1. Introduction

Over the last decades, fiber Bragg gratings (FBGs) have been utilized as optical sensors to measure a wide range of physical parameters including temperature, strain, pressure, loading, bending, etc [1-4]. Since the light coupling takes place between well-bound core modes that are screened from the influence of the surrounding-medium refractive index (SRI) by the cladding, normal FBGs are intrinsically insensitive to SRI. In contrast, long-period fiber grating (LPFG) couples the light between the core and cladding modes. The cladding acts as an interface with the external medium making LPFG intrinsically sensitive to SRI, thus sensitive to solution-based chemical and biological materials and processes. To date, a number of SRI sensors have been realized using LPFG structures to measure concentrations of some chemicals [5-8]. Despite high SRI sensitivity, LPFGs exhibit several disadvantages as deployable devices, including high temperature and bending cross-sensitivities, broad resonance and relatively long device size, and measurement in transmission only.

Alternatively, also tilted FBGs have been employed as in fiber refractometers mainly based on the SRI modulation of the coupling efficiency towards contra-propagating cladding modes or on the wavelength shift of the cladding modes resonances [9-10]. Similarly to LPG refractometers, high sensitivity to fiber bending represents a key limitation of such devices. In comparison, FBG structures are less associated with these problems. The unique properties of small size and measurement-in-reflection of FBGs make them ideal as point-probe sensors. Thus, it is a challenge to make FBGs as optical chemsensors. One needs to exploit a new technique which can effectively sensitize FBG structures to SRI, to detect and analyze chemical/biochemical materials and processes.

The first attempt to use FBGs as chemical transducers via evanescent wave interaction was demonstrated in 1996 by Meltz et al. [11]. To address this issue, the basic idea was to use fiber gratings written in D-shaped fibers (DFs) sensitized to SRI through post processing cladding removal. Successively, also planar side polishing was demonstrated as effective technique to develop reliable chemical sensors based on FBGs written in DFs with normal birefringence and depressed cladding [12-14].

Even if DF accommodation allows short-time etching to sensitize the FBG to SRI, its circular asymmetry makes the guiding properties polarization dependent and thus appropriate equipment for sensor interrogation is requested. To overcome this limitation, uniform thinned FBGs represented a suitable solution. They were first demonstrated by Asseh et al. in 1998 [15] and in 2003, have been widely investigated to the aim of assessing the fabrication process [16-17]. Successively, Chryssis et al. demonstrated that SRI sensitivity can be enhanced by ultra-thinned FBGs where the etching process acts also within the fiber core [18] and considering high order modes [19]. Also, the first demonstration of biosensor concept based on etched core FBGs where single stranded DNA oligonucleotide probes of 20 bases were immobilized on the surface of the fiber grating by using relatively common glutarahyldehyde chemistry was provided in Ref. [20] and a highly sensitive liquid-level sensor based on etched FBG was explored in Ref. [21].

In last years, novel configurations involving post-fabrication micro-structuring of FBG devices have been successfully designed and experimentally validated. First approach based on photonic band-gap engineering in FBGs was demonstrated in 2005 [22-23]. The micro-structured FBG (MSFBG) involves a uniform grating with a localized stripping of the cladding layer with azimuthal symmetry along the grating structure [24]. The introduction of the defect leads to strong changes in the reflected spectrum: a band-gap is induced within the stop-band of the pristine grating, similarly to the effect observed in phase-shift FBGs [25-26]. Differently from them, MSFBGs exhibit a spectral response dependent on the SRI, able to tune the defect state within the grating stop-band. Further advantages could be obtained enabling the fabrication of more complex structures based on multi-defect MSFBGs and chirped MSFBGs [27-30]. Moreover, high refractive index coated MSFBGs could be used to enhance the sensing performances of the basic device as reported in Ref. [28]. A high resolution lithographic fabrication method for MSFBGs was recently proposed [31]. A polymeric coating uniformly deposited along the grating length acts as protective layer for wet chemical etching process. An UV laser micromachining tool was then used to properly pattern the coating enabling the cladding removal in well defined regions along the grating length. Optimization was carried out by using polyamide coatings as protective layers and a special designed UV laser micromachining tool operating at 193 nm able to accurately remove polymeric layers with azimuthal symmetry and a precise spatial control [31]. However it requires the use of expensive post-fabrication tools whereas the practical use in sensing field reveals additional drawbacks principally related to the presence of the masking layer. In particular, the fiber weakening in the thinned part, together with the large diameter mismatch in comparison to the thicker lateral coated regions, causes a poor mechanical robustness for practical sensing applications. Besides, isotropic chemical etching causes undercutting effects that are undesired when MSFBG has to be combined with micro-fluidic technology and/or sensitive coating layers.

