Open Access
Add to BibSonomy
Abstract
A glucose-responsive photonic crystal integrated optical sensor that can detect glucosuria and urinary tract infection was designed by comprehensive numerical analysis. We showed a comparative study of nanoparticle shapes and materials and discussed the performance parameters of the sensor, such as the signal-to-noise ratio, the sensitivity, and the detection limit for various pH levels. Normally distributed nanoparticles were embedded within a functionalized microgel matrix in this proposed structure, and the theoretical study was conducted using the finite element method. Gold was selected as a photonic crystal material due to its nontoxicity and better sensing performance, and triangular-shaped nanoplates were chosen due to their better surface plasmon resonance performance. The sensor showed improved sensitivity of <85.65 nm/mM for pH 7.4, and <110.60 nm/mM for pH 8.0. It depicted a lower detection limit of >0.34 mM for pH 7.4 and >0.30 mM for pH 8.0 than those of previously reported glucose sensor studies. Moreover, the proposed glucose sensor exhibited a practicable signal-to-noise ratio and sensor resolution range. This label-free sensor manifested additional features such as reusability of the sensor, high sensing reversibility, and being readily usable at point-of-care medical facilities.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Biosensors play a crucial role in disease detection, environmental monitoring, drug discovery, and food control [1]. Point-of-care (POC) diagnostic testing is gaining popularity due to the capability of rapid diagnosis at any location. Refractive index of urine at various pH changes due to the different concentrations of excreted glucose. Hence, evaluating urine has diagnostic effectiveness in POC testing. A high level of urine glucose alludes to familial renal glycosuria because of the decrease of tubular reabsorption [2], whereas a low level of urine glucose refers to urinary tract infection (UTI) as a result of bacterial metabolism of glucose [3]. Diabetic patients with UTI can suffer fatal complications such as bacteremia, renal abscess, and renal papillary necrosis.
A typical method of glucose detection is the inaccurate test strip method because of its airborne and finger-borne contamination [4]. The fluorescence glucose sensor has higher sensitivity but is not commercially feasible due to its high price, and it necessitates carrying a light source for an implanted sensor [5]. Electrochemical sensors centered on glucose oxidase reactions have a low sensitivity [6,7]. To detect glucose, optical sensors are better than fluorescent molecules [5], traditional dyes [8], electroluminescent [9], and electrochemical [10] assays in the sense that they are label-free and can be involved in real-time monitoring [11]. Optical sensors can be made of photonic crystal (PC) with wide applications in optical devices and sensing materials [12]. The periodicity of the PC lattice and repetition of macroscopic dielectric media in the lattice control light propagation [13]. This photonic bandgap concept has been widely used in diverse photonic and optoelectronic applications [12]. Recently, hydrogel-based PC sensors were studied extensively [14]. Silver spherical nanoparticles (NPs) based sensor was reported where the effects of structural parameters were presented [15]. However, PC incorporated microgel-based glucose sensors are yet to be reported. Moreover, a research gap exists on the effect of NP shape and material in these PC sensors.
In PC-based devices, gel matrix acts as a protecting layer that monitors diffusion and enhances biocompatibility. Gel matrix has its application in the biomedical sector, especially in drug delivery and sensing the environment. Gel matrices such as microgel and hydrogel are biocompatible, flexible, and soft [16]. Microgel has similar polymer chemistry as hydrogel but has different physical molecular arrangements [17]. A proper solvent swells microgels that are crosslinked latex particles [18]. These materials have an immensely high surface area and lower viscosity and consequently, respond faster to the environment [17]. In contrast, hydrogels, which lack grafted side chains, take a long time to swell and shrink due to their larger size. To make this gel matrix sensitive to varying glucose concentration at physiological conditions, copolymerization of (3- acrylamidopropyl)trimethylammonium chloride (ATMA) and 3- acrylamidophenylboronic acid (3-APB) with PNIPAM can be used [17,19]. Silver nanospheres embedded in the hydrogel matrix [15] was reported as a PC glucose sensor with the disadvantage of slow response to glucose [17]. Though hydrogel-based PC sensors were studied extensively, there is a scope of significant performance improvement of PC glucose sensor using NP implanted microgel-based sensor.
