Fig. 1. Optical fiber sensing scheme composed of a broadband light source and a spectrometer to measure the plasmonic NP response on the fiber optic sensing probe. This probe is based on a 1 cm core-exposed section near the optical fiber cleaved tip (onto which a silver mirror was previously deposited).

Fig. 2. A: Schematic representation of the simulation setup where the random ensemble of Au NPs immobilized over the fiber core was approximated by a gradient effective index method (GEMA). This method divides the NPs into multiple homogeneous layers with an effective RI given by Maxwell-Garnett theory. B: Real component of the effective index profile for the multiple GEMA layers at several NP coverage densities for 50 nm NPs.

Fig. 3. A: Au NP deposition kinetics onto the optical fiber sensor, showing two plasmonic bands dependent on the deposition time. The Au NP diameter is 50 nm. B: Time evolution of the wavelength shift of both plasmonic modes during 50 nm Au NP immobilization. C: Scanning electron microscope (SEM) images of the Au NPs on optical fibers at different deposition times (5, 20, and 40 min). The estimated value of NP densities on the fiber surface (fNP) is included. D: Extinction cross section of Au NPs deposited on fiber for different NP densities (from 2% to 20%) calculated using a scattering-matrix formalism. As a model, hundreds of discretized Au NPs randomly positioned on 1  μm2 were employed. E: Calculated absorbance for different Au densities (from 2% to 20%) obtained via TMM using the GEMA. A strong correlation between the red-shifts and absorbance increases of the plasmonic band with NP density. F: Comparison on wavelength shifts as a function of the NP fiber surface coverage calculated through a scattering-matrix formalism and TMM.

Fig. 4. A: Calculated dispersion curves for a 50 nm effective index plasmonic layer on the fiber surface at various NP densities (fNP), ranging from 0 to 30%. The dispersion curves intersect with the fiber light cone at wavelengths from 720 to 950 nm depending on fNP. The dotted line represents the wavevector of light in vacuum. B: E-field spatial distribution for 50 nm NPs using the GEMA with different fNP, showing a steep increase towards the metal layer and long decay towards the external medium. C: Dispersion curves for several NP sizes at a constant fNP of 20%, causing the intersection with the fiber light cone at successively larger wavelengths. D: Field distribution within the GEMA method with a constant fNP of 20% for different NP diameters as indicated, showing stronger E-fields for larger NPs. E: Refractive index sensitivity for optical fibers with different NP densities. NP diameters range between 20 and 80 nm. The values were measured around 1.340.

Fig. 5. A: Refractive index sensitivity of optical fibers immobilized with 51 nm Au NPs with fNP of 18%, 20%, and 22%, as indicated. B: Refractive index sensitivity of optical fibers immobilized with 87 nm Au NPs with fNP of 9% and 14%. C: Refractive index sensitivity of optical fibers immobilized with 21 nm Au NPs with fNP of 16%. The white filled markers correspond to experimental data, while the solid curves represent a second-order polynomial fitting curve. Dashed curves are the numerical results obtained through the GEMA method. D: Representative SEM images of optical fibers with immobilized Au NPs of different sizes (DNP). The NP density on the fiber surface (fNP) is also indicated.

Fig. 6. A: Experimental optical properties recorded with a 50 nm Au NP immobilized optical fiber immobilized upon changing the refractive index of the surrounding medium from water (n=1.333) to air (n=1.000). B: Dispersion curves of the Au meta-layer calculated for different refractive indexes (ranging from 1.333 to 1.000) via the GEMA method. C: Comparison between the experimental and numerical results on RIS of 50 nm Au NPs immobilized on a planar glass substrate under perpendicular illumination. The numerical results were obtained through the scattering-matrix method formalism for an NP density of 15%.

Fig. 7. Experimental results obtained for optical fibers coated with 70 nm Au NPs at fNP of 7% and 14%: absorption spectra (A) and RIS comparison (B). C: Real-time tracking of the plasmonic band of two Au NP immobilized optical fibers during the functionalization with PLL and TBA, as well as the THR binding events with TBA. The diameter of Au NPs was 70 nm and fNP was 7% and 14%, as indicated. D: Wavelength shift as a function of THR concentration, as measured at the final of each incubation period (in the Tris solution), for both optical fibers covered with NP at a total fNP of 7% and 14%. The dashed curves represent the Hill fitting of the experimental data. LODs are indicated. E: Cross-sensitivity test with HTR showing negligible interference for the optical fiber sensor configurations covered with an fNP of 14%.

Fig. 8. Calculated effective index through the Maxwell-Garnett effective index approximation for distinct Au densities. The complex material refractive index is shown for its real (A) and imaginary (B) components. C: Wavelength of the effective index maxima for the Au densities ranging between 0 and 0.3.

Fig. 9. Calculated optical response from the GEMA structure for a 20 nm layer under s-polarization, p-polarization, and the sum of both (p+s) for increasing fNP from 0 to 30% in steps of 2.5% (represented from the lighter to the darker curves).

Fig. 10. Comparison of the RI sensitivity at fNP of 10% (blue curve) and 20% (red curve) as calculated through the GEMA for NP diameters between 20 and 100 nm. The RI sensitivity difference (ΔRIS) between both curves is shown for NP diameters of 20 and 80 nm.