On the other side, recently, Zhou et al. proposed a novel micro-slot based FBG refractometer [32]. Microtunnels were created in standard optical fiber using tightly focused femtosecond laser inscription and chemical etching. A 1.2×125×500 μm micro-slot engraved along a FBG was used to construct liquid core waveguide by filling the slot with index matching oils. The device was used to measure refractive index and sensitivity up to 10^-6/pm was obtained.

However, also in this case, complex post processing procedures are required to obtain micro structuring of the grating device enabling SRI sensitization.

In this work we propose a novel technique to fabricate MSFBGs with the same functionalities of the device presented in Ref. [31] without using expensive post processing tools. The novel method involves the use of the electric arc-discharge (EAD) procedure as suitable technique for local fiber pre-treatment combined with mask-less chemical etching. In particular, EAD approach was selected to taper the fiber along the FBG region in correspondence of a precise localization, while mask-less wet etching allows a uniform fiber thinning preserving the taper profile. In this case, a selective SRI sensitization along the grating is ensured by the fiber taper if the difference between waist and unperturbed diameter is larger than the SRI sensitivity threshold of 20 μm for single mode fibers [16-17].

It is worth highlighting that appropriate EAD-based treatments on common FBGs have been recently demonstrated to be an efficient tool allowing complete photonic band-gap engineering of UV grating devices [32]. In the present work, EAD-based procedure was arranged to provide a suitable geometrical treatment enabling mask-less wet chemical etching with consequent advantages in comparison with lithographic procedure [31]. First of all, the absence of masking layers does not require the use of laser tool to mask patterning. Additionally, it leads to consequent advantages for the chemical etching procedure, particularly regarding the etching stopping and acid neutralization. Also, the mechanical strength of the final device should benefit from the lower diameter mismatch between SRI sensitized and not sensitized regions. Finally, the absence of undercutting effect leads to an intuitive advantage when the MSFBG is combined with micro-fluidic technology.

In this paper, a detailed description of the fabrication procedure and the first prototype are presented and discussed. In addition, the proposed procedure allowed the first experimental realization of multi-defect configurations.

2. Not-lithographic MSFBGs fabrication procedure

The method proposed to realize MSFBGs relies on the use of the EAD approach for fiber grating tapering with a suitable waist diameter in combination with mask-less wet chemical etching. On this line of argument, the fabrication stage can be organized as the sequence of two main steps:

The following subsections accurately present both steps as well as the details of the adopted tools.

2.1 Electric arc discharge

The first step of the proposed approach is aimed to induce a localized taper along the grating in order to enable a selective SRI sensitization through mask-less HF etching. To this aim EAD stage capable to locally heat the fiber combined with a pulling setup was arranged as schematized in Fig. 1(a). Here EAD procedure was provided by a standard fusion splicer (Fujikura FSM-50S) whereas one fiber end is clamped to a fixed holder and a small weight is attached to the other end through a pulley to keep the fiber under constant axial tension. More details on the EAD-based setup are reported elsewhere [33]. The fusion current, arc duration and weight need to be properly adjusted to achieve desired taper profiles [33].

Figure 1(b) shows a schematic diagram of the FBG after the tapering procedure where L_taper and D_waist indicate the length and lowest diameter of the tapered region and D_core and D_clad indicate core and cladding diameters of the original fiber, respectively. Additionally, according to previous works [33-34], EAD treatment locally erases the core refractive index modulation forming a Fabry-Perot like structure enabling the formation of defect states within the grating bandwidth. The pulling procedure is added to control on one side the perturbation length and on the other side to control the optical losses for interferometer balancing [33]. It is reasonable to believe that due to non uniformities of the heated region, at the edges of the thinned part of the grating on both sides of the cavity, weak refractive modulation combined with the pitch chirping due to the fiber pulling will occur. Moreover, these transition regions (few hundred microns long) and characterized by a refractive index modulation partially erased are not able to modify the device spectra accordingly to Ref. [33].

Here, EAD-based procedure was selected for its capability to locally taper the fiber rather than its spectral effect. The main requirement of the EAD stage is, in fact, to provide a fiber thinning (D_clad-D_waist) significantly higher than the SRI sensitivity threshold of 20 μm [16-17] enabling selective SRI sensitization through HF etching of the whole structure without using complex masking procedures.