Here, we proposed a PNIPAM microgel-based Au nanoplates integrated sensor for urine glucose detection. We performed a detailed two-dimensional simulation analysis of the effect of NP shapes and their constituent materials using the finite element method (FEM). Additionally, we calculated the wavelength shift due to the glucose concentration changes at different pH levels. We determined its performance parameters such as signal-to-noise ratio (SNR), sensitivity (S), detection limit (δn), figure of merit (FOM), and signal resolution (SR). Furthermore, we proposed an easy, rapid, and efficient fabrication technique for the proposed sensor. Our study elucidated the ability of the proposed urine glucose optical nanosensor to measure glucose in urine for the detection of glucosuria and UTI. This study provides an outlook for our nanosensor and exhibits its promise as a POC testing sensor in primary clinics.
2. Design of sensor and simulation methodology
Figure 1 shows our proposed optical biosensor consisting of Au nanoplates embedded in a PNIPAM microgel with the incident transverse electric polarized light. Though hydrogel-based PC sensors were reported previously [14,15,20], microgel was utilized in our design as hydrogel typically shows a slower swelling or shrinking rate. Functionalized microgel matrix were on the silicon substrate, and patterned gold nanoplates were embedded within the matrix. Depending on the glucose concentration of the surrounding medium, the sensor swells or shrinks. Inside the gel matrix, the gel matrix had 36 blocks, and the NPs were implanted into the blocks.
Fig. 1. (a) 3D representation of triangular-shaped gold (Au) nanoplates embedded within the functionalized PNIPAM matrix deposited on the silicon substrate at pristine condition. (b) The swelled state due to increase of either pH or glucose concentration.
FEM analysis was carried out using COMSOL Multiphysics a simulation software. Extensive analysis was performed on a two-dimensional model of our proposed biosensor. At first, patterns of NPs of particular shapes were generated by writing a script in MATLAB software. It is expected that during the fabrication process, NPs will be normally distributed. A detailed analysis of the effect of nanoparticle edge length was presented in Supplement 1. To imitate the real sensor system while having minimal computational requirements, NPs with an average size of 45.22 nm with a deviation of 3.38 nm were generated following normal distribution in each sensor block. Each block consists of two NPs. Each vertical stack contains six blocks in the proposed design structure, with six stacks in total. These blocks were half of the stack spacing in the x-direction. Additionally, We simulated a 3D model of gold nanospheres embedded in a microgel-based optical sensor for a fixed stack spacing of 216 nm. The glucose concentration of 1 mM for pH 7.4 was considered, and block thickness was varied from 340 nm to 600 nm (see Supplement 1). It is noteworthy that the dip of transmittance spectra remained the same irrespective of block thickness. It was found that lesser thickness resulted in higher reflectance, which implied better performance. Hence, in our modeling of the nanoparticles, 340 nm was chosen to be the thickness of each block which can be inferred as the thickness of nanoparticles.
In y-direction, the width of these blocks was 375 nm and they were separated from one another by 150 nm. In x-direction, the stack spacing or stack constant, l had a relation with the wavelength of diffraction peak or resonance, $\lambda _d$ given by,
Here, n is the refractive index of the microgel. In this Bragg diffraction grating mechanism, the alternating refractive index media consisted of stacks of triangular nanoplate embedded microgel and empty microgel. Resonance peak wavelength depends strongly on the size of NPs, their shapes, and the number of stacks. Therefore, the desired tuning can be achieved by varying these parameters. The MATLAB script-produced NP structures were exported to COMSOL Multiphysics to conduct characteristics and performance analysis. Noble metals such as gold, silver, and aluminum were used as the material of the NPs. The Lorentz-Drude model was followed to get the complex refractive index of these metals. A rectangular microgel matrix that enclosed the crystal array was built. A wavelength-dependent complex refractive index of the microgel previously reported by Brasse et al. was utilized in the simulation [21]. Time-averaged power outflow, to get the spectra of transmittance (T) and reflectance (R), was recorded using an average type boundary probe at the right boundary, as can be seen in Fig. 2. Scattering boundary conditions were used in all directions, and the light was incident on the left boundary (see Fig. 2). Ten samples of NP embedded microgel were randomly generated to test the reproducibility. The average wavelength of the diffraction peaks for ten samples was calculated. The finer non-uniform mesh was used, and the spectra were calculated for a wavelength range of 400 nm to 1000 nm with a 0.5 nm step size. Different morphologies of nanoparticles with cross-sectional shapes such as circular, triangular, and square were exhaustively studied. Experimental data of gel complex swelling and shrinking with glucose concentration was collected [14] and accordingly, a framework of stack spacing with glucose concentration and theoretical operating wavelength was created. Afterward, the wavelength shift was calculated for sensing glucose at different pH levels. During the fabrication process, the particles may be randomly oriented, which is beyond the scope of current work, and the results of current work will be beneficial to conducting a rigorous analysis of the effect of the tilted and rotated nanoparticles and their functionality in the optical glucose sensor.Fig. 2. 2D Schematics of gel matrix with nanoparticles of different cross-sectional geometries - (a) triangular, (b) square, and (c) circular. The edge length of the triangle in (a), the edge length of square in (b) and the diameter of the circle in (c) were considered to be 45.22 nm with a deviation of 3.38 nm. (d-f) The electric field distribution at 641 nm, 617 nm, and 615 nm diffraction peak wavelength for nanoparticles with triangular, square, and circular cross-section, respectively (see Visualization 1). Glucose concentration of 1.0 mM at pH 7.4 was considered.
