Open Access
Add to BibSonomy
Abstract
A surface plasmon resonance (SPR) prism coupler is proposed for the high-resolution non-invasive (NI) measurement of the circular birefringence (CB) properties of turbid media. The feasibility of the proposed device is demonstrated by means of numerical simulations. It is shown that the SPR sensor enables the CB properties to be detected with a resolution of up to 8.9 × 10−7 RIU (refractive index units) for refractive indices in the range of 1.3∼1.4. Moreover, for tissue phantom solutions containing 2% lipofundin, the device has a detection limit of 3.72 mg/dL. This resolution performance satisfies the detection limit of 10 mg/dL stipulated by the U.S FDA for point-of-care glucose monitoring devices. Thus, the proposed SPR sensor has significant potential for NI glucose sensing in such applications as diabetes detection and management.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Surface plasmon resonance (SPR) relies on the resonance oscillation of conduction electrons at the interface between two dielectric materials, and has significant potential for biosensing applications [1–2]. SPR sensing exploits the change in refractive index which occurs at the sensing surface in the presence of the target analyte and is typically performed using either indirect techniques based on prism couplers or diffraction gratings [3–5], or direct techniques based on plasmonic gold/silver nanoparticles [6–7]. SPR-based sensors have high sensitivity, good resolution and excellent repeatability. Furthermore, they provide a strong anti-interference ability and offer the potential for label-free sensing and dynamic monitoring. Consequently, they are used for many different applications nowadays, including gas sensing [8], biochemistry [9–10], medicine [11], food processing [12], environmental monitoring [13], and biomedicine [14]. Many studies have confirmed the effectiveness of SPR as a tool for disease diagnosis, drug discovery, and foodborne pathogen detection [15]. Diao et al. [16] proposed a method for detecting HIV-related DNA based on entropy-driven strand displacement reaction (ESDR) amplification and an SPR prism coupler sensor. The experimental results showed that the proposed biosensor was capable of detecting DNA over a measurement range of 1 pM to 150 nM with a detection limit of 48 fM and a detection time of less than 60 min. Stojanovic et al. [17] presented an SPR imaging-based technique for analyzing the expression of different cancer cell lines in a 44-plex antibody array. The combined correlation of the positive markers across all of the cell lines was shown to be 0.76. Aslan et al. [18] used plasmonic nanogold sensing aggregates to perform the continuous detection of millimolar changes in the glucose concentration of solutions with physiological pH. The results revealed that the glucose sensing range was determined primarily by the molecular weight, the size of the gold colloids, and the concentration of the medium used to form the sensing aggregates.
Glucose monitoring has attracted significant attention for the diagnosis and treatment of diabetes. Besides SPR techniques, the literature contains many other proposals for glucose biosensing, including optical polarimetry [19] and surface enhanced Raman spectroscopy (SERS) [20]. However, SPR has a higher sensitivity than optical polarimetry and a simpler setup and procedure than SERS. Accordingly, many SPR-based measurement systems have been proposed for glucose monitoring. Hsieh et al. [21] developed an SPR sensor for performing the direct detection of glucose by monitoring the binding affinity of glucose to a glucose/galactose-binding protein (GGBP) covalently attached to the SPR surface. The sensor was shown to have a reliable glucose measurement range of 0∼25 μM. Wang et al. [22] proposed an SPR sensor for glucose detection based on the change in the longitudinal plasmon band (LPB) of Au nanorods. The experimental results showed that the sensor was capable of detecting the glucose concentration over a measurement range of 0∼500 mM. Li et al. [23] presented an SPR sensor based on a self-assembled optical glucose-sensitive membrane and an osmotic protection membrane for the detection of glucose over the concentration range of 0-80 mg/dL. The studies in [21–23] provide a valuable contribution toward the development of effective techniques for performing glucose sensing using SPR methods. However, the accuracy and resolution of the proposed methods are not yet sufficiently high for practical glucose sensing applications. Moreover, the methods do not support the non-invasive (NI) measurement of glucose. However, simple and NI techniques for performing glucose monitoring are essential in detecting diabetes and managing its treatment if required. Consequently, much work remains to be done in developing NI glucose monitoring techniques with an accuracy and resolution as high as those of conventional invasive glucose measurement devices.
