Hydrogen peroxide and glucose concentration measurement using optical fiber grating sensors with corrodible plasmonic nanocoatings

We propose and demonstrate hydrogen peroxide (H2O2) and glucose concentration measurements using a plasmonic optical fiber sensor. The sensor utilizes a tilted fiber Bragg grating (TFBG) written in standard single mode communication fiber. The fiber is over coated with an nm-scale film of silver that supports surface plasmon resonances (SPRs). Such a tilted grating SPR structure provides a high density of narrow spectral resonances (Q-factor about 10) that overlap with the broader absorption band of the surface plasmon waves in the silver film, thereby providing an accurate tool to measure small shifts of the plasmon resonance frequencies. The H2O2 to be detected acts as an oxidant to etch the silver film, which has the effect of gradually decreasing the SPR attenuation. The etching rate of the silver film shows a clear relationship with the H2O2 concentration so that monitoring the progressively increasing attenuation of a selected surface plasmon resonance over a few minutes enables us to measure the H2O2 concentration with a limit of detection of 0.2 μM. Furthermore, the proposed method can be applied to the determination of glucose in human serum for a concentration range from 0 to 12 mM (within the physiological range of 3-8 mM) by monitoring the H2O2 produced by an enzymatic oxidation process. The sensor does not require accurate temperature control because of the inherent temperature insensitivity of TFBG devices referenced to the core mode resonance. A gold mirror coated on the fiber allows the sensor to work in reflection, which will facilitate the integration of the sensor with a hypodermic needle for in vitro measurements. The present study shows that Ag-coated TFBG-SPR can be applied as a promising type of sensing probe for optical detection of H2O2 and glucose detection in human serum. © 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement OCIS codes: (060.2370) Fiber optics sensors; (240.6680) Surface plasmons; (060.3735) Fiber Bragg gratings. References and links 1. W. Chen, S. Cai, Q. Q. Ren, W. Wen, and Y. D. Zhao, “Recent advances in electrochemical sensing for hydrogen peroxide: a review,” Analyst (Lond.) 137(1), 49–58 (2012). 2. C. Amatore, S. Arbault, D. Bruce, P. de Oliveira, L. M. Erard, and M. Vuillaume, “Characterization of the electrochemical oxidation of peroxynitrite: relevance to oxidative stress bursts measured at the single cell level,” Chemistry 7(19), 4171–4179 (2001). 3. E. W. Miller, A. E. Albers, A. Pralle, E. Y. Isacoff, and C. J. Chang, “Boronate-based fluorescent probes for imaging cellular hydrogen peroxide,” J. Am. Chem. Soc. 127(47), 16652–16659 (2005). 4. R. Raman and A. Shukla, “A direct borohydride/hydrogen peroxide fuel cell with reduced alkali crossover,” Fuel Cells (Weinh.) 7(3), 225–231 (2007). 5. S. R. Sarathy and M. Mohseni, “The impact of UV/H2O2 advanced oxidation on molecular size distribution of chromophoric natural organic matter,” Environ. Sci. Technol. 