3. Results and discussion
By varying the material of the nanoparticles, the shape of the nanoparticles, and initial stack spacing during the fabrication process, the sensor can be optimized. To analyze which shape and material show better prospects in designing a glucose optical nanosensor, a detailed study was shown in Subsections 3.1 and 3.2. To determine the response of the sensor to various glucose concentrations at different pH levels, further analysis was executed in Subsection 3.3. To calculate the ability of the sensor, we presented the performance parameters analysis in Subsection 3.4 and comparative analysis among the reported studies and our proposed work in Subsection 3.5. Through these rigorous analyses, we proposed a sensor that has a great potential for detecting urinary tract infection and glucosuria.
3.1 Effect of nanoparticle shape
NPs, such as nanospheres, triangular nanoplates, and nanorod, show plasmon resonance tunability which can be referred to as "shape tuning" [22]. The two-dimensional model of triangular, cuboidal, and cylindrical nanoparticles embedded in the gel matrix are illustrated in Figs. 2(a), (b), and (c), respectively. To understand the result of varying shapes of the NPs, the distribution of the electric field inside the structure was needed to study extensively to comprehend the behavior of the sensor [23]. The material of the NPs was considered to be gold. Figures 2(d), (e), and (f) show electric field wave propagation at the resonance frequency for triangular, cuboidal, and cylindrical nanoparticles incorporated microgel sensor, respectively. The stack spacing, l was kept fixed at 216 nm, which corresponds to glucose-free urine condition.
Figure 3(a) shows the spectra of normalized transmittance for different shapes of NPs. This study demonstrated that for 216 nm stack spacing, the average resonance wavelengths for triangular, cuboidal, and cylindrical nanoparticles were located at 641 nm, 617 nm, and 615 nm, respectively. Plasmonic characteristics of the noble metal NPs associated with different morphologies were widely inspected previously [24]. Transmittance spectra can be explained by Mie theory for spherical shapes and discrete dipole approximation (DDA) for nonspherical shapes [25]. When the evenness of the NPs is reduced, and the shape goes from circular to cubic to triangular, the highest degree of charge separation is obtainable with sharp corners [26]. Because of sharper tips or edges, significant change in the spectra is visible, and more surface plasmon resonance (SPR) shift occurs. For triangular nanoplates, the peak wavelength was much closer to the theoretical operating wavelength than cylindrical and cuboidal nanoparticles. Fig. 3(b) shows the average resonance peak wavelength for different stack spacing of triangular nanoplates integrated sensor. The empirical relation between the resonance peak wavelength and stack spacing was deduced as,
Fig. 3. (a) Normalized transmittance spectra for triangular, cuboidal, and cylindrical nanoparticles embedded in the PNIPAM gel matrix. (b) Diffraction peak of triangular nanoplates incorporated biosensor for different stack spacings. The average wavelength of the diffraction peaks for ten samples are presented here and deviations from the average values are illustrated with error bars.
Here, the adjusted R-square was 0.99 for this empirical relation deduction. Moreover, the relation between the glucose concentration and stack spacing was reported by Yetisen $et~al.$[14]. From the amount of swelling, the peak wavelength can easily be determined by the aforementioned equation, and consequently, the concentration of glucose in the urine can be determined. We found that the amount of reflectance was significantly less for triangular-shaped plates. In other words, the trough was shallower than the other two spectra. The volume of the cuboidal nanoparticles was the greatest, whereas the volume of the triangular nanoplate was the smallest. This signified that a cuboidal nanoparticle covered more space inside the microgel than a triangular nanoplate. Therefore, the concentration of the PCs in a stack was less when in a triangular structure. Less concentration dictated a decrease in effective refractive index (RI) contrast. Hence, the weakest reflectance was found in this morphology. Effective RI contrast can be calculated from the RI of NP and microgel and the volume fraction of the NP (see Supplement 1). Nevertheless, the more corner sharpness was, the more redshift of the reflectance band occurred. Thus, we got triangular nanoplates to be more sensitive to RI changes. As can be seen in Fig. 2(e), the electric field enhancement occurred at the regular interval due to the triangular nanoplate stacks. For cylindrical and cuboidal PC sensors, the electric fields were more localized and non-uniform, and the electric field enhancement was much less. Therefore, we selected triangular nanoplates embedded in the microgel matrix to achieve superlative performance.