In previous studies, the present group proposed an SPR prism coupler for detecting the circular dichroism (CD) with a sensitivity of 10−5∼10−6 RIU [24] and circular birefringence/circular dichroism (CB/CD) properties of complex turbid media with a resolution of 10−4∼10−5 RIU [25]. In addition, the present group proposed a differential Mueller matrix formalism for NI glucose detection with a resolution of 20 mg/dL [26]. However, the resolution of proposed technique in [26] do not meet the requirement of U.S FDA for blood glucose monitoring devices. Furthermore, the mathematical model of SPR for detecting only CB properties or glucose concentration has not been derived yet. In order to fill the gap, in the present study, the analytical model of SPR prism coupler-based sensor was derived for enhanced CB properties sensing and application to NI glucose monitoring. The feasibility of the proposed method is demonstrated numerically. The suitability of the SPR sensor for practical applications is then confirmed by performing the NI measurement of the optical rotation angle and degree of polarization (DEP) of glucose phantom solutions containing 2% lipofundin.
2. SPR prism coupler and polarimetry model for extracting CB properties
Figure 1 presents a schematic illustration of the proposed SPR prism coupler [24–25]. As shown in Fig. 1(a), the device consists of a half-ball glass lens, a Cr-Au isotropic thin-film layer (thickness d1 and refractive index n1), a Ta2O5 anisotropic layer (thickness d2 and refractive index n2) and CB sample. In the proposed device, total internal reflection (TIR) is generated at the half-ball lens while the incident angle greater than critical angle and formed the evanescent wave at the sensed surface. The metal and anisotropic layers enhanced the evanescent wave and induced SPR condition at the resonance angle resulting in absorption of light. The change in glucose concentration results in changing refractive index of sample thus will be sensed by SPR sensor. It is noted that, as compared to the conventional Krestschmann configuration [27], an introducing of addition Ta2O5 anisotropic layer over the metal thin film in the proposed configuration formed a guided-wave surface plasmon resonance (GWSPR) structure to enhance the sensitivity of the sensor [28–29]. The half-ball lens used in the present study was fabricated of B270 (Thorlabs ACL1210U) with a refractive index of n0=1.52 at wavelength of 632.8 nm. The refractive indices and thickness of the isotropic were n1=0.36-2.9i, and d1=20 nm, respectively, at wavelength of 632.8 nm [24–25]. The thickness d1 is comparable to the mean free path of the electron in Au metals [30], the surface scattering and grain boundaries of conduct electrons will affect the resistivity or dielectric function of the metal [31–32]. However, the thickness of the isotropic layer was designed to achieve the highest resolution detection only, thus the effect of conduction electron scattering was ignored to simplify the mathematical model. In addition, the anisotropic layer Ta2O5 were deposited on B270 substrate by electron-beam evaporation using the oblique angle decomposition method. When performing the coating, the deposition angle was set at 70 deg. with column angle of 42 deg., the thickness of d2=13 nm, and the principal refractive index of n21=1.637; n22=1.449, n23=1.589 at wavelength of 632.8 nm [33]. It is noted that when the anisotropic material has a large refractive index and very thin thickness, it can support guided waves, enhance the electric field at the interface of the analyte and thus increase the sensitivity of the sensor [34]. As shown in Fig. 1(b), the resonance angle of the prism coupler was calculated with refractive indices ranging from 1.3 to 1.4 (step size of 0.2) and to be approximately 60° at a wavelength of 632.8 nm and resulted in a reflectance coefficient Rpp of less than 0.1. (Note that all of the physical parameters of the prism coupler were specifically designed with the aim of improving the measurement sensitivity of the CB properties and enhancing the suitability of the sensor for NI glucose detection.)
Fig. 1. Schematic illustration of (a) four layers structure and (b) simulated resonance angle of SPR prism coupler with refractive indices ranging from 1.3 to 1.4 (step size of 0.2). Note that the arrows in (b) show the direction of increasing refractive index.