It is worth highlighting that, in principle, the EAD step alone could allow a local SRI dependence via evanescent wave if D_waist lower than 20 μm is reached. Unfortunately, it would induce significant power losses and a poor diameter control capability, whereas the consequent high diameter mismatch along the fiber would lead to poor advantages in terms of mechanical robustness with respect to the lithographic method [31].

 

Fig. 1. (a). Schematic diagram (not in scale) of the EAD setup; (b). Locally tapered FBG (not in scale); (c). Schematic diagram (not in scale) of the FBG arrangement for the etching procedure; (d). FBG after uniform HF based wet chemical etching (not in scale).

Download Full Size | PDF

2.2 HF chemical etching

The objective of this step is to uniformly thin the pre-treated FBG up to sensitize only the tapered region to SRI changes. To this aim, a simple and low cost approach based on wet chemical etching in HF solution [16-17] was selected. In particular, the presence of fiber taper allows to achieve selective SRI sensitivity along the grating by uniformly thinning the EAD-treated grating without using masking procedure. Besides, the isotropic nature of the etching process, not conditioned by any mask, is able to maintain the fiber profile created by the EAD pre treatment avoiding any undercutting effect (as observable in lithographic approach Ref. [31]). As matter of fact, the objective of this step is a uniform thinning of the EAD processed grating until the SRI sensitization of the fiber taper region is achieved while the maintained diameter mismatch ensures the not sensitization of the lateral regions.

The complete absence of any masking procedure significantly facilitates the chemical process, the tapered FBG was arranged on a Teflon stick and left free from any mechanical stress (see Fig. 1(c)). The stick was then inserted into an etching bath at room temperature. The isotropic nature of HF based operation preserves the taper profile after the fiber etching.

Moreover, the fiber taper ensures a spatial separation between the tapered region itself and the support stick preserving the azimuthal symmetry of the SRI sensitive area during the etching process. This avoids birefringence effects and polarization induced dependence of the defect state. On the contrary, because of the direct contact with the Teflon stick, the etching process could not symmetrically act on the lateral not-tapered regions. However, since these regions will preserve the original guiding behavior no birefringence effect can be observed.

Figure 1(d) shows a schematic diagram of the MSFBG. Here, D_out-taper and D_waist-etched indicate HF thinned cladding and waist diameters respectively, whereas the taper length L_taper was preserved if compared to Fig. 1(b), according with experimental data.

3. MSFBG Prototyping

The FBG utilized presents central wavelength of 1542.05 nm, Full Width Half Maximum (FWHM) bandwidth of about 0.19 nm, peak reflectance of about 98% and physical length of 9 mm.

For the first step, fusion current and arc duration were manually selected to 15 mA and 200 ms, respectively, and a weight of 12.2 g was used, to obtain waist diameters ranging in 60-50 μm, total length in 600-700 μm and insertion losses of 1.0-1.5 dB. Besides, according to Ref. [33], here the EAD location was shifted of 1 mm from the grating centre towards the FBG interrogation end to achieve a quasi balanced Fabry-Perot structure.

A photograph of the taper obtained along the FBG is shown in Fig. 2(a). By digital image analysis, D_waist of about 49 μm and total longitudinal extension L_taper = 680 μm +/- 5 μm were measured. Moreover, a quasi-uniform tapered region length (diameter changes lower than 10% with respect to D_waist) of 160 μm +/-5 μm was also estimated.

Reflected spectra were thus monitored using an optical spectrum analyzer (Ando AQ6317C) with resolution of 10 pm, a broadband super-luminescent diode operating at 1550 nm with 40 nm FWHM and a 3 dB coupler. The optoelectronic setup was completely computer assisted by a GPIB controller and a LabView plug-in. Figure 2(b) compares the reflected spectra of the pristine FBG and after the EAD operation. According to previous work [33], the spectral response after EAD-treatment shows the formation of a defect state inside the original grating stop-band due to local grating erasure and fiber tapering. The measured spectrum shows a defect state centered at 1541.98 nm with a bandwidth of 40 pm whereas 16.5% minimum reflectivity was achieved due to the not perfect compensation of the power losses induced by the taper and asymmetric EAD positioning. Nevertheless, it is important to stress that a full defect state through accurate photonic bandgap engineering can be achieved by acting on the taper waist and shape as demonstrated elsewhere [33]. Finally, after EAD also the bandwidth is 60 pm larger than that of the original grating due to the shorter length of the lateral grating regions compared to the original grating.