41(24), 8315–8320 (2007). Vol. 9, No. 4 | 1 Apr 2018 | BIOMEDICAL OPTICS EXPRESS 1735 Journal © 2018 Received 19 Jan 2018; revised 1 Mar 2018; accepted 8 Mar 2018; published 15 Mar 2018 #320127 https://doi.org/10.1364/BOE.9.001735 6. O. S. Wolfbeis, M. Schäferling, and A. Dürkop, “Reversible optical sensor membrane for hydrogen peroxide using an immobilized fluorescent probe, and its application to a glucose biosensor,” Mikrochim. Acta 143(4), 221–227 (2003). 7. S. Hanaoka, J. M. Lin, and M. Yamada, “Chemiluminescent flow sensor for H2O2 based on the decomposition of H2O2 catalyzed by cobalt (II)-ethanolamine complex immobilized on resin,” Anal. Chim. Acta 426(1), 57–64 (2001). 8. B. Li, Z. Zhang, and L. Zhao, “Chemiluminescent flow-through sensor for hydrogen peroxide based on sol–gel immobilized hemoglobin as catalyst,” Anal. Chim. Acta 445(2), 161–167 (2001). 9. E. Hurdis and H. Romeyn, Jr., “Accuracy of determination of hydrogen peroxide by cerate oxidimetry,” Anal. Chem. 26(2), 320–325 (1954). 10. C. Matsubara, N. Kawamoto, and K. Takamura, “Oxo [5, 10, 15, 20-tetra (4-pyridyl) porphyrinato] titanium (IV): an ultra-high sensitivity spectrophotometric reagent for hydrogen peroxide,” Analyst (Lond.) 117(11), 1781–1784 (1992). 11. Y. Song, L. Wang, C. Ren, G. Zhu, and Z. Li, “A novel hydrogen peroxide sensor based on horseradish peroxidase immobilized in DNA films on a gold electrode,” Sens. Actuators B Chem. 114(2), 1001–1006 (2006). 12. T. Endo, Y. Yanagida, and T. Hatsuzawa, “Quantitative determination of hydrogen peroxide using polymer coated Ag nanoparticles,” Measurement 41(9), 1045–1053 (2008). 13. E. Filippo, A. Serra, and D. Manno, “Poly (vinyl alcohol) capped silver nanoparticles as localized surface Plasmon resonance-based hydrogen peroxide sensor,” Sens. Actuators B Chem. 138(2), 625–630 (2009). 14. G. L. Wang, X. Y. Zhu, H. J. Jiao, Y. M. Dong, and Z. J. Li, “Ultrasensitive and dual functional colorimetric sensors for mercury (II) ions and hydrogen peroxide based on catalytic reduction property of silver nanoparticles,” Biosens. Bioelectron. 31(1), 337–342 (2012). 15. H. Li, X. Ma, J. Dong, and W. Qian, “Development of methodology based on the formation process of gold nanoshells for detecting hydrogen peroxide scavenging activity,” Anal. Chem. 81(21), 8916–8922 (2009). 16. X. Gao, L. Y. Jin, Q. Wu, Z. C. Chen, and X. F. Lin, “A nonenzymatic hydrogen peroxide sensor based on silver nanowires and chitosan film,” Electroanalysis 24(8), 1771–1777 (2012). 17. B. Yu, J. C. Feng, S. Liu, and T. Zhang, “Preparation of reduced graphene oxide decorated with high density Ag nanorods for non-enzymatic hydrogen peroxide detection,” RSC Advances 3(34), 14303–14307 (2013). 18. S. Chen, R. Yuan, Y. Chai, and F. Hu, “Electrochemical sensing of hydrogen peroxide using metal nanoparticles: a review,” Mikrochim. Acta 180(1–2), 15–32 (2013). 