3.2 Effect of nanoparticle constituting material
We studied electric field distribution in gold, silver, and aluminum NPs embedded microgel matrix-based biosensors at resonance wavelengths as can be seen in Figs. 4(a), (b), and (c), respectively. The electric field patterns were similar for gold and silver. The electric field enhancement occurred at the regular interval due to constructive interference. Interestingly, the electric field looked more localized and uniform for gold than silver. However, the electric field enhancement did not follow any specific pattern for aluminum (see Fig. 4(c)). Fig. 4(d) shows normalized transmittance spectra for gold, silver, and aluminum as constituent materials of nanoplates. For 216 nm stack spacing, the transmittance dips of the materials were located at 641, 636 nm, and 610 nm wavelength, respectively. The peak wavelength for gold nanoplates was much closer to the theoretically estimated wavelength than those of silver and aluminum nanoplates. Bulk material and nanomaterial have remarkable dissimilarities, and one of these dissimilarities is exhibiting the SPR phenomenon. Specifically, noble metal NPs draw much attention because of this phenomenon. Silver is widely used because of its good performance on SPR shift, and it has a higher reflectance quantity. However, silver spreads toxicity after disposal of the device that pollutes the environment, and silver also shows low sensing reversibility [27]. Moreover, silver is chemically less stable. Gold is not only chemically more stable, nontoxic, and biocompatible but also reliable and reversible. Though aluminum is less expensive, aluminum-based sensors have less sensing capability due to their refractive index. Based on the study, our NP constituent material of choice was gold. These nanoplates can be grown via a one-pot seedless growth process [20]. To fine-tune the resonant frequency, we varied the edge length, the number of stacks, and the number of NPs in a stack. Considering fabrication size limitations, we chose an optimal edge length of 45.22 nm with 3.38 nm deviation for nanoplates, and 36 blocks with two NPs in each block were considered for the study. Fig. 4(e) shows the average diffraction peak wavelength with error bars for different stack spacings of these noble metals.
Fig. 4. (a-c) The electric field distributions for gold, silver, and aluminum nanoparticles-based sensors, respectively (see Visualization 2). (d) Normalized transmittance spectra for gold, silver, and aluminum as a material of nanoplates. Average diffraction peaks were at 641, 636 nm, and 610 nm wavelength, respectively. (e) Diffraction peaks of gold, silver, and aluminum nanoparticles incorporated biosensors for different stack spacings, respectively. The average wavelength of the diffraction peaks for ten samples is presented here, and deviations from the average values are illustrated with error bars.
3.3 Glucose sensing
Wavelength shifting occurred with a change in glucose concentration. The increase in shift occurred because increasing glucose concentration made the microgel swell and consequent increase in stack spacing [14]. Conversely, microgel shrank with decreasing glucose concentration. Glucose concentrations for pH 7.4 and pH 8.0 were considered because the average normal urine pH level is $\sim$ 6.0, and higher pH is an indication of UTI. During UTI infection, bacteria metabolize glucose, and thus, the concentration of urine glucose reduces below 1.0 mM [3]. High-level glucose concentration is an indicator of glucosuria. A 0 mM glucose concentration can be considered glucose-free urine, And we considered 0.4 mM and 0.8 mM as low-level glucose concentration, and 1.0 mM, 4.0 mM, and 8.0 mM for high-level glucose concentration. Fig. 5 shows the normalized reflectance spectra at various concentrations of glucose in urine. By fitting with peaks of the reflectance spectra, we can evaluate symmetry and prominence of peaks and optimize the diffraction peak (for details see Supplement 1). The average diffraction peaks were found at 555.5, 576.2, 613.0, 641.1, 733.4, and 834.7 nm for 0, 0.4, 0.8, 1.0, 4.0, 8.0 mM of glucose, respectively (see Figs. 5(a), and (b)) at physiological pH condition (pH 7.4) which is high compared to normal urine pH level. The average diffraction peak positions of ten samples were 555.5, 590.5, 631.9, 666.1, 821.9, and 916.0 nm wavelength for 0, 0.4, 0.8, 1.0, 4.0, and 8.0 mM of glucose at pH 8.0, respectively as can be seen in Figs. 5(c), and (d).