 

Fig. 2. (a). Optical photograph of the FBG region tapered by EAD; (b). Comparison between unperturbed and tapered FBG spectra.

Download Full Size | PDF

Successively the EAD treated FBG was chemically etched in a 12% HF acid solution whereas the reflected spectrum was in situ monitored. Figure 3 plots the reflected spectra corresponding to different etching times. Likewise to the etching step of the lithographic method presented in Ref. [31], the spectrum of the FBG remained unchanged in the first phase of the etching, indicating that the core mode was still well bounded by the cladding also in correspondence of the taper region. The defect state started to blue shift when the thickness of the residual cladding layer on the tapered region was reduced to just a few microns, thereby allowing the evanescent field to penetrate to the surrounding medium. As the etching went on, the defect state induced by the EAD step shifted towards shorter wavelengths up to its disappearance from the left stop-band edge with the formation of a new defect state at the right edge. The chemical etching was easily stopped after 125 minutes by using aqueous solution of calcium oxide (CaO) [31] demonstrating an etching rate of about 0.3 μm/min. It is worth noting that the absence of mask layer and thus undercutting effects significantly facilitates the etching stop providing better etching control than the lithographic approach. The experimental evidence in case of lithographic procedure, in fact, reveals slower acid neutralization probably due to presence of acid solution in the undercutting regions [31].

Figure 4(a) shows an image of the tapered region after the wet etching. By digital image analysis, it was estimated that the taper longitudinal length and its quasi-uniform region length were practically unchanged after the etching process. Differently, the structure appears uniformly thinned exhibiting D_waist-etched of 9 μm, significantly lower than the SRI sensitivity threshold, and D_out-taper of 84 μm, far from SRI sensitivity regime.

Finally, Fig. 4(b) compares the original FBG spectrum with that of the fabricated MSFBG when air is considered as surrounding medium. As observable, the defect state is located at 1542.10 nm and presents a minimum reflectivity of 25.5% and a bandwidth of 50 pm.

A critical issue that needs to be taken into account in micro-thinned structures regards the optical power losses. Figure 4(b) reveals that the maximum reflected power decreases less than 20% in comparison with the unperturbed grating spectrum. Consequently common grating interrogation units can be adopted for MSFBG spectral measurements avoiding the use of high power light sources.

 

Fig. 3. FBG spectral evolution during the etching process.

Download Full Size | PDF

 

Fig. 4. (a). Optical photograph of the MSFBG perturbation; (b). Comparison between unperturbed FBG and MSFBG spectra in air.

Download Full Size | PDF

As well documented in previous works [22-24, 31], the most attractive MSFBGs feature is the SRI tunability of the defect state inside the grating stop-band. In this section, the response of the fabricated MSFBG to SRI changes was tested with a series of glycerin solutions at different concentrations. The test solutions were characterized by a commercial Abbe refractometer with resolution of 10^-4. For glycerin concentration from 0% to 90%, the refractive index increases more or less linearly from 1.33 to 1.46. The realized MSFBG was immersed into each glycerin solution and its spectral response was monitored step-by-step. In Fig. 5(a) the spectra of the grating for several SRIs are shown. In agreement with theoretical analysis [24] and previous experimental results [31], a red shift of the defect state inside the FBG stop-band occurs as the SRI increases.

Figure 5(b) displays the SRI sensitivity characteristic in terms of defect state wavelength changes versus SRI variation. It is clear from the figure that the defect state wavelength red-shifts at different rates with increasing SRI because higher evanescent wave interaction of the guided light with the external medium occurs [24, 31]. Here, circles represent experimental data whereas the solid line is a numerical fitting. Experimentally, the defect state red-shifts from 1542.09 nm to 1542.17 nm (total red shift of 80 pm) as the SRI passes from 1.333 to 1.4493. In particular, red shifts of 20 and 30 pm correspond to SRI changes from 1.333 to 1.3812 and from 1.4098 to 1.4392, respectively. If we define the SRI sensitivity as the defect state wavelength shift induced by refractive index unit (RIU) change, the defect state tuning sensitivity passes from 0.41 nm/RIU around the water refractive index (SRI=1.333) to 1.02 nm/RIU for SRI of about 1.44.