19. J. Ju and W. Chen, “In situ growth of surfactant-free gold nanoparticles on nitrogen-doped graphene quantum dots for electrochemical detection of hydrogen peroxide in biological environments,” Anal. Chem. 87(3), 1903– 1910 (2015). 20. Y. Li, J. B. Zheng, Q. L. Sheng, and B. N. Wang, “Synthesis of Ag@AgCl nanoboxes, and their application to electrochemical sensing of hydrogen peroxide at very low potential,” Mikrochim. Acta 182(1–2), 61–68 (2015). 21. H. Razmi, R. Mohammad-Rezaei, and H. Heidari, “Self-assembled prussian blue nanoparticles based electrochemical sensor for high sensitive determination of H2O2 in acidic media,” Electroanalysis 21(21), 2355– 2362 (2009). 22. C. K. Tagad, H. U. Kim, R. Aiyer, P. More, T. Kim, S. H. Moh, A. Kulkarni, and S. G. Sabharwal, “A sensitive hydrogen peroxide optical sensor based on polysaccharide stabilized silver nanoparticles,” RSC Advances 3(45), 22940–22943 (2013). 23. C. Zhang, L. Li, J. Ju, and W. Chen, “Electrochemical sensor based on graphene-supported tin oxide nanoclusters for nonenzymatic detection of hydrogen peroxide,” Electrochim. Acta 210, 181–189 (2016). 24. C. Caucheteur, T. Guo, and J. Albert, “Review of plasmonic fiber optic biochemical sensors: improving the limit of detection,” Anal. Bioanal. Chem. 407(14), 3883–3897 (2015). 25. C. K. Tagad, S. R. Dugasani, R. Aiyer, S. Park, A. Kulkarni, and S. Sabharwal, “Green synthesis of silver nanoparticles and their application for the development of optical fiber based hydrogen peroxide sensor,” Sens. Actuators B Chem. 183(5), 144–149 (2013). 26. P. Bhatia, P. Yadav, and B. D. Gupta, “Surface plasmon resonance based fiber optic hydrogen peroxide sensor using polymer embedded nanoparticles,” Sens. Actuators B Chem. 182, 330–335 (2013). 27. J. Albert, L. Y. Shao, and C. Caucheteur, “Tilted fiber Bragg grating sensors,” Laser Photonics Rev. 7(1), 83– 108 (2013). 28. T. Erdogan and J. E. Sipe, “Tilted fiber phase gratings,” J. Opt. Soc. Am. A 13(2), 296–313 (1996). 29. G. Laffont and P. Ferdinand, “Tilted short-period fiber-Bragg grating induced coupling to cladding modes for accurate refractometry,” Meas. Sci. Technol. 12(7), 765–770 (2001). 30. C. Caucheteur, C. Chen, V. Voisin, P. Berini, and J. Albert, “A thin metal sheath lifts the EH to HE degenercacy in the cladding mode refractometric sensitivity of optical fiber sensors,” Appl. Phys. Lett. 99(4), 041118 (2011). 31. T. Guo, F. Liu, Y. Liu, N. K. Chen, B. O. Guan, and J. Albert, “In-situ detection of density alteration in nonphysiological cells with polarimetric tilted fiber grating sensors,” Biosens. Bioelectron. 55, 452–458 (2014). 32. T. Guo, F. Liu, X. Liang, X. Qiu, Y. Huang, C. Xie, P. Xu, W. Mao, B. O. Guan, and J. Albert, “Highly sensitive detection of urinary protein variations using tilted fiber grating sensors with plasmonic nanocoatings,” Biosens. Bioelectron. 78, 221–228 (2016). 33. Y. Y. Shevchenko and J. Albert, “Plasmon resonances in gold-coated tilted fiber Bragg gratings,” Opt. Lett. 32(3), 211–213 (2007). Vol. 9, No. 4 | 1 Apr 2018 | BIOMEDICAL OPTICS EXPRESS 1736


