Novel Detection of Diesel Adulteration using Silver coated Surface Plasmon Resonance Sensor

: In this study, we propose a silver-coated (SPR-surface plasmon resonance) based biosensor for the detection of diesel adulteration for the ﬁrst time in the ﬁeld of bio-sensing. Here, the numerical analysis is done by COMSOL Multiphysics V-5.1 and fully simulation software MATLAB-V16 on 30190 mesh elements, 2792 boundary elements, and 206593 degrees of freedom using ﬁnite element method (FEM) at 25 0 C (298 K) for different concentration level. More-over, the entire experiment on SPR-PCF is done for the major optical parameters for instance-Birefringence ( B i ), Coupling-length ( L c ), Power-fraction ( P f ), Conﬁnement-loss ( α c ), Amplitude-sensitivity ( S a ), Wavelength-sensitivity ( S w ), Resolution ( R l ), Transmittance ( T x ), Transmittance-variance ( T v ), Relative-sensitivity ( S r ), Figure-of-merit ( F OM ) and Resonance ( R 2 ) etc which are respectively to their corresponding maximum performance proﬁles of 2 . 5 × 10 − 3 , 1300 µ m, 99.93%, 1200 –1600 dB/cm , -7950 RIU − 1 , 49941.169 nm/RIU , 6 . 319 × 10 − 4 RIU , -180 dB , 170 dB/RIU , 97.25%, 980, and 0.98312 at operating wavelength 0.5–2 µ m. Therefore, it is noticed that the proposed sensor can be able to establish itself as one of the toughest candidates in ﬁber sensing. Also, diesel adulteration sensing including the entire working procedure using a SPR platform will be regarded as a novel procedure for diesel sensing in the history of photonics.


The Geometry and Internal structure of SPR PCF
The cross-sectional view of the proposed diesel sensor is displayed in Fig. 1. Therefore, the corecladding boundary as well as PML is constructed with fused silica as a base material. Besides, it can be described with 3 layers likely-outer layer, inner layer and an intermediate layer. Moreover, the outer layer is described as a perfectly matched layer (PML) having the diameter of d 1 = 5.80 µm and the intermediate layer consists of analyte layer forms with the diameter of d 2 = 5.235 µm. Therefore, the rest of the portion contains a cladding layer with a diameter of d 3 = 4.60 µm as well as a plasmonic metal layer of silver (Ag) having a thickness of t g = 35 nm. Also, two rectangular air holes along with the same measurements of x 1 = 3.80 µm, and y 1 = 0.50 µm and pitch-constant P = 2.0 µm forms a cutting edge chord at the inner layer that helps to increase the area of analyte layer.
= Analyte = Gold = Silica = Air To sum up, the internal geometry contains a rectangular and circular structure that makes it implementable by using existing technology for instance-sol-gel, extrusion, and drilling methods that help to overcome all the issues (material selection, processing condition, alloy bonding technique, melting procedure, drilling, fabrication failure, and fiber dicing [1]- [23]) of real-life implementation.

Numerical analysis
The entire numerical analysis depends on polarization modes. Therefore, the proposed sensor depicts 2 polarization modes: (i) x-axis pole, and (ii) y-axis pole for spp modes, core modes, and transmission modes as well. Basically, the entire numerical analysis is done on user-defined mesh elements and starts with refractive indices analysis. The base material of the suggested sensor is fused silica and its refractive index is varied by wavelength λ and some constants. So, Sellmeier's empirical equation is described as [25], Here, the constants term used for fused silica are described as the Sellmeier's constants which are respectively for 0.69616, 0.407942, 0.897479, 0.0046914826 µm 2 , 0.0135120631 µm 2 and 97.9340025 µm 2 corresponding to B 1 , B 2 , B 3 , C 1 , C 2 ,and C 3 .
However, both the carrier-sender and the receiver-sender interactions totally depend on modal birefringence profiles of the proposed PCF model that is calculated by the contrast of indices of polarization modes [26,27], Here, n x and n y define the effective RI for the fundamental x and y-polarizations respectively. In fact, the minimum fiber length at which the maximum amount of electromagnetic power of light interaction occurs through the fiber core is determined as the coupling-length and can be estimated by [27], Where, B i defines as the birefringence of the suggested model at the proposed operating wavelength.