Even if the achieved results are able to demonstrate the effectiveness of the proposed approach it is worth noting that some improvements aimed to increase the SRI sensitivity should be adopted. In fact, the reported sensitivities values are quite low if compoared with other refractometers configurations involving LPGs or TFBGs [8, 10]. Nevertheless, according to theoretical studies reported elsewhere [24] and previous experimental results [18, 31] longer defect and/or thinner waist diameter can strongly enhance the SRI sensitivity more than one order of magnitude [31]. Also, the author theoretically studied that the SRI sensitivity can be further increased up to one order of magnitude by using polymeric overlays with refractive index higher than the silica one and appropriate thickness [28]. Currently, the last approach is experimentally under investigation.

 

Fig. 5. MSFBG dependence on the SRI: (a) Spectral response for different SRIs; (b) Defect state wavelength shift versus the SRI.

Download Full Size | PDF

4. Multi-defect MSFBG

This section reports on the experimental realization of a multi-defect MSFBG prototype for the first time to our best knowledge. MSFBGs with more than one defect or SRI sensitized region have been widely theoretically investigated demonstrating appealing features such as simultaneous multi SRI measurements in sensing field or multi-channel filtering in communication applications [27-30]. However, up to now unsuccessful prototyping has been obtained by the previously reported fabrication techniques. For instance, the undercutting effects of adjacent defects limit the use of the lithographic procedure for the fabrication of multi-defect configurations.

On the contrary, the EAD-based not-lithographic procedure seems to be the most suitable solution to prototype multi-defect MSFBGs. To this aim, EAD treatment needs to be repeated at various precise locations along the grating length providing multi-tapered regions to be uniformly etched in the etching step. This enables the possibility to realize multi-defect structure with defect separation also of a few hundred microns.

Here, experimental steps devoted to the first two-defect MSFBGs fabrication and successive SRI characterization are reported. To this aim, a commercial 9 mm long uniform FBG, demonstrating peak reflectance of 98%, central wavelength of about 1542.04 nm and a bandwidth FWHM of 0.19 nm was selected.

Tapered regions positioned at 3 mm and 4 mm from the interrogation end of the grating were selected with regards to the EAD process. Again, the asymmetric structure was designed to compensate taper insertion losses according to Ref. [33].

The EAD stage was carried out in order to achieve waist diameters in both defects as close as possible with one another, enabling the possibility to sensitize to SRI both tapered regions with a single and uniform chemical step. However, it is obvious that the characteristics of the two tapered regions can be ad hoc selected by acting on the EAD parameters as already demonstrated [33]. Here, the EAD procedure were repeated in correspondence of each defect to obtain the desired waist diameter where fusion current, arc duration and weight were step-by-step adjusted ranging within 15-17.1 mA, 100-200 ms and 4-12.6 g respectively. Figure 6 reports digital image of the fiber grating focusing on both tapered regions (at 3 and 4 mm from left to right). The first tapered region (at 3 mm) shows D_waist of 66 μm and L_taper of approximately 550 μm +/- 5 μm, whereas the quasi-constant tapered region is about 120 μm +/- 5 μm long. D_waist of 69 μm, Ltaper of 560 μm +/- 5 μm and a quasi-constant tapered region length of 110 μm +/- 5 μm, instead, correspond to the second tapered region.

 

Fig. 6. Optical photograph of the EAD-treated grating at 3 mm and 4 mm from the interrogation end.

Download Full Size | PDF

Figure 7 shows the FBG spectral evolution after each EAD treatment. After the first EAD process, the reflected spectrum (red line) shows a larger stop-band (bandwidth of about 0.27 nm) and the formation of a single defect state located approximately at 1542.05 nm and characterized by a minimum reflectivity of 22.5% and a bandwidth of about 30 pm. After the second EAD process, instead, it is possible to observe the presence of two different defect states inside the FBG stop-band. The one at lower wavelength is located at 1542.02 nm and the second one at 1542.15 nm. They present minimum reflectivity of 12% and 8%, respectively, while both of them have 30 pm bandwidth.

It is worth highlighting that the global spectral response is the result of the optical beating of the three unperturbed grating regions modulated by the optical paths of the EAD treated regions or cavities [28, 34]. It is important to remark, any physical or geometrical change in correspondence of each tapered region would induce the shift of both defect states.

 

Fig. 7. FBG spectral evolution during the EAD operations.

Download Full Size | PDF

After the tapers fabrication, the structure was subjected to HF based wet chemical etching similarly to the single defect MSFBG described in the previous section. During the etching process, the grating spectrum remained unchanged in the first phase of the etching. As expected, as the tapered regions became SRI sensitive, both defect states induced by the EAD operations started to blue shift.