Introduction
The detection of hydrogen peroxide, H 2 O 2 , at low concentration is an important analytical task due to its importance in pharmaceutical, clinical, environment, mining, textile and food manufacturing applications.For example, H 2 O 2 is a toxic by-product of many biological oxidases [1] in which H 2 O 2 has been considered to be an indicator of diseases such as Parkinson's disease, cancer, stroke, arteriosclerosis and Alzheimer's disease [2,3].In addition, H 2 O 2 has been used in liquid-based fuel cells and in organic synthesis because it is a powerful oxidizing agent [4,5].Therefore, it is important for both academic research and industrial purposes to be able to detect and accurately measure H 2 O 2 concentration.To date, various analytical techniques have been developed for this task, such as fluorescence [6] chemiluminescence [7,8], titration [9], spectrophotometry [10], and electrochemical techniques [11], Some surface plasmon resonance (SPR) devices based on different surface functionalized nanomaterials such as nanoparticles [12][13][14] nanoshells [15] nanowires [16] and nanorods [17] have been used to quantitatively determine H 2 O 2 concentration.In the case of SPR, the absorption strength decreases because of the metal oxidation by the H 2 O 2 .Most of above methods provide benefits of simplicity, speed, high sensitivity and reliability, and therefore are widely used for in vitro detection with a limit of detection (LOD) of about a micromole per liter (μM) in the laboratory [18][19][20][21][22][23].However, the sample preparation for existing measurement techniques is relatively time-consuming and the measurements are vulnerable to environmental influences such as temperature, dust, spatial light coupling fluctuations.
The optical fiber sensors have significant advantages, such as small size, costeffectiveness, multiplexing ability and remote operation ability, and have been widely studied in pharmaceuticals, healthcare, biomedical research, and environmental protection.Therefore, SPR based fiber optic sensors have gained considerable attention in bio-chemical fields due to their unique characteristics such as fast response, high sensitivity and label-free detection [24].For example, H 2 O 2 measurement based on etching of metal on a polymer fiber stripped of its cladding and bent into a U-shape [25] and polyvinyl alcohol (PVA) embedded silver nanoparticles over the silver coated core of the optical fiber [26] have been reported.However, such configurations considerably weaken the mechanical strength of the fiber and require careful individual fabrication, thereby limiting their use in practical applications.To overcome this problem, a new kind of fiber grating sensor, the tilted fiber Bragg grating (TFBG), has been developed and studied [27] based on original work in other fields [28,29].These are short period (~500 nm) gratings with light-induced refractive index modulation planes slightly inclined with respect to the cross-section of the optical fiber.There is no structural damage to the fiber, so the reproducibility and sensing stability can be guaranteed.The tilted grating also breaks the cylindrical symmetry of standard fibers, resulting in a strong polarization selectivity of the TFBG's excited cladding modes.When the fiber cladding over the TFBG is coated with a thin metal layer, radially polarized cladding modes can excite surface plasmon resonances (SPRs) at the interface between the metal and the surrounding dielectric.
Such TFBG-SPR structures enable measurements of changes in the refractive index (RI) in proximity to the surface of the fiber with resolution down to 10 −5 refractive index units (RIUs) [30][31][32][33].Finally, the most important feature of TFBG-SPR devices for sensing is the presence in the transmission spectrum of a resonance that is insensitive to the surrounding RI but sensitive to the temperature (with a very well calibrated sensitivity of ~10 pm/°C).This resonance can be used as an in situ thermometer to calibrate out this effect on the results, and as a power level reference in the same measurement that is used to sense the surface refractive index.
Most of the reported SPR based optical fiber sensors so far are based on a metal film with fixed thickness [34][35][36][37].However, the thickness of the metal film itself can be used as a changeable parameter for high resolution measurement, especially for surrounding refractive index (SRI) measurement [38][39][40] Instead of the traditional measurement technique, where the wavelength shift of the SPR attenuation is monitored, here we propose a much simpler sensing method of monitoring the rate of change of the attenuation induced by the SPR.We will show that the rate of change is directly related to the H 2 O 2 concentration.The operating principle is that H 2 O 2 will etch the silver coating of the TFBG, reducing its thickness, so that the SPR-induced attenuation of the fiber cladding resonances decreases and eventually disappears.In this way, the attenuation of the narrow cladding modes whose wavelengths are matched to the SPR attenuation band can be used to monitor the concentration of the H 2 O 2 .
Besides monitoring H 2 O 2 , we further propose a simple and in vitro method to measure glucose in human serum by the same approach.The continuous interest in sensing glucose, mainly in blood, is due to the increase of incidence of diabetes resulting from aging and unhealthy diets.Diabetic mellitus, which results from insulin deficiency and hyperglycemia can be detected when blood glucose concentrations deviate from the normal range of 80-120 mg/dL (4.4-6.6 mM) [41].The proposed detection scheme is based on the effect of enzymatic oxidation, whereby the glucose in human serum will react with added glucose oxidase (Gox) to produce H 2 O 2 .Since the etching rate of the silver coating is proportional to the concentration of the glucose, this can be detected by spectral monitoring of the SPR attenuation in reflection.Compared with the traditional electrochemical method for sensing glucose, our method is capable of remote operation ability and multiplexing [42].Moreover, in comparison with the other optical fiber methods, the response time of our proposed method is only 20 minutes and no enzyme treatment is required on the surface of the fiber [43][44][45].