In general, an optical sample can be described by the matrix formulation S = MS′, where S is the Stokes vector of the output light, M is the 4 × 4 Mueller matrix of the sample, and S′ is the Stokes vector of the input light. For the sensing configuration considered in the present study, the sample Mueller matrix is given by
where MCB, MR and MD are the Mueller matrices of the CB property of the sample, the reflectance of the prism coupler, and the scattering-induced depolarization effect, respectively. The Mueller matrix of reflectance MR is obtained by Berreman’s 4 × 4 matrix method for three-layer thin film structure and has the form [35]A series of simulations was performed using MATLAB to compare the values obtained from Eqs. (5) and (12) for parameters γ and Δ, respectively, of a hypothetical CB sample with the known values inserted into the sample matrix in Eq. (3). In performing the simulations, the refractive index of the sample was set as 1.33 and the incident angle, θi, was set equal to the SPR angle of the prism coupler (60°). Furthermore, the thickness and refractive index of the isotropic/anisotropic layers were assigned the values given above; while the values of p1, p2 and p3 in Eq. (3) were set randomly as 0.02, 0.13 and 0.06, respectively. Finally, e1, e2 and e3 in Eq. (3) were assigned randomly as 0.89, 0.64 and 0.42, respectively. It is noted that the values of p1-3 and e1-3 are able to assign randomly within the range of 0< p1-3, e1-3 <1. Figure 2 shows the extracted values of γ and Δ for each input value. As shown in Fig. 2(a), (a) good agreement exists between the two sets of values for γ in every case. Furthermore, the inset in the top-left hand corner of the figure shows that the error bar of the extracted value of γ for a given input value of γ = 15° deviates from the input value by just ±5×10−4 degree over the full range. (Note that this value assumes the use of a commercial Stokes polarimeter with an accuracy of ±0.25% in measuring the Stokes vectors of the output light (PAX1000VIS/R, Thorlabs Co.)). Figure 2(b) compares the extracted and input values of Δ over the considered range of Δ = 0 ∼ 1. A good agreement is again observed between the two sets of results. In addition, the inset in the top-left hand corner of the figure shows that the error bar of the extracted value of Δ for a given input value of Δ = 0.322 deviates from the input value by only ±2×10−4 over the full measurement range. In other words, the ability of the proposed model to extract the values of γ and Δ over the full range is confirmed. It is noted that the same results are obtained by using different value of refractive index of sample over the range 1.3 to 1.4.
Fig. 2. Comparison of extracted values and input values of: (a) γ and (b) Δ. Note that the inset in the top-left corner show the error bar of the extracted value of: (a) γ for a given input of 15° and (b) Δ for a given input of 0.322.
3. Sensitivity of CB measurements to glucose concentration
Further simulations were performed to investigate the sensitivity of the extracted optical rotation angle and depolarization index values to changes in the glucose concentration. In performing the simulations, the refractive index was increased from 1.3∼1.4 in increments of 0.01 to mimic glucose concentrations in the range of 0∼500 mg/dL. Figure 3 shows the results obtained for optical rotation angle and depolarization index using Eqs. (5) and (12), respectively, given polarization scanning angles of 0∼180°. It is seen in Fig. 3(a) that the extracted optical rotation angle, γ, is highly sensitive to changes in the refractive index (i.e., the glucose concentration) at polarization scanning angles of 30° and 150°, respectively. In particular, the value of γ increases with an increasing refractive index at a scanning angle of 150° (as indicated by the red arrow). The sensitivity and resolution of the sensor are determined as S=Δγ/Δn (°/RIU) and R=δγ/S (RIU), where Δγ and Δn are the variation of optical rotation angle γ and refractive index n, respectively; δγ is the maximum deviation of γ given an accuracy of ±0.25% in the output Stokes vector measurements obtained using commercial Stokes polarimeter. A close inspection of the inset in the bottom-right hand corner of the figure shows that the sensitivity and resolution of the extracted optical rotation angle are 560°/RIU and 8.9 × 10−7 RIU, respectively. It is noted that the same resolution is obtained at a scanning angle of 30°, however the values of optical rotation angles are negative thus inapplicable. This resolution value is once again estimated on the basis of an assumed output Stokes vector measurement accuracy of ±0.25%. Notably, this resolution is comparable with that of the SPR sensor proposed in [24] and two orders higher than that of the SPR sensor proposed in the previous study of the present group [25]. As shown in Fig. 3(b), the depolarization degree, Δ, is insensitive to the refractive index. This finding is reasonable since the active layer of the SPR sensor is limited to just several nanometers, and hence the depolarization effect is inevitably very small. In practice, the measurement sensitivity of depolarization index can be improved through an appropriate design of the prism coupler structure. However, in the present study, the prism coupler is designed only to maximize the measurement sensitivity of γ over the considered measurement range and hence such a redesign is unnecessary
Fig. 3. Variation of: (a) extracted optical rotation angle value and (b) extracted depolarization index value for polarization scanning angles in range of 0∼180° and simulated glucose samples with refractive indices ranging from 1.3 to 1.4 (step size of 0.01). Note that the arrows in (a) and (b) show the direction of increasing refractive index.