However, the entire evanescent field's power-spectrum confines through the core of the structure. And, for calculating the mode confinement, the total output-power-spectrum (OPS) that passes through the core of silica-based fiber [28] can be examined by, Here, L = the assumed length of the SPR-PCF. Nonetheless, the total efficiency of the fiber is examined by the power-fraction and [28] can be expressed by, Here, output and input optical power is distinctly presented by P in and P out . P f is the ration of input power over the output power. However, the entire mode-confinement creates a huge power loss that may be varied for mediums to mediums and [27] can be evaluated by, Here, λ defines as the operating wavelength for the proposed structure. On the other hand, the total plasmonic incidents greatly rely on the transmittance profiles of the nominated PCF based SPR-PCF sensor. So, the transmittance performance [28] of the proposed sensor can be defined by, Where, P in defines as the maximum power fraction. Transmittance variance T v [28] can be calculated by, Here, ∆n b defines as the refractive-index (RI) difference between 2 transmission modes (i) xaxis polarization and (ii) y-axis polarization, and tx (b1,b2) defines as the maximum amplitude of transmittance (T x ) profile for b1 and b2 respectively for the assumed analytes. Here, the absolute difference between two fundamental polarizations can be express by ∆n b and nx (b1,b2) defines as the RI profiles at which the maximum amplitude of transmittance T x takes place respectively to b1 and b2 for the assumed analytes.
But, the overall problems of the amplitude-interrogation-method (AIM) can be minimized by amplitude sensitivity evaluation. The amplitude sensitivity of the proposed sensor can be evaluated by [27], Where, α c and ∆α c = the confinement losses profile and loss deviation for different diesel concentrations respectively, and ∆n defines as the refractive index (RI) contrast between the x and y polarization.
However, the wavelength-sensitivity can be determined through the steep peak of loss profile for an operating wavelength according to the variation of the refractive index (RI). And, the wavelength sensing [27,28] profile for different diesel concentration can be defined as, Here, ∆λp defines as the variety of wavelengths for different peaks and ∆n = the variation of refractive index (RI) of different diesel mixtures. On the other hand, any alteration the occurred in the RI of diesel can easily be noticeable by the amplitude-resolution variation of the proposed PCF. So, the entire resolution of the proposed PCF should be expressed by [27], Here, ∆n defines as the RI difference. And, ∆λ min , and ∆λ peak are respectively for the minimum wavelength contrast and peak wavelength-difference for a specific refractive index of the x and y-axis polarization. Moreover, the relative-sensitivity is totally dependent on the effective refractive index profile (RI). And, the relative-sensitivity of the suggested PCF based SPR sensor can be measured by varying the index profile of core in it can be expressed by [29] Rs(%) = n core n ef f × P out (12) Here, n core and n ef f define the effective refractive index of core and effective refractive index of operating wavelength, and P out denotes as the power-spectrum of the output power.
Nonetheless, the actual measurement of the efficiency of the sensor is depicted by the Figure-Of-Merit (FOM) and can be shown by [30], Here, m = the peak resonance slope per refractive index unit (RIU) and FWHMP defines as the half of maximum peak resonance full-width properties. These above-mentioned methods play a significant role to estimate sensor performances. Now based on these methodologies the results and discussion part are included below.