Materials and equipment preparation
Serum samples were collected from the Cancer Center of the First Affiliated Hospital of Jinan University (Guangzhou, China).The phosphate buffer solution (PBS, PH 7.4), Hydrogen peroxide (H 2 O 2 , 30 wt. %) and glucose sample were bought from GZ Chemical Reagent (Guangzhou, China), and the glucose oxidase (GOx) was purchased from Sigma-Aldrich.Ultrapure water (Milli-Q grade, Millipore) with a resistivity of 18.2 MΩ/cm was used throughout this study.A scanning electron microscope and a centrifuge (Beckman, Germany) were used, as well as an Atomic force microscope (AFM, Bruker, USA).Silver coatings on fibers were produced with a magnetron sputtering coating machine (SKY TEACHNOLOGY DEVELOPMENT, China).

Sensor configuration and principle
The experimental setup for H 2 O 2 measurement is based on an Ag-coated 18°-tilt TFBG (diameter: 125 μm; length: 2 cm) as presented in Fig. 1(a).Here, the plasmonic TFBG sensor was powered by a broadband source (BBS) with wavelength range of 1460-1560 nm and its reflection spectrum was monitored by an optical spectrum analyzer (OSA) with wavelength resolution of 0.02 nm.A linear polarizer and a polarization controller (PC) were placed upstream of the circulator to adjust the orientation of input light polarization (P-polarization relative to the tilted planes of the grating, in Fig. 1(b)), so as to provide the strongest SPR excitation, i.e. selective excitation of cladding modes with almost 100% radially polarized light at the cladding surface.The measured results were recorded continuously every 1.2 seconds.Figure 1(b) shows the configuration of the Ag-coated TFBG probe.The tilt of the grating is an important parameter since it can be used to choose which set of cladding modes is to be excited [27].As a result, it is possible to adjust the operating range of the sensor in order to optimize the response for certain refractive indices.In this work, the tilt angle of TFBG was selected to be 18°, which provides the maximum amplitude for cladding modes in aqueous solution measurements with RI range of 1.32-1.34.
A 32 nanometer thick silver film was deposited on the fiber by radio frequency magnetron sputtering.To ensure adequate adhesion of the silver film, we deposited a 2 to 3 nm film of chromium on the fiber prior to silver deposition.Azimuthal uniformity of the silver film thickness over the fiber surface was achieved by rotating fiber around its axis during the sputtering process.A thin gold mirror was coated on fiber tip surface to enable interrogation of the sensor in reflection, allowing the sensor to be simply inserted into the solution under test.Finally, the gold mirror is protected with a deposited layer of 500 nm of SiO 2 on the fiber tip by radio frequency (RF) magnetron sputtering.This effectively eliminates external influences and keeps the reflection spectrum much more stable.

Theoretical analysis and demonstration
To further study the relationship between the SPR mode and the thickness of silver film coated over the fiber, we numerically calculate the SPR excitation through a finite-difference complex mode solver using the FIMMWAVE software.This software will model fibers of arbitrary (real) RI, using a rigorous solution to the vectorial wave equation in cylindrical coordinates.The solution assumes isotropic media and either an infinite cladding layer or with a perfect electrical conductor as the outer boundary condition [46].
A four-layered waveguide was constructed with the following parameters: fiber grating with title angle of 18 degree and grating pitch of 557 nm; fiber core with refractive index of 1.4498 and diameter of 8.2 μm; fiber cladding with refractive index of 1.4440 and diameter of 125 μm; fiber cladding surrounded by a silver sheath with a complex refractive index (n silver = 0.139-i10.963)[47] and thickness over the range from 0 to 30 nm; the silver sheath is surrounded by 80 μm water layer with refractive index of 1.3154 (the nominal RI of water was calculated at the wavelength of 1550 nm, not 589 nm Abbe refractometer).We numerically simulated radially polarized modes ("EH" in the accepted mode nomenclature) to get the mode intensity profiles shown in Fig. 2(a).As we can see, a cladding mode with strong SPR-induced field localization near the coating boundary can be found when the silver layer has a thickness of 30 nm.The SPR nature of this mode can be further confirmed by the fact that its propagation constant is found to have the largest imaginary part, corresponding to the maximum loss.Also, evident from the simulation is that the SPR excitation becomes weaker and weaker when the thickness of silver layer decreases.Such strong modulation of SPR efficiency numerically demonstrates the feasibility of our proposed sensor.Further simulations of actual TFBG-SPR transmission spectra (carried out with an in house numerical tool) are shown in part I of Fig. 2(b), for four different thicknesses of silver, EH modes, and water surroundings.The SPR-induced attenuation of the cladding mode resonance becomes weaker and spectrally broadened.Therefore, a selected cladding mode located near the center of the SPR attenuation band becomes stronger and deeper as the silver thickness decreases (Fig. 2bII).As expected however, low order cladding modes (the socalled "ghost" mode) and the core mode resonance remain unchanged (Fig. 2bIII).