4. Experimental setup and results
The practical feasibility of the proposed SPR sensor for NI glucose concentration measurements was investigated using the experimental Stokes-Mueller matrix polarimetry system shown in Fig. 4(a). The system consisted mainly of a polarization state generator (PSG) and a commercial Stokes polarimeter (PAX1000VIS, Thorlab Co.). The PSG comprised a He-Ne laser (SL 02/2, SIOS Co., central wavelength 632.8 nm), a polarizer (GTH5M, Thorlab Co.) to produce linear polarized light, and two electro-optic (EO) modulators (ONESET Co., Model 350-50) with principal angles of 45° and 0°, respectively, to provide a full dynamic scanning angle range of θ=0∼180°. Figure 4(b) shows the input signals provided to the EOs of the PSG to generate the required scanning polarization angles. Note that all of the signals were produced by an arbitrary waveform generator (TGA 1244, Tektronix). The signals applied to the EOs were designed in such a way as to produce the four required states of polarization of the input light, namely 0°, 45°, 90°, and R-, respectively. For all of the experiments, the angle of the incident light was set equal to the prism coupler resonance angle of 60°. Moreover, the samples were stored in quartz cuvettes with dimensions of 10×10×1 mm. The coupler was attached to the cuvettes by means of industrial glue and a layer of silicon around the border. Prior to mounting the coupler, the cuvettes were drilled with a small hole with a diameter of 6 mm such that the sample made (thereby avoiding optical interference by the cuvette material).
Tissue phantom samples with glucose concentrations ranging from 0 ∼ 500 mg/dL (in 100 mg/dL increments) were prepared by mixing appropriate quantities of de-ionized (DI) water with 10 ml glucose solution (100 mg/ml-Merck Ltd) and 2% lipofundin (lipofundin MCT/LC1 20%, B|Braun). Additional samples were also prepared consisting of DI water, glucose solutions with concentrations ranging from 0∼100 mg/dL in 20 mg/dL increments and 2% lipofundin in order to determine the minimum detection limit of the proposed SPR prism coupler. (Note that previous studies have reported that 2% lipofundin is sufficient to accurately reproduce the scattering effect of human skin and tissue [36–37].)
Figures 5(a) and 5(b) show the experimental results obtained for the γ and Δ properties of the samples with glucose concentrations ranging from 0∼500 mg/dL. (Note that the scanning angle is 150° in every case.) The simulation results are also presented for comparison purpose. As shown, optical rotation angle increases linearly with a correlation coefficient of R2=0.9904 as the glucose concentration increases over the considered range. By contrast, depolarization index remains approximately constant. The average standard deviation of the measured values of γ obtained over four repeated tests was found to be 6.7×10−2 degree. From inspection, the sensitivity of the measured γ value is equal to approximately S=Δγ/Δn=0.018°/(mg/dL). The standard deviation value of the experimental results (6.7×10−2 degree) indicates that the proposed technique is able to detect glucose concentrations as low as R=6.7×10−2 deg./S=3.72 mg/dL. It is noted that, this resolution is higher than that of the differential Mueller matrix technique proposed in [24] with a detection resolution of 20 mg/dL. As a result, the feasibility of the proposed method for enhanced CB sensing is confirmed. It is noted that the variation of the two set results shown in Figs. 5(a) and 5(b) are most likely the results of conduction electron surface and grain boundaries scattering ignorance. Table 1 describes the FDA requirements for glucose monitoring test systems (GMTDs). As shown, the point-of-care devices required a minimum accuracy of ± 12% (or ±12 mg/dL), while the over-the-counter devices required a minimum accuracy of ±15% (or ±10 mg/dL). With a measured resolution of 3.72 mg/dL, the system proposed in the present study is able to meet the FDA requirements for both types of device.