Fig. 5. Normalized reflectance spectra at pH 7.4 for different levels of concentrations of urine glucose: (a) low level glucose, and (b) high level glucose. Normalized reflectance spectra at pH 8.0 for (c) low level glucose, and (d) high level glucose concentrations in urine.
Expanding stack spacing lowered the concentration of the crystals that were the gold nanoparticles. As the sensor was based on PCs, the increase of the resonance peak originated from the photonic bandgap. When the propagating waves and the reflecting waves interfere destructively, few frequencies get filtered out, which is called the photonic bandgap. Consequently, the peak frequency or wavelength could not pass through the crystal. This bandgap became narrow, i.e., the peak shifted towards a longer wavelength, with an increasing concentration of glucose. Hence, whenever the PC array swelled, redshift occurred, and full width at half maximum (FWHM) increased. Contrarily, FWHM decreased with the crystal array shrinkage due to decreasing glucose concentration. It is evident from Figs. 5(a), (b), (c), and (d) that typically, the reflectance peak position redshifted, and the intensity increased with the increase of glucose concentration level with an increment of pH. Wavelength shift owing to variation in concentration is shown in Fig. 6. The peak redshifted with the increase of pH level, as can be seen from the normalized reflectance spectra shown in the inset of Fig. 6(b). Based on our analysis, we formulated an empirical equation for the concentration dependence of resonance wavelength shift, δλres given by,
Here, $C_g$ is the concentration of glucose. The adjusted R-squares for the equations mentioned above were 0.99.Fig. 6. Wavelength shift for increasing concentration at (a) pH 7.4 and (b) pH 8.0. The wavelength shift for ten samples are presented here and deviations from the average values are illustrated with error bars. Inset shows Normalized reflectance spectra for 8.0 mM urine glucose concentration at different pH levels.
3.4 Performance analysis
Performance of a PC sensor can be evaluated using SNR, δn, S, FOM, and SR. The SNR represents the ratio of peak wavelength shift with peak bandwidth and is calculated by,
where, δλ$\frac{1}{2}$ is the FWHM of the resonance peak. The SNR increased irrespective of pH with the increase of glucose concentration.The sensitivity of the sensor, S can be calculated by,
The sensitivity of our proposed sensor were 51.87, 71.93, 85.65, 44.47, 34.90 nm/mM for 0.4, 0.8, 1.0, 4.0, 8.0 mM of glucose at pH 7.4. The average sensitivity of the sensor over the considered glucose concentration was found to be 57.76 nm/mM. At pH 8.0, the sensitivity of our sensor were 87.62, 95.56, 110.6, 66.60, 45.06 nm/mM sensitivity for 0.4, 0.8, 1.0, 4.0, 8.0 mM of glucose. The sensor showed better sensitivity at a higher pH level with an average sensitivity of 81.09 nm/mM at pH 8.0. At both pH levels, when glucose concentration increased at a lower level (0 to 1.0 mM), sensitivity gradually increased, whereas when the concentration increased at a higher level (1.0 to 8.0 mM), sensitivity gradually decreased. The δn, a performance parameter that exhibits the precision of the sensor, is given by, The lowest limit of detection was found to be 0.34 mM at pH 7.4 and 0.30 mM at pH 8.0.The figure of merit, FOM of a sensor is given by,
The figure of merit was the highest at 1.0 mM glucose concentration for both pH levels. The sensor resolution were calculated using the following equation, Table 1 shows the calculated values of these parameters at different glucose concentrations and different pH levels.
Table 1. Performance parameters for the sensor at various glucose concentrations
3.5 Comparative analysis
Table 2 shows a detailed comparative analysis of the performance among the previously reported studies and the proposed work. A 3D gold film electrode-based sensor showed an SNR of 3.00 and a low detection limit [32]. However, N$_{2}$ environment is required during this electrochemical measurement to avoid oxidation. Kim et al. presented a nicotinamide adenine dinucleotide-dependent (NAD) glucose dehydrogenase-based sensor, which had inferior performance compared to our structure [29]. Moreover, an electrode transfer mediator compound was required, which may not be readily available at POC. A terahertz metamaterial was proposed to distinguish glucose concentration [30]. However, the measurement of this sensor required an expensive and sophisticated terahertz time-domain spectroscopy system. Moreover, our proposed nanosensor exhibited improved sensitivity than previously reported PC, metamaterial, optic fiber-based sensors [28,30,31]. Our proposed design showed a better detection limit than the gold/1D photonic crystal sensor, NAD-glucose dehydrogenase-based sensor, and terahertz metamaterial sensor [28–30]. Additionally, our proposed sensor is reusable after cleaning the urine solution [15].
Table 2. Performance parameters comparison of our theoretical study with previous reports
4. Proposed fabrication technique
Microgel [P(NIPAM-co-3-APB-ATMA)] can be prepared via surfactant-free emulsion polymerization [17,19]. In this process, 3-APB, ATMA, NIPAM, and crosslinker (N,N’-methylenebisacryl-amide) were dissolved in the required amount of DI water, and potassium peroxydisulfate (KPS) was used as an initiator. Gold nanoplates can be grown using a rapid one-pot seedless growth process as reported by Lei Chen et al. [20]. Hexadecyltrimethylammonium chloride (CTAC) is taken as a surfactant material in this process. For shape-directing agent, iodide ions are used. The pH of the growth solution is governed by Sodium hydroxide. Au$^{3+}$ is reduced in the mixture of CTAC, potassium iodide, sodium tetrachloroaurate solution, ascorbic acid, and NaOH, and monodispersed triangular gold nanoplates are produced.
Microgel can be deposited on top of the substrate as can be seen in Fig. 7(a). The etching process can be done on a polymeric microgel via standard photolithography (see Figs. 7(b), (c), and (d)). Microgel can be patterned by dry or wet etching. Dry etching uses RIE plasma, mask, and UV stepper. It gives commendable anisotropic etching and is greatly controllable. However, This approach is costlier. On the other hand, wet etching is cost-effective and simplest but lacks the aforementioned qualities. In our structure, six trenches can be created by the etching process. The exposed resist portion can be chemically washed, as can be seen in Fig. 7(e). Trenches are separated by 150 nm from one another. These trenches act as a block filled with NP incorporated sol-gel (see Fig. 7(f)). After spin-coating and evaporation, NP mixed colloidal microgel can be turned into a sol-gel. Microgel can be deposited on top of this layer as shown in Fig. 7(g). In this way, one stack can be produced. Afterward, this whole process can be repeated five more times, and the stacks can be deposited on top of each other to create six stacks in total (see Fig. 7(h)). This kind of sensor can be reused by removing the urine solution through cleaning and evaporation [15].
Fig. 7. Proposed fabrication procedure for the optical nanosensor. (a) Microgel film was deposited on the substrate. (b) Positive photoresist and mask used for etching via a photolithography process. (c) Etched photoresist. (d) Etched microgel. (e) Photoresist removed. (f) Nanoparticle incorporated sol-gel deposited on the trenches using spin coating and evaporation methods. (g) Microgel deposited on top of this layer. (h) Stack deposited by repeating the same process mentioned in (a)-(g) and complete structure after successive depositions. The structural parameters shown here are for glucose-free urine.
5. Conclusion
In this work, we proposed an ingenious optical sensor for UTI and glucosuria detection based on triangular-shaped gold nanoplates embedded within 3-APB and ATMA functionalized PNIPAM microgel. This portable sensor is promising for POC testing due to its better performance, rapid detection, and reusability. Among the different nanoparticle shapes studied here, triangular nanoplates showed better prospects of being chosen as the PC of the sensor because of their corner sharpness. Gold was chosen as the potential material of the PC due to its nontoxicity, better chemical stability, and biocompatibility. The SNR increased regardless of different pH levels with increased urine glucose concentration. This sensor demonstrated the lowest detection limit to be 0.34 mM at pH 7.4 and 0.30 mM at pH 8.0. The glucose sensor exhibited improved sensitivity at a higher pH level. It posed better signal-to-noise ratio (<2.92 nm/mM for pH 7.4 and <2.61 nm/mM for pH 8.0), sensitivity (<85.65 for pH 7.4 and <110.60 for pH 8.0), sensor resolution (>29.80 nm for pH 7.4, 33.90 nm for pH 8.0) than those of reported glucose sensor studies. We also proposed a rapid, efficient, and easy fabrication technique for our proposed device. This sensor can be applied to detect urine glucose of different concentrations at various pH levels, and thus, it can detect glucosuria and UTI. The better performance, reusability, portability, and rapid detection process will guarantee this sensor to be beneficial in biochemical applications, specifically in POC medical service centers.