Surface morphology of silver film
Figure 3 gives us the initial thickness of silver on TFBG which is not etched by H 2 O 2 solution.The film includes chromium and silver, and the total thickness is about 32 nm.Optimizing thickness of silver coatings on TFBG for H 2 O 2 application is demonstrated by the simulation.Figure 4 shows the surface morphologies of a 30-nm thick silver film (obtained by atomic force microscopy) with etching times of 0, 10 and 30 minutes in 2 mM H 2 O 2 in water.It is clear that the originally smooth Ag film gradually changes to a coarse surface with different sizes of nanoparticles.The longer the etch time the rougher and the thinner of the silver film is.As the silver film becomes thinner, the SPR attenuation profile (dashed lines) becomes weaker and spectrally broadened as predicted by simulations.

Detection of H 2 O 2
Figure 5 presents the spectral response of a 30 nm Ag-coated TFBG sensor when it was immersed in a 2000 μM H 2 O 2 solution.Figure 5(a) shows the full reflection spectrum of the sensor.There are two regions of interest: From 1450 to 1550 nm the Ag film etching induced SPR modulation and the corresponding cladding modes changes are evident (gray shadow area), and near 1559 nm, we observe the unchanged "ghost" mode (a group of strongly guided cladding modes which interact with cladding-metal interface but little with the outside medium) and core mode used for self-calibration (blue shadow area).Figure 5(b) presents the zoomed spectral response of the selected cladding mode measured each minute during a 30minute H 2 O 2 etch.The selected cladding mode becomes stronger and deeper as the SPR attenuation decreases due to the silver etching.Whilst the cladding modes in the SPR area change significantly with time due to silver film etching, the ghost and core modes remain unchanged, as shown in Fig. 5(c), because they are insensitive to the SRI.The ghost and core modes can be used to calibrate the effect of temperature and to eliminate power fluctuation from the measurement, thus improving the stability and accuracy of the sensor.The optical fiber sensor response for different concentrations of H 2 O 2 is shown in Fig. 6. Figure 6(a) depicts the SPR intensity change versus H 2 O 2 concentration during a one hour etch.Note that the selected cladding mode for monitoring is chosen as the most attenuated resonance on the original spectrum.We can see that the reaction speeds up as the concentration increases but that it saturates after a certain time.Therefore, for the detection of lower concentrations of H 2 O 2 , a much longer response time is required.Furthermore, we can see that when the concentration is as low as 0.2 μM, it becomes difficult to observe an amplitude change on the scale of the figure, which means that the concentration of H 2 O 2 is too low to etch the silver film significantly within a limited time (one hour here).To calibrate the sensor response in terms of amplitude recovery rate vs concentration, the resonance amplitude at the arbitrarily determined time of 20 minutes was chosen.The resulting resonance amplitude recovery (normalized to the initial resonance amplitude) as a function of concentration of H 2 O 2 is shown in Fig. 6(b).This clearly shows that the SPR attenuation of the resonance amplitude decreases nonlinearly but monotonically with the concentration of H 2 O 2 in the range from 0.2 μM to 5000 μM.In particular, it will be shown in Section 4.4 that for glucose concentrations in the physiological range from 3 to 8 mM, the concentration of H 2 O 2 produced by enzymatic oxidation falls between 1 and 20 μM, corresponding to amplitude recoveries between 1 and 5 dB (cross-hatched area in Fig. 6(b)).Of course, the proposed biosensor is designed as a disposable sensor (once the silver film is etched, the sensor no longer works).However, it is re-usable simply by depositing a new coating.In our lab, we use a mixed solution of 50 g of ceric ammonium nitrate (CAN) dissolved in 13 ml perchloric acid (PA) and 200ml deionized water solution to refresh the fiber surface (including removal the residual Cr) prior to re-deposition.