Figure 8 reports the digital image of both defects when the etching process was stopped. The chemical treatment preserved the EAD-step fiber profile whereas thinned waist diameters of D_waist-etched = 9 and 12 μm, respectively, and out of taper thinning up to D_out-taper= 65 μm were measured.

 

Fig. 8. Optical photograph of the etched device. It focuses on defects at 3 mm and 4 mm from the interrogation end.

Download Full Size | PDF

The spectral response of the realized multi-defect MSFBG in air is compared with the original grating spectrum in Fig. 9. Due to the blue shift caused by the taper diameter reduction during the etching process now the defect states are located at 1541.91 and 1542.06 nm, respectively. They reveal minimum reflectivity of 20% and 8% and bandwidth of 30 and 40 pm, respectively.

The spectral characterization in terms of defect states shifts of the two-defect MSFBG as function of different SRIs is also investigated. To this aim, the experimental procedure used for the single defect MSFBG was repeated in order to investigate the following cases:

 

Fig. 9. Comparison between unperturbed FBG and multi-defect MSFBG spectra in air

Download Full Size | PDF

 

Fig. 10. Defect states wavelength shift versus the SRI of right (a) and left (b) defect state.

Download Full Size | PDF

Figure 10 shows the shift of the defect states wavelength (λ_DS,left and λ_DS,right for the defect state at lower and higher wavelength, respectively) versus SRI changes in the range 1.333 to 1.4663 for all the investigated cases. Accordingly with single defect behaviour and theoretical analysis, as the SRI increases (in correspondence of one or both defects) a consequent red shift of all defect states occurs. Also, theoretical studies predict that higher SRI sensitivities of both defect states are expected passing from case (a) to (b) and then to (c) due to increasing interaction between evanescent wave and surrounding medium [28]. Figure 10, in fact, reveals the lowest sensitivity for both defect states in case of SRI changes only on the thicker defect (case a). Wavelength shift amount of 50 pm was measured for both defects as SRI changes from 1.333 to 1.4663. By acting on the SRI in correspondence of the thinner defect (case b), instead, total red shifts of about 60 and 70 pm were registered for λ_DS,left and λ_DS,right, respectively. Highest sensitivity was measured as SRI changes on both defects where red shifts of 70 and 90 pm correspond to λ_DS,left and λ_DS,right respectively. In particular as SRI moves around 1.45 sensitivities of approximately 0.6nm/RIU, 0.8 nm/RIU and 1.6 nm/RIU for the left defect state and 0.6 nm/RIU, 1.3 nm/RIU and 2.1 nm/RIU for the right defect state are measured in (a), (b) and (c) cases, respectively. Assuming a wavelength resolution of 1pm, the lowest refractive index variations that can be detected have been estimated to be 1.7·10^-3, 7.7·10^-4 and 4.7·10^-4 for (a), (b) and (c) cases respectively.

These results demonstrate the real feasibility and correct working of first two-defect MSFBG prototype. The successful realization of such device offers new interesting perspectives for multi-parameters sensing approaches by using a single in-fiber device. On the base of well-documented theoretical studies [28], in fact, such structure could enable the possibility to provide dual refractive index measurements. Also, new perspectives are expected since the proposed fabrication stage should provide a valid processing technology for multi-defect micro-structured chirped FBG [29-30]. Further experimental tests will be addressed to tailor the defect state spectral position and sensitivity to specific applications and to integrate MSFBG with micro-fluidic systems.

5. Conclusion and discussion

In conclusion, in this work, we propose a novel processing procedure for the fabrication of in fiber refractometers based on MSFBGs by adopting a two-step procedure with consequent benefits in terms of reliability, robustness, cost-effectiveness and integration with sensitive overlays and microfluidics components In particular, local EAD treatment and successive wet chemical etching have been used to local SRI sensitization. In comparison to alternative grating micro-structuring procedures, the proposed approach does not require expensive or complex equipments such as UV excimer laser for mask patterning [31] or femtosecond laser to assist selective chemical etching [32].

Additionally, in comparison to the lithographic approach for MSFBG fabrication [31], the EAD-based procedure overcomes some drawbacks enhancing MSFBG functionalities in practical use. In particular, the absence of the mask layer leads to:

Finally, the experimental demonstration of the potentiality of the proposed approach in case of single and multi defect refractometers have been also carried out. It is expected that the higher yielding of the not lithographic approach and the aforementioned advantages enable MSFBGs technology to be a valid technological platform for next generation n of chemical and biological sensors and systems.