Acknowledgments
The authors acknowledge the technical support of the Department of Electrical and Electronic Engineering at Bangladesh University of Engineering and Technology, Dhaka, Bangladesh, for the completion of the work.
Disclosures
The authors declare no conflict of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. P. Estrela, N. Bhalla, P. Jolly, N. Formisano, and P. Estrela, “Introduction to biosensors,” Essays Biochem. 60(1), 1–8 (2016). [CrossRef]
2. E. Ferrannini, “Learning from glycosuria,” Diabetes 60(3), 695–696 (2011). [CrossRef]
3. A. Toivanen, B. Nielsen, H. Fritz, I. Lorenzen, O. Edhag, O. Selroos, P. Siltanen, W. Bleifeld, L. Köhler, and B. Scherstén, Assessment of Subnormal Urinary Glucose as an Indicator of Bacteriuria in Population Studies: An Investigation of 3, 911 Subjects Between the Ages of Four and Sixty-five Years, by Hans Fritz, Lennart Köhler, and Bengt Scherstén, no. nos. 496-505 in Acta medica Scandinavica: Supplementum (Almqvist & Wiksell, 1968).
4. P. Sharma and P. Sharan, “Design of photonic crystal-based biosensor for detection of glucose concentration in urine,” IEEE Sens. J. 15(2), 1035–1042 (2015). [CrossRef]
5. B. M. Cummins, J. T. Garza, and G. L. Coté, “Optimization of a concanavalin a-based glucose sensor using fluorescence anisotropy,” Anal. Chem. 85(11), 5397–5404 (2013). PMID: 23627407. [CrossRef]
6. A. Heller and B. Feldman, “Electrochemical glucose sensors and their applications in diabetes management,” Chem. Rev. 108(7), 2482–2505 (2008). PMID: 18465900. [CrossRef]
7. J. Wang, “Electrochemical glucose biosensors,” Chem. Rev. 108(2), 814–825 (2008). PMID: 18154363. [CrossRef]
8. S. B. Bankar, M. V. Bule, R. S. Singhal, and L. Ananthanarayan, “Glucose oxidase — an overview,” Biotechnol. Adv. 27(4), 489–501 (2009). [CrossRef]
9. K. Wu, H. Zhang, Y. Chen, Q. Luo, and K. Xu, “All-silicon microdisplay using efficient hot-carrier electroluminescence in standard 0.18 μm cmos technology,” IEEE Electron Device Lett. 42(4), 541–544 (2021). [CrossRef]
10. M. A. Invernale, B. C. Tang, R. L. York, L. Le, D. Y. Hou, and D. G. Anderson, “Microneedle electrodes toward an amperometric glucose-sensing smart patch,” Adv. Healthcare Mater. 3(3), 338–342 (2014). [CrossRef]
11. M.-S. Steiner, A. Duerkop, and O. S. Wolfbeis, “Optical methods for sensing glucose,” Chem. Soc. Rev. 40(9), 4805–4839 (2011). [CrossRef]
12. E. Ozbay, I. Bulu, K. Aydin, H. Caglayan, and K. Guven, “Physics and applications of photonic crystals,” Photonics Nanostructures - Fundamentals Appl. 2(2), 87–95 (2004). [CrossRef]
13. V. E. Babicheva, N. Kinsey, G. V. Naik, M. Ferrera, A. V. Lavrinenko, V. M. Shalaev, and A. Boltasseva, “Towards cmos-compatible nanophotonics: Ultra-compact modulators using alternative plasmonic materials,” Opt. Express 21(22), 27326–27337 (2013). [CrossRef]
14. A. K. Yetisen, Y. Montelongo, F. da Cruz Vasconcellos, J. Martinez-Hurtado, S. Neupane, H. Butt, M. M. Qasim, J. Blyth, K. Burling, J. B. Carmody, M. Evans, T. D. Wilkinson, L. T. Kubota, M. J. Monteiro, and C. R. Lowe, “Reusable, robust, and accurate laser-generated photonic nanosensor,” Nano Lett. 14(6), 3587–3593 (2014). PMID: 24844116. [CrossRef]
15. C. P. Tsangarides, A. K. Yetisen, F. da Cruz Vasconcellos, Y. Montelongo, M. M. Qasim, T. D. Wilkinson, C. R. Lowe, and H. Butt, “Computational modelling and characterisation of nanoparticle-based tuneable photonic crystal sensors,” RSC Adv. 4(21), 10454–10461 (2014). [CrossRef]
16. R. Deka, S. Sarma, P. Patar, P. Gogoi, and J. K. Sarmah, “Highly stable silver nanoparticles containing guar gum modified dual network hydrogel for catalytic and biomedical applications,” Carbohydr. Polym. 248, 116786 (2020). [CrossRef]
17. M. Muratalin, “Stimuli-responsive nanogels for environmental and pharmaceutical application,” Ph.D. thesis, Imperial College London (2012).