Fig. 5. Experimental results for: (a) γ and (b) Δ of CB samples with glucose concentrations ranging from 0∼500 mg/dL. Note that the scanning angle is θ = 150° and the incident angle is 60° in every case.
Table 1. FDA requirements for GMTDs (2016).
Figure 6 shows the extraction of optical rotation angle from an oral glucose tolerance tests (OGTTs). It is noted that the OGGTs were performed on four healthy human volunteers aged between 20 and 25 years old, male Asian. The subjects were asked to fast for 8 hours and the body glucose concentration was then measured. The volunteers then drank 750 ml of aqueous solution containing 75 g sugar and the glucose measurement process was performed once again after an interval of 2 hours. It is noted that the body glucose concentration was measured by pressing the fingertip against the flat surface of the SPR prism coupler. As shown, the extracted values of the optical rotation angle increased following glucose ingestion for all of the volunteers. In other words, the results are consistent with the measurement results obtained for tissue phantom solutions shown in Fig. 5. The average deviation over 4 repeated tests of fasting and 2 h after the ingestion of rich sugared water are 0.39° and 0.36°, respectively. In general, the results confirm the practical feasibility of the proposed noninvasive SPR-polarimetry system for point-of-care or over-the-counter glucose monitoring applications. In the future, a method for extracting absolute value of glucose concentration from measured optical rotation angle will be developed. Further experiments on human fingertip tissue will be performed for diabetes and healthy volunteers by using conventional invasive and proposed NI method to confirm the practical application of the proposed technique for NI glucose sensing.
5. Conclusions
This study has presented a novel CB measurement technique based on an SPR prism coupler sensor. The validity of the proposed method has been demonstrated by numerical simulations. The results have shown that the proposed method enables the optical rotation angle to be measured with a resolution of 8.9 × 10−7 RIU for refractive indices in the range of 1.3∼1.4. Furthermore, the experimental results obtained for tissue phantom samples have shown that the measured optical rotation value is linearly related to the glucose concentration over the measurement range of 0 ∼ 500 mg/dL with a correlation coefficient of R2=0.9904. The average deviation of the measurement results over four repeated tests is equal to approximately ±6.57 × 10−2 degree, while the measurement resolution is as fine as 3.72 mg/dL. Consequently, the proposed SPR sensor satisfies the detection limit of 10 mg/dL stipulated by the U.S FDA for point-of-care glucose monitoring devices.
Funding
Ministry of Science and Technology, Taiwan (108-2218-E-239-002-MY2.).
Acknowledgments
The authors gratefully acknowledge the financial support provided to this study by the Ministry of Science and Technology of Taiwan (MOST) under Grant No. 108-2218-E-239-002-MY2. The research was also supported in part by the MOEMS lab of Prof. Yu-Lung Lo in the Mechanical Engineering Department of National Cheng Kung University, Taiwan.
Disclosures
No conflicts of interest, financial or otherwise, are declared by the authors.
References
1. H. H. Nguyen, J. Park, S. Kang, and M. Kim, “Surface plasmon resonance: a versatile technique for bisosenor applications,” Sensors 15, 10481–10510 (2015). [CrossRef]
2. J. Y. Jing, Q. Wang, W. M. Zhao, and B. T. Wang, “Long range surface plasmon resonance and its sensing applications: A review,” Opt. Laser. Eng. 112, 103–118 (2019). [CrossRef]
3. J. Homola, I. Koudela, and S. S. Yee, “Surface plasmon resonance sensors based on diffraction gratings and prism couplers: sensitivity comparison,” Sens. Actuators, B 54(1-2), 16–24 (1999). [CrossRef]
4. H. Arwin, M. Poksinski, and K. Johasen, “Total internal reflection ellipsometry: principles and applications,” Appl. Opt. 43(15), 3028–3036 (2004). [CrossRef]
5. A. Karabchevsky, S. karabchevsky, and I. Abdulhalim, “Fast surface plasmon resonance imaging sensor using Radon transform,” Sens. Actuators, B 155(1), 361–365 (2011). [CrossRef]
6. H. T. Ngo, H. N. Wang, A. M. Fales, and D. T. Vo, “Plasmonic SERS bisosensing nanochips for DNA detection,” Anal. Bioanal. Chem. 408(7), 1773–1781 (2016). [CrossRef]
7. D. Rodrigo, O. Limaj, D. Janner, D. Etezadi, F. Javier, V. Pruneri, and H. Altug, “Mid-infrared plasmonic biosensing with graphene,” Science 349(6244), 165–168 (2015). [CrossRef]
8. Q. H. Phan, P. M. Yang, and Y. L. Lo, “Surface plasmon resonance prism coupler for gas sensing based on Stokes polarimetry,” Sens. Actuators, B 216, 247–254 (2015). [CrossRef]
9. Q. Wu, D. G. Song, D. Zhang, and Y. Sun, “An enhanced SPR immunosensing platform for human IgG based on the use of silver nanocubes and carboxy functionalized graphene oxide,” Microchim. Acta 183(7), 2177–2184 (2016). [CrossRef]
10. A. Karabchevsky, L. Tsapovsky, R. S. Marks, and I. Abdulhalim, “Study of immobilization procedure on silver nanolayers and detection of estrone with diverged beam surface plasmon resonance SPR imaging,” Biosensors 3(1), 157–170 (2013). [CrossRef]
11. Y. B. Miao, H. X. Ren, G. Ning, Y. Zhou, Y. T. Cao, T. H. Li, and Y. J. Chen, “A triple-amplification SPR electrochemilumines- cence assay for chloramphenicol based on polymer enzyme-linked nanotracers and exonuclease-assisted target recycling,” Biosens. Bioelectron. 86, 477–483 (2016). [CrossRef]
12. D. T. Tran, K. Knez, K. P. Janssen, J. Pollet, D. Spasic, and J. Lammertyn, “Selection of aptamers against Ara h 1 pro- tein for FO-SPR biosensing of peanut allergens in food matrices,” Biosens. Bioelectron. 43, 245–251 (2013). [CrossRef]
13. G. D. Giustina, A. Sonato, E. Gazzola, G. Ruffato, S. Brusa, and F. Romanato, “SPR enhanced molecular imprinted sol–gel film: a promising tool for gas-phase TNT detection,” Mater. Lett. 162, 44–47 (2016). [CrossRef]
14. J. F. Masson, “Surface plasmon resonance clinical biosensors for medical diagnostics,” ACS Sens. 2(1), 16–30 (2017). [CrossRef]
15. H. Daraee, A. Eatemadi, E. Abbasi, S. F. Aval, M. Kouhi, and A. Akbarzadeh, “Application of gold nanoparticles in biomedical and drug delivery,” Artif. Cells, Nanomed., Biotechnol. 44(1), 410–422 (2016). [CrossRef]
16. W. Diao, M. Tang, S. Ding, S. Li, W. Cheng, F. Mo, X. Yan, H. Ma, and Y. Yan, “Highly sensitive surface plasmon resonance biosensor for detection of HIV related DNA based on dynamic and structural DNA nanodevices,” Biosens. Bioelectron. 100, 228–234 (2018). [CrossRef]
17. I. Stojanovic, C. Ruivo, T. Velden, R. Schasfoort, and L. Tersappen, “Multiplex label free characterization of cancer cell lines using surface plasmon resonance imaging,” Biosensors 9(2), 70 (2019). [CrossRef]
18. K. Aslan, J. R. Lakowicz, and C. D. Geddes, “Nanogold plasmon resonance-based glucose sensing,” Anal. Biochem. 330(1), 145–155 (2004). [CrossRef]
19. T. T. H. Pham and Y. L. Lo, “Extraction of effective parameters of turbid media utlilizing the Mueller matrix approach: study of glucose sensing,” J. Biomed. Opt. 17, 097002 (2012). [CrossRef]
20. J. M. Yuen, N. C. Shah, J. T. Walsh, M. R. Glucksberg, and R. P. V. Duyne, “Transcutaneous glucose sensing by surface enhanced spatially offset Raman spectroscopy in a rat model,” Anal. Chem. 82(20), 8382–8385 (2010). [CrossRef]
21. H. V. Hsieh, Z. A. Pfeiffer, T. J. Amiss, D. B. Sherman, and J. B. Pitner, “Direct detection of glucose by surface plasmon resonance with bacterial glucose/galactose binding protein,” Biosens. Bioelectron. 19(7), 653–660 (2004). [CrossRef]
22. X. Wang, L. Dong, J. Li, G. Shan, Y. Chen, and Y. Liu, “A glucose biosensor based on detecting longitudinal surface plasmon resonance of gold nanorods,” J. Nanosci. Nanotechnol. 16(7), 6925–6929 (2016). [CrossRef]
23. N. Li, D. Gong, Y. Yuan, and M. Yang, “Surface plasmon resonance sensing performance and adsorption law of self-assembly glucose sensitive membrane,” IEEE Sens. 2019 (early access).
24. Q. H. Phan, Y. L. Lo, and C. L. Huang, “Surface plasmon resonance prism coupler for enhanced circular dichroism sensing,” Opt. Express 24(12), 12812–12824 (2016). [CrossRef]
25. Q. H. Phan and Y. L. Lo, “Stokes-Mueller matrix polarimetry technique for circular dichroism/birefringence sensing with scattering effects,” J. Biomed. Opt. 22(4), 047002 (2017). [CrossRef]
26. Q. H. Phan and Y. L. Lo, “Differential Mueller matrix polarimetry technique for non-invasive measurement of glucose concentration on human fingertip,” Opt. Express 25(13), 15179–15187 (2017). [CrossRef]
27. E. Kretschmann and H. Raether, “Nadiative decay of nonradiative surface plasmon excited by light,” Z. Naturforsch., A: Phys. Sci. 23(12), 2135–2136 (1968). [CrossRef]
28. A. Shalabney and I. Abdulhalim, “Sensitivity enhancement methods for surface plasmon sensors,” Laser Photonics Rev. 5(4), 571–606 (2011). [CrossRef]
29. S. Hayashi, D. V. Nesterenko, and Z. Sekkat, “Fano resonance and palsmon-induced transparency in waveguid coupled surface plasmon resonance sensors,” Appl. Phys. Express 8(2), 022201 (2015). [CrossRef]
30. D. Gall, “Electron mean free path in elemental metals,” J. Appl. Phys. 119(8), 085101 (2016). [CrossRef]
31. T. Zhou and D. Gall, “Resistivity scaling due to electron surface scattering in thin metal layers,” Phys. Rev. B 97(16), 165406 (2018). [CrossRef]
32. K. Moors, B. Soree, and W. Magnus, “Resistivity scaling in metallic thin films and nanowires due to grain boundary and surface roughness scattering,” Microelectron. Eng. 167, 37–41 (2017). [CrossRef]
33. C. F. Lin and Y. J. Jen, “Use of Ta2O5 biaxial thin film as a high efficiency polarization coverter,” J. Nanophotonics 6(1), 061507 (2012). [CrossRef]
34. A. Lahav, M. Auslender, and I. Abdulhalim, “sensitivity enhancement of guided wave surface plasmon resonance sensors,” Opt. Lett. 33(21), 2539–2541 (2008). [CrossRef]
35. W. Berreman, “Optics in stratified and anisotropic media 4 ( 4 matrix formulation,” J. Opt. Soc. Am. 62(4), 502–510 (1972). [CrossRef]
36. T. L. Troy and S. N. Thennadil, “Optical properties of human skin in the near infrared wavelength range of 1000 to 2200 nm,” J. Biomed. Opt. 6(2), 167–176 (2001). [CrossRef]
37. G. Zaccanti, S. D. Bianco, and F. Martelli, “Measurement of optical properties of high-density media,” Appl. Opt. 42(19), 4023–4030 (2003). [CrossRef]
