. Further, A case study has been done by using various 2-D materials over the Ag (56 nm)-Si (3 nm) layer and the Au (50 nm)-Si (3 nm) layer for efficient glucose detection in a urine sample. This sample consists of a range of glucose concentrations of 0–15 mg/dL, .625 gm/dL, 1.25 gm/dL, 2.5 gm/dL, 5 gm/dL, and 10 gm/dL, and the corresponding refractive indices are 1.335, 1.336, 1.337, 1.338, 1.341, and 1.347. Here, the thickness of the bio-sample has been taken as 1,500 nm. However, we know the SPR phenomenon correlates only with a metal-dielectric interface and the operating wavelength of the monochromatic incident light. This way, analysis has been done by making two structures. Structure-I demonstrates the analysis of 2-D materials over the Ag (56 nm)-Si (3 nm) layer and forms the device with the following layers: Prism-Ag (56 nm)-Si (3 nm)-2D-Bio-sample, whereas Structure-II illustrated the analysis of 2-D materials over the Au (50 nm)-Si (3 nm) and created the device structure as Prism-Au (50 nm)-Si (3 nm)-2D-Bio-sample. Refractive indices of metal layers are complex in nature and have been calculated using the Drude–Lorentz model by using Eq. 3 (Sharma et al., 2015) n = 1 − λ 2 * λ c λ p 2   λ c + λ * i 1 / 2 (3) where λ, λ_{p, and} λ_c denote the operating wavelength of the monochromatic source of light, plasma wavelength, and collision wavelength respectively. Values of λ_{p and} λ_c for metals used are given in Table 1.

The refractive index of the silicon (Si) layer has been taken as 3.9160 (Malitson, 1965) with an optimised thickness of 3 nm. Table 2 provides the details of the refractive indices of 2-D materials that have been used in the simulation analysis. We can summarise our structure as a prism: the first layer is made up of BK7 glass to couple the entire incident light, the metal (Ag or Au) layer is the second layer for generating surface plasmons at the interface, Si is the third layer for enhancing the plasmonic effects at the metal surface, and the next 2-D material layer is for further improving the sensing performance of the biosensor.

This section represents the mathematical modelling of the proposed biosensor. This has been done by using the Transfer Matrix Method and Fresnel equations to analyse the reflectivity of the sensor (Yamamoto, 2015). Reflectivity (R) measurement of the reflected light at the output is the major requirement for any biosensor used for sensing purposes. To find out the reflectivity, we have to consider the reflection coefficient (r) for p-polarized monochromatic light. Here, for the five-layered device structure, the relation between R12345 and r12345 for the first interface can be expressed by Eq 4 (Barchiesi, 2013): R 12345 = r 12345 2 = r 12 + r 2345 e 2 i K 2 d 2 1 + r 12 r 2345 e 2 i K 2 d 2 2 (4)

Expression for the second interface can be represented by Eq. 5 (Yamamoto, 2015): r 2345 = r 23 + r 345 e 2 i k 3 d 3 1 + r 23 r 345 e 2 i K 3 d 3 (5)

Expression for the third interface can be expressed by Eq. 6 (Yamamoto, 2015): r 345 = r 34 + r 45 e 2 i k 4 d 4 1 + r 34 r 45 e 2 i K 4 d 4 a n d

Reflection coefficient expression for interface 1 and 2, 2 and 3, 3 and 4, and four and five can be written by using Eq. 6 (Yamamoto, 2015): r 12 = ε 2 / k 2 − ε 1 / k 1 ε 1 / k 1 + ε 1 / k 1 ,   r 23 = ε 3 / k 3 − ε 3 / k 3 ε 2 / k 2 + ε 2 / k 2 , r 34 = ε 4 / k 4 − ε 4 / k 4 ε 3 / k 3 + ε 3 / k 3 ,   r 45 = ε 5 / k 5 − ε 5 / k 5 ε 4 / k 4 + ε 4 / k 4 (6) Here r 12 , r 23 ,   r 34 , and   r 45 represents the amplitudes of reflected light from the interfaces 1–2, 2–3, three to four, and four to five, respectively, and   d n represent the thickness of each layer (where n = 2,3,4,5).

Eq. 7 (Barchiesi, 2013) represents the wave vector (k) component perpendicular to the layer interface k j = ε j ⍵ c + k z (7) where ε j is the dielectric constant of the corresponding medium (jth layer), ⍵ = 2πc/λ, is the angular velocity, with λ as the wavelength of incident light (Raether, 1988). Here, the incident wave vector component is parallel to the interface and can be written by Eq. 8 (Raether, 1988): k z = ε 1 ⍵ c sin ⁡ θ (8) and the dielectric constant of jth layer can be noted as: ε j = n j 2 (9) where θ is the incident angle, c is the velocity of light, and n j is the refractive index of the jth layer.

In this study, the angular modulation approach has been used to find the sensitivity of the proposed biosensor. In this approach, the incident angle, with a fixed wavelength, is scanned continuously. For a different refractive index of the bio-sample, the shift of the SPR angle is thus measured. The shift of dip in SPR curve, directly reflects the mass change on the SPR sensing surface (Liedberg et al., 1983).

In this study, a finite element method (FEM) based on the “COMSOL Multiphysics” software has been employed for the detection of glucose concentration in a urine sample with changes in the refractive indices. The light has been incident on the prism BK-7, and the angular interrogation method has been performed by using the parametric sweep in the range from 65° to 89° with a .01° increment. Reflection intensity has been measured for each incident angle. Optimization of biosensors has been carried out in the following subsections through a selection of metals and 2-D materials and by optimising their layer thickness.

In this present study, a numerical analysis has been carried out to find out the effect of 2-D nanomaterial (Graphene/BP/MoS₂/MoSe₂/WS₂/WSe₂) layers on a conventional (BK7/Ag or Au/Bio sample) biosensor based on Si layer, aiming towards the design of a highly sensitive biosensor for the efficient detection of urine glucose concentration levels. For this analysis, the angular modulation method has been done for these two structures, viz., Structure-I (Prism/Ag/Si/Bio-sample) and Structure-II (Prism/Au/Si/Bio-sample). A 2-D nanomaterial layer with optimised thickness has been placed over the Si layer, and a case study has been done for both structures. It is shown that among the 2-day nanomaterials, the addition of a monolayer of WS₂ enhances the sensitivity up to 200 °/RIU for Structure-I because the large real part of the refractive index of WS₂ can absorb a huge amount of light energy. On the other hand, a monolayer of BP plays a vital role in Structure-II because of its high absorption and adsorption coefficients, which enhance the sensitivity up to 273.4 °/RIU. Additionally, the obtained DA and FOM for Structure-I are .5617 degree⁻¹ and 112.35/RIU, respectively. However, DA and FOM for Structure-II are .134 degree⁻¹ and 36.89/RIU, respectively. Thus, Structure-II (Bk7/Au (50 nm)/Si (3 nm)/BP (.53 nm)/Bio-sample) shows better sensing performance parameters as compared to Structure-I (BK7/Ag (56 nm)/Si (3 nm)/WS2 (.8 nm)/Bio-sample). Therefore, we expect this proposed work based on 2-D nanomaterials has the potential to be used in continuous glucose monitoring as it can detect the minute change of .001 in the refractive index of a urine sample.