Glucose detection in human serum
Unlike the traditional plasmonic fiber-optic sensors discussed, here we can also take advantage of the weakening of the surface Plasmon resonance in the presence of H 2 O 2 for glucose measurements.When the D-glucose contained in human serum is mixed with a proper oxidant (GOx.),H 2 O 2 will be produced by enzymatic oxidation and etch the silver coating as described above.
The enzymatic oxidation between the D-glucose and GOx to generate H 2 O 2 can be described by Eqs. ( 1) and ( 2 ) Experiments were carried out with glucose at different concentrations from 0 mM to 12 mM in serum solutions (serum in PBS, diluted 10 times).In each case, 500 μL of 10 μM glucose oxidase was transferred into 4.5 mL of the mixed solution.The glucose and GOx concentrations are allowed to reach steady state in solution before the sensor is introduced.
For each glucose mix solution, the silver coated TFBG is immersed and its spectrum monitored for 20 minutes.The change in amplitude of the 1500 nm resonance over this period is plotted in Fig. 7, as a function of the actual glucose concentration measured separately with a commercial glucometer (SANNUO quasi).Three times repeated measurements were used to quantify the measurement performance with a relative standard deviation (RSD) of less than 15%.
These results can be used to calibrate the responsivity of our proposed plasmonic fiberoptic sensor, yielding a glucose concentration sensitivity of 0.5 dB/mM at a detection time of 20 minutes.

Conclusion
We have proposed and experimentally demonstrated an Ag-coated plasmonic TFBG sensor for H 2 O 2 concentration measurement.Through monitoring the amplitude change of a selected cladding mode within the wavelength band of the Ag-water SPR, a concentration of H 2 O 2 as low as 0.2 μM can be detected, thanks to inherent self-calibration using other, un-sensitive resonances in the measured data.Furthermore, taking advantage of the unique optical properties of the etching effect of H 2 O 2 generated from the enzymatic oxidation of glucose, we have successfully demonstrated a new method for in vitro measurement of glucose in human serum.The high sensitivity together with the miniaturized size, remote operation and label-free detection may make the proposed sensor a good candidate for biochemical detection involving the generation of H 2 O 2 .

Fig. 1 .
Fig. 1.Ag-coated TFBG-SPR system: (a) experimental setup of glucose detection; (b) schematic diagram of the Ag-coated TFBG-SPR biosensor interrogated with tunable linearly polarized light; (c) the reflection spectrum before and after etching.

Figure 1 (
c) illustrates the sensing principle of the H 2 O 2 sensor.Following immersion of the TFBG coated with 30 nm of silver into the H 2 O 2 solution, the thickness of the silver film progressively decreases with the H 2 O 2 etching time.The decreasing thickness of the metal film broadens and weakens the SPR resonance of the metal-SRI boundary and causes a progressive recovery of the amplitudes of the SPR attenuated cladding mode resonances.Since the etching rate of silver shows a clear relationship with the prepared H 2 O 2 concentration, we can measure this concentration by monitoring the rate of recovery of the cladding mode amplitudes, following a suitable calibration.

Fig. 2 .
Fig. 2. a) Simulated mode intensity profiles of a cladding mode phase matched to the silverwater interface SPR; b) Simulated transmission spectra of 18° tilted FBG for different Ag thicknesses according to the legends: I full spectrum; II SPR matched cladding mode; III low order cladding modes and core mode (details in the text).

Fig. 3 .
Fig. 3. SEM of the initial thickness of silver on TFBG without etching.

Fig. 4 .
Fig. 4. Experimentally measured spectral response of Ag-coated TFBG and the corresponding surface morphologies of silver film versus different etching times in 2 mM H 2 O 2 in water.

Fig. 5 .
Fig. 5. Experimental spectra during H 2 O 2 etching at times indicated in the legend: (a) Response of 30 nm Ag-coated TFBG immersed in a 2000 μM H 2 O 2 solution; (b) and (c) are the zoomed spectra of the SPR attenuated cladding mode and the ghost mode, respectively.

Fig. 6 .
Fig. 6.Responses of the plasmonic TFBG versus H 2 O 2 solutions at different concentrations: (a) real time intensity responses of cladding mode near 1500 nm in Fig. 5; (b) Normalized amplitude recovery, in dB relative to the initial amplitude, for different H 2 O 2 concentrations after 20 minutes of etch time.

Fig. 7 .
Fig. 7. Amplitude recovery of the SPR resonance near 1500 nm for different glucose concentrations in human serum at 20 minutes response time.