Results and Discussion
Fig. 3 exhibits the effective refractive index (RI) variation vs. the operating wavelength compared to the density variation of oil mixtures. In fact, it depicts a straight downward shifting curve ranges from 0.50 µm to 2.00 µm wavelength with a mild bend at the curve portion of the specific wavelength. For all density levels of diesel mixtures, the maximum effective refractive index (RI) varies between 1.42 and 1.47. To sum up, the x-polarization delineates the larger RI profiles compared to y-polarization and lower concentration levels (10%) of diesel maintains a better refractive index (RI) profiles than any other concentration levels. Fig. 4 exhibits the birefringence variation alteration respectively to the operating wavelength for different density levels of diesel mixtures. Moreover, birefringence is absolutely proportional to the shifting of wavelength. Particularly, the lower density level allows the minimal birefringence profile and vice versa. And, at the same optimum point of 1.50 µm wavelength, the achieved birefringence profile is nearly 1.39×10 −3 , 1.365×10 −3 , 1.275×10 −3 and 1.245×10 −3 respectively to 10%, 20%, 30%, 40% and 50% density level. It is evident that the whole birefringence profile holds slower upward when λ is shifting from 0.50 µm to 0.80 µm, sharp upward shifting between  Table 2 The maximum possible sensing performance and the adulteration measurements of the PCF sensor. 1.6 µm and 2 µm wavelength and rest displays moderate upward shifting. To sum up, a higher concentration level maintains a higher birefringence profile and so on. Fig. 5 delineates the alteration of the coupling-length respectively to the controlling wavelength for the introduced micro-structure SPR-PCF. Also, the entire coupling length profiles of the proposed PCF sensor delineates a smooth downward trending curve-line with a slight bending manner. Besides, it varies for the maximum values of the range between 1300 µm and 1000 µm that smoothly moves downward with a concave bend to the minimum values of ranges from 1000 µm to 700 µm. Furthermore, it is examined that 10% adulteration of diesel displays the minimum coupling-length and that ranges between 950 µm and 480 µm. Besides, 50% shows the lowest coupling-length varies from 1300 µm to 700 µm. In fact, the coupling-length profiles of the proposed PCF depict a smoothly downward manner curve respectively to the linear manner of wavelength. Fig. 6 shows power-spectrum variations vs. wavelength variation which follows the rhythm of a sinusoidal curve. From this figure, it is noticed that the power-spectrum of 50% density level of adulteration is lagged most and 10% density level of adulteration is forward mostly by a few phase-angles than the rest of power spectrum. The power spectrum amplitude varies inside the range of (0, 1). It is noticed that the power of the 50% adulteration level provides a dense spectrum Table 3 The maximum possible major performance profile of the adulteration measurement sensor. than the others. Fig. 7 describes the power fraction as a function of operating wavelength with the range of wavelength from 0.5 µm to 2.0 µm. Besides, it displays a smooth upward bending curve in a concave manner. However, the power fraction can be varied for the minimum values ranges from 87.91% to 35% and the maximum values varies between 99.93% and 57.41%. Nonetheless, it is seen that the power fraction of 40% adulteration provides a better profile than others. Fig. 8 depicts the chromatic confinement loss vs. wavelength respectively to x-y polarization for different concentration levels of adulteration from 0.5 µm to 1.5 µm wavelength. In fact, the confinement losses are 1200 dB/cm, 1360 dB/cm, 1450 dB/cm, 1510 dB/cm and 1600 dB/cm respectively to 10-50% concentration level. From the figure, it is identified that 50% mixture level has a higher loss profile of 1600 dB/cm and the 10% mixture level has a minimum loss performance of 1200 dB/cm respectively to the following case that is described in Table. 2 and  Table. 3. However, all the loss peaks are established at the perpendicular cross-section of their corresponding core and spp mode. Fig. 9 delineates the amplitude sensitivity alteration as a function of wavelength respectively for different density levels of adulteration from 0.5 µm to 1.5 µm operating wavelength. However,  Table. 2 and Table. 3, the amplitude sensitivity of the SPR-PCF is 3550 RIU −1 , 4510 RIU −1 , 5902 RIU −1 , 7810 RIU −1 and 7980 RIU −1 respectively to 10%, 20%, 30%, 40% and 50%. It can be noticed that y-axis polarization has a greater amplitude profile compared to x-axis polarization. It is noticed that the 50% adulteration level has a huge sensing performance of 7980 RIU −1 . On the other hand, the 10% adulteration level has a lesser sensing performance of 3550 RIU −1 respectively for the following case. of the refractive index. Conversely, the maximum possible wavelength sensitivity varies from 29611.76nm/RIU to 49941 nm/RIU along with the resolution level that varies between 2.31×10 −4 RIU and 6.32 × 10 −4 RIU with respect to the RI variation of from 1.4580 to 1.4660. It is noticed that the total resolution profile of the proposed sensor displays a smoother convex upward curve with respect to the refractive index. Alternatively, wavelength sensitivity may vary with an upward concave trend for the RI variation. To sum up, the lower refractive index (RI) possesses to have lesser wavelength-sensitivity and the resolution for the proposed PCF described in Table. 2 Fig. 11 delineates the transmittance profile as a function of optical wavelength for the proposed sensor. Although, the transmittance profile may alter from 0 dB to -180 dB having a steep declining peak for each concentration level. Therefore, the optimum possible transmittance performances are -180 dB, -140 dB, -120 dB, -115 dB and -105 dB respectively to 10-50% adulteration level. It is commented that the maximal transmittance profile is -180 dB at the 10% mixture level that promotes more sensitivity and the lower transmittance is -105 dB at the 50% mixture level that promotes less sensitivity and so on.