18. B. Saunders and B. Vincent, “Microgel particles as model colloids: theory, properties and applications,” Adv. Colloid Interface Sci. 80(1), 1–25 (1999). Publisher: Elsevier. [CrossRef]
19. S. Kabilan, A. J. Marshall, F. K. Sartain, M.-C. Lee, A. Hussain, X. Yang, J. Blyth, N. Karangu, K. James, J. Zeng, D. Smith, A. Domschke, and C. R. Lowe, “Holographic glucose sensors,” Biosens. Bioelectron. 20(8), 1602–1610 (2005). Selected Papers from the Eighth World Congress on Biosensors, Part I. [CrossRef]
20. L. Chen, F. Ji, Y. Xu, L. He, Y. Mi, F. Bao, B. Sun, X. Zhang, and Q. Zhang, “High-yield seedless synthesis of triangular gold nanoplates through oxidative etching,” Nano Lett. 14(12), 7201–7206 (2014). PMID: 25412030. [CrossRef]
21. Y. Brasse, M. B. Müller, M. Karg, C. Kuttner, T. A. F. König, and A. Fery, “Magnetic and electric resonances in particle-to-film-coupled functional nanostructures,” ACS Appl. Mater. Interfaces 10(3), 3133–3141 (2018). PMID: 29256586. [CrossRef]
22. H. Malekzad, P. S. Zangabad, H. Mirshekari, M. Karimi, and M. R. Hamblin, “Noble metal nanoparticles in biosensors: recent studies and applications,” Nanotechnol. Rev. 6(3), 301–329 (2017). [CrossRef]
23. K. Xu, “Silicon electro-optic micro-modulator fabricated in standard CMOS technology as components for all silicon monolithic integrated optoelectronic systems,” J. Micromech. Microeng. 31(5), 054001 (2021). [CrossRef]
24. M. Kupresak, X. Zheng, G. A. E. Vandenbosch, and V. V. Moshchalkov, “Benchmarking of software tools for the characterization of nanoparticles,” Opt. Express 25(22), 26760–26780 (2017). [CrossRef]
25. B. T. Draine and P. J. Flatau, “Discrete-dipole approximation for scattering calculations,” J. Opt. Soc. Am. A 11(4), 1491–1499 (1994). [CrossRef]
26. C. Cobley, S. Skrabalak, D. Campbell, and Y. Xia, “Shape-controlled synthesis of silver nanoparticles for plasmonic and sensing applications,” Plasmonics 4(2), 171–179 (2009). [CrossRef]
27. A. Loiseau, V. Asila, G. Boitel-Aullen, M. Lam, M. Salmain, and S. Boujday, “Silver-based plasmonic nanoparticles for and their use in biosensing,” Biosensors 9(2), 78 (2019). [CrossRef]
28. M. Shaban, A. M. Ahmed, E. Abdel-Rahman, and H. Hamdy, “Tunability and sensing properties of plasmonic/1d photonic crystal,” Sci. Rep. 7(1), 41983 (2017). [CrossRef]
29. D.-M. Kim, M.-y. Kim, S. S. Reddy, J. Cho, C.-h. Cho, S. Jung, and Y.-B. Shim, “Electron-transfer mediator for a nad-glucose dehydrogenase-based glucose sensor,” Anal. Chem. 85(23), 11643–11649 (2013). PMID: 24199942. [CrossRef]
30. J. Yang, L. Qi, B. Li, L. Wu, D. Shi, J. Ahmed Uqaili, and X. Tao, “A terahertz metamaterial sensor used for distinguishing glucose concentration,” Results Phys. 26, 104332 (2021). [CrossRef]
31. S. Singh and B. D. Gupta, “Fabrication and characterization of a surface plasmon resonance based fiber optic sensor using gel entrapment technique for the detection of low glucose concentration,” Sens. Actuators, B 177, 589–595 (2013). [CrossRef]
32. Y. Bai, W. Yang, Y. Sun, and C. Sun, “Enzyme-free glucose sensor based on a three-dimensional gold film electrode,” Sens. Actuators, B 134(2), 471–476 (2008). [CrossRef]