However, transmittance-variance vs wavelength is displayed in Fig. 12. And, the entire transmittancevariance might range between 110 dB/RIU and 170 dB/RIU having a steep inclining peak for each estimate. Moreover, transmittance-variance of the suggested sensor is 170 dB/RIU, 135 dB/RIU, 165 dB/RIU , 165 dB/RIU and 110 dB/RIU respectively to 10%, 20%, 30%, 40% and 50% adulteration level. It is noticed that 10% has higher transmittance-variance of 170 dB/RIU and 50% has a lower transmittance-variance of 110 dB/RIU . In fact, a higher percentage of adulteration has a better average profile than others in terms of transmittance variance. Fig. 13 shows the relative sensitivity responses for alteration comparison with respect to operating wavelength for the PCF based SPR. It depicts that the maximum possible sensitivity of 97.25%, 97.19%, 97.15%, 97.08%, and 97.06% for respectively adulteration level from 10% to 50%. Therefore, it depicts an upward shifting of sensing and continues up to the end in a convex manner. As a result, the minimum adulteration level provides a more relative sensing profile than the maximum adulteration levels and so on.  to the maximum detection limit of 0.0020. So, it is noticed that higher adulteration level of fuels has more FOM than others from Table. 3. Here, the greater value of FOM confirms the better efficiency of the introduced sensor. Fig. 15 defines the regression model of adulteration vs. the variation of RI. Furthermore, the more fitting of a regression sample indicates the more better performance indication for a higher linearly response. And, the resonance of fitting line varies according to the real part of RI as optimum parameters. So, the fitting functions are described as-y = 7.05x + 0.0034, y = 7.06x + 0.0034, y = 7.04x + 0.0034, y = 7.07x + 0.0034 and y = 7.03x + 0.0034 with R 2 values of 0.9831, 0.9842, 0.9719, 0.9852 and 0.9702 respectively to the 10% to 50% adulteration level as an optimum parameter. So, the lower adulteration level has a better fitting function than the higher adulteration level. The better fitting function in low precision adulteration level provides evidence of more accuracy of a SPR sensor. As a result, this sensor can be applied in the areas of aqueous analyte sensing or bio-sensing purposes where a low precision is so important.

Conclusion
In this letter, we propose a surface plasmon resonance (SPR) based biosensor for the detection of diesel adulteration for the first time in the field of bio-sensing. Though fabrication of the proposed model remains undone, the numerical analysis is done by COMSOL Multiphysics V-5.1 and fully simulation software MATLAB-V16 on 30190 mesh elements, 2792 boundary elements, and 206593 degrees of freedom using finite element method (FEM) at 25 • C (298 K) for different concentration level. However, the fabrication of the PCF can be done by using existing technologies like-sol-gel technique etc. Moreover, the entire experiment on SPR-PCF is justified through more than 12 the major optical measurements with tremendous results −2.5 × 10 −3 , 1300 µm, 99.93%, 1200-1600 dB/cm, -7950 RIU −1 , 49941.17 nm/RIU , 6.32×10 −4 RIU , -180 dB, 170 dB/RIU , 97.25%, 980, and 0.98312 respectively to the modal birefringence, coupling-length, modal powerfraction, confinement-loss, amplitude-sensitivity, wavelength-sensitivity, resolution, transmittance profile, transmittance-variance, relative-sensitivity, FOM and total resonance etc. In summary, the proposed sensor might be able to establish itself as a tough competitor in the area of